Animal Behavior

How and Why Animals Do the Things They Do

Volume 1: History, Causation, and Development

Ken Yasukawa and Zuleyma Tang-Martı́nez, Editors

Copyright 2014 by Ken Yasukawa
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted, in any form or by any means, electronic, mechanical,
photocopying, recording, or otherwise, except for the inclusion of brief quotations in a
review, without prior permission in writing from the publisher.
Library of Congress Cataloging-in-Publication Data
Animal behavior : how and why animals do the things they do / Ken Yasukawa, editor.
pages cm
Includes bibliographical references and index.
ISBN 978–0–313–39870–4 (hard copy : alk. paper) — ISBN 978–0–313–39871–1
(ebook) 1. Animal behavior. I. Yasukawa, Ken, 1949–
QL751.A6498 2013
591.5—dc23 2013012228

ISBN: 978–0–313–39870–4
EISBN: 978–0–313–39871–1
18 17 16 15 14 1 2 3 4 5
This book is also available on the World Wide Web as an eBook.
Visit www.abc-clio.com for details.
Praeger
An Imprint of ABC-CLIO, LLC
ABC-CLIO, LLC
130 Cremona Drive, P.O. Box 1911
Santa Barbara, California 93116-1911
This book is printed on acid-free paper
Manufactured in the United States of America
 Contents

Preface v

1. This Is How We Do It: The Scientific Study of Animal Behavior 1

Ken Yasukawa

2. In the Beginning: A History of Animal Behavior 33

Lee C. Drickamer

3. Being Direct: Can Pigeons Plan Ahead? 79

Brett M. Gibson

4. Social Learning, Tradition, and Culture: Data and Debate 115

Bennett G. Galef

5. Behavioral Genetics: Beyond Nature and Nurture 151

Bronwyn H. Bleakley and Anne Danielson-François

6. Neural Systems, Behavior, and Evolution: Integrating Proximate

and Ultimate Analyses of Behavior 185
Laura Smale

7. Hormones and Animal Behavior 217

Matthew J. Fuxjager, Sarah Jane Alger, and Catherine A. Marler
iv CONTENTS

8. Epigenetics and Animal Behavior 263

David Crews

9. Behavioral Development and Ontogenetic Adaptation 289

Jeffrey R. Alberts and Christopher Harshaw

10. The Individual in Context: A Social Network Approach to the

Study of Behavior 325
David J. White, Andrew P. King, and Meredith J. West

11. A Time for Every Purpose: Biological Clocks and Animal

Behavior 351
Megan Hastings Hagenauer and Sean P. Bradley

12. Getting There: Animal Orientation, Navigation, and Migration 387

James L. Gould

Glossary 413
About the Editors and Contributors 435
About the Editorial Board 439
Index 441
 Preface

Most people are interested in the behavior of animals, but the scientists who
study animal behavior, exemplified by the authors of the volumes of this book,
use time-honored methods of hypothesis testing in their attempts to under-
stand why and how animals do the things they do. As stated in many of the
chapters in this book, the scientific study of animal behavior owes a tremen-
dous debt to Niko Tinbergen, who turned his boyhood naturalist’s curiosity
about the world of animals into a career that was both highly productive and
extremely influential. This book is dedicated to Tinbergen.
Tinbergen proposed his famous four questions in his 1963 paper “On Aims
and Methods in Ethology” (Zeitschrift für Tierpsychologie, 20, 410–433). Tin-
bergen noted that a full understanding of behavior must include both “proxi-
mate” and “ultimate” explanations. “How” questions are answered by
proximate explanations of the developmental history and mechanisms that
control behavior. “Why” questions are answered by ultimate explanations for
the adaptive value and evolutionary history of behavior. Taken together, this
approach has produced a tremendous amount of research and lots of answers,
but also even more questions. The chapters in this book have been written to
demonstrate the dynamic nature of the scientific study of animal behavior. We
hope that you will find each chapter an informative and enjoyable glimpse into
the curious minds of behavioral scientists.
As behavioral scientists, we have been inspired and influenced by many
who came before us. To paraphrase Sir Isaac Newton, we have seen a little fur-
ther by standing on the shoulders of giants. But we have also benefitted from
vi PREFACE

many contemporaries, some of whom have recently passed away much too
soon. We could name many of them, but a few who stand out to us are Val
Nolan, Devra Kleiman, Penny Bernstein, Al Dufty, and Chris Evans. We
should note that Chris died while writing a chapter for Volume 3, and we
are therefore proud that one of his last contributions to animal behavior will
appear in this book. We also dedicate this book to Val, Devra, Penny, Al,
and Chris, our friends and colleagues.

SPECIAL FEATURES
Each chapter is written by active researchers who are experts in the subjects
they cover. Because science tends to read like a foreign language, we have
included some special features to help you understand what our contributors
are telling you. First, our contributors try to explain why they are interested
in their topics and how they go about asking and answering the questions they
consider. Second, the technical terms (vocabulary) are listed in bold italics
and are defined in the glossary at the end of each volume. Finally, each chapter
also includes a list of the references cited within it. If you are interested, you
can try to read some of these research papers. Some can be found using a com-
puter search—Google Scholar is one way to find some of them. Another way
is to use the name of the author to search for the author’s website—many
authors provide PDF copies of their papers on their websites. One last way is
to visit a university library and ask to do a computer search using their elec-
tronic data bases. You will probably need to get special permission to do so,
however.
 1

This Is How We Do It: The Scientific Study

of Animal Behavior
Ken Yasukawa

INTRODUCTION
Interest in animal behavior is at an all-time high. Animal behavior is the subject of
documentaries (e.g., March of the Penguins), animated children’s adventures (e.g.,
Finding Nemo), TV shows (e.g., Dog Whisperer), TV series (e.g., Shark Week), and
entire cable networks (e.g., Animal Planet). But where does the information come
from? Information about the behavior of animals comes from scientific study,
and the field of animal behavior is now well established. In this chapter I will
briefly describe the three major methods (observations, experiments, and model-
ing) that researchers use to study animal behavior, and I will organize it around
the process by which research is designed. My goal is a general introduction with
some examples and suggestions for further reading rather than an exhaustive dis-
cussion with endless references. I write this chapter in honor of my role model,
Niko Tinbergen, who established the framework for the study of animal behavior
(see Chapter 2). Figure 1.1 shows some of the animals he studied.

WHO DOES RESEARCH IN ANIMAL BEHAVIOR?

Animal behaviorists are a diverse lot. They range in age from elementary
school children (Blackawton et al., 2010) to retirees (several chapters in this
book). Those with advanced degrees come primarily from biology,
2 ANIMAL BEHAVIOR

Figure 1.1. Drawings of some of the animals studied by Niko Tinbergen. (Redrawn by
Emmerson Fuller)

psychology, and anthropology, but disciplines such as mathematics, physics,

sociology, economics, and philosophy are also represented. And they reside
throughout the world, although North America, Europe, and Australia are dis-
proportionately represented.

FUNDING FOR RESEARCH

Some research projects require few financial resources. My doctoral thesis
study of red-winged blackbirds (Agelaius phoeniceus), for example, required
travel to my study site, a few items of equipment (a telescope, which I bor-
rowed; binoculars; a stop watch; and a clipboard), and a few supplies (data
sheets, pencils, coffee, and Snickers bars). In contrast, however, before most
animal behaviorists can conduct their research projects, they must secure
funding to pay for travel, housing, personnel, equipment, supplies, and animal
care. Authors published in a single recent issue of Animal Behaviour, a highly
respected, international journal, acknowledge funding from a variety of
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 3

governmental and nongovernmental sources, including the National Science

Foundation and National Institutes of Health (USA), Natural Science and
Engineering Research Council (Canada), Consejo Nacional de Ciencia y
Technología de México, Biotechnology and Biological Sciences Research
Council, and Natural Environment Research Council (UK), Northern Ireland
Environment Agency, Centre National de la Recherche Scientifique (France),
Fundação para a Ciência e a Technologia (Portugal), Max-Planck-Institut
für Verhaltensphysiologie (Germany), Austrian Science Fund, Hungarian
National Science Foundation, Swedish Research Council, Swedish Envi-
ronmental Protection Agency, Academy of Finland, Netherlands Organisation
for Scientific Research, Swiss National Science Foundation, Commission
of the European Communities, Agency for Innovation by Science and Tech-
nology (Lithuania), Israel Science Foundation, National Research Foundation
(South Africa), Australian Research Council, New Zealand Marsden Fund,
Fujian Province Nature and Science Project (China), National Geographic
Society, Volkswagenstiftung, Birds Australia, Canadian Society of Ornitholo-
gists, and the Leverhulme Trust. This listing of funding sources also empha-
sizes the worldwide reach of animal behavior and the variety of foundations
and government agencies that support animal behavior research.
This financial support was absolutely necessary, but funding is extremely
competitive—most research proposals are not funded. Professional animal
behaviorists constantly deal with the stress of securing funding.

HOW DO ANIMAL BEHAVIORISTS GET STARTED?

Every journey must start with the first step, but that step can be very diffi-
cult. Paul Martin and Patrick Bateson (2007) suggest that the first step is to
ask a question about behavior. Anything from a very general “What does this
animal do?” to a narrow and more hypothetical “Do females prefer conspicu-
ous male behavior?” will do. In many cases, research tends to generate more
and more specific questions, and the questions emerge from earlier observa-
tions, questions, and potential answers.
Here is an example. Early observations of the behavior of my study species,
the red-winged blackbird, showed that males defend territories (i.e., they try
to keep other males out of small portions of nesting habitat). So, an initial
question might be, “How do male red-winged blackbirds defend their
territories?”
With the first step taken, a next step would be to gather more information.
Typically, this next step involves doing some background reading or making
4 ANIMAL BEHAVIOR

preliminary observations and then using this information to propose some

possible answers, which we call working hypotheses.
For our initial question, doing a computer search, reading classic papers
(Nero, 1956a, 1956b; Orians & Christman, 1968), and observing males at a
marsh in spring would be helpful. Our reading and observation would tell us
that males scan the habitat constantly from prominent perches, frequently
produce a song that sounds like “o-ka-lee,” and show the red-and-yellow wing
patches (epaulets) for which the species is named. At this point we have at
least two potential answers to our question: (1) Song is used to defend the
territory. (2) Epaulets are used to defend the territory. To make it easier to
talk about these two hypotheses, we give them names. Although it might be
tempting to name each one after the person who proposed it, it is more helpful
to give them descriptive names such as the “song hypothesis” and the “epaulet
hypothesis” for territory defense.
These potential answers are working hypotheses because each one is a
testable explanation (answer to our question), and having more than one pos-
sibility means we are using the method of multiple working hypotheses
(Chamberlin, 1890). Once we have working hypotheses, we must test their
predictions, but what does that mean? In our context, a prediction has nothing
to do with telling the future. Each working hypothesis must predict the results
of research that someone could perform. We then do the research to see
whether we get the predicted results or some other results. This hypothesis test-
ing is critical to designing research in animal behavior or any other scientific
discipline.
But what predictions do our two hypotheses make? One suggestion to identify
predictions is to use the “if–then” construction.
1. If red-winged blackbird song is used to defend territory (hypothesis), then males
should sing when they are on territory but not when they away from the territory,
and males that are unable to sing should be unable to hold their territories
(predictions).
2. If red-winged blackbird epaulets are used to defend territory (hypothesis), then
males should show their epaulets when they are on territory but not when they
are away from the territory, and males that lack epaulets should be unable to hold
their territories (predictions).

Once testable predictions have been identified, the next step is to choose a
research design, including a statistical method, to test the predictions.
Research design and statistical analyses are both very large and complex topics
well beyond the scope of this chapter.
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 5

HOW DO ANIMAL BEHAVIORISTS TEST HYPOTHESES?

You were probably taught that the scientific method involves making
observations, formulating hypotheses, and testing them with experiments.
Although this description is correct in its broad outline, it does little to explain
what happens in actual practice. Note that, despite what you learned about the
scientific method, hypothesis testing is not limited to experiments—observa-
tions and modeling can also be used to test hypotheses. The key aspect of
hypothesis testing is whether the research design is appropriate to test a spe-
cific prediction and whether the prediction, and therefore the hypothesis itself,
can be rejected (Popper, 1959). According to this view, tests of hypotheses are
attempts to falsify them. In other words, when we design tests of predictions,
we always choose designs that could produce results that are contrary to the
hypothesis and its prediction. Although it will be tempting to design research
to confirm hypotheses, such research will not provide strong or rigorous sup-
port for the hypothesis. Instead, it is important to understand that only a study
with the potential to falsify a working hypothesis has the ability to generate
rigorous results that add to our knowledge of animal behavior. I offer a simple
thought experiment to illustrate the process of hypothesis testing.

The Numbers Game

In this game your task is to guess the rules used to choose the next number
in a series. We will start with the numbers 1, 2, and 4. Your first task is to
design experiments to gather data by proposing numbers to test potential rules
(hypotheses). I have used this exercise in many of my classes and even with the
faculty at my college. In every case, students (and my colleagues) pick an
obvious rule and then propose numbers that fit it. So, they propose 8, 16,
32, and so on to test the rule “double the previous number.” What are my stu-
dents doing? In one respect, they are testing hypotheses. They have a particular
rule (hypothesis) in mind, and they are proposing numbers (making predic-
tions) that can be tested (they fit or not). Unfortunately, they will never get
any closer to identifying the rule because they are attempting to confirm their
hypothesis instead of trying to disprove it. What’s the difference? All of their
numbers fit their hypothesis and will do so to infinity (or until the class ends).
To illustrate this point, suppose someone else believes that the rule is “a num-
ber larger than the previous one.” The predictions 8, 16, 32, and so on also fit
that hypothesis, so which one is correct? Unfortunately, attempts to confirm
either hypothesis do not allow us to choose one over the other.
To test the hypothesis critically, you must propose a number that does not
fit the hypothesis. Why does that help? Because if the proposed number does,
6 ANIMAL BEHAVIOR

in fact, fit the rule, then you have showed that the hypothesis is incorrect—you
have disproved the hypothesis. In contrast, even if you double the last number
100 times, you still have not proven the doubling rule because the predictions
fit lots of other rules (e.g., “a number larger than the last”). It is only by elimi-
nating some possible hypotheses that we learn something. So, in our example,
if you want to test the doubling hypothesis, then you should propose a num-
ber that is not double the last number. For example, after 32, you could pro-
pose 63. What happens when you find out that 63 fits? Obviously, the rule is
not “double the last number.”

Testing the Epaulet and Song Hypotheses

Let us return to our two working hypotheses for territory defense in male
red-winged blackbirds. How have predictions of the epaulet hypotheses been
tested? Andrew Hansen and Sievert Rowher (1986) observed that male red-
winged blackbirds conceal their epaulets while they trespass on other terri-
tories and when they are first establishing their territories, but once ownership
is established, they show their epaulets during encounters with other males. In
addition, Frank Peek (1972) found that males whose epaulets are blackened
with hair dye are more likely to lose their territories than males treated simi-
larly but whose epaulets were not changed. These tests have the potential to
disprove the hypothesis, but both the observations and experiments support
the predictions of the epaulet hypothesis.
What about the song hypothesis? Gordon Orians and Gene Christman (1968)
observed that song is the most common and conspicuous vocalization male red-
winged blackbirds give on their territories, and that trespassing males do not sing.
Douglas Smith (1976) showed that males that are surgically prevented from sing-
ing have much more difficulty holding their territories than males that are given
sham operations. These observational and experimental studies also have the
potential to disprove the song hypothesis, but the results support the hypothesis.
Before describing observations and experiments in more detail, it is important to
discuss the advantages and limitations of observations and experiments.

EXPERIMENTS AND OBSERVATIONS: DIFFERENT KINDS OF VALIDITY

Many people believe that experimental studies are better than observation-
al studies and that experimental researchers are more rigorous (scientific) than
those who rely on mere observations. Fortunately for animal behavior, this
viewpoint is misinformed. Marian Stamp Dawkins (2007) offers a thorough
discussion of the value of observational studies, and Alan Kamil (1988)
cogently discusses this issue in his explanation of validity.
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 7

Research in animal behavior (and in science in general) that involves gath-

ering data can be broadly divided into experimental and descriptive work,
but empirical research is really a continuum, with purely descriptive, observa-
tional study in the animal’s natural habitat (fieldwork) at one end and tightly
controlled laboratory experiments at the other. Between the extremes are other
kinds of empirical research including natural experiments and quasi-
experiments. Kamil (1988) asserts that no one kind of research is inherently
better than another because each has advantages and limitations. The critical
feature is not where or how data are gathered but the implications of the
results. The study of animal behavior has always used a combination of
methods.

Internal and External Validity

Validity has two distinct meanings in scientific research. External validity
describes how well results of a study can be generalized to other situations or
conditions. Internal validity, in contrast, is the extent to which an effect
can be attributed to a specific cause. In an ideal world, we would want to know
what causes a particular behavior and how the cause-and-effect relationship
works in all other situations. In the real world, however, we cannot have it
both ways.
It should be clear that a descriptive field study has external validity because it is
conducted in the situations and conditions normally experienced by the animals
being studied. Field researchers want as many different conditions as possible to
see a full range of behavior, so they make no attempt to control conditions.
Observations of red-winged blackbird behavior by Robert Nero (1956a,
1956b) and Orians & Christman (1968) are well-known examples. The condi-
tions that make such studies externally valid, however, also prevent us from
knowing with complete confidence what causes a particular behavior to occur.
In contrast, a well-controlled laboratory experiment achieves the rule of
one variable because control and experimental groups are identical in all ways
but one (the experimental variable), so any difference between them must
occur because of the experimental manipulation. Ideally, in the perfect experi-
ment, there are no confounding variables (other things that might explain the
differences), so controlled laboratory experiments have high internal validity.
William Searcy’s (1988) study of male red-winged blackbird song approaches
the rule of one variable. Searcy played recorded songs of males to captive
females in a laboratory and found that four different song types (a repertoire)
are more stimulating to the female than a single song type. The control neces-
sary to achieve internal validity, however, makes it impossible to generalize the
results because we have no idea what would happen if other things also varied.
8 ANIMAL BEHAVIOR

A natural experiment, in which the researcher takes advantage of some

change in the environment, falls between the extremes but is closer to descrip-
tive field study. There is a weak sense of control in that the researcher compares
behavior before and after some natural event. A spectacular example is the
eruption of Mount St. Helens, which was used to study the behavior of male
red-winged blackbirds (Orians, 1985). Prior to the eruption, males rarely fed
their nestlings, but they did so in the next breeding season, perhaps because
there were fewer females than usual for them to attract as mates or because
the volcanic ash that “fertilized” the insect supply improved the food supply.
A quasi-experiment is closer to the other extreme. Here the researcher
manipulates a variable while attempting to control some conditions, but
because not all possible confounds are controlled, the researcher cannot say
with complete confidence that a difference between control and experimental
groups is caused by the experimental manipulation. I tested predictions of the
song hypothesis by conducting an experiment in the field in which I compared
the ability of “singing” and silent loudspeakers to defend otherwise empty red-
winged blackbird territories (Yasukawa, 1981). Singing speakers were more
effective than silent ones in discouraging trespassers, but uncontrolled con-
founds included time of day, day of the season, territory quality, age and expe-
rience of the removed male red-winged blackbirds and their neighbors, and
many others.
Although no one method can produce both internal and external validity, a
combination of methods can produce great confidence in explaining a particu-
lar behavior. Taken together, the many studies of red-winged blackbird song
and epaulet function are a good example.

PRELIMINARY CONSIDERATIONS IN TESTING BEHAVIORAL HYPOTHESES

Once we have a question, some preliminary information, at least one work-
ing hypothesis, and predictions, we can begin hypothesis testing. Behavioral
data can be used to address Nikolaas Tinbergen’s (1963) four central ques-
tions of animal behavior.
1. What causes the behavior to occur? (causation)
2. How does the behavior develop? (development)
3. How does the behavior affect survival, mating ability, and reproductive success?
(function)
4. What is the evolutionary history of the behavior? (evolution)

Before any observations are done or behavior measured, it is very important

to make a series of decisions (Martin & Bateson, 2007). The level of analysis
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 9

must be chosen to provide enough detail to be worthwhile but not so much

that it is overwhelming. A species with sufficient background information
and appropriate life history and social characteristics that is reasonably avail-
able, easy to observe, and tolerant of human observers must be identified.
Finally, a good location, whether in the field, zoo, aquarium, farm, or labora-
tory, must be found and appropriate times to do the research chosen. With
these decisions made, observer effects, anthropomorphism, and ethics must
also be considered.
Without meaning to, you could have a subtle or even substantial effect on
the behavior of your animals. Observer effects can be mitigated by the use of
blinds (hides) in which the observer conceals him- or herself or by making a
video recording of the behavior, but being restricted to a blind or using a video
camera might make observing more difficult. An alternative is to spend time
making the animals accustomed to your presence, but it is difficult to assess
the effectiveness of such habituation. Therefore, observer effects are some-
thing all animal behaviorists keep in mind and attempt to minimize in any
way possible.
It is easy to misinterpret the actions of animals by assuming that they are just
like us, with our thought processes and emotions. People say, “my dog is feel-
ing guilty” or “my cat is jealous,” and most movies and TV programs about ani-
mal behavior are rife with such anthropomorphism. But animals are not just
like us—many species differ dramatically from us in their sensory abilities,
behavioral responses, and ability to learn. Using human emotions and inten-
tions to explain the behavior of (nonhuman) animals can thus prevent us from
understanding their behavior, but viewing animals as machines is not produc-
tive either. A bit of projection might lead to interesting hypotheses to test.
Any study of animal behavior should balance the information you might
gain against potential harm to the animals. There are three important ques-
tions to ask when examining the ethics of behavioral research:
1. Will the research increase scientific understanding?
2. Will the research produce results beneficial to humans or to the animals
themselves?
3. How much discomfort or suffering, if any, will the research inflict on the animals?

The benefits addressed by the first two questions must be weighed against
the cost considered by the third. A valuable tool in determining this balance
is the Guidelines for the Treatment of Animals in Research and Teaching, pro-
duced jointly by the Animal Behavior Society (ABS) and the Association for
the Study of Animal Behaviour and published each January in the journal Ani-
mal Behaviour. These guidelines are also included in the ABS Handbook,
10 ANIMAL BEHAVIOR

which can be found on the website of the Animal Behavior Society

(www.animalbehaviorsociety.org). In addition, colleges and universities in
the United States have institutional animal care and use committees that
examine and approve research protocols, and journals such as Animal Behav-
iour require potential authors to stipulate that their research conforms to eth-
ical care and use guidelines. I should also mention that in many cases
researchers must also get permission (e.g., from a property owner) and permits
(e.g., state and federal) to do approved research.
Keeping your question or hypothesis in mind, you next need to make pre-
liminary observations, identify the behavioral variables to measure, and choose
suitable recording methods for making the measurements.

HOW DO ANIMAL BEHAVIORISTS PERFORM OBSERVATIONAL STUDIES?

Dawkins (2007) discusses three principles of good observational design.
First, replication must be independent, meaning that one observation or ani-
mal must not influence or affect another one. For example, if you observe one
individual many times, each observation is not independent of the others
because the same animal is involved—we would not get a good picture of
the differences in behavior that can occur in this species because the one indi-
vidual may, by chance, behave strangely or differently from the normal pattern
of the species. Such an improper use of repeated observations is called pseu-
doreplication, and it leads to improper statistical analysis and interpretation
of results. Unfortunately, attempts to avoid pseudoreplication can also lead
us astray. A hypothetical example follows (Dawkins, 2007).
Suppose we want to know whether schooling fish respond differently to
large and small predators. If we use a single school of 20 fish to observe reac-
tions to large and small predators, then obviously each animal is not an inde-
pendent replicate because each school member is affected by the other fish in
the school, so we end up with only one unit of replication (the school). To
avoid pseudoreplication and to generate a more useful sample size (number
of independent replicates), we decide to observe each fish separately, thus pro-
ducing 20 independent replicates. Unfortunately, although we have generated
a statistically valid design, we have also produced a biologically meaningless
(invalid) one because these fish are schooling animals, so they do not encoun-
ter predators singly. Ideally, then, we would need to study enough different
schools of fish to allow for valid statistical analysis.
Second, variables must not be confounded. Although we mentioned con-
founding variables in the context of the internal validity of controlled labora-
tory experiments, confounds can complicate observational studies as well.
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 11

Going back to the fish example, if we observed schools of fish responding to

large and small predators but did the large-predator observations in a plastic
wading pool and the small-predator observations in a small aquarium, then we
could not attribute a difference to predator size because size of the testing arena
may also have affected the behaviors. Fortunately for us, there are several sam-
pling designs that address this issue. One is to randomize the order of observa-
tions, for example of the large- and small-predator observations. This is because
if, for example, we were to test all of the small predators first, followed by tests
with large predators, exposure to and experiences with the small predators may
subsequently bias responses to the large predators, and the results obtained
would not be reliable. Note that random is not the same as haphazard—we need
to use a randomization method. Some observers use simple methods such as
flipping a coin or rolling dice, but statisticians recommend using a table of ran-
dom numbers (e.g., odd number ¼ large predator, even number ¼ small preda-
tor) or a computer’s random number generator. You may wonder why a
number from a table or a computer is random given that there must have been
some specific process to choose it. The answer is that the sequence of numbers
has the properties of a random sequence even if the numbers are chosen accord-
ing to an algorithm (and for this reason they are called pseudorandom). Ran-
dom sampling turns out to be an important requirement for statistical testing,
so we have another reason to use randomization procedures. Of course, if a
sequence is truly random, then sometimes we will get a long series of the same
result (e.g., large predator six times in a row) so we might be tempted to alter-
nate observations of large and small predators, but statisticians will surely object,
and there is always the remote possibility that some other process that would
confound our results is also alternating in the same way. A solution to this prob-
lem is to balance in combination with randomization. For example, randomly
choose a large or small predator, observe the opposite next, and then repeat
the process. This procedure produces balanced pairs of observations. Further-
more, if more than one observer is involved in the study, then confounding
applies to observers as well as to subjects, so do not have one observer watch
responses to large predators and the other watch reactions to small predators.
Finally, known but unwanted sources of variation must be removed. A way
to deal with these known confounds is by blocking (or matching). Block
designs allow us to compare like with like or matched observations. So, in
our fish and predator example, we would observe each school’s reaction to
both large and small predators (but of course not always in the same order
because order effects are another source of confounding). Block designs can
be quite complex, so a full discussion of them is beyond us here. You may
want to read the relevant section of Dawkins (2007).
12 ANIMAL BEHAVIOR

Our next two steps in performing an observational hypotheses test are to

identify appropriate behavioral variables and then to choose suitable methods
to record them.

Observational Sampling
Behavior occurs in a continuous stream, but we must somehow break it
into categories to make measurements, and we need names for these categories
of behavior. Behavior can be described by its structure (the postures and
movements) or by its consequences (the effects). Structural descriptions are
objective in that they do not involve an interpretation by the observer, but
they can be needlessly detailed and subtle. Describing behavior by its conse-
quences is not as detailed, but it runs the risk of misinterpreting the conse-
quences altogether. Neutral or descriptive labels avoid this problem. For
example, nestling red-winged blackbirds adopt specific postures and call and
are subsequently fed by their parents. This behavior could be described in
great detail as a series of movements, postures, and sounds, or it could be
called “begging,” assuming that an unambiguous operational definition is
included. Another form of description uses spatial relationships (where and
with whom) rather than what an animal does. For example, the parent red-
winged blackbird bringing food to its nestlings could be said to approach
and depart the nest.
Observations of behavior can be divided into three types of measurements.
Latency is the amount of time until a behavior occurs, whereas duration is
how long a behavior lasts. Frequency (rate) is how often a behavior occurs in
a given amount of time. Measurements can also be categorized into nominal,
ordinal, interval, and ratio scales. A nominal scale measurement is a set of
names or mutually exclusive categories (e.g., resting, displaying, fighting).
Ordinal measurements can be ranked from first to last (e.g., highest, second,
lowest). Truly numeric measurements can be on an interval scale or a ratio
scale. Both interval and ratio measurements can be placed along a scale that
gives specific meaning to differences (i.e., the difference between 1 and
2 minutes is exactly the same as between 11 and 12 minutes), but the interval
scale lacks a true zero point (e.g., time of day, day of the year) whereas the
ratio scale has a true zero point (e.g., duration, distance).

Recording Method
Choosing rules for systematic recording of behavior is critical to designing
observational studies, and choices involve two distinct levels: sampling rules
(which subjects to watch and when) and recording rules (how behavior is
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 13

recorded). One commonly used sampling rule is focal animal sampling,

which limits observations to specific, randomly selected (focal) individuals
(or groups), each one for a specified period. One problem with focal animal
sampling, especially in the field, is that the focal animal may move out of view
during the specified sampling period. One way to reduce this problem is to use
scan sampling, in which a group of individuals is scanned at specified intervals
and the behavior of each individual at that instant is recorded. Recording can
be continuous or periodic. Martin and Bateson (2007) discuss these sampling
and recording rules in more detail, and Bonnie Ploger provides an exercise in
which you can practice the different kinds of sampling and recording (Ploger
& Yasukawa, 2003).

Recording Medium
Our next consideration is the medium used to make recordings. These days,
video recording with camera-equipped cell phones is common (think You-
Tube), but other alternatives include voice recorders, automatic data recorders,
and check sheets. Note that high-tech methods are not necessarily better than
paper-and-pencil methods (field notes and check sheets). For example, video
and audio recordings have the advantage of instant replay, which allows you to
go back and recheck the recording to obtain more accurate measures of the
behaviors. Unfortunately, however, the field of view of a video camera is lim-
ited, and once you make a recording you then have to analyze it, which can be
extremely time consuming. When I began studying red-winged blackbird
behavior in 1973, I wanted to construct a time budget of male activities as well
as to calculate frequencies of specific behavior categories. As a graduate student
with limited resources, I used paper-and-pencil methods to do continuous
recording by setting up a data sheet with 15 rows, each representing one minute
and consisting of 60 equally spaced dots, and a shorthand code for each behavior
(Figure 1.2). I recorded time budget categories by noting when each started and
ended (duration), and events by writing a letter code for each category (e.g., “S”
for song, “3” for high-intensity song spread display). Armed with a clipboard,
wind-up stopwatch (think of the TV show 60 Minutes), and lots of pencils (first
rule of field work—always have more than one pencil), I spent my mornings
observing and recording the behavior of male red-winged blackbirds. These
days, however, there are automated ways to record behavioral data.

Data Analysis
The final methodological questions to answer are “How much data should
I collect?” and “How should I analyze my data?” Answering these questions
14 ANIMAL BEHAVIOR

Figure 1.2. A data sheet used to record the behavior of male red-winged blackbirds.
Each row of dots represents one minute of observation. Time budget data were
recorded with codes (e.g., F ¼ flight) and notations (e.g., FORAGE, CHASE,
COURT) for each category, and durations were denoted with squiggly lines (e.g.,
during the first minute the male was in flight from the 9th to the 16th second). Events
were recorded with codes (e.g., FS ¼ flight song, S ¼ unseen song, 0 ¼ incipient song
spread, 1 ¼ low-intensity song spread, 2 ¼ medium-intensity song spread, 3 ¼ high-
intensity song spread, FCC ¼ flight call complex, Tsy ¼ trespass by a second-year
male) to indicate when they occurred. Other information (e.g., approach neighbor, ti-
ti-ti courtship call) was also noted. I then used the information on each data sheet to
calculate the proportion of time the male spent on his territory and to construct a time
budget (how much time was devoted to: singing and defending the territory, foraging,
flying, courting, preening, and resting) while he was on his territory. I also calculated
rates of singing, alarm calling, and trespassing and the average intensity of song spread
displays (Yasukawa, 1979).

thoroughly is beyond our scope, but a rule of thumb is to gather as much data
as possible given the logistical constraints. Beginning students of behavior
often make the mistake of focusing so narrowly on only one behavior that they
miss out on other potentially interesting or important information. For exam-
ple, say that you are studying whether a hypothetical male insect is more likely
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 15

to sing in the presence of a known, neighboring male or an unknown intruder.

You record only the frequency of song in each case and find no difference.
However, because you only recorded frequency of singing, you miss the fact
that males sing for longer periods (duration) and also accompany singing with
vigorous kicks of their hind legs (potentially an aggressive display) when in the
presence of intruders, but not neighbors. (This is one place where video
recordings might come in handy and allow you to retrieve additional informa-
tion that you did not originally record!)
Once the data are in hand, an appropriate statistical inference test must be
used. Although proper statistical analysis is critical, this topic is too large and
complex for us to consider here, but I can use a parable to illustrate its impor-
tance. If you are interested in more information, Philip Lehner (1996) pro-
vides an excellent resource.

The Magic Coin

One way to explain the need for statistical analysis is with the following
story. You are walking near home when a stranger approaches. Although you
are worried about what might happen, you are relieved to learn that the
stranger is also interested in animal behavior. Because you share a common
interest, he wants to do you a great favor. He says, “I have a great deal for
you. I have a magic silver dollar, and it is magic because when you flip it, it
comes up heads most of the time.”
You do not reply, so the guy continues. “You could make a lot of money
with this coin. All you have to do is bet heads every time.”
“Wow,” you say, “I could use some extra money! How much do you want
for it?”
The guy says, “Well, because we are both interested in animal behavior, I’ll
charge you only $10.”
Would you be willing to buy the coin? Almost all of my students say “no,”
but I ask them whether they are absolutely, positively sure that they have made
the correct decision. The fundamental problem is that there are two ways that
you could make the wrong decision. (1) You could decide to buy the coin only
to find out that it is a regular silver dollar, in which case you are out $9, or (2)
you could refuse to buy the coin even though it really is magic (i.e., it is a
loaded coin that comes up heads most of the time), in which case you stand
to lose whatever net profits you would make with the coin.
Is there any way that you might improve your decision-making process? Of
course—you should gather data, that is, you should flip the coin, so you say,
“I’m no fool! I want to flip the coin first to see if it is magic.”
16 ANIMAL BEHAVIOR

“OK,” the guy says. “You can flip the coin four times, but then you have to
decide whether to buy it or not.”
That may seem reasonable, but a bit of math will show that four flips will
not provide much useful information. If you flip a regular (not magic) coin
four times, what possible outcomes could you get? There are 16 unique
arrangements of four coin flips.
0 heads: TTTT
1 head: H T T T, T H T T, T T H T, TTTH
2 heads: H H T T, H T H T T H T H T T H H H T T H T H H T
3 heads: HHHT HHTH HTHH THHH
4 heads: HHHH

Thus, of the 16 possible outcomes, 5 have at least three heads, and 5/16 ¼
0.3125. In other words, almost 1/3 of the time you would expect to get a
result that would seem to support the magic-coin hypothesis even though
you have flipped an ordinary coin.
Let us put the parable of the magic coin in statistical testing terms. Tradi-
tional statistical analyses use a null hypothesis to calculate probabilities, as we
did above with coin flips. Our null hypothesis is that we are flipping an ordi-
nary coin. These analyses also employ an alternate hypothesis, in this case that
we are flipping a magic coin. Statistical inference tests are used to help us to
decide whether to choose the null or alternate hypothesis. Unfortunately for
us, any decision could be right or wrong, and because we could either decide
to buy or not, there are two ways that we could be wrong. In statistical terms,
there are two types of error: type I error (that we buy an ordinary coin because
we think it is magic), and type II error (that we do not buy a magic coin
because we think it is ordinary). In the case of research, this means we could
conclude that our hypothesis is supported when, in fact, it the hypothesis is
wrong, or we could reject a hypothesis when, in fact, it is supported.
Statistical inference tests are designed to calculate the probability of making
a type I error, so they tell us how likely it is that an ordinary coin would pro-
duce results that seem magic. We want to avoid making a type I error (we do
not what to pay $10 for an ordinary silver dollar), so we decide to buy the coin
only if it is very unlikely that a regular coin would produce magic results. In
our calculation above, that probability was 0.3125 that if we flip a regular coin
(one that comes up heads exactly half the time) four times, we would get three
or four heads. About 31 percent seems too risky for us to conclude that the
coin is actually magic, but three or four heads are the only results that would
support the magic coin hypothesis. What happened?
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 17

What we need is a way to calculate the probability that the coin is magic,
but we have no way to do that calculation because we would need to know
exactly how often the magic coin comes up heads, so we cannot calculate the
type II error. We can calculate the statistical power of the test, however. Power
is the probability of avoiding a type II error, so we want lots of power because
we really want to buy a coin that is magic. In this example, power is very low,
which is why flipping four times was not very useful. What would increase
power? Increasing the sample size by flipping the coin 40 times would provide
more power because the probability of a regular coin coming up heads 30 or
more times is low (0.0032). Would you buy the coin if it came up heads 30
out of 40 times? Perhaps you would, but can you say with absolute certainty
that you got a magic coin? No, because 32 of 10,000 times a regular coin
would also produce that result.
One final complication is that, like internal and external validity, type I and
type II errors trade off—when one is low, the other is high. If we want very low
probability of making a type I error, then we will have a very high probability of
making a type II error, and vice versa. For this reason, convention dictates that
we choose the alternate hypothesis when type I error is ≤0.05 (5%)—meaning
that if we repeated the experiment 100 times, only 5 times or fewer would we
obtain the same results simply by chance—and we choose the null hypothesis
when type I error is >0.05. When animal behaviorists test hypotheses, they
almost always use statistical inference tests to help them interpret their results.
I often have students say that they want to study animal behavior, but they
are not interested in statistics in particular and math in general. These days a
working knowledge of statistical inference testing is required to study animal
behavior from a scientific perspective.
I should emphasize that, even when P < 0.05 (or even < 0.001), we can
never be absolutely, positively sure that our hypothesis is correct because there
is always some small probability that we did, in fact, obtain these results by
chance. Therefore, although we can definitely reject hypotheses, we can never
absolutely prove them (Popper, 1959). At best, we can only say that our
results strongly support our hypothesis. For example, consider the hypothesis
that all elephants are gray. We can go out and count thousands of elephants
and find that they all are gray (this strongly supports our hypothesis). But sup-
pose that suddenly we find one more elephant (perhaps an albino mutant) and
it is white. That one elephant would be enough to disprove our hypothesis—
we can say with certainty that not all elephants are gray. And the fact that we
initially found 10,000 gray elephants cannot prove that all elephants are gray.
Once the methodological decisions have been made, it is finally time to
observe behavior, and then to analyze the data statistically. Throughout these
18 ANIMAL BEHAVIOR

last stages, it is important to remember that your purpose remains testing

hypotheses to answer the four principal questions of animal behavior.

HOW DO ANIMAL BEHAVIORISTS PERFORM EXPERIMENTAL STUDIES?

Experimental design is yet another huge and complex topic, but we need
some understanding of basic principles to conduct even a simple experiment.
Lehner (1996) provides a useful summary of experimental methods in parti-
cular and research in animal behavior in general. A good starting point is a list
of the desirable properties of experimental design: good estimation of treat-
ment effects, good estimation of random variation, absence of bias, preci-
sion and accuracy, wide applicability, and simplicity in both execution and
analysis.
Our first consideration is the meaning of experimental variables. A variable
is something that takes on different values. In experimental studies, we make a
clear distinction between two kinds of variables. An independent variable is
suspected to affect the dependent variable. The independent variable is what
the experimenter manipulates and is hypothesized to affect behavior. A depen-
dent variable changes in response to changes in the independent variable and
is a measured behavior. Our discussion of ways to measure behavior in obser-
vational studies is also relevant to experimental studies. Our next considera-
tion is that of control (of confounding effects). People often use the word
experiment in an informal way (i.e., try something to see what happens).
I emphasize the importance of proper control and the rule of one variable in
experimental design. TV ads provide useful examples. An old favorite of mine
was an ad for an air freshener. It showed a smoke-filled room into which a
hand holding a spray can of room deodorizer suddenly appeared. Amazingly,
as the deodorizer was sprayed, the smoke cleared! Of course, blowing anything
from an aerosol can just might clear the smoke anyway. What is needed is an
appropriate control such as a can of propellant without any room deodorizer.
As in observational studies, replication (using more than one subject per
experimental treatment group) is important because it gives us a way to
account for random variation. Thus, differences among replicates all experi-
encing the same treatment tell us about random differences among individuals
in general, and that information allows us to examine treatment effects over
and above random variation. The greater the number of replicates, the better
our measurements and the more power in our statistical tests. The concept if
replication is both deceptively simple and subtly complex. I discuss this topic,
including the risks of pseudoreplication, in more detail elsewhere (Yasukawa,
2010).
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 19

Experimental Designs for Treatment Groups

The design of treatments is basic but crucial because it defines our hypoth-
esis tests. Treatments can be broadly divided into random effects and fixed
effects designs, but we will focus on fixed effects designs, of which there are
many. The design of layout is a complementary consideration to treatment
design. We have already encountered this concept in observations, but here
layout is the way we assign treatments to experimental subjects. I briefly dis-
cuss some basic designs used in animal behavior research.
A completely randomized design is conceptually simple and allows exami-
nation of the effect of a single factor with any number of treatment groups on
a behavior. For example, we could compare aggressive behavior (as measured
by display duration) of male Siamese fighting fish (Betta splendens) in response
to (1) a neutral object, (2) a mirror image, and (3) a male behind a clear parti-
tion. In this case we have a control (neutral object) and two experimental
manipulations (mirror and male). If we have 30 potential subjects, each would
be assigned randomly to one treatment group, perhaps using a procedure to
ensure equal replication (e.g., fish are randomly assigned in groups of three,
with each treatment getting one subject from each group).
A randomized block, repeated measures design is used to control addi-
tional sources of variance that would otherwise obscure our treatment effects.
Returning to our fighting fish example, rather than dividing our 30 males
among the three treatments, we could expose each fish to all three treatments
(as long as we account for order effects by balancing or randomizing the order
in which the treatments are applied). In this case, each individual fish is treated
as a “block,” so blocking allows us to reduce the effect of individual differences
in aggressive behavior.
Completely randomized two-factor designs can be used to assess the
effects of two distinct sets of treatment groups. Using our fighting fish study
once again, we could add color of fish (red, blue, brown) as a second factor
to our stimulus treatments (neutral object, mirror image, other male behind
a clear partition) to produce a 3 × 3, two-factor design. In this two-factor
design we want to know whether red, blue, and brown males differ in aggres-
sive behavior, as well as whether aggressive behavior varies in response to the
three stimulus objects. The result is nine different groups representing all
combinations of male color and stimulus objects. One additional question
that can be addressed in this design is whether or not the two factors interact
(i.e., that the effect of color depends on which stimulus treatment and vice
versa). If, for example, red males are always most aggressive and brown males
least aggressive regardless of stimulus object, then the effect of color is inde-
pendent of object type. On the other hand, if the most aggressive color
20 ANIMAL BEHAVIOR

changes with different stimulus objects, then the two factors are not indepen-
dent. Even more complex (e.g., three-factor and four-factor) designs are also
possible but are beyond our scope.

Experimental Designs for Numeric Variables

In contrast to the previous designs, which use specific levels of each factor
(e.g., stimulus treatment categories or colors of males), in gradients we attempt
to assess behavioral response to a more continuous range of treatments (e.g., a
temperature gradient). Analysis of gradients uses methods such as correlation
or regression because both variables are numeric. Regression attempts to iden-
tify a functional relationship (e.g., cricket calling rate as a function of temper-
ature). In contrast, correlation measures how tightly the two variables vary
together (e.g., do calling rates of two species increase together, which we say
is a “positive correlation,” or does one increase as the other decreases, which
we call a “negative correlation”). Note that regression and correlation are dis-
tinct methods. Only one is appropriate for a given situation.

Experimental Designs for Counts

When we count how many times something occurs in one of several dis-
tinct categories, enumeration methods are appropriate. For example, we
could ask whether the number of breeding males is the same as the number
of breeding females (Do the numbers of males and females fit a 1:1 ratio?) or if
the presence of a rival affects whether a subject performs a particular display
(Is display performance independent of presence of a rival?). As with regression
and correlation, these two methods (goodness of fit and test of independence,
respectively) are not interchangeable.
Once the experimental methods are set, it is time to collect data (to do the
experiment) and analyze them statistically. Choosing the proper statistical test
depends on the experimental design, so careful planning is essential. The stat-
istical references that I mentioned for observational studies are equally relevant
for experiments, and it is important to remember once again that your purpose
remains testing hypotheses to answer the four principal questions of animal
behavior.

HOW DO ANIMAL BEHAVIORISTS USE MODELS?

In addition to observations and experiments, animal behaviorists can also
use modeling, which often involves the behavior of mathematical systems or
computers and is thus more theoretical in its approach. Rather than studying
the behavior of animals per se, we study the behavior of the model. You may
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 21

not realize how common models are in everyday life (or in animal behavior).
My favorite analogy for a model is a map (Singer, 1996). Think about giving
directions to your house to someone who has never been there. A map of
the route could be a few lines representing the streets to take and maybe a
few major landmarks. This simple map is a representation of reality, but it is
not meant to be real. No one would confuse this map with the part of the
Earth that is travelled to get to your house, but it does an adequate job of
describing relevant aspects. Like maps, models are simplified versions of real-
ity. In the early days of animal behavior, some models were literally that—
you could build them out of real materials and then investigate their proper-
ties. Konrad Lorenz created a famous example with his psychohydraulic model
of motivation, which resembled a traditional water closet (a water storage tank
above and connected to a toilet bowl with a pull handle to flush it). Simple
models can also be more conceptual such as diagrams with boxes and connect-
ing arrows to indicate related processes and effects, or graphical representa-
tions of functional relationships (e.g., the polygyny-threshold model of
Gordon Orians [1969] or Robert Trivers’s [1972] model of parental
investment).
These days, modeling is more formal and mathematical. A mathematical
model uses equations to describe the essential aspects of behavior. The equa-
tions can be solved mathematically to examine how behavior might operate
under very clearly described circumstances, which are called the model
assumptions. Despite the mathematical nature of these methods, they are still
used to test predictions of hypotheses, so they are relevant to us. In animal
behavior, a commonly used method is game theory, although other methods
are used as well.
Game theory was devised by John Nash (made famous in the movie A
Beautiful Mind) to describe economic conflict between countries, businesses,
and even individual people. This modeling approach was introduced to animal
behavior by John Maynard Smith (1976), who first used it to analyze contests
between rivals who are competing for an important resource such as food,
territory, or mates. Maynard Smith was trying to answer a question that had
been puzzling animal behaviorists for many years: “Why do animals use dis-
play (like disputing neighbors shaking their fists at each other) to settle dis-
putes rather than more violent means?” At one time the answer was,
“Because fighting would produce lots of injuries, which would be bad for the
species.” Explanations that rely on advantages to the species or other groups
of individuals are called group selection hypotheses, but evolutionary analyses
in the 1960s and 1970s showed that these hypotheses were usually inadequate
—they could not explain why display was the best way for an individual to
22 ANIMAL BEHAVIOR

settle contests. If a displayer comes up against a fighter, the fighter would win
every time, even if fighting were disadvantageous for the species as a whole.
Maynard Smith’s (1976) solution was the classic “hawk-dove game.”
As with all modeling studies, the hawk-dove game starts with assumptions:

• Pairs of animals engage in contests over a resource, and in each contest a winner
takes possession of the resource.
• Winning increases the fitness (survival, mating success, or reproductive success) of
the winner.
• An injury sustained in a contest reduces fitness.
• If a contest continues for a long time, both contestants experience a reduction in
fitness as a result of wasted time and energy.
• Finally, each animal always employs (plays) a particular strategy (a method of com-
peting) in all contests.

To be clear, it is important that you understand the meaning of the term

strategy in this context. A behavioral strategy is simply a fixed and predictable
way of behaving in a contest. It does not imply that contesting animals make
conscious decisions. Although contests involve two contestants, the purpose
of a game-theory model is to compare alternate strategies to each other to see
whether one is best. In this case, we compare the contest strategies hawk
and dove.
A dove uses threat display in a contest but never fights. If the opponent also
displays, then the dove continues to display as well, but if the opponent
attacks, the dove retreats immediately, losing the contest but avoiding injury.
Thus, a contest between two doves wastes a lot of time for both contestants,
although neither contestant is injured. In contrast, a hawk always attacks
immediately. If a hawk plays against a dove, the hawk always wins and the
dove always loses because the dove retreats immediately. On the other hand,
hawks can also fight each other, and both contestants risk injury as a result.
These written descriptions of what happens in particular contests are stated
formally in equations that describe the fitness payoffs to each strategy when
played against the same strategy (dove versus dove and hawk versus hawk) or
the other strategy (dove versus hawk and hawk versus dove).
The concept that makes game theory models useful in the study of animal
behavior is the evolutionarily stable strategy (ESS). An ESS cannot be
“invaded” by any other strategy. Suppose that all animals in a group play a par-
ticular strategy. What would happen if an individual with a different strategy
joins this group? If the new strategy wins against the old one, it will begin to
spread (that is, this new individual will be successful in reproducing, so its off-
spring will become more and more prevalent over time). Eventually, the new
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 23

strategy will become so common that most contests involve the new strategy
playing itself. At this point, if the new strategy still has higher fitness than
the now rare old one, the new strategy will continue to spread. Eventually,
all animals in the group will play the new strategy. In this case, the old strategy
is clearly not an ESS.
Is there an ESS in the hawk-dove game? Suppose that we have a group of
doves. If hawk cannot invade this population, then dove is an ESS. Inciden-
tally, this situation is the one that group selection (for the good of the species)
explanations would predict, so our game theory model allows us to test the
group selection hypothesis. It should be clear to you, however, that dove can-
not resist invasion by hawk because hawk would win every encounter with
dove, and therefore, hawk will spread. If we start with an all-hawk group, we
get a similar result. Dove can invade because it is rare initially, so it plays
against hawk almost all of the time. Under these conditions, dove initially does
better than hawk because dove does not pay the cost of injury, while in con-
trast, every time hawk plays hawk, there is a high probability of injury. Our
game-theory model predicts that neither strategy is an ESS as long as the cost
of injury to hawk is high and dove can never win against hawk.
If neither strategy is an ESS, what will happen to our all-dove and all-hawk
groups? Maynard Smith (1976) showed that a mixture of the two strategies is
a stable equilibrium. What is a stable equilibrium? In this case it is the mixture
of hawks and doves at which neither strategy does better than the other on
average. In game-theory terms, this equilibrium is called a mixed ESS.
Hawk and dove are certainly not the only ways that an animal might
behave in a contest. Another possible strategy is called retaliator, which dis-
plays against dove but fights (retaliates) when attacked by hawk. Again assum-
ing that the cost of injury to hawk is high, retaliator is an ESS against both
dove and hawk.
In some cases, however, the relationships are too complex for mathematical
(analytical) solutions, so researchers can also use simulations to investigate
behavior. Like mathematical models, simulations attempt to model a particu-
lar behavioral system to gain insight into how the system works, but they
require a computer (or even a network of computers) programmed to perform
the tedious calculations and to display the results in a useful way. The first
simulation was of a nuclear explosion for the Manhattan Project of World
War II. Simulation was used because the scale and complexity of a nuclear
explosion was far greater than blackboards and mathematical models could
handle. Some simulations such as stochastic dynamic programs (Mangel &
Clark, 1988, Hutchinson & McNamara, 2000) and genetic algorithms
(Holland, 1992) follow a specified procedure, but others are purpose-built to
24 ANIMAL BEHAVIOR

test hypotheses for specific circumstances or species. What all simulations

share is a set of representative situations (scenarios) for which a complete list-
ing of all possible states would be impossible. Like mathematical models, sim-
ulations start with assumptions and are typically run under different
conditions to investigate what happens when those assumptions or other con-
ditions change. Searcy and I used a computer simulation to study the process
by which female red-winged blackbirds search for mates (Yasukawa & Searcy,
1986).
Steve and Melinda Pruett-Jones’s (1994) work on bowerbirds (Ptilono-
rhynchidae) provides an example of both mathematical modeling using game
theory and computer simulation. In many species of bowerbirds the males
build amazing structures (bowers) and decorate them with colorful objects
to attract mates. Females choose mates by assessing bower quality, so a male
needs a good one to reproduce successfully. Females visit many bowers before
they choose, so there is tremendous competition among the males. The
Pruett-Joneses modeled three male strategies to determine what a male bower-
bird should do to be successful. A male could spend lots of time constructing a
great bower and then defend it against raiding by other males (defender). Or,
he could split his time between constructing and defending his own bower and
visiting other bowers to steal their decorations (stealer). Or, he could split his
time between constructing and defending his own bower and visiting other
bowers to destroy them (destroyer). By measuring the costs and benefits of
these strategies in terms of access to females, the game-theory model shows
that both destroyer and stealer are stable against defender under most circum-
stances. Simulations show that the range of conditions in which defender is
stable increases if intruders have to travel long distances between bowers or if
residents are able to repair damaged bowers quickly.
Regardless of the modeling method used, as with empirical methods, our
purpose remains testing hypotheses to answer the four principal questions of
animal behavior.

AFTER THE RESEARCH IS DONE

The primary goal of research is the production of new information, but in a
very real sense, that information does not exist until it is published (made pub-
lic in a scientific journal). Publications are the principal means by which scien-
tists communicate their findings, and publication success is often the most
important evaluation criterion for professional animal behaviorists. For the
most part, we publish in peer-reviewed journals, meaning that other experts
in animal behavior must review and approve a potential article. Successful
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 25

publication is no easy task. Most peer-reviewed journals only accept a fraction

(5–40%) of the papers submitted to them. In addition, some journals ask their
authors to pay to publish their research.

CONCLUSIONS
For those of us who have dedicated our lives to the study of animal behavior,
the goal remains explaining how and why animals do what they do. An excel-
lent example of this process, written without technical obfuscation, is More
Than Kin and Less Than Kind by Douglas Mock (2004). Our use of the meth-
ods outlined in this chapter have produced lots of valuable information, but
perhaps the most important point to make is that much more remains poorly
understood or completely unknown. Animal behaviorists will continue to use
the general methods described here and elsewhere to produce answers to that
most general question, “How and why do animals do the things they do?”
You will read many accounts of this process in the chapters that follow.

ACKNOWLEDGMENTS
I would like to thank the many people who contributed to my development
as a behavioral biologist, most especially my thesis advisors Val Nolan Jr. and
John M. Emlen, my postdoctoral advisor Peter Marler, and my friends and
colleagues at the Rockefeller University Field Research Center.

REFERENCES AND SUGGESTED READING

Bart, J., M. A. Fligner, & W. I. Notz (1998). Sampling and Statistical Methods for
Behavioral Ecologists. Cambridge, UK: Cambridge University Press.
Blackawton, P. S., S. Airzee, A. Allen, S. Baker, A. Berrow, C. Blair, et al. (2010).
Blackawton bees. Biology Letters, 7, 168–172.
Chamberlin, T. C. (1890). The method of multiple working hypotheses. Science, 15,
92–96.
Dawkins, M. S. (2007). Observing Animal Behaviour: Design and Analysis of Quantita-
tive Controls. Oxford, UK: Oxford University Press.
Dugatkin, L. A. & H. K. Reeve (1998). Game Theory and Animal Behavior. Oxford,
UK: Oxford University Press.
Hansen, A. J. & S. Rowher (1986). Coverable badges and resource defense in birds.
Animal Behaviour, 34, 69–76.
Holland, J. H. (1992). Adaptation in Natural and Artificial Systems: An Introductory
Analysis with Applications to Biology, Control, and Artificial Intelligence. Cam-
bridge, MA: MIT Press [Paperback reprint of 1975 original].
26 ANIMAL BEHAVIOR

Hutchinson, J. M. C. & J. M. McNamara (2000). Ways to test stochastic dynamic

programming models empirically. Animal Behaviour, 59, 665–676.
Kamil, A. C. (1988). Experimental design in ornithology. In R. F. Johnston (ed.),
Current Ornithology, Volume 5 (pp. 313–346). New York: Plenum Press.
Lehner, P. N. (1996). Handbook of Ethological Methods. Second Edition. Cambridge,
UK: Cambridge University Press.
Mangel, M. & C. W. Clark (1988). Dynamic Modeling in Behavioral Ecology. Prince-
ton, NJ: Princeton University Press.
Martin, P. & P. Bateson (2007). Measuring Behaviour: An Introductory Guide. Third
Edition. Cambridge, UK: Cambridge University Press.
Maynard Smith, J. (1976). Evolution and the theory of games. American Scientist, 64,
41–45.
Mock, D. W. (2004). More Than Kin and Less Than Kind. Cambridge, MA: Belknap
Press.
Nero, R. W. (1956a). A behavior study of the red-winged blackbird. I. Mating and
nesting activities. Wilson Bulletin, 68, 5–37.
Nero, R. W. (1956b). A behavior study of the red-winged blackbird. II. Territoriality.
Wilson Bulletin, 68, 129–150.
Orians, G. H. (1969). On the evolution of mating systems in birds and mammals.
American Naturalist, 103, 589–603.
Orians, G. H. (1985). Blackbirds of the Americas. Seattle: University of Washington
Press.
Orians, G. H. & G. M. Christman (1968). A comparative study of the behavior of
red-winged, tricolored, and yellow-headed blackbirds. University of California
Publications in Zoology, 84, 1–81.
Peek, F. W. (1972). Experimental study of territorial function of vocal and visual dis-
play in the male red-winged blackbird (Agelaius phoeniceus). Animal Behaviour,
20, 112–118.
Platt, J. R. (1964). Strong inference. Science, 146, 347–353.
Ploger, B. J. & K. Yasukawa (2003). Exploring Animal Behavior in Laboratory and
Field. San Diego, CA: Academic Press.
Popper, K. (1959). The Logic of Scientific Discovery. London: Hutchinson and
Company.
Pruett-Jones, S. & M. Pruett-Jones (1994). Sexual competition and courtship disrup-
tions: Why do male bowerbirds destroy each other’s bowers? Animal Behav-
iour, 47, 607–620.
Searcy, W. A. (1988). Dual intersexual and intrasexual functions of song in red-
winged blackbirds. Proceedings of the XIX International Ornithological Congress,
1, 1373–1381.
Singer, F. (1996). J. B. S. Haldane and the evolution of the Hardy-Weinberg model.
In J. Hagen, D. Allchin, & F. Singer (eds.), Doing Biology (pp. 163–173). New
York: Harper Collins College Publishers.
Smith, D. G. (1976). An experimental analysis of the function of red-winged black-
bird song. Behaviour, 56, 136–156.
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 27

Tillberg, C. V., M. D. Breed, & S. J. Hinners (2007). Field and Laboratory Exercises in
Animal Behavior. London: Academic Press.
Tinbergen, N. (1963). On aims and methods of ethology. Zeitschrift für Tierpsycholo-
gie, 20, 410–433.
Trivers, R. L. (1972). Parental investment and sexual selection. In B. Campbell (ed.),
Sexual Selection and the Descent of Man, 1871–1971 (pp. 136–179). Chicago:
Aldine.
van Emden, H. (2008). Statistics for Terrified Biologists. Malden, MA: Blackwell
Publishing.
Whitlock, M. C. & D. Schluter (2009). The Analysis of Biological Data. Greenwood
Village, CO: Roberts and Company.
Yasukawa, K. (1979). Territory establishment in red-winged blackbirds: Importance
of aggressive behavior and experience. Condor, 81, 258–264.
Yasukawa, K. (1981). Song and territory defense in the red-winged blackbird. Auk,
98, 185–187.
Yasukawa, K. (2010). Experiment, observation, and modeling in the lab and field. In
M. Breed & J. Moore (eds.), Encyclopedia of Animal Behavior (pp. 679–685).
Oxford, UK: Academic Press.
Yasukawa, K. & W. A. Searcy (1986). Simulation models of female choice in red-
winged blackbirds. American Naturalist, 128, 307–318.
Zar, J. H. (2010). Biostatistical Analysis. Fifth Edition. Upper Saddle River, NJ:
Prentice Hall.

APPENDIX: A PRIMER OF INFERENTIAL STATISTICS

Assuming that the sampling or experimental design is correct, statistical
analysis consists of answering a series of questions such as the following. (Note
that these questions will not apply to all research in animal behavior.)

• What is the biological question or hypothesis?

• What predictions can you make from this hypothesis?
• For one of the predictions, what are the null and alternative hypotheses of a statis-
tical test?
• Is your hypothesis one-tailed (expected effect in one direction) or two-tailed
(expected effect in either direction)?
• Does your hypothesis involve correlation or regression analysis?
• If not, how many groups were sampled, and were they related or unrelated?
• Was there a dependent variable, and if so, what was it?
• Was the effect that you measured nominal, ordinal, or interval/ratio?
• What category of statistical inference tests is appropriate?
• What are the available inference tests in that category?
• What are the assumptions of each test, and are they met?
• Which test is most appropriate?
28 ANIMAL BEHAVIOR

• What are the values of the test statistic and their associated probability?
• Are the results statistically significant?
• How powerful was the test?
• What is the answer to your biological question?

An Example
One way to help you to understand the process of statistical analysis is to work
through the following example. Suppose you want to compare the mating suc-
cess of two kinds of male frogs. Your data consist of numbers of mates acquired
by two groups of subjects: males in one group were calling when observed, but
those in the other group were not calling when observed. On average, 20 calling
frogs attracted five mates each, although some calling males attracted more and
others fewer (mating success varied). In contrast, the 20 silent males attracted a
mean of three mates, again with variation in mating success.

• What is the biological question or hypothesis?

Question: Does male calling behavior affect mating success?
Hypothesis: Calling by males functions as a means of attracting mates.
Note that the hypothesis answers the question “How or why does the behavior
occur?”

• What predictions can you make from this hypothesis?

Here are a few examples: loud males attract more mates than soft callers;
males who call frequently attract more females than infrequent callers; calling
males attract more females than silent males. Note that predictions are state-
ments of what would result from a specific observational or experimental study
if the hypothesis is correct. So, if calling functions to attract mates, then males
who call frequently will attract more females than infrequent callers.

• For one of the predictions, what are the null and alternative hypotheses of a statistical
test?
• Is your hypothesis one-tailed (expected effect in one direction) or two-tailed (expected
effect in either direction)?

For a two-tailed statistical hypothesis test,

HO: The number of mates attracted will not differ between calling males and
silent males.
HA: The number of mates attracted will differ between calling males and silent
males. (Note: this alternative is two-tailed because a difference in either
direction will cause us to reject HO.)
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 29

The hypothesis should be one-tailed, however, because we were testing the

hypothesis that calling enhances mating success:
HO: Calling males will attract fewer than or as many mates as silent males.
HA: Calling males will attract more mates than silent males.
The two-tailed hypothesis is preferable when there is no a priori reason to
expect results in a particular direction.
The null and alternative hypotheses written above clearly specify the inde-
pendent and dependent variables. When first attempting to write these hypoth-
eses, my students often make vague statements such as “the groups will differ.”
My students sometimes also incorrectly pair a two-tailed null with a one-tailed
alternative hypothesis (or vice versa). For example, a student might pair “HO:
the number of mates attracted will not differ between calling males and silent
males” with “HA: calling males will attract more mates than silent males” (the
correct HA would be “the number of mates attracted by calling males will differ
from that attracted by silent males”). With clearly written hypotheses, you will
have a much easier time determining how many treatment groups (levels) you
have and whether you have nominal, ordinal, or interval/ratio data.

• Does your hypothesis involve correlation or regression analysis?

• If not, how many groups were sampled, and were they related or unrelated?
The hypothesis does not involve correlation or regression. Only the dependent
variable is numeric (number of females). The independent variable, calling, is
divided into just two discrete categories, calling and not calling. Two indepen-
dent groups (calling and silent) were sampled. The groups are independent
because each subject either called or did not call.

• Was there a dependent variable and, if so, what was it?

Yes, the dependent (measurement) variable was number of mates per male.
Students often have trouble distinguishing the dependent from the inde-
pendent variables. You need to consider which variable responds to changes
(dependent) in the other (independent).

• Was the effect that you measured nominal, ordinal, or interval/ratio?

Because the number of mates per male is an integer (we could also have a fre-
quency: number/hour), this variable can be treated as interval/ratio data.

• What category of statistical inference tests is appropriate?

One-tailed tests of two unrelated samples are appropriate for interval/ratio data.
30 ANIMAL BEHAVIOR

• What are the available inference tests in that category?

Two-sample Student’s t-test (a parametric test)

Mann-Whitney U-test, Wilcoxon two-sample test (nonparametric tests)

• What are the assumptions of each test, and are they met?
A standard t-test assumes random samples, equal variances, and normality.
(Note: a variant of the t-test allows us to test two samples with unequal var-
iances.) Although I have not discussed doing so, you would need to test
both assumptions to decide whether they are met in this case. For example,
you could use a Bartlett F-test for equality of variances, and a Shapiro-Wilk
W-test of normality. Such testing is beyond the scope of this chapter,
however.
Nonparametric tests such as the Mann-Whitney U-test do not have these
restrictive assumptions, although they do assume random sampling. Nonpara-
metric tests are also less powerful than corresponding parametric tests if the
assumption of normality is met.

• Which test is most appropriate?

We would prefer to use the t-test if its assumptions are met because it is most
powerful, but if the variances are unequal or if either sample is not normal,
then we should use a nonparametric test such as the Mann-Whitney
U-test.

• What are the values of the test statistic and their associated probability?
You would calculate (or have a computer calculate) the value of t or U,
depending on which test we have chosen. This value of this test statistic is then
used to calculate an associated probability, the P-value (probability of a type I
error—incorrectly rejecting HO).

• Are the results statistically significant?

We would conclude that the results are significant if P ≤ α (typically α ¼ 0.05).
In our example, a significant result (P ≤ 0.05) tells us that the two means are
significantly different (probably not the same). Note that with α ¼ 0.05, we
can expect to reject the null hypothesis incorrectly (i.e., to make a type I error)
1 time in 20. In other words, even though the difference was significant, there
is still a chance that the two groups do not differ in mating success. In other
words, “statistically significant” does not mean “truly different” (because there
is an α probability that we made a type I error).
THIS IS HOW WE DO IT : THE SCIENTIFIC STUDY OF ANIMAL BEHAVIOR 31

• How powerful was the test?

Power analyses will tell us, for a specified set of conditions, how likely we are
to reject the null hypothesis when we should reject it (i.e., to avoid making a
type II error), what sample size would be necessary to reject the null hypothe-
sis, and how large a difference we might be able to detect.

• What is the answer to your biological question?

Remember that the purpose of performing statistical inference testing is to
help us answer a question about animal behavior; a statistical conclusion is
not the answer—it is a means to help us choose the answer.
If P ≤ α, then the two kinds of males differed significantly in mating suc-
cess. In this case, we would say, “The two kinds of males differed significantly
in mating success (t ¼ 2.71, df ¼ 38, P ¼ 0.005).”
On the other hand, if P > α, then the two kinds of males did not differ sig-
nificantly in mating success. We would say, “The difference in mating success
of the two groups of males was not significant (t ¼ 0.130, df ¼ 38, P ¼
0.20).”
Note that when reporting results we use the past tense, and we give the value
of the test statistic (e.g., t ¼ 0.130), the degrees of freedom (e.g., df ¼ 38) or
sample size (depending on the test), and the probability (e.g., P ¼ 0.20). Also,
we say “not significantly different” rather than “the same” (after all, the groups
means were 5 and 3, which are not the same).
This page intentionally left blank
 2

In the Beginning: A History of Animal

Behavior
Lee C. Drickamer

INTRODUCTION
When we think of animal behavior today, visions of zoological parks, Animal
Planet or National Geographic Channel television programs, or some wildlife
observation experiences of our own come to mind. Many individuals reading
an essay such as this are likely interested in the modern study of animal behav-
ior. The purpose of this chapter is to explain the history of animal behavior
and how we got to the current, rich history of a discipline known by different
groups as animal behavior, ethology, or comparative psychology, depending
on training and perspective. This includes the acquisition and accumulation of
knowledge concerning animals and their actions over many millennia and led
eventually to the systematic collection of natural history information begin-
ning before Greek and Roman times and continuing today. The topics
covered build from this early natural history to the scientific study of animal
behavior, which effectively commenced in the nineteenth century, grew
stronger in the first half of the twentieth century, and exploded along several
pathways during the final decades of that century. In the chapter, we also
explore some of the ways that information about animal behavior is acquired
and some of the techniques, both methodological and statistical, that scientists
use to investigate animal activities and their importance. The chapter
34 ANIMAL BEHAVIOR

concludes with brief comments on what lies ahead for the study of animal
behavior.

EARLY HUMAN HISTORY

As humans emerged as Homo sapiens they faced the same issues as most ani-
mals. They needed to find shelter and protection from weather and changes
occurring in their habitat caused by major climatic events. They organized
into some form of social system and engaged in patterns of reproduction and
rearing of offspring to produce subsequent generations. They had to locate
foods providing nourishment and nutrition—a diet that generally included
both animals and plants. Becoming proficient in methods to catch and kill
animal life as a food source required trial-and-error attempts; knowledge
gained was passed to subsequent generations. Animals that served as food
sources included a vast array of invertebrates as well as all types of vertebrates,
ranging from fish to mammals, and were found in both terrestrial and aquatic
habitats. Last, and perhaps most important, early humans faced the constant
threat of being eaten by a variety of carnivorous organisms. Knowing the hab-
its and haunts of the larger mammals, as well as some reptiles and large fish,
was requisite for survival. Adopting behavioral strategies for dealing with pred-
ators might include having individuals as lookouts, avoiding areas of potential
predator concentration, or even devising ways to kill predators to reduce their
numbers.
While no writing yet existed in this period, from before 50,000 years ago
until about 5000 BCE, we do have information concerning early knowledge
of animal behavior. Many artifacts are associated with areas of early human
habitation. These include physical objects made from bone, wood, and stones,
as well as depictions of people and animals on cave walls. Among the utilitar-
ian objects are hunting tools such as arrowheads, scrapers, spears, and atlatls.
These tell us a great deal about the types of animals that were hunted and
how the prehistoric peoples caught, killed, and processed the prey to use its
various parts to meet their needs. For pictures and descriptions of tools used
by early humans during prehistoric times see, for example, (1) http://www
.templeresearch.eclipse.co.uk/bronze/ab.htm and (2) http://irisharchaeology
.ie/work.
Among the items excavated from early human sites are carvings and sculp-
tures, which represent early forms of human religion. These include small
pieces made from stone, wood, bone, and other materials that depict animals,
providing insights into how they were significant to these peoples. Other
objects are vessels with carvings, incising, or artwork, again depicting, in some
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 35

instances, animals. Jewelry for body adornment contains animal forms and
representations, emphasizing the importance of these animals with which the
prehistoric humans shared their world. For examples of these art forms and
information on their interpretation, see: (1) http://en.wikipedia.org/wiki/
Prehistoric_art, (2) http://ezinearticles.com/?Religious-Jewelry:-Prehistoric
-But-Persistent-Symbol-of-Spirituality&id=5880415, and (3) http://art
historyresources.net/ARTHprehistoric.html.
Cave paintings are known from a variety of locations on all of the conti-
nents except Antarctica and date from as far back as 30,000 BCE to recent
centuries. These artworks are a primary resource for interpreting the interac-
tions of early humans with the other animals sharing their world. Two major
types of art are known: (1) pictographs are, as the name implies, pictures
painted on rocks, cave walls, or other generally protected surfaces and
(2) petroglyphs are areas where rock has been chipped, pecked, or scraped to
produce a recognizable form or symbol. Excellent examples of the ways in
which animals were depicted can be found on the Internet. For examples,
see (1) http://www.petroglyphs.us/ and (2) http://en.wikipedia.org/wiki/
Cave_painting.
Often the animals shown are species such as ungulates, bears, or similar-
sized mammals that served as food. Other depictions represent symbolic
connections between the people and particular groups of animals; the clan
concept used by many Native American tribes often involves animal symbols,
and these are depicted in pictographs and petroglyphs. From some of the rock
and cave art, scientists can interpret aspects of both the hunting techniques
used, such as cooperation in driving animals off cliffs, and the social structure
of the group. In other cases the rock art has clear religious meaning. For in-
stance, we know of depictions of creation myths, gods and goddesses, and
what likely are religious rituals. These finds tell us things about human-
animal interactions and also about the behavior of our ancestors.
A key feature of early human civilizations was the beginning of domestica-
tion of both animals and plants. Knowledge of plants resulted in better and
more reliable stocks of food. A cumulative knowledge of such things as the
behavior of insect pests, rodent damage to crops, and consumption of grains
and other foods by wild animals all contributed to the ability of the early peo-
ples to engage in successful, though by today’s standards primitive, agriculture.
Animals were domesticated as well; much of this occurred over many genera-
tions and involved considerable knowledge gained via observations of wild ani-
mals. Some animals, such as goats and cattle, provided meat, milk, hides, and
bones for tool making. Others animals, horses for example, served to carry
goods and people. Finally, some wild canids, and later likely felids as well,
36 ANIMAL BEHAVIOR

associated with human groups and soon became companion animals, a mutu-
alistic relationship in which the canids were cared for by the humans and, in
turn, provided both extra eyes and ears for detection of prey or possible pred-
ators and true companionship.
In all of these instances of domestication, some nascent forms of selective
breeding were introduced as a means of ensuring that desirable traits were
preferentially passed on to the next generation. Early humans had no knowl-
edge of genetics, but nonetheless they were quite capable of putting their
observations to advantage when doing controlled matings. Because the process
of human cultural evolution is generally best viewed as a continuum and not a
step process, some of what we have just discussed carried over into and was
refined during subsequent years, covered in the next section.

EARLY CIVILIZATIONS IN ASIA AND THE MIDDLE EAST

It is perhaps questionable to put these widely disparate geographic areas,
comprising almost all of Asia from the Middle East to the Far East, into a gen-
eral commentary. However, during the period from about 8000 BCE until the
start of the Christian era, events in these locations occurred in parallel, with
regional interactions and some groups supplanting others over time.
In places such as Mesopotamia and Egypt, animals were a common part of
daily life, including both livestock and household companions. Domestication
of plants resulted in true agriculture with irrigation, beginning about
6500 BCE. At the same time, planned breeding of animals with specific desir-
able traits, for example less aggression or larger litter size, became common
practice, resulting in full domestication of animals ranging from cats and dogs
to various ungulates, egg-laying birds, and birds of prey such as falcons. This
was the beginning of what we now call animal husbandry, combining knowl-
edge of the animals’ traits and needs with human uses for the animals; animal
husbandry continues as a key aspect of daily life even now. Many animals were
considered as representations of gods and were often given special treatment
during their lives and special burial ceremonies. We are all familiar with the
myriad breeds of dogs existing today; these initially began as human attempts
to produce dogs with specific traits providing efficient aid for particular func-
tions. Some were bred as watchdogs, others to assist in hunting, some to help
herd livestock, and so forth. The critical aspect of almost all dog breeds is that
they are selected to possess desirable behavioral traits, requiring concentrated
observations of and knowledge about daily activities and physical traits.
Cats were a favorite in Egypt and are frequently depicted in various art
forms, rising to the level of gods. In addition to the mammals, birds and
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 37

certain insects were important to the Egyptians. A phenomenon we call

anthropomorphism enters into many aspects of the human–nonhuman ani-
mal relationship. The term anthropomorphism means the attribution of human
traits or characteristics to nonhuman animals (see Chapter 1) and sometimes
to inanimate aspects of nature such as wind or fire. Particular animals gained
status as they came to represent traits of humans; this process points directly
to human observations of the animals leading to the connection and trait attri-
bution. Examples include jackals to represent protection, cattle representing
good aspects of motherhood, and falcons that signified Horus, a primary deity.
Over time, knowledge of astronomy, mathematics, and the physical sci-
ences was far advanced in these Middle Eastern locations relative to what
was known in Europe and even in the Far East. An outcome of this was a basic
framework for conducting scientific studies that involved data collection and
testing of hypotheses, later expanded by the Greeks and Romans. While
aspects of life sciences, including studies of anatomy, were subjected to the sci-
entific process, animal behavior does not appear to have been a subject of these
investigations.
All groups of humans experience phenomena, including birth, death, disas-
ters, and similar events, for which they lack satisfactory explanations. Even
things we take for granted, such as the daily appearance of the sun and the
changing seasons, raised key questions. Most of the phenomena were
explained through mythology, which consists of a collection of stories belong-
ing to a people and addressing their origin, history, and deities, as well as key
life events. Though we lack the space here to provide details, there are many
resources available that emphasize that the human understanding of behavior
lead to incorporation of animals with specific traits into myths used to describe
these otherwise inexplicable events.
In India, the cow was of primary significance. It was the basis for many
beliefs but was also the focus of much early work on veterinary medicine.
Knowledge of animal behavior was critical to the husbandry practices devel-
oped by early Indian subcontinent peoples. Given their importance for breed-
ing, bulls were also the subject of behavioral studies and were highly revered in
early Indian cultures. The origins of the cow as a sacred being derive from this
period in the first millennium BCE. Specialized breeding programs lead to
cows that produced more milk and cows that were better mothers to their
calves.
India was home to domesticated elephants, trained to perform a variety of
tasks, including both practical, everyday uses like plowing fields or carrying
burdens and service in warfare. Elephants served to carry kings and other roy-
alty. A key element of these practices was the studies of trackers, individuals
38 ANIMAL BEHAVIOR

skilled at reading elephant signs and behavior to aid in the capture and domes-
tication of the elephants. As in the other cultures developing at this time, dogs
and horses were domesticated and bred for specific purposes.
In the Far East, including China, Japan, and Korea, animal behavior was
significant in events as diverse as birth months and annual calendars. Early
forms of astrology were prevalent in Asia, using specific constellations named
for animals to ascribe powers and behaviors of the particular animal to an indi-
vidual born at a time of the year when that constellation was prominent or
when the sun was passing through the constellation. The Chinese calendar
uses a sequence of 12 animals, each with different behavior patterns; each year
is assigned a particular animal, and the nature of that creature provides a basis
for foretelling and interpreting the events of that year. The Chinese calendar
and the zodiac used by astrologers in more Western cultures are often
compared.
The prediction of earthquakes by the actions of nonhuman animals, which
continues today as a topic of scientific interest, likely has its origins in ancient
China. Animals of a variety of types were often observed engaging in what
might be described as unusual or aberrant activities in advance of an earth-
quake. It is possible that the association of these activities with earthquakes
builds upon the knowledge about major storms that we associate with changes
in animal behavior.
Divination, or practicing the art of foretelling future events or discovering
knowledge by interpreting omens or signs, is known from all of the foregoing
regions among early people in Africa and Europe, and later in the Americas.
Those who practiced divination were said to have supernatural powers and
to be capable of seeing into the future. Among the literally hundreds of meth-
ods of divination, several use animals and knowledge of animal behavior as
sources of information to make predictions. Theriomancy is a general term
for the use of animals in divination. A wide variety of animals from insects
to mammals are used in this practice. Examples include birds, where flight
and migration patterns, flock size, and various calls are used as information
sources. Another example is rodents, whose destructive capabilities were inter-
preted as an evil omen.

GREECE AND ROME

Overlapping with the civilizations just discussed and extending into the
first millennium CE, peoples in Greece, and later Rome, provided advances
in science, including animal behavior. Some individuals made significant con-
tributions, and the interactions among those who were our first scientists
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 39

contributed to the progress made during the period from early Greece in about
2700 BCE until the onset of the Byzantine Empire in about 600 CE.
Aristotle (384–322 BCE) put forth a scheme for the classification of living
organisms, and though his ladder-like proposal was later replaced, it is a mea-
sure of overall progress in understanding the natural world. He was a key figure
in terms of systematic recordings of natural history observations, preserving
findings that could be replicated by others. His extensive records on marine
organisms such as starfish and mussels produced what we may call the first
ethograms; an ethogram is a record of all of the behavior patterns and activities
of an organism and may be in the form of text or illustrations. Another aspect
of Aristotle’s work concerned use of the comparative method—combining
and contrasting observations on organisms of the same species over time or of
organisms from different species. Finally, the underpinnings of major aspects
of modern philosophy are part of Plato’s (429–347 BCE) and Aristotle’s leg-
acy. Science is based on philosophical principles, and, indeed, modern science
is an outgrowth of philosophy.
Studies of anatomy and of domestic animal activities furthered the under-
standing of behavior during the Greco-Roman era. Better knowledge of
anatomy lead to a better grasp of behaviors related to locomotion and feeding;
differences in animal structures were related to their functions in movement
and feeding. After humans solved problems pertaining to growing food and
avoiding predators, they consistently worked on aspects of animal domestica-
tion and husbandry. Indeed, for this author, the roots of animal behavior as
a science lie in the field of animal husbandry. Even in the early years of the
primary journal in our field, Animal Behaviour, in the 1950s, a significant pro-
portion of the published articles pertained to the behavior of domestic
animals.
A quotation from Aristotle provides insight regarding the perceptiveness of
animal observations:
They say that the cuckoos in Helice, when they are going to lay eggs, do not
make a nest, but lay them in the nests of doves or pigeons, and do not sit, nor
hatch, nor bring up their young; but when the young bird is born and has
grown big, it casts out of the nest those with whom it has so far lived. (1909,
p. B1)

This is exactly what happens with parasitic birds such as cuckoos and cow-
birds. However, not all that was recorded was correct, as the next quote
illustrates:
Men say that tortoises, when they have eaten part of a viper, eat marjoram as an
antidote, and, if the creature fails to find it at once, it dies. (1909, p. B2)
40 ANIMAL BEHAVIOR

In addition to the extensive writings of Aristotle, Pliny the Elder (23–79

CE) compiled a 37-volume compendium of all aspects of the world from
astronomy to zoology. Considerable sets of observations of a variety of animals
are included in several of these volumes. Pliny’s Natural History remained an
importance source of information for 1,500 years. A quote from Pliny con-
cerning peacocks gives the flavor of his writings:
When it hears itself praised, this bird spreads out its gorgeous colors, and espe-
cially if the sun happens to be shining at the time, because then they are seen in
all their radiance and to better advantage. At the same time, spreading out its
tail in the form of a shell, it throws the reflection upon the other features which
shine all the more brilliantly. (Natural History, Book 10, 22–23)

As with Aristotle, not everything written in the animal volumes of Pliny’s

treatise was true:
Hyenas can imitate the human voice among the stalls of the shepherds; and
while there, learns the name of some one of them and then calls him away
and devours him. It is said also that it can imitate a man vomiting and that,
in this way, attracts the dogs and then falls upon them. (Natural History, Book
8, 30)

A LOST MILLENNIUM
The period from the decline of the Roman Empire (ca. sixth century CE)
until the start of the Renaissance was, as some have called it, the Dark Ages.
Other sources define this period as the millennium from 200 to 1200 CE.
Progress in many areas of science was stagnant or proceeded only in small
stages, although many areas of human endeavor were quite productive and
the label “Dark Ages” is something of a misnomer. Some science relapsed into
the use of mythology and supernatural explanations for natural history obser-
vations. Religions, primarily Christianity in Europe and Islam in the Middle
East (beginning in the seventh century CE), were dominant and served as
the bases for interpretation of both human life events and the activities of all
living creatures.
With respect to knowledge about and interest in animal behavior, there
were several noteworthy developments. A modest increase in trade with areas
that included Asia and Africa resulted in accumulated knowledge of these
“exotic” locations, including their animals. In the fifteenth century, explora-
tion of other geographic locales began on a large scale with ships traveling to
the Americas and around the globe, further expanding collections of animals,
living and preserved, in the nations of Europe.
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 41

The Renaissance that followed was built upon events of the fourteenth and
fifteenth centuries, including the expanded knowledge of other parts of the
Earth. The invention of the printing press in the mid-1400s replaced the labo-
rious process of hand-copying text for information dispersal and storage. At
about the same time, new methods of illustration were developed that initially
involved woodcuts and later etching and engraving. The former lead to wide
dissemination of knowledge about topics including animal behavior, and the
latter resulted in means for diagramming and depicting animals engaged in
their daily activities, individually and in groups.

RENAISSANCE, THE RISE OF NATURAL HISTORY, AND THE INITIAL STUDIES OF

ANIMAL BEHAVIOR
The term Renaissance is generally ascribed to a period of renewed intellec-
tual and creative activity that lasted from the mid-1400s to about 1650, pre-
dominantly in Europe but also spreading to other regions of the globe. A
key figure during the Renaissance was René Descartes (1596–1650), whose
work influenced future science in at least three principle ways. His approach
was to divide a problem into as many smaller, constituent problems as practi-
cal and then explore each of those. Many studies of animal behavior, as for
example understanding annual reproduction in doves, follow this scheme.
By looking at hormonal effects, courtship, nest construction, interactions
between the female and male parents, egg incubation, and many other topics,
investigators can piece together the complete story.
A second aspect of Descartes’s approach is that science works through deduc-
tive reasoning. This implies that to fully test an idea we need to gather informa-
tion, based on a hypothesis and using standardized methods, and then arrive at
conclusions (see also Chapter 1). In an example from animal behavior, we
hypothesize that testosterone is a key factor in male hamster aggression. To test
this we need to conduct several experiments. First, we measure aggression and
testosterone levels in groups of male hamsters. Our prediction is a positive cor-
relation between higher testosterone levels and more aggression. Next, we cas-
trate males to “level the playing field,” and we predict significant declines in
aggression. Finally, we use exogenous hormone to provide varying levels of tes-
tosterone to males. Here we predict that the levels of exogenous hormone pre-
dict behavior—higher testosterone results in more aggression. Thus, the result
is a function of a set of experiments, not a simple test but a sequence designed
to assess the possible causal relationship between the hormone and behavior.
A third tenet of Descartes’s approach involves the need to be absolutely
confident that the information gathered during the course of an investigation
42 ANIMAL BEHAVIOR

is obtained objectively. Today this can best be divided into two parts. Our
methods should be designed to ensure that we do not exert any undue bias
in our data collection. Particularly in animal behavior, there can be a tendency
to see what we want to see—observation is at the heart of much of what we do,
and thus there is a need for both observer objectivity and, often, the use of
multiple observers to ensure accuracy. Also, today data must be verifiable;
some other investigator or laboratory should be able to conduct exactly the
same steps we did, using the same types of test animals, apparatus, and meth-
ods, to obtain identical or nearly identical results.
During the Renaissance, individuals such as Leonardo da Vinci (1452–
1519), William Harvey (1778–1657), and Andreas Vesalius (1514–1564)
contributed to a better understanding of aspects of anatomy and physiology.
Additional work on aspects of veterinary medicine in the Middle East and
India provided insights into the behavior of some domestic animals. The Arab
agricultural revolution produced several ideas that are forerunners of ideas in
modern ecology that have strong behavioral implications. One of these was
the notion of food chains, which can be used to describe the links between
predator and prey and also imply some understanding of the behavior patterns
of animals either seeking food or avoiding being eaten. A second idea, the
notion that there is a constant struggle among animals for food and other
resources, can be viewed as a precursor of the struggle for existence among
animals of the same and different species.
By the seventeenth century, considerable knowledge about natural history
was recorded and accumulated, including many observations of and thoughts
about animal behavior. However, no separate studies or writings existed spe-
cifically for animal behavior. Ideas about how best to conduct observations
and the beginnings of experimental science set the stage for new developments
in what we know today as animal behavior during the latter portion of the
seventeenth century.
Among the naturalists who contributed to the growing body of information
were John Ray (1627–1705) and his student Francis Willoughby (1635–
1672), both Englishmen. Together they produced volumes on the plants
and animals of areas of England, most notably from natural observations.
Ray believed that similarities among organisms were significant but that these
similarities required careful examination of the traits on which the compari-
sons were based. So, for example, earlier connections made between bats and
birds because they both flew in the air were changed to placing the animals
with kin according to the existence of fur or feathers. Ray was among the first
to write about instinct, based in part on his observation that birds, even when
reared by hand, built nests using similar materials to their wild counterparts,
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 43

even though they had not had prior access to these materials. Though religion
was still a strong factor in terms of the role of science in seeking to explore
God’s creations, the underlying thinking was critical to both observing and
writing about behavior of these and many other animals.
Other naturalists also contributed to the growing trend of making careful
examinations of animal traits in natural settings. Maria Sibylla Merian
(1647–1717) produced considerable information on the life history traits of
various insects in Europe and South America. Life history traits are founda-
tional for the investigation of many aspects of animal behavior. Antoine
Reamur (1683–1757) studied the habits and social aspects of ants, and Jan
Swammerdam (1637–1680) worked out aspects of insect metamorphosis.
The discovery of microscopy by Anton von Leeuwenhoek (1632–1723)
enabled naturalists and scientists to examine, in fine detail, the anatomy and
behavior of smaller organisms.
In addition to the work of the naturalists and other scientists, some key
developments portended the rise of science. One group of changes involved
the appearance of various venues for scholars to present and record their find-
ings. What began as a sharing of abstracts among colleagues in England and
Europe developed into societies where presentations of findings were dis-
cussed. Published articles, which began as summaries of previously presented
work, evolved into written accounts of new information and eventually to cov-
erage of experimental work. Much later, in the eighteenth and nineteenth cen-
turies for some disciplines but not until the twentieth century for animal
behavior, specialized journals emerged, covering topics in more depth and
with common themes. The first journal devoted specifically to behavior, the
Journal of Animal Behaviour, appeared in the first decade of the last century
and lasted for just five years. By midcentury, several journals for studies of ani-
mal behavior existed, including Behaviour, British Journal of Animal Behaviour
(later Animal Behaviour), and the Zeitschrift für Tierpsychologie (later Ethology).
As writings about scientific topics increased, some individuals began accu-
mulating personal libraries, saving and organizing copies of publications. A
few of the nascent scholarly societies also started library collections. Institu-
tional libraries came into existence at various colleges and universities and
often benefited from donations of libraries from individuals or estates. Collec-
tions of publications gave opportunities for a broad range of scholars with
interest in scientific topics to peruse what was known on a particular topic,
perhaps stimulating their own thinking.
A third seventeenth-century change involved expansion of animal collec-
tions, including living specimens that could be observed and studied. Some
wealthy individuals created zoos or “menageries” as collections of exotic
44 ANIMAL BEHAVIOR

animals housed on their estates. An early example was the Menagerie du Parc
at Versailles (1660s), though many others were created at about this same time
in England, Europe, and areas as geographically widespread as China and
India. These menageries were the province of wealthy aristocrats with an
emphasis on animal husbandry issues, but some observations valuable to
understanding behavior formed a part of these efforts. True zoological parks,
where the public could see and learn about captive animals, did not come into
existence until the last half of the eighteenth century with, for example, the
Tiergarten Schonbrunn at Vienna. As we know today, zoos are an integral part
of work on animal behavior, particularly with regard to conservation efforts
and education.

EIGHTEENTH-CENTURY ANIMAL STUDIES

The years of the eighteenth century witnessed the expansion of the prac-
tices of the previous century, the development of both classification schemes
for living organisms and early ideas about evolution, and the first work that
can be clearly defined as an experimental study of animal behavior.
Those who engaged in continuing and expanding natural history work
included George Louis Leclerc Comte de Buffon (1707–1788), Carolus
Linnaeus (1707–1778), and Jean Baptiste de Monet Lamarck (1744–1829).
Buffon produced 44 volumes of detailed observations and careful drawings,
containing much information about animal behavior. He was one of the first
to examine the distributions of living organisms across various regions, a fore-
runner of what we now call biogeography. Gilbert White (1720–1793) wrote
a collection of natural history observations published as The Natural History of
Selbourne, a book still commonly available and read by people embarking on
making natural history observations and recording them.
The early menageries expanded, some becoming game parks associated with
estates, and as the century proceeded, true zoological parks, where the public
was welcome, became more common. In related developments, some animals
were raised for the purpose of engaging in fighting, as with dogs, bulls, and roost-
ers. In the course of working with animals in these contexts, more was learned
about husbandry, and much of this had clear components of animal behavior:
What sorts of traits would make the best fighting cock, and how could these be
bred into subsequent generations? What animals adapted best to captive condi-
tions in zoos and game parks, and how well did they reproduce? It is noteworthy
that early issues of Animal Behaviour, from the 1950s, contain a strong selection of
articles on the behavior of domestic animals, a continuation of scientists working
on applied aspects of animal behavior as was the case over many earlier centuries.
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 45

Linnaeus, from Sweden, is best known for developing a classification

scheme to organize information about living organisms. His hierarchical
scheme began with a species name combined with a second level, the genus;
together these terms are known as the binomial nomenclature. He organized
groups of organisms into the various levels of his scheme using physical char-
acteristics, internal anatomy, and, in some instances, behavioral traits. Having
such a scheme enabled scientists to begin comparative work, examining, for
instance, the similarities of animals living in forested habitats, or, specifically,
birds with regard to the entire range of characteristics associated with the
annual cycle of reproduction.
Lamarck developed the idea that organisms could pass along to the next
generation those traits that were acquired during a lifetime and that provided
an advantage to those who possessed the trait. A central feature of his formu-
lation was the idea that organisms could adapt in the course of their lifetime.
That is, changes could occur in both form and function across generations of
organisms of a particular species. Over several centuries, the central idea con-
cerning traits that could occur as adaptations in one generation and could then
be passed on as acquired traits to the next generation has been discredited in
the framework and with the mechanisms that Lamarck proposed. However,
as sometimes happens in science, the notion of inheritance of acquired charac-
teristics, which was part of Lamarck’s formulation, has reappeared in a differ-
ent form in terms of modern ideas about epigenesis, which we will discuss
later in this chapter.
One other individual, Thomas Malthus (1766–1834)—whose life, like
that of Lamarck, spanned the end of the eighteenth and beginning of the nine-
teenth century—made a key contribution that aided, later in the nineteenth
century, in the formulation of the theory of natural selection by Charles
Darwin. Malthus, concerned primarily with issues of human economics and
resources, postulated that as populations grow and resources come under
increasing demand the combination could pose a threat to the continued exis-
tence of humans. Erasmus Darwin (1731–1802), the grandfather of Charles
Darwin and a naturalist as well as a physician, wrote ideas about the interrelat-
edness of all living organisms and the possibility of evolution as a form of
change over time. Though he died prior to the birth of his famous grandson,
it seems likely that some of what he wrote was passed along to the young
Darwin.
For our examination of the early history of animal behavior, a key individ-
ual is Charles Leroy (1723–1789), a Frenchman who was the gamekeeper at
the menagerie maintained at the Versailles Palace outside Paris. He eventually
wrote a book detailing observations of the animals under his care. His writings
46 ANIMAL BEHAVIOR

support the notion that he was primary figure in the history of the field. First,
he claimed, in a fashion perhaps reminiscent of Descartes, that all observations
must include clear definitions of the behaviors being recorded and that all such
information should be recorded accurately. Second, he likely produced the
first “complete” catalogues of behavior and life history traits for various ani-
mals, what we would later call an ethogram. Third, he made comparisons
between animals and their differences in behavior, as between herbivores and
carnivores. Fourth, he felt that animal actions were not necessarily entirely
mechanistic but rather were influenced by differing needs and motivations.
In this way he is, in a sense, an early participant in the debate that continues
to this day about the varying and complex roles of genetics and experience in
affecting observed behaviors.

NINETEENTH CENTURY, PART I—TO DARWIN

During the first six decades of this century there were significant increases
in exploration of far-flung locations around the globe. Many expeditions
involved in-depth coverage of areas only partially known to Europeans, such
as the trip of the Corps of Discovery under the leadership of Meriwether Lewis
(1774–1809) and William Clark (1770–1838) in the United States, as well as
many other travelers in the western part of the continent. Others explored
areas of South America, the Pacific Islands, and regions of Asia. The increase
in both museum specimens and live animals housed in zoos resulted in an
enormous accumulation of materials for study and research. Large collections
developed at museums and educational institutions in North America and
Europe. The voyage of the Beagle, on which Charles Darwin (1809–1882)
sailed as a naturalist, is a prime example of the exploratory voyages and expe-
ditions that characterized the first half of the nineteenth century.
An ongoing controversy, exemplified by a debate between Georges Cuvier
(1769–1832) and Etienne Geoffrey Saint-Hilaire (1772–1844), centered on the
issue of whether species were fixed in terms of their traits or were continual adap-
tations to changing conditions occurring in nature. Cuvier argued that similarities
between organisms were due to common functions and did not derive from
common ancestry. He was a strong proponent of the idea that form followed
function. Saint-Hilaire was representative of an alternative viewpoint, namely that
animal characteristics are influenced by environmental conditions and thus can
exhibit transformations over time. While we know today that the latter viewpoint
is primarily the case, the discussions surrounding these ideas, which encompassed
aspects of animal behavior, were critical to the growing body of thought on how
species arose and how they might adapt to different environments.
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 47

The field of ecology, though not quite as fully developed as we know it to-
day, began to emerge during these decades based on the findings from the
explorations, providing a solid basis for developing principles that could be
applied across a wide range of locations and habitat types. These principles con-
cerned topics such as populations, communities of organisms of different
species, and interactions among organisms of the same and different species.
The nitrogen (N) cycle, an example of a biogeochemical cycle, was partially
deciphered and included key roles for behavioral processes like feeding and pre-
dation where N was passed from plants to animals and then to other animals.
If one examines what I like to call the “ladder of life,” atoms and chemicals are
at the bottom end and biomes and the biosphere are at the top end. In between,
the lower steps involve cells, physiology, and organ systems, arriving at the ani-
mal and its behavior near the center of the climb. Above the organism we find
populations, communities, and ecosystems before reaching the higher steps.
This is all by way of noting that the animal is the key both to the ways in which
form and function result in animal actions and to the ways those actions are then
the critical element in putting together the various ecological processes that
emerge from organisms living together. Beginning to understand the ecological
portion of the ladder and the functions of individual animals in an ecological set-
ting emerged as critical developments in the first half of the nineteenth century.
Other events happening in the life sciences over the same period were
important for animal behavior. Johannes Muller (1801–1858) provided key
insights into the functioning of the nervous system and also performed early
embryology experiments; his Handbook of Physiology brought together, for
the first time, work on biology (anatomy and physiology) with principles from
chemistry and physics. Principles of energetics and metabolic functions were
fashioned and tested by individuals such as Claude Bernard (1813–1878).
The clear relationships between muscle structure and function were elucidated
by William Bowman (1816–1892) and Albert von Kolliker (1817–1905),
among others. Key aspects of the sensory-motor systems of animals were
developed from work by Luigi Galvani (1737–1798) in the seventeenth cen-
tury and expanded upon by Emil du Bois-Reymond (1818–1896), who dem-
onstrated the passage of a change of electrical state as an impulse passes along a
neuron, and Francois Magendie (1783–1855), who helped separate the sen-
sory nervous system and the motor nervous system. These scientists worked
with combinations of anatomy and physiology, using careful dissections of
nerves and then studying the phenomena through the use of experimental
techniques such as nerve-muscle dissections and the kymograph. The work-
ings of the nervous system were summarized by T. H. Huxley (1825–1895)
in his 1877 treatise, Manual of the Anatomy of the Invertebrate Animals.
48 ANIMAL BEHAVIOR

Though direct observations of the behavior of the whole organism were not
a regular part of the experiments conducted by these individuals, behavior in
some form was often an endpoint. The experimental techniques and principles
developed during this period form the basis for more thorough examination of
underlying physiological and developmental aspects of behavior in the latter
half of the century.
For all of biology, the seminal event of this time period was the formulation
of the idea of evolution by natural selection put forward by Charles Darwin
in his 1858 paper and 1959 book, On the Origin of Species by Means of Natural
Selection. It should be noted that Alfred Russell Wallace (1823–1913), who
worked in the Malay Archipelago, came up with the same notions about evo-
lutionary processes as did Darwin. Darwin’s conception was developed ini-
tially in the decades following his voyage on the Beagle, but he did not
publish the idea until correspondence with Wallace revealed that both men
were on the same path with their ideas.
The framework for understanding evolution by natural selection grew, as
happens with many key ideas in science, from a mix of earlier ideas. In this
case, those prior ideas certainly included the work of Lamarck, Darwin’s
grandfather, and Malthus. Also important was the work of a geologist, Charles
Lyell (1797–1875), who, upon examining rocks and rock strata at a number
of locations, developed the notion of uniformitarianism. He proposed that
geological history encompassed long periods of time with gradual changes,
something that was important to Darwin’s later formulation.

NINETEENTH CENTURY, PART II—ETHOLOGY EMERGES AS A SCIENCE AND

DEVELOPS SEPARATE PATHWAYS
Commencing with Darwin and including his further writings and their
effects on science, the field of animal behavior emerged as a separate, distinct
concern for scientists during the final decades of the nineteenth century. Many
avenues of research blossomed involving both laboratory and field investiga-
tions. At this same time, the science of psychology was emerging as a discipline
and included a behavioral component, especially as regarding the processes
involved in human thoughts and actions. We should examine briefly each of
these threads.
The titles of additional books published by Charles Darwin included The
Expression of the Emotions in Man and Animals and The Descent of Man and
Selection in Relation to Sex. These two volumes, and other writings, contain
ideas that are fundamental to certain paths in the study of animal behavior,
particularly sexual selection but also various forms of social behavior and
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 49

examination of individual differences in behavior. Among those who followed

Darwin, a number engaged in studies pertaining to animal behavior.
Charles O. Whitman (1843–1910), an American, studied the evolutionary
bases for animal behavior. His early years were characterized by many stuffed
animals and some live ones sharing his parental home and dormitory room.
Whitman was both a committed naturalist and a keen observer of events
involving the whole animal and, at the other end of the size range, at levels
of embryological development studied through the microscope. He felt that
instinct preceded intelligence, based on his observations of a variety of organ-
isms, though his favorite group was the birds. His interests spanned genetics
and evolution, and though Whitman did not accept the new precepts from
Mendelian genetics, he did conduct extensive breeding experiments with
pigeons, selecting for both anatomical and behavioral traits.
Douglas Spaulding (1841–1877), an Englishman, was among the first to
conduct experimental studies of behavior. Some of his key work was on early
flight behavior in fledgling birds. He is generally accorded the honor of being
the first scientist to describe imprinting in birds. From his work, Spaulding
concluded that observed behavior was a combination of developmental experi-
ences and instinct. George John Romanes (1848–1894) used invertebrates as
his primary subjects, and his major focus was on the evolution of intelligence.
Among his published books are Animal Intelligence and Mental Evolution in
Animals. Romanes is credited as one of those who developed the field of exper-
imental psychology.
Conway Lloyd Morgan (1862–1936) published the first textbook on ani-
mal behavior, An Introduction to Comparative Psychology (1894), and his work
influenced the nascent fields of both comparative psychology and ethology.
Further, and partly in response to the use of anthropomorphic or anecdotal
evidence by some of his colleagues, Morgan put forward what is now known
as Morgan’s canon. Simply put, this is the idea that “in no case may we inter-
pret an action as the outcome of the exercise of a higher psychical faculty, if it
can be interpreted as the outcome of one which stands lower on the psycho-
logical scale” (Morgan, 1903, p. 59). This rule is still important in empirical
studies of behavior today and is taught to all students of animal behavior.
From the natural history perspective, the careful work on ants by William
Morton Wheeler (1865–1937), over many decades, revealed aspects of the
social life of a number of species. He is the author of a statement published
in Science in which he defined ethology as the best term for describing the sub-
field of zoology that examines the behavior of organisms toward the biotic and
abiotic aspects of their environment. Another general principle, the Umwelt,
was proposed by Jakob von Uexkill (1864–1944): the best way to understand
50 ANIMAL BEHAVIOR

the actions of animals is to examine the world in which they live from their
perspective, with their sensory and brain systems. This too is a key aspect of
most training of students in their thinking about animal behavior today.
In the United States, a long tradition of naturalists whose observations
included many aspects of animal behavior included individuals like John
Bartram (1699–1777), his son William Bartram (1739–1823), and John J.
Audubon (1785–1851). In addition, there were many naturalists attached to
expeditions of the nineteenth century exploring areas of the western United
States and Alaska who added to the legacy of important information, both at
the level of species discovery and identification and in terms of behavioral
observations. These included Thomas Say (1787–1834), who worked pri-
marily on descriptions of insects (more than 1,400 new species) but also on
birds; John E. LeConte (1784–1860), who studied a variety of vertebrates
and invertebrates and also plants; and Elliott Coues (1842–1899), a father of
modern ornithology from earlier in the nineteenth century. In the later deca-
des of the century were John Muir (1838–1914), who traveled throughout
the western United States and to Alaska, writing on many topics, including
animal behavior; John Burroughs (1827–1921), who is often credited with
being the progenitor of the natural history essay and who wrote broadly on
many animal species; and C. Hart Merriam (1855–1942), a zoologist whose
field work, largely in the American Southwest, lead to the life zones concept.
This principle explains the fact that as one traverses from lower elevations to
higher climes, the changes in vegetation and animal life parallel similar
changes recorded when one travels from lower to higher latitudes. There are
many aspects of animal behavior where this concept continues to stimulate
research.
Within the emerging discipline of animal psychology, in addition to the
contributions of Romanes and Morgan, other important ideas emerged that
became part of the foundations for this approach to behavior. The idea of
organic selection was proposed by James M. Baldwin (1861–1934) and
others. This concept involves the development of variable phenotypes not
based on genetics, followed by some form of hereditary traits with clear adap-
tive advantages. This could lead to selection that would favor developmental
pathways resulting in a phenotypic trait like that which had appeared origi-
nally without a genetic basis. This has become known as the Baldwin effect
and has again met with interest from scientists in the past several decades.
Linus W. Kline (1866–?) was an early proponent, in several papers, of psy-
chology as a source of new methods as well as the use of a wide variety of
organisms for studying comparative behavior. He is also known for one of
the earliest attempts to provide a course in laboratory instruction for students
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 51

in this field. It was during this period that the use of the white rat (Rattus nor-
vegicus) as a subject first occurred and the first mazes and similar test apparatus
were introduced. Comparative psychology remained a two-pronged approach;
some studies examined animal behavior for the sake of understanding the
activities of the subjects, and others were designed to explore behavior using
animals as models for humans.
On an organizational level, T. Wesley Mills (1845–1915) originated the
Association for the Study of Comparative Psychology, dedicated to providing
for wide-ranging inquiry on topics pertaining to the animal kingdom. William
James (1842–1910) published The Principles of Psychology in 1890, a textbook
that was in standard use for several decades. The American Journal of Psychology
was founded by G. Stanley Hall (1844–1924) in 1887. All three of these indi-
viduals were adherents of the new theory of evolution by natural selection and
combined that set of principles with work on physiology, education, and child
development, respectively.
By the end of the nineteenth century, the study of animal behavior was well
grounded and could, in some manner, be ascribed to three different fields of
study. In England and Europe, what later became known as ethology was
the primary focus. In the United States, most attention in zoology was focused
on studies of natural history and physiology. In both locations, though pri-
marily in the United States, work progressed on aspects of animal psychology.
In the next three sections, we will explore in more detail how each of these
three approaches resulted in a considerable body of information about animal
behavior as well as a number of theoretical ideas concerning animal actions. It
should be noted that in many writings about the origins of modern animal
behavior, only ethology in Europe and animal psychology in the United States
are considered as foundational. I claim here that the tradition of natural his-
tory and some accompanying explorations of pertinent anatomy and physiol-
ogy are a third major pillar supporting the later appearance of modern
animal behavior.
Science does not recognize boundaries such as decades or centuries. Thus,
with particular reference to the next three sections on approaches to the study
of behavior, there will, of necessity, be some coverage that extends backward
to the nineteenth century, just as some of those whose work has just been cited
worked into the twentieth century. Of course, it also is true that one cannot
fully assign individual scholars exclusively to one of the three approaches;
many worked at intersections or were trained in one tradition and eventually
made significant contributions in one or more of the other traditions. Finally,
the history of science is a mixture of tracking the origin and evolution of ideas
but includes a considerable emphasis on the people who conducted the
52 ANIMAL BEHAVIOR

research and wrote the papers and books providing a published record of their
findings.

COMPARATIVE PSYCHOLOGY 1900–1960

The first half of the twentieth century was characterized by several themes:
(1) learning, motivation, and brain function; (2) development of behavior
through continuing struggles with the notions of instinct versus environment,
with the latter represented by behaviorism; and (3) new methods and forms of
data analysis. Throughout the period psychologists embraced evolution by
natural selection and applied the tenets of this theory to their work. During
the later decades, new topics, such as behavior genetics, social behavior, repro-
duction (including endocrinology), and more emphasis on physiology, based
on new discoveries in that field, emerged and became foci for new areas of
investigation.
The first prominent comparative psychologist to study learning in the
twentieth century was Edward L. Thorndike (1874–1949), whose earliest
work dealt with instincts and was based on pecking behavior in chickens. As
his career evolved it included studies of problem solving in various animals,
including cats, fishes, rats, and monkeys. He was a tireless advocate for more
careful experimentation and rigorous testing, replacing the tendency to use
anecdotal evidence. His summary conclusion focused on the idea that there
were considerable similarities in the learning processes across a wide variety
of animal species.
Among those who studied learning, James B. Watson (1878–1958) may be
the most familiar. A portion of his work revolved around behaviorism, the
notion that all animal actions can explained purely by objective observations
with no recourse to or need for involving introspection (popular among those
who studied mental evolution in animals during the prior century) or other
internal mechanisms. Humans were thought to be the same as all other ani-
mals in this regard. The quote that is often used to summarize Watson’s idea
is as follows:
Give me a dozen healthy infants, well formed and my own specified world to
bring them up in and I’ll guarantee to take any one at random and train him to
become any type of specialist I might select—doctor, lawyer, artist, merchant-
chief and, yes, even beggar-man and thief, regardless of his talents, abilities, voca-
tions, and race of his ancestors. (Watson, 1930, p. 82)

We should note that Watson was an earlier proponent of and engaged in con-
siderable field research, including trips to distant locations. B. F. Skinner
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 53

(1904–1990) developed the operant conditioning method for studying learn-

ing, and though he also is generally seen as favoring a strong environmental
component influencing behavior and its development, he combined those
views with aspects of evolutionary influence on behavior. Overall, psychology
took a strong position that environmental effects were significantly more
important for behavior than instinct or genetic control during development.
Robert M. Yerkes (1876–1956) began his work with a series of explorations
of comparative learning in species such as crabs, turtles, and mice. His career,
however, is best exemplified by a broad, highly influential interest in compara-
tive studies involving physiological processes, anatomy, genetics, and embryol-
ogy. He was an early champion of the study of abnormal psychology using
animal models. He was a pioneer in establishing a center for work on pri-
mates; this eventually helped to spawn a series of primate centers throughout
the United States and no doubt stimulated similar endeavors in other
countries.
Margaret Floy Washburn (1871–1939) first published The Animal Mind—
A Textbook of Comparative Psychology in 1908 as part of a series by Macmillan
Press on animal behavior topics; the book went through several editions over a
more than 20-year period. The book is noteworthy because it provided cover-
age of a wide range of then-current psychological perspectives rather than
adopting a particular theme, accepting one set of concepts and completing dis-
carding others. Washburn’s own considerable experimental research covered
topics pertaining to emotions, aesthetics, and spatial perception.
Primates as subjects were a popular trend during this 60-year interval,
including field work by individuals such as C. Ray Carpenter (1905–1975)
and laboratory investigations by Harry Harlow (1905–1981) and others.
The major focus of Carpenter’s work on Central American howler monkeys
was the basis for understanding primate behavior for more than three decades.
He played a critical role in the establishment of the Cayo Santiago rhesus
macaque (Macaca mulatta) colony off the eastern end of Puerto Rico. Harlow
is best known for his work on infant rhesus macaques deprived of normal
maternal contact and interactions. Varying degrees of social isolation, includ-
ing the use of surrogate mothers and total isolation in metal binds, produced
evidence for the importance of care giving and of social companionship with
peers during the course of early development.
Significant new developments in the 1940s were highlighted by efforts to
study the genetic bases underlying behavior by John Paul Scott (1909–2000)
using dogs and examination of their personality traits. Scott is also credited
with first proposing the term sociobiology to describe the study of the social
systems of animals in relation to their ecology and biology. New explorations
54 ANIMAL BEHAVIOR

of reproductive behavior by Frank A. Beach Jr. (1911–1988) resulted in a

series of students and papers that continued into the 1970s. Going beyond
the initial work on careful descriptions of male and female mammals engaging
in copulation, Beach and others moved into studies of the underlying physio-
logical bases for these behavior patterns; his edited 1948 book, Hormones and
Behavior, provided a summary of what was then known about endocrines and
animal activities, not just limited to reproductive behavior but including
aggression, migration, and other topics.
The neural bases for behavior, particularly in terms of brain function, were
explored by Karl Lashley (1890–1958) and Donald O. Hebb (1904–1985).
Lashley used preparations involving damage to specific areas of rat brains to
study both memory and motor patterns. His summary conclusion was that
memories were not localized in specific brain areas but rather existed in a dis-
tributed form throughout the brain. This has been expressed as two principles:
(1) mass action is the idea that the cerebral cortex acts as a single entity for
many types of learning, and (2) equipotentiality embodied the idea that when
there is brain damage in certain areas, other parts of the brain can assume some
of the roles normally carried out by that damaged area. Hebb’s book, The
Organization of Behavior, published in 1949, was comparable to and parallel
in time with Beach’s book on hormones and behavior, providing an excellent
and thorough summary of what was known about the nervous system at the
start of the second half of the century. Hebb also made significant research
contributions in terms of neural plasticity, memory storage, understanding
brain damage, and overall brain functions.

EUROPEAN ETHOLOGY 1900–1960

An approach that was based in evolutionary biology and examined behavior
in its natural context blossomed in several European countries during the first
half of the twentieth century. In 1973, the Nobel Prize for Physiology or
Medicine was awarded to three pioneers in animal behavior, Niko Tinbergen
(1907–1988), Konrad Lorenz (1903–1989), and Karl von Frisch (1886–
1982). Of these, Tinbergen and Lorenz are considered to be the founding
fathers of modern ethology. Lorenz’s mentor, Oskar Heinroth (1871–1945),
working at about the same time as C. O. Whitman in North America, pub-
lished a massive treatise on the birds of Central Europe and contributed to
our understanding of species-specific behavior patterns observed in rituals
and other forms of communication. He was among the first to realize that
behavior patterns could be used for establishing systematic relationships
among closely related species, as for example with duck courtship behavior.
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 55

Ethology, based on natural history and, at least initially, on observations,

focused on questions pertaining to the functional purposes of the behavior in
its natural context and on the manner in which the behavior pattern(s)
evolved. Tinbergen was a gifted experimentalist, using observations to gener-
ate ideas for testing specific hypotheses. Perhaps two of his best-known studies
concerning the use of spatial cues by digger wasps to locate holes in the ground
where they were provisioning eggs and the practice of egg-shell removal from
the nest soon after hatching in gulls. A book, The Study of Instinct (1951),
and a paper on the “Aims and Methods of Ethology” (1963), both by Tinber-
gen, provide excellent summaries of the findings in ethology up to those
points in time. Lorenz is known by his successors for the fact that he never
actually collected data points or did statistical analyses but nevertheless made
seminal contributions during the emergence of ethology. Among his efforts
were expanded studies on imprinting, models for motivation, and work on
the concept of instinct and the control of behavior. Lorenz promoted the use
of the comparative method as a tool for understanding behavior.
Other scientists also contributed to the body of knowledge produced by
European ethologists. Jacob von Uexkill, whose work included studies on sen-
sory systems, muscles, and the nervous system, put forward a principle that is
pertinent today and is taught to most beginning students of animal behavior.
He posited that in order to understand fully the actions of an organism, it is
necessary, insofar as possible, to try to examine the world of that animal
through its sensory systems. Of course, this is not possible in a complete sense,
but many who study behavior find themselves “living” in the environments of
their study animals to obtain ideas for testing what they are observing. Karl
von Frisch is best known for his work on the dance language of the honey
bee, where his many meticulous experiments provided detailed information
on how bees communicate their knowledge about food sources to hive mates.
He also made important contributions to our understanding of color vision,
particularly among fish and some invertebrates.
At midcentury, ethology was poised for rapid expansion. Among those who
fostered this growth was W. H. Thorpe (1902–1986), whose work spanned
the questions surrounding the study of both learning and instinct in animals.
Thorpe conducted research on insects and birds, but his greatest efforts were
to summarize and synthesize the earlier and ongoing work in ethology, which
he did in several books and numerous articles. Erich von Holst was a pioneer
in what today we call neuroethology, studying possible brain and sensory sys-
tem mechanisms underlying ethological concepts like species-specific beha-
vior patterns as well as stimulus-response systems. The Dutchman Gerard
Baerends (1916–1999) was responsible for initiating studies on egg
56 ANIMAL BEHAVIOR

recognition as well as factors controlling incubation; through these studies, he

developed the concept of an innate releasing mechanism. His laboratory also
became involved in studies of the nervous system and behavior, including
brain lesions and measurements of neuronal action potentials, as well as
aspects of metabolism and the effects of insulin.
The foundation principles of ethology were developed in England and
continental Europe. The notion of instinct and its underlying mechanism(s)
was a principal focus. So, too, the modes of learning and modification of
behavior formed a second major focus. The evolution of behavioral traits
and how they served particular functions was a foundation for both the design
and interpretation of research investigations. A vocabulary based on concepts
derived from observations of animals emerged. This included ideas like sign
stimulus or releaser, innate releasing mechanism, specific action potential,
appetitive and consummatory behavior, and displacement activity, among
others. An ethogram is a complete description of the behavioral repertoire of
an organism, both individually and in social situations. Beginning students
in laboratories of ethology were often assigned the task of completing an etho-
gram for a particular species, as for example using a tank of fish or watching
birds in their natural habitat.

AMERICAN ZOOLOGY AND PHYSIOLOGY 1900–1960

As noted in an earlier section of this chapter, the exploration of North and
South America led to a tradition of naturalists who traveled widely and pro-
vided written records of their observations of the plant and animal life of these
vast continents. From these beginnings, an approach to animal behavior, not
often fully appreciated but certainly an integral part of modern animal behav-
ior, developed during the early decades of the twentieth century in the United
States and Canada. The term ethology is often applied to early animal behavior
research in North America as well as in Europe, but I see some distinctions
that warrant separate consideration.
There are perhaps four threads that, when woven together, provide a pic-
ture of the development of animal behavior in North America as distinct from
and parallel to events during the same period in England and Europe and also
developments in comparative psychology. These include (1) guide books and
zoology textbooks, (2) individuals and groups doing animal behavior research,
(3) animal ecology, and (4) intersections with British scientists and those
studying comparative animal psychology in North America.
Building upon the naturalist tradition and the earlier work by individuals
like Whitman, Wheeler, and others, the study of behavior evolved in North
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 57

America, encompassing a combination of field and laboratory work. Zoology

textbooks of this period contained considerable content concerning the activ-
ities of animals, both in relation to their ecological environment and with
respect to internal mechanisms. Among these are Animal Studies (1909) by
David Starr Jordan, Vernon Lyman Kellogg, and Harold Heath; Elements of
Zoology (1921) by Charles Benedict Davenport; and Practical Zoology (1915)
by Robert William Hegner. By the 1930s virtually all textbooks in zoology
and biology included one or more chapters on animal behavior, often in con-
junction with coverage of ecology. Many young people were attracted to the
study of behavior by the early guides to birds and mammals. Examples include
Field Guide of Wild Birds and Their Music (1904) by Ferdinand S. Mathews
and A Field Guide to the Birds (1934 and later) by Roger Troy Peterson for
birds and American Mammals (1905) by Witmer Stone and William E. Cram,
Field Book of North American Mammals (1928) by Harold E. Anthony, and
Mammals of the Eastern United States (1943 and later) by William J. Hamilton
for mammals. These books and others, some of which are still in print in
revised editions, served as successors to the earlier compilations of natural
history.
One pathway to the study of behavior in North America emerged from
ecology. To understand ecological processes, a critical element is the cadre of
animals, from diverse groups, that are components of any functioning system.
This is true for studies of almost all major ecological process, including succes-
sion, energy relations, populations and community structure, biogeochemical
cycles, and animal interactions. Those who began the scientific study of ecol-
ogy in the nineteenth century were drawn to investigations of the roles of ani-
mals in these processes. This was one factor in the inclusion of animal work
in the early zoology textbooks. Books published on Animal Communities in
Temperate America (1931) by Victor E. Shelford and Bio-Ecology (1939) by
Frederic E. Clements and Shelford cover many aspects of animal activities,
including reproduction, migration, social systems, habitat selection, feeding,
and many more. This line of work culminated in the 1949 book Principles of
Animal Ecology by Warder Clyde Allee, Alfred E. Emerson, Orlando Park,
Thomas Park, and Karl P. Schmidt, a classic and the forerunner of almost all
aspects of animal ecology even today. It should be noted that what became
the Animal Behavior Society in 1964 emerged from sections within several
other organizations, including the Section of Animal Behavior and Sociobiol-
ogy of the Ecological Society of America. Another was the Division of Animal
Behavior within the American Society of Zoologists.
Several key figures provided foundations for the study of behavior in North
America. An early laboratory scientist was Jacques Loeb (1859–1924), whose
58 ANIMAL BEHAVIOR

book Forced Movements, Tropisms, and Animal Conduct (1918) summarized

lengthy experimental tests of the manner in which animals responded to cues
from the environment such as light, temperature, or chemicals. The book has
an exhaustive bibliography of 554 citations, likely a complete compilation of
all known research in the area up to that date. As part of his work, Loeb dis-
cusses instinct and memory and their possible connections to his general thesis
about animal activities being a function of tropisms.
Warder Clyde Allee (1885–1955) worked primarily at the University of
Chicago, where he produced numerous articles on behavior and ecology and
important books such as Animal Aggregations, The Social Life of Animals, and
Cooperation among Animals, with Human Implications. He was a student of
Victor Shelford (1877–1968), a founder of animal ecology. Allee may be best
remembered for his work on cooperation and what he called protocooperation
in relation to competition in the evolution of different forms of social organi-
zation in both invertebrates and vertebrate organisms. A portion of Allee’s
focus was on human social systems and the relationships to the processes he
studied in nonhuman animals. During the course of his career, Allee also men-
tored several individuals who would go on to productive careers furthering
the development of animal behavior as a separate discipline; these include
Nicholas Collias, Edwin M. Banks, and Edgar B. Hale.
Gladwyn K. Noble (1894–1940) investigated behavioral and physiological
systems in reptiles, though he also did work on several other vertebrate groups.
He established the Department of Experimental Biology at the American
Museum of Natural History in New York, the precursor of a significant group
of animal behaviorists that worked there from the 1960s to the 1980s. As his
career progressed, he became interested in the roles of hormones affecting
behavior patterns involved with courtship and reproduction.
In addition to the emphasis on physiology that characterized the work of
Allee, others working on what we would call today physiological ecology con-
tributed to aspects of animal behavior. Those who study animal adaptations
were often concerned with responses to climate, including temperature,
humidity, and weather events. Others looked at specific types of adaptations
for different diets or the effects of seasonal changes, including hibernation by
invertebrates and vertebrates. In these instances and many others, examining
the adaptations of the animals led to discoveries about behavior. Physiology
is the intersection of form and function, and behavior is, in some large mea-
sure, a product of physiological processes, both sensory and motor. Though
certainly not exclusive to North American animal behavior, the advances in
physiological ecology and related behavioral phenomena were a hallmark of
this half century and form part of the fabric of a third approach to behavior.
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 59

The work of these American animal behaviorists intersected both with

work ongoing in England and with comparative psychology. Previously, we
examined the early work by some prominent comparative psychologists; their
work often involved a variety of species, problems based on animals observed
in natural settings, and issues pertaining to physiology and development. This
broad scope had shifted by the 1940s. A key paper by Frank Beach, entitled
“The Snark Was a Boojum,” argued that the focus of comparative psychology
had become too narrow, concentrating on learning and generally using only
the white rat as a test species. Though perhaps somewhat overstated, his main
point was true. His paper was significant in opening the door to more truly
comparative research in animal psychology. The lines between comparative
psychology and American zoology with respect to studies of animal behavior
were now becoming blurred. Several individuals, including Daniel Lehrman
(1919–1972) and J. P. Scott, conducted studies that combined elements of
both approaches.
Lehrman and his students produced an elegant series of experiments on the
reproduction of ring doves (Streptopelia risoria). Their findings illustrated the
stepwise processes and interactions of behavior and internal hormonal states
involved at each stage of the reproductive cycle beginning with courtship
and extending to the feeding of crop milk to squabs. Lehrman also wrote a
paper in 1952 that was critical of the approach being taken by European ethol-
ogists, stating they were too wedded to the concept of instinct, giving insuffi-
cient attention to other factors affecting behavior. Though initially viewed
negatively by many ethologists, this and other papers began the process of
integrating the fields of comparative psychology, ethology, and American zoo-
logical studies of behavior.
J. P. Scott, trained as a zoologist, spent his career doing research that
involved topics ranging from behavior genetics to social relationships and
social organization. His studies of the development of behavior included major
advances using the critical-periods concept. After founding the behavior pro-
gram at the Jackson Laboratories in Maine, his academic appointment was
in psychology, something that was happening more frequently with animal
behaviorists during the 1960s and 1970s, including individuals trained in psy-
chology with appointments in zoology. Scott was a principal participant in the
formation of the Animal Behavior Society and wrote an early textbook of ani-
mal behavior. Like many of his contemporaries, one focus of Scott’s endeavors
was to better understand the behavioral biology of our own species; he co-
authored several books pertaining to this topic.
European connections influencing American zoologists (and comparative
psychology) during this period include Robert A. Hinde (1923–), Irenaus
60 ANIMAL BEHAVIOR

Eibl-Eibesfeldt (1928–), and Peter Marler (1928–). Hinde was a primary fix-
ture of the animal behavior research group at Cambridge University for more
than half a century, mentoring numerous students, writing one of the first
textbooks on animal behavior, and adding considerable knowledge to the field.
His major foci included social and reproductive behavior of birds, animal
communication, and a body of work that paralleled the studies by Harlow
on mother-infant separation in rhesus macaques. He, like others of the time,
related many of his findings to human biology and behavior.
Eibl-Eibesfeldt studied the development of behavior, primarily in mammals
and involving studies in the field in Africa. Later he worked on comparative
aspects of animal communication in vertebrates, including humans, where
his work intersected with linguistics. He was a cofounder of the International
Society for Human Ethology. His 1970 textbook, Ethology, The Biology of
Behavior, is noteworthy for its integration of all approaches to the study of
animal activities.
Marler, who was Hinde’s first doctoral student, started his career in the
United Kingdom but spent many years at the Rockefeller University and the
University of California at Berkeley and Davis. He is thus a prime example
of the sort of cross-fertilization that occurred during this period. Marler co-
authored, with William J. Hamilton, a widely used early textbook of animal
behavior (1964). His research contributions center on bird vocalizations and
include discovering that white-crowned sparrows (Zonotrichia leucophrys) had
regional song dialects, a phenomenon now known from a variety of verte-
brates. He worked, in fine detail, on factors affecting the development of bird
vocalizations, working out the different strategies used for song learning and
defining stages in the process, such as plastic song and crystallized song.
Many other scientists and their contributions could be provided here as
examples of connections between Europe and America involving approaches
to animal behavior. The result of these connections, a period of integration
of ethology, comparative psychology, and zoology, is the subject of the next
section.

1950–1975—RAPPROCHEMENT AND INTEGRATION OF APPROACHES

The quarter century between the early 1950s and mid-1970s witnessed an
unprecedented growth for all aspects of animal behavior. This is evidenced by
the number of books produced, both as scholarly monographs and early text-
books on the subject; by the new classes offered in behavior in both biology
and psychology departments at many colleges and universities; and by support
for research on many topics in animal behavior through external grants,
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 61

Table 2.1. Decadal summary of new and total numbers of journals with a primary
focus on animal behavior.
Year interval Number of new journals Total number of journals

Before 1960 7
1960–1969 7 14
1970–1979 16 30
1980–1989 12 42
1990–1999 2 44
2000–2009 1 45

primarily from government agencies such as the National Institutes of Health

(NIH—initially through the National Institute of Mental Health) and the
National Science Foundation (NSF). In addition, a number of new journals
appeared, beginning with Behaviour in 1948 and continuing with the British
Journal of Animal Behaviour (soon changed to Animal Behaviour) in 1952.
Many more publications followed in the ensuing two decades (Table 2.1).
Where there were previously subgroups within larger professional, scholarly
associations devoted to animal behavior research, some became separate, dis-
tinct organizations. In England, the Association for the Study of Animal
Behaviour originated in 1936, and in North America, the Animal Behavior
Society was founded in 1964. Division 6 of the American Psychological Soci-
ety, devoted to the comparative study of behavior and later to physiological
aspects of behavior and neuroscience, was formed in 1944. Of great impor-
tance was the initiation of International Ethological Conferences, starting in
1949 at Cambridge, England, as a gathering of primarily English and Euro-
pean scholars meeting to have lengthy discussions about pertinent research
topics. In the ensuing years, several scientists from North America joined these
gatherings. Over the next 50 years, this organization held biennial meetings
that eventually encompassed all nations where work on animal behavior
occurs. The meetings in the 1960s, held in Europe, and in the1970s, held
alternately in Europe or at various non-European locations, were critical to
bridging some of the differences in approaches for those in ethology, compara-
tive psychology, and American zoology.
What about scholarly endeavors during this critical quarter century? As the
field expanded, many more individuals conducted work in all aspects of ani-
mal behavior. This was a period between the founders of the first half of the
century and the split between subareas of animal behavior that would occur
62 ANIMAL BEHAVIOR

in the late 1970s and onward. Major areas of research included topics like
behavior genetics, behavior development, social behavior (including ecological
and evolutionary aspects of social organization), social and ecological aspects
of learning, neuroethology, neuroendocrinology, and the evolution and func-
tion of play behavior. Generally, these and other research foci fit within the
four-question scheme proposed by Tinbergen: (1) causation—the underlying
physiological mechanisms; (2) development, wherein the inherited blueprint
unfolds in the context of environment and experience; (3) evolution of behav-
ior; and (4) the function or ecological context of behavior.
Looking at the work of several individuals should provide a flavor of the
types of research being conducted during this quarter century, a mix of some
things from the past, some new areas, and many new investigators. Several of
the key figures had experience in both North America and Europe. George
Barlow (1929–2007) trained first at the University of California—Los Angeles
and then worked with Konrad Lorenz, and later with Niko Tinbergen in
Europe and England. His approach to cichlid fishes and their behavior drew
from both approaches to behavior. The complete, comparative picture of
cichlid fish behavioral and social systems is a model for thoroughness and
detail.
Richard Dawkins (1941–), now known more for his championing aspects
of evolutionary biology, was a Tinbergen student at Oxford and then spent
time in a faculty position at the University of California at Berkeley. His
own research combined studies on animal communication and evolutionarily
stable strategies, with synthetic contributions including genes and behavior,
evolution, and science and religion. Donald Griffin (1915–2003), an Ameri-
can and the father of cognitive ethology, began work on echolocation in bats
early in his career and then devoted a major portion of his efforts to the idea
that animals can think, form mental concepts, and use thought process during
their daily activities and interactions with their surroundings. This latter
worked spawned an entire area of research within animal behavior that contin-
ues to this day. Stephen T. Emlen (1940–) spent his academic career at
Cornell, where his studies span the globe, both in terms of the locations where
he conducted investigations (Africa, Central and South America, United
States) and the topics that he explored (bullfrog [Rana catesbeiana] territories,
navigation in nocturnally migrating songbirds, mating systems and infanticide
in jaçanas, and breeding biology and parental behavior in bee eaters).
Patrick Bateson (1938–), an Englishman, did his doctoral work with
Tinbergen and, like several others just mentioned, spent time early in his
career in the United States. His studies involved early behavior development
in cats and early experience effects on sexual preferences in fowl, and he was
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 63

a key person with regard to techniques used for measuring behavior and analy-
sis of findings from observations of animals. He coedited a series of volumes
on Perspectives in Ethology and wrote numerous articles popularizing aspects
of behavior. A fellow Englishman, Geoffrey Parker (1944–), is a leading pro-
ponent of using game theory concepts to explain behavior and generate
hypotheses for investigations, and he has worked on aspects of sexual selection
and other tenets of Darwinian natural selection. He is, in effect, a bridge to the
idea of behavioral ecology that came to dominate a segment of animal behav-
ior research during the late 1970s.
Textbooks published during this period, including those by Marler and
Hamilton, Hinde, Eibl-Eibesfeldt, and Jerram Brown, all evidence a broader
approach to the study of behavior. The material covered involves all topics,
ranging from physiology and development to ecology and evolution. The liter-
ature cited and examples used reinforce the large scope of the field of ethology
or animal behavior during its period of maturation.
An important feature of the quarter century we are discussing was the
beginning of greater involvement of women studying animal behavior. To
be sure, there were females who worked in this field during the 1950s and
1960s, such as Evelyn Shaw, Ethel Tobach, and Lee Ehrman, but the field
was largely dominated by males. The 1986 volume Studying Animal Behavior:
Autobiographies of the Founders by Donald A. Dewsbury, with autobiographies
of the founders of ethology, contained essays only from men. Among those
whose contributions were recognized in the second volume of autobiographies
(Leaders in Animal Behavior, The Second Generation [2010] by Lee C. Drick-
amer and Dewsbury), which recognizes individuals whose efforts began by
the 1970s, were Jeanne Altmann (1940–), Sarah Hrdy (1946–), Mary Jane
West-Eberhard (1941–), Patricia A. Gowaty (1945–), Marian Stamp Dawkins
(1945–) and Meredith West (1947–). Altmann’s (1974) paper on sampling
methods for the study of animal behavior is the most cited paper in our field.
She continues long-term contributions on the social biology and behavior of
baboons, studied in the wild in Kenya. Hrdy investigated the significance of
paternal infanticide, based on her work on langurs, and is a leading figure in
writing about the importance of understanding female roles in social systems.
West-Eberhard is a pioneer in terms of studying and writing about phenotypic
plasticity and epigenetic effects on behavior. Gowaty has investigated blue-
birds, fruit flies, and other organisms, with several foci. One of these is the
connection between evolutionary biology and feminism and the importance
of variation in life history traits of females and males and their interrelation-
ships. Marian Dawkins is a principal figure in one of the newer subfields of
animal behavior, the examination of animal welfare, particularly in regard to
64 ANIMAL BEHAVIOR

agriculture, with particular attention to poultry. West’s contributions center

on several decades of work on cowbirds and, in particular, their songs. The
research is a story of combination—field and lab, development and evolution,
male and female—and gene-environment interactions. Approximately 50 per-
cent of the principal officers of the Animal Behavior Society and also the Asso-
ciation for the Study of Animal Behaviour have been women over the past
30 years, and there is a clear tendency to find more than 50 percent of the doc-
toral degrees in North America for work in animal behavior awarded to
women; the longstanding male bias is on the wane. However, some forms of
perception bias remain. For example, one need only look at the composition
of editorial boards for major journals like Science and Nature to find that just
a small fraction of the members are women; there is progress, but much more
concerted effort is needed to properly encourage the role(s) women play in sci-
ence, and in animal behavior in particular.
With the publications of works by individuals like George C. Williams
(1926–2010), Robert Trivers (1943–), William D. Hamilton (1936–2000),
and Edward O. Wilson (1929–), combined with other work on endocrinology
and neuroscience in relation to behavior by people such as Kenneth Roeder
(1908–1979), William C. Young (1899–1965), and Donald W. Pfaff
(1939–), animal behavior began to split into at least two separate subfields:
(1) behavioral ecology and sociobiology and (2) neuroscience and neuroendo-
crinology. Evidence for this split can be seen, for example, in the content of Ani-
mal Behaviour (Table 2.2), which shows a clear shift from years with perhaps a
slight bias in the number of papers dealing with mechanisms relative to those
covering topics in behavioral ecology for the 1950s and 1960s. The focus shifts
dramatically in favor of papers with behavioral ecology content beginning in the
1970s. At this same time, a host of new journals appeared during the period
from the late 1960s until the 1980s. In addition, a number of new academic
societies formed during this interval reflect the split, with organizations such
as Society for Neuroscience (1971), International Society for Neuroethology
(1986), and a group now known as Society for Behavioral Neuroendocrinology
(1969) encompassing topics involving mechanisms and other groups such as the
International Society for Behavioral Ecology (1986) and Journal of Insect
Behavior (1988) with strong emphasis on ecological aspects of animal behavior.
For a period of nearly three decades, lasting into the early twenty-first century,
these approaches remained distinct and grew further apart. Many, though not
all, of those working in comparative and physiological psychology shifted to
an emphasis on neurobiology and endocrinology. We should thus examine each
of these and then conclude with some thoughts on how the two different
approaches can be brought back together in the next several decades.
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 65

Table 2.2. Content of Animal Behaviour by type of article using one randomly
chosen issue for each of the last five years of a decade.
Number
of Behavioral
Interval articles Number/year Mechanism ecology Other

1956–1960 129 35 57 57 15
1966–1970 135 85 78 40 17
1976–1980 140 110 29 92 19
1986–1990 124 190 20 100 4
1996–2000 103 240 17 79 7
2006–2010 157 320 14 136 7

BEHAVIORAL ECOLOGY AND SOCIOBIOLOGY

We can define this system for the study of behavior as an approach to inves-
tigations of how different ecological circumstances affect animal behavior and the
evolution of traits with functions that fit these different environments. The use
of the comparative method is a hallmark of behavioral ecology. Sociobiology
is effectively a category within behavioral ecology referring to issues pertaining
to social systems and pertinent ecological constraints.
In addition to the books already noted, an important textbook by Jerram
Brown (1930–), The Evolution of Behavior, published in 1975, contributed
to the education of a new generation of students, now trained to think pri-
marily in terms of behavioral ecology. Of the 19 individuals whose autobiog-
raphies appear in Leaders in Animal Behavior: The Second Generation, which
covers those whose influence spanned the late 1960s through the 1990s and
up to the present time, 12 or 13 are behavioral ecologists. The growing domi-
nance of this approach as the focus of mainline animal behavior is well illus-
trated by Table 2.2 summarizing articles in Animal Behaviour. Two series of
books, which each went through multiple editions, both written or edited by
John Krebs (1945–) and Nicholas Davies (1952–), are good summaries of
behavioral ecology and of the changes that happened over a 15- to 20-year
period in terms of the concepts and approaches used by scientists in this area.
One set involves edited volumes on Behavioral Ecology: An Evolutionary
Approach, with each edition containing chapters by the leaders in the field.
The other set, the textbook An Introduction to Behavioral Ecology, provided
the basic framework and also, through revisions, kept pace with the develop-
ment of the field. All of these volumes are excellent resources for anyone
66 ANIMAL BEHAVIOR

wishing to gain a fuller understanding of the basic tenets of behavioral ecology.

Scores of other monographs appeared during the 35-year period from 1975
through 2010, and more are being written on all of the topics noted in the
following paragraphs.
What are some of the major themes of behavioral ecology and sociobiology? Each
of these themes is interwoven with the others and, in some instances there is
considerable overlap. One of the earliest emphases was on foraging behavior.
What strategies do animals use to locate food, and how do they decide when a par-
ticular resource has diminished to a point where it is necessary to find additional
locations for feeding? Some animals seek food individually, while others work
in various cooperative schemes. What ecological factors contribute to the evolu-
tion of these different patterns? Over several decades, the interest in foraging
behavior has waned, though the topic still receives some attention.
A second major area of interest involved social systems. Wilson’s book was
the major impetus for questions surrounding topics like What different forms of
social systems exist in nature, ranging from individuals who live solitary lives,
except to meet for reproduction, to social insects with their complex caste systems?
Using the comparative method it is possible to explore the different ecological
factors that influence the evolution of these widely varying social systems. In
all cases, basic needs such as food, shelter, protection from predators, and
reproduction must be met. The roles of aggression and assessment of an oppo-
nent’s capabilities can fit with this theme.
The third area, and by far the most important as measured by attention
given to the various subtopics that follow, is reproduction. Of course, sexual
selection, in its many manifestations, is the most studied topic in all of
behavioral ecology. Beginning with Darwin, there are myriad issues pertain-
ing to topics like assessing mate quality, mate choice and its consequences,
and testing ideas related to the work of Ronald A. Fisher (1890–1962). What
forms of parental investment and parental care are there, and how has evolution
shaped these traits in different types of animals? When and why does cooperative
breeding evolve? What is parent-offspring conflict, and how is this resolved?
The entire study of the evolution of life history traits (e.g., progeny sex ratio,
maturation, and litter or clutch size) is part of the behavioral ecology of
reproduction.
Genetics is at the heart of evolution. Thus, as we would expect, a fourth
collection of research topics centers on ideas like kin selection, altruism,
and group selection. Kin selection is the hypothesis that because close relatives
share significant proportions of their genes they can, in effect, pass along traits
to subsequent generations even without reproducing themselves. Providing an
alarm call, potentially resulting in capture by a predator, could save relatives
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 67

who carry your same genes. These, and possibly other types of altruistic acts,
sometimes referred to as reciprocal altruism, provide many testable hypothe-
ses in relation to social systems, cooperative breeding, and similar topics.
Group selection is built upon the idea that traits that benefit the group,
though not always individuals, can be selected for in the course of evolution.
This notion was elucidated in the 1960s, received considerable criticism in
ensuing decades, and has, in recent years, been reconsidered, as for example
in terms of gene-culture evolution in humans.
A fifth area, which grew considerably by the 1990s, involves animal com-
munication. Animal signals often are complex, multimodal, and convey infor-
mation concerning many different aspects of the social and environmental
situation. Communication, the use of signals to change or modify the behav-
ior of other individuals, is constrained by ecological conditions. Thus, a pri-
mary area of investigation is exploring these constraints and comparing
communications system in animals in similar and different environments.
Other topics include, for example, specificity of animal warning systems, traits
used in mate selection, parent-offspring interactions, honesty of signals, and
the use of different sensory modalities to communicate.
A constant theme, which began with the early writings that led to behav-
ioral ecology and continues today, concerns methods and models. Many new
methods or adaptations of existing methods contribute to behavioral ecology.
These are as diverse as techniques for marking or tracking individual animals,
to remote sensing, to genetics and DNA testing. Today, they also include pro-
cedures for noninvasive measurements of hormones and other physiological
parameters. Much of the work in behavioral ecology was stimulated by mod-
els, sometimes borrowed from other fields such as economics. Optimality
theory is one such example. Game theory, which involves predictions about
individual and group behaviors in competitions for resources, has widespread
applications to problems in animal behavior. We have already mentioned the
comparative method, a standard tool for examining virtually all topics in
behavioral ecology. Evolutionarily stable strategies, which are systems of
behavior adopted by an individual or group that, if invaded by an alternative,
cannot be “defeated,” are another model system that was an impetus for work
on topics ranging from foraging to reproduction.
Today, some of those studying behavioral ecology have adapted to the
changing landscape, employing ever more genetic techniques, including some
scientists using genetic manipulations. Also, investigators are working on ener-
getics and related physiological parameters, often using combinations of field
and laboratory experiments. These are the harbingers of the future, and we
will turn to them in the last section of the essay.
68 ANIMAL BEHAVIOR

NEUROSCIENCE AND BEHAVIORAL ENDOCRINOLOGY

At the same time behavioral ecology and sociobiology emerged and grew to
dominate the field of animal behavior, a second series of events resulted in an
equally strong research area. Thus, what I call the Big Split occurred in the
1980s when several major fields exploring mechanisms coalesced in varying
degrees to make a collective presence. These fields are physiological psychol-
ogy, behavioral endocrinology, and neuroscience, and include many who
became disaffected with the dominant trend toward behavioral ecology at
meetings of the Animal Behavior Society, the Association for the Study of Ani-
mal Behaviour, and the International Ethological Conferences. These latter
individuals brought with them a whole-animal perspective and, in some cases,
strong interests in the development of behavior and behavior genetics. The
1990s were the “Decade of the Brain,” which gave added impetus and funding
to the emerging mechanism studies.
The work in endocrinology and behavior traces its origins to individuals
like Frank Beach; the early work is summarized in the volume edited by Carter
(1974). Separate conferences on hormones and behavior began in the late
1960s and grew into the Society for the Study of Reproduction, now known
as Behavioral Neuroendocrinology. This latter name reflects a degree of con-
vergence between those working on hormones and those studying neurosci-
ence. Several early books on neuroethology appeared in the 1970s and
spurred research that served as the foundation for parts of the new field of
investigation. Physiological psychology narrowed the focus to primarily the
brain and nervous system. Basic research in neuroscience and endocrinology,
as well as the intersection of the two fields, both provides information for
those looking at behavior and supplies new technologies like measurements
for cell receptor functions in the endocrine and neural systems and refined
hormone assays. Of key importance are new ways to measure brain activity
during normal behavioral processes. These include functional magnetic reso-
nance imaging (fMRI) and positron emission tomography (PET) scans as
well as others.
Some major themes in this broad area of investigation include (1) sex and
reproduction; (2) social behavior and social systems; (3) parental behavior
and development; (4) processes related to homeostasis, dealing with stress,
and biological rhythms; and (5) learning, memory, and motivation.
Studies on sex and reproduction ask questions like What regions of the brain
and what types of neuron cell receptors are involved in the processes of sexual repro-
duction? How do specific environmental factors affect the expression of these recep-
tors and levels of hormones involved in the various stimulus-and-feedback-loop
systems regulating aspects of sexual behavior? What sorts of developmental processes
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 69

influence the adult organism in terms of its reproductive capacities and success?
These studies rely on newer technologies such as receptor binding assays,
field-based systems for hormone measurement, and chemical assays for hor-
mone precursors. A research example is measuring the levels of luteinizing hor-
mone and estrogen in migratory birds to provide a better understanding of the
start of the annual breeding season after the birds arrive at their summer
locations.
The study of social behavior includes how the animals of a particular spe-
cies array themselves in space, the overall nature of their social system and
reproductive system (e.g., whether they are solitary, monogamous, or live in
groups), aggression, and affiliative interactions among group members. A
comparative study of the architecture of various brain regions might suggest
some basic brain structural differences or similarities among different species
that share similar social systems. Are there different neuroendocrinological pro-
cesses, perhaps modulated by species differences in hormone receptors, associated
with brain cells that can predict whether a species is monogamous or has some other
form of system with multiple mates? Are there individual differences within a spe-
cies in terms of brain regions that are involved in aggressive (or submissive) inter-
actions? A research example would be the elucidation of the pathways that
begin with chemical signals released by animals, such as a dominant individ-
ual, trace through the neurons of the olfactory system to particular sensory
regions of the brain, and continue then to the motor pathways that result in
acts of aggression. It may be possible to manipulate levels of chemical signals
or interrupt brain pathways to further test these findings.
When an animal reproduces there are generally either eggs or live young
produced. In the 1960s Lehrman worked out the details of a sequence of
behavior in reproducing ring doves. The steps included (1) courtship, (2) nest
construction, (3) egg laying, (4) incubation, and (5) feeding crop milk to the
squabs after hatching. Feedback systems involving behavior and endocrine sys-
tems of the male and female ring dove coordinated their efforts to reproduce.
It is now possible to dissect this sequence much further using a combination of
the techniques already mentioned. The result for the ring dove, as well as other
species that have been studied, including reptiles and mammals, is an under-
standing of the underlying neural and hormonal sequences that occur with
an almost lock-step sequence between and within each member of the pair
of ring doves.
Other questions that fit this theme include the following: What are the
neurohormonal correlates of the cessation of parental care when weaning or fledg-
ing occurs? What about processes such as metamorphosis where it occurs in insects
and amphibians in terms of the interactions of the environmental conditions,
70 ANIMAL BEHAVIOR

behavior—both social and individual—and neural and hormonal events? As with

the other themes, the possibilities for developing new hypotheses seem almost
limitless when we can apply both the new technologies and our increasing
understanding of hormonal and neural systems to a host of observed processes
in natural settings.
Many of an animal’s activities are directed at homeostasis, maintaining
internal and external conditions within certain limits that ensure its physio-
logical systems remain operational within physical and chemical limits. A sim-
ple example is maintaining water balance so as to avoid severe dehydration.
Sensory systems within the animal trigger changes in behavior and physiology,
with attempts to locate sources of fluids and enhanced water retention mech-
anisms. There is a clear connection between the physiology and behavior,
and we can now examine, in detail, the neuroendocrinology of these events.
Similar sequences occur with regard to maintaining body temperature or
energy levels. We have long been able to record animal actions related to
thermoregulation by either exposing body surfaces to sunlight to gain heat or
going to cooler locations to reduce heat, but now we can assess the internal
sequences that regular the behavior. Energy management is a complex opera-
tion with both internal changes in terms of mobilizing existing resources and
external changes in terms of the search for food. Here we can, for example,
manipulate energy levels in terms of food restrictions to examine the neural
and hormonal consequences at the level of individual cells.
Learning, memory, and motivation, topics that have stirred interest among
psychologists and ethologists for more than a century and a half, still are
important topics. With advanced technologies for measuring nerve impulses,
neuronal pathways, and the level of neuronal activity in specific brain regions,
we can now explore these topics in much greater detail. The imaging tech-
niques we just mentioned are critical to sorting out the brain regions where
neural activities occur during learning processes of various types. What areas
are most responsible for information storage and retrieval for performing particular
tasks? Studies of neuron growth and the formation of new connections provide
significant information on, for example, how birds learn their songs; in some
cases, there are seasonal changes in the neurons corresponding to periods when
singing is critical for maintaining a territory and securing a mate.
Animal cognition, sometimes called cognitive ethology, arose, in terms of
interest from animal behaviorists, beginning with books by Griffin in the
1980s. This area of investigation is concerned with the mental processes that
characterize nonhuman animals and how we can explain them through exami-
nation of the impact of evolution on relevant brain processes in connection
with underlying neural processes and development. How do birds that store
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 71

seeds locate their caches when winter conditions dictate a need for food? What
about the neural processes related to language skills as studied, for example, in Afri-
can grey parrots (Psittacus erithacus)? Many bird species migrate seasonally:
How do they navigate during these sometimes very long-distance movements? These
and many other topics are the focus of investigators exploring cognitive pro-
cesses in nonhuman animals, with comparisons to similar processes in humans.
One final area that relates internal body mechanisms with observed behavior
arises from immunology. We know that the parasite load of an organism, if
revealed through external features, can be a factor in processes like mate selec-
tion. We also know that genetic traits, such as variations in the major histocom-
patability locus, influence mate choice, possibly through odor cues. The
immune system has close interconnections with both the endocrine and nervous
systems, effectively making a triangle where each of the three points is connected
to the others. As animal behavior progresses to more studies of mechanisms, this
triangle will be an important source of thinking about these interrelationships.
Several topics in current animal behavior research include components
from both studies of mechanisms and behavioral ecology. These involve, for
example, conservation work and applied animal behavior. Knowing the behav-
ior and ecology of an endangered species may give some clues as to the reasons
for its population decline, but establishing appropriate breeding programs for
reintroductions requires a thorough knowledge of the neuroendocrine systems
of each species. Knowledge of a species’s habitat and needs in a natural envi-
ronment is also critical both for preserving limited existing populations and
for reintroductions. For domestic livestock, issues regarding both better pro-
duction and animal welfare occasion the need to examine, in more detail than
just external behavioral observations, the underlying mechanisms correspond-
ing to stress reactions or successful reproduction.

TWENTY-FIRST CENTURY—NEW THREADS

Several recent developments are influencing the study of animal behavior.
These new threads come from different directions and will, perhaps, bring
the studies of ecology and evolution together with the investigations of the
underlying physiological mechanisms, development, and molecular genetics.
What are some of these themes?

Integration of Proximate and Ultimate Causation

The integration of proximate and ultimate causation of behavior is not
new; there have been several books and symposia on this topic over the last
20 years. However, there has been only modest effort to actually do science
72 ANIMAL BEHAVIOR

that brings together the external ecological and evolutionary aspects of behav-
ior with the internal mechanisms. The sorts of investigations that target inte-
gration include, for example, the analysis of the genetics underlying mating
systems in voles and social systems in honey bees, the study of hormones relat-
ing to mating and stress in birds living in the natural environment, and the
examination of longer-term effects on fitness of manipulation of hormone lev-
els in field enclosures. With behavioral scientists from differing backgrounds
working together it is now possible to study behavioral phenomena at several
levels simultaneously (see Volume 3 of these volumes).

Epigenetics
Epigenetics is defined as events that influence trait expression, in addition to
the genetic blueprint, during the development of an organism (see Chapter 8).
That is, there are forms of inheritance that go beyond simple DNA-based
effects that we have long considered as the mainstay of development and evolu-
tion. The mechanisms of these effects include such things as methylation pat-
terns on the DNA, RNA interference with genes, and inheritance of aspects
of the structure from mother cells to daughter cells. The key to these epigenetic
inheritance systems is that they influence the expression of traits and are,
because they are inherited, subject to natural selection. Because behavior is a
product of the structure and physiology of an organism, which in turn are
shaped by both genetic and epigenetic effects during the course of develop-
ment, and indeed into maturity, the study of epigenetics will be critical to
understanding the mechanisms underlying animal actions and activities.
A closely related concept, phenotypic plasticity, sometimes referred to as
ecological developmental biology, involves a combination of epigenetics and
thorough examination and testing of environmental (ecological) effects on
the developing organism. A key feature of this approach is an emphasis on real,
natural world environments; hence the incorporation of “ecology” in the
name given to this emerging area of research. In effect, the age-old dichotomy
and seemingly endless discussions concerning the effects of genetics and expe-
rience on observed behavior is resolved in favor of an approach that combines
an understanding of the breadth of inherited components of this matter with
knowledge of the detailed interactions of the inheritance with dynamic pro-
cesses that occur during development and on into adult life.

Manipulating Genetics and Epigenetics

Closely related to the previous theme is the notion of manipulating both
genetic and epigenetic aspects of inheritance and thus the course of
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 73

development of particular traits. We can now substitute genes to alter behavior

in, for example, voles. Prairie voles (Microtus ochrogaster) have monogamous
mating systems, whereas meadow voles (M. pennsylvanicus) mate in a polygy-
nous system. When specific gene transfers are done from prairie voles to
meadow voles, the latter become more monogamous. This is not a simple
effect—no one-to-one operation exists—but occurs through several steps. In
this case, the gene change results in small changes to brain receptors for par-
ticular hormones, which alter the animal’s sensitivity to the hormone and
result in different cellular outcomes, which in turn influence other hormones
and the nervous system, eventually leading to small behavioral changes. As
genes and other epigenetic markers are found for various multiple-step mech-
anisms underlying behavioral traits, it will be possible to elucidate other effects
like that just described. We will begin to understand how changes in brain
receptors, cells that produce hormones, neural transmitter production and
reception, or other systems all influence behavior.

Strong Inference
Work in behavioral ecology and all areas of animal behavior should return
to strong inference approaches—not affirmational tests. What is strong infer-
ence? This approach has several important features. Rather than engaging in
testing a single hypothesis, which has an inherent confirmational bias, science
should be done with alternative hypotheses as a basis for the experimental
design (Chapter 1). This promotes the use of one or more critical experiments
to differentiate between alternative hypotheses. And, as we noted more than
once in our historical retrospective earlier in this chapter, the techniques and
methods for data collection must provide a clean and unbiased sample of the
phenomena under investigation. Where possible, it is best to measure multiple
dependent variables to assess the full effects of any treatments.
As an example, consider the problem of possible outcomes of mating pref-
erences in mice. If a female mouse is given a choice among several males, we
might hypothesize that if she mates with the preferred male, she will have
higher reproductive success. Two alternatives would be that she might, when
mated with the preferred male, have lower reproductive success, or her success
may not differ from matings where she is paired with a nonpreferred male. To
provide the strongest test among these alternatives we need a third test group
with random matings between males and females—a control. Only by using
all three test groups can we fully differentiate among the effects of mating with
different males. In this case, it is also possible to measure a number of depen-
dent variables. These include success in producing litters, litter size, survival of
pups to weaning, and, to take the process a step further, the effects of the
74 ANIMAL BEHAVIOR

various matings on behavioral traits such as aggression and nest building that
may relate to reproduction in the next generation.

Modeling
There are several types of models that we can use in the study of behavior
(Chapter 1). They serve both to generate new ideas for testing and to summa-
rize information. One type of model involves a schematic diagram, often
involving arrows connecting components of the model. Such diagrams can
be used for depicting dominance hierarchies, movements of groups of ani-
mals including interactions between groups, and, in an ecological sense, for
presenting food webs and the complex feeding interactions that involve
behavior.
Mathematical models attempt to describe processes using equations, which
can include, in our case, behavioral activities and their consequences as well as
contingent probabilities. At another level mathematical models can be used for
predicting outcomes, such as the effects of different mating patterns on repro-
ductive success and possible evolutionary implications. Simulations are
another form of modeling, usually involving a set of components that can
either be assigned specific values or with values that change according to some
formula or a random sequence. Working with animal populations and includ-
ing features like changes in reproduction due to environmental conditions,
social variables, and age, for example, can be modeled in this fashion.
The use of models is certainly not new for those who study animal behav-
ior. However, renewed attention to the types of models just described as well
as others will become a hallmark of the next steps for animal behavior as we
integrate internal mechanisms and external functions. Evaluating models can
be a multistep process, involving examining how well they fit previous obser-
vations as well as using them to design experiments to test how well the model
predicts findings from new tests.
With advances in statistics, we now have available a wider variety of pos-
sible ways to analyze data. Examples of newer statistical approaches developed
and refined in the past 20 years include trajectory analysis, new types of time-
series testing, and various forms of logistic regression. The use of these new
statistical methods and advanced versions of existing software entails a full
understanding of the assumptions that underlie the tests as well as a good
grasp of the ways in which planning for the statistical tests to be used in data
analysis can influence the experimental design.
Many of these new methods involve better ways to look at multivariate sys-
tems; because natural systems are inherently complex, the use of these tools is
a great way to make progress. However, because there is also considerable
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 75

variation in natural systems, it is almost always necessary to have large sample

sizes for multivariate analysis techniques; many scholars run afoul of this prob-
lem, using sample sizes that are too small. Also, a key assumption for the col-
lection of most data points, whatever the dependent variable that we are
measuring, involves the requirement of independence. That is, when we mea-
sure a trait in one animal, we cannot assume that when we measure another
animal in the same group it will behave in a manner that is completely inde-
pendent of the actions of the first animal we measured. Thus, for example, lit-
ter mates, or animals living in the same social group, cannot usually be
considered as independent data points. There are usually statistical techniques
for handling these potential confounds (e.g., general linear mixed models), but
all too often they are ignored, and a key assumption is violated. Younger scien-
tists will be well advised to learn both the new statistical techniques and the
assumptions that must be met or tested to properly use each technique.

ACKNOWLEDGMENTS
I thank Ken Yasukawa for the opportunity to write this chapter. My thanks to
Patricia A. Gowaty, Donald Dewsbury, John Byers, Gordon Burghardt, and Jack
King for helpful discussions about the history of the study of animal behavior.

REFERENCES AND SUGGESTED READING

Adkins-Regan, E. (2005). Hormones and Animal Social Behavior. Princeton, NJ:
Princeton University Press.
Adkins-Regan, E. & C. S. Carter. (2010). Neurobiology, endocrinology and behav-
ior. In M. Breed & J. Moore (eds.), Encyclopedia of Animal Behavior, Volume
2 (pp. 549–556). Amsterdam: Academic Press.
Allee, W. C. (1951). The Social Life of Animals. Boston: Beacon Press.
Allee, W. C., A. E. Emerson, O. Park, T. Park, & K. P. Schmidt. (1949). Principles of
Animal Ecology. Philadelphia: W. B. Saunders.
Aristotle. (1909). De mirabilibus auscultationibus. Oxford, UK: Clarendon Press.
Balda, R. P., I. M. Pepperberg, & A. C. Kamil. (1998). Animal Cognition in Nature.
New York: Academic Press.
Banks, E. M. (1985). Warder Clyde Allee and the Chicago school of animal behavior.
Journal of the History of Behavioral Sciences, 21, 345–353.
Beach, F. A. (1948). Hormones and Behavior. New York: Paul B. Hoeber.
Bliss, E. L. (ed.) (1962). Roots of Behavior. New York: Harper and Brothers.
Brown, J. L. (1975). The Evolution of Behavior. New York: W. W. Norton.
Brown, J. L. (2010). Behavioral ecology and sociobiology. In M. Breed & J. Moore
(eds.), Encyclopedia of Animal Behavior, Volume 1 (pp. 186–190). Amsterdam:
Academic Press.
76 ANIMAL BEHAVIOR

Burghardt, G. M. (2010). Comparative animal behavior—1920–1973. In M. Breed

& J. Moore (eds.), Encyclopedia of Animal Behavior, Volume 2 (pp. 340–
344). Amsterdam: Academic Press.
Burkhart, R. W., Jr. (2005). Patterns of Behavior: Konrad Lorenz, Niko Tinbergen, and
the Founding of Ethology. Chicago: University of Chicago Press.
Carter, C. S. (1974). Hormones and Sexual Behavior. Stroudsburg, PA: Dowden,
Hutchinson & Ross.
Clemmons, J. R. & R. Buchholz. (1997). Behavioral Approaches to Conservation in the
Wild. Cambridge, UK: Cambridge University Press.
Clutton-Brock, T. H. (1991). The Evolution of Parental Care. Princeton, NJ: Prince-
ton University Press.
Cutright, P. R. & M. J. Brodhead. (1981). Elliott Coues, Naturalist and Frontier His-
torian. Urbana: University of Illinois Press.
Dewsbury, D. A. (1984). Comparative Psychology in the Twentieth Century. Strouds-
burg, PA: Hutchinson Ross.
Dewsbury, D. A. (1986). Studying Animal Behavior: Autobiographies of the Founders.
Chicago: University of Chicago Press.
Dewsbury, D. A. (1988). A brief history of the study of animal behavior in North
America. In P. P. G. Bateson & P. H. Klopfer (eds.), Perspectives in Ethology,
Volume 8: Whither Ethology? (pp. 85–122). New York: Plenum Press.
Dewsbury, D. A. (1999). The proximate and the ultimate: Past, present, and future.
Behavioral Processes, 46, 189–199.
Drickamer, L. C. (2010). Animal behavior: Antiquity to the sixteenth century. In M.
Breed & J. Moore (eds.), Encyclopedia of Animal Behavior, Volume 2 (pp. 63–
67). Amsterdam: Academic Press.
Drickamer, L. C. (2010). Animal behavior at a crossroads. Distinguished Animal
Behaviorist Lecture, Animal Behavior Society, Williamsburg, VA.
Drickamer, L. C. (2010). Animal behavior: The seventeenth to the twentieth centu-
ries. In M. Breed & J. Moore (eds.), Encyclopedia of Animal Behavior, Volume
2 (pp. 68–72). Amsterdam: Academic Press.
Drickamer, L. C. & D. A. Dewsbury. (2010). Leaders in Animal Behavior: The Second
Generation. Cambridge, UK: Cambridge University Press.
Drickamer, L. C. & P. A. Gowaty. (2010). Future of animal behavior: Predicting
trends. In M. Breed & J. Moore (eds.), Encyclopedia of Animal Behavior,
Volume 2 (pp. 764–769). Amsterdam: Academic Press.
Eibl-Eibesfeldt, I. (1970). Ethology: The Biology of Behavior. New York: Holt, Rine-
hart and Winston.
Griffin, D. R. (1981). The Question of Animal Awareness. New York: Rockefeller Uni-
versity Press.
Griffin, D. R. (1984). Animal Thinking. Cambridge, MA: Harvard University Press.
Hamilton, W. D. (1964). The genetical evolution of social behavior, I, II. Journal of
Theoretical Biology, 7, 1–52.
Jablonka, E. & M. J. Lamb. (2006). Evolution in Four Dimensions. Cambridge, MA:
MIT Press.
IN THE BEGINNING : A HISTORY OF ANIMAL BEHAVIOR 77

Klopfer, P. H. (1973). An Introduction to Animal Behavior: Ethology’s First Century.

Englewood Cliffs, NJ: Prentice-Hall.
Krebs, J. R. & N. B. Davies. (1993). An Introduction to Behavioral Ecology. Oxford,
UK: Blackwell Scientific Publications.
Krebs, J. R. & N. B. Davies (eds.). (1997). Behavioral Ecology: An Evolutionary
Approach. Oxford, UK: Blackwell Scientific Publications.
Kruuk, H. (2003). Niko’s Nature. Oxford, UK: Oxford University Press.
Ley, W. (1968). Dawn of Zoology. Englewood Cliffs, NJ: Prentice-Hall.
Lorenz, K. Z. (1981). The Foundations of Ethology. New York: Springer-Verlag.
McGill, T. E. (1965). Readings in Animal Behavior. New York: Holt, Rinehart and
Winston.
Morgan, C. L. (1900). Animal Behaviour. London: Edward Arnold.
Morgan, C. L. (1903). An introduction to comparative psychology, 2nd ed. London: W.
Scott.
Nelson, R. J. (2000). An Introduction to Behavioral Endocrinology. Sunderland, MA:
Sinauer Associates.
Nordenskiold, E. (1928). The History of Biology. New York: Tudor.
Pepperberg, I. M. (2002). The Alex Studies: Cognitive and Communicative Abilities of
Grey Parrots. Cambridge, MA: Harvard University Press.
Romanes, G. J. (1883). Animal Intelligence. New York: D. Appleton and Co.
Singer, C. (1959). A History of Biology. London: Abelard-Schuman.
Stephens, D. W. & J. R. Krebs. (1986). Foraging Theory. Princeton, NJ: Princeton
University Press.
Taborsky, M. (2010). Ethology in Europe. In M. Breed and J. Moore (eds.), Encyclo-
pedia of Animal Behavior, Volume 1 (pp. 649–651). Amsterdam: Academic
Press.
Thorpe, W. H. (1966). Learning and Instinct in Animals. Cambridge, MA: Harvard
University Press.
Thorpe, W. H. (1979). The Origins and Rise of Ethology. New York: Praeger.
Tinbergen, N. (1951). The Study of Instinct. Oxford, UK: Oxford University Press.
Tinbergen, N. (1963). On aims and methods of ethology. Zeitschrift für Tierpsycholo-
gie, 20, 410–433.
Toates, F. (2001). Biological Psychology: An Integrated Approach. Reading, MA: Addi-
son Wesley.
Trivers, R. L. (1972). Parental investment and sexual selection. In B. Campbell (ed.),
Sexual Selection and the Decent of Man, 1874–1974 (pp. 136–179). Chicago:
Aldine.
Vessey, S. H. & L. C. Drickamer. (2010). Proximate and ultimate causation in the
study of animal behavior. In M. Breed and J. Moore (eds.), Encyclopedia of
Animal Behavior, Volume 1 (pp. 180–185). Amsterdam: Academic Press.
Watson, J. B. (1930). Behaviorism, Rev. ed. Chicago: University of Chicago Press.
Wilson, E. O. (1975). Sociobiology: The New Synthesis. Cambridge, MA: Harvard
University Press.
This page intentionally left blank
 3

Being Direct: Can Pigeons Plan Ahead?

Brett M. Gibson

INTRODUCTION
As a young adult I recall canoeing in the Boundary Waters in northern
Minnesota and closely attending to the shape of a lake or river to help deter-
mine my current location on a map. Our group often designated me as the
navigator because of my strong spatial skills on the open water. However,
being a parent of two young children has taken its toll on my spatial skills
recently. A good example of such a deficit in spatial cognition occurred the
other day. My wife and I were walking with our children, and we had to run
two errands—returning a children’s video to the library and shopping for
groceries—before returning home. The library was to the northeast of our cur-
rent position and the store was to the northwest (thus, our position at the time
was at the base of a virtual T). Being the primary navigator in our household,
I confidentially recommended traveling to the library first and then to the gro-
cery store before heading home. Unfortunately, my wife pointed out that the
proposed route was inefficient as it would require us to backtrack past the
library on the way home. She suggested the shorter route of going to the gro-
cery store first and then to the library on the way home. Note that my wife’s
route not only was more efficient in terms of the distance traveled but also
may have reduced other travel costs (e.g., minimizing travel time and reducing
the opportunity for a child breakdown away from home).
80 ANIMAL BEHAVIOR

THE TRAVELING SALESPERSON PROBLEM

Determining a route to several locations such as in the example above is
called the traveling salesperson problem (TSP). One can use the formula N!,
given that N is the number of places or nodes to be visited, to determine the
number of possible solutions to a TSP. Researchers have used TSP-like prob-
lems to examine the spatial cognitive abilities of humans (MacGregor &
Ormerod, 1996; MacGregor, Ormerod, & Chronicle, 1999, 2000; Vickers
et al., 2006) and nonhuman animals (MacDonald & Wilkie, 1990; MacDon-
ald, 1994; Gallistel & Cramer, 1996; Cramer & Gallistel, 1997; Gibson et al.,
2007; Miyata & Fujita, 2010). Like people, many nonhuman animals (hence
forth referred to as animals) need to travel to multiple locations during the
course of a day. For example, many animals need to travel to well-known loca-
tions where food, water, or mates might be found. Animals may have to find
an efficient route to these locations in order to minimize the costs associated
with a longer route, such as increased exposure to predators or a reduced
amount of time spent in other activities like feeding. Indeed, much work in
behavioral ecology has shown that animals are sensitive to different travel
costs. In some well-cited studies, Alex Kacelnik and colleagues (Kacelnik,
1984; Kacelnik & Cuthill, 1987) examined the number of invertebrates that
European starlings (Sturnus vulgaris) would take in their bills back to the nest.
They found that starlings would take fewer invertebrates back to their nest if
the outward trip was short but would increase their load as the distance
between the foraging site and the nest was longer. It would appear that many
animals are able to appreciate the costs associated with different traveling
routes and adjust their behavior accordingly. One important question is
whether animals are sensitive to the travel costs associated with a TSP.
Emil Menzel (1973) was one of the first to investigate whether animals can
solve TSP-like problems efficiently. In Menzel’s study, individual chimpan-
zees (Pan troglodytes) were carried around a large compound by an experi-
menter. Food was hidden in each of 18 locations inside the compound as
the chimpanzee watched. Next, the focal animal and other chimpanzees that
were “blind” to (did not know) the location of the food were released and
allowed to recover the hidden food. The informed animal searched at locations
where the researcher had placed the food using new but efficient routes to
travel from one location to the next. More recent work—including that
reported in the first part of this chapter—has also shown that animals in addi-
tion to chimpanzees adjust their travel routes based on the economics of trav-
eling to multiple locations as in a TSP (e.g., Blaser & Ginchansky, 2012;
Janson, 1998, 2007; Miyata & Fujita, 2010).
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 81

On the basis of the work showing that a variety of animal travelers take effi-
cient routes, a second line of research has started to examine how animal trav-
elers might solve TSPs in a way that would allow them to be efficient. One
possibility is that a traveler might represent the problem globally and encode
into memory the multiple locations that need to be visited. Such a representa-
tion might allow the traveler to consider multiple travel routes and also plan a
route several steps ahead before setting off on the journey (Janson, 1998,
2007; Miyata et al., 2006; Miyata & Fujita, 2008, 2010). Solving the problem
from a more global perspective may allow the traveler to plan beyond just the
next step in the route. Here we define planning in its broadest sense, that is,
developing a plan for future movements before they actually occur (Friedman
& Scholnick, 1997). Planning several steps ahead may require an allocentric
representation such as a cognitive map—an internal representation of impor-
tant places that indicates the geometric relationships between them (O’Keefe
& Nadel, 1978; Gallistel, 1990; see Shettleworth, 2010, for a review). Such
a representation could be used to determine an efficient route to multiple
locations.
The efficient traveling behavior of the chimpanzees in Menzel’s study is
consistent with cognitive mapping (Tolman, 1948; O’Keefe & Nadel, 1978;
Gallistel, 1990; Shettleworth, 2010). Likewise, Randy Gallistel and Audrey
Cramer (1996) had monkeys find food that was hidden at four goal locations
positioned at the corners of a diamond. During some occasions the monkeys
had to make a round trip (going to each location once before returning to
the start location), and at other times the monkeys had to make a one-way trip
from the start and visit all locations but were not required to return to the
starting point. During the round-trip condition the monkeys tended to use
an efficient perimeter route, but during the one-way trip condition they
changed their behavior and used a Z-like route in going to each location.
Thus, it would appear that the monkeys planned a travel route in advance
based on the configuration of the four locations and the task requirements.
While the monkeys’ behavior in Gallistel and Cramer’s study was consistent
with a stored representation of the spatial relationships among several loca-
tions, it is not clear these representations are required for efficient traveling.
Cognitive maps remain controversial (Bennett, 1996), and evidence for such
internal representations can often be explained by alternative mechanisms,
such as vector addition or learning (e.g., Brown et al., 1993; Bennett, 1996;
Gibson & Kamil, 2001; Hamilton et al., 2007; Shettleworth, 2010).
In TSPs with a limited number of locations, like the one in Figure 3.1, it is
pretty easy for a person (one would think, at least!) to determine most of the
solutions and select an efficient route. However, a traveler’s ability to generate
82 ANIMAL BEHAVIOR

Figure 3.1. Four locations—(C) current, (G) grocery store, (L) library, and (H) home—
as potential destinations along a hypothetical route. The arrows indicate some potential
routes of travel to get from the current location (C) to home (H). The arrow in the top-
right corner indicates north.

and peruse most of the solutions becomes more demanding as the number of
locations (nodes) in the problem increases. In the example shown in Figure 3.1
there are three locations (if the location of home is counted as a destination),
so there are 3! (3 × 2 × 1) or 6 potential routes from which to choose. But sup-
pose my kids were starting to get grumpy while I was formulating my route so
I also wanted to stop for lunch, and then later pick up the dry cleaning and
also get a chai tea for the walk home. Now there would be 6! or 720 possible
routes to choose from! If I had a problem selecting an efficient route with 12
possible solutions, I am sure to struggle with discriminating among 720 solu-
tions in the expanded problem. My decision is made somewhat easier by the
fact that half of the solutions are the same route taken in the other direction,
and some of these may be eliminated quite easily (for example, I hope I would
not select going back to my home from the starting location as the first leg of
my route for my remaining errands).
Finding an efficient route, for example one that minimizes the distance
traveled, for problems with a large number of nodes would appear to be more
challenging and potentially cognitively demanding. Instead of using a global
strategy, some travelers may instead solve TSPs and minimize the amount of
cognitive demand by using a rule-based solution to determine a route. The
use of a simple rule would not require that the traveler plan one or more steps
ahead along the route. An example of a rule-based solution is the local nearest-
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 83

neighbor solution. The nearest-neighbor solution specifies that a traveler select

the location that is closest to its current location and then continue to travel to
the next closest location until all locations have been visited once. The neighbor
solution is a more local solution as it uses proximity between the current and next
set of locations as a rule for identifying the next leg in the route and therefore
does not require that the traveler plan beyond the next step (see Kong & Schunn,
2007). It is important to point out that a solution that occurs as a result of the
neighbor algorithm and one from planning a route several steps ahead of the cur-
rent location may both be quite efficient. The cognitive abilities and require-
ments that are required for each type of solution may be different, however.
In this chapter I describe two attempts to study how humans and animals
solve TSPs. The first examines to what extent people and pigeons (Columba
livia) can efficiently solve virtual TSPs on computer monitor in an operant
chamber (“Skinner box”). People were used as a comparison group for the
performance of pigeons. In addition, in this work my colleagues and I exam-
ined whether pigeons would be sensitive to different costs associated with tak-
ing an inefficient route. While past work has shown that nonhuman primates
may select efficient routes, little work has been done exploring route-selection
behavior in birds. In the second set of studies my colleagues and I examined
how pigeons solved TSPs in a 3-D laboratory environment. We examined
the extent to which pigeons were efficient in solving an ambulatory version
of this task, but also whether they were using a global or local solution. That
is, we were curious whether pigeons were capable of planning beyond the next
step along a route. It remains unclear the extent to which animals are using
local compared to global solutions when solving TSPs and how route perfor-
mance may vary in virtual and real environments. Data suggesting that
pigeons can plan a route beyond its next step would indicate that they have
an understanding of the geometric relationships among multiple locations that
can be used to direct sophisticated spatial behavior.

ARE PEOPLE AND PIGEONS EFFICIENT TRAVELERS IN 2-D SPACE?

In an initial study we investigated the ability of people and pigeons to
solve TSPs that were presented on a computer monitor (Gibson et al.,
2007). People have been examined for their ability to solve TSPs (Ormerod &
MacGregor, 1996; MacGregor et al., 1999, 2000), but most of these tests
have had the participants solve the problems by connecting nodes drawn on
a piece of paper. Our goal was to explore the extent to which people and
pigeons can solve 2-D TSPs efficiently and whether or not the potential solu-
tions each “traveler” used were comparable to each other.
84 ANIMAL BEHAVIOR

People
We recruited 18 adult female undergraduate students for the study, who
gave informed consent; credit was given as part of a course research-
experience requirement. The experimenters had people sit in a chair in front
of a 15-inch high-resolution LCD monitor; the participant could adjust the
chair so that she could view the stimuli from a distance of 0.5 m (3.67 degrees
of visual angle). During the start of each trial we presented a black cross in the
center of the display area of the monitor to signal the trial onset. The partici-
pant used a cursor controlled by a mouse to click the stimulus. Next, the com-
puter program displayed three or more identical 2 cm square nodes on the
monitor. The computer randomly determined the position of the nodes in
the display area with the constraint that no node could be closer than 1 cm
to the border of the display area. Participants therefore received different prob-
lems on each trial of the session. Participants were instructed to click an initial
node and then each of the remaining nodes just once to complete the trial.
The shape of the node changed from light to dark to indicate to the partici-
pant that a node had been selected. An ominous tone sounded if the partici-
pant returned to a previously selected node, and the screen also went dark
for 20 seconds prior to the trial being repeated. The participants were told to
select a node and try and find the shortest route to pass through each of the
remaining nodes on the screen. We also told the participants that route effi-
ciency rather than speed was important in solving the problem. We presented
the participants with 96 trials in a daily session: 32 problems each with three,
four, and five nodes that were randomly intermixed.

Pigeons
We also examined four adult pigeons. The pigeons were kept at 85 percent
of their free feeding weight to ensure motivation for food as an outcome
during testing in the operant chamber. We trained and tested the pigeons in
operant chambers that had a 15-inch monitor positioned in the front wall.
We positioned a touchscreen over the monitor to record the Cartesian loca-
tion (coordinates in a 2-D plot) of the pecks the birds made to the stimuli pre-
sented on the monitor. The computer dispensed food pellets into a small cup
located on the floor next to the rear wall of the chamber following appropriate
responses. We initially trained the pigeons to respond to one and then two
nodes (the same as described previously) in the display area to familiarize them
with the “rules” of the traveling task. As was the case for people, we presented
a start stimulus in the center of the display area to signal the beginning of a
trial. A peck to the start stimulus advanced the trial; the screen went black,
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 85

and initially a single node (as described for people) then appeared. The com-
puter randomly determined the position of the node prior to the start of a trial.
The node became dark after being pecked. The computer then cleared the dis-
play area and two food pellets were delivered. We required that the pigeons
peck each node once during this two-node phase of training. The computer
cleared the display area after the second node was pecked, and responses to
previously pecked nodes resulted in the trial being redone as described previ-
ously. Each daily session consisted of 120 trials.
Next, during the testing phase we presented the pigeons with problems that
included three nodes. Each pigeon could select any of the three nodes as the
starting node. The birds then needed to peck the two remaining nodes just
once without returning to a previously responded-to node. Each daily session
with three nodes consisted of 120 problems, and we continued testing for
12 days. We subsequently conducted testing with four nodes in a problem
for a block of 12 sessions and then a block of 12 sessions with five nodes.
These daily sessions also had 120 trials.

Determining the Routes Chosen by People and Pigeons

We determined the sequence of nodes that the people and pigeons selected
during each trial. We were able to construct the “route” that both people and
pigeons choose to take to connect the nodes on the screen and subsequently
determined the total distance of the route for each trail. While the distance
of the routes selected by the travelers is only one of several dependent measures
for route efficiency, this measure is likely to be strongly correlated with other
potential measures of performance (e.g., time, energy expenditure). As men-
tioned previously, the number of possible one-way routes for a given trial or
problem can be determined by the equation N!, where N is the number of
nodes. There are six possible solutions (3! = 6) for a problem with three nodes;
however, half of these are the same route taken in the opposite direction.
Thus, for the purpose of this experiment we identified three routes that had
the same distance and generally ignored the direction in which the route was
implemented. In subsequent work discussed later in this chapter the direction
in which the route was taken is considered and plays an important role in our
analysis.
We developed Monte Carlo (random), nearest-neighbor, and optimal
models that were each given the same exact problems (three, four, and five
nodes) that the pigeons and people encountered. We then compared the dis-
tance of the routes selected by each model with that selected by the people
and pigeons. We used the Monte Carlo, or random, model to determine the
distance of a solution for each problem given a random selection of nodes.
86 ANIMAL BEHAVIOR

Specifically, we programed the Monte Carlo model to select a starting node

(randomly determined) and then choose each remaining node in the problem
until all nodes had been selected just once. For the nearest-neighbor model we
created a program that determined the identity of the node that the participant
had first selected during a trail (for both people and pigeons). Next, the
nearest-neighbor program selected the node closest to first node in the prob-
lem as its second choice. The neighbor rule was subsequently applied to
all the remaining unselected nodes until the route was completed. We also cal-
culated the proportion of routes that the people and pigeons selected that
corresponded with the route that was selected by the nearest-neighbor model
as an additional measure. Finally, for the optimal model we programed the
computer to select the single shortest possible route for each problem.
We also were curious if the travelers were selecting nodes that were clus-
tered together to improve overall route efficiency or perhaps because they are
attracted to objects that are spaced close together. We calculated the internode
distance between all possible combinations of two nodes in a problem and
ranked them from shortest to longest. If the travelers choose clusters of nodes,
then they should regularly choose the two nodes with the shortest internode
distance at a high frequency. As part of this cluster analysis we also compared
the proportion of trials in which people and pigeons selected a cluster of nodes
as a function of the traveler’s first choice in the route to the proportion of trials
that a Monte Carlo simulation given the same problem selected clusters
during its first choice.
One important feature of the current experimental design was that each
participant encountered a different configuration of nodes on each trial; thus,
the problem set and solutions that each participant encountered were unique.
Because we programed the computer to randomly generate the coordinates of
the nodes for each trial, we had little control over the resulting node configu-
ration on the screen. We developed a measure of solution difficulty (disparity)
that quantified the degree of difference in path length among the solutions for
a given problem. For some trials the traveling problem may have been rela-
tively easy, for instance for problems with three nodes when all three nodes
were on the same line. If two of the nodes in the problem are clustered
together and the third happens to be farther away, then the difference in the
total distance among the three solutions could be quite large, and an efficient
solution may be relatively easy to determine. Detecting which of the three sol-
utions is the most efficient may be less cognitively demanding in these cases.
In contrast, it may be much harder to distinguish the most efficient of the
three solutions for a problem in which the three nodes are positioned near
the corners of an equilateral triangle. The three solutions would all have nearly
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 87

the same length, and determining the single best route would be harder than
when the three nodes were in a line. Solution disparity was determined for
each problem by calculating the mean path length of all possible one-way sol-
utions for a trial and dividing that mean by the standard deviation of the
solutions. The disparity score increases as the difference among the solutions
increases, whereas the disparity score decreases when the distances of the
possible routes to a solution become more comparable.

What Did People and Pigeons Do?

As can be seen in Figure 3.2, the routes that the people selected were much
shorter than those of the Monte Carlo model but typically longer than those of
the optimal route for TSPs with three, four, and five nodes. Thus, people gen-
erally selected an efficient route even though they did not always take the sin-
gle most efficient one. The routes that the people choose had approximately
the same length as those selected by the nearest-neighbor model for these same
problems, suggesting that people may be employing proximity as part of their
solution. The mean proportion of routes selected by people that were identical
to the nearest-neighbor model was 0.82  0.05 (standard error) for prob-
lems with three nodes, 0.78  0.05 for problems with four nodes, and 0.67
0.09 for problems with five nodes. Thus, as the problems potentially became
more challenging, people used the neighbor solution somewhat less often.
The total length of the routes selected by people was much shorter than the
length of the routes selected by the Monte Carlo model for the 25 percent of
the problems that had the highest solution disparity. This pattern was
true whether the problems had three, four, or five nodes. However, as the
solution disparity for the problems decreased (moving from left to right in
Figure 3.3)—and the differences between the routes were reduced—the
distance of the routes selected by the Monte Carlo model and the people gen-
erally became more comparable, particularly for TSPs with five nodes. The
distance of the routes selected by people again was longer than optimal but
comparable with that of the nearest-neighbor model. The only deviation from
this pattern was again for problems with five nodes, in which the distance of
the routes selected by the people tended to increase beyond that of the neigh-
bor model as solution disparity decreased. Thus, even though people could
have used the neighbor solution more often, they failed to do so even though
it would have resulted in a more efficient route. Perhaps for TSPs with five
nodes with the smallest amount of solution disparity the difference in the dis-
tance between the nodes became less apparent for people and performance
declined.
Figure 3.2. The mean length of the routes selected by the Monte Carlo model, people,
the nearest-neighbor model, and the optimal model for problems with three nodes
(top), four nodes (middle), and five nodes (bottom).

88
Figure 3.3. The mean length of the routes selected by the Monte Carlo model, people,
the nearest-neighbor model, and the optimal model for problems with three nodes
(top), four nodes (middle), and five nodes (bottom) across decreasing levels of solu-
tion disparity.
89
90 ANIMAL BEHAVIOR

People tended to choose the shortest possible leg of the TSP as the first
leg of their route when compared to the Monte Carlo model (Figure 3.4).
This pattern was true for problems with three, four, and five nodes. People
tended to select clusters of nodes early during the formation of the route rather
than later in the process. The reason why the people tended to select the clus-
ters early in the trial remains unclear. The routes would be just as efficient by
including the cluster at the end compared to the beginning of the route, so
perhaps the propensity to select a cluster early has more to do with a percep-
tual tendency when selecting a route during a 2-D task.
The mean distance of the routes selected by the pigeons was shorter than
that of the Monte Carlo model and longer than optimal for problems with
three, four, and five nodes (Figure 3.5). Thus, like people, the pigeons were
not randomly selecting the nodes to form a route. The routes selected by the
pigeons were longer than those selected by the nearest-neighbor model for
problems with three, four, and five nodes (Figure 3.5). In contrast, the length
of the routes selected by people tended to overlap with those of the Monte
Carlo model. Likewise, the proportion of the pigeons’ routes that were identi-
cal to the nearest-neighbor model was substantially lower than that observed
for people. The mean proportion of neighbor solutions for the pigeons was
0.35  0.03 for problems with three nodes, 0.31  0.01 for problems with
four nodes, and 0.33  0.04 for problems with five nodes. The difference in
the mean distance of the routes selected by the pigeons and the other models
described above appeared to remain consistent across solutions with high to
modest levels of solution disparity (Figure 3.6). When solution disparity
was most extreme, the distance of the pigeons’ routes overlaps considerably
with that of the Monte Carlo model (Figure 3.6, far right).
Like people, the pigeons tended to select clusters of nodes for their first
choice (Figure 3.7). The pigeons choose the two nodes with the shortest inter-
node distance as part of their first choice of the route at a high level for TSPs
with three and four nodes. The tendency to select a cluster as part of the first
choice diminished for problems with five nodes, in contrast to what was
observed for people.

A More Challenging Problem

The cost of selecting an inefficient route would appear to be relatively small
for the pigeons and may account for why they selected longer routes on some
trials compared to people. Therefore, in a follow-up study, we attempted to
improve the efficiency of the routes selected by the pigeons by requiring that
their routes meet a minimum standard of performance (criterion). The same
pigeons were used in the follow-up. The general procedures for this study were
Figure 3.4. The proportion of trials in which each possible leg of a problem was
selected as the first leg of a route by people for problems with three nodes (top), four
nodes (middle), and five nodes (bottom). The possible legs of a problem are rank
ordered on the x axis from shortest (left) to longest (right).
Figure 3.5. The mean length of the routes selected by the Monte Carlo model,
pigeons, the nearest-neighbor model, and the optimal model for problems with
three nodes (top), four nodes (middle), and five nodes (bottom).

92
Figure 3.6. The mean length of the routes selected by the Monte Carlo model,
pigeons, the nearest-neighbor model, and the optimal model for problems with
three nodes (top), four nodes (middle), and five nodes (bottom) across decreasing
levels of solution disparity.
93
Figure 3.7. The proportion of trials in which each possible leg of a problem was
selected as the first leg of a route by pigeons for problems with three nodes (top),
four nodes (middle), and five nodes (bottom). The possible legs of a problem are rank
ordered on the x axis from shortest (left) to longest (right).
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 95

similar to those used previously, except that now the pigeons only encountered
problems with four nodes. Additionally, the route a pigeon selected during the
course of a trial was evaluated to determine if it met a criterion for efficiency.
The pigeons were required to select a route whose solution was greater than
or equal to that of the criterion before continuing onto the next trial. If the
route that the pigeon selected during the first pass through the trial did not
meet or exceed the criterion, a correction trial occurred until the pigeon had
a solution that was above the criterion. Specifically, for an initial block of 10
daily sessions (120 trials/session) the route that a pigeon selected during the
course of a trial had to be shorter than 42 percent of all possible routes for that
problem. The birds were required to select a route that was shorter than
50 percent of all routes and then a route that was shorter than 66 percent of
all possible routes during a third and a fourth block of 10 sessions,
respectively.
To see if our procedures had an impact on the pigeons’ routes, we calculated
two difference measures. For the first measure, we subtracted the mean length of
the routes selected by the pigeon from the mean length of the routes selected by
the Monte Carlo model (MC–P score). The MC–P score should increase if the
routes the pigeons select become shorter. For the second measure we subtracted
the mean length of the routes selected by the pigeon in each daily session
from the mean length of the routes selected by the nearest-neighbor model
(NN–P difference score). This difference score should decrease as the distance
of the routes declines.
Figure 3.8 (top panel) shows that as the criterion was imposed the routes
the pigeons selected became more efficient compared to the earlier study and
a baseline condition with four nodes that was conducted prior to the start of
the follow-up. Specifically, the difference in distance between the routes that
the pigeons and the Monte Carlo model selected became larger (MC–P, black
bars), whereas the difference in distance between the routes the pigeons and
the nearest-neighbor model selected decreased. In addition, as the perfor-
mance criterion was increased the changes in the difference scores become
more robust. The numbers in the white bars in Figure 3.8 (top panel) show
the proportion of the pigeons’ routes that conformed to the nearest-neighbor
route. As the performance criterion increased the pigeons used a proximity
rule with more regularity. Thus, one way that the pigeons became more effi-
cient was to utilize proximity cues during the route formation process. The
bottom panel in Figure 3.8 indicates that as the criterion increased so too
did the proportion of trials in which the pigeons selected a cluster as the first
leg of their route. The fact that the birds were selecting the shortest leg as
the first leg of a route does not necessarily indicate that this was the
Figure 3.8. (Top) The difference in the length of the routes selected by pigeons and
the Monte Carlo model (MC - P, filled bars) and by pigeons and the nearest neigh-
bor model (NN - P, unfilled bars) during the different criterion conditions. The
proportion of trips selected by the pigeon that were identical to those taken by the
nearest neighbor model is indicated by the unfilled bar. (Bottom) The proportion of
trials that each possible leg of a problem was selected as the first leg of a route by
pigeons for problems with four nodes for the different criterion conditions (baseline,
42%, 50%, 66%) and by the Monte Carlo model. The possible legs of a problem are
rank ordered on the x axis from shortest (1) to longest (6).
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 97

Figure 3.9. The proportion of routes taken by the pigeons that conformed to the
neighbor model during Experiment 3 as a function of the rank (1 to 6) of the first
leg used in the neighbor route and performance criterion (baseline, 42%, 50%, 66%).

mechanism for improved efficiency, only a tendency to select two nodes that
were close together. However, Figure 3.9 indicates that not only were the
pigeons selecting the shortest leg of a problem as the first leg of their route,
they also had a greater tendency to utilize a neighbor route (a more efficient
route) when doing so.

HOW DO WE INTERPRET THESE RESULTS?

The results from people in the current study are consistent with past work
indicating that human “travelers” will select an efficient route when solving
2-D TSPs (e.g., Blaser & Ginchansky, 2012 MacGregor & Ormerod, 1996,
2000; Vickers et al., 2001, 2003, 2006). James MacGregor and Tom
Ormerod (1996) tested people using paper-and-pen versions of the TSP in
two experiments. In the first experiment of their study, people were given a
sheet of paper with 10 points as part of the TSP. The participants could select
any point as the starting point but had to travel to each subsequent point once
(without revisits) before returning to their starting point to complete the
“trip.” In a second experiment the procedures were the same except the partic-
ipants were given problems with 20 points. People selected highly efficient sol-
utions that were within the 99 percentile of the optimal solution and that were
consistently better than the local nearest-neighbor model. The efficient choice
of routes was true regardless of the total number of points in the problem. The
performance of people in the current study was efficient but not as optimal as
98 ANIMAL BEHAVIOR

that reported by MacGregor and Omerod (1996) and other comparable work
(Vickers et al., 2001). In our study, the routes selected by people were signifi-
cantly less efficient than optimal. Likewise, the routes selected by people
tended to be more comparable to those of the local nearest neighbor for prob-
lems with the highest solution disparity. There are at least two possible
differences between our study and those conducted by other researchers
using pen-and-paper tasks that might account for the reduced efficiency of
people in the current study. In our study the participants selected a node
and it changed color (so the participants did not have to remember which
node they had visited), but there were no visible lines connecting the nodes,
as in the pen-and-paper studies. Thus, our task may have placed a greater
demand upon memory. Additionally, in our study the participants were
required to take a one-way trip, while in the other studies mentioned above
the participants had to complete a round trip. It may be the case that effi-
ciently solving one-way TSPs may be more challenging than completing a
round trip and also may account for some of the performance differences that
were observed.
The routes selected by the pigeons in our initial experiment, though more
efficient than those predicted by the Monte Carlo model, were less efficient
that those selected by people and the nearest-neighbor model. The fact that
the pigeons selected relatively efficient routes is consistent with other work
(MacDonald & Wilkie, 1990; MacDonald, 1994; Gallistel & Cramer,
1996; Cramer & Gallistel, 1997; Miyata & Fujita, 2010) and is impressive
given the diverse population of problems the pigeons were presented with in
our task. The difference in efficiency between the pigeons and the Monte
Carlo model appeared to be due to the fact that the pigeons, but not the
Monte Carlo model, rarely selected the most inefficient route. Notably, the
pigeons tended to be less efficient than would be expected if using a nearest-
neighbor solution. The finding is similar to that reported for rats (Rattus nor-
vegicus) given TSP-like problems (Blaser & Ginchansky, 2012). In that study,
rats were allowed to travel to each of 10 baited targets in any sequence after
departing a start location. Like the pigeons in the current study, the rats
selected routes that were more efficient than those produced by a Monte Carlo
model but less efficient than those of the nearest-neighbor solution. However,
in both our initial study with pigeons and the study by Rachel Blaser and
Rachel Ginchansky (2012) the animals’ motivation for producing efficient
routes may have been low.
In a second experiment we made the traveling demands more challenging,
and the pigeons responded by improving the efficiency of their solutions, pri-
marily using two techniques. Pigeons tended to start using the nearest-
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 99

neighbor solution with greater frequency, choosing nodes that were closer to
their current position on the monitor. Other work has also demonstrated that
animals will use the nearest-neighbor solution to solve TSP-like problems.
Hiromitsu Miyata and Kazuo Fujita (2010) required that pigeons travel to
two or three nodes before reaching a goal location on a computer screen.
The configuration of the nodes changed such that the nodes were on a line
or at the corners of a virtual square across experiments. The pigeons often
choose the nearest location to the starting point first before moving onto other
neighboring nodes, though this resulted in a less efficient route (similar to the
pigeons in the current study). Similarly, Charles Janson (1998) found that in
the natural environment monkeys tended to go to the closest available food
resource even when a more distant resource was substantially more useful
(Janson, 2007).
Related to the neighbor strategy, the pigeons in the current study also
tended to select clusters of nodes as part of their solution to improve route effi-
ciency in our second study. The tendency of the pigeons to initially seek out a
cluster of nodes is consistent with other work indicating that animals given
similar types of problems will try and travel to clusters of nodes before moving
onto nodes that are spaced farther apart and farther away. For instance,
Audrey Cramer and Randy Gallistel (1997) had vervet monkeys (Chlorocebus
pygerythrus) travel to several destinations to collect food. During some condi-
tions the food was clustered so that one side of the test environment had a
large cluster of locations, whereas the other side had relatively fewer locations.
The vervet monkeys tended to travel to the large cluster of locations first
before completing the remainder of their route, similar to the pigeons in the
current study. Importantly, this result also suggests that both groups of trav-
elers represented—perhaps in a “map”—most if not all of the nodes in the
problem before selecting a route.

Did Pigeons Use a Cognitive Map?

Menzel (1973) reported that chimpanzees took efficient routes that mini-
mized the distance traveled when returning to up to 18 sites where food could
be hidden. In addition, Gallistel and Cramer (1996) found that monkeys con-
sider the nature of the route requirements and adjust their behavior depending
upon whether they need to make a one-way or round trip to optimize the dis-
tance of their route. Miyata, Fujita, and colleagues (2006, 2008) have found
evidence that pigeons may preplan routes in a virtual maze task in an operant
chamber. In their initial study (Miyata et al., 2006) the pigeons viewed a
square target in the middle of a square frame defining movement space. Next,
an L-shaped barrier appeared between the target and a newly appearing square
100 ANIMAL BEHAVIOR

goal. The birds previewed the entire maze before the color of the screen
changed, cuing the pigeons to move the target to the goal using a pecking
response. During some tests the maze changed shaped following the preview
phase. In one condition the route used to get to the goal remained the same
(despite the change in the maze), but in another condition the route changed.
Across a variety of measure the performance of the pigeons declined when the
maze changed and the pigeons were required to implement a new route to
solve the maze compared to when the route did not change. Likewise, in a
related study Miyata and Fujita (2008) used a similar procedure, but testing
was conducted on a radial-arm-type maze on the computer screen. In that
study the pigeons moved a target from the distal end of one arm of the maze
through the center and then to a goal located at the distal end of another
arm. During some test conditions the goal switched locations to another arm
compared to the preview phase. The goal switched one or more “steps” on
the outbound arm before the pigeon moved the target through the center.
The authors found evidence indicating that the pigeons would change their
solutions after the change in goal location several steps prior to advancing to
the center of the maze.
These findings and others have suggested that some nonhuman animals
have a representation of the entire set of destinations and plan their routes
before starting the trip. That the pigeons in our study were more efficient than
the Monte Carlo model indicates that they were not selecting nodes at random
on most trials. However, this does not necessarily require the use of a cognitive
map or an ability to preplan. We also found that our pigeons tended to select
clusters of nodes and that this tendency increased as a criterion for efficiency
was implemented. The fact that pigeons selected the clusters early in the route
suggests the pigeons scanned a large portion of the monitor for clusters of
nodes before starting the route. Indeed, the clusters may have been a salient
stimulus for the pigeons, which would have resulted in an improvement in
efficiency. The pigeons could use proximity either in the form of detecting
clusters or forming a route based on the nearest-neighbor solution to form a
relatively efficient route. As mentioned in the introduction, a local solution
based on proximity would suggest a relatively local knowledge of the problem
rather than the use of a more global representation that might be consistent
with planning one or more steps ahead.
Some readers may be concerned that the magnitude of the disparities
between the travelers and the various models was relatively small.
My colleagues and I have argued that small differences in efficiency will multi-
ply over time. Another concern is that while the travelers appear to solve these
problems efficiency, the TSP tests here are not a valid measure of traveling
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 101

behavior in a 3-D environment. To addresses these concerns and to examine

whether or not pigeons can plan ahead, we examined the ability of pigeons
to solve TSPs while ambulating in a laboratory room.

OK, BUT CAN PIGEONS PLAN AHEAD?

The results from our first study indicated that pigeons can take relatively
efficient routes when “traveling” to multiple locations. In addition, the results
from these first studies also provide some clues as to what cues the pigeons
may be using to create a route. When the performance criterion was imple-
mented the pigeons showed an increased tendency to use proximity. However,
the use of proximity may only require a relatively local understanding of the
problem rather than considering the entire problem, which would be consis-
tent with planning beyond the next step. In a second body of work, we set
out to examine the extent to which pigeons might have knowledge of the geo-
metric relationships among multiple destinations and plan beyond the next
step. This second body of work also explores whether the efficiencies we
observed in a 2-D version of the TSP task would also be observed in a 3-D test
conducted in a laboratory environment.

The 3-D Test

We used eight adult White Carneaux pigeons that were experimentally naïve
for this study. Four of the pigeons were randomly assigned to a one-way group
and the other four to a round-trip group. During each trial one pigeon was
placed in a clear plastic start box (S) in the experimental room (2.1 m × 2.7 m;
see Figure 3.10). An end box (E, Figure 3.10) was in the corner of the room
opposite S. We placed three 60 cm (height) × 5 cm (width) cylindrical PVC
feeders in the room; peas were placed in shallow food wells at the base of each
cylinder. During each trial, the door to the start box was opened and the bird
was required to visit each feeder just once and consume the food. A buzzer
sounded after the bird consumed the food at the last feeder to indicate the end
of the trial. At this point the birds in the round-trip group learned to return to
S to complete a round trip, whereas the birds in the one-way group had to travel
to the end box (E). Once the birds in either group reached their final box a
bucket feeder that was filled with mixed grain and attached to the rear of the
box was raised for 2 seconds. If a pigeon returned to S (or E) before visiting the
second feeder, or to the first feeder again after visiting the second feeder, the
lights in the room were turned off, the trial ended, and a correction trial ensued
(errors occurred during less than 10% of the trials—data not included).
Figure 3.10. The performance of pigeons in the round-trip (dashed line) and one-way
(solid line) groups during the 16 three-feeder test configurations. Each of the panels
A-P shows the first or second choices (see text) for that configuration. The position
of the feeders is indicated by the filled circles. The location of the start box is indicated
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 103

by S, and the location of the end box for the birds in the one-way group is indicated
by E. The Cartesian grid used to position the feeders is shown but was not visible to
the birds. The number of choices and direction of choices to each feeder are indicated
by the arrows. There were a total of 16 first feeder choices for each configuration for
each group. The width of the arrow corresponds with the number of choices to a
particular feeder; bolder arrows indicate more choices to a feeder (arrow width of 4
font ¼ 16 choices). The arrow with the N to the far right of the panels indicates a
northerly direction.

We set up the floor of the room as a Cartesian grid (x–y coordinate system)
to help the experimenters position the feeders. Each cell was 30.5 cm wide,
and there were a total of 63 cells. We had two types of trials; during
random-condition trials the position of the three feeders was determined ran-
domly from the total set of cells with the rule that the feeders could not occupy
adjacent cells. During test-condition trials we arranged the feeders into one of
16 different predetermined test configurations designed to explore whether
the pigeons could plan one step ahead (see Figure 3.10). Center-near versus
peripheral-far configurations (Figure 3.10, A–B) were designed to see whether
the birds would prefer to go to a nearby central feeder or one of two more dis-
tant feeders in the periphery after departing S. We used the one-way/round-
trip global configurations (Figure 3.10, C–E) to examine if the pigeons would
choose an efficient global solution instead using a more proximal strategy. For
the global strategy configurations test, configurations F–H (Figure 3.10), the
feeders for the second choices were always the same distance and angle from
the feeder that we predicted the birds were likely to choose first. Proximity was
less likely to play a role in the second choices for either group for these tests.
On the basis of the work with the virtual TSP task reported earlier, we developed
cluster configurations to see if clusters of feeders would be preferred compared to
a lone feeder (Figure 3.10). We developed arc configurations, in which all three
feeders were spaced equally on an arc whose geometric center was either towards
(configuration O) or away (configuration P) from S, which allowed us to exam-
ine for choice preferences given a constant angle and distance relative to S.
There were eight trials in each daily session. During each daily session four of
the trials were random-condition trials and the other four were test-condition
trials; the order of the trials was randomly determined. The random trials men-
tioned above were used in part to disperse the test trials and maintain traveling
performance. Sampling from the pool of test configurations for each daily ses-
sion was done randomly without replacement. Testing continued until the birds
completed three sets with each test configuration (3 × 16 = 48 total test trials)
and a corresponding number of trials with the random configurations.
104 ANIMAL BEHAVIOR

Determining the Routes Chosen by Pigeons

As in the 2-D study reported earlier, we recorded the order of the nodes
(feeders) that the birds went to for each trial. Using this information we calcu-
lated the total distance of the pigeons’ route. We again used the Monte Carlo
and optimal models given the same problems as the pigeons to compare per-
formance. In a second set of analyses we examined the first or second choices
that the birds made of the feeders after departing S.

What Did Pigeons Do?

As with the 2-D traveling study, the mean distance of the routes that the
pigeons selected was shorter than the mean distance of the routes indicated
by the Monte Carlo model (Figure 3.11). In contrast to the 2-D study the
routes that the pigeons selected were not different in length from the opti-
mal solutions. The travel costs associated with the ambulatory task, or per-
haps the similarities of the 3-D task with natural foraging behavior, may
have facilitated the improvement in performance compared to the study
in the operant chamber.

Figure 3.11. The distance of the routes taken by the pigeons and the three models. The
values are taken across all groups and conditions for each traveler.
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 105

Center-near versus peripheral-far configurations. We observed that the

pigeons had a strong preference for the central and near feeder (Figure 3.10)
during these tests (A [feeder 1 ¼ 1 choice; feeder 2 ¼ 10 choices; feeder 3 ¼
21 choices] and B [feeder 1 ¼ 1 choice, feeder 2 ¼ 2 choices, feeder 3 ¼ 29
choices]). By choosing the feeder closest to S the birds had to travel less dis-
tance to get to food initially. After departing S the birds that traveled to the
most proximal feeder first saved 0.26–1.22 m in the distance traveled com-
pared to if they had traveled to one of the other two feeders first. While this
first choice was the most efficient in the short term, going to the closest feeder
from S first resulted in a relatively inefficient overall route for both groups of
birds. Specifically, choosing the most proximal feeder first after departing S
resulted in a 0.6 m and a 1.83 m increase in the length of the route for the
round-trip group and between a 0 m and a 1.21 m increase in the total length
of the trip for the one-way group.
One-way/round-trip global configurations. For test configurations C and D the
birds distributed their first choices approximately equally between feeders 1 and
2. For the second choice, the birds in the round-trip group should select feeder
3 based on proximity but also because such a choice would be consistent with
an efficient global solution. In contrast, the one-way group should travel across
the diagonal to feeder 1 if at feeder 2 (or vice versa) as part of a more global effi-
cient route if they are planning ahead. However, it is clear that the birds in both
groups traveled to feeder 3 during these conditions (64 of 64 choices). Configu-
ration E was used to examine whether the round-trip birds would select a feeder
consistent with planning ahead compared to a more proximal feeder at a critical
decision point. The first choices for the birds in both groups were again distrib-
uted between feeders 1 and 2 for test configurations E (Figure 3.10). The birds
in the round-trip group should choose feeder 3 if using a global solution and
choose feeder 1 if at feeder 2 (or vice versa) if using a neighbor or proximity based
solution. Thirteen of the 16 second choices for the birds in the round-trip group
were to feeder 3, which is consistent with planning ahead.
Global strategy configurations. The birds in the one-way group all went to
feeder 2 after departing S when presented with configuration F. During the
second choice, 15 of the 16 birds went to feeder 1 instead of feeder 3, even
though both of these choices were the same distance and angle from feeder
2. By traveling to feeder 1 instead of feeder 3 the pigeons reduced the total dis-
tance of their route by 1.87 m. Thus, these choices were consistent with plan-
ning ahead. The birds in the round-trip group made 13 choices to feeder 2 as
their first choice for the same configuration. Likewise, the pigeons in the
round-trip group should select feeder 3 instead of feeder 2 for their second
choice if planning ahead. Choosing feeder 3 would result in a savings of
106 ANIMAL BEHAVIOR

0.75 m for the round-trip group compared to selecting feeder 1. The round-
trip group made 10 of their 13 second choices to feeder 3, consistent with a
global knowledge of the remaining destinations and planning ahead. For con-
figurations A–E, when the pigeons primarily appeared to go to the next closest
feeder, proximity and other geometric features related to the second choice
were not held constant. When proximity to feeders was kept constant during
the second choice for testing with configuration F the pattern of choices for
both groups was consistent with planning ahead.
Test configurations G and H were similar to configuration F in that prox-
imity was held constant for the second choice. For these two configurations,
the second choices for the birds in the one-way group were highly consistent
with planning ahead, as 25 of 31 of their choices were to feeder 2 (a savings
of 1.32 and 1.83 m, respectively). The second choices for the birds in the
round-trip group were more mixed, however, as 14 of their choices were to
feeder 2 and 15 were to feeder 3.
While the results from configurations A–D indicated that the pigeons used
a proximity rule to make their first choice, the results from testing with con-
figurations E–H indicated that the pigeons appeared capable of planning one
step ahead. We observed that during testing with configuration E the pigeons
in the one-way group consistently avoided traveling to a proximal feeder first,
which resulted in a more efficient route. Likewise, for configurations F–H we
observed that the birds in the one-way group consistently chose a feeder that
was consistent with the most efficient global route when proximity between
second choices was held constant (configurations F–H).
The evidence from configurations E–H suggests that the pigeons can plan
one or more steps ahead, yet the results from configurations A–D for the
pigeons indicate that they prefer to go to a proximal feeder even when it results
in a less efficient overall route. One explanation for these apparent inconsis-
tencies is that in one case the saliency of the proximity option (configuration
B) was likely much larger than for the other problems. Also, for configurations
A–D the net difference in the distance of the local/proximal solution com-
pared to the global solution was relatively small (0.66 m, 1.21 m, 0.31 m,
and 0.16 m for configurations A–D, respectively). The small differences
between the routes are consistent with the idea of low solution disparity that
was introduced with the operant work reported earlier. Either the pigeons may
have had a difficult time discerning the difference in distance between the proxi-
mal and global solutions or the traveling costs associated with either solution may
have been quite small (e.g., energy costs or delays to food in the end location).
In contrast, for configurations F, G, and H, in which the pigeons in the
one-way group showed evidence of planning ahead, the difference between
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 107

the local and global solutions was notably larger at 1.83 m for F, 1.32 m for G,
and 1.83 m for H. The results from configurations F–H for the pigeons in the
round-trip group also showed evidence of planning ahead. The evidence for
planning ahead by the pigeons in the round-trip group was more variable than
for the one-way group, however. The net difference in the distance of the sol-
utions between the local and global strategies for problems F–H was larger for
the one-way group than the round-trip group (0.76–1.17 m round-trip versus
1.32–1.83 m for the one-way group) and may have accounted for the addi-
tional variability in choices by the round-trip group. In general, the net differ-
ence in distance in the routes based on proximity or planning ahead was
higher for these problems than for configurations A–D, perhaps accounting
for more reliance on planning.
Cluster configurations. For the cluster configuration tests the length of the
routes for the round-trip group would be the same regardless of whether or
not the birds traveled to the cluster or the single feeder first. Therefore, the
pigeons in the round-trip group might be expected to travel to the cluster of
two nodes first (assuming that both are visited in sequence) after departing S
to maximize their reward early rather than later in the trip. The birds in the
one-way group had a bit more difficult task in that they had to potentially bal-
ance going to the cluster first with the efficiency of the total route.
Across cluster test configurations the birds in the round-trip group showed no
preference for going to either the cluster or lone feeder first. Sixteen of the 32
choices were to the cluster across both test configurations for this group. The birds
in the one-way group had a strong preference to go to the location in the room
that would minimize the distance of their route regardless of whether a cluster
or lone feeder was present (Figure 3.10). Because of the geometric relationship
between S, the feeders, and E, it was 1.21 m shorter for the birds in the one-
way trip to travel to the feeder(s) that was positioned vertically relative to E (Figure
3.10) independent of the position of the cluster. Correspondingly, all 32 of the
choices were to the location (16 choices to feeders 1, configuration M; 16 choices
to the cluster for configuration N) that minimized the total distance of their route.
The results from the operant work reported above indicated that the
pigeons had a strong tendency to select a cluster of two nodes as their first
leg of their trip. In contrast, the results from the tests in the open room indi-
cated that pigeons did not have a preference to travel to the cluster. One pos-
sibility that might account for the apparent difference is that the birds
identified the two feeders as being part of a cluster when viewing them on a
monitor but not when moving in the room. We positioned the feeders
approximately 30 cm apart in the room, so the birds may not have categorized
these feeders as being part of a cluster.
108 ANIMAL BEHAVIOR

Arc configurations. The pigeons had a slight preference for choosing feeder 2
compared to feeders 1 and 3 for configuration O (7 choices to feeder 1, 15
choices to feeder 2, 10 choices to feeder 3). For configuration P, the pigeons
had a preference for traveling to feeder 3 as their first choice.

SO, DO ANIMALS PLAN AHEAD?

Past work with primates has revealed that both monkeys and chimpanzees
tested in laboratory environments and in the field display efficient routes when
traveling to multiple locations (Menzel, 1973; Gallistel & Cramer, 1996;
Cramer & Gallistel, 1997; Janson, 1998, 2007). Notably, the researchers
who have conducted this past work with primates have also examined the
nature of the spatial information that primates may be using as they travel to
multiple locations. These researchers have observed that primates will travel
to clusters of food locations as part of a larger route. As discussed in the intro-
duction, this may be part of a local strategy for which the proximity of the next
location to the current location plays an important role in the formation of the
route. A local proximity-based solution would appear to be less cognitively
demanding, as the traveler would not be required to represent multiple loca-
tions in memory or use cognitive resources in planning ahead. Instead, using
proximity, the traveler could stick with a simple rule to guide its behavior,
which in many cases would lead to an efficient route. Such a proximity-
based strategy may require that the traveler primarily tend to a local and
immediate portion of the environment. However, the results from the studies
with primates have indicated that they may be planning one or more steps
ahead when forming a route. The fact that primates can take efficient routes
during these tests has been used as evidence for cognitive mapping (e.g.,
Gallistel & Cramer, 1996).
In our studies we examined the extent to which pigeons, a bird with a very
different evolutionary trajectory than that of primates, could also take an effi-
cient route to multiple locations. We also were interested in the extent to
which pigeons might have a global consideration of multiple locations and
plan ahead. In our initial work we observed that pigeons and people were more
efficient than a random model of performance when given a computerized
version of a traveling problem. However, both travelers often took routes
that were suboptimal. People tended to use the neighbor solution for a high
proportion of their routes, whereas pigeons did so to a lesser extent.
We hypothesized that the pigeons’ routes may have been less efficient because
the cost of selecting a route was relatively low. We increased the cost of
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 109

selecting an inefficient route during a follow-up study with the pigeons in

which we used a criterion for performance. The pigeons’ routes became pro-
gressively more efficient than the Monte Carlo model and simultaneously
more consistent with the nearest-neighbor model as the criterion was made
increasingly demanding. Thus, the neighbor strategy appeared to be a solution
that was utilized by both species as a mechanism for improving performance.
Whether this implies that both species only had local knowledge of the prob-
lem during the course of a trail remains unclear. Both the people and the
pigeons selected clusters of nodes early during the formation of the route.
Thus, the travelers may have observed the entire display prior to selecting their
route and may have represented the information in some type of cognitive
map (for reviews see Gallistel, 1990; Shettleworth, 2010) that would allow
for forward planning of efficient routes. Alternatively, the tendency of pigeons
to select clusters of nodes might not indicate the use of a map-like representa-
tion for forward planning. The pigeons, and in fact people, may have simply
been “attracted” to large clusters of nodes on the display and continued on
their route after starting with these clusters.
In a second body of work we examined if the skill set required to solve TSPs
on a computer screen would also apply to pigeons solving TSPs in a 3-D envi-
ronment. Additionally, we wanted to explore further whether or not pigeons
were capable of planning one step ahead. As with the operant work, we found
that the pigeons display a preference for choosing feeders (nodes) that are close
to the current location. The results from configurations A and B (and to a
lesser extent configurations C and D) revealed that the pigeons preferred to
travel to a nearby feeder even if it resulted in a longer route in most instances.
In general animals show a preference when given a choice for a small but
immediate reward rather than waiting for a large one (see Stevens & Stephens,
2008, for a review), so perhaps our pigeons were also applying this rule and
maximizing their initial intake of food. Our results from the ambulatory task
indicated that the pigeons might be going beyond using proximity as part of
an efficient solution, however.
We observed evidence from both groups that the pigeons could plan one or
more steps ahead. The birds in the round-trip group showed evidence for
planning during testing with configurations E–G. Likewise, during testing
with configurations E–H we observed that the pigeons in the one-way group
showed strong evidence of planning. Our data suggest that pigeons were likely
looking one or two steps ahead as part of determining their route. The pigeons’
choices early during the trial constrained the efficiency of the route. Thus, birds
likely had to consider the geometry of the remaining two legs of the journey
110 ANIMAL BEHAVIOR

when making their first or second choice along the route. Our work is consis-
tent with that of other researchers suggesting that pigeons may be capable of
planning ahead (Miyata & Fujita, 2006, 2008). For these problems E–H the
additional costs associated with choosing a route based on proximity only
may have facilitated the pigeons planning where to go beyond the next feeder.
Although the pigeons may have planned ahead, the nature of the spatial rep-
resentation that the pigeons may have formed remains elusive. The pigeons’
second choices during testing with configurations E–H indicates that the birds
were likely aware of the locations of the remaining feeders and planned accord-
ingly. If a cognitive map is an understanding of the geometric relationships
between objects in an environment that can be used to determine efficient
and new routes, then the pigeons’ behavior suggests that they may have had
such a representation. Of course, other possible mechanisms that might
account for such performance remain. Michael Brown and Michael Drew
(1998) found that the ability of rats to discriminate between familiar and unfa-
miliar spatial locations improved with perceptual learning. In that study one
group of rats was exposed to the visual environment surrounding an eight-
arm radial maze, whereas another group of rats was exposed to a nearby control
room in which testing was not conducted. Later during testing, both groups of
rats performed a memory task on the same maze in the test environment. Both
groups of rats acquired the task at the same rate and performed at comparable
levels during testing. Notably, the rats that were preexposed to the test environ-
ment were somewhat better at discriminating between arms that had and had
not been previously visited during the memory test. Thus, the effect of expo-
sure and subsequent improvement in discrimination of location may be due
to a perceptual process rather than one of planning.
Another possibility is that the pigeons may have used a general response sol-
ution during Experiment 2 that resulted in behavior that was also consistent
with planning ahead. The pigeons in the round-trip group may have learned
to make a circular pattern after departing S, whereas the pigeons in the one-
way group may have learned to make a Z-like route to get from S to E. This
later explanation seems unlikely as the pigeons did not have extensive experi-
ence with the arrays. Likewise, the insertion of the random configurations
with the test configuration may have disrupted a general type of response from
developing.
As a pigeon moves from one node to the next, particularly for the study in
the experimental room, it is faced with the task of determining what feeders
have been visited and which remain. Robert Cook and colleagues tested rats
on a 12-arm radial-maze memory task and found that the rats likely used a
dual-code memory process (Cook et al., 1985). Early during the trial, when
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 111

only a few of the arms had been selected, the rats appeared to be using retro-
spective information, but as the trial progressed prospective memory processes
appeared to play a stronger role. It is unclear if the pigeons in the current study
were similarly using a dual-code process, but it would not be an unreasonable
expectation. Future work that looks at when the pigeons encode the feeders
into memory and whether or not they are using both prospective and retro-
spective memory processes would be useful.
To conclude, we have discovered that pigeons, like primates, can generate
efficient routes to travel to multiple destinations. Like primates, the pigeons
appear to employ multiple solutions to finding an efficient route. The proximity
of a node relative to the pigeons’ current location had a large impact on the
pigeons’ route choices, particularly when the initial gain was large and the detri-
ment to overall route efficiency small. Pigeons, like primates, appear capable of
planning beyond the next step when necessary. The nature of the geometric
knowledge used to make such decisions is consistent with a cognitive map; how-
ever, other possible mechanisms cannot be ruled out. Future research using TSPs
will be important and useful to delineate the nature of the spatial cognitive infor-
mation that pigeons and other animals used to travel to multiple destinations.

ACKNOWLEDGMENTS
I would like to acknowledge all of the undergraduate students who have
worked in my laboratory over the years. Your dedication and commitment
to exploration and research has been inspiring.

REFERENCES AND SUGGESTED READING

Bennett, A. T. D. (1996). Do animals have cognitive maps? Journal of Experimental
Biology, 199, 219–224.
Blaser, R. E. & R. R. Ginchansky. (2012). Route selection by rats and humans in a
navigational traveling salesman problem. Animal Cognition, 15, 239–250.
Brown, M. F. & M. R. Drew. (1998). Exposure to spatial cues facilitates visual dis-
crimination but not spatial guidance. Learning and Motivation, 29, 367–382.
Brown, M. F., P. A. Rish, J. E.VonCulin, & J. A. Edberg. (1993). Spatial guidance of
choice behavior in the radial-arm maze. Journal of Experimental Psychology:
Animal Behavior Processes, 19, 195–214.
Cook, R. G., M. F. Brown, & D. A. Riley. (1985). Flexible memory processing by
rats: Use of prospective and retrospective information in the radial maze. Jour-
nal of Experimental Psychology: Animal Behavior Processes, 11, 453–469.
Cramer, A. E. & C. R. Gallistel. (1997). Vervet monkey as traveling salesman.
Nature, 387, 464–464.
112 ANIMAL BEHAVIOR

Friedman, S. L. & E. K. Scholnick (eds.). (1997). The Developmental Psychology of

Planning: Why, How, and When Do We Plan? Mahwah, NJ: Erlbaum.
Gallistel, C. R. (1990). The Organization of Learning. Cambridge, MA: MIT Press.
Gallistel, C. R. & A. E. Cramer. (1996). Computations on metric maps in mammals:
Getting oriented and choosing a multi-destination route. Journal of Experimen-
tal Biology, 199, 211–217.
Gibson, B. M. & A. C. Kamil. (2001). Tests for cognitive mapping in Clark’s nut-
crackers (Nucifraga columbiana). Journal of Comparative Psychology, 115,
403–417.
Gibson, B. M., E. A. Wasserman, & A. C. Kamil. (2007). Pigeons and people select
efficient routes when solving a one-way “traveling salesperson” task. Journal
of Experimental Psychology: Animal Behavior Processes, 33, 244–261.
Hamilton, D. A., K. G. Akers, M. P. Weisend, & R. J. Sutherland. (2007). How do
room and apparatus cues control navigation in the Morris water task? Evidence
for distinct contributions to a movement vector. Journal of Experimental Psy-
chology: Animal Behavior Processes, 33, 100–114.
Janson, C. H. (1998). Experimental evidence for spatial memory in foraging wild
capuchin monkeys, Cebus paella. Animal Behaviour, 55, 1229–1243.
Janson, C. H. (2007). Experimental evidence for route integration and strategic plan-
ning in wild capuchin monkeys. Animal Cognition, 10, 341–356.
Kacelnik, A. (1984). Central place foraging in starlings (Sturnus vulgaris), I. Patch res-
idence time. Journal of Animal Ecology, 53, 283–299.
Kacelnik, A. & I. C. Cuthill. (1987). Starlings and optimal foraging theory: Modeling
in a fractal world. In A. C. Kamil, J. R. Krebs, & H. R. Pulliam (eds.), Foraging
Behavior (pp. 303–333). New York: Plenum Press.
Kong, X. & C. D. Schunn. (2007). Global vs. local information processing in visual/
spatial problem solving: The case of traveling salesperson problem. Cognitive
Systems Research, 8, 192–207.
MacDonald, S. E. (1994). Gorillas’ (Gorilla gorilla gorilla) spatial memory in a forag-
ing task. Journal of Comparative Psychology, 108, 107–113.
MacDonald, S. E. & D. Wilkie. (1990). Yellow-nosed monkeys’ (Cercopithecus asca-
nius whitesidei) spatial memory in a simulated foraging environment. Journal
of Comparative Psychology, 104, 382–397.
MacGregor, J. N. & T. C. Ormerod. (1996). Human performance on the traveling
salesman problem. Perception and Psychophysics, 58, 527–539.
MacGregor, J. N. & T. C. Ormerod. (2000). Evaluating the importance of the con-
vex hull in solving the Euclidean version of the traveling salesperson problem:
Reply to Lee and Vickers (2000). Perception and Psychophysics, 62, 1501–1503.
MacGregor, J. N., T. C. Ormerod, & E. P. Chronicle. (1999). Spatial and contextual
factors in human performance on the traveling salesperson problem. Percep-
tion, 28, 1417–1427.
MacGregor, J. N., T. C. Ormerod, & E. P. Chronicle. (2000). A model of human
performance on the traveling salesperson problem. Memory and Cognition,
28, 1183–1190.
BEING DIRECT : CAN PIGEONS PLAN AHEAD ? 113

Menzel, E. W. (1973). Chimpanzee spatial memory organization. Science, 182,

943–945.
Miyata, H. & K. Fujita. (2008). Pigeons (Columba livia) plan future moves on com-
puterized maze tasks. Animal Cognition, 11, 505–516.
Miyata, H. & K. Fujita. (2010). Route selection by pigeons (Columba livia) in “trav-
eling salesperson” navigation tasks presented on an LCD screen. Journal of
Comparative Psychology, 124, 433–446.
Miyata, H., T. Ushitani, I. Adachi, & K. Fujita. (2006). Performance of pigeons
(Columba livia) on maze problems presented on the LCD screen: In search
for preplanning ability in an avian species. Journal of Comparative Psychology,
120, 358–366.
O’Keefe, J. & L. Nadel. (1978). The Hippocampus as a Cognitive Map. Oxford, UK:
Oxford University Press.
Shettleworth, S. J. (2010). Cognition, Evolution, and Behavior. Second Edition. New
York: Oxford University Press.
Stevens, J. R. & D. W. Stephens. (2008). Patience. Current Biology, 18, R11–12.
Tolman, E. C. (1948). Cognitive maps in rats and men. Psychological Review, 55,
189–208.
Vickers, D., M. Butavicius, M. Lee, & A. Medvedev. (2001). Human performance
on visually presented traveling salesman problems. Psychological Research/
Psychologische Forschung 65, 34–45.
Vickers, D., M. D. Lee, & M. Dry. (2006). The aesthetic appeal of minimal struc-
tures: Judging the attractiveness of solutions to traveling salesperson problems.
Perception and Psychophysics, 68, 32–42.
Vickers, D., M. D. Lee, M. Dry, & P. Hughes. (2003). The roles of the convex hull
and the number of potential intersections in performance on visually presented
traveling salesperson problems. Memory and Cognition, 31, 1094–1104.
This page intentionally left blank
 4

Social Learning, Tradition, and Culture:

Data and Debate
Bennett G. Galef

INTRODUCTION
The last three decades have seen a remarkable increase in interest in the pos-
sibility that animals’ acquisition of adaptive patterns of behavior can be facili-
tated by their observing the behavior of others. Such social learning, as it is
often called, is currently seen as playing an important role in animals as diverse
as ants and chimpanzees learning almost anything from how to choose a mate
to how to handle potentially dangerous prey.
Despite the ever-expanding knowledge of the ways in which social learning
affects the development of adaptive behaviors, important questions remain
concerning the mechanisms, functions, and evolution of social learning. Con-
sequently, in this chapter, I not only review a sample of instances in which
social interactions have been clearly shown to influence the ways in which ani-
mals come to perform biologically important behaviors but also discuss several
aspects of social learning that remain controversial.

EXAMPLES OF SOCIAL LEARNING

It is perhaps best to begin with a few of the many possible examples of phe-
nomena that those who study social learning in animals have examined in
detail. More thorough discussions of both the facts of social learning and
116 ANIMAL BEHAVIOR

controversial issues in the area are available in a number of books and special
issues of journals devoted to studies of social learning (e.g., Zentall & Galef,
1988; Heyes & Galef, 1996; Fragaszy & Perry, 2003; Galef & Heyes, 2004;
Laland & Galef, 2009; Kendal et al., 2010; Whiten et al., 2011; Nielsen
et al., 2012), as well as in innumerable refereed articles in scientific journals,
a selection of which are cited in the reference section at the end of this chapter.

Learning How to Court

Brown-headed cowbirds (Molothus ater) are particularly interesting subjects
for studies of social influence on behavioral development because cowbirds are
obligate brood parasites. Adult cowbirds always lay their eggs in nests con-
structed by birds of other species and leave rearing of their young to such
foster parents. As a result, early in life, young cowbirds have no opportunity
to interact with members of their own species.
In many other species of songbird (see Volume 3, Chapter 2), exposure to
the songs of conspecific adult males during a sensitive period (see Chapter
9) early in life is necessary if young males are to develop songs that will be
attractive to females of their species. Because during this sensitive period in
development young male cowbirds do not hear adult males of their species
sing, it was long thought that courtship songs of male cowbirds must develop
independently of social experience. However, the results of numerous studies
by Meredith West and Andrew King and their students (for review, see Free-
berg, 2004), have shown that, as in other species of songbird, social learning
plays a central role in young males developing songs that will be effective in
courting females.
Male cowbirds from different areas of North America sing regional dialects,
each a variant of a prototypic cowbird song. These regional dialects are differ-
entially effective in causing females to assume a copulatory posture, with
females from each region responding preferentially to the songs of males from
that region.
Outside the breeding season, cowbirds live in mixed-sex flocks, and while
in these flocks each young male cowbird sings a number of different song
variants, ordinarily without immediately repeating any particular song. Female
cowbirds respond with a distinctive “wing-stroke” display only to male song
variants that will be most effective in causing them to accept a male as a mate
during the breeding season. After a female responds to one of a male’s songs
with a wing-stroke display, the male repeats that song three or four times in
succession and increases the frequency with which he subsequently sings it
(West & King, 1996). Thus, female cowbirds shape the song displays of males
of their species.
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 117

The songs of male cowbirds that are most effective in eliciting positive
responses from females are also the songs that most frequently elicit attacks
by other male cowbirds, causing submissive males to abandon singing highly
effective songs. As a result, only the dominant males in a group can continue
to produce song variants that have a high probability of eliciting positive
responses from females, and females therefore have a high probability of
mating with socially dominant males.

Learning Where, What, and How to Eat

Perhaps the most thoroughly investigated instances of social learning are
those influencing development of various aspects of foraging (e.g., choice of
foods and feeding sites, methods of exploiting heavily defended foods, etc.).
Consequently, examples of social influences on foraging behaviors—ranging
from the development of flavor preferences in Norway rats (Rattus norvegicus)
to the methods that chimpanzees (Pan troglodytes) use to feed on army ants—
are scattered throughout this chapter.
The relative frequency of studies of social influences on various aspects of
foraging behavior reflects both the considerable time and energy that members
of many species spend acquiring food and the resultant ease with which feed-
ing behaviors can be studied in both captive and free-living animals. Further,
because the feeding behavior of wild rodents has a substantial impact on
human success in producing food, and because many rodents (e.g., rats, mice,
hamsters, voles, etc.) thrive in captivity (where behavior can be most easily
studied under controlled conditions), progress in analysis of social influences
on the foraging behavior of rodents (and of honey bees, which share both
characteristics [see Chapter 12; von Frisch, 1967]) has progressed relatively
rapidly. Here, the focus is on social influences on the feeding of rodents,
though a similar review could be provided with respect to social effects on
foraging in avian species (e.g., Galef & Giraldeau, 2001).
Social influences on foraging in rats—More than a half-century ago, von Fritz
Steiniger (1950), an applied ecologist, tried various ways to decrease the cost
of controlling pest populations of Norway rats. One tactic Steiniger evaluated
involved introducing large portions of poison bait into a rat-infested area, thus
avoiding some of the expense associated with the usual procedure of constantly
replenishing small portions of bait as rats consumed them. Much to his dis-
may, Steiniger found that despite initial reduction in rat numbers shortly after
he introduced a large portion of bait into a rat-infested area, the target popu-
lation soon completely stopped eating the poisoned bait and eventually
returned to its initial size. The reasons for this failure to provide long-term
reduction in rat numbers proved interesting.
118 ANIMAL BEHAVIOR

First and most obviously, although most rats in a target population ingested
a lethal dose of poison when it was first introduced, a few ate less than a lethal
amount of bait, became ill, and learned, as a result of that single experience
with the poisoned bait, to avoid further ingestion of it (Garcia & Koelling,
1966). Second, and far more discouraging to Steiniger, young rats born to sur-
vivors of initial contact with a poisoned bait totally rejected that bait without
ever even tasting it. Bait avoidance learned by surviving rats appeared to be
somehow transmitted to their young.
Mertice Clark and I (Galef & Clark, 1971a) brought adult wild rats
captured on garbage dumps into the laboratory, established the rats in small
colonies, and offered them two foods, one of which we had contaminated with
a mild toxin. As expected, the captive rats rapidly learned to eat only the
untainted food and continued for weeks to avoid the previously tainted food
even when provided with untainted samples of it. Further, as Steiniger had
found, young subsequently born to such trained colonies refused to even taste
the previously tainted food that the adults of their colony had learned to avoid
(Galef & Clark, 1971a).
Because the phenomenon of social transmission of bait avoidance could be
studied under controlled conditions, investigation of social-learning processes
leading to transmission of food choices from adult rats to their offspring
became possible, and several different types of interaction between young
and adult rats were found to result in transmission of learned food preferences
from adult rats to their offspring (for review, see Galef, 1985).
Selecting a place to eat—Young wild rats emerging from their nest site to
take their first meals of solid food use visual cues to find adults at a distance
from the nest (blind rat pups do not show the effect) and approach them
(Galef & Clark, 1971b). In fact, placing an anesthetized adult rat in the vicin-
ity of one of two otherwise identical feeding sites leads pups to both visit and
feed at the site close to the anesthetized rat far more frequently than at the
alternative (Galef, 1981).
Indeed, adult rats do not even have to be present at a feeding site to increase
its attractiveness to their young. As adult rats leave a feeding site and return to
their nest, they deposit scent trails that young rats follow when seeking food
(Galef & Buckley, 1996). Furthermore, while feeding, rats defecate and uri-
nate in the area where they eat, and these residual cues, like the actual presence
of an adult at a feeding site, attract pups and cause them to eat in locations that
adults have exploited (Galef, 1985; Laland & Plotkin, 1990).
Of course, in natural circumstances foods are often patchily distributed.
Consequently, learning to eat in a location where adults have eaten often leads
to eating a food that adults have eaten.
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 119

Selecting a food to eat—Even before birth, young rats can obtain informa-
tion from their mother concerning at least some of the foods that she has eaten
while gestating them. Peter Hepper (1988) fed garlic to pregnant rats and
then, shortly after garlic-fed mothers delivered their litters, transferred the
young to foster mothers that had never eaten garlic. When subsequently
offered a choice between dishes containing garlic and onion, pups that had
been gestated by a mother eating garlic spent more time near the dish contain-
ing garlic than pups gestated by mothers not exposed to garlic. Similarly,
Smotherman (1982) found that injecting lemon flavor into the amniotic fluid
of a rat mother resulted in her offspring, when adult, increasing their prefer-
ence for lemon-flavored food.
The flavor of foods that a dam eats affects not only the flavor of her amni-
otic fluid but also the flavor of her milk. At weaning, rat pups prefer foods
with flavors that they have experienced in their mother’s milk (for review,
see Galef, 1985).
My students and I have also found that after a naïve young rat (an observer
rat) interacts for a few minutes with an adult rat that has recently eaten a distinc-
tively flavored food (a demonstrator), the observer rat shows a markedly
enhanced preference for whatever food its demonstrator ate (for review, see
Galef, 1988, 1996). Similar effects have been found in other mammalian spe-
cies ranging from bats (Ratcliffe & ter Hofstede, 2005) to hyenas (Yoerg, 1991).
Analyses of the processes underlying this last type of social induction of fla-
vor preference have repeatedly shown that, if an observer rat is to develop a
preference for the food that its demonstrator ate, the observer rat has to expe-
rience the odor of a food at the same time that it experiences the breath of a
demonstrator rat. Simple exposure to a food odor in isolation does not pro-
duce the effect.
Chemical analysis of rat breath reveals significant concentrations of carbon
disulfide (CS2), and experiments show that experience of a food odor together
with CS2 results in enhanced preference for the food almost as great as experi-
ence with a demonstrator rat that has eaten that food (Galef, 1988). Further,
recent electrophysiological studies of the olfactory system of mice have
revealed a specialized olfactory subsystem expressing the receptor guanylyl
cyclase that is highly sensitive to CS2. Gene-targeted mice with disruptions
of the transduction cascade that transforms receptor response to CS2 into elec-
trical impulses in the nervous system not only lose sensitivity to CS2 but also
fail to show social influences on their food choices (Munger et al., 2010).
Thus, there appears to be a specialized portion of the olfactory system in mice
(and presumably other rodent species as well) dedicated to social learning of
food preferences.
120 ANIMAL BEHAVIOR

Learning How to Eat

There are no squirrels in the pine forests of either Israel or Corsica, and
consequently there are no specialized mammals present in these forests eating
pine seeds, the sole food suitable for mammals found there. In both Israel and
Corsica, and nowhere else, roof rats (Rattus rattus) have learned to eat pine
seeds and have established colonies that have thrived in pine forests for many
generations (Terkel, 1996).
Laboratory studies of the feeding behavior of rats captured in pine forests in
Israel reveal that in order for rats to gain more energy from eating pine seeds
than they expend in removing the tough scales that protect the seeds from
potential predators, the rats must take advantage of the physical structure of
the pinecones, first removing scales from the base of a cone and then removing
in succession the spiral of scales circling a cone’s shaft from base to apex.
Almost all young rats reared by either their natural mother or a foster mother
who efficiently stripped the scales from pinecones while young were present
acquire the efficient technique. To the contrary, young rats reared by either
their natural mother or a foster mother that does not know how to open pine-
cones efficiently almost never learn to open pinecones efficiently for themselves.
The failure of young rats reared by dams that do not strip cones efficiently to
learn to open pinecones efficiently and the success of young rats reared by dams
that do strongly suggest that the behavior of mothers somehow facilitates young
rats learning to strip pinecones of their seeds. Indeed, further studies showed
that the experience of young rats completing the stripping of scales from cones
properly started by an adult rat allowed most young rats to acquire the efficient
method, even if the young were reared by a female rat that did not know how to
open cones efficiently. Most informative, a majority of young rats reared by a
dam who did not strip cones efficiently and given access to cones started by a
human experimenter (who had used a pair of pliers to remove a few scales from
the base of cones) also learned the efficient technique (Terkel, 1996).
The type of social learning involved in learning to strip pinecones of their
seeds is not particularly sophisticated. It involves nothing more than directing
the attention of young rats to the base of cones. Nonetheless, such simple
social learning is sufficient to establish a tradition of pinecone opening in roof
rats that allows them to thrive in a portion of the environment that would oth-
erwise be closed to them.

Learning to Attack “Toxic” Prey

Food choices can be socially influenced by exposure not only to olfactory
and visual cues but to acoustic cues as well. The group nesting, socially
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 121

foraging, neotropical fringe-lipped bat (Trachops cirrhoses) is unique among

bats in using the calls that male frogs use to attract mates to locate potential
food (Tuttle & Ryan, 1981). Most other species of bat use their sonar to
detect prey rather than using the sounds made by prey to locate and capture
them.
In the wild, fringe-lipped bats prey frequently on túngara frogs (Physalae-
mus pustulosus) and avoid contact with poisonous cane toads (Bufo marinus).
When brought into the laboratory for study, recently captured, hungry
fringe-lipped bats readily approach a concealed loudspeaker playing the mat-
ing calls of túngara frogs but avoid one playing the easily distinguished mating
calls of cane toads.
Rachel Page and Michael Ryan (2006) captured fringe-lipped bats and
exposed them to a series of trials in which the bats were rewarded with food
when they approached a concealed loudspeaker playing an artificial call made
by electronically combining a cane-toad call with that of a túngara frog. Over
trials, Page and Ryan (2006) gradually decreased the loudness of the frog-call
portion of the compound auditory stimulus while gradually increasing the
loudness of its toad-call portion until hungry bats would fly for food to a loud-
speaker broadcasting pure toad calls.
Once a bat had been trained to approach cane-toad calls, it was placed
together with a second fringe-lipped bat that had been captured and then
tested for its response to a loudspeaker broadcasting cane-toad calls (none
responded positively). This second bat was then allowed to observe the trained
bat fly to and acquire food at the concealed speaker playing cane-toad calls.
Observer bats quickly began to fly to loudspeakers playing cane-toad calls
and could subsequently be used as demonstrators for other wild-caught bats
that initially avoided cane-toad calls. It is easy to imagine that such social
learning could help fringe-lipped bats track the rapid seasonal changes in the
availability of various prey species of frog typical of neotropical rain forest.

Learning Where to Nest

Information contained in the behavior of knowledgeable individuals is
likely to be of greater value than information obtained by observing the less
well informed (Laland & Galef, 2009). Consequently, in some circumstances,
natural selection seems to have acted to predispose animals to learn socially
from members of species other than their own.
Blue tits (Parus caeruleus) and great tits (Parus major)—both close relatives
of North American chickadees—as well as pied flycatchers (Ficedula hypoleuca)
living in northern Europe nest in holes in tree trunks (or in suitable nest boxes
provided by humans). During the nesting season, all three species forage on
122 ANIMAL BEHAVIOR

similar foods, fall prey to similar predators, and utilize similar nesting sites.
The tits are year-round residents, but the flycatchers come to Europe only to
breed in the spring and summer. Consequently, the tits are likely to be well
informed as to local conditions, and migrant flycatchers might obtain useful
information from the tits when selecting locations in which to nest and rear
their young.
During the winter, Janne-Tuomas Seppanen and Jukka Forsman (2007)
provided great and blue tits in eight locations with an excess number of nest
boxes. Further, before the flycatchers arrived in the spring, the experimenters
decorated all the nest boxes in each area in which the tits had chosen to nest
with either a circle or square so that it would appear to arriving migrant fly-
catchers that all tits in an area had chosen nest sites marked with one symbol
or the other. Also, before the flycatchers arrived, the experimenters placed a sec-
ond nest box a few meters from each nest box the tits were occupying and
marked the second nest box with the opposite symbol so that it would look
to arriving flycatchers as though tits were avoiding nest sites with that mark.
When the first flycatchers arrived on the study sites in spring, the experi-
menters placed pairs of empty nest boxes, one marked with a triangle and
one marked with a circle, in each of the eight study sites so that arriving fly-
catchers could choose between them. The flycatchers clearly preferred nest
boxes of the same type apparently preferred by tits in their area and by late
in the mating season were three times as likely to nest in this type of nest
box as in the other. Male flycatchers were seen to visit tits’ nests early in the
breeding season, and information gathered during these visits may have played
a role in providing the flycatchers with information (Forsman & Thomson,
2008).

Learning What to Fear

Recognizing and responding appropriately to potential predators is a chal-
lenge many birds and mammals face, and given the range of predators that
prey especially on smaller species, innate recognition of each potentially preda-
tory species is not likely. In vertebrates ranging from European blackbirds
(Turdus merula; Curio, 1988) to Japanese macaques (Macaca fuscata; Mineka
& Cook, 1988), naïve individuals learn to respond to potential predators by
attending to the behavior of knowledgeable individuals (reviewed in Griffin,
2004).
Although we tend to think of flight or concealment as the most appropriate
response to the appearance of a potential predator, many species of bird
respond to potential predators with a behavior known as mobbing. When
mobbing, an individual discovering a predator produces a special vocalization
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 123

(a mobbing call), and begins to behave aggressively toward (mob) the preda-
tor. Other birds approach and join in (hence the term mobbing), both giving
mobbing calls and attacking the predator. Owls frequently prey on European
blackbirds, and when they see an owl, wild-caught, adult blackbirds give mob-
bing calls and harass the potential predator.
Using a simple but ingenious apparatus in which demonstrator and
observer blackbirds were separated by an opaque barrier, Curio (1988) allowed
an experienced blackbird to look at a stuffed owl and direct mobbing calls
towards it while a naïve blackbird looked at a stuffed, harmless (nonpredatory)
bird. In response to the mobbing vocalizations of the experienced bird looking
at the owl, the naïve bird started to give mobbing vocalizations of its own and
to act aggressively, but towards the harmless object at which it was looking,
not towards the owl that was invisible to it.
Days later, when the previously naïve bird was again shown the stuffed,
harmless bird, it gave mobbing vocalizations and attacked it. Indeed, this pre-
viously naïve bird had learned so well that it could now serve as a model for a
new naïve bird, inducing it also to mob the harmless object. Curio was thus
able to establish a tradition of mobbing a harmless stuffed bird in a captive
population of blackbirds.
Recent field studies by Heather Cornell and colleagues (2011) provide evi-
dence of similar social learning of mobbing responses both within and
between generations of American crows (Corvus brachyrhyncos) living outside
the laboratory. Individual marked crows that had been trapped, banded,
and released by an experimenter wearing a distinctive mask gave mobbing
vocalizations to the masked figure and attracted other crows. Both other adult
crows and offspring of the banded crows subsequently mobbed any human
wearing the distinctive mask while not mobbing individuals wearing a
different mask.

Learning with Whom to Mate

Sampling among potential mates to decide which of many potential part-
ners would make the best parent for one’s offspring can be an expensive
undertaking, both consuming time and energy that could be devoted to other
activities and increasing the risk of predation on individuals focused on evalu-
ating mates. Information that might reduce such costs can be valuable.
Lee Dugatkin (1996) was the first of several investigators to examine the
possibility that animals, especially those that are young and inexperienced in
such matters, might copy the mate choices of others, thus both reducing the
potential cost associated with identifying an appropriate mate and taking ad-
vantage of experienced individuals’ possibly superior knowledge. If you copy
124 ANIMAL BEHAVIOR

someone else’s choice of a partner, you will do no worse than they did while
avoiding any potential costs of evaluating potential mates.
Dugatkin and colleagues worked with laboratory-born descendants of wild
guppies (Poecilia reticulata), a small freshwater fish familiar to all aquarium
keepers. Guppies were an appropriate choice for such experiments because in
their natural habitat (streams in Trinidad, where Dugatkin captured them)
guppies select partners and breed under conditions that would allow them to
observe and copy one another’s mate choices.
Dugatkin conducted his experiments with guppies in a simple apparatus
consisting of a large aquarium placed between two smaller aquaria. He posi-
tioned a male “target” guppy in each of the small aquaria and a “focal” female
guppy, restrained in a transparent tube, in the middle of the central aquarium.
He then introduced a second female (a “model” female) into one of the two
smaller aquaria but separated from the target male by a transparent barrier.
The focal female, constrained in the transparent tube in the central aquar-
ium, then watched while one of her target males courted the model female
and the other remained alone. Dugatkin then removed the model female from
the apparatus, released the focal female from the tube constraining her, and
allowed her to choose between her two target males. Seventeen of 20 focal
females tested in this way spent more time near the target male they had seen
courting a model female than near the target male that had remained alone.
Although this result is surely consistent with the hypothesis that female
guppies copy one another’s choices of partner, it is also possible that female
guppies, members of a species that gathers in shoals in nature, simply prefer
a place where they have previously seen two fish to a place where they have
seen only one fish or that the behavior of a male who has recently courted is
different from and more attractive to females than that of a male that has
not, and so forth. Experiments to test such alternative explanations are easy
to carry out, and their outcomes allowed Dugatkin to exclude many such
alternative explanations of the females’ preferences for males they had seen
courting other females. Female guppies, like females of some other species of
fish, as well as of some avian species, prefer to mate with individuals that they
have seen courting others (for review, see White, 2004).
Such laboratory evidence of a tendency to copy the mate choices of conspe-
cifics suggests an explanation for some unusual behavior seen in species (e.g.,
sage grouse [Centrocercus urophasianus] or prairie chickens [Tympanuchus
cupido]) in which males compete for females on communal breeding grounds
(leks) that females visit for no purpose other than to choose a male and mate.
Often, only a small percentage of the males on any lek get to mate, and these
fortunate males each mate with many female visitors.
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 125

Human investigators have been unable to identify characteristics that dif-

ferentiate successful from unsuccessful males on a lek. However, if females in
lekking species copy one another`s mate choices, as laboratory findings of
mate-choice copying in other avian species suggest they may (Galef, 2009),
then the extreme skew in female choice of male partners is not difficult to
understand. Indeed, the observed sequence of mating behavior on sage-
grouse leks is consistent with the hypothesis that female grouse observe and
copy one another’s mate choices, thus producing the observed inequality in
male success (Gibson et al., 1991).

DEFINING TYPES OF SOCIAL LEARNING: THE ANALYSIS OF BEHAVIOR

The preceding examples provide evidence of effects of socially acquired
information on the behavior of many species in many circumstances. How-
ever, as a moment’s reflection about such examples makes clear, social learning
can occur in quite different ways. There have been many attempts to both cat-
egorize the behavioral processes involved in social learning (for examples, see
Galef, 1988; Whiten & Ham, 1992; Heyes, 1994; Zentall, 1996) and to
determine the type of social learning processes responsible for particular
instances of social influence on behavior.
Sometimes an animal actively shapes the behavior of another (for example,
when female cowbirds give a wing stroke in response to a male’s song). As illus-
trated below, sometimes social learning appears to involve true imitation (learn-
ing to do a motor act from seeing another perform that act; Thorndike, 1898).
In other cases, social learning appears to rest on emulation—learning by watch-
ing others that a goal can be achieved without learning very much about the
behavior a knowledgeable individual used to achieve the goal (Tomasello, 1994).
As we have seen in many of the examples discussed above (for example,
when rats deposit olfactory cues at a feeding site), local enhancement (“appar-
ent imitation resulting from directing the animal’s attention to a particular
object or part of the environment” [Thorpe, 1963, p. 134]) often suffices to
increase the probability that a naïve animal will come to exhibit the same
behavior as others of its social group.
Why should we make the effort to distinguish among processes involved in social
learning? It can be argued that all social learning processes serve similar func-
tions, for example, permitting animals to learn adaptive behaviors without
incurring whatever costs are involved in learning for oneself by trial and error.
Fine-grain distinctions differentiating local enhancement from emulation and
emulation from imitation may seem an unnecessary burden. However, there
are contexts in which such distinctions have proven useful.
126 ANIMAL BEHAVIOR

One of the important questions faced by those interested in social learning

is why social learning has resulted in impressive cumulative cultures in the
human species but not in any other. All agree both that nonhuman animals
exhibit socially learned behavioral traditions and that these traditions never
approach the level of sophistication and complexity of the cultures seen in all
human social groups. Is it because the processes supporting social learning in
humans differ from those seen in other animals?
As discussed more fully in the section of this chapter concerned with animal
culture, experts have both asserted (e.g., Galef, 2009; Tomasello, 2009) and
denied (e.g., Heyes, 1993) that an ability to imitate or teach is critical for pro-
duction of the kind of cumulative culture typical of human societies. How-
ever, if cumulative culture does depend on an ability to teach or imitate, the
question of whether animals in general, and chimpanzees and orangutans in
particular, engage in teaching or imitation becomes central to our understand-
ing of the obvious difference in the extent and complexity of the traditions
exhibited by our own species and by other animals, even our closest relatives.
The only distinctions used in the remainder of this chapter involve deter-
mining, first, whether, in any particular instance, a naïve individual (1)
acquires by observation a pattern of behavior exhibited by another (imitates);
(2) has its attention directed towards important aspects of the environment
or environmental affordances by the behavior of others (for example, as a
result of local enhancement or emulation) and then learns for itself how to
achieve a desired outcome; or (3) whether knowledgeable individuals modify
their behavior in the presence of naïve individuals and thus facilitate acquisi-
tion of behavior by pupils (teaching). In cases of imitation, emulation, and
local enhancement, naïve individuals extract useful information as a result of
observing the normal behavior of accomplished individuals, whereas in teach-
ing, the knowledgeable individual behaves differently in the presence than in
the absence of a naïve individual, thus both facilitating learning by pupils
and incurring some cost from teaching (Caro & Hauser, 1992).
The distinctions being made may be clarified by a hypothetical example.
Consider an adolescent chimpanzee watching its mother as she uses her left
hand to lift a log and then takes insect larvae from the log’s underside with
her right hand. The juvenile chimp might learn from watching its mother
(1) to investigate the log (local enhancement), (2) to look for food by turning
over logs (emulation), or (3) to put its left hand under a log, move that hand
upward, and pick up food with its right hand (imitation). If the mother chimp
were to slow or exaggerate her movements when her infant was present (thus
sacrificing efficiency), allowing her “pupil” to learn to find grubs under logs
sooner than it would have by simply watching its mother hunt for grubs in
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 127

the normal way, then teaching would be involved. Specific, real examples of
the more complex forms of social learning (teaching, imitation, emulation)
are discussed in the next few paragraphs.

Teaching
As mentioned above, teaching differs from other types of social learning in
that, when teaching, a teacher suffers some short-term cost (e.g., a loss of effi-
ciency or delay or loss of reward) with the result that a pupil acquires some
behavior more rapidly than it would have if it weren’t taught (Caro & Hauser,
1992). In most social-learning processes involving a knowledgeable demon-
strator and naïve observer, it is the behavior of the observer that changes.
In teaching both the demonstrator’s and the observer’s behaviors change
(Hoppitt et al., 2008).
There has been some question as to the importance of teaching even in
human preindustrial societies (Laland & Hoppitt, 2003), and, of course, it is
inappropriate in the absence of careful studies to assume that we somehow
know how hunter-gatherers acquire skills. Perhaps surprisingly, less is known
of the role of social learning in the acquisition of life skills by humans than
by many other animals, and formal, language-based instruction of the kind
typical in industrial societies seems to be rare in the preindustrial world. Still,
recent anthropological observations suggest that teaching, as defined in the
preceding paragraph (Caro & Hauser, 1992), plays an important role in main-
tenance of the complex behavioral traditions that are characteristic of human
populations in pre- as well as postindustrial societies (e.g., Ruddle & Chester-
field, 1997; Diamond, 2001).
Although a handful of possible instances of teaching by apes and cetaceans
have been provided in the literature, many find these examples relatively
unconvincing (for review, see the discussion of Rendell & Whitehead,
2001). Perhaps most interesting, with the exception of two instances of pos-
sible of adult chimpanzees teaching their young to crack nuts (Boesch,
1991), there have been no reports of teaching in free-living chimpanzees or
bonobos (Pan paniscus) despite tens of thousands of hours of observation.
Consequently, recent reports of teaching in animals with much less sophisti-
cated nervous systems than those of great apes, whales, or porpoises, far
exceeding the criteria for teaching proposed by Caro and Hauser (1992), came
as something of a surprise (for review, see Thornton & Raihani, 2008).
Perhaps the most compelling case of teaching in a nonhuman species
involves wild meerkats (Suricata suricatta). Young meerkats are largely depen-
dent on the adult members of their colony for sustenance, although by the
128 ANIMAL BEHAVIOR

time young are three months of age, they can forage for themselves, handling a
variety of prey, including potentially dangerous scorpions.
Observations of free-living meerkats, together with experimental interven-
tions in wild populations, show that the transition to nutritional indepen-
dence is facilitated by adults teaching the young necessary skills (Thornton
& McAuliffe, 2006). Normally, adults consume whatever prey they capture.
However, when in the presence of begging young, adults often kill or disable
prey (scorpions are disabled by removing their stingers), then carry the food
to juveniles.
As pups become older and more experienced at handling scorpions, adults
give the young intact scorpions increasingly often. Apparently, adults judge
the ability of young to handle scorpions from the maturity of their begging
calls, and playing recordings of begging calls of older pups to adult meerkats
providing scorpions to younger individuals causes the adults to bring intact
scorpions, even though the young being fed are not yet ready to handle such
dangerous items. Conversely, playing recordings of the calls of young pups
to adults provisioning older individuals causes adults to increase the frequency
with which they deliver disabled scorpions to older pups.
Further, as the Caro and Hauser (1992) definition of teaching requires, the
provisioning strategy of adult meerkats accelerates the acquisition of scorpion-
handling skills by juveniles. Pups given disabled scorpions by a human experi-
menter learned to handle intact scorpions effectively at a younger age than
pups that had been artificially provisioned with either dead scorpions or
hard-boiled eggs.
Lest you think that teaching is confined to big-brained mammals, it is
worth looking closely at such phenomena as the waggle dance of the honey
bee (see Chapter 12) and tandem running in ants, where successful foragers
guide their nest mates to newly discovered food sources (Franks & Richard-
son, 2006). When in the company of a naïve nest mate, an ant of the species
Temnothorax albipennis that knows the location of a food source moves slowly
toward it, travelling in spurts and pauses. During such tandem runs to food,
the leader slows its pace, waiting while the follower looks around, apparently
to examine landmarks, and the leader moves rapidly toward food only after
its follower taps it with its antennae. The result of this unusual behavior is that
a leader ant sacrifices efficiency in returning to food (an ant acting as leader of
a nest mate takes four times as long to travel to food as it would if it were trav-
elling to food alone), while followers show clear benefits from following, both
finding food far faster when following a leader than when searching on their
own and returning more directly to their nest than did their leader after its ini-
tial discovery of food.
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 129

Imitation
Interest in the possibility that animals are able to imitate, defined as learn-
ing to do an act from seeing it done (Thorndike, 1898), dates to the late nine-
teenth century when George Romanes (1884), a disciple of Charles Darwin,
provided anecdotal observations consistent with the view, widely held at the
time, that animals could acquire complex patterns of behavior by observing
and then imitating the activities of either humans or members of their own
species. Late in the nineteenth century, the view that animals could imitate
was challenged by Edward Thorndike (1898), an early comparative psycholo-
gist, who was the first to examine the ability of animals to imitate under con-
trolled conditions.
Thorndike studied the ability of his subjects to learn, from observing
trained demonstrators, how to escape from various “puzzle boxes” that
required hungry animals to make simple responses (e.g., pull a string or step
on a lever) to gain access to food. Thorndike’s failure to find evidence that
cats, dogs, or chickens could learn such simple acts by watching trained
conspecifics demonstrate them led him to conclude that animals cannot
imitate.
Imitation in animals was of great interest because an ability to imitate was
seen as indicating the presence of complex cognitive abilities (though more
recently that view has been challenged; see Heyes, 1994, 2012). Conse-
quently, subsequent generations of students of animal learning found Thorn-
dike’s failure to provide evidence of imitation in animals an irresistible
challenge, and for 90 years innumerable experiments were undertaken using
more sophisticated versions of Thorndike’s puzzle boxes in the attempt to
demonstrate that animals could, in fact, “from an act witnessed, learn to per-
form that act” (Thorndike, 1911, p. 79) and were therefore capable of com-
plex forms of mental activity.
However, it proved impossible using methods akin to Thorndike’s to
exclude the possibility that some less cognitively demanding social learning
process (such as local enhancement) was responsible for any facilitation of
naïve animals’ acquisition of a behavior that it had seen demonstrated by a
trained conspecific. For example, in a typical study (and one that was widely
cited at one time as demonstrating learning by imitation), kittens given an
opportunity to observe their mother pressing a lever to obtain food sub-
sequently learned to press the lever far more rapidly than kittens that had
watched a strange female press the lever for food. The data can be interpreted
as showing either that (1) kittens imitate their mothers or (2) mother cats are
better than unfamiliar adult cats at calling their kittens’ attention to some por-
tion of the environment. Not surprisingly, all serious discussions found
130 ANIMAL BEHAVIOR

evidence of imitation provided by the results of such studies relatively uncon-

vincing (for review, see Galef, 1988).
The two-action method—Persuasive evidence of imitation in animals
became available only in the latter part of the twentieth century, some decades
after it was first found that a naïve bird (an observer) that had watched a
trained bird (a demonstrator) use either its foot or bill to depress a lever and
obtain food tended to use the same appendage to press the lever as had its
demonstrator (Dawson & Foss, 1965). The power of this two-action method
lay in the fact that demonstrators directed different behaviors toward the same
object and location, making it impossible to attribute any tendency of observ-
ers to copy the behaviors of their respective demonstrators to local
enhancement.
Introduction of the two-action method resulted in a dramatic change in the
experimental paradigm used to study imitation learning. Observer pigeons
and quail watched others of their species either step on or peck at treadles;
observer starlings watched as other starlings pulled or pushed at stoppers con-
cealing food (for review, see Zentall, 2004); observer chimpanzees and chil-
dren watched adult humans as they pushed or pulled at stoppers blocking
access to food in artificial fruit (for review, see Whiten et al., 2004). Results
of such studies provided convincing evidence that a surprising range of avian
and mammalian species could imitate simple behaviors, and, consequently,
that contrary to the conclusion Thorndike (1911) had reached, many animals
could learn to do an act from seeing it done.
Diffusion chain experiments—The two-action method has been widely used
in diffusion-chain laboratory experiments to produce behavioral traditions
in groups of animals that are similar in important respects to the traditions
seen in human social groups. In a diffusion-chain experiment (we have already
seen examples in Curio’s work on social transmission of mobbing in black-
birds and Page and Ryan’s work with fringe-lipped bats) in which imitation
is involved, a demonstrator that has been explicitly trained to exhibit one of
two behaviors directed toward some object is observed by a naïve individual.
When the originally naïve individual acquires the behavior that its trained
demonstrator exhibited, it becomes a demonstrator for a second naïve individ-
ual, and the process is repeated, potentially indefinitely (e.g., Dindo et al.,
2008). The ability of apes to faithfully transmit behaviors across cultural gen-
erations in diffusion-chain studies suggests that they, like humans, may be
bearers of culture (for review, see Whiten, 2009), the subject of the final sec-
tion of this chapter.
Do this—A second method used to explore the possibility that animals are
able to imitate involves first training subjects to repeat arbitrary actions
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 131

performed by a human demonstrator when the demonstrator says “Do this”

and then presenting these trained subjects with new arbitrary actions to see if
these are also copied when the human demonstrator says “Do this.” Obvi-
ously, in order for the do-this experimental strategy to work, the animal sub-
ject has to be able to produce motor patterns similar to those produced by
its human demonstrator. Consequently, such experiments have generally used
monkeys or apes as subjects (for review, see Whiten & Custance, 1996),
though an interesting exception, Moore’s (1992) work with an African gray
parrot, is also described below.
Using the do-this method Tomasello and colleagues (1993) studied imita-
tion in two-year-old human infants and chimpanzees that had been raised
either by humans or by their natural mothers. All subjects watched while a
human demonstrated a series of novel actions directed toward objects. For
example, subjects saw a human demonstrator place an object on her head or
use a lever to pry open the lid of a paint can. The children were told, “Do
this,” and the apes had been previously trained to reproduce familiar actions
when they saw a human engage in them. Apes reared by humans and two-
year-old children imitated the experimenter’s actions both frequently and
equally often, whereas chimpanzees reared by members of their own species
did not imitate the actions of their demonstrators (for review, see Whiten &
Custance, 1996; Tomasello, 2009).
Some animals seem to imitate human actions even without explicit train-
ing. Bruce Moore (1992) housed an African gray parrot (Psittacus erithacus)
that he named Okichoro alone in a room that contained, along with the usual
perches and toys, a microphone and video camera that allowed Moore to
observe Okichoro’s behavior while she was undisturbed.
Moore visited with Okichoro several times each day for more than five
years and during each visit repeatedly performed several distinct movements,
accompanying each with a different spoken word or phrase. For example,
Moore often said, “Look at my tongue!” and then stuck out his tongue. And
each time Moore left, he waved good-bye and said, “Ciao!”
Okichoro rapidly learned to say “caio” and after a year was observed while alone
in her room saying “ciao” and then waving her foot. A few months later Okichoro
started saying “Look at my tongue,” then opening her mouth and raising her tongue.
Over the years, the parrot spontaneously copied six different actions involv-
ing six different parts of her body while coupling each action with the appro-
priate word or phrase. Inappropriate pairings of words and actions were rare
(for review, see Moore, 1996). The performance of Okichoro, like that of Alex
(a fellow gray parrot), was so extraordinary that it has yet to be incorporated in
general discussions of imitation in animals.
132 ANIMAL BEHAVIOR

Anecdote—Much of the evidence consistent with the view that animals can imi-
tate comes from serendipitous observations made outside the laboratory, suggesting
that individual animals from chimpanzees (seen to open paint cans, sharpen
pencils, and use sandpaper) to dolphins (reported to copy the behavior of a diver
cleaning the windows on their aquarium) have learned complex patterns of behav-
ior by imitating human caretakers’ behavior (for review, see Moore, 1992).
Perhaps the most famous of the many anecdotal examples of animals imi-
tating human behavior involved a cat that had learned to open a garden gate
by jumping up and depressing the thumb piece of the gate latch with one fore-
paw while simultaneously pushing at the gate post with its hind legs. The
observer of the cat’s behavior concluded that the cat must have seen humans
open the gate by depressing the latch and simultaneously pushing at the gate
and reasoned, “If a hand can do it, why not a paw,” then produced the
required behavior (Romanes, 1884)—maybe, maybe not.
Although Romanes surely observed the behavior he described, there are at
least two reasons to be cautious in attributing such behaviors to animals being
able to learn to do acts from seeing others engage in them. First and most
important, anecdotal observations in uncontrolled environments rarely pro-
vide insight into how improbable, human-like behaviors develop in animals.
Did the cat observe the gate being opened and then reproduce the behavior it had
seen, or did the behavior of a caretaker focus the cat’s attention on the gate latch
and result in numerous contacts with the handle that eventually led to the cat
learning to open the gate by trial and error? It is impossible to tell from just
watching the cat after it has learned the relevant actions.
Second, tens of millions of people watch tens of millions of pets for hours
each day, but it is only in those rare instances when an animal appears to copy
a behavior of its owner that the incident is reported to others. No one talks
about the countless millions of times when pets see their owners engage in a
potentially useful behavior without copying (Visalberghi & Fragaszy, 1990).
Such selective reporting of positive instances cannot exclude chance corre-
spondences between the behavior of an animal and that of humans with
whom the animal interacts.
One approach to avoiding problems inherent in using chance observation
of possibly interesting behaviors is to formalize the observation of behavior
in uncontrolled environments. For example, Anne Russon and Birute
Galdikas (1993) worked at a camp in Borneo where orangutans that had been
captured by poachers and raised in captivity were reintroduced to life in the
wild. The apes were free to come and go from the jungle to the camp, where
they both interacted with the camp’s human inhabitants and were provided
with food while they made the transition to life in the wild.
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 133

The behavior of the apes was systematically observed, and video recordings
were made of dozens of behavior sequences in which the apes appeared to copy
human behaviors. For example, a female orangutan unscrewed caps from two
fuel drums, then inserted the end of a hose into one fuel drum, inserted the
other end into her mouth, bellowed her cheeks, and after much fooling
around with the end of the hose that had been in her mouth, placed it in
the other fuel drum. Although the fuel drum the ape had selected to start
her procedure was empty and the timing between sucking on one end of the
hose and inserting it into the second rum was inappropriate, the sequence of
acts was similar to that of a human siphoning fuel from one can to another,
an act regularly carried out by camp staff.
As with the earlier field studies of imitation, although it is impossible to
know just how the ape came to exhibit such a complex series of human-like
behaviors, the numerous cases of potential imitation that Russon and Galdikas
filmed suggest that, under some circumstances, orangutans can imitate.
Perhaps most important, videotapes of the apparently imitative behaviors are
available to those interested in viewing them so that others may make their
own judgments concerning the apes’ behavior. Some will surely be impressed
by the similarity of the apes’ performance to that of humans. Others may won-
der at the apes’ apparent lack of directedness as they move, often painfully
slowly, through a behavior sequence. Indeed, some have argued that apes do
not understand the goal of a demonstrator in the same way that humans do
and consequently do not imitate with the same intentionality, instead simply
reproducing behaviors without connecting those behaviors with their goal
(Tomasello, 2009).
To the contrary, others have suggested that free-living animals learn com-
plex patterns of behavior by imitation and that such imitation provides the
foundation for purported behavioral traditions that have been observed in
many species. However, important questions remain as to how such behaviors
develop in the individuals and populations that exhibit them and the relation-
ship of such patterns of behavior in animals to human culture. Both issues are
discussed in the final sections of this chapter.

TRADITIONS AND CULTURE

Historical Examples of Animal Tradition: Japanese Macaques and British Birds
Japanese macaques—Perhaps the best-known instance of the possible spread
of a complex behavior through a population of animals by imitation came
from studies of a free-living (though provisioned) troop of Japanese macaques
resident on Koshima Islet, a small island in the Sea of Japan (Kawai, 1965).
134 ANIMAL BEHAVIOR

In 1953, an 18-month-old female macaque (Imo) began to take pieces of

sweet potato left on a beach by a provisioner (and consequently covered with
sand) to a stream and to wash the sand from the potato pieces before eating
them.
Most Japanese macaques use their hands to brush any sand or dirt from
pieces of sweet potato left for them. However, Imo started washing potato
pieces in water, and the habit of sweet-potato washing gradually spread
through her troop. First Imo’s playmates and then Imo’s mother learned the
behavior, and over the next nine years 14 of 15 juveniles and 2 of 11 adults
in Imo’s troop started washing their sweet potatoes. This sweet-potato wash-
ing, as well as several other unusual behaviors seen in the Koshima troop (from
eating fish to bathing in the ocean), have been interpreted in innumerable sec-
ondary sources as products of imitation, though those who first described
them were considerably more cautious (Kawai, 1965).
Of course, we will never know with certainty what caused sweet-potato
washing to spread through the Koshima troop more than 50 years ago. How-
ever, there are reasons to question the common interpretation of the behavior
as spreading by imitation: (1) the behavior spread extremely slowly (Galef,
1992) with individuals taking, on average, more than two years each to acquire
it; (2) the behavior of washing food is not so unlikely in macaques as intuition
would suggest (Visalberghi & Fragaszy, 1990); and (3) provisioners may have
inadvertently trained the monkeys to exhibit the behavior (Green, 1975).
British birds—In the years following World War II several species of titmice
residing in Great Britain acquired the habit of opening milk bottles left on
doorsteps by milk delivery services common at the time, pecking at the card-
board lids used to seal the bottles and then drinking cream from the surface
of the milk (which was not homogenized). The original investigators of the
geographic spread of the behavior were cautious in their interpretation of its
causes. After all, simple local enhancement might suffice (Fisher & Hinde,
1949). However, as with the Koshima macaques’s sweet-potato washing, text-
books frequently attributed the spread of milk-bottle opening to learning by
imitation. More recent, controlled experiments with chickadees, North
American birds closely related to the British titmice, have shown that simply
drinking cream from milk bottles opened by others can induce birds to start
pecking at closed milk bottles and opening them for themselves (Sherry &
Galef, 1984).
Obviously, we have some way to go before we understand the development
of unusual behaviors in free-living populations and the role, if any, of social
learning in producing such differences. The use of tools to capture insects by
woodpecker finches in the Galápagos Islands and crows in New Caledonia
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 135

provide further examples of problems encountered in determining the role of

social learning in development of complex behaviors.

More Recent Examples of Animal Tradition in Birds

Galápagos finches—Galápagos woodpecker finches (Camarhynchus pallidus)
are famous for their use of twigs and cactus spines as tools to pry arthropods
out of crevices in the bark of trees. If such a complex pattern of behavior were
to be observed in a cetacean or primate, it would surely be viewed as a possible
tradition resting on imitative learning. However, when Sabine Tebbich and
colleagues (2001) examined the development of tool use by woodpecker
finches, they found no evidence that social learning played any role at all.
Juvenile woodpecker finches used tools to look for prey even if they had
never seen tool use demonstrated by others, and adults that had been captured
in damp areas, where tool use is rarely seen, never learned to use tools either by
trial and error or after observation of conspecifics demonstrating the tech-
nique. Tebbich and colleagues (2001) interpreted their findings as indicating
that Galapagos woodpecker finches are predisposed to learn to use tools by
trial and error, but only during a sensitive period early in development.
New Caledonian crow—In nature, New Caledonian crows (Corvus monedu-
loides) (relatives of North American crows) manufacture tools from leaves and
use these tools to retrieve food hidden in crevices, much as woodpecker finches
in the Galapagos use twigs for the same purpose. Differences in the types of
tool and number of types of tool found in different locations on New Caledo-
nia suggest that there may be traditions of tool use in crow populations (for
review, see Holzhaider et al., 2010).
Ben Kenward and colleagues (Kenward et al., 2006) hand-reared four New
Caledonian crows in captivity, giving the maturing birds no opportunity to
interact with tool-using adults of their species. All four hand-reared birds
developed the ability to use twig tools, and one of them both cut a simple tool
from a leaf and used it to obtain hidden food. Observation of a human dem-
onstrator handling potential tools increased the attention that hand-reared
young crows paid to the tools they saw handled but had no influence on their
tool manufacture or use (Kenward et al., 2006).
The general message in examples such as cats opening garden gates, mon-
keys washing sweet potatoes, and birds opening milk bottles and using tools
to get insects out of crevices is that, although discovery of unusual behavior
in an individual or population is suggestive and interesting, such discovery
tells us nothing definitive about the processes responsible for the development
of a behavior in the individuals and populations exhibiting it. In some cases,
136 ANIMAL BEHAVIOR

analyses under controlled conditions of instances of apparent tradition in free-

living animals have raised questions about the validity of explanation in terms
of social learning. In other cases, controlled studies have strengthened inter-
pretation of an unusual behavior as dependent on social learning.

Evidence of Tradition and Culture in Apes: The Method of Exclusion

Discovery of multiple differences in the behavioral repertoires of various
troops of free-living chimpanzees and orangutans has led to the controversial
proposal that they, like humans, are bearers of culture. Part of the controversy
over ape culture reflects disagreement as to how culture should be defined.
However, all involved in such controversy are agreed that, at a very minimum,
for observed differences in the behavioral repertoires of diverse groups of apes
to be considered as even potentially cultural, such differences have to be a
result of social learning (for review, see Laland & Galef, 2009).
In 1999, Andy Whiten, at St. Andrews University in Scotland, and numer-
ous collaborators, who together had accumulated more than 150 person-years
of observation of seven geographically distinct populations of chimpanzees in
their home environments in East and West Africa, presented evidence that
each chimpanzee population exhibits a unique behavioral repertoire in food-
processing techniques and social customs (Whiten et al., 1999). Indeed,
description of the behavioral repertoire of a population is sufficient to identify
that population, much as one could identify human populations as, say, North
American or Asian on the basis of description of their social customs and
methods of processing food. Carel van Schaik and collaborators (van Schaik
et al., 2003) have provided a similar analysis of variation in the behavioral rep-
ertoires of populations of orangutans in Indonesia. Because the issues are sim-
ilar and the data from chimpanzees is more elaborated than that for
orangutans, I focus here solely on the question of whether chimpanzees have
been shown to exhibit culture.
Whiten and his coauthors used what they called the method of exclusion to
attribute to social learning the distribution of 39 behaviors that are customary
(i.e., all relevant individuals show the behavior) or habitual (i.e., several indi-
viduals show the behavior) in some of the seven chimpanzee communities
they examined but absent in others. The label “method of exclusion” refers
to the fact that attribution of a behavior to social learning depended upon
exclusion of alternative explanations of interpopulation differences in behav-
ior, for example, differences in the genetic makeup of communities or differ-
ences in behavior in response to differences in the environments that
different populations inhabit.
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 137

Ecological Effects on “Traditional” Behaviors of Chimpanzees

Perhaps the most carefully studied of the 39 possibly cultural behaviors
described by Whiten and his collaborators concerns chimpanzees’ use of tools
to capture and eat army ants. At Gombe, in Tanzania in East Africa, chimpan-
zees “dipping” for ants hold a long branch that they have stripped of leaves in
one hand, introduce the distal end of the branch into a nest of army ants, and
then quickly withdraw the branch as the ants charge up it to attack. The chim-
panzee then sweeps the length of the wand with its free hand, collecting the
ants into a loose ball, which it then pops into its mouth.
In the Tai Forest in the Ivory Coast in West Africa, ant-dipping chimpan-
zees use a short stick to collect a relatively small number of ants and strip the
ants from the stick by directly pulling the stick through their mouths. The
Tai technique, using the short stick, results in a significantly lower rate of
ant capture than the long-stick technique seen at Gombe.
Chimpanzees at Bossou, in Guinea in West Africa, use both the Tai and
Gombe techniques but use the apparently less efficient short-stick-directly-
into-the-mouth technique considerably more frequently than they use the
apparently more efficient long-stick-hand-to-mouth technique. Chimpanzees
in the Budungo forest in Uganda (East Africa) do not feed on army ants even
though there are plenty of army ants present there for chimps to feed on. In
accord with the cultural hypothesis, simple description of whether and how
ants are exploited suffices to identify four chimpanzee communities.
A second widely studied, purportedly traditional behavior involves the meth-
ods used by some chimpanzees to crack open hard nuts to gain access to their
nutritious kernels. For example, most chimpanzee populations to the west of
the N’Zo-Sassandra River in the Ivory Coast use stone hammers and anvils to
crack open nuts, whereas those to the east of the river, which forms a barrier to
gene flow between populations, rarely do so. The distribution of nut-cracking
behavior could not be explained by differences in the density of chimpanzees,
the density of nut-bearing trees, or the frequency with which objects suitable
for use as hammers and anvils are encountered to the east and west of the river.
Although between-population variation in such complex and apparently
socially learned behaviors as ant dipping and nut cracking provides evidence
consistent with the assertion that, like their human cousins, chimpanzees are
bearers of culture, that assertion has not been universally accepted. There are
two general reasons for such challenges; the first involves a fundamental prob-
lem with the method of exclusion itself.
Because it is always possible that some as yet undiscovered ecological or
genetic difference between communities may explain any observed difference
in their behavior, conclusively demonstrating by the method of exclusion that
138 ANIMAL BEHAVIOR

behavioral divergence between two populations results from social learning is

logically impossible (Laland & Janik, 2006). Indeed, there is evidence that
such exclusion may be more difficult than it seems.
Ant dipping—Tatyana Humle and Tetsuro Matsuzawa (2002) investigated
the role of nonsocial factors in chimpanzees’ use of long and short sticks when
dipping for army ants. They found that the aggressiveness of the ant species
being preyed upon by chimpanzees predicted which of the two techniques
chimpanzees at Bossou use when dipping for ants. To feed on aggressive black
ants, chimpanzees use a long wand and their hand to remove ants from the
probe, whereas chimpanzees feeding on less-aggressive red ants use a short
stick and their mouths to remove ants directly from the stick.
A simple experiment in which humans used wands of different lengths to
dip into nests of red and black ants revealed that the black ants swarmed up
the probes in greater numbers and delivered far more painful bites than did
the red ants. Consequently, it is not too surprising that chimpanzees used long
probes when dealing with the more aggressive and harmful black ants than
when dealing with relatively benign red ants and collected more ants per unit
time when feeding on the former than the latter.
Of course, the finding at Bossou that the method used to dip for ants is
affected by prey behavior does not mean that the different ant-dipping tech-
niques used by chimpanzees at Tai and Gombe are also responses to differ-
ences in prey behavior at the two sites. In fact, a comparison of the behavior
of chimpanzees at Bossou and at Tai (where only the short-stick technique is
seen) revealed that although both aggressive and nonaggressive ants are present
at Tai, Tai chimps feed only on the less aggressive red ants (Mobius et al.,
2008). Thus, social custom determining which ants are eaten at Tai and
Bossou might be indirectly responsible for the difference in feeding techniques
seen at the two locations. Alternatively, some as yet undescribed ecological fac-
tor might determine the kinds of ants that chimpanzees choose as prey. If so,
the methods chimpanzees use to prey on ants might not be traditional at all.

Genetic Effects on Traditional Behaviors of Chimpanzees

Recent research has also provided evidence of a strong statistical correlation
between the genetic distance between chimpanzee populations (determined by
number of nucleotide differences) and the extent of their behavioral diversity
(Langergraber et al., 2011). However, even such strong correlations between
genetic distance and behavioral dissimilarity cannot be interpreted uncritically
as showing that social transmission processes are less important than previ-
ously believed in producing observed differences in behavior patterns. High
rates of short-distance migration between chimpanzee communities of females
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 139

carrying the behavioral variants of their natal troops could result in a positive
correlation between genetic and behavioral distance, even if behavioral var-
iants were entirely socially learned. At the time of this writing, interpretation
of Langergraber and colleagues’ (2011) data remains controversial.

Possible Resolution
Evidence appears to be gradually accumulating indicating that behavioral
differences among chimpanzee populations are likely to result from the com-
bined action of genetic, ecological, and social factors. As Kevin Laland and
Vincent Janik (2006, p. 545) have proposed, “Researchers studying animal
culture would be better advised to think in terms of partitioning variance to
alternative sources. The prime issue in the animal culture debate is not
whether a given behavior is learned socially or asocially, but rather how much
of the variance in the behavior can be attributed to social learning.” Some field
researchers seem to agree (Mobius et al., 2008), and a resolution of the “cul-
ture wars,” similar to that which defused the controversy regarding the role
of genes and environment in development of the behavior of individuals,
seems possible. In the final analysis, the cultural hypothesis of the origins of
behavioral diversity among chimpanzee populations is a hypothesis concern-
ing the development of relevant behaviors. Consequently, anticipating parallel
resolutions of the “nature-nurture controversy” and the culture wars is reason-
able (see Chapter 5 in this volume).

Development of Traditional Behaviors of Chimpanzees

Termite fishing—The cultural hypothesis asserts, almost invariably in the
absence of direct evidence, that social interactions among group members are
the source of the behaviors present in some populations but absent in others.
For example, following a four-year field study of development of termite fish-
ing (a tool-using behavior akin to ant dipping) in young chimpanzees,
Elizabeth Lonsdorf (2006) found that there is substantial variation in both
the time females spend termite fishing in the presence of their young and the
ages at which their offspring become proficient fishers of termites. However,
proficiency was best predicted not by variation in the time mothers spend ter-
mite fishing in the presence of their offspring but by the gender of the young,
clearly a genetic factor.
Young of both sexes appear to learn from their mother both that the termite
mound is an object to which attention should be directed and that the goal of
activity at the mound is to capture termites. Male offspring subsequently
develop their own fishing technique without much reference to the details of
140 ANIMAL BEHAVIOR

their mothers’ behavior while young females seem to learn something of the
proper form of the behavior from watching their mothers. In particular, daugh-
ters, but not sons, insert probes to the same depth into termite nests as did their
mothers. Young of both sexes then had to learn by trial and error how to with-
draw a tool from the nest without knocking off termites (Lonsdorf et al.,
2004). Social interaction, genetic predisposition, and trial-and-error learning
each seem to play some role in development of termite fishing, although the
extent of social influence on development of the behavior was not great.
Nut cracking—Observations of interactions of adult and young chimpan-
zees at nut-cracking sites, like observation of young at termite mounds, are
only partly consistent with the view that nut cracking is socially learned.
Adults frequently chase away juveniles trying to take nuts and stones, and rear-
ing by a mother who did not engage in nut cracking did not slow acquisition
of nut cracking by her offspring (Inoue-Nakamura & Matsuzawa, 1997).
Indeed, Noriko Inoue-Nakamura and Matsuzawa (1997, p. 172) conclude
that “the members of a community provided the infants only with the oppor-
tunity to freely access nuts and stones. These opportunities could facilitate the
individual experience of stone-nut manipulation and result in the apparent
social transmission of tool use behavior among the wild chimpanzees.” Provid-
ing infants with the opportunity to freely access nuts and stones, a form of
local enhancement, can provide the basis for a tradition in a population. How-
ever, as discussed in the next section, the distinction between such social
enhancement of behavior and learning directly by imitation or teaching may
be important in understanding the differences between animal and human
culture.

Concluding Remarks on the Method of Exclusion

The point of the present section is not that social learning plays little role in
development of the 39 potential cultural behaviors identified by Whiten and
colleagues (1999). Rather, it seems reasonable to conclude that knowledge of
the development of purportedly traditional behaviors is currently insufficient
to determine either which of the 39 behaviors identified by Whiten and col-
leagues (1999) as cultural are influenced in their development by social learn-
ing or the kind of social learning involved in development of those behaviors
that are socially influenced.

“Culture” in Chimpanzees and Humans: Analogy or Homology?

There is considerable diversity in views regarding the relationship between
the culture of humans and of other animals (Laland & Galef, 2009). Some
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 141

see no reason to make a distinction, whereas others prefer to restrict use of the
term culture to our own species. All agree that both animal traditions and
human culture function similarly, both providing naïve individuals with access
to innovative adaptive behaviors that others of their species have developed
and reducing potential costs to naïve individuals of acquiring adaptive behav-
iors. However, some argue that human culture serves additional functions
never seen in animal traditions (for example, use of arbitrary social conven-
tions, such as hairstyle or facial decoration, to define group identity [Perry,
2009] or regulation of individual behavior by enforced adherence to specific
systems of rules [Hill, 2009]). Others are less concerned with differences in
function than with differences in both the processes producing traditions in
animals and humans and the nature of the products of those processes (Galef,
2009; Hill, 2009; Tomasello, 2009).
Human culture, particularly human material culture, cumulates over gener-
ations, and as a result can produce increasingly complex and sophisticated arti-
facts and patterns of behavior (Boyd & Richerson, 1996; Tomasello, 2009). A
scant handful of chimpanzee behaviors have been interpreted as providing
possible instances of cumulative culture (for review, see Whiten et al., 2003).
However, each such example is open to alternative, simpler interpretations,
and no one has claimed that any ape has learned any behavior socially that it
could not learn for itself from interacting with its asocial environment.
Much of the behavior in which humans engage is, to the contrary, clearly a
product not of individual learning but of knowledge accumulated across cul-
tural generations. This ratcheting (Tomasello, 1990), seen in all human pop-
ulations, depends upon a naïve observer being able to acquire a close copy of
the behavior of a proficient demonstrator. Only a precise copy can serve as a
base (or scaffold) for subsequent modifications, allowing an iterative process
in which each new modification becomes a scaffold for further modification,
in turn allowing the emergence of gradually increasing complexity in behavior.
Existence of ratcheting in human traditions and its absence in animal tradi-
tions suggests a fundamental difference in the developmental processes sup-
porting animal traditions and human culture (Galef, 1992). For example,
György Gergely and Gergely Csibra (2006) have provided experimental evi-
dence that children will copy the behavior of a model with greater fidelity
when provided with explicit pedagogical cues and suggest that such guided
transmission is essential to the faithful social learning of complex skills. Laur-
aeno Castro and Miguel Toro (2004) suggest that both increased efficiency in
imitation and development of a capacity to approve or disapprove of the learned
behavior of others (a simple form of teaching) were necessary for the emergence
of cumulative culture. In either case, uniquely human processes, particularly
142 ANIMAL BEHAVIOR

teaching and precise imitation (which, like teaching, can at least in principle
result in precise copying of behavior), are seen as providing a necessary substrate
for cumulative culture. Current evidence suggests that chimpanzees never teach,
are not as proficient imitators as are humans (Whiten, 2005), and have no sym-
bolic language (unless taught one by humans) with which to instruct others.
Much of animal social learning depends on local enhancement or emu-
lation. As a result of either, a naïve individual has to develop for itself the
behaviors that it directs towards portions of the environment to which the
behavior of others has directed its attention. In local enhancement and emu-
lation, there is no precise copying of behavior and consequently no possibility
of cumulative improvement in performance over behavioral generations.
The extent to which members of traditional human societies engage in
teaching has been controversial (e.g., Whiten et al., 1980; Laland & Hoppitt,
2003), possibly because of undue focus in the ethnographic literature on
explicit linguistic instruction. Such focus tended to overlook nonverbal behav-
iors of potential teachers that facilitating acquisition by their pupils. A recent
review of the archaeological and ethnographic evidence of teaching that adopts
a definition of teaching similar to that employed by students of animal behav-
ior (Caro & Hauser, 1992) suggests “the gradual scaffolding of skills in a nov-
ice through demonstration, intervention and collaboration . . . has played an
essential role in securing the faithful transmission of skills across generations
and should be regarded as the central mechanism through which long-term
and stable material culture traditions are propagated and maintained”
(Tehrani & Riede, 2008, p. 316).

CONCLUSION
Perhaps in time we shall discover that apparent differences between the
products of social learning in humans and animals are trivial rather than pro-
found and that small differences in capacity have resulted in huge differences
in performance. Until then, we need to explore both differences and similar-
ities in the behavioral substrate(s) of human and animal “culture.”
Laland (2004) has suggested that an ability to evaluate the relative utility of
behavioral alternatives may be a cognitive requirement for the emergence of
human-like cumulative culture. Tomasello (1999) has proposed that imitation
and teaching, which he sees as critical to establishment of cumulative culture,
require a capacity to take the perspective of a model or pupil, a capacity that—
like that for culture—is far more fully developed in humans than in other spe-
cies. Gergely and Csibra (2006) suggest that human cultural capacities
coevolved with the ability to learn and transfer knowledge through teaching,
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 143

which they suggest is a process independent from, and possibly antecedent to,
either language or the ability to attribute mental states. Vaesen (2011) has dis-
cussed eight cognitive abilities of humans but lacking in apes that make cumu-
lative culture difficult, if not impossible, for apes to achieve.
It is, of course, far easier to generate than to test hypotheses concerning the
necessary cognitive substrate of cumulative culture. However, if any such sug-
gestions are correct, culture of the human kind may well have arisen only in
the ancestral hominid line after its divergence from that of the great apes some
6 or 7 million years ago. If so, despite the close phylogenetic relationship
between Homo sapiens and other extant great apes, the traditions of animals
may provide little insight into the evolution of the human capacity for cumu-
lative culture.

ACKNOWLEDGMENTS
I thank the Natural Sciences and Engineering Research Council of Canada
for 43 years of continuous financial support and the tens of technicians and
students, both undergraduate and graduate, whose participation through the
years in the work of the laboratory was indispensable to whatever success it
enjoyed and whose presence made going to work each day such a great
pleasure.

REFERENCES AND SUGGESTED READING

Bateson, P. P. G. & P. Martin. (1999). Design for a Life: How Behavior Develops.
London: Jonathan Cape.
Boesch, C. (1991). Teaching among wild chimpanzees. Animal Behaviour, 41,
530–532.
Boyd, R. & P. J. Richerson. (1996). Why culture is common but cultural evolution is
rare. Proceedings of the British Academy, 88, 77–93.
Campbell, F. M., C. M. Heyes, & A. R. Goldsmith. (1999). Stimulus learning and
response learning by observation in a two-object/two-action test. Animal
Behaviour, 58, 151–158.
Caro, T. M. & M. D. Hauser. (1992). Is there teaching in non-human animals?
Quarterly Review of Biology, 67, 151–174.
Castro, L. & M. A. Toro. (2004). The evolution of culture: From primate learning to
human culture. Proceedings of the National Academy of Sciences, USA, 101,
10235–10240.
Cornell, H. N., J. M. Marzluff, & S. Pecoraro. (2011). Social learning spreads knowl-
edge about dangerous humans among American crows. Proceedings of the Royal
Society of London, B, 279, 499–508.
144 ANIMAL BEHAVIOR

Curio, E. (1988). Cultural transmission of enemy recognition by blackbirds. In T. R.

Zentall & B. G. Galef, Jr. (eds.), Social Learning: Biological & Psychological Per-
spectives (pp. 75–98). Hillsdale, NJ: Lawrence Erlbaum Associates.
Dawson, B. V. & B. M. Foss. (1965). Observational learning in budgerigars. Animal
Behaviour, 13, 470–474.
Diamond, J. (2001). Unwritten knowledge. Nature, 410, 521.
Dindo, M., B. Thierry, & A. Whiten. (2008). Social diffusion of novel foraging
methods in brown capuchin monkeys (Cebus apella). Proceedings of the Royal
Society of London, B, 275, 187–193.
Dugatkin, L. (1996). Copying and mate choice. In C. M. Heyes & B. G. Galef, Jr.
(eds.), Social Learning in Animals: The Roots of Culture (pp. 85–106). San
Diego, CA: Academic Press.
Fisher, J. & R. A. Hinde. (1949). The opening of milk bottles by birds. British Birds,
42, 347–357.
Forsman, J. T. & R. L. Thomson. (2008). Evidence of information collection from
heterospecifics in cavity-nesting birds. Ibis, 150, 409–412.
Fragaszy, D. M. & S. Perry. (2003). The Biology of Traditions. Cambridge, UK:
Cambridge University Press.
Franks, N. R. & T. Richardson. (2006). Teaching in tandem-running ants. Nature,
439, 153.
Freeberg, T. M. (2004). Social transmission of courtship behavior and mate prefer-
ence in brown-headed cowbirds, Molothrus ater. Learning & Behavior, 32,
122–130.
Galef, B. G., Jr. (1977). Mechanisms for the transmission of acquired food prefer-
ences from adult to weanling rats. In L. M. Barker, M. R. Best, & M. Domjan
(eds.), Learning Mechanisms in Food Selection (pp. 123–150). Waco, TX:
Baylor University Press.
Galef, B. G., Jr. (1981). The development of olfactory control of feeding site selection
in rat pups. Journal of Comparative and Physiological Psychology, 95, 615–622.
Galef, B. G., Jr. (1985). Social learning in wild Norway rats. In T. D. Johnston & A.
T. Pietrewicz (eds.), Issues in the Ecological Study of Learning (pp. 143–166).
Hillsdale, NJ: Lawrence Erlbaum Associates.
Galef, B. G., Jr. (1988). Communication of information concerning distant diets in a
social, central-place foraging species: Rattus norvegicus. In T. R. Zentall & B.
G. Galef Jr. (eds.), Social Learning: Biological & Psychological Perspectives
(pp.119–139). Hillsdale, NJ: Lawrence Erlbaum Associates.
Galef, B. G., Jr. (1992). The question of animal culture. Human Nature, 3, 157–178.
Galef, B. G., Jr. (1996). Social enhancement of food preferences in Norway rats: A
brief review. In C. M. Heyes & B. G. Galef Jr. (eds.), Social Learning and Imi-
tation: The Roots of Culture (pp. 49–64). New York: Academic Press.
Galef, B. G. (2009). Culture in animals? In K. N. Laland & B. G. Galef Jr. (eds.), The
Question of Animal Culture (pp. 222–246). Cambridge, MA: Harvard Univer-
sity Press.
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 145

Galef, B. G., Jr. & L. L. Buckley. (1996). Use of foraging trails by Norway rats. Ani-
mal Behaviour, 51, 765–771.
Galef, B. G., Jr. & M. M. Clark. (1971a). Social factors in the poison avoidance and
feeding behavior of wild and domesticated rat pups. Journal of Comparative
and Physiological Psychology, 75, 341–357.
Galef, B. G., Jr. & M. M. Clark. (1971b). Parent-offspring interaction determine
time and place of first ingestion of solid food by wild rat pups. Psychonomic
Science, 25, 15–16.
Galef, B. G., Jr. & L.-A. Giraldeau. (2001). Social influences on foraging in verte-
brates: Causal mechanisms and adaptive functions. Animal Behaviour, 61,
3–15.
Galef, B. G., Jr. & C. M. Heyes. (2004). Social learning & imitation. Learning and
Behavior, 32, 1–140.
Garcia, J. & R. A. Koelling. (1966). The relation of cue to consequence in avoidance
learning. Psychonomic Science, 5, 123–124.
Gergely, G. & G. Csibra. (2006). Sylvia’s recipe: The role of imitation and pedagogy
in the transmission of human culture. In N. J. Enfield & S. C. Levinson (eds.),
Roots of Human Sociality: Culture, Cognition and Human Interaction (pp.
229–255). Oxford, UK: Berg.
Gibson, R. M., J. W. Bradbury, & S. L. Vehrencamp. (1991). Mate choice in lekking
sage grouse revisited: the rolws of vocal display, female site fidelity, and copy-
ing. Behavioral Ecology, 2, 165–180.
Green, S. (1975). Dialects in Japanese Monkeys: vocal learning and cultural transmis-
sion of locale-specific vocal behavior? Zeitschrift für Tierpsychologie, 38,
304–314.
Griffin, A. S. (2004). Social learning about predators. Learning & Behavior, 32,
131–140.
Hepper, P. G. (1988). Adaptive fetal learning: prenatal exposure to garlic affects post-
natal preference. Animal Behaviour, 36, 935–936.
Heyes, C. M. (1993). Imitation, culture and cognition. Animal Behaviour, 46,
999–1010.
Heyes, C. M. (1994). Social learning in animals: Categories and mechanisms. Biologi-
cal Reviews of the Cambridge Philosophical Society, 69, 207–231.
Heyes, C. M. (2012). What`s social about social learning? Journal of Comparative Psy-
chology, 126, 193–202.
Heyes, C. M. & B. G. Galef, Jr. (1996). Social Learning in Animals: The Roots of
Culture. San Diego, CA: Academic Press.
Hill, K. (2009). Animal “culture”? In K. N. Laland & B. G. Galef Jr. (eds.), The Question
of Animal Culture (pp. 269–287). Cambridge, MA: Harvard University Press.
Holzhaider, J. C., G. R. Hunt, & R. D. Gray. (2010). Social learning in New Calado-
nian crows. Learning & Behavior, 38, 206–219.
Hoppitt, W. J. E., G. R. Brown, R. Kendal, L. Rendell, A. Thornton, M. M.
Webster, et al. (2008). Lessons from animal teaching. Trends in Ecology and
Evolution, 23, 486–493.
146 ANIMAL BEHAVIOR

Horner, V., A. Whiten, E. Flynn, & F. B. M. de Waal. (2006). Faithful replication of

foraging techniques along cultural transmission chains by chimpanzees and
children. Proceedings of the National Academy of Sciences, USA, 103,
1378–1383.
Humle, T. & T. Matsuzawa. (2002). Ant dipping among the chimpanzees of Bossou,
Guinea, and some comparisons with other sites. American Journal of Primatol-
ogy, 58, 133–148.
Inoue-Nakamura, N. & T. Matsuzawa. (1997). Development of stone tool use by
wild chimpanzees (Pan troglodytes). Journal of Comparative Psychology, 111,
159–173.
Kawai, M. (1965). Newly acquired pre-cultural behavior of the natural troop of
Japanese monkeys on Koshima Islet. Primates, 6, 1–30.
Kendal, R. L., B. G. Galef, & C. P. van Schaik. (2010). Social learning research out-
side the laboratory: How and why. Learning and Behavior, 38, 187–386.
Kenward, B., C. Rutz, A. A. S. Weir, & A. Kacelnik. (2006). Development of tool use
in New Caledonian crows: Inherited action patterns and social influences.
Animal Behaviour, 72, 1329–1343.
Laland, K. (2004). Social learning strategies. Learning and Behavior, 32, 4–14.
Laland, K. N. & B. G. Galef. (2009). The Question of Animal Culture. Cambridge,
MA: Harvard University Press.
Laland, K. N. & W. Hoppitt. (2003). Do animals have culture? Evolutionary
Anthropology, 12, 150–159.
Laland, K. N. & V. Janik. (2006). The animal culture debate. Trends in Ecology and
Evolution, 21, 542–547.
Laland, K. N. & H. C. Plotkin. (1990). Social learning and social transmission of for-
aging information in Norway rats (Rattus norvegicus). Animal Learning &
Behavior, 18, 246–251.
Langergraber, K. E., C. Boesche, M. Inoue-Muryama, J. Mitani, T. Nishida, A.
Pusey, V. Reynolds, G. Schubert, R. W. Wrangham, E. Wroblewski, & L.
Vigilant. (2011). Genetic and “cultural” similarity in wild chimpanzees. Pro-
ceedings of the Royal Society of London, B, 278, 408–416.
Lonsdorf, E. (2006). What is the role of mothers in the acquisition of termite-fishing
behaviors in wild chimpanzees? Animal Cognition, 9, 36–46.
Lonsdorf, E. V., A. E. Pusey, & L. Eberly. (2004). Sex differences in learning in chim-
panzees. Nature, 428, 715–716.
Mineka, S. & M. Cook. (1988). Social learning and the acquisition of snake fear in
monkeys. In T. R. Zentall & B. G. Galef, Jr. (eds.), Social Learning: Biological
& Psychological Perspectives (pp. 99–116). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Mobius, Y., C. Boesch, K. Koops, T. Matsuzawa, & T. Humle. (2008). Cultural
differences in army ant predation by West African chimpanzees? A compara-
tive study of microecological variables. Animal Behaviour, 76, 37–45.
Moore, B. R. (1992). Avian movement imitation and a new form of mimicry: Tracing
the evolution of a complex form of learning. Behaviour, 122, 231–263.
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 147

Moore, B. R. (1996). The evolution of imitative action. In C. M. Heyes & B. G.

Galef, Jr. (eds.). Social Learning in Animals: The Roots of Culture (pp. 245–
266). San Diego, CA: Academic Press.
Munger, S. D., T. Leinders-Zufall, L. M. McDougall, R. E Cockerham, A. Schmid,
P. Wandernoth, G. Wennemuth, M. Biel, F. Zufall, & K. R. Kelliher.
(2010). An olfactory subsystem that detects carbon disulfide and mediates
food-related social learning. Current Biology, 20, 1438–1444.
Nielsen, M., F. Subiaul, A. Whiten, B. G. Galef, & R. R. Zentall. (2012). Social
learning in humans and non-human animals: theoretical and empirical dissec-
tions. Journal of Comparative Psychology, 126, 109–113.
Page, R. A. & M. J. Ryan. (2006). Social transmission of novel foraging behavior in
bats: Frog calls and their referents. Current Biology, 16, 1201–1205.
Perry, S. (2009). Are nonhuman primates likely to exhibit cultural capacities like
those of humans? In K. N. Laland & B. G. Galef (eds.), The Question of Ani-
mal Culture (pp. 247–268). Cambridge, MA: Harvard University Press.
Ratcliffe, J. M. & H. M. ter Hofstede. (2005). Roosts as information centres: Social
learning of food preferences in bats. Biology Letters, 1, 72–74.
Rendell, L. & H. Whitehead. (2001). Culture in whales and dolphins. Behavioral and
Brain Sciences, 24, 309–324.
Romanes, G. J. (1884). Mental Evolution in Animals. New York: AMS Press.
Ruddle, K. & R. Chesterfield. (1997). Education for traditional food procurement in
the Orinoco delta. Ibero-American No. 53. Berkeley, CA: University of Califor-
nia Press.
Russon, A. E. & B. M. F. Galdikas. (1993). Imitation in free-ranging rehabilitant
orang-utans (Pongo pygmaeus). Journal of Comparative Psychology, 107,
147–161.
Seppanen, J.-T. & J. T. Forsman. (2007). Interspecific social learning: novel prefer-
ence can be acquired from a competing species. Current Biology, 17, 1248–
1252.
Sherry, D. F. & B. G. Galef Jr. (1984). Cultural transmission without imitation: Milk
bottle opening by birds. Animal Behaviour, 32, 937–938.
Smotherman, W. P. (1982). Odor aversion learning by rat fetus. Physiology & Behav-
ior, 29, 769–771.
Steiniger, von F. (1950). Beitrage zur Soziologie une sonstigen Biologie der Wander-
ratte. Zeitschrift für Tierpsycholgie, 7, 356–379.
Stoinski, T. S., L. A. Drayton, & E. E. Price. (2011). Evidence of social learning
in black-and-white ruffed lemurs (Varecia variegata). Biology Letters, 7,
376–379.
Tebbich, S., M. Taborsky, B. Fessl, & D. Blumqvist. (2001). Do wood-pecker
finches acquire tool use by social learning? Proceedings of the Royal Society of
London, B, 268, 2189–2193.
Tehrani, J. J. & F. Riede. (2008). Towards an archaeology of pedagogy: Learning,
teaching and the generation of material culture traditions. World Archaeology,
40, 316–331.
148 ANIMAL BEHAVIOR

Terkel, J. (1996). Cultural transmission of feeding behavior in the black rat (Rattus
rattus). In C. M. Heyes & B. G. Galef Jr. (eds.), Social Learning in Animals:
The Roots of Culture (pp. 17–48). San Diego, CA: Academic Press.
Thorndike, E. L. (1898). Animal intelligence: an experimental study of the associative
process in animals. Psychological Review Monographs, 2, No. 8.
Thorndike, E. L. (1911). Animal Intelligence. New York: Macmillan.
Thornton, A. & K. McAuliffe. (2006). Teaching in wild meerkats. Science, 313,
2227–2229.
Thornton, A. & N. Raihani. (2008). The evolution of teaching. Animal Behaviour,
75, 1823–1826.
Thorpe, W. H. (1963). Learning and Instinct in Animals. London: Methuen.
Tomasello, M. (1990). Cultural transmission in tool use and communicatory signal-
ing of chimpanzees? In S. Parker & K. Gibson (eds.), Comparative Develop-
mental Psychology of Language and Intelligence in Primates (pp. 274–311).
Cambridge, UK: Cambridge University Press.
Tomasello, M. (1994). The question of chimpanzee culture. In R. W. Wrangham,
W. C. McGrew, F. B. M. de Waal, & P. G. Heltne (eds.), Chimpanzee Cul-
tures (pp. 301–317). Cambridge, MA: Harvard University Press.
Tomasello, M. (1996). Do apes ape? In C. M. Heyes & B. G. Galef Jr. (eds.), Social Learn-
ing in Animals: The Roots of Culture (pp. 320–346). San Diego, CA: Academic Press.
Tomasello, M. (1999). The human adaptation for culture. Annual Review of
Anthropology, 28, 509–529.
Tomasello, M. (2009). The question of chimpanzee culture, plus postscript (chim-
panzee culture, 2009). In K. N. Laland & B. G. Galef (eds.), The Question of
Animal Culture (pp. 198–221). Cambridge, MA: Harvard University Press.
Tomasello, M., S. Savage-Rumbaugh, & A. C. Kruger. (1993). Imitative learning of
actions on objects by children, chimpanzees, and enculturated chimpanzees.
Child Development, 64, 1688–1705.
Tuttle, M. D. & M. J. Ryan. (1981). Bat predation and the evolution of frog vocal-
izations in the neotropics. Science, 214, 677–678.
Vaesen, K. (2011). The cognitive basis of human tool use. Behavioral and Brain
Sciences, 35, 203–218.
van Schaik, C. P., M. Ancarenaz, G. Borgen, B. Galdikas, C. D. Knott, I. Singleton,
A. Suzuki, S. S. Utami, & M. Merrill et al. (2003). Orangutan cultures and the
evolution of material culture. Science, 299, 102–105.
Visalberghi, E., & D. Fragaszy. (1990). Do monkeys ape? In S. T. Parker & K. T.
Gibson (eds.), Language and Intelligence in Monkeys and Apes (pp. 247–273).
New York: Cambridge University Press.
von Frisch, K. (1967). The Dance Language and Orientation of Bees. Cambridge, MA:
Belknap Press.
West, M. J. & A. P. King. (1996). Social learning: Synergy and song birds. In C. M.
Heyes & B. G. Galef Jr. (Eds.) Social Learning in Animals: The Roots of Culture
(pp. 155–178). San Diego, CA: Academic Press.
SOCIAL LEARNING , TRADITION , AND CULTURE : DATA AND DEBATE 149

White, D. (2004). Influences of social learning on mate-choice decisions. Learning

and Behavior, 32, 105–113.
Whiten, A. (2005). The second inheritance system of chimpanzees and humans.
Nature, 437, 52–55.
Whiten, A. (2009). The identification and differentiation of culture in chimpanzees
and other animals: from natural history to diffusion experiments. In K. N.
Laland & B. G. Galef Jr. (eds.), The Question of Animal Culture (pp. 99–
124). Cambridge, MA: Harvard University Press.
Whiten, A. & D. Custance. (1996). Studies of imitation in chimpanzees and chil-
dren. In C. M. Heyes & B. G. Galef Jr. (eds.), Social Learning in Animals:
The Roots of Culture (pp. 291–318). San Diego, CA: Academic Press.
Whiten, A., J. Goodall, W. C. McGrew, T. Nishida, V. Reynolds, Y. Sugiyama, et al.
(1999). Cultures in chimpanzees. Nature, 399, 682–685.
Whiten, A. & R. Ham. (1992). On the nature and evolution of imitation in the ani-
mal kingdom. Advances in the Study of Behavior, 21, 239–283.
Whiten, A., R. A. Hinde, C. B. Stringer, & K. N. Laland. (2011). Culture evolves.
Philosophical Transactions of the Royal Society of London, B, 366, 935–1187.
Whiten, A., V. Horner, C. A. Litchfield, & S. Marshall-Pescini. (2004). How do apes
ape? Learning and Behavior, 32, 36–52.
Whiten, A., V. Horner, & S. Marshall-Pescini. (2003). Cultural panthropology.
Evolutionary Anthropology, 12, 92–105.
Whiten, A. & A. Mesoudi. (2008). Establishing an experimental science of culture:
Animal social diffusion experiments. Proceedings of the Royal Society of London,
B, 363, 3477–3488.
Whiten, A., J. Moseley, P. Milner, & I. Agiobukemmer. (1980). Infant education in
Britain and Nigeria. Bulletin of the British Psychological Society, 33, 99.
Yoerg, S. I. (1991). Social feeding reverses learned flavor aversions in spotted hyena
(Crocuta crocuta). Journal of Comparative Psychology, 105, 185–189.
Zentall, T. R. (1996). An analysis of imitative learning in animals. In C. M. Heyes &
B. G. Galef Jr. (eds.), Social Learning in Animals: The Roots of Culture
(pp. 221–244). San Diego, CA: Academic Press.
Zentall, T. R. (2004). Action imitation in birds. Learning & Behavior, 32, 15–23.
Zentall, T. R. & B. G. Galef, Jr. (Eds.). (1988). Social Learning: Biological and Psycho-
logical Perspectives. Hillsdale, NJ: Lawrence Erlbaum Associates.
This page intentionally left blank
 5

Behavioral Genetics: Beyond Nature

and Nurture
Bronwyn H. Bleakley and Anne Danielson-François

INTRODUCTION
The public has been fascinated with behavior genetics, for both humans and
other animals, as long as there has been a public consciousness of genetics.
The discussion is typically framed around determining whether “nature versus
nurture,” or in a genetic context “genes versus environment,” explain the ori-
gins of any particular behavior. We see this dichotomy used throughout the
public discourse on topics as diverse as the causes of language, schizophrenia,
parental care, sexual orientation, aggression, and pair bonding, to name only
a few (Parens et al., 2006). Understanding the underlying causes and evolu-
tion of these behaviors has important implications for advancing our under-
standing of the natural world as well as many areas of public health, ethics,
and policy. For example, autism is a neurodevelopmental disorder character-
ized by impairments in three areas of behavior, including social interaction
(Abrahams & Geschwind, 2008). Developing treatments and potential pre-
vention depends on understanding the underlying causes of the disorder,
including the effects of genes and environments (Auffray et al., 2009). A few
behavioral traits have been shown to result from the action of a single gene
that explains virtually all of the variation observable in that trait. However,
the vast majority of behavioral traits, including autism, have been found to
have complex etiology, with measurable effects of several or many genes
152 ANIMAL BEHAVIOR

combined with environmental influences. As a result, the study of behavior

genetics most often focuses on understanding the relative influences of both
genes and environments that generate observable variation in behavior. These
studies typically utilize a quantitative genetic framework to partition these
sources of variation (Falconer & Mackay, 1996). Recent advances in molecu-
lar genetics, genomics, and epigenetics are now providing us with mechanisms
to understand the role environments play in shaping the expression of
genes. As a result, the underlying causes of behavior cannot be described using
the dichotomy of nature and nurture but rather must incorporate an explicit
understanding of the complex inheritance patterns of behavior that result
from the influences of genes interacting with environments. In this
chapter, we adopt a quantitative genetic model for understanding variation
in behavior.
A devil, a born devil, on whose Nature nurture can never stick, on whom my
pains, humanely taken, all, all lost, quite lost. (Shakespeare, The Tempest, 4.1)
For good nurture and education implant good constitutions, and these good
constitutions taking root in a good education improve more and more, and this
improvement affects the breed in man as in other animals. (Plato, The Republic)

The study of behavior genetics is really the study of individual variation in

and patterns of behavior of both humans and other animals, and it has a long
history. Philosophers have worked to describe the causes of behavior since at
least 300 BCE. Aristotle wrote about psychology, a term he applied to all
living things, in De Anima. Aristotle identified two general causes for individ-
uals to be or to act as they do: causes that are essential to what an organism is
and causes that are incidental to that organism (Shields, 2011). Aristotle was
by no means considering nature versus nurture or genes versus environments,
but essential and incidental causes imply internal and external forces analogous
to those of our modern views. René Descartes, a seventeenth-century philoso-
pher and, among other pursuits, an anatomist, suggested that animal bodies
and behavior could be described in the same ways as machines. He believed
all behaviors were invariant reactions to chemical and mechanical signals,
and consequently, animals acted as “automatons” (Huxley, 1874; reviewed
in Grandin & Deesing, 1998, p. 5). The action of genes as the single control-
ler of behavior is implicit in this view, even in the absence of any knowledge of
actual genetics on the part of Descartes. It is worth noting that this determin-
istic perspective is still prevalent, although not universally applied, for example
in public debates about the “inherent viciousness” of particular dog breeds or
in the common characterization of sharks as nothing but “killing machines.”
While philosophical debates about the “nature” of humans are still going on,
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 153

by the eighteenth century true scientific exploration of the biological causes of

behavior were beginning to occur.
In the mid-nineteenth century, Charles Darwin observed behavior in many
animals, especially domesticated varieties. The observation that humans and
other animals vary substantially within species was integral to his theory of
evolution by natural selection. In addition, Darwin postulated that this varia-
tion must be at least partly heritable, although he did not know the specific
mechanism by which this would be accomplished (Darwin, 1859, 1871).
At the same time, Darwin’s cousin Francis Galton sought to understand how
natural and predictable ranges of phenotypic differences in a variety of human
traits, including behavior, might be inherited (reviewed in Parens et al., 2006).
Galton describes the essential formulation of quantitative genetics when he
discusses his first book Hereditary Genius (1869) by saying, “The object of that
book was to assert the claims of one of what may be called the ‘pre-efficients’
[e.g., genes] of eminent men, the importance of which had been previously
overlooked; and I had yet to work out more fully its relative efficacy, as com-
pared with those of education, tradition, fortune, opportunity, and much else”
(Galton, 1874, pp. v–vi). Despite acknowledging the multitude of factors that
might influence behavior, Galton ultimately rejected the effects of the envi-
ronment as inconsequential to the development of human behavior (Galton,
1874, p. v). Darwin, however, acknowledged that environments might con-
tribute significantly to variation in many traits and did not subscribe to such
a deterministic view as his cousin (reviewed in Kutschera & Niklas, 2004).
In the early twentieth century, the genetics research of Gregor Mendel was
rediscovered and melded with the work of other geneticists and evolutionary
biologists in the “modern synthesis.” The field of quantitative genetics was
blossoming at that time, making use of new statistical techniques to describe
observed variation and beginning to partition the relative influences of genes
and environments in generating such diversity (Falconer & Mackay, 1996).
In particular, Ronald Fisher and Sewell Wright helped to bridge the apparent
gap between the inheritance of categorical traits that could be described as
kinds, such as yellow or green and round or wrinkled peas, and the continuous
variation in traits that formed the basis of Darwin’s theory of evolution by
natural selection. Studies of behavior were developing throughout this period,
first as behaviorism and later as ethology, with a wide range of opinions about
the importance of genes and environments. B. F. Skinner believed behavior
could be fully understood as responses to external stimuli, including the
animal’s own previous behavior (Skinner, 1984). Konrad Lorenz and Niko
Tinbergen, the fathers of the study of ethology, believed that most if not all
behavior was innate (nature/genetic), with environments acting only to trigger
154 ANIMAL BEHAVIOR

a behavioral response (reviewed in Sokolowski & Levine, 2010). By the late

1970s, quantitative genetic approaches that were first widely applied to
improving traits in crops and livestock began to make their way into the tool
kits of behavioral ecologists (reviewed in Boake, 1994b).
But are nature/nurture, innate/learned, and genetics/environment really
dichotomies? This chapter will explore in detail what we know about the inher-
itance of behavior, but as you might have guessed by now, there is no
dichotomy. Genes play an essential role in the development of all traits, but
the action of genes does not happen in a vacuum, so the actions of genes are
not only not independent of environmental influences but may depend explic-
itly on environmental influences. As a result, behavioral traits, like all other
kinds of traits, are influenced by both genes and environment in sometimes
complex interactions. One of the tasks of this chapter is to sort out the degree
to which either is important for explaining observable variation in behavior.
We will begin with simpler cases of inheritance, where a single gene can
explain a large proportion of the variation observed in a trait. We will then
explore traits with more complex patterns of inheritance for quantitative traits
that result from the actions of many genes. We will spend some time disentan-
gling the points at which genes and environments, namely those provided by
relatives and social partners, become difficult to differentiate altogether. And
finally, we will introduce the molecular methods that allow us to explore
explicitly the links between genes and environments in generating behavior
that are covered extensively in Chapter 8, on epigenetics, in this volume.

SINGLE-GENE EFFECTS ON BEHAVIOR

There is a common misconception that there is a gene for every behavior—
a gene for love and a gene for ruthlessness. But genes encode proteins, not
behaviors. Behavior results from a complex interaction between the environ-
mental and genetic factors. These genetic factors can be a protein, mRNA,
or a transcription factor that regulates a feedback loop or disrupts signaling
in the brain, which causes the change in behavior. When these signal pathways
are conserved across taxa, one is likely to find homologous gene sequences that
influence similar behaviors in organisms as disparate as flies and humans.
Behavior is an emergent trait that is influenced by genetic factors and the
environment, but sometimes, very rarely, a single gene can be found that
explains a large amount of variance in a behavioral trait. To locate such a gene,
one first needs to find individuals that differ in their behavior. Behavioral
scientists have taken three main approaches to finding variation in beha-
vior among individuals: (1) creating mutants using technological processes,
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 155

(2) observing behavioral variants in natural populations, and (3) selectively

breeding over generations for a behavioral trait. We will return to the selective
breeding techniques for behavioral variants in later sections of this chapter, so
for now we will start with the first method of finding individuals with differ-
ences in behavior: mutagenesis. By the 1920s, geneticists such as Hermann
Muller were exposing Drosophila fruit flies to X-rays to create large numbers
of mutant flies to scan for behavioral and morphological oddities (reviewed
in Muller, 1973). For his discovery that x rays can induce mutations, Muller
received the 1946 Nobel Prize in Physiology or Medicine. Mutagenesis
(chemical induced or radiation induced) spreads mutations randomly
throughout the genome, causing the inactivation of different genes depending
on where the mutations were located. With hopes of unraveling the mystery of
the brain, mutagenesis was used to generate the first behavioral mutants.

Keeping Time
In the late 1960s Seymour Benzer and his students at Cal Tech began using
mutagenesis combined with behavioral assays to generate behavioral mutants,
using mutations as “microsurgical tools” to dissect the complexity of the nerv-
ous system of Drosophila.
Since it is now possible to produce very high mutation rates in Drosophila, one
can, instead, use an inbred strain and isolate mutants in which a behavioral
change occurs by a single step, so that direct relationships between individual
genes and the nervous system may be investigated. (Benzer, 1967, p. 1112)

Seymour Benzer used this methodology along with a behavioral assay (the
“countercurrent” process) to uncover the first behavioral mutants in Dro-
sophila, which lacked phototaxis—movement in response to light (Benzer,
1967). Subsequent experiments with his students at Cal Tech resulted in the
discovery of many behavioral mutants, including one that showed that the
biological clock was under genetic control, which has fascinated behavioral
scientists ever since (see Chapter 11).
Similar to most animals on earth, flies have daily rhythms for when they are
active, when they eat, and when they sleep. These daily rhythms are called cir-
cadian rhythms (from the Latin circa—about and diem—day). One of the
behavioral assays devised by Ronald Konopka, a graduate student of Benzer,
allowed him to determine the rhythmicity of a fly’s daily movements. After
running hundreds of flies through this assay, Konopka and Benzer discovered
three mutants that had different patterns of eclosion and locomotion. One mutant
had a shortened day (perS), another a longer day (perL), and a third (per0) exhi-
bited no pattern at all—it was arrhythmic (Konopka et al., 1971). These three
156 ANIMAL BEHAVIOR

mutants were evidence that the biological clock was under genetic control.
Since that discovery many more clock genes that each significantly impact an indi-
vidual’s circadian rhythm have been found, including timeless, clock, cycle,
doubletime, and others (reviewed in Vosshall, 2007). For example, period, often
abbreviated per, sets the speed of the Drosophila biological clock and also regulates
its function. Per messenger RNA peaks in concentration at one point during the
day, and the PER protein peaks at a second point, later than the mRNA from
which it is produced. The patterns of accumulation and decrease occur over the
course of 24 hours, creating a feedback loop to regulate the biological clock
(Bargiello et al., 1984; Zehring et al., 1984). There are both positive and negative
transcriptional feedback loops that regulate the biological clock as well as other
regulatory factors, including a cytoplasmic timing mechanism (reviewed in
Vosshall, 2007).
Although we have progressed tremendously in understanding how the Dro-
sophila biological clock functions, and it appears to be a good model for other
insect biological clocks, we are just beginning to unravel the functioning of the
biological clock in vertebrates. Homologs to per have been discovered in mam-
mals, such as the mice homologs mPer1 and mPer2. A human homolog has
been found (Per2), which may explain familial advanced sleep phase syn-
drome, where individuals go to sleep and wake up very early in a very similar
manner to flies with the perS allele (Toh et al., 2001). As more genetic discov-
eries are made, we will likely find more homologs to insect clock genes as well
as new genes involved with the biological clock in vertebrates.

Finding Food
Our next example of a behavioral gene was discovered in a more natural
way. Rather than mutagenizing flies, Marla Sokolowski decided to watch the
behavior of Drosophila larvae feeding on rotten fruit in an orchard. She noticed
a pattern: when food was present under crowded conditions, most of the lar-
vae (~70%) would move long distances, making trails in the food or moving
to another food source, but a subset of the population (~30%) would stay
and eat where they were (Sokolowski, 1980). She named these behavioral
phenotypes “rover” and “sitter” and began the long search for the gene respon-
sible for this difference in foraging behavior. Nearly 20 years later, her persis-
tence paid off, and Sokolowski and her students were able to clone the
forager gene and generate the sequence for the underlying alleles, forR and
forS, respectively (Osborne et al., 1997). She determined that the gene forager
codes for a cyclic guanosine monophosphate (cGMP)–dependent protein
kinase (PKG). Sokolowski found that when exposed to crowded conditions,
rovers have a higher expression of the gene forager and higher levels of PKG
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 157

in their brains. By manipulating the density of larval populations in the labo-

ratory, Sokolowski and her team were able to show that foraging behavior
evolves under density dependence: high-density populations favored rover
phenotypes and low-density populations favored sitters (Sokolowski et al.,
1997). Larvae responded to crowded conditions and food levels, and their
behavior (wandering versus staying put) was influenced by subtle changes in
forager expression levels (reviewed in Sokolowki, 2001). Here, even with the
action of a single gene strongly impacting phenotype, the complexity of an
emergent trait like behavior is more apparent: the environment acts as a trigger
for the type of food-searching behavior exhibited by an individual, and indi-
viduals are plastic in their behavioral response.
Once again, as in the case of per, we have found a gene that is conserved
across many taxa. Homologs of the for gene have since been found in every-
thing from honey bees to humans (Fitzpatrick & Sokolowski, 2004). In honey
bees (Apis mellifera) the timing of food-searching behavior has been influenced
by a homolog of the forager gene, Amfor (Ben-Shahar et al., 2002). Young
adult honey bees start as nurses in the hive, and then as they get older they
become foragers that search for nectar and pollen to bring back to the hive.
Honey bees that have become foragers have higher levels of expression of
Amfor RNA in their brains and higher levels of PKG enzyme activity. When
nurse bees are treated with PKG, they become foragers (Ben-Shahar et al.,
2002). Many genes are involved in the switch from nurse to forager (Whitfield
et al., 2003), but clearly Amfor is critical to the transition.
Although we do not yet know the influence of homologs of forager in
humans, it is certainly a candidate gene for examining food-searching behav-
iors in animals. A homolog of for, called egl-4, has been found in the nematode
Caenorhabditis elegans and influences locomotion related to food-searching
behavior (Fujiwara et al., 2002). Another C. elegans gene that influences
food-searching behavior is the neuropeptide Y receptor homolog, npr-1
(de Bono & Bargmann, 1998). Natural variants of npr-1 are similar to
Drosophila rover/sitter phenotypes, and it would be reasonable to hypothesize
that it lies in the same cGMP signaling pathway as for (reviewed in Sokolowksi,
2001).

Nurturing Offspring
More recently, technological advances have allowed scientists to use the
techniques of “reverse genetics” to knock out the function of a particular gene
in order to test its function. In model systems where behavior could be quan-
tified and genetic tools were present, certain genes could be targeted and ren-
dered inactive (“knocked out”). This technique was used for the next
158 ANIMAL BEHAVIOR

behavioral genes we shall introduce, genes that showed behavioral scientists

that parental nurturing behavior and pair bonding were under genetic control.
In the mid-1990s, a gene regulating parenting behavior—fosB—was dis-
covered within the fos family of transcription factors, which are induced by
the environment. This was the first discovery that nurturing in mammals
had a genetic component. The researchers at Harvard Medical School created
a mouse with an inactive form of the gene fosB (a knockout) to test its func-
tion. When the gene was inactive, postpartum female mice did not exhibit
the nurturing behavior, such as licking and nursing, needed for their pups to
survive, although in every other way the mothers appeared normal (Brown et
al., 1996). Indeed, even mice that had not yet been parents, young female
and male fosB knockout mice, were also observed to lack the nurturing pheno-
type. The researchers could not say with certainty which genes fosB was acti-
vating. However, they suggested that oxytocin receptor genes were good
candidate genes to examine next, based on the recently published structure
of the rat oxytocin receptor gene, which was shown to contain an AP-1 site
within its promoter region (Rozen et al., 1995).
An interesting possibility is that FosB, which may be induced by oxytocin bind-
ing to receptors on POA neurons, then leads to increased expression of the oxy-
tocin receptor gene. Thus, by a feedforward mechanism, POA neurons would
express more oxytocin receptors and would therefore have enhanced sensitivity
to oxytocin. (Brown et al., 1996, p. 306)

Later work would show that their hunch was correct. Oxytocin is often called
the “love hormone” for its role in social bonding. The brain neuropeptides
oxytocin (OXT) and arginine vasopressin (AVP) are similar in structure (dif-
fering by only two amino acids), and both play a role in nurturing and pair-
bonding behavior (reviewed in Young et al., 2008). Vasopressin varies across
species and influences pair bonding between males and females. Sensitivity
to vasopressin is influenced by the avpr1a gene that encodes one of the AVP
receptor subtypes (vasopressin 1 receptor, V1aR) and a polymorphic repeat
sequence in the 5’ flanking region of this gene.
Long-term pair bonding is critical for monogamous mating systems, which
are fairly rare in mammals. Approximately 3 percent of all mammalian species
have monogamous mating systems (Kleiman, 1977). The prairie vole (Microtus
ochrogaster) is monogamous, with both males and females forming lasting pair
bonds. Both OXT and vasopressin are important for partner preferences and pair
bonding between male and female prairie voles (Cho et al., 1999). When the
genes influencing vasopressin receptors, the vole avpr1a gene and its 5’ flanking
regions, were inactivated, it transformed the once-monogamous prairie vole
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 159

males to become polygamous with weak pair bonding to their mates (Winslow
et al., 1993). Comparative studies with polygamous meadow (M. pennsylvanicus)
and montane (M. montanus) voles showed that the gene sequence was similar to
that in prairie voles, although the 5’ flanking region was not (Young et al., 1996,
1999). When the entire prairie vole gene sequence and the 5’ flanking region was
inserted into the formerly promiscuous vole species, pair bonding and mate pref-
erence was enhanced (Lim et al., 2004). For the first time, behavioral scientists
found that entire mating systems were under genetic control and influenced
whether voles were monogamous or polygamous (reviewed in Young et al.,
2008).
Once again, we find conservation of behavioral genes across taxa. There is a
homolog of the avpr1a gene in humans, AVPR1A, although there is no homo-
log of the polymorphic repeat sequence in the 5’ region of the avpr1a gene as is
found in prairie voles (Walum et al., 2008). But there are different polymor-
phic microsatellite repeats in that region, and one of those polymorphic
repeats (RS3) is associated with pair-bonding variation among humans
(Walum et al., 2008). Of the tested RS3 alleles, one of them (allele 334) was
associated with variation in male pair bonding. Males homozygous for the
334 allele had significantly lower scores on the Partner Bonding Scale, had
increased marital crisis scores, and were more likely to be unmarried than
males with one or no copies of the allele (Walum et al., 2008). The authors
made the important point that this allele does not predict the behavior of indi-
vidual males but is associated at the group level with variation in pair bonding.
So, similar to prairie voles, genetic factors associated with the AVPR1A gene
influence pair bonding in humans. But this is not the end of the story for
social bonding; rather it is the start of a whole new line of inquiry. Later in this
chapter, and with more detail in Chapter 8, we shall learn that how much
nurturing an offspring receives from its mother alters the fate of that off-
spring’s gene expression and future parenting behavior through epigenetic
mechanisms.

COMPLEX INHERITANCE AND QUANTITATIVE TRAITS

Traits such as those described in the previous section, for which the major-
ity of variation can be attributed to a single gene with two or more alleles, are
relatively easy to describe. However, they are unfortunately the rare case for
most phenotypes, including behavior. The vast majority of behavioral pheno-
types are best described by continuous variation between a minimum and a
maximum. For example, it does not make a great deal of sense to describe high
and low categories of feeding rates when a bird might provision its young
160 ANIMAL BEHAVIOR

anywhere from 12 to 60 times per hour (Freeman-Gallant & Rothstein,

1999). We must therefore modify our genetic descriptions to allow for con-
tinuous variation. Continuously distributed traits are typically referred to as
quantitative traits, and they result from the combined actions of several to
many genes. The distribution of phenotypes for quantitative traits is not really
different than for categorical traits; it is simply that there are many small “cat-
egories” within the range of variation for the trait. For example, a single gene
that shows incomplete dominance will give us three categories of a trait, which
already looks a bit like a normal distribution (Figure 5.1a). Adding in the
effects of just two more loci, each with two alleles (e.g., Bb, Cc) gives us 64
possible genotypes. If each capital allele adds one feeding event per hour, then
our population has seven possible phenotypes that will be distributed as an
approximately normal distribution (Figure 5.1b). With even small variations
in the effects of particular alleles or environmental influences blurring the
distinction between the classes of the phenotype, we end up with a smooth,
continuous distribution (Figure 5.1c).

Figure 5.1. The phenotypic classes that result from three simple genetic models. In all
cases, any capital allele adds a unit to the phenotype (e.g., one feeding event per hour).
(a) Single gene with two alleles, (b) three loci with two alleles each, and (c) three loci
with two alleles each that have slightly unequal effects or environmental effects. In (c),
the unequal effects of the alleles in addition to the environment blur the distinction
between classes, yielding a quantitative trait.
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 161

In these examples, each of the loci had approximately equal effects on the phe-
notype that were independent of each other. As a result, we could predict how
many feeding events a parent would engage in simply by adding up the number
of capital alleles. This is termed additive effects of alleles. Whenever alleles inter-
act, the effects are termed nonadditive effects. The simplest interaction between
alleles is dominance, where the effect of one allele at a locus is partially or entirely
hidden by another allele at the same locus. Epistasis works in much the same
way, but the effects are seen across loci (Falconer & Mackay, 1996; Lynch &
Walsh, 1998; Brodie, 2000). This is analogous to viewing the alleles at one locus
as the environment experienced by the alleles at a second locus and the effects of
the alleles at the second locus are context dependent (Brodie, 2000).
Breadcrumb sponge isopods (Paracerceis sculpta) exhibit a unique set of mat-
ing behaviors that are linked to the presence of three genetically and behaviorally
distinct male morphs. Alpha males are large and aggressively defend harems of
females. Beta males are female mimics, able to infiltrate a harem and mate because
they fool the Alpha male by looking and acting exactly like females. Finally,
Gamma males are much smaller than either Alpha or Beta morphs, steal into
the harem, and engage in sneak mating with females (Shuster, 1987). Alpha
males have an ams-a/ams-a for the alternative mating strategy gene. Gamma
males are ams-g/ams-a or ams-g/ams-g. Finally, Beta males carry at least one
ams-b allele. A second gene, transformer (trf), can change males to females, but
whether a male develops as a female is dependent on its genotype at the ams locus.
Alpha males become female if they have at least one copy of the trf-2 allele. Beta
males are never transformed, and Gamma males are transformed if they have at least
one copy of the trf-1 allele. As if that were not complicated enough, sex in these iso-
pods may also be influenced by a maternal effect transferred from a mother through
the cytoplasm of an egg to her offspring (Shuster & Sassaman, 1997).
Because there is no single gene for feeding rate, alternative male behavior
in Paracerceis isopods, or other quantitative traits, the genes that contribute
to the quantitative traits are called quantitative trait loci, or QTL for short
(Falconer & Mackay, 1996). QTL may combine in various ways, many with
minor effects and some with major effects on the phenotype. The collective
actions of all QTL, including the magnitude of their effects, their interactions,
and the influences of the environment on the expression of those loci, are
summed as the genetic architecture of a trait (reviewed in Mackay, 2001).

Quantitative Trait Loci

QTL are identified by looking for statistical associations between particular
markers within a genome and phenotypic variation (Falconer & Mackay,
1996; Lynch & Walsh, 1998). QTL analysis relies on identifying molecular
162 ANIMAL BEHAVIOR

markers that are distributed throughout the genome. These markers do not
have to code for any particular trait; instead, they function like flags used to
identify physical locations throughout the genome. Individuals are scored for
variation in all of the markers and for variation in the trait being studied. Par-
ticular alleles for some markers will be statistically associated with a particular
phenotype. Imagine, from our hypothetical example above, that allele 1 of
marker 1 is found in a large proportion of birds who feed their babies 60 times
per hour while a different allele is often found in birds that feed less often. The
statistical association between that marker and the behavior signals that a locus
that is physically near the marker influences feeding behavior (QTL analysis
reviewed in Cheverud, 2000).
QTL analysis provides information about the genetic architecture of a trait,
including the number of loci that influence phenotypic variation and among
those loci, which have major or minor and dominance or epistatic effects
(Falconer & Mackay, 1996). Any number of genes might influence the ultimate
outcome of behavior but are not the direct cause of the behavior (Hoekstra,
2010). For example, Grl1 controls the development of glucocorticoid receptors
on cells, which bind and initiate the response to cortisol and corticosterone,
sometimes referred to as stress hormones in vertebrates (Tronche et al., 1999).
Glucocorticoids and their interactions with receptors influence many beha-
viors, from aggression to parental care and mating to migration (e.g., Leary et
al., 2006; Soma et al., 2008; Horton & Holberton, 2010; Onuma et al.,
2010), but the Grl1 gene is not a gene for a particular behavior. In a QTL analy-
sis, Grl1 might show up as a gene of minor effect (e.g., Crabbe et al., 1994).
Conversely, variation in a gene of major effect is associated with a large propor-
tion of phenotypic variation. For example, a single gene, or a very tightly linked
set of QTL, is associated with up to 73 percent of the variation in feeding behav-
ior of pea aphids (Caillaud & Via, 2012).
QTL analysis also provides information about how alleles and loci interact,
whether through dominance, epistasis, or additive effects (Phillips et al.,
2000). With the advent of genomic sequencing, QTL analysis may be taken
a step further. Specific candidate genes that fall within a QTL region can be
sequenced and matched to existing knowledge about the action of that gene.
Increasingly sophisticated genomic tools allow gene networks that influence
complex traits to be mapped. Genetic variation may then be tied to specific
proximate mechanisms, such as changes in receptor density or the efficiency
of peptide synthesis. Identifying the action of particular genes is particularly
important for applying animal behavioral genetics to humans, understanding
the evolution of behavior across multiple species, or developing treatments
for conditions linked to behavioral genes (reviewed in Crabbe et al., 1994;
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 163

Mackay, 2004, 2009). Of particular interest to behavioral ecologists is linking

genetic variation to phenotypic variation that may be the target of selection.
The genetic architecture of traits affects the evolutionary response, and under-
standing the genetics of behavior is therefore necessary to understand
behavioral evolution (reviewed in Phillips et al., 2000; Bleakley et al., 2010;
Wolf & Moore, 2010b).

Drosophila : A Model for QTL Analysis

QTL for many different behaviors, including stress resistance, olfactory
behavior, drug tolerance, courtship, fertility, and longevity, have been identi-
fied for fruit flies (Drosophila melanogaster). As with many other aspects of
biology, research focuses on D. melanogaster in part because the genome has
been fully sequenced and because Drosophila stock centers carry hundreds of
isogenic (genetically identical) lines of flies that can be compared to match
differences in behavior with known differences in genetics (Mackay, 2004).
Three examples will serve to illustrate the breadth of research on identifying
QTL for behavior in Drosophila. Trudy Mackay’s lab at the University of
North Carolina used isogenic lines of D. melanogaster to identify QTL for
locomotor behavior. They identified four QTL that corresponded to 12
chromosomal regions encompassing 13 candidate genes, including the gene
Ddc that controls the last step in the synthesis of serotonin and dopamine.
As a result, they suggest that serotonin and dopamine might be related to loco-
motor behavior in other animals, including humans, and could be important
for developing therapeutic treatments for diseases that affect motion. It is
worth stopping to note that just as some of the single genes described earlier
in the chapter were sensitive to environmental effects, so were all four of these
identified QTL (Jordan et al., 2006). Aggression in D. melanogaster is associ-
ated with five QTL encompassing four candidate genes, at least three of which
interact with each other in complex epistatic ways (Edwards & Mackay,
2009). QTL analysis, followed by sequencing and association mapping, iden-
tified 12 polymorphisms in two genes that influence fertility and the rate at
which female D. melanogaster remate. The differences in female mating
behavior are likely to have consequences for the strength of sexual selection
experienced by males, and QTL analyses like these help to explain fitness
differences associated with particular phenotypes and the evolution of behav-
ioral traits (Giardina et al., 2011).
Larger genetic differences that might contribute to sex differences or species
differences, including divergent selection pressures among them, have also
been investigated using QTL in Drosophila. For example, the Mackay lab
164 ANIMAL BEHAVIOR

substituted 43 X chromosomes and 35 third chromosomes from wild D. mel-

anogaster into a common genetic background and exposed the flies to a chemi-
cal odorant, benzaldehyde. They found that relatively few QTL, each of large
effect, influenced odor-guided behavior. Surprisingly, the QTL identified for
males differed from the QTL identified for females, despite being in the same
species (Mackay et al., 1996). Cuticular hydrocarbons, which serve as phero-
mones and contribute to courtship behavior, are influenced by 25 QTL in
female and 15 QTL in male D. melanogaster, but as with odor-guided behav-
ior, they are not the same loci in the two sexes. Such genetic differences
between males and females in the underlying causes of their behavior are inter-
esting because they suggest sex-specific behavior experiences different pres-
sures from natural and sexual selection (Foley et al., 2007). Reproductive
isolation between closely related species is often accomplished by behavioral
differences, and QTL analysis can contribute to our understanding of the
genetics of speciation. For example, hybrids of the closely related species
of D. simulans and D. sechellia were used to identify six QTL explaining
41 percent of variation in reproductive behaviors that produced reproductive
isolation between the two parental species (Gleason & Ritchie, 2004).

Linking QTL for Behavior to Evolution Using Nonmodel Systems

QTL have also been identified in many nonmodel organisms, including
honey bees and Russian foxes, a domesticated form of the silver fox (Vulpes
vulpes). Beyond highlighting the sometimes complex interactions that emerge
within gene networks, these systems provide information about broader evolu-
tionary and ecological questions, including the genetics of division of labor in
eusocial systems, the processes of adaptive evolution and domestication, and
genetics associated with invasive species. Answering these questions relies on
measuring natural genetic variation for relevant behaviors by quantifying
differences within and between populations or species. Measuring such varia-
tion is multistaged, often involving controlled breeding crosses, fine-scale
mapping, and measures of natural and manipulated gene expression, but it
all begins with identification of QTL and candidate genes within the QTL
(Mackay, 2004; Hunt et al., 2007b).
Africanized honey bees are an invasive subspecies that has largely replaced
European honey bees throughout much of their New World range and is
responsible for several hundred human deaths across Europe and North
America (Guzmán-Novoa et al., 2002; Scott Schneider et al., 2004). Success-
ful invasion of new territory is linked to highly aggressive stinging behavior,
where the bees swarm threats to their hives. The genetics of stinging behavior
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 165

have been explored using QTL analysis: F1 hybrids between European honey
bees and Africanized honey bees were created in both directions, and the
hybrids were then backcrossed to both parental populations. As a result, both
parental types, F1 hybrids, Africanized backcrosses and European backcrosses
could all be compared for stinging behavior. The bees were then genotyped
for a QTL locus associated with stinging, sting1. The marker linked to sting1
was significantly more likely to be found in individuals that stung more often
in the European backcross genotypes but not in the Africanized backcross
genotype, and the effects of the allele appeared to be dominant. In this case,
sting1 is the gene of major effect but cannot by itself account for all variation
in stinging behavior (Guzmán-Novoa et al., 2002). The dominant effects of
a single gene of large effect may help to explain the rapid, invasive spread of
the Africanized aggressive behavior throughout much of the North American
range of European honey bees.
Genomic data for the honey bee allows analysis of QTL to go a step further
in analyzing the evolution of social behavior and caste systems. Honey bee
workers act as nurses for the developing larvae in the hive. Later in life, their
behavior changes radically, and the nurses become foragers, who may further
specialize in collecting pollen or nectar. The change from nurse to forager
must be accomplished through gene regulation, switching from a nurse set
of genes to a forager set of genes. Behavioral studies identified correlated
behaviors, and expression assays and comparative bioinformatics studies
together allow researchers to identify the functions of and correlations between
genes necessary for foraging behavior (Hunt et al., 2007b). Several studies
have found that the size of workers’ ovaries is genetically linked, through
pleiotropic effects of several genes of major effect, to the specific behavior of
worker bees provisioning the colony known as pollen-hoarding syndrome.
These genes may also play a role in caste determination (Wang et al., 2009;
Graham et al., 2011). Pollen-foraging behavior therefore appears to have
evolved from reproductive behavior by co-opting a gene network that supports
ovarian development and insulin signaling in reproductive individuals (Hunt
et al., 2007a). This is particularly fascinating because, of course, workers bees
do not reproduce; evolution simply “borrowed” an existing network of genes
and tweaked it for a different purpose.
Defensive and aggressive behaviors appear to have followed a similar evolu-
tionary path. QTL that are important for activity of the central nervous sys-
tem, neurogenesis, and sensory structures also strongly influence aggressive
behavior. Once again, evolution has co-opted existing gene networks and
modified them slightly for additional functions (Hunt et al., 2007a). This also
provides the perfect example for why the actions of genes cannot realistically
166 ANIMAL BEHAVIOR

be separated from environmental influences: by definition, sensory signaling

and the actions of the central nervous system respond to immediate and
long-term environmental information but do so according to the genetic back-
ground of the individual receiving the information.
Russian foxes provide our last example of QTL analysis. Russian foxes
might not seem like an obvious candidate species for genetic analysis, but they
have been the focus of a long-term quantitative genetic experiment to under-
stand the genetics of domestication that began in an unlikely way. Russian
foxes have been reared for the fur trade for many years, and in the 1950s fur-
riers became interested in developing tamer foxes that could be more easily
handled. At the same time, Russian geneticist Dmitri Belyaev was helping to
create the Siberian Department of the Soviet Academy of Sciences, where he
served as the director of its Institute of Cytology and Genetics from 1959 until
his death in 1985. Belyaev had made many observations of domesticated ani-
mals, and he noted that regardless of species—guinea pigs to goats, dogs to
horses—one often found a particular set of physical changes, such as drooping
ears, shorter tail length paired with altered carriage, and coat color and style,
accompanying behavioral changes. He hypothesized that these correlated
changes might result from selecting on genetic variation for physiological char-
acteristics, such as stress response, that provide the foundation for both behav-
ioral and morphological traits. Belyaev, over the course of the next 50 years,
performed selection on Russian foxes, first for approachability and then later
for more and more domestic behavior. The surprisingly fast end result of this
process was the Russian silver fox (Trut, 1999).
Silver foxes and their wild relatives, which remain completely undomesti-
cated without the process of selecting for friendly behavior, even in captivity,
are now used to study the process of domestication and the evolution of
behavior in general. The tame foxes show a variety of developmental, neuro-
chemical, and morphological changes associated with the behavioral changes,
including many of the morphological changes Belyaev predicted (Trut,
1999; Lindberg et al., 2005). For example, the tame foxes develop silver coat
color, compared to the wild black color; their tails shorten a bit and become
curled; and their ears droop. They also very rapidly develop dog-like behavior,
barking and wagging their tails and responding to human verbal and body-
language cues. The behavioral and morphological changes in foxes turn out
to be linked in large part by pleiotropic changes in adrenal gland production
of corticosteroids, the stress hormone from our hypothetical example at the
beginning of this section (Trut, 1999).
F1 foxes resulting from crosses of wild and tame foxes exhibit a wide range
of behaviors that overlap the behavior of both parent strains, suggesting that
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 167

many genes, each of small effect, are influencing tame behavior (Kukekova et
al., 2008). Genomic work has more precisely identified where genetic changes
occur via domestication. Tame foxes maintained both with and without selec-
tion vary for how quickly they approach a human, how long they stay in con-
tact, and other measures of domestication. These relatively large evolutionary
changes in behavior are not, however, accompanied by large changes in gene
expression within three key brain regions (hypothalamus, amygdala, and fron-
tal lobe). These results are contrary to observations of Drosophila, in which
behavioral changes appear to require large changes in the transcriptome,
which can be thought of as the full library of transcription products being
made by an individual’s genes (Mackay et al., 2005). Evolutionary changes
in behavior from wild to domestic behavior in foxes also required big
differences in transcriptomes, but differences between the selected and nonse-
lected tame lines required very few changes (Lindberg et al., 2005). It remains
to be discovered why some behavioral changes require large differences in
gene expression while others require only slight tweaking, even in the same
species.

Quantitative Genetics: Breaking the Stick of Variance

As alluded to in the introduction, the goal of quantitative genetics is to
describe the proportion of variation attributable to different effects. We can
describe this as “breaking the stick of total phenotypic variation” (Arnold,
1994, p. 30). In practical terms, this allows quantitative geneticists to partition
the effects of genetics, including identifying the roles of additive, dominance,
and epistatic variation and environmental influences on phenotypes. How-
ever, it becomes ever trickier to separate the influences of genes and environ-
ments. For example, genes and environments may interact, generating plastic
traits, and sometimes the environment contains genes of its own that we must
consider, in particular the social environment. To understand how all of these
factors add to up to create variation in behavior, it is easiest to start with the
basic quantitative genetic model:
z=aþe [Equation 1]
where z refers to phenotype, a to the additive effects of genes, and e to the
effects of the environment (Falconer & Mackay, 1996). It is worth pausing a
moment to tell you that in quantitative genetic models, capital letters refer
to the variation attributable to the effects of that particular component; for
example, E represents variation in the phenotype attributable to environmen-
tal effects. Lowercase letters refer to the effects themselves.
168 ANIMAL BEHAVIOR

In general, quantitative geneticists try to identify the sources of pheno-

typic variation because they are ultimately interested in using that information
in other ways (Falconer & Mackay, 1996). Behavioral ecologists use informa-
tion about the genetic architecture of a trait to better understand how the trait
evolved in the past and how it might change in the future (Boake, 1994a,
1994c). They are typically interested in measuring additive genetic variation
specifically, because it provides the best information about how a single trait
will respond to selection (Brodie, 2000; Wolf, 2000).
Sexual behavior has been well studied in many species, in part because
sexual selection often leads to incredibly fast and flamboyant evolution (see
Volume 2, Chapters 4 and 6). For female preferences to evolve there must
be additive genetic variation for choosiness or other aspects of female choice,
but strong selection may erode genetic variation (reviewed in Miller & Moore,
2007; Danielson-François et al., 2009). A common first step in understanding
a mating system is to measure additive genetic variation (Cade, 1984; Boake,
1989). For example, Klaus Reinhold and colleagues identified significant addi-
tive genetic variation, estimated through repeatability analysis, for female
grasshopper (Chorthippus biguttulus) mating behavior, a prerequisite for the
evolution of female preferences (Reinhold et al., 2002). Robert Brooks and
John Endler identified significant additive genetic variation for two aspects
of choosiness, responsiveness to particular males and discrimination among
males, for female guppies (Poecilia reticulata). They were able to identify some
male traits that females uniformly preferred and others for which female pref-
erence varied. They were then able to incorporate those data into a model to
investigate whether it is important to include information about mating pref-
erences in models for trait evolution (Brooks & Endler, 2001). There appears
to be substantial variation across species and traits, however. For example,
Chris Boake (1989) estimates zero additive genetic variation for mating behav-
ior in flour beetles (Tribolium castaneum).
As you know from previous parts of the chapter, however, the simplest
description of genetics is often insufficient to explain all the variation we
observe in behavioral phenotypes. Our model can be expanded to include
various kinds of effects, including dominance (d) and epistasis (i):
z=aþdþiþe [Equation 2]
These additions require more sophisticated experiments to identify. They are
usually measured through what are termed breeding designs, where
information about who is related to whom can be used to statistically
distinguish between additive, dominance, and epistatic effects of genes. For
example, if you look at how closely half-siblings created from mating a father
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 169

to multiple mothers resemble one another, you should see only the additive
effects of alleles. This works because offspring cannot inherit an entire geno-
type from their father, only one of each of his alleles, and siblings from multi-
ple mothers should not often get the same second allele at any particular locus.
As a result, dominance and epistatic relationships are different across half-
siblings and what remains to explain resemblance is simply additive effects
(Falconer & Mackay, 1996). Data from breeding designs can be collected
purely at the phenotypic level by measuring observed variation in traits, or it
may be collected by incorporating QTL analyses that identify particular loci.
Let us return briefly to two examples from earlier in the chapter to illustrate
how measures of epistasis, obtained through QTL analysis, have informed our
understanding of the inheritance of behavior. We previously discussed genes
responsible for establishing and maintaining an individual’s biological clock,
including period, timeless, and clock. While variation in any one of these loci alone
can have significant effects on an individual’s circadian rhythm, they ultimately
interact in complex ways to generate variation in biological timing both within
and across species. Using F2 hybrid mice obtained by crossing two inbred lines
that differed in their biological clocks, Kazuhiro Shimomura and his colleagues
in Joseph Takahashi’s lab carried out genome-wide complex trait analysis for five
aspects of circadian rhythm phenotypes. Fourteen loci contributed significantly
to variation in three of the phenotypes: period, phase, and amplitude of the bio-
logical clock. However, they also found two additional pairs of loci that interact
epistatically to influence significantly the other two aspects of clock phenotype,
dissociation and activity levels. They also detected at least some interactions
among the loci influencing period, phase, and amplitude. As a result, epistatic
variation is an important contributor to variation in circadian rhythms
(Shimomura et al., 2001). Odor-guided behavior in Drosophila is also strongly
influenced by epistatic interactions. Trudy Mackay’s lab combined measures of
transcription with quantitative genetic analysis on five smell-impaired (smi)
mutant lines of D. melanogaster and identified an astonishing 530 loci that are
coregulated in response to smi mutations, providing many loci among which epis-
tasis could occur. In addition, epistasis at the transcription level translated into
epistatic effects on the phenotype 67 percent of the time (Anholt et al., 2003).

Phenotypic Plasticity and Genotype × Environment Interactions

One of the more interesting and yet difficult aspects of behavior is that it
is so variable. That is, in one environment an animal might do one food-
searching behavior, but in another environment the same animal might do a
different food-searching behavior. Behavior can change over time and
170 ANIMAL BEHAVIOR

environments. Behavior can be refined with experience (learning), it can

change based on the environment (both physical and social) that an individual
interacts with, and it can change during development or over the lifetime of an
individual (reviewed in West-Eberhard, 2003).
In order to study the influence of genes on behavior, one must control the
environment that an individual is in. In order to study the influence of the
environment on a behavior, one must control for the particular genes in an
individual. In order to study the interaction of the genes and the environment,
one must control for all of those factors, including genes carried in social part-
ners! The study of behavior gets complicated very quickly. Behavioral scien-
tists have used several different approaches to study plasticity. To control for
environmental effects, animals are typically brought into the laboratory and
reared in climate-controlled chambers where an environment can be held con-
stant. To control for genetic effects, inbred lineages of animals are produced
that have little or no genetic variation, or geneticists use breeding designs to
provide information about relatives, and sometimes one or two particular
genes are inserted into animals (or rendered inactive) in order to test their
function.
Phenotypic plasticity occurs when the same genotype produces different
phenotypes in response to different environments. Genotype by environment
interactions (G×E interactions) arise when the same environmental
influences result in different responses from different genotypes. G×E
interactions are usually measured as the covariance between additive genetic
variation and the environment; including this type of interaction in our model
yields
z = a þ e þ 2cov(a, e) [Equation 3]
In Equation 3, the G×E interaction is encompassed by the 2cov(a, e) term
because it represents a nonrandom association between genetic and environ-
mental effects that changes the resemblance between relatives (Wolf & Moore,
2010a).
One way of describing plasticity is to measure behavior in two different
environments. The resulting regression line through the two data points is
called a reaction norm, and the slope of the reaction norm can serve as a mea-
sure of plasticity (Figure 5.2; Lynch & Walsh, 1998). In order to generate a
reaction norm, many individuals are measured because behavior is quite vari-
able, even within the same individual over time, so having repeated measures
of a behavior is important (for more on repeatability, see Boake, 1989). One
of the benefits of using inbred lines to measure behavioral traits is that both
the genetic factors and the environment are controlled so one can get repeated
Figure 5.2. Genotype by environment interaction for male attractiveness in an acoustic
moth: evidence for plasticity and canalization (Danielson-François et al., 2006). (a)
Genotypes A and B exhibit different levels of trait development in both environments
1 and 2, but the levels expressed by both genotypes do not change across environ-
ments; that is, reaction norms (solid and dashed lines) are flat, and phenotypic plas-
ticity is absent. (b) Genotypes A and B exhibit comparable reactions in development
in environment 2, that is, reaction norms are steep but parallel, and both genotypes
exhibit phenotypic plasticity. (c) Genotypes A and B exhibit different reductions in
development in environment 2, that is, reaction norms are not parallel, and a geno-
type by environment (G×E) interaction is present. (d) Reaction norms of A and B
intersect, that is, ecological crossover is present, with each genotype exhibiting the
greater trait development in only one of the two environments.
172 ANIMAL BEHAVIOR

measures of the same behavior across many different individuals that share a
particular genotype.
There are various types of phenotypic plasticity, ranging from differences
that are purely environmental to complex interactions between genes and
environments (Figure 5.2; reviewed in Schlichting & Pigliucci, 1998). When
the reaction norms of two different genotypes are not parallel, we have a
G×E interaction (Figures 5.2c and 5.2d). In some cases, a special type of
G×E interaction can occur when reaction norms cross, called ecological cross-
over (Figure 5.2d). In this case, no one genotype performs the best across all
the environments. Ecological crossover may be able to maintain genetic varia-
tion for traits under directional selection in different environments, such as
variation in courtship songs in wax moths grown under different larval den-
sities (Danielson-François et al., 2006, 2009; for a review see Ingleby et al.,
2010). If we return to Drosophila and the example of the forager gene with
its two alleles forS and forR, this is an example of ecological crossover:
low-density environments favored sitters and high-density environments
favored rovers (Sokolowski et al., 1997). Each genotype is favored under dif-
ferent environmental conditions, so ecological crossover maintains the varia-
tion at the forager locus. Note that the expression of the rover phenotype is
also plastic—it is only expressed under high-density, crowded conditions!
The interaction between genes and the environment can be even more com-
plex than what we find in G×E interactions. In some cases, the genes themselves
can respond to environments in ways that change their expression patterns in a
non-Mendelian fashion. We will examine this topic more in the section of this
chapter on epigenetics, and see Chapter 8 on epigenetics in this volume.

When Environments Have Genes of Their Own

The effects of environments and genes become even more difficult to disen-
tangle when the environment with which an individual interacts also contains
genes, namely the social environment. All animals interact with conspecifics at
least a few times throughout their lives, whether for mating, territory defense,
rearing offspring, or acquiring or competing for food. Interactions with social
partners have the potential to define or alter an individual’s behavior, creating
what are termed interacting phenotypes. For example, cockroach (Nauphoeta
cinerea) dominance status can only be seen when a cockroach is interacting
with another individual (Moore, 1991), and individuals may only be cannibal-
istic when interacting with a conspecific (Bleakley et al., 2013). Interacting
phenotypes may result from indirect genetic effects, which are the effects of
genes carried in social partners on the phenotype of a focal individual. Indirect
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 173

genetic effects appear in our quantitative genetic model as the additive effects
of genes (a), just as in Equation 1, but they are modified slightly because they
are carried in a social partner (Equation 4; Moore et al., 1997; Wolf et al.,
1999; McGlothlin et al., 2010). Including indirect genetic effects in our
model looks like this:
z = a þ e þ a 0, [Equation 4]
where the prime indicates that the genes are carried in a social partner. The
coefficient of the interaction, , determines how much the social environment
influences the behavior we observe in a focal individual. If the social environ-
ment completely determines the behavior of a focal individual, then = −1
or 1. If the social environment has no influence whatsoever on an individual’s
behavior, then = 0 and we can simplify right back to Equation 1, where the
individual’s phenotype results just from its own genes and environments
(reviewed in Bleakley et al., 2010; Wolf & Moore, 2010b).
Indirect genetic effects are harder to measure than other quantitative
genetic effects because one must control the direct effects of genes carried
in focal individuals, the effects of genes carried in social partners, and the
effects of the general environment. Guppies (Poecilia reticulata) are a model
system for understanding the evolution of cooperation because they perform
antipredator behavior, including predator inspections, in cooperative pairs or
groups. Because individual guppies respond strongly to the behavior of their
social partners, Bronwyn Bleakley and Butch Brodie predicted that indirect
genetic effects might influence their anti-predator behavior (Bleakley et al.,
2007). Inbred lines of guppies that differed genetically and in their anti-
predator were used to control the genetic component of the social environ-
ment and additive genetic effects in focal females (Bleakley et al., 2006;
Bleakley et al., 2008). Females were tested for their responses to a predator
model in the presence of one of four social “context” strains. Changes in
their behavior, analogous to measures of plasticity, can be attributed to the
only thing that changed between trials: the genes present in their social part-
ners. For several behaviors, indirect genetic effects were as strong, or nearly
so, as the effects of a fish’s own genes. Somewhat surprisingly, indirect
genetic effects explained nearly all of the variation in inspection behavior in
the guppies while a female’s own genes mattered very little (Bleakley &
Brodie, 2009).
To date, indirect genetic effects have been infrequently explicitly mea-
sured but have been shown to influence wax moth courtship and competi-
tive behavior (Achroia grisella, Danielson-François et al., 2009), Drosophila
courtship behavior (Chenoweth & Blows, 2006; Kent et al., 2008;
174 ANIMAL BEHAVIOR

Chenoweth et al., 2010), parental care in burrowing beetles (Nicrophorus

pustulatus; Rauter & Moore, 2002), dominance in red deer (Cervus elaphus;
Wilson et al., 2011) and aggression in deer mice (Peromyscus maniculatus;
Wilson et al., 2009). Many examples of interacting phenotypes that probably
result from indirect genetic effects are reviewed in Bleakley and colleagues
(2010). Maternal effects are a special case of indirect genetic effects where
genes in a mother (or father) influence the phenotype of the offspring. For
example, milk production has a strong additive genetic component in cattle.
but the effects of the milk production are seen in how quickly and how well
a calf grows. Maternal effects have been well studied, particularly in species
where offspring can be separated from their biological parent and fostered
by another parent to separate out the effects of the genes they inherited from
their parents from the effects of being reared by their parents (reviewed in
Mousseau & Fox, 1998).

Adding Up the Effects for Quantitative Traits

As we said at the beginning of this section, we can think about separating
the sources of phenotypic variation. These included additive effects of genes,
including single genes and QTL, as well as environmental sources of variation.
We accounted for dominance relationships among alleles and epistatic interac-
tions among loci, and to the extent possible, studies try to factor those sources
of genetic variance out of the equation because they are less useful for under-
standing the evolution of behavioral traits. We also had to include interactions
between genes and environments and consider the special case of indirect
genetic effects, where environments have their own genes! We turn our math-
ematical model into a diagram representing these effects in Figure 5.3. Notice
that indirect genetic effects, including maternal effects, are actually grouped in
the portion of variance explained by environmental effects. As you can see
from the interplay of all these sources of variation in behavioral traits, the
distinction between nature and nurture is getting quite murky!

Epigenetics: Linking Genes and Environments

The exciting field of behavioral epigenetics has recently emerged. and scien-
tists are studying how changes to gene expression can be inherited without
changing the underlying DNA sequence. Epigenetic changes also directly link
gene expression to environmental influences and provide a mechanism for
plasticity, G×E interactions, and potentially even indirect genetic effects (see
Mutic & Wolf, 2007, for an example of gene expression changes in response
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 175

to “social partners” in plants). Genes that are responsive to environmental

conditions can have their expression patterns changed by the environment
and (sometimes) be inherited. Genes and the environment constantly feed
back into each other. Sometimes just the act of performing a behavior can
change the regulation of genes that influence that behavior.
Recently, scientists discovered that how a mother nurtures her offspring
alters their development as well as their future parenting behavior, impacting
subsequent generations. For mammals, the postnatal period immediately after
birth is critical. Until weaning, mammalian mothers produce milk to sustain
their offspring. This period of intense mother-offspring interaction can have
a profound impact on offspring development. In rats, maternal care is respon-
sible for the female offspring’s future maternal behavior and response to stress,
mediated through oxytocin receptor gene expression (reviewed in Meaney,
2001). In fact, Frances Champagne and Michael Meaney identified a mecha-
nism for transgenerational inheritance of epigenetic marks (Champagne &
Meaney, 2007; reviewed in Champagne, 2008): Rat offspring that were

Figure 5.3. Total phenotypic variation (VP) for feeding behavior can be broken into its
constitutive parts, first the variation attributable to genetic effects (VG) and then that
attributable to environmental effects (VE). Variation that results from genetic effects
can be further broken into variation attributable to additive genetic effects (VA),
dominance effects (VD), and epistatic effects (VI). Environmental variance can be
broken into the effects of the general environment (VEG), as well as effects of the
social environment (VES), maternal effects (VMAT), and the effects of growing up in
the same environment as your relatives, known as common environmental effects
(VCOM). Last, we see the effects of G×E interactions. In this example, the amount a
mother feeds her babies reflects all of these influences, including genes carried in the
father of her offspring and those carried in her offspring (for example for begging).
176 ANIMAL BEHAVIOR

nurtured by an attentive mother were more likely to grow up to be good

mothers themselves. Offspring of an unattentive mother were more likely to
grow up to be poor mothers, but their phenotype could be “rescued” by being
fostered with a good mother. Later it was discovered that maternal nurturing
alters DNA methylation at the gene encoding the glucocorticoid receptor,
which modulates the stress response (Weaver et al., 2004). This example of
epigenetics is also another example of a G×E interaction, where the environ-
ment is the “nurturing quality” that an offspring receives.
Methylation is not the only way in which epigenetic transformations can
occur. We hope that our short introduction to the field of behavioral epige-
netics has intrigued you enough to continue on to Chapter 8 in this volume
on epigenetics.

CONCLUSION
The study of behavioral genetics has come a long way since Plato first con-
templated the interplay of nature and nurture in human behavior. While changes
in single genes can have profound effects on behavior, we now know that com-
plex interactions among genes, as well as interactions between genes and environ-
ments, determine the variation we observe in behavioral traits. The environment
is particularly difficult to distinguish from the genes when animals interact with
social partners whose genes may also influence what that animal does. And
finally, environments can alter the ways in which genes are expressed in some-
times heritable ways. These complex inheritance patterns break down the barriers
between nature and nurture and suggest that for most behaviors simply asking
“Is there a gene for it?” is not sufficient. We hope this chapter has illustrated
why behavioral geneticists believe Shakespeare got it wrong.

ACKNOWLEDGMENTS
We wish to thank Ken Yasukawa for helpful comments on the chapter.

REFERENCES AND SUGGESTED READING

Abrahams, B. S. & D. H. Geschwind. (2008). Advances in autism genetics: On the
threshold of a new neurobiology. Nature Reviews Genetics, 9, 341–355.
Anholt, R. R. H., C. L. Dilda, S. Chang, J. J. Fanara, N. H. Kulkarni, I. Ganguly,
et al. (2003). The genetic architecture of odor-guided behavior in Drosophila:
Epistasis and the transcriptome. Nature Genetics, 35,108–184.
Anway, M. D., A. S. Cupp, M. Uzumcu, & M. K. Skinner. (2005). Epigenetic trans-
generational actions of endocrine disruptors and male fertility. Science, 308,
1466–1469.
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 177

Arnold, S. (1994). Multivariate inheritance and evolution: A review of concepts. In

C. R. B. Boake (ed.), Quantitative Genetic Studies of Behavioral Evolution
(pp. 17–48). Chicago: University of Chicago Press.
Auffray, C., Z. Chen, & L. Hood. (2009). Systems medicine: The future of medical
genomics and healthcare. Genome Medicine, 1, 2.1–2.11.
Bargiello, T. A., F. R. Jackson, & M. W. Young. (1984). Restoration of circadian
behavioural rhythms by gene transfer in Drosophila. Nature, 312, 752–754.
Ben-Shahar, Y., A. Robichon, M. B. Sokolowski, & G. E. Robinson. (2002). Influence
of gene action across different time scales on behaviour. Science, 296, 741–744.
Benzer, S. (1967). Behavioral mutants of Drosophila isolated by countercurrent distri-
bution. Proceedings of the National Academy of Sciences, USA, 58, 1112–1119.
Bleakley, B. H. & E. D. Brodie III. (2009). Indirect genetic effects influence antipre-
dator behavior in guppies: Estimates of the coefficient of interaction psi and the
inheritance of reciprocity. Evolution, 63, 1796–1806.
Bleakley, B. H., A. C. Eklund, & E. D. Brodie III. (2008). Are designer guppies
inbred? Microsatellite variation in five strains of ornamental guppies, Poecilia
reticulata, used for behavior research. Zebrafish, 5, 39–48.
Bleakley, B. H., C. M. Martell, & E. D. Brodie III. (2006). Variation in anti-predator
behavior among five strains of inbred guppies, Poecilia reticulata. Behavior
Genetics, 36, 783–791.
Bleakley, B. H., D. J. Parker, & E. D. Brodie III. (2007). Nonadditive effects of
group membership can lead to additive group phenotypes for anti-predator
behaviour of guppies, Poecilia reticulata. Journal of Evolutionary Biology, 20,
1375–1384.
Bleakley, B. H., S. M. Welter, K. McCauley-Cole, S. M. Shuster, & A. J. Moore.
(2013). Cannibalism as an interacting phenotype: Pre-cannibalistic aggression
is influenced by social partners in the endangered Socorro isopod (Thermos-
phaeroma thermophilum). Journal of Evolutionary Biology, 26, 832-842.
Bleakley, B. H., J. B. Wolf, & A. J. Moore. (2010). The quantitative genetics of social
behaviour. In T. Szekely, A. J. Moore, J. Komdeur, & M. Griffiths (eds.),
Social Behaviour: Genes, Ecology and Evolution (pp. 29–54). Cambridge, UK:
Cambridge University Press.
Boake, C. R. B. (1989). Repeatability—Its role in evolutionary studies of mating
behavior. Evolutionary Ecology, 3, 173–182.
Boake, C. R. B. (1994a). Applications of quantitative genetics to behavioral evolution.
In C. R. B. Boake (ed.), Quantitative Genetic Studies of Behavioral Evolution
(pp. 305–325). Chicago: University of Chicago Press.
Boake, C. R. B. (1994b). Outlining the issues. In C. R. B. Boake (ed.), Quantitative Genetic
Studies of Behavioral Evolution (pp. 3–16). Chicago: University of Chicago Press.
Boake, C. R. B. (ed.). (1994c). Quantitative Genetic Studies of Behavioral Evolution.
Chicago: University of Chicago Press.
Brodie, E. D., III. (2000). Why evolutionary genetics does not always add up. In
J. B. Wolf, E. D. Brodie III, & M. J. Wade (eds.), Epistasis and the Evolution-
ary Process (pp. 3–19). New York: Oxford University Press.
178 ANIMAL BEHAVIOR

Brooks, R. & J. A. Endler. (2001). Female guppies agree to differ: Phenotypic and
genetic variation in mate-choice behavior and the consequences for sexual
selection. Evolution, 55, 1644–1655.
Brown, J. R., H. Ye, R. T. Bronson, P. Dikkes, & M. E. Greenberg. (1996). A defect
in nurturing in mice lacking the immediate early gene fosB. Cell, 86, 297–309.
Cade, W. H. (1984). Genetic variation underlying sexual behavior and reproduction.
American Zoologist, 24, 355–366.
Caillaud, M. C. & S. Via. (2012). Quantitative genetics of feeding behavior in two
ecological races of the pea aphid, Acyrthosiphon pisum. Heredity, 108, 211–218.
Cedar, H. & Y. Bergman. (2009). Linking DNA methylation and histone modifica-
tion: Patterns and paradigms. Nature Reviews Genetics, 10, 295–304.
Champagne, F. A. (2008). Epigenetic mechanisms and the transgenerational effects of
maternal care. Frontiers of Neuroendocrinology, 29, 386–397.
Champagne, F. A. (2011). Maternal imprints and the origins of variation. Hormones
and Behavior, 60, 4–11.
Champagne, F. A., D. D. Francis, A. Mar, & M. J. Meaney. (2003). Variations in
maternal care in the rat as a mediating influence for the effects of environment
on development. Physiology and Behavior, 79, 359–371.
Champagne, F. A. & M. J. Meaney. (2007). Transgenerational effects of social envi-
ronment on variations in maternal care and behavioral response to novelty.
Behavioral Neuroscience, 121, 1353–1363.
Chenoweth, S. F. & M. W. Blows. (2006). Dissecting the complex genetic basis of
mate choice. Nature Reviews Genetics, 7, 681–692.
Chenoweth, S. F., H. D. Rundle, & M. W. Blows. (2010). Experimental evidence for
the evolution of indirect genetic effects: Changes in the interaction effect
coefficient, psi ( ), due to sexual selection. Evolution, 64, 1849–1856.
Cheverud, J. M. (2000). Detecting epistasis among quantitative trait loci. In
J. B. Wolf, E. D. Brodie III, & M. J. Wade (eds.), Epistasis and the Evolution-
ary Process (pp. 58–81). New York: Oxford University Press.
Cho, M. M., A. C. DeVries, J. R. Williams, & C. S. Carter. (1999). The effects of
oxytocin and vasopressin on partner preferences in male and female prairie
voles (Microtus ochrogaster). Behavioral Neuroscience, 113, 1071–1079.
Ciccone, D. N., H. Su, S. Hevi, F. Gay, H. Lei, J. Bajko, G. Xu, E. Li, & T. Chen.
(2009). KDM1B is a histone H3K4 demethylase required to establish mater-
nal genomic imprints. Nature, 461, 415–418.
Cooney, C. A., A. A. Dave, & G. L. Wolff. (2002). Maternal methyl supplements in
mice affect epigenetic variation and DNA methylation of offspring. Journal of
Nutrition, 132, 2393S–2400S.
Crabbe, J., J. Belknap, & K. Buck. (1994). Genetic animal models of alcohol and
drug abuse. Science, 264, 1715–1723.
Danielson-François, A. M., J. Kelly, & M. D. Greenfield. (2006). Genotype x envi-
ronment interaction for male attractiveness in an acoustic moth: Evidence for
plasticity and canalization. Journal of Evolutionary Biology, 19, 532–542.
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 179

Danielson-François, A. M., Y. Zhou, & M. D. Greenfield. (2009). Indirect genetic

effects and the lek paradox: Inter-genotypic competition may strengthen geno-
type x environment interactions and conserve genetic variance. Genetica, 136,
27–36.
Darwin, C. (1859). The Origin of Species by Means of Natural Selection or the Preserva-
tion of Favoured Races in the Struggle for Life. New York: Penguin.
Darwin, C. (1871). The Descent of Man and Selection in Relation to Sex. London: Murray.
de Bono, M. & C. I. Bargmann. (1998). Natural variation in a neuropeptide Y recep-
tor homolog modifies social behavior and food response in C. elegans. Cell, 94,
679–689.
Donaldson, Z. & L. Young. (2008). Oxytocin, vasopressin, and the neurogenetics of
sociality. Science, 322, 900–904.
Edwards, A. C. & T. F. C. Mackay. (2009). Quantitative trait loci for aggressive
behavior in Drosophila melanogaster. Genetics, 182, 889–897.
Falconer, D. S. & T. F. C. Mackay. (1996). Introduction to Quantitative Genetics.
Harlow, UK: Pearson Prentice Hall.
Fitzpatrick, M. J. & M. B. Sokolowski. (2004). In search of food: Exploring the evo-
lutionary link between cGMP-dependent protein kinase (PKG) and behav-
iour. Integrative and Comparative Biology, 44, 28–36.
Foley, B., S. F. Chenoweth, S. V. Nuzhdin, & M. W. Blows. (2007). Natural genetic
variation in cuticular hydrocarbon expression in male and female Drosophila
melanogaster. Genetics, 175, 1465–1477.
Freeman-Gallant, C. R. & M. D. Rothstein. (1999). Apparent heritability of parental
care in savannah sparrows. Auk, 116, 1132–1136.
Fujiwara, M., P. Sengupta, & S. L. McIntire. (2002). Regulation of body size and
behavioural state of C. elegans by sensory perception and the EGL-4 cGMP-
dependent protein kinase. Neuron, 36, 1091–1102.
Galton, F. (1874). English Men of Science: Their Nature and Nurture. London: Mac-
millan & Co.
Giardina, T. J., A. Beavis, A. G. Clark, & A. C. Fiumera. (2011). Female influence on
pre- and post-copulatory sexual selection and its genetic basis in Drosophila
melanogaster. Molecular Ecology, 20, 4098–4108.
Gleason, J. M. & M. G. Ritchie. (2004). Do quantitative trait loci (QTL) for a court-
ship song difference between Drosophila simulans and D. sechellia coincide with
candidate genes and intraspecific QTL? Genetics, 166, 1303–1311.
Graham, A., M. Munday, O. Kaftanoglu, R. Page, G. Amdam, & O. Rueppell.
(2011). Support for the reproductive ground plan hypothesis of social evolu-
tion and major QTL for ovary traits of Africanized worker honey bees (Apis
mellifera L.). BMC Evolutionary Biology, 11, 95.
Grandin, T. & M. J. Deesing. (1998). Genetics and the Behavior of Domestic Animals.
San Diego, CA: Academic Press.
Guzmán-Novoa, E., G. J. Hunt, J. L. Uribe, C. Smith, & M. E. Arechavaleta-
Velasco. (2002). Confirmation of QTL effects and evidence of genetic
180 ANIMAL BEHAVIOR

dominance of honeybee defensive behavior: Results of colony and individual

behavioral assays. Behavior Genetics, 32, 95–102.
Hajkova, P., K. Ancelin, T. Waldmann, N. Lacoste, U. C. Lange, F. Cesari, C. Lee,
G. Almouzni, R. Schneider, & M. A. Surani. (2008). Chromatin dynamics
during epigenetic reprogramming in the mouse germ line. Nature, 452,
877–881.
Hoekstra, H. (2010). In search of the elusive behavior gene. In P. R. Grant & B. R.
Grant (eds.), In Search of the Causes of Evolution: From Field Observations to
Mechanisms (pp. 192–210). Princeton, NJ: Princeton University Press.
Horton, B. M. & R. L. Holberton. (2010). Morph-specific variation in baseline cor-
ticosterone and the adrenocortical response in breeding white-throated spar-
rows (Zonotrichia albicollis). Auk, 127, 540–548.
Hunt, G., G. Amdam, D. Schlipalius, C. Emore, N. Sardesai, C. Williams, O. Rueppell,
E. Guzman-Novoa, M. Arechavaleta-Velasco, & S. Chandra. (2007a). Behav-
ioral genomics of honeybee foraging and nest defense. Naturwissenschaften, 94,
247–267.
Hunt, G. J., G. V. Amdam, D. Schlipalius, C. Emore, N. Sardesai, C. E. Williams,
O. Rueppell, E. Guzmán-Novoa, M. Arechavaleta-Velasco, S. Chandra, M.
K. Fondrk, M. Beye, & R. E. Page. (2007b). Behavioral genomics of honeybee
foraging and nest defense. Naturwissenschaften, 94, 247–267.
Huxley, T. H. (1874). On the hypothesis that animals are automata, and its history.
Nature, 10, 362–366.
Ingleby, F. C., J. Hunt, & D. J. Hosken. (2010). The role of genotype-by-environment
interactions in sexual selection. Journal of Evolutionary Biology, 23, 2031–2045.
Jordan, K. W., T. J. Morgan, & T. F. C. Mackay. (2006). Quantitative trait loci for
locomotor behavior in Drosophila melanogaster. Genetics, 174, 271–284.
Kent, C., R. Azanchi, B. Smith, A. Formosa, & J. D. Levine. (2008). Social context
influences chemical communication in D. melanogaster males. Current Biology,
18, 1384–1389.
Kleiman, D. G. (1977). Monogamy in mammals. Quarterly Review of Biology, 52, 39–69.
Konopka, R. J. & S. Benzer. (1971). Clock mutants of Drosophila melanogaster. Pro-
ceedings of the National Academy of Sciences, USA, 68, 2112–2116.
Kukekova, A., L. Trut, K. Chase, D. Shepeleva, A. Vladimirova, A. Kharlamova, I.
Oskina, A. Stepika, S. Klebanov, H. Erb, & G. Acland. (2008). Measurement
of segregating behaviors in experimental silver fox pedigrees. Behavior Genetics,
38, 185–194.
Kutschera, U. & K. J. Niklas. (2004). The modern theory of biological evolution: An
expanded synthesis. Naturwissenschaften, 91, 255–276.
Leary, C. J., A. M. Garcia, & R. Knapp. (2006). Elevated corticosterone levels elicit
non-calling mating tactics in male toads independently of changes in circulat-
ing androgens. Hormones and Behavior, 49, 425–432.
Lim, M. M., Z. Wang, D. E. Olazabal, X. Ren, E. F. Terwilliger, & L. J. Young.
(2004). Enhanced partner preference in a promiscuous species by manipulat-
ing the expression of a single gene. Nature, 429, 754–757.
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 181

Lindberg, J., S. Björnerfeldt, P. Saetre, K. Svartberg, B. Seehuus, M. Bakken, C. Vilà,

& E. Jazin. (2005). Selection for tameness has changed brain gene expression
in silver foxes. Current Biology, 15, R915–R916.
Lynch, M. & B. Walsh. (1998). Genetics and Analysis of Quantitative Traits. Sunder-
land, MA: Sinauer Associates.
Mackay, T. (2009). The genetic architecture of complex behaviors: Lessons from Dro-
sophila. Genetica, 136, 295–302.
Mackay, T. F. C. (2001). The genetic architecture of quantitative traits. Annual
Review of Genetics, 35, 303–339.
Mackay, T. F. C. (2004). The genetic architecture of quantitative traits: Lessons from
Drosophila. Current Opinion in Genetics and Development, 14, 253–257.
Mackay, T. F. C., J. B. Hackett, R. F. Lyman, M. L. Wayne, & R. R. H. Anholt.
(1996). Quantitative genetic variation of odor-guided behavior in a natural
population of Drosophila melanogaster. Genetics, 144, 727–735.
Mackay, T. F. C., S. L. Heinsohn, R. F. Lyman, A. J. Moehring, T. J. Morgan,
& S. M. Rollmann. (2005). Genetics and genomics of Drosophila mating
behavior. Proceedings of the National Academy of Sciences, USA, 102,
6622–6629.
McGlothlin, J. W., A. J. Moore, J. B. Wolf, & E. D. Brodie III. (2010). Interacting
phenotypes and the evolutionary process III: Social evolution. Evolution, 64,
2558–2574.
Meaney, M. J. (2001). Maternal care, gene expression, and the transmission of indi-
vidual differences in stress reactivity across generations. Annual Review of
Neuroscience, 24, 1161–1192.
Miller, C. W. & A. J. Moore. (2007). A potential resolution to the lek paradox
through indirect genetic effects. Proceedings of the Royal Society of London, B,
274, 1279–1286.
Moore, A. J. (1991). Genetics, inheritance and social behavior. Animal Behaviour, 42,
497–498.
Moore, A. J., E. D. Brodie, & J. B. Wolf. (1997). Interacting phenotypes and the evo-
lutionary process. I. Direct and indirect genetic effects of social interactions.
Evolution, 51, 1352–1362.
Mousseau, T. A. & C. W. Fox (eds.). (1998). Maternal Effects as Adaptations. New
York: Oxford University Press.
Muller, H. J. (1973). Modern Concept of Nature. Albany, NY: SUNY Press.
Mutic, J. J. & J. B. Wolf. (2007). Indirect genetic effects from ecological interactions
in Arabidopsis thaliana. Molecular Ecology, 16, 2371–2381.
Onuma, T. A., M. Ban, K. Makino, H. Katsumata, W. Hu, H. Ando, M.-a. Fuku-
waka, T. Azumaya, & A. Urano. (2010). Changes in gene expression for
GH/PRL/SL family hormones in the pituitaries of homing chum salmon
during ocean migration through upstream migration. General and Comparative
Endocrinology, 166, 537–548.
Osborne, K. A., A. Robichon, E. Burgess, S. Butland, R. A. Shaw, A. Coulthard, H.
S. Pereira, R. J. Greenspan, & M. B. Sokolowski. (1997). Natural behavior
182 ANIMAL BEHAVIOR

polymorphism due to a cGMP-dependent protein kinase of Drosophila.

Science, 277, 834–836.
Parens, E., A. R. Chapman, & N. Press. (2006). Introduction. In E. Parens, A. R.
Chapman, & N. Press (eds.), Wrestling with Behavioral Genetics: Science, Ethics
and Public Conversation (pp. xiii–xxxv). Baltimore, MD: Johns Hopkins
University Press.
Phillips, P. C., S. P. Otto, & M. C. Whitlock. (2000). Beyond the average: The evo-
lutionary importance of gene interactions and variability of epistatc effects. In
J. B. Wolf, E. D. Brodie III, & M. J. Wade (eds.), Epistatis and the Evolution-
ary Process (pp. 20–38). New York: Oxford University Press.
Rauter, C. M. & A. J. Moore. (2002). Evolutionary importance of parental care per-
formance, food resources, and direct and indirect genetic effects in a burying
beetle. Journal of Evolutionary Biology, 15, 407–417.
Reinhold, K., K. Reinhold, & K. J. Jacoby. (2002). Dissecting the repeatability of
female choice in the grasshopper Chorthippus biguttulus. Animal Behaviour,
64, 245–250.
Rozen, F., C. Russo, D. Banville, & H. H. Zingg. (1995). Structure, characterization,
and expression of the rat oxytocin receptor gene. Proceedings of the National
Academy of Sciences, USA, 92, 200–204.
Schlichting, C. D. & M. Pigliucci. (1998). Phenotypic Evolution: A Reaction Norm
Perspective. Sunderland, MA: Sinauer Associates.
Scott Schneider, S., G. DeGrandi-Hoffman, & D. R. Smith. (2004). The African
honey bee: Factors contributing to a successful biological invasion. Annual
Review of Entomology, 49, 351–376.
Shields, C. (2011). Aristotle’s psychology. In E. N. Zalta (ed.), The Stanford Encyclo-
pedia of Philosophy. Metaphysics Research Lab, Center for the Study of
Language and Information, Stanford University. http://plato.stanford.edu/
archives/spr2011/entries/aristotle-psychology/.
Shimomura, K., S. S. Low-Zeddies, D. P. King, T. D. L. Steeves, A. Whiteley,
J. Kushla, P. D. Zemenides, A. Lin, M. H. Vitaterna, G. A. Churchill, &
J. S. Takahashi. (2001). Genome-wide epistatic interaction analysis reveals
complex genetic determinants of circadian behavior in mice. Genome Research,
11, 959–980.
Shuster, S. M. (1987). Alternative reproductive behaviors: Three discrete male
morphs in Paracerceis sculpta, an intertidal isopod from the northern Gulf of
California. Journal of Crustacean Biology, 7, 318–327.
Shuster, S. M. & C. Sassaman. (1997). Genetic interaction between male mating
strategy and sex ratio in a marine isopod. Nature, 388, 373–377.
Skinner, B. F. (1984). An operant analysis of problem solving. Behavioral and Brain
Sciences, 7, 583–591.
Sokolowski, M. B. (1980). Foraging strategies of Drosophila melanogaster: A chromo-
somal analysis. Behavior Genetics, 10, 291–302.
Sokolowski, M. B. (2001). Drosophila: Genetics meets behaviour. Nature Reviews, 2,
879–890.
BEHAVIORAL GENETICS : BEYOND NATURE AND NURTURE 183

Sokolowski, M. B. & J. D. Levine. (2010). Nature-nurture interactions. In

T. Szekely, A. J. Moore, & J. Komdeur (eds.). Social Behaviour: Genes, Ecology
and Evolution (pp. 11–28). Cambridge, UK: Cambridge University Press.
Sokolowski, M. B., H. S. Pereira, & K. Hughes. (1997). Evolution of foraging behav-
ior in Drosophila by density dependent selection. Proceedings of the National
Academy of Sciences, USA, 94, 7373–7377.
Soma, K. K., M.-A. L. Scotti, A. E. M. Newman, T. D. Charlier, & G. E. Demas.
(2008). Novel mechanisms for neuroendocrine regulation of aggression.
Frontiers in Neuroendocrinology, 29, 476–489.
Toh, K. L., C. R. Jones, Y. He, E. J. Eide, W. A. Hinz, D. M. Virshup, L. J. Ptacek,
& Y. H. Fu. (2001). An hPer2 phosphorylation site mutation in familial
advanced sleep phase syndrome. Science, 291, 1040–1043.
Tronche, F., C. Kellendonk, O. Kretz, P. Gass, K. Anlag, P. C. Orban, R. Bock,
R. Klein, & G. Schutz. (1999). Disruption of the glucocorticoid receptor gene
in the nervous system results in reduced anxiety. Nature Genetics, 23, 99–103.
Trut, L. (1999). Early canid domestication: The farm-fox experiment. American
Scientist, 87, 160–169.
Vosshall, L. (2007). Into the mind of a fly. Nature, 450, 193–197.
Waddington, C. H. (1942). Canalization of development and the inheritance of
acquired characters. Nature, 150, 563–565.
Waddington, C. H. (1953). Genetic assimilation of an acquired character. Evolution,
7, 118–126.
Walum, H., L. Westberg, S. Henningsson, J. M. Neiderhiser, D. Reiss, W. Igi, J. M.
Ganiban, E. L. Spotts, N. L. Pedersen, E. Eriksson, & P. Lichtenstein. (2008).
Genetic variation in the vasopressin receptor 1a gene (AVPR1A) associates
with pair-bonding behavior in humans. Proceedings of the National Academy
of Sciences, USA, 105, 14153–14156.
Wang, Y., G. V. Amdam, O. Rueppell, M. A. Wallrichs, M. K. Fondrk, O. Kaftano-
glu, & R. E. Page, Jr. (2009). PDK1 and HR46 gene homologs tie social
behavior to ovary signals. PLoS ONE, 4, e4899.
Weaver, I. C., N. Cervoni, F. A. Champagne, A. C. D’Alessio, S. Sharma, J. R. Seckl,
S. Dymov, M. Szyf, & M. J. Meaney. (2004). Epigenetic programming by
maternal behavior. Nature Neuroscience, 7, 847–854.
Weaver, I. C., F. A. Champagne, S. E. Brown, S. Dymov, S. Sharma, M. J. Meaney,
& M. Szyf. (2005). Reversal of maternal programming of stress responses in
adult offspring through methyl supplementation: Altering epigenetic marking
later in life. Journal of Neuroscience, 25, 11045–11054.
West-Eberhard, M. J. (2003). Developmental Plasticity and Evolution. New York:
Oxford University Press.
Whitfield, C.W., A. M. Cziko, & G. E. Robinson. (2003). Gene expression profiles in
the brain predict behaviour in individual honey bees. Science, 302, 296–299.
Wilson, A. J., U. Gelin, M.-C. Perron, & D. Réale. (2009). Indirect genetic effects
and the evolution of aggression in a vertebrate system. Proceedings of the Royal
Society of London, B, 276, 533–541.
184 ANIMAL BEHAVIOR

Wilson, A. J., M. B. Morrissey, M. J. Adams, C. A. Walling, F. E. Guinness, J. M.

Pemberton, T. H. Clutton-Brock & L. E. B. Kruuk. (2011). Indirect genetics
effects and evolutionary constraint: An analysis of social dominance in red
deer, Cervus elaphus. Journal of Evolutionary Biology, 24, 772–783.
Winslow, J., N. Hastings, C. S. Carter, C. Harbaugh, & T. Insel. (1993). A role for
central vasopressin in pair bonding in monogamous prairie voles. Nature,
365, 545–548.
Wolf, J. B. (2000). Indirect genetic effects and gene interactions. In J. B. Wolf, E. D.
Brodie III, & M. J. Wade (eds.), Epistasis and the Evolutionary Process (pp.
158–176). New York: Oxford University Press.
Wolf, J. B., E. D. Brodie, & A. J. Moore. (1999). Interacting phenotypes and the
evolutionary process. II. Selection resulting from social interactions. American
Naturalist, 153, 254–266.
Wolf, J. B. & A. J. Moore. (2010a). A brief introduction to quantitative genetics.
In D. F. Westneat & C. W. Fox (eds.), Evolutionary Behavioral Ecology
(pp. 75–79). New York: Oxford University Press.
Wolf, J. B. & A. J. Moore. (2010b). Interacting phenotypes and indirect gene-
tic effects: A genetic perspective on the evolution of social behavior. In
D. F. Westneat & C. W. Fox (eds.), Evolutionary Behavioral Ecology (pp.
225–245). New York: Oxford University Press.
Young, K. A., Y. Liu, & Z. Wang. (2008). The neurobiology of social attachment: A
comparative approach to behavioral, neuroanatomical, and neurochemical
studies. Comparative Biochemistry and Physiology Part C: Toxicology & Pharma-
cology, 148, 401–410.
Young, L. J., B. Huot, R. Nilsen, Z. Wang, & T. R. Insel. (1996). Species differences
in central oxytocin receptor gene expression: Comparative analysis of promoter
sequences. Journal of Neuroendocrinology, 8, 777–783.
Young, L. J., R. Nilsen, K. G. Waymire, G. R. MacGregor, & T. R. Insel. (1999).
Increased affiliative response to vasopressin in mice expressing the V1a receptor
from a monogamous vole. Nature, 400, 766–768.
Zehring, W. A., D. A. Wheeler, P. Reddy, R. J. Konopka, C. P. Kyriacou, M. Rosbash,
& J. C. Hall. (1984). P-element transformation with period locus DNA restores
rhythmicity to mutant, arrhythmic Drosophila melanogaster. Cell, 39, 369–376.
 6

Neural Systems, Behavior, and Evolution:

Integrating Proximate and Ultimate
Analyses of Behavior
Laura Smale

INTRODUCTION
When we consider what behavioral neuroscience is about, one might start
with the question: What is it that sets a slug, a lamprey, and an elephant apart
from life forms outside of the animal kingdom, such as slime molds, a bush of roses,
a mushroom, and blue-green algae?
In my view, the most interesting difference that sets animals distinctly apart
from the other forms of life is the system most directly specialized to coordi-
nate behavior: the nervous system. It comes in many forms and can be as sim-
ple as a net of interconnected neurons that allows a sea anemone to move its
tentacles. But a nervous system can also be unimaginably complex and can
include a brain, such as those of birds and mammals, that has many organized
populations of neurons, each talking to others through what adds up to astro-
nomical numbers of junctions called synapses. This communication is at the
root of all behavior including those behaviors that form the basis of networks
of social relationships and the cognitive processes and emotional experience
associated with such behavior. The nervous system is at the center of our inter-
actions with the world around us. It allows us to respond to, and anticipate,
changes in our environment. The nature of the relationships between the
186 ANIMAL BEHAVIOR

patterns of interconnectedness among those neurons and behavior, cognition,

and emotion is what behavioral neuroscience is about. In this chapter I hope
to provide you with some new insights into the nature of those relationships.
I will begin by providing an extremely brief and selective review of some of
the basic elements and structures of the vertebrate brain. I will then describe
in greater depth two areas in the study of brain-behavior relationships that I
believe are particularly illustrative of how neuroscience can inform us about
basic issues in the more general study of animal behavior. But first, to under-
stand that objective more clearly, it pays to think for a moment about Niko
Tinbergen.
In 1960 Tinbergen wrote about the distinction between what are referred
to now as ultimate and proximate levels of analysis in efforts to understand
behavior. He saw the former as focusing on the evolutionary processes that
have operated over many generations to shape behavior patterns typically seen
in all members of a species, while the latter addresses questions about processes
regulating behavior within the lifespan of an individual. Investigators typically
focused on questions about behavior at one level of analysis or the other, not
both. For decades, little attention was paid to another important point that
Tinbergen had made: to fully understand behavior we must understand rela-
tionships among levels of analysis. At that time, he could say little else about
this integration because the field was simply not yet at that point in its devel-
opment. That has changed. The two examples of behavioral neuroscience that
I have chosen to emphasize below are ones that can illustrate efforts to under-
stand relationships between the processes that have shaped the evolution of
behavior and the mechanisms that operate within a lifespan to regulate the
expression of that behavior. More specifically, the issues that I will consider
are (1) How have brain mechanisms supporting social and sexual bonds been
altered at evolutionary transitions that led from one mating system to another?
and (2) How does a brain produce the rhythms in behavior that occur across a
24-hour day, and how did those neural mechanisms change as animals underwent
evolutionary transitions from a nocturnal to a diurnal niche? These two very dif-
ferent issues have in common a focus on an integration of proximate and ulti-
mate levels of analysis, and they illustrate how an understanding of the neural
mechanisms regulating behavior can inform both phylogenetic and functional
perspectives on its evolution, and vice versa. One of my goals here is to con-
vince you, even those of you that are most taken by evolutionary processes,
that an understanding of brain-behavior relationships is an essential and fasci-
nating part of behavioral biology. However, I will begin with an extremely
brief introduction to some basics of the nervous system in order to provide
some context for our discussions of bonds and rhythms; more details can be
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 187

found in a variety of introductory textbooks in behavioral neuroscience

(e.g., Bear et al., 2007; Breedlove et al., 2007).

BRAIN BASICS
The most basic, elemental units of the mammalian nervous systems are the
neurons and the glial cells. The latter get less attention, as reflected in their
name, which comes from the Greek work for “glue.” Glial cells actually do
much more than stick things together; they also help neurons send their mes-
sages more rapidly, modulate their development and their structure, provide
them with some molecules, and protect them from others. Nonetheless, glial
cells tend to be overlooked, primarily because they do not fire action poten-
tials, sometimes referred to as nerve impulses. Neurons use these to carry mes-
sages in a binomial code from one cell to another along extensions from the
cell body that are referred to as axons. While there are only four basic types
of glial cells, there are hundreds of distinct types of neurons. Their structures,
compositions, and interconnections enable them to receive and integrate
incoming signals and to transform them into new patterns that they send to
an array of target cells near and far. The signals that transmit information from
one neuron to another are molecules referred to as neurotransmitters, whose
effects are rapid and short lasting, and neuromodulators, whose effects are
slower and last for longer periods of time. Examples include oxytocin (OT),
vasopressin (AVP), and dopamine (DA), all of which will come up in our dis-
cussion of the formation of social bonds (see also Chapter 7 in this volume).
Transmitters can have either excitatory effects on another neuron, increasing
the rate at which they fire their own action potentials, or inhibitory ones that
decrease the rate of such signals.
The vertebrate nervous system contains a peripheral component at the
interface of the animal and its environment (i.e., the peripheral nervous sys-
tem) that transmits signals into and out of its central component, the spinal
cord and brain (i.e., the central nervous system). Circuits within the spinal
cord are complex and fascinating and can do quite sophisticated things with-
out the brain. Interactions between central pattern generators there, for
example, coordinate the rhythms in contraction and relaxation of a host of
limb muscles in a manner that produces the different gaits of a four-legged ani-
mal. But ordinarily these spinal circuits do not act alone. The brain, which
receives special patterns of signals that emerge from circuits within the spinal
cord, also sends signals to the populations of cells and pattern generators
within the spinal cord that can initiate them, modulate them, and shut them
down.
188 ANIMAL BEHAVIOR

The vertebrate brain can be thought of as consisting of three divisions: the

hindbrain, midbrain, and forebrain. We will spend most of our time
focused on the latter two, but it pays to put it all in context. The hind portion
of the brain, continuous with the spinal cord but inside the skull, contains
structures essential for life that keep us breathing and our hearts beating.
It also includes populations of cells that wake us up and put us on high alert,
some of which release an adrenalin-like neurotransmitter, norepinephrine,
from terminals that reach widely distributed regions of the forebrain. Another
hindbrain structure is referred to as the cerebellum because it looks a bit like a
little brain at the back end of a big brain. It is essential for balance and coordi-
nated movement and is also where the experience of classical conditioning can
lead to changes in circuits that underlie lasting changes in behavior (i.e., where
a memory can be stored). Some of these basic structures are beautifully illus-
trated at http://rws2012.files.wordpress.com/2012/01/mid-brain-labels.png.
The midbrain contains a “tectum” or roof, which, when viewed from
above, appears to have four little hills, or “colliculi,” on it; the front (rostral)
two are referred to as the optic tectum and the back (caudal) two as the audi-
tory tectum. As is suggested by the names, these are structures where visual
and auditory information, respectively, are processed. In frogs the midbrain
tectum is huge relative to the rest of the brain, but in mammals it is relatively
small, as other portions of the forebrain have expanded and taken sensory
processing to whole new levels. Beneath the tectum lies the tegmentum, the
“floor” of the midbrain, where we find a structure that plays a major role in
reward and motivation, the ventral tegmental area (VTA). This population
of cells has axons extending into several forebrain structures, including the
nucleus accumbens (NAc), where their terminals release DA. This signal mod-
ulates behaviors such as eating, caring for offspring, and sex; it is sometimes
described as having a more general and basic role in motivation. The VTA also
plays a key role, as we will see, in mediation of a suite of behaviors that
together represent a social bond. Another structure contained within the mid-
brain is the tuberomamilary nucleus, whose cells help maintain wakefulness
during the night in nocturnal species and during the day in diurnal ones.
Moving on to the forebrain, we find the largest part of the mammalian cen-
tral nervous system. It is divided into two regions on the basis of their patterns
of development: the diencephalon and the telencephalon. One could spend
several lifetimes learning about these highly interconnected structures, but
here I will simply highlight a few things about them.
In the diencephalon, we find the hypothalamus, a relatively small structure
that has a major impact on a far-ranging suite of behaviors. It influences what
and how much animals eat and drink, how and with whom they copulate, and
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 189

at what time of day they are most likely to engage in all of these behaviors (see
Chapter 11 in this volume). The primary hypothalamic structure that contrib-
utes to the latter is referred to as the suprachiasmatic nucleus, or SCN. The
hypothalamus also includes one population of cells specialized to stimulate
wakefulness and another that puts us to sleep. Both are heavily influenced by
the SCN, and both exhibit daily rhythms that are in an antiphase relationship
in diurnal and nocturnal species. The other division of the forebrain, the telen-
cephalon, contains the NAc, noted earlier, that helps motivate animals to
engage in a wide range of behaviors. The NAc, like much of the brain, also
changes in systematic and predictable ways from day to night. The positions
of the VTA and the NAc, and the connections between them, are depicted
here in the human brain: http://webspace.ship.edu/cgboer/limbicsystem.gif.
Finally, the forebrain is capped by the cortex, with its “paleo” (old) and “neo”
(new) divisions that are distinctly different with respect to their structures and
functions. The paleocortex includes the hippocampus, a well-studied structure
critically involved in the formation of some kinds of memories. The neocortex
is a six-layered structure in all mammals, but its size varies wildly among them.
Its organization includes “modules” with astonishingly specific functions that
are organized into distributed circuits that bring those modules together in ways
that support integrated and complex behaviors and perceptions. Examples of
these discrete modules include one whose activity produces laughter and another
in which individual neurons fire in response to specific faces but not anything
else. The distributed nature of cortical function is reflected in interconnected
regions that figure prominently in language production and comprehension;
the pathways between some of them are massive in humans and minuscule in
our close relatives, chimpanzees, with their more limited language abilities.

SOCIAL SYSTEMS
The evolution and diversification of social behavior, and the social relation-
ships that follow from it, are based on changes in the brain mechanisms that
lead animals to interact with others in particular ways. A female spotted hyena
(Crocuta crocuta), for example, has a brain that enables her to establish domi-
nance relationships with many other individuals; however, like most other
mammals, her brain does not have within it the neural mechanisms that
would enable her to establish a pair bond with any given individual. There
are, however, some mammals that engage in a suite of behaviors associa-
ted with a monogamous mating system. The question that I want to turn to
now is how their brains support these behaviors, as I believe that this is one
of the most fascinating stories to emerge from behavioral neuroscience in
190 ANIMAL BEHAVIOR

recent times. This work has centered on four species of North American voles,
the prairie (Microtus ochrogaster) and pine (M. pinetorum) voles, which are rel-
atively monogamous, on the one hand, and the meadow (M. pennsylvanicus)
and montane (M. montanus) voles, which are far more polygynous, on the
other. The story has led to a remarkable level of understanding of relationships
among small snippets of DNA, brain structure and function, and, ultimately,
patterns of behavioral interactions between individuals. However, it began
with fieldwork focused on relationships between social behavior and ecology
(see also Volume 2, Chapter 3).
In the 1970s and 1980s ecologists were conducting trapping studies of
small rodents in their natural habitats, characterizing distributions and den-
sities of populations and how they changed over time and space. Such studies
revealed substantial differences among the different species of voles, particu-
larly during their breeding seasons. Among meadow voles, found across a large
swath of North America from the East Coast well into Alaska (Hoffmann &
Koeppl, 1985), males have home ranges that are about five times larger than
those of females (Gaulin & FitzGerald, 1989), and they nest alone. Bonds
between the sexes appear to be fleeting at best, lasting just long enough for a
male to pass his sperm along to the female. These basic patterns of social
behavior are also evident in the closely related montane vole. Among prairie
voles, from the central grasslands of North America, it is not so simple. While
all populations exhibit characteristics typical of a monogamous mating system,
some of the features of their social behavior vary from one population to
another. It is in Illinois that pair bonds are most evident in prairie voles.
Among these animals, copulation is not a brief and simple matter. If all goes
well when unpaired males and females meet up, they will interact in a friendly
manner and explore each other’s scents. Until this point the female has never
experienced a period of estrus, but following contact between her vomeronasal
organ and the male’s urine, her hormones begin to flow. Within two to three
days her uterus doubles or triples in size; copulation begins, and it keeps going,
intermittently, for two to three days. The male and female move in together,
their home ranges become virtually identical, and they raise offspring together
(Carter et al., 1980, 1988; Dluzen et al., 1981; Gavish et al., 1983; Hofmann
et al., 1984; Gaulin & FitzGerald, 1989). Though less is known about the
pine voles, they too are socially monogamous, and many basic features of their
social behavior are the same. These four microtine rodents have come to be
used as “models” for the study of the neural substrates of the behaviors that
form the basis of divergent social and mating systems.
Neuroscientists have explored the substrates underlying an array of behav-
iors that distinguish these voles from each other. This has included behaviors
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 191

associated with the formation of preferences of animals for their mates, as well
as those coordinating the use of space, territoriality, and paternal care. With
respect to the latter, for example, the prairie vole males help build the nest,
huddle over the pups in that nest, groom the pups, and return them to the
nest when they have wandered too far. They contribute to all forms of parental
behavior except nursing (Oliveras & Novak, 1986). The brain mechanisms
coordinating such behaviors in the male prairie vole have been studied exten-
sively. This has not been the case for the meadow vole for the simple reason
that, as noted above, the males of this species do none of this (Gruder-
Adams & Getz, 1985) as they are too busy wandering far and wide in search
of prospective mates. For them, a more important adaptation is the ability to
find their way to these females. In what follows I will not try to tell you about
the stories that have emerged from studies of brain mechanisms supporting all
of the adaptations distinguishing the sociality of monogamous and polygynous
voles. I will, instead, zero in on two of them: the ability of the meadow vole
to navigate through the territories of multiple females, and the ability of the
male and female prairie voles to form pair bonds involving lasting partner
preferences.

Space
To be successful, a male meadow vole has to remember how to find his way
to each of several females, and those spatial memories have to be integrated
with memories of the ever-changing reproductive conditions of these females.
Females of this species simply do not face those challenges. Among the prairie
and pine voles the male home range is the same as his mate’s, and the demands
for spatial memory are consequently quite similar between the sexes (Gaulin
et al., 1990). So, the question here is, what is it about their brains that enables
the male meadow vole to find his way through a home range that can encompass
those of several females, and how might this be different in their pair-bonding
counterparts?
Exploration of this issue began with a direct examination of the spatial abil-
ities of male and female meadow and prairie voles. After their work in the field
had allowed Steven Gaulin and Randall FitzGerald (1989) to measure the size
of the animals’ home ranges, they brought the voles into the lab where they
could assess their spatial abilities. The animals were placed into a standardized
series of mazes, each with the promise of food at the end, if they were able to
find their way there. This work revealed that, indeed, performance was quite
similar in male and female prairie voles, whereas male meadow voles were con-
siderably better than their female counterparts. This finding was subsequently
192 ANIMAL BEHAVIOR

extended to tests of lab-born voles in a Morris water maze (Kavaliers et al.,

1998); here, an animal is dropped into a large tank of opaque water and has
to learn to locate a hidden platform and remember where that platform is.
So the question became, what is it in the brains of these animals that might
account for the differences in their spatial abilities? The answers came from
examination of the size of the hippocampus, which is depicted here in a
human brain: http://webspace.ship.edu/cgboer/limbicsystem.gif. An under-
standing of relationships between this forebrain structure and memory forma-
tion had already begun in the 1950s with reports of a man, known as HM,
who after suffering severe damage to his hippocampus also experienced strik-
ing impairments in his ability to form new memories of events that occurred
after the hippocampal damage. Since that time, the hippocampus has been at
the center of much animal research into the neural basis of memory, and
numerous lines of evidence from many species have converged to implicate
this area of the brain in the establishment of new memories (reviewed in
Nakazawa et al., 2004; Squire, 2004). But not all memories are the same.
HM could learn with repetition to perform new procedures, but he could
never remember that he had actually done the procedure before. Lesions of
the hippocampus of laboratory animals have relatively little impact on some
forms of learning, such as passive avoidance or classical conditioning, but they
have a devastating impact on the ability of rodents to form new spatial memo-
ries. This was first discovered by John O’Keefe and Lynn Nadel (1978), who
used a “radial arm maze,” a contraption with eight arms extending outwards
from a central chamber. Food is placed at the end of each arm, and the chal-
lenge for the animal as it searches for the food is to remember which arms it
has already traveled down so that it can go directly to the remaining ones that
still have food waiting there. Lesions of the hippocampus rendered an animal
completely incapable of remembering anything about where it had been in
the maze and where it should go next. Since that first report, innumerable
studies of the hippocampus have converged to provide a truly remarkable level
of understanding of its role in memory processes. Recently, for example,
molecular methods have been used to insert or delete genes within hippocam-
pal cells, which have dramatically altered the animals’ ability to navigate the
Morris water maze (e.g., Nicholas et al., 2006; Heldt et al., 2007).
Let us go back to voles and to a time before this form of molecular magic
had begun. Lucia Jacobs and colleagues (1990) were beginning to wonder
about the evolution of sex differences in behavior and their underlying neural
substrates. Jacobs (1996, p. 83) wrote, “success in scramble competition
would be improved with spatial navigation skills to find receptive mates and,
once found, with enhanced spatial memory of their locations. If learning
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 193

abilities are sexually selected then we would predict that, within a given spe-
cies, sex differences in the learning ability and its neural basis should be
enhanced in the sex that competes more strongly for mates.” The sex differ-
ence in learning abilities had already been described in the meadow voles, so
Jacobs and colleagues (1990) went on to test the prediction that these differ-
ences were associated with hippocampal size, structure, or function. This is
exactly what they found. Specifically, the male meadow voles had a signifi-
cantly larger hippocampus than did the females. This sex difference was not
seen in the monogamous pine vole, in which there was no sex difference in
home-range size or spatial learning abilities.
An enhancement of spatial memory has occurred in different contexts and
in multiple independent lineages, raising the question of whether these
changes were brought about via transformation of the same underlying neural
mechanisms. The answer appears to be “yes.” One interesting example has
come from studies of brown-headed cowbirds (Molothrus ater), a species in
which females behave as parasites of sorts, depositing their eggs in the nests
of unwitting “hosts” of other species that then raise their offspring for them.
For this strategy to be effective, female cowbirds must find and remember
locations of appropriate host nests, and they must integrate that memory with
their assessment of when those nests will be ready to have new eggs deposited
in them. Male cowbirds do not face these challenges, and in this species the
hippocampus of the female is substantially larger than that of the male (Sherry
et al., 1993). Another example comes from birds that stash their food in prepa-
ration for the arrival of winter and must later, sometimes much later, find
exactly where they have hidden it (e.g., Lucas et al., 2004). These birds have
a better spatial memory and a larger hippocampus than do sister species that
face no such demands (see Volume 3, Chapter 10).
Taken together, the data appear to suggest that independent evolutionary
routes towards enhanced spatial memory, whether it is in the service of nest
parasitism, storage of food, or finding females, have involved transformation
of the same basic brain mechanisms. That is, regardless of the selection pres-
sures or the phylogeny, the evolutionary transition has involved the same
changes in the proximate neural mechanisms.

Partner Preferences and Pair Bonds

Much of the more recent work on the brain mechanisms differentiating
members of what have come to be thought of as monogamous and polygynous
voles has focused on the formation of bonds between males and females.
As noted above, pairs of male and female prairie voles share a nest within their
194 ANIMAL BEHAVIOR

virtually identical, and otherwise exclusive, home ranges. These findings led
Sue Carter and her colleagues to wonder if the males and females shared a
basic social and sexual bond that held them together. This question was more
difficult to test in the field, where systematic observations of these relatively
nocturnal rodents are difficult. Therefore breeding colonies of prairie voles
from Illinois were established, and the animals could then be tested in the
lab as well as in seminatural enclosures (Carter et al., 1995). This is when it
became clear that males contribute to all forms of parental care except nursing.
Lab studies also revealed that a mere six hours of cohabitation with a female
could cause an increase in aggressive behavior directed by a male towards unfa-
miliar animals. In the field this presumably helps prevent other males from
mating with the newly acquired partner, and after she has been impregnated,
such aggression could help prevent a mates’ pregnancy from being interrupted
by contact with an unfamiliar male, something that can occur for as long as
17 days postconception in this species. Overall, the patterns of aggression
and space use, as well as paternal care, all suggest that pair bonding may be a
critical feature of prairie vole sociality, but is there an actual preference for the
putative partner over others?
Development of a partner-preference test was the first step in what has
become a rich and fascinating line of research into the neural basis of relation-
ships between cohabiting male and female voles (Williams et al., 1992). The
test is conducted in a three-chambered arena with a central compartment con-
nected with open tubes to two other chambers, one on each side. The test ani-
mal is placed in the center, and two stimulus animals are tethered within the
chambers connected to it. The test animal has no experience with one of the
animals but has varying degrees and types of experience with the other. Over
a three-hour period under these conditions, observers monitor where the test
animal spends its time and how much of that time is spent in side-by-side con-
tact with the stimulus animal. The answer turned out to depend on the
amount and type of experience the animals had had with each other. If a test
female had a mere six hours of experience with a male, and the two had not
copulated, then she showed no particular preference for him. If, however,
she had spent a full 24 hours with him and they had copulated, then she did
subsequently choose to spend most of her time snuggling up to him. The same
basic pattern was seen with respect to the development of a male preference for
a particular female. What made this particularly important was the fact that
development of such partner preferences was not seen in the polygynous
meadow voles.
Pair bonds have been defined more broadly than partner preferences as “a
stable relationship between a breeding pair of animals that share a common
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 195

territory and parental duties” (Wang & Aragona, 2004, p. 319). The patterns
of behavior described above provide good evidence for the development of pair
bonds between adult male and female prairie voles, and similar studies have
yielded evidence that this is the case in pine voles as well. This is clearly not
the case for meadow voles or montane voles, species in which male and female
promiscuity reigns—territories of males are far larger than those of females,
males provide no care for their young, and there appears to be no preference
for members of the opposite sex with which an animal has mated (Oliveras
& Novak, 1986). The patterns revealed by these studies, in the field as well
as the lab, have provided the foundation for over 20 years of work into the
neural and endocrine mechanisms of pair bonding. Here, I focus on the
research into one element of that bond: partner preferences.
Investigation of these mechanisms began with a focus on two peptides in
the brains of these animals, OT and AVP (see also Chapter 7 in this volume).
Both of these strings of nine amino acids are hormones, released by the pitui-
tary gland into the general circulation; the same peptides also function as neu-
romodulators produced within the brain and operating as signals traveling
across synapses from one neuron to another. Efforts to understand pair bonds
were focused on OT and AVP because these peptides had been associated with
several behaviors related to formation of social, emotional, and sexual bonds
between individuals (reviewed in Donaldson & Young, 2008). For example,
the surge in OT release from the pituitary that occurs at parturition contrib-
utes not only to uterine contractions that help the birth process along but also
to the development of general patterns of maternal behavior in laboratory rats
(Rattus norvegicus) and of bonds that develop between female sheep and their
newborn offspring. Another association between OT and bonding is its rise
in the bloodstream of women as they experience orgasm, something that has
been recorded in the context of sexual intercourse. AVP was known to con-
tribute to other behaviors associated with formation of bonds between males
and females, such as the rise in aggressive behavior noted above (Ferris & Pote-
gal, 1988). It is against this background that investigators developed the
hypothesis that OT and AVP might play a role in establishment of partner
preferences.
The hypothesis was first tested in prairie voles by determining if intracranial
administration of OT could influence the establishment of partner preferences
in the paradigm described above (Williams et al., 1994). Cannulae were
implanted that could deliver OT or vehicle (i.e., a control substance) into
the third ventricle of the brain of females during a period of cohabitation with
a male. The ovaries had been removed so there was no mating involved. When
tested after six hours, females showed a clear preference for the male that they
196 ANIMAL BEHAVIOR

had spent that time with, but only if they had received the OT infusions
during that time! Among males treated in the same manner, there was no
effect of OT on partner preference, but there was an effect on another element
of the pair bond: aggression directed towards outsiders. That is, infusions of
OT during the period of cohabitation led to an increase in aggression directed
towards other adult males.
The next question was a simple one: could behavioral differences between spe-
cies with and without pair bonds emerge from differences in where OT or AVP act
in the brain? This issue was initially investigated by incubating slices of brain
tissue in a solution that contained a form of the peptide with a radioactive
label attached it (Insel & Shapiro, 1992; Insel et al., 1994). The peptide binds
to its receptors in the brain, film is placed over the sections, and the radioactiv-
ity leaves its imprint. When the film is developed, one can effectively see
where the receptors for that peptide are located. Those first studies revealed
stunning differences between pair-bonding and polygynous species with
respect to the distributions and densities of the binding. The differences were
so large and so clear that one could hold the images up to the light and see
from them what type of social system the animals would have had in nature!
One of the early follow-up questions was whether the differences in recep-
tor distribution were due to differences in translation or in posttranslational
regulatory processes. The latter hypothesis was ruled out by examination of
the distribution of the mRNA for the different forms of the AVP and OT
receptors in males and females of the different species (Young et al., 1996,
1997). These studies, conducted using a procedure known as in situ hybridi-
zation, revealed that whereas sex differences were minimal, the species differ-
ences were striking. This approach also enabled investigators to see more
precisely where the cells are and to distinguish different forms of the receptors.
One striking difference was in the ventral pallidum, where V1a receptor den-
sity was considerably higher in the monogamous voles than the polygynous
ones. This region, located in the telencephalon, receives input from the
NAc, which, as noted earlier, receives dopaminergic input from the VTA. This
pathway is part of a circuit involved in reward associated with a number of
highly motivated, and addictive, behaviors. There was also a large difference
in the NAc, in that OT receptor mRNA was elevated in the pair bonding
compared to the polygynous voles. It started to become easy to imagine how
the distribution of these receptors might be able to explain the different social
systems of these voles. However, correlational data can only go so far. A true
test of the hypothesis that evolutionary changes in these peptidergic systems
contributed to the current differences in social behavior would require exper-
imental manipulations of the DNA.
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 197

The first step was to sequence the genes for the receptors, something that
has become virtually routine these days. When this was done, it became clear
that the coding regions are extremely similar across species, which means that
the receptor proteins themselves were essentially identical (Hammock &
Young, 2002). However, the regulatory regions were quite different, and these
are what determine where, when, and how much the gene will be transcribed.
These studies revealed, for example, that the promoter region of the V1aR
gene of prairie and pine voles contains several repetitive sequences that are
simply absent in the DNA of the meadow and montane voles. The next ques-
tion was, could the differences in these little snippets of DNA be responsible for
differences in the distribution of the mRNA for the receptors and, ultimately, for
the different patterns of social interactions seen between voles that establish pair
bonds and those that do not? With the sequences in hand, it became possible
to test this hypothesis by experimentally altering DNA and examining effects
that this might have on partner preferences.
Molecular methods that made it possible to insert and delete segments of
DNA in mice had been developed by 1999 to the point where Larry Young
and colleagues (1999) could actually create transgenic mice with the prairie
vole V1aR sequence. In the brains of these animals the distribution of the
receptor was actually similar to that seen in prairie voles! Then came the big
question: what about the behavioral response to AVP—would the mice form pair
bonds? When the transgenic and control mice were injected with AVP and put
into a two-chambered testing arena with a female on one side, the transgenic
mice responded to the AVP with a rise in the levels of olfactory investigation
and grooming directed towards the females, while the wild-type mice did
not (Young et al., 1999). A full-fledged pair bond was not described, but the
behavior showed signs of moving in that direction. By 2004, transgenic voles
had been created (Aragona & Wang, 2004). Now it was actually possible to
ask whether the promoter region of the V1a gene of a prairie vole could cause
a male meadow vole to develop partner-preference patterns like those seen in
the prairie voles. Miranda Lim and her colleagues used a viral vector to transfer
this segment of the prairie vole DNA into the ventral forebrain of a meadow
vole. They were targeting the ventral pallidum. She and her coauthors wrote
that they were “in essence recreating a singular evolutionary event in the labo-
ratory” (Lim et al., 2004, p.754). The procedure successfully changed the dis-
tribution of the V1aR so that the ventral forebrain of the meadow vole now
looked like that of a prairie vole. The behavior was assessed by putting the ani-
mals through the standard partner-preference protocol after 24 hours of
cohabitation and sex with a partner. In the three-chambered choice apparatus
the control males spent as much time in contact with the unfamiliar female as
198 ANIMAL BEHAVIOR

the ones with which they had been paired, whereas transgenic animals spent far
more time huddling with their partners than with the strangers! I still feel as if this
study is one of the most extraordinary that I have learned about in the many
years that I have been in the business of behavioral neuroscience!
As occurs in most scientific advancements, as this experiment answered old
questions it raised new ones. Some were about the proximate mechanisms.
Perhaps the central question was, how did that change in V1aR distribution lead
to a change the behavior? Lim and colleagues (2004) suggested that the answer
lies in integration between sensory pathways and reward circuits involving
dopamine and the projection from the VTA to the ventral forebrain, where
the NAc lies. In their words, “We propose that, during pair bond formation,
concurrent activation of individual recognition and reward pathways results
in convergent V1aR and D2 receptor activation in the ventral forebrain, lead-
ing to an association between the rewarding nature of sex and the olfactory sig-
nature of the partner and thus the development of a conditioned partner
preference” (Lim et al., 2004, p. 756). A great deal of work in this area has
focused on this hypothesis, and it has received considerable support. However,
in other ways the story of the mechanisms behind pair-bond formation has not
been so simple (e.g., Phelps et al., 2010).
For example, while pair bonding is the norm among prairie voles, not all of
these animals are bound so tightly to their partners that they will not take
opportunities to copulate with others; in fact, approximately 25 percent of
these animals are conceived through extra-pair copulations (i.e., copulations
with individuals other than the socially bonded partner). Given the story of
the V1aR gene and reward circuits, one might have expected to see a relation-
ship between fidelity and expression of the gene in the ventral forebrain. Stud-
ies of these animals in large, seminatural outdoor enclosures revealed that
differences in the degree of sexual fidelity were completely unrelated to
V1aR distribution in the reward circuits of the ventral forebrain. However,
they were related to V1aR densities in parts of the neocortex and the hippo-
campus, a very unexpected finding (Ophir et al., 2008)! Variation in the
degree of monogamy among prairie voles does not seem to be as straightfor-
ward as variation between prairie voles and meadow voles. And then there is
the question of what happens when we move beyond the microtine rodents? Did
independent evolutionary transitions to a pair-bonding lifestyle all occur via
changes in genes that regulate the distribution of V1aRs? This is an interesting
and complex issue that investigators are still grappling with. But I will leave
the pair-bonding story here and turn to another area of behavioral neurosci-
ence where questions about evolutionary processes and about proximate
mechanisms have merged. We will be stepping back from a focus on the
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 199

interactions between two individuals within a species to consider interactions

between an individual and the movements of the planet that it lives on.

THE PUZZLE OF A DIURNAL MAMMAL

As the Earth spins on its axis the world in which animals find themselves
changes in truly striking ways. Temperatures from day to night may vary by
well over 40 degrees Fahrenheit, and luminance may dip to 1 or 2 lux at night
and climb to blinding levels of 100,000 lux in the midday sun. The changes in
an animal’s world that occur over time, on a daily basis, strike me as being far
more dramatic than many of the differences that exist across space or, more
generally, than other features of an animals’ niche. Yet animals can adapt,
and they have a remarkable ability to anticipate and prepare for the coming
events. The reason is that they have clocks in their brains, circadian clocks that
produce circadian rhythms with a period of approximately 24 hours. The
internal clock that drives these rhythms is ordinarily shifted on a daily basis
such that it becomes synchronized with external rhythms of exactly 24 hours;
the process is referred to as entrainment. One of the great success stories of
behavioral neuroscience in recent years has been the elucidation of the mech-
anisms responsible for the internally driven circadian regulation of behavior
and for the entrainment of those rhythms by photic (light) stimuli. Chapter
11 beautifully lays out this story. In addition to this circadian system, animals
have mechanisms that enable them to respond directly to the daily changes in
their environments, such as the levels of illumination they experience. The
acute effects of the photic environment reflect a process called masking. These
direct effects of light on behavior can sometimes override the influence of the
internal clock and can at other times reinforce the drive produced by that
clock. Our understanding of the neural basis of masking is still in its infancy.
One fascinating aspect of the coordination of activity patterns by these two
processes is the way in which they operate in day-active (diurnal) compared to
night-active (nocturnal) animals. Before elaborating on this point, I want to take
a moment to consider some of the issues more generally. When mammals first
evolved from their reptilian ancestors, many novel adaptations arose. One that
appears to have been key—that made the full radiation of the mammals pos-
sible—was the transition from a diurnal to a nocturnal pattern of adaptation to
the day-night cycle (Crompton et al., 1978). The night provided a huge, wide-
open, unoccupied, niche for them to walk right into! Most mammals remain
nocturnal today, but diurnality has resurfaced in a variety of independent line-
ages. At each of these transitions, changes occurred in the coordination of a far-
ranging suite of functions. The mechanisms regulating general activity produced
200 ANIMAL BEHAVIOR

new patterns, but so did those that coordinate such things as body temperature,
hormone secretion, liver function, kidney function, as well as the time of day at
which animals mate, ovulate, eat, drink, sing, and sleep. Somehow at each evolu-
tionary transition from one temporal niche to another the brain mechanisms
coordinating the timing of this multitude of functions were transformed.
A gradual, step-by-step, incremental, evolutionary transition from nocturnal-
ity to diurnality, or vice versa, would seem to have required marching into a very
maladaptive valley between two optimal “adaptive peaks.” Wouldn’t many gener-
ations of animals have had to live and reproduce in an intermediate state of temporal
confusion with disorganized daily patterns of physiology, brain function, and behavior
to make it from the nocturnal to the diurnal state? Consider sleep and wakefullness,
for example. In a nocturnal rodent the daily patterning of these states is regulated
by the coordination between neural pathways that produce arousal at night, on
the one hand, and systems that push for sleep during the day. So, how would such
animals survive an evolutionary transition to diurnality if first the wake-promoting
mechanisms changed their activity pattern to keep the animals up during the day,
but arousal-promoting pathways continued to keep them up at night? The poor
things would suffer from serious insomnia! What if the mechanisms that influence
the time of day at which mate seeking occurs were now out of phase with the rhythms
in neuroendocrine function that make it possible for the females to ovulate and get
pregnant? No matter how much mating was going on, there might be very few
babies! Most brain functions are rhythmic, and though their patterns are not
identical, they are beautifully coordinated to produce an adaptive pattern opti-
mized for either a diurnal or a nocturnal existence, but not both. This general
issue represents one of the puzzles of any major evolutionary transition from
one niche to another. How were the maladaptive valleys between the adaptive peaks
bridged? There are no fossils of a biological rhythm or of the brain regions respon-
sible for their coordination, and even if there were they would be hard to study.
So the only way to begin to answer the ultimate question of how diurnality
evolved is to learn about the proximate mechanisms that produce these rhythms
and how they may differ in modern-day diurnal and nocturnal species. Our
understanding of that is far from complete, but it is getting better, and that is
what we will be considering here. But we need to start with the question of where
the rhythms are generated.

In the Beginning . . .
The study of brain mechanisms generating and coordinating circadian
rhythms in behavior really took off after a 1972 paper was published by Fred
Stephan and Irving Zucker. It was clear at that time that light could influence
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 201

these endogenously driven rhythms in a host of very predictable ways such

that under normal circumstances they would be entrained to a 24-hour cycle
of light and darkness. Stephan and Zucker set out to find the internal clock
by following the neural pathways through which information about the light
outside could reach it. A new technique had just been developed for more
detailed mapping of the inputs from the eyes to the brain and had revealed
one pathway that extended from the retina to the hypothalamus; its discover-
ers, Robert Moore and Nicholss Lenn (1972), gave it the name retinohypotha-
lamic tract (RHT). The most prominent projection of the RHT is to two
small nuclei sitting on top of the optic chiasm (OX), where the nerves bring-
ing signals from the eyes into the brain cross at the base of the hypothalamus.
These were called the SCN. When they learned about the RHT, Stephan and
Zucker (1972) set out to determine if it, and the SCN, might be responsible
for entrainment, and they produced lesions to test their hypothesis in rats.
What they found was not what they had predicted—it was much more stun-
ning. Their original question about entrainment was moot because there were
no longer any rhythms to entrain! The nice, predictable daily rhythms in
drinking behavior that they recorded before the lesions had been produced
were replaced with randomly distributed bursts of drinking across the day.
That seminal discovery opened the floodgates, and today there may be no
other area of behavioral neuroscience where our understanding of the mecha-
nisms, from the level of the behavior to the dynamics of the molecular interac-
tions themselves, is more sophisticated. This story is addressed in Chapter 11,
which describes such things as transplant studies that established that the SCN
is not simply necessary for behavioral rhythms but actually produces them.
A very simple and fundamental model of the role that the SCN plays in
regulation of rhythms has provided a framework that has been used for several
decades when thinking, writing, and speaking about circadian mechanisms.
It presents the system as one that contains three fundamental components:
(1) the SCN at the center, with its oscillators that produce the rhythm; (2) a
pathway from the eye to the SCN that ordinarily entrains the oscillators to a
24-hour light/dark cycle in the environment; and (3) a pathway extending
outwards from the SCN to the systems that become rhythmic in response to
the signals that it carries. The basic elements of the model and the connections
between them are depicted here: http://upload.wikimedia.org/wikipedia/
commons/thumb/5/55/Circadian_rhythm_labeled.jpg/350px-Circadian_rhythm
_labeled.jpg.
To understand where in the system the key differences between diurnal and
nocturnal species reside, it is useful to start with this framework. There are sev-
eral theoretical possibilities. The simplest are (1) the oscillator within the SCN
202 ANIMAL BEHAVIOR

could respond differently to the retinal input such that its pattern of entrain-
ment is 180 degrees out of phase in diurnal and nocturnal animals; (2) the
oscillator in the SCN could be linked to the light/dark cycle in the same
way, but it could be coupled differently to processes that determine the pat-
tern of signals that the SCN sends out; or (3) the SCN could emit the same
signals, and the differences could lie downstream of it. For example, cells that
the SCN projects to could respond differently to its signals, or SCN outputs
could extend to different populations of target cells, or differences between
nocturnal and diurnal animals could emerge from processes that lie down-
stream of SCN targets. These possibilities are not mutually exclusive, as there
are many distinctly different subpopulations of cells within the SCN, as well as
in regions that receive input from it. Next, I will consider each of these basic
possibilities in turn.

From the Light into the Oscillator

There are two very general kinds of evidence that argue against the first pos-
sibility noted above (reviewed in Smale et al., 2008). One comes from behav-
ioral studies and the other from direct examination of the SCN. Both involve
administering light at different times of day and observing the effects it has on
the phase of a circadian rhythm. This procedure reveals what we call a phase
response curve (PRC), a term invented by Patricia DeCoursey, who produced
one for her PhD thesis in 1960. Her fundamental question was whether there
are rhythms in how the circadian oscillator itself responds to light. To answer
this, she kept her animals, flying squirrels (Glaucomys volans), in a cage with a
running wheel and turned out the lights (DeCoursey, 1986). As she expected,
the endogenous activity rhythms “ran free,” unconstrained by a light/dark
cycle, and adopted periods close to, but not exactly, 24 hours. She then
exposed her squirrels to10-minute pulses of light at different times of day.
What she saw was that the light could shift the animals’ circadian rhythms
in activity and, more importantly, that there was a rhythm in how much and
in what direction that internally driven rhythm shifted! The PRC is a repre-
sentation of that rhythm. So, what does this have to do with diurnality? If the
circadian oscillator is coupled to a light/dark cycle in opposite ways in noctur-
nal and diurnal species, then their PRCs should be 180 degrees out of phase.
That is not what DeCoursey saw in her squirrels or what others have seen over
the years since 1960 in a wide range of both diurnal and nocturnal species.
The PRCs are all essentially the same.
The second, more recent, line of evidence against the hypothesis that diur-
nality could be explained by a difference in the coupling between rhythms in
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 203

the primary circadian oscillator and in ambient light comes from studies of
molecular processes within the SCN. The details of exactly how a sequence
of changes in gene expression can repeat itself according to a rhythm with a
period of approximately 24 hours are described in Chapter 11. The point for
us here is simply that the pattern of those rhythms, and their phase relative
to a light/dark cycle, are the same in virtually all nocturnal and diurnal animals
that have been looked at. This includes nocturnal mice, rats, and hamsters as
well as diurnal ground squirrels, grass rats, degus (Octodon degus), and sheep
(Ovis aries) (reviewed in Smale et al., 2008).

Outwards from the Oscillator

So the phase relationship between the circadian oscillator in the SCN and
the light/dark cycle is the same in nocturnal and diurnal species. Let us move
on to the next question, could that molecular oscillator be coupled differently to
downstream processes that determine the pattern of signals that the SCN sends
out to other brain regions that are under its influence? Indirect lines of evidence
bearing on this issue started coming in long before the molecular biologists
had even begun to attend to the clock. One of the first of these involved a sim-
ple study of energy metabolism. Action potentials represent a key step in the
transmission of a signal from one neuron to another, and they take energy. If
the SCN sends out signals at different times of day in nocturnal and diurnal
species, then one would expect that its rhythms in energy utilization should
be different. William Schwartz (1991) examined this prediction by measuring
uptake of glucose by the SCN in diurnal and nocturnal animals sacrificed at
different times of day. In these classic studies, a radioactive label was first
attached to glucose to make C-14-labeled 2-deoxy-D-glucose (2DG), and this
was then injected into animals at different times of day. When they were sac-
rificed a short time later, the 2DG had accumulated in the brain regions that
had been most metabolically active. The distribution and density of the radio-
active label was imprinted on film placed over the sections of brain; when the
film was processed the resulting “autoradiograph” showed which brain regions
had been especially active and whether this depended on the time of day at
which the animals were sampled. This procedure revealed a very striking
rhythm in the SCN, with maximal uptake of the 2DG during the day, whether
the animal was diurnal or nocturnal! The same temporal patterns were seen in
the SCN of nocturnal and diurnal species from widely varying taxa that
included, for example, laboratory rats, Turkish hamsters (Mesocricetus
brandti), golden hamsters (M. auratus), mice, opossums, squirrel monkeys,
house sparrows (Passer domesticus), and cats (Felis catus) (Schwartz, 1991).
204 ANIMAL BEHAVIOR

This was taken by many as proof that the SCN is the same across species
regardless of their activity patterns and that differences between them must
emerge downstream of that central “clock.”
Another early line of evidence consistent with this hypothesis came from
direct recordings of the rate at which SCN neurons fire off action potentials
and how that rate changes over a 24-hour period (e.g., Inouye & Kawamura,
1979; Green & Gillette, 1982). These studies revealed rhythms with their
peaks during the day and troughs at night in rats, chipmunks, and guinea pigs
(Cavia porcellus). Some of these signals were certainly involved in local circuits
processing information within the SCN, but others must have been playing a
role in getting the message about time of day outwards to regions beyond the
SCN. Interestingly, when the electrodes were misplaced and picked up on
neuronal activity outside of the SCN, rhythms were also apparent, but here
the peak was during the day in chipmunks (which are diurnal) and at night
in the rats (which are nocturnal)! The case that the SCN clock is the same,
from its inputs through its outputs, was getting to be pretty convincing, and
it was looking like cells beyond the SCN must be responding to SCN signals
differently.
Still, for the holdouts, there was the possibility that those action potentials,
though they reach the terminals at the same time of day in nocturnal and diur-
nal animals, might be triggering the release of different neurotransmitters or
neuromodulators. Screens of neuropeptides in the SCN, and of rhythms
within the SCN cells that produced those molecules, revealed some differences
among species but nothing that distinguished a diurnal from a nocturnal SCN
(Smale et al., 2003). One of these is actually AVP, which is depicted in the
SCN of a Nile grass rat (Arvicanthiss niloticus): http://rhythms.psy.msu.edu/
uploads/images/avp_scn.jpg.
Most importantly, rhythms in synthesis of the mRNA for each of three out-
put molecules were found to be identical in the nocturnal and diurnal species
in which they were examined (Smale et al., 2008). Specifically, this was true
for comparisons between one diurnal species, the Nile grass rat, and two noc-
turnal ones, lab rats and mice.
This brings us to an important caveat and an interesting question: Will the
conclusions drawn concerning a transition to diurnality along one phylogenetic line
inform us about how it happened along others? Did all evolutionary pathways to
diurnality involve the same changes in the same neural mechanisms? Did they all
leave the SCN alone and only alter downstream, direct or indirect, responses to
its signals? There are some suggestions that this might not be the case. Let us
consider the degu. These caviomorph rodents from South America are active
primarily during the day in nature, though they may become crepuscular at
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 205

some times of year (reviewed in Hagenauer & Lee, 2008). The possibility that
the SCN of degus might function differently from that of nocturnal rodents
was examined in several studies of these animals. One that was particularly
interesting involved directly comparing neuronal activity recorded from elec-
trodes placed in the SCN of degus and lab rats (Vosko et al., 2009). Two
differences between these species were identified. The first was the likelihood
that a light pulse would stimulate an increase or a decrease in the rate at which
action potentials were fired by individual SCN neurons. In rats, approximately
six times more cells were activated than suppressed by the light, whereas in
degus approximately three times more were suppressed than were activated.
This is quite a striking difference! Second, some neurons did not respond to
light at all in either species, and among rats, these cells fired at higher rates
during the day than night, whereas in degus these rates were absolutely identi-
cal. These fascinating results suggest that signaling between neurons within
the SCN, and/or signals sent from the SCN to its targets, are different in degus
and rats. Could these differences reflect changes that supported a transition to diur-
nality along the evolutionary pathway to modern-day degus? We do not know
yet, but the data are certainly consistent with that hypothesis. But when we
turn to a representative of another group of diurnal rodents, the sciurids, this
does not appear to be the case. In their study of chipmunks (Eutameous siberi-
cus), Tetsu Sato and Hiroshi Kawamura (1984) found that SCN cells fired at
higher rates during the day than night, just as they do in nocturnal rats.
So, what message can we take home from the few studies that provide informa-
tion bearing on the question of whether the oscillator in the SCN is coupled differ-
ently to SCN output signals in day- and night-active animals? There are no
definitive answers. However, the data do suggest the possibility that this may
have contributed to diurnality in some species (e.g., degus), but that it is very
unlikely to have played a role in others (e.g., grass rats and sciurids).

Beyond the Clock

The third general hypothesis, which I would like to turn to now, is, I
believe, the one likely to go furthest in accounting for how the brain can pro-
duce the suite of rhythm patterns that distinguish a diurnal species from a noc-
turnal one. That is, the evolution of diurnality may have involved changes in
networks that mediate patterns of responsiveness to signals sent outward from
the SCN (Smale et al., 2003). Such differences could be in cells that receive
signals directly or indirectly from the SCN, or the “wiring diagram” could
be different such that cell populations receiving direct input from the SCN
are not all the same in day- and night-active species.
206 ANIMAL BEHAVIOR

The diurnal animals in which circadian mechanisms beyond the SCN have
been examined most fully are grass rats from Africa. These members of the genus
Arvicanthis include the unstriped Nile grass rat and the Sudanese grass rat
(A. ansorgei). Each of these species has made its way into the labs of circadian
biologists, though on opposite sides of the Atlantic Ocean. Although studies of
the former species, in the United States, have been more extensive, the only
direct examination of responsiveness to an SCN signal was done with the Suda-
nese grass rat, in Holland (Kalsbeek et al., 2008). That experiment looked at
secretion of the hormone corticosterone from the adrenal gland, which rises in
anticipation of the beginning of the most active period of an animal’s day. One
SCN signal that plays a role in this is a peptide that we encountered in our dis-
cussion of pair bonds, AVP. It had been known for some years that when AVP
is infused into a region of the hypothalamus that receives input from the SCN,
it stimulates a rise in corticosterone in the blood of nocturnal lab rats. The ques-
tion now was whether such infusions might have the opposite effect in the Suda-
nese grass rat, and the answer was “yes”! This was a truly elegant and pioneering
study, the first and only direct evidence that responses to SCN signals can differ
in nocturnal and diurnal species. Mechanisms mediating behavioral responses to
SCN signals, however, are not as well understood as those producing rhythms in
hormone secretion. In fact, no direct studies on them have been conducted in
diurnal species. However, in the case of the Nile grass rat, we do know something
about rhythms in a suite of brain areas that mediate a variety of behaviors and
that receive direct or indirect input from the SCN.
Before going any further, however, I need to fill you in a bit about these Nile
grass rats, which my students and I have been studying for some years now,
because much of the research described below comes from this work. These are
small (60–80 g) murid rodents that look very much like a mouse or a rat (though
I like to think they are a bit more handsome, and they smell better). They are also
quite beautiful, for those who love animals, as can be seen in the photograph
here: http://www.cons-dev.org/GVS/images/hab/252.jpg.
They come from East Africa, where I was living from 1988 to 1992.
During that time I often saw these animals out and about during the brightest,
hottest periods of the day, something that is relatively rare for a murid. My
interest in behavior, its neural substrates, and the evolutionary processes that
shaped them came together as I thought, what was it that drove them out of
their burrows during the day and into them at night? So I brought them home,
into my lab at Michigan State University, where we have been working with
them to try to figure out the key(s) to diurnality—which brings us back to
the idea that it seems to reside downstream of the SCN.
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 207

So where does the SCN send its signals? Tract-tracing studies have revealed
that axons extending outward from the SCN project to the same basic target
regions in grass rats as in nocturnal rodents (Novak et al., 2000; Schwartz
et al., 2010). So, at a regional level, the wiring-diagram hypothesis does not
hold up. What about rhythms within those regions—are they different in noctur-
nal and diurnal species? To address these questions, a series of studies has been
conducted to assess temporal patterns of change in numbers of cells that con-
tain two kinds of proteins. One, cFos, is a product of an immediate early gene
whose expression is linked to the activity of cells (Hoffman et al., 1993). The
others, Period 1 and Period 2 (Per1/2), are elements of the molecular oscillator
in the SCN, which is necessary for, and drives, circadian rhythms in behavior
(Ko & Takahashi, 2006). However, as described in Chapter 11, molecular
oscillators are present in other brain regions as well. These “extra-SCN oscilla-
tors” almost all depend on the SCN, and their rhythms damp out if it is
lesioned. Therefore, a series of studies was launched to investigate the tempo-
ral patterns of cFos and Per1/2 in order to determine if a simple switching
mechanism could have shifted rhythms beyond the SCN in a uniform man-
ner, thereby transforming a diurnal brain into a nocturnal one. The answer
seems to be that it is considerably more complicated.
The temporal patterns of change in cFos in the brains of Nile grass rats
reveal that some areas that receive input directly from the SCN exhibit
rhythms that are 12 hours out of phase with those seen in nocturnal lab rats,
but that others are shifted by just 9 hours or 4 hours (Smale et al., 2008).
Studies of Per1/2 beyond the SCN also reveal rhythms with varying phase
relationships in grass rats relative to lab rats, but most are shifted by a full
12 hours (Ramanathan et al., 2008, 2010). These areas include, for example,
the VTA, the NAc, and the hippocampus. One cannot help but wonder if pat-
terns in the first two regions could lead to different rhythms in reward-related
processes (perhaps even those associated with sex behavior and social bonds),
or if the patterns in the third area could influence daily fluctuations in abilities
of animals to learn to find their way around.
Overall, the data on cFos and Per1/2 suggest that the evolution of diurnal-
ity in grass rats did not involve changes in a single structure or pathway. The
system through which the SCN produces a range of rhythms in behavior
and physiology includes multiple intersecting output pathways, and the pic-
ture emerging for the grass rat is one in which direct and indirect components
of this output system differ in a multitude of ways from those seen in noctur-
nal rats. Taken as whole, the product is a coordinated suite of rhythms in
behavior and physiology exquisitely adapted to a diurnal niche.
208 ANIMAL BEHAVIOR

Bridging the Maladaptive Valley?

This brings us back to the question of how such animals could have ever
made the evolutionary journey from a nocturnal niche to a diurnal one with-
out being caught in the middle. Here, we dive even deeper into a world of
speculation—could part of the answer be related to plasticity in this expanded cir-
cadian network? The general idea that a maladaptive valley can be bridged with
help from flexibility in a biological system has been around for some time.
Here, the more specific question is how the interconnected oscillators, whose
phase relationships have changed over evolutionary time, can respond to
changing circumstances within an animal’s lifespan. The question is not
whether an animal can be perfectly adapted to both a diurnal and a nocturnal
niche—that would not be possible—but how, and to what extent, can its
temporal organization adjust to changing circumstances?
Plasticity in circadian systems is evident in a variety of contexts (Ronnenberg
& Merrow, 2003). Rhythms can change on a seasonal basis, as in the case of
reindeer (Rangifer tarandus), which become arrhythmic in the summer, or voles,
which become more diurnal in the winter. Developmental changes are evident in
many species, and female mammals undergo changes in the pattern of myriad
rhythms as they transition from one reproductive state to another. An animal’s
activity pattern may also change in response to the presence of competitors, pred-
ators, and the availability of food (Kronfeld-Schor & Dayan, 2008). Perhaps the
most striking, and strange, is that extreme forms of plasticity along the nocturnal-
diurnal dimension have been seen in the laboratory. Activity patterns of golden
spiny mice (Acomys russatus), for example, can switch abruptly from being noc-
turnal in the field to becoming day-active in the laboratory (Cohen &
Kronfeld-Schor, 2006). In several diurnal rodents, including gerbils, degus, and
grass rats, the introduction of a wheel into a cage can trigger a strikingly noctur-
nal pattern of activity. The animal is day-active, the wheel is introduced, and
there is a flip to a nocturnal pattern—and when the wheel is removed the diurnal
pattern returns (Blanchong et al., 1999). Wheels do not elicit this nocturnal pat-
tern in all grass rats, and this form of plasticity is heritable in these animals. The
night-active pattern persists when they are held in constant darkness, indicating
that the endogenous circadian system is driving it. The fundamental difference
in these patterns is depicted in a figure that can be found here: http://arvicanthis
.psy.msu.edu/pics/nixon_activity_graph.png.
So, can this strange phenomenon, this laboratory artifact, tell us anything
about natural processes? Animals did not evolve in a world strewn with running
wheels, and we have no idea why rodents seem to be so addicted to running in
them, but they are, at certain times of day. Even if it is a laboratory phenome-
non, the wheel-induced switch from a diurnal pattern of behavioral
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 209

organization to a nocturnal one can reveal something about the nature of plas-
ticity in the mechanisms that organize the temporal program of an animal.
The question is, what other aspects of temporal organization in these animals
are shifted in the nocturnal direction by the presence of the wheel? Patterns of
feeding and drinking behavior, and the day/night ratio of behavioral sleep,
all shift with the wheel running (Schwartz & Smale, 2005). In fact, unpub-
lished data from our laboratory show that masking responses are altered such
that light now triggers a decrease in activity rather than an increase, as it does
ordinarily. However, other features of the animals’ temporal program are not
so easily bounced around. For example, although the individuals that run
most at night sleep more during the day, that sleep is highly fragmented,
whereas their nighttime sleep is consolidated into longer bouts (Schwartz &
Smale, 2005). This day-night difference is indistinguishable from that seen
in the grass rats that maintain a diurnal pattern of behavior.
When looking into the brains of these animals, one also sees plasticity in
some aspects of the temporal program but not in others. A suite of rhythms
in both Per1 and Per2 in regions beyond the SCN undergoes a full 12-hour
reversal in phase when running occurs at night (Ramanathan et al., 2010).
These extra-SCN oscillators become indistinguishable from those seen in noc-
turnal laboratory rats! This is the case, for example, in the hippocampus. Yet
some of these rhythms are more stubbornly stuck in the diurnal program. This
is the case, for example, for Per2 in the NAc. These patterns suggest that
rhythms in behaviors regulated by these structures, such as memory formation
and sex behavior, might be able to adapt relatively easily as animals move
towards a nocturnal niche. A similar story seems to come through when one
examines rhythms in Fos (Smale et al., 2008). Rhythms in one arousal-
related population of cells in the hypothalamus (those containing a neuropep-
tide called orexin) are perfectly reversed in the day- and night-active grass rats,
which is not the case in one region of the hypothalamus that promotes sleep
(the ventrolateral preoptic area, or VLPO; Schwartz & Smale, 2005). Here,
one can easily imagine that the former could help during adjustment of a diur-
nal animal to a new temporal niche by maintaining arousal at night, while the
latter would lag behind and interfere with a simple transition by preventing
sleep consolidation during the day.
So, let us get back to evolution. Cyclical behavior patterns have been
shaped by natural selection, which acts on heritable variation. When the
grass rats were brought to the United States from Kenya, only one in 15
showed the night-active pattern of wheel running (Katona & Smale, 1997).
However, 13 of her 25 offspring did as well. A subsequent study confirmed
that the pattern is very highly heritable (Blanchong et al., 1999). Using the
210 ANIMAL BEHAVIOR

pattern of wheel running as our criterion for selection had thus revealed herit-
ability of a surprising degree of plasticity in a linked suite of features of the
temporal program. Other aspects of timekeeping mechanisms in the brain
may be more resistant to change within a lifespan and less variable or less her-
itable. This may also be the case for other traits associated with adaptation to a
diurnal niche, such as those associated with sensory processing. The cone/rod
ratio in the retina of the eye is 10 times higher in grass rats than lab rats, and
the size of the optic tectum is the same, though the lab rats are four times
the size of the grass rats. It is hard to know for sure, but structures that are this
different might be less plastic and require more time for natural selection to
chip away at.
What can the mosaic of patterns of plasticity and resistance tell us about the evo-
lution of adaptive change from one temporal niche to another? What it offers is
simply a suggestion of the kinds of plasticity that might have helped bridge
that valley between a diurnal and a nocturnal adaptive peak. The plasticity of
some features of the circadian system could have provided time for natural
selection to act on the traits that were more resistant to environmental influen-
ces during an individual’s lifetime.

CONCLUSIONS
The central focus of behavioral neuroscience has always been on efforts to
understand the relationships between the brain and behavior. The field has
traditionally asked such questions as, how does the brain enable animals to avoid
predators and find food, to have sex, and to take care of babies? It represents the
analysis of behavior at Tinbergen’s proximate level. However, behind all of
these behaviors is an evolutionary history from which these mechanisms have
emerged. Behavioral neuroscientists appreciate this more than ever and share
the excitement of a behavioral ecologist focused on the evolutionary forces,
such as natural selection, that may have shaped the evolution of new patterns
of behavior over many generations. At the same time, scientists focused at the
ultimate level are appreciating more and more how our understanding of
behavioral evolution can be so much deeper if we know something about the
mechanisms that produce it. The merging of Tinbergen’s levels has been
occurring at a rapidly accelerating rate. I could have written about this integra-
tion by telling you about bird song, learning and memory, social cognition, or
a host of other topics in behavioral biology, but I did not have the space! The
stories that I chose to tell, of pair bonds and of diurnality, may seem quite
unrelated, but what they have in common is that efforts to understand them
are beginning to bring together an appreciation of the fascinating bidirectional
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 211

relationships between the evolution of behavior and the proximate neural

mechanisms underlying its expression.

ACKNOWLEDGMENTS
I would like to thank Zuleyma Tang Martínez and Ken Yasukawa for their
enormous help in the preparation of this manuscript, and also Antonio A.
Nunez and Lily Yan for their helpful discussions on many of the issues in it.

REFERENCES AND SUGGESTED READING

Aragona, B. J. & Z. Wang. (2004). The prairie vole (Microtus ochrogaster): An animal
model for behavioral neuroendocrine research on pair bonding. Institute for
Laboratory Animal Research Journal, 45, 35–45.
Bear, M. F., B. W. Connors, & M. A. Paradiso. (2007). Neuroscience: Exploring the
Brain. Philadelphia: Lippincott Williams and Wilkins.
Blanchong, J. A., T. L. McElhinny, M. M. Mahoney, & L. Smale. (1999). Nocturnal
and diurnal rhythms in the unstriped Nile rat, Arvicanthis niloticus. Journal of
Biological Rhythms, 14, 364–377.
Breedlove, S. M., M. R. Rosenzweig, & N. V. Watson. (2007). Biological Psychology:
An Introduction to Behavioral, Cognitive, and Clinical Neuroscience. Fifth
Edition. Sunderland, MA: Sinauer Associates.
Carter, C. S., A. C. Devries, & L. L. Getz. (1995). Physiological substrates of mam-
malian monogamy: The prairie vole model. Neuroscience and Biobehavioral
Reviews, 19, 303–314.
Carter, C. S., L. L. Getz, L. Gavish, J. L. McDermott, & P. Arnold. (1980). Male-
related pheromones and the activation of female reproduction in the prairie
vole (Microtus ochrogaster). Biology of Reproduction, 23, 1038–1045.
Carter, C. S., D. M. Witt, E. G. Thompson, & K. Carlstead. (1988). Effects of hor-
monal, sexual, and social history on mating and pairbonding in prairie voles.
Physiology and Behavior, 44, 691–697.
Cohen, R. & N. Kronfeld-Schor. (2006). Individual variability and photic entrain-
ment of circadian rhythms in golden spiny mice. Physiology and Behavior, 87,
563–574.
Crompton, A. W., C. R. Taylor, & J. A. Jagger. (1978). Evolution of homeothermy
in mammals. Nature, 272, 333–336.
DeCoursey. P. J. (1986). Light-sampling behavior in photoentrainment of a rodent
circadian rhythm. Journal of Comparative Physiology A: Sensory, Neural, and
Behavioral Physiology, 159, 161–169.
Dluzen, D. E., V. D. Ramirez, C. S. Carter, & L. L. Getz. (1981). Male vole urine
changes luteinizing hormone-releasing hormone and norepinephrine in female
olfactory bulb. Science, 212, 573–575.
212 ANIMAL BEHAVIOR

Donaldson, Z. R. & L. J. Young. (2008). Oxytocin, vasopressin, and the neuroge-

netics of sociality. Science, 322, 900–904.
Ferris, C. F. & M. Potegal. (1988). Vasopressin receptor blockade in the anterior
hypothalamus suppresses aggression in hamsters. Physiology and Behavior, 44,
235–239.
Gaulin, S. J. C. & R. W. FitzGerald. (1989). Sexual selection for spatial learning
ability. Animal Behaviour, 37, 322–331.
Gaulin, S. J. C., R. W. FitzGerald, & M. S. Wartell. (1990). Sex differences in spatial
ability and activity in two vole species (Microtus pennsylvanicus and
M. ochrogaster). Journal of Comparative Psychology, 104, 88–93.
Gavish, L., C. S. Carter, & L. L. Getz. (1983). Male-female interactions in prairie
voles. Animal Behaviour, 31, 511–517.
Green, D. J. & R. Gillette. (1982). Circadian rhythm of firing rate recorded from sin-
gle cells in the rat suprachiasmatic brain slice. Brain Research, 245, 198–200.
Gruder-Adams, S. & L. L. Getz. (1985). Comparison of the mating system and pater-
nal behavior in Microtus ochrogaster and M. pennsylvanicus. Journal of Mammal-
ogy, 66, 165–167.
Hagenauer, M. H. & T. M. Lee. (2008). Circadian organization of the diurnal cavio-
morph rodent Octodon degus. Biological Rhythm Research, 39, 269–289.
Hammock, E. A. D. & L. J. Young. (2002). Variation in the vasopressin V1a receptor
promoter and expression: Implications for inter- and intraspecific variation in
social behaviour. European Journal of Neuroscience, 16, 399–402.
Heldt, S. A., L. Stanek, J. P. Chhatwal, & K. J. Ressler. (2007). Hippocampus-
specific deletion of BDNF in adult mice impairs spatial memory and extinc-
tion of aversive memories. Molecular Psychiatry, 12, 656–670.
Hoffman, G. E., M. S. Smith, & J. G. Verbalis. (1993). c-Fos and related immediate
early gene products as markers of activity in neuroendocrine systems. Frontiers
in Neuroendocrinology, 14, 173–213.
Hoffmann, R. S. & J. W. Koeppl, (1985). Zoogeography. In R. H. Tamarin (ed.),
Biology of New World Microtus (pp. 84–115). Stillwater, OK: American
Society of Mammalogists.
Hofmann, J. E., L. L. Getz, & L. Gavish. (1984). Home range overlap and nest
cohabitation of male and female prairie voles. American Midland Naturalist,
112, 314–319.
Inouye, S. T. & H. Kawamura. (1979). Persistence of circadian rhythmicity in a
mammalian hypothalamic “island” containing the suprachiasmatic nucleus.
Proceedings of the National Academy of Sciences, USA, 76, 5962–5966.
Insel, T. R. & L. E. Shapiro. (1992). Oxytocin receptor distribution reflects social
organization in monogamous and polygamous voles. Proceedings of the
National Academy of Sciences, USA, 89, 5981–5985.
Insel, T. R., Z.-X. Wang, & C. F. Ferris. (1994). Patterns of brain vasopressin recep-
tor distribution associated with social organization in microtine rodents. Jour-
nal of Neuroscience, 14, 5381–5392.
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 213

Jacobs, L. F. (1996). Sexual selection and the brain. Trends in Ecology and Evolution,
11, 82–86.
Jacobs, L. F., S. J. C. Gaulin, D. F. Sherry, & G. E. Hoffman. (1990). Evolution of
spatial cognition: Sex-specific patterns of spatial behavior predict hippocampal
size. Proceedings of the National Academy of Sciences, USA, 87, 6349–6352.
Kalsbeek, A., L. A. W. Verhagen, I. Schalij, E. Foppen, M. Saboureau, B. Bothorel,
R. M. Buijs, & P. Peyet. (2008). Opposite actions of hypothalamic vasopressin
on circadian corticosterone rhythm in nocturnal versus diurnal species. Euro-
pean Journal of Neuroscience, 27, 818–827.
Katona, C. & L. Smale. (1997). Wheel-Running Rhythms in Arvicanthis niloticus.
Physiology and Behavior, 61, 365–372.
Kavaliers, M., K.-P. Ossenkoop, L. A. M. Galea, & B. Kolb. (1998). Sex differ-
ences in spatial learning and prefrontal and parietal cortical dendritic mor-
phology in the meadow vole, Microtus pennsylvanicus. Brain Research, 810,
41–47.
Ko, C. H. & J. S. Takahashi. (2006). Molecular components of the mammalian cir-
cadian clock. Human Molecular Genetics, 15, R271–R277.
Kronfeld-Schor, N. & T. Dayan. (2008). Activity patterns of rodents: The
physiological ecology of biological rhythms. Biological Rhythm Research, 39,
193–211.
Lim, M. M., Z. Wang, D. E. Olazábal, X. Ren, E. F. Terwilliger, & L. J. Young.
(2004). Enhanced partner preference in a promiscuous species by manipulat-
ing the expression of a single gene. Nature, 429, 754–757.
Lucas, J. R., A. Brodin, S. R. de Kort, & N. S. Clayton. (2004). Does hippocampal
size correlate with degree of caching specialization?Proceedings of the Royal Soci-
ety of London, B, 271, 2423–2429.
McGuire, B. & M. Novak. (1984). A comparison of maternal behaviour in the
meadow vole (Microtus pennsylvanicus), prairie vole (M. ochrogaster), and pine
vole (M. pinetorum). Animal Behaviour, 32, 1132–1141.
Moore, R. Y. & N. J. Lenn. (1972). A retinohypothalamic projection in the rat.
Journal of Comparative Neurology, 146, 1–14.
Nakazawa, K., T. J. McHugh, M. A. Wilson, & S. Tonegawa. (2004). NMDA recep-
tors, place cells and hippocampal spatial memory. Nature Reviews Neuroscience,
5, 361–372.
Nicholas, A., C. D. Munhoz, D. Ferguson, L. Campbell, & R. Sapolsky. (2006).
Enhancing cognition after stress with gene therapy. Journal of Neuroscience,
26, 11637–11643.
Novak, C. M., J. A. Harris, L. Smale, & A. A. Nunez. (2000). Suprachiasmatic
nucleus projections to the paraventricular thalamic nucleus in nocturnal rats
(Rattus norvegicus) and diurnal Nile grass rats (Arviacanthis niloticus). Brain
Research, 874, 147–157.
O’Keefe, J. & L. Nadel. (1978). The Hippocampus as a Cognitive Map. Oxford, UK:
Oxford University Press.
214 ANIMAL BEHAVIOR

Oliveras, D. & M. Novak. (1986). A comparison of paternal behaviour in the

meadow vole Microtus pennsylvanicus, the pine vole M. pinetorum and the prai-
rie vole M. ochrogaster. Animal Behaviour, 34, 519–526.
Ophir, A. G., P. Campbell, K. Hanna, & S. M. Phelps. (2008). Field tests of cis-
regulatory variation at the prairie vole avpr1a locus: Association with V1aR
abundance but not sexual or social fidelity. Hormones and Behavior, 54,
694–702.
Phelps, S. M., P. Campbell, D.-J. Zheng, & A. G. Ophir. (2010). Beating the
boojum: Comparative approaches to the neurobiology of social behavior.
Neuropharmacology, 58, 17–28.
Ramanathan, C., A. A. Nunez, & L. Smale. (2008). Daily rhythms in PER1 within
and beyond the suprachiasmatic nucleus of female grass rats (Arvicanthis niloti-
cus). Neuroscience, 156, 48–58.
Ramanathan, C., A. Stowie, L. Smale, & A. A. Nunez. (2010). Phase preference for
the display of activity is associated with the phase of extra-suprachiasmatic
nucleus oscillators within and between species. Neuroscience, 170, 758–772.
Ronnenberg, T. & M. Merrow. (2003). The network of time: understanding the
molecular circadian system. Current Biology, 13, R198–R207.
Sato, T. & H. Kawamura. (1984). Circadian rhythms in multiple unit activity inside
and outside the suprachiasmatic nucleus in the diurnal chipmunk (Eutamias
sibiricus). Neuroscience Research, 1, 45–52.
Schwartz, W. J. (1991). SCN metabolic activity in vivo. In Klein, D. C., R. Y. Moore,
& S. M. Reppert (eds.), Suprachiasmatic Nucleus: The Mind’s Clock (pp. 144–
156). Oxford, UK: Oxford University Press.
Schwartz, M. D. & L. Smale. (2005). Individual differences in rhythms of behavioral
sleep and its neural substrates in Nile grass rats. Journal of Biological Rhythms,
20, 526–537.
Schwartz, M. D., H. F. Urbanski, A. A. Nunez, & L. Smale. (2010). Projections of
the suprachiasmatic nucleus and ventral subparaventricular zone in the Nile
grass rat (Arvicanthis niloticus). Brain Research, 1367, 146–161.
Sherry, D. F., M. R. L. Forbes MARK R. L., M. Khurgel, & G. O. Ivy. (1993).
Females have a larger hippocampus than males in the brood-parasitic brown-
headed cowbird. Proceedings of the National Academy of Sciences, USA, 90,
7839–7843.
Smale, L., T. Lee, & A. A. Nunez. (2003). Mammalian diurnality: Some facts and
gaps. Journal of Biological Rhythms, 18, 356–366.
Smale, L., A. A. Nunez, & M. D. Schwartz. (2008). Rhythms in a diurnal brain.
Biological Rhythm Research, 39, 305–318.
Squire, L. R. (2004). Memory systems of the brain: A brief history and current
perspective. Neurobiology of Learning and Memory, 82, 171–177.
Stephan, F. K. & I. Zucker. (1972). Circadian rhythms in drinking behavior and
locomotor activity of rats are eliminated by hypothalamic lesions. Proceedings
of the National Academy of Sciences, USA, 69, 1583–1586.
NEURAL SYSTEMS , BEHAVIOR , AND EVOLUTION 215

Vosko, A. M., M. H. Hagenauer, D. L. Hummer, & T. M. Lee. (2009). Period gene

expression in the diurnal degu (Octodon degus) differs from the nocturnal
laboratory rat (Rattus norvegicus). American Journal of Physiology: Regulatory,
Integrative and Comparative Physiology, 296, R353–R361.
Wang, A. & B. J. Aragona. (2004). Neurochemical regulation of pair bonding in male
prairie voles. Physiology and Behavior, 83, 319–328.
Williams, J. R., K. C. Catania, & C. S. Carter. (1992). Development of partner
preferences in female prairie voles (Microtus ochrogaster): The role of social
and sexual experience. Hormones and Behavior, 26, 339–349.
Williams, J. R., T. R. Insel, C. R. Harbaugh, & C. S. Carter. (1994). Oxytocin
administered centrally facilitates formation of a partner preference in female
prairie voles (Microtus ochrogaster). Journal of Neuroendocrinology, 6, 247–250.
Young, L. J., B. Huot, R. Nilsen, Z. Wang, & T. R. Insel. (1996). Species differences
in central oxytocin receptor gene expression: Comparative analysis of promoter
sequences. Journal of Neuroendocrinology, 8, 777–783.
Young, L. J., R. Nielsen, K. G. Waymire, G. R. MacGregor, & T. R. Insel. (1999).
Increased affiliative response to vasopressin in mice expressing the V1a receptor
from a monogamous vole. Nature, 400, 766–768.
Young, L. J., J. T. Winslow, R. Nilsen, & T. R. Insel. (1997). Species differences in
V₁a receptor gene expression in monogamous and nonmonogamous voles:
Behavioral consequences. Behavioral Neuroscience, 111, 599–605.
This page intentionally left blank
 7

Hormones and Animal Behavior

Matthew J. Fuxjager, Sarah Jane Alger, and Catherine A. Marler

INTRODUCTION
In an ever-changing environment, animals frequently adjust their behavioral
traits or tactics to maximize their survival and reproductive output. This
means that there is plasticity in the physiological and neurological mecha-
nisms that govern how an individual processes and responds to sensory
stimuli, samples its environment, assesses its internal condition, and produces
motor responses. To make matters more complicated, it is sometimes neces-
sary for more than one of these mechanistic channels to change simultaneously
so that animals can modify more complex behavioral repertoires. How do these
changes in internal physiology (and thus behavioral output) occur, and how are
they coordinated with an animal’s external surroundings? The endocrine system
plays an important role in these processes, and we can study the diverse ways
in which components of the endocrine system have evolved to affect behavior
both within individuals and across species.
In this chapter, we explore a diversity of social paradigms that illustrate
interactions between endocrine communication within the body and social
stimuli in the environment. Our primary focus will be on the roles of steroid
hormones and behavior, but we will also examine the effects of a few other
peptide hormones that have proven integral to how behavior is modulated.
With this framework in mind, we will search for evidence of how natural
and sexual selection have interacted with hormonal systems to sculpt adaptive
behavioral diversification.
218 ANIMAL BEHAVIOR

HORMONES
Hormones are generally considered chemical messengers released from cells
into the bloodstream to exert action on a distant target tissue. Usually, target
tissues contain receptors that bind to specific hormones, such that each
hormone-receptor complex triggers its own unique cascade of reactions that
eventually modify cellular activity. In some cases, the effects of hormones are
relatively slow (hours to days), and these effects can persist for a relatively long
time. However, in other instances, the effects of hormones can be incredibly
rapid (seconds to minutes), and the effects can be similarly quick to fade.
Numerous factors, including the chemical properties of a given hormone, its
receptor, and physiological constraints on intracellular reactions, determine
the speed and duration of hormone effects. We will not delve into the nuances
of hormone biochemistry; instead, we will base our discussion on the relatively
simple model of hormone action described above and draw on hormone bio-
chemistry when it is germane to a specific case or example.

Steroid Hormones
Steroids are a category of hormone known to have pronounced effects on
animal behavior. The adrenal glands and gonads (testes and ovaries) are the
two primary tissues that secrete steroid hormones, though there are a few other
tissues, such as the brain, that are sometimes also considered endocrine
glands. The molecular precursor to all vertebrate steroid hormones is choles-
terol, which is obtained from plasma lipoproteins and stored in steroidogenic
(steroid-producing) glands. Cholesterol and its steroid-hormone products
share a fundamental structure of three carbon hexagons and one carbon penta-
gon. However, steroids all differ with respect to the carbon and oxygen atoms
attached to these rings and whether the carbons in each ring contain single or
double bonds. Within all endocrine organs (as well as other nonendocrine tis-
sues), a vast network of enzymes is expressed to synthesize the various types of
steroidal precursors and steroidal hormones (Figure 7.1). Also important is
that steroid hormones are lipophilic, which means that steroids can pass
through cellular membranes with relative ease.
In the first major section of this chapter, we will mainly discuss the effects
of three classes of steroid hormones: androgens, estrogens, and glucocorti-
coids (Figure 7.1). Of these, androgens and estrogens are considered sex ste-
roids, given their “masculinizing” and “feminizing” effects, respectively (their
overlap in function is discussed later in the chapter). Androgens predomi-
nantly include the steroids testosterone (T), 5α-dihydrotestosterone (DHT),
androstenedione (ANDRO), and dehydroepiandrosterone (DHEA), though
HORMONES AND ANIMAL BEHAVIOR 219

Figure 7.1. Steroid hormone biosynthesis pathway. The name of each steroidal com-
pound is given below the corresponding molecular structure, with abbreviations
used in text indicated as well. Names above arrows denote the enzymes that catalyze
the conversion of precursor to product.

there are other types of androgens that we will not discuss here. Estrogens
include estradiol (E), estriol, and estrone. Glucocorticoids are loosely consid-
ered “stress” hormones because their function is frequently associated with
acute and chronic stress stimuli. In most vertebrates, the main glucocorticoid
is corticosterone (CORT), although in humans and other primates the main
glucocorticoid is cortisol.
Control of androgen and estrogen secretion occurs mostly through the
hypothalamic-pituitary-gonadal (HPG) axis. Top-down regulation of this
axis begins with a part of the brain called the hypothalamus, which emits
gonadotropin-releasing hormone (GnRH). GnRH then travels to the
220 ANIMAL BEHAVIOR

pituitary gland, where it binds and induces the release of luteinizing hor-
mone (LH) and follicle-stimulating hormone (FSH) into circulation. These
two hormones eventually reach the gonads and regulate the production of
sex steroids. The HPG axis is under negative feedback control, meaning that
sex steroids typically act on the hypothalamus and pituitary to decrease
GnRH, LH, and FSH production, respectively. Thus, high levels of sex ste-
roids are capable of suppressing further production via the HPG system.
Control of glucocorticoids occurs mostly through the hypothalamic-
pituitary-adrenal (HPA) axis. Top-down regulation of this axis begins with
hypothalamic secretion of corticotropin-releasing hormone (CRH), which
travels and binds to the pituitary gland. Next, the pituitary gland releases
adrenocorticotropic hormone (ACTH) into the bloodstream. ACTH acts
on the adrenal glands and causes them to release glucocorticoids.

Nonapeptide Hormones
In the second major section of this chapter we focus on neuropeptide hor-
mones that also have a strong influence on behavior. These hormones are just a
few amino acids in length. The prehormones are processed in the endoplasmic
reticulum, packaged into secretory vesicles that are released from the cell when
the membrane-bound packages fuse with the cell membrane, and are secreted
from cells via exocytosis. We focus here only on the nonapeptides, so named
because they have nine amino acids: oxytocin (OT) and vasopressin (AVP).
The nonmammalian homologues of interest are mesotocin (MT) and vasotocin
(AVT). These peptide hormones function as both hormones and neuromodula-
tors (substances that alter nerve-impulse transmission). For example, oxytocin
functions in milk letdown in response to nursing or stimulation of the nipple.
Oxytocin is released from the posterior pituitary into the blood, where it can in-
fluence distant targets, thus acting as a hormone. AVP is released into the blood
as well, and one of its well-known functions is to regulate the water balance of an
organism. In addition to these physiological functions, both peptides are also syn-
thesized and released in other brain areas and can therefore function as neuromo-
dulators by influencing the corresponding receptors in the brain.
In the following paragraphs we link these steroid and peptide hormones to
a diversity of social behaviors.

STEROID HORMONES AND SOCIAL BEHAVIOR

Sex Behavior
Activation of sex behavior—Historically, one of the clearest connections
between steroid hormone action and behavior occurs with respect to sexual
HORMONES AND ANIMAL BEHAVIOR 221

behavior (though we will see later in this chapter that this is not always the case).
In male rodents, for example, removal of the testes (castration) reduces sexual
behavior, and this effect is rescued by treatment with either T or ANDRO. It
is interesting to note, however, that treatment with DHT by itself does not res-
cue male sex behavior entirely (Davidson, 1966a, 1966b; Nelson, 2000). DHT
in general does not have estrogenic effects because it is not converted to E via
the enzyme aromatase, whereas both T and ANDRO can be converted. Overall,
this suggests that some of the effects of androgens on sex behavior occur through
estrogenic channels after T is converted to E. Studies have indeed shown that this
idea has merit, as E treatment in rodents activates sexual behavior (Beyer et al.,
1976; Davidson, 1969). Birds and reptiles show similar pathways through which
androgens (via conversion to E) regulate male sexual behavior (Hutchison, 1970;
Hutchison & Steimer, 1983; Woolley et al., 2004).
The story of how steroids control sexual behavior in females is slightly more
complicated, because sex involves coordination between reproductive cycles and
neuronal pathways that underlie sex behavior. In female mammals, the two ste-
roids E and progesterone (P) usually act in a coordinated manner to regulate
sex behavior. For example, in female rodents, E needs to act roughly one day
before P to induce what is considered normal sexual behavior (McCarthy &
Becker, 2002). If either of these hormones is given by itself or in the reverse
order, then female sexual behavior is abnormal or is not shown. In general, it is
thought that E primes parts of the female brain that facilitate proper behavior
during sexual encounters, though the exact nature of this priming is not yet
completely understood (Erskine, 1989; Nelson, 2000; see Powers, 1970;
Whalen, 1974).
The way E and P interact in a natural context to induce sex behavior is also
quite intriguing, and one of the best examples comes from golden hamsters
(Mesocricetus auratus) (Lisk et al., 1983; Lisk & Reuter, 1980). Males and
females of this species live solitary lives; they inhabit burrows by themselves
and vigorously protect these homes from all intruders. This, of course, poses
a problem when individuals want to reproduce, because they are predisposed
to being solitary. However, when E is released from the ovaries, it causes
females to begin rubbing a vaginal secretion that attracts males around the
entrance of their burrow. Next, P acts to induce sexual receptivity, such that
females allow enticed males into their burrow for copulation. Once ovulation
occurs, E and P levels drop and females again become aggressive and readily
evict any intruder from their burrow (described in Adkins-Regan, 2005).
Suppression of sex behavior—Another steroid that is important for sexual
behavior is CORT, which in many species suppresses the ability or willingness
to reproduce. A good example of this is found in male rough-skinned newts
222 ANIMAL BEHAVIOR

(Taricha granulosa), which typically clasp onto a prospective mate for several
days and fertilize her eggs at some point during this amplexus. While T helps
activate male clasping (and there is little that can disrupt clasping), the percep-
tion of a predator can quickly cause a male to release. This occurs because
acute stressors, such as the presence of a predator, generate an immediate rise
in circulating CORT, which in turn rapidly alters the signaling properties of
neurons in the brain stem and spinal cord that control motoric elements of
clasping (Coddington & Moore, 2003; Deviche & Moore, 1988; Moore &
Miller, 1984; Moore & Orchinik, 1994).
An additional example of CORT-induced reproductive suppression comes
from research in white-crowned sparrows (Zonotrichia leucophrys), particularly
populations that inhabit the northern Rocky Mountains of the United States.
These individuals sometimes experience snowstorms during the early spring
when males are beginning to compete for and defend territories. Such cli-
matic events cause CORT levels to increase, which is associated with an
abrupt change from defending a territory to searching for food. To test the
hypothesis that CORT causes this behavioral switch, researchers first gave
male song sparrows (Melospza melodia) (close relatives to white-crowned spar-
rows) CORT implants and then measured their territorial behavior. CORT-
implanted males showed virtually no aggression toward staged territorial
intruders, despite no change in T levels, compared to males with normal
CORT levels (Wingfield et al., 1998; Wingfield & Sapolsky, 2003). Thus,
this strongly supports the idea that stressful environmental events, such as
the onset of late-spring blizzards, can restrain sexual behavior via the actions
of stress hormones like CORT.
When we consider much of the work investigating the relationship between
sex steroids, stress steroids, and reproduction from a broader standpoint, an
interesting pattern emerges. It appears that acute activation of the HPA axis
and the subsequent secretion of glucocorticoids does not necessarily “turn
off” the HPG axis and secretion of gonadal steroids; rather, when it is on,
the HPA axis has the potential to outcompete the HPG axis. Recall that in
white-crowned sparrows, a snowstorm increases circulating CORT but does
not reduce circulating T from levels that are sufficient to induce reproduction.
This means that an acute stressor is able to physiologically redirect an animal’s
priorities from sex to survival, but once this stressor is removed and HPA axis
activity is diminished, the HPG axis is still functioning and can immediately
switch priorities back toward sex (reviewed by Adkins-Regan, 2005; Wingfield
& Sapolsky, 2003). Given this, the HPG and HPA axes seem to have evolved
an ability to interact with each other such that reproduction is maximized
HORMONES AND ANIMAL BEHAVIOR 223

within the bounds imposed by an organism’s own physiology and surrounding

environment.
Evolution of steroid mediated sex behavior—Though we have focused on the
activational and repressive relationships between steroids and sex behavior, we
must also emphasize that such relationships do not always exist. This has been
well documented in a number of species and is generally referred to as sex ste-
roid–independent sexual behavior (in contrast to the species mentioned above
that show sex steroid–dependent behavior).
One of the first studies that investigated sex steroid–independent control of
mating was on the red-sided garter snake (Thamnophis sirtalis perietali). In this
species, males emerge from hibernation in the spring before females. Once the
females begin to appear, males mob and attempt to mate with them. In males,
this entire process does not depend on the actions of peripheral T; that is, it is
affected neither by castration nor by the removal of other endocrine tissues
that synthesize androgens or androgenic precursors, such as the adrenals and
pancreas. Instead, male copulation behavior occurs when the testes are
regressed in size, and males use sperm that was made the previous autumn
before hibernation to inseminate females. Only in the summer, after mating
ceases, do males increase the size of their testes and begin making sperm
(which is stored for the upcoming spring mating season) (Camazine et al.,
1980; Crews et al., 1984; Garstka et al., 1982).
Steroid-independent sex behavior also occurs in some females. For example,
in big brown bats (Eptesicus fuscus), removal of the ovaries does not appear to
reduce sex behavior (especially in a natural setting) compared to intact controls
(Mendonca et al., 1996). This result suggests that steroids secreted from the
ovaries are not necessary for female sex behavior, though it is not completely
clear whether other steroidogenic glands play a role. Other studies in common
marmosets (Callithrix jacchus) show that removal of the ovaries does not com-
pletely abolish female copulation solicitations (Kendrick & Dixson, 1984),
highlighting the reduced importance of sex steroids in regulating sex in some
species.
The importance of the dichotomy between sex steroid–dependent and sex
steroid–independent behavior cannot be overstated because it helps uncover
why evolution has potentially coupled the actions of these hormones with
reproductive behavior in the first place. In a seminal paper, David Crews
and Michael Moore (1986) formulated the logic behind this argument. They
suggest that there may be an evolutionary advantage for the brain to evolve
mechanisms to detect elevated levels of peripheral sex steroids when mating
occurs at the same time as gamete production. This ability, in theory, allows
224 ANIMAL BEHAVIOR

organisms to coordinate their own mating behavior with processes underlying

gamete maturation to maximize reproductive success. On the other hand,
Crews and Moore (1986) suggest that there may be less of an evolutionary ad-
vantage for the brain to detect high levels of sex steroids when mating and
gamete production are temporally dissociated because these two events do
not have to be accurately or closely coordinated with each other to maximize
fitness. In such instances, other mechanisms may evolve to govern sex behav-
ior. Based on this framework, it is predicted that sex steroids are more likely
to mediate behavior that is tightly linked to reproduction rather than behavior
that is not. We will see that this relationship (more or less) applies to numer-
ous types of behavior, such as courtship, aggression, and parenting, and greatly
helps explain rich species variation in behavioral phenotypes.

Courtship Behavior
Courtship behavior is often an essential component of reproduction and
has numerous functions, such as species recognition, coordination of repro-
ductive physiology, and assessment of mate quality. In many species, different
features of courtship behavior have evolved to be immensely complex and
involve vocalization displays, visual displays, and stereotyped movements of
the body. Given the tight link between courtship and sex, it is unsurprising
that many types of courtship behavior are guided by the actions of steroids,
particularly sex steroids.
Steroids and vocalizations—Bird song is arguably the best-studied form of
courtship behavior in wild animals. This is particularly true in passeriform birds
(perching birds), which have evolved a unique and complicated network of brain
regions that control how and when a bird sings. Nearly all of the areas within this
neural song system are sensitive to sex steroids, and it is widely understood that
androgens, estrogens, and glucocorticoids act on these regions to dramatically
affect song learning and production. A relatively simple, yet elegant, example of
the relationship between sex steroids and bird song occurs in the grey partridge
(Perdix perdix) (Fusani et al., 1994). Males and females flock together in large
social groups during the winter but then pair off and form pair bonds during
the spring. Males sing a species-specific song throughout the entire year; this song
helps gather individuals in the winter when birds flock, and its helps attract mates
in the spring when birds begin to pair. Experiments show that endogenous
plasma T, which is basal in the winter and rises in the spring, helps mediate struc-
tural changes to the male song that help distinguish a “noncourtship song” from
a “courtship song” (Fusani, 2008; Fusani et al., 1994).
HORMONES AND ANIMAL BEHAVIOR 225

In general, it is unclear whether the observed effects of T on the song sys-

tem are universal among species. For instance, work in zebra finches (Taenio-
pygia guttata) shows that long-term T treatment decreases the fundamental
frequency of the male’s courtship song (Cynx et al., 2005), whereas studies
in Bengalese finches (Lonchura striata) reveal that T treatment increases male
singing rate in the presence of a female but does not alter structural compo-
nents of song (Ritschard et al., 2011). It is also not well understood how this
variation influences mate choice, because female preferences for components
of male song can be species specific. Nevertheless, it is also quite clear that
sex steroids, such as T, play an instrumental role in modulating the motivation
and ability for males to sing and court females (Pinxten et al., 2002; Riters et
al., 2000).
Birds, however, are not the only animals that court mates via acoustic
signals. The Gulf toadfish (Opsanus beta) is a species in which males build
and protect nests in the shallow waters of the Gulf of Mexico. While defend-
ing their nest, males emit calls known as boatwhistles to attract females for
spawning and to interact with other territorial males. Studies suggest that the
11-ketotestosterone (11KT), an androgen that cannot be converted to E and
that is found exclusively in fish, modulates male calling behavior by acting
rapidly (within minutes) on vocal motor circuits found in the brain
(Remage-Healey & Bass, 2006a). Estrogens, in contrast, cannot activate such
circuits (Remage-Healey & Bass, 2006b). Taken together, these data imply
that reproductive calling in male Gulf toadfish is partially governed by andro-
genic pathways and not by estrogenic pathways.
Even more interesting is that the steroidal mechanism of vocal control in Gulf
toadfish appears to differ from that of their close relative, the plainfin midship-
man fish (Porichthys notatus). Male midshipman fish are similar to Gulf toadfish
in that they also build and defend nests for reproduction and then entice females
to their nest by broadcasting courtship calls (known as hums). However, unlike
Gulf toadfish, the calls of midshipman males are affected by actions of both
androgens and estrogens on the brain’s vocal control centers (Remage-Healey
& Bass, 2004, 2006b). To this end, midshipman males also express abundant
aromatase around these brain regions, which in turn suggests that androgens
are locally converted to estrogens for rapid action within the brain (Schlinger
et al., 1999). Exactly why Gulf toadfish and midshipman fish differ in terms of
steroidal control of courtship calling behavior is not well understood, but
researchers speculate that such variation may result from other evolved differences
in mating systems. For example, the midshipman fish has also evolved a male
morph that does not act territorial and instead “steals” fertilizations by sneaking
226 ANIMAL BEHAVIOR

into a territory-holder’s nest and quickly spawning (Remage-Healey & Bass,

2006b).
Another fascinating aspect of the hormone-courtship relationship in mid-
shipman fish is observed in females. Roughly one month before female mid-
shipman fish begin spawning, they show peaks in circulating androgens and
estrogens. Both of these hormones act to fine-tune the encoding properties
of neural cells in the inner ear so that they optimally respond to male hums
(Sisneros, Forlano, Deitcher, et al., 2004; Sisneros, Forlano, Knapp, et al.,
2004). This study supports the idea that steroids are capable of modulating
sensory-system plasticity in females for optimal assessment of courtship
behavior.
Steroids and elaborate displays—Production of acoustic signals is not the
only means of courting mates. Instead, males of some species display for
females as a way of “showing off” their quality, and birds are often studied in
this regard because they show some of the most elaborate displays. Manakins,
for example, are a family of tropical birds that includes nearly 60 separate spe-
cies. Most male manakins perform highly complex displays that involve
remarkable behavioral coordination. The golden-collared manakin (Manacus
vitellinus) has been studied the most, especially in regard to the influence that
steroids have on display behavior. Males of this species aggregate at leks
(assembly areas where animals display and court potential mates) in the low-
land rainforest, and each male makes his own display arena by clearing away
the leaf litter from a small patch of forest floor. When a female arrives at the
lek, the males all begin to display by rapidly jumping among saplings around
their arena and forcefully flipping their wings above their head to produce a
loud, firecracker-like snap. Studies show that while T activates this courtship
repertoire, pharmacologically blocking androgen action virtually eliminates it
(Day et al., 2007; Fusani et al., 2007). This suggests that androgens, not estro-
gens, mediate golden-collared manakin displays. Also intriguing is a study that
shows that the skeletal muscles that control wing and leg movement during
courtship are especially adapted in this species to be extremely sensitive to
androgens (Feng et al., 2010). This is consistent with the idea that androgens
signal these muscles to enhance their strength or motor performance during
courtship. The data are also consistent with the notion that the action of
T per se provides a mechanism through which signal honesty evolves.
The lance-tailed manakin (Chiroxiphia lanceolata) is yet another species in
this family that has an extraordinary courtship display. In this case, breeding
males are socially dominant, and they co-opt socially subordinate males to par-
ticipate in a courtship dance. The dance itself consists of the males leaping
over each other on a tree branch as though they are playing leapfrog, while a
HORMONES AND ANIMAL BEHAVIOR 227

nearby female watches the two males (but only mates with the dominant
male). The dominant males have higher plasma androgens than subordinate
males (DuVal & Goymann, 2010). Because subordinate males are able to par-
ticipate in the dance even though they have relatively low androgens, these
data suggest that androgens do not necessarily mediate the ability to court.
Instead, androgens may indirectly influence courtship by affecting a male’s
status compared to others within the population (DuVal, 2007; DuVal &
Kempenaers, 2008).
Taken together, experiments that investigate the hormonal control of mana-
kin courtship behavior suggest an interesting pattern—it appears that selection
may couple the effects of sex steroids with courtship traits that most closely relate
to male copulation, as opposed to those traits that trigger the onset of courtship.
In other words, in golden-collared manakins, evidence suggests that androgens
are tightly associated with the motor skills required for male displays, and other
work suggests that females base their mate choice decisions on these attributes
(Barske et al., 2011). On the other hand, in lance-tailed manakins, evidence sug-
gests that androgens are tightly associated with male social status because this ulti-
mately dictates whether a male mates with the female for whom he performs
(DuVal, 2007; DuVal & Kempenaers, 2008).

Parenting Behavior
To grow and mature into reproducing adults, newly hatched or born off-
spring sometimes require care from their parents. The amount and type of care
that parents give typically varies across species. Thus, selection has sculpted
numerous mechanisms to govern how and when animals tend to their young,
and steroid action is but one of these mechanisms. Keep in mind, however, that
steroids do not singlehandedly influence parent-offspring interactions. Rather,
steroids act in concert with a host of other hormones, such as prolactin and other
neuropeptides, that can be equally important in affecting parental behavior.
Steroids and parental behavior—In mammals, females provide most of the
care given to their offspring. To elucidate the mechanisms that control this
care, researchers first examined the effects of hormones (e.g., steroids) that
linked with pregnancy itself. Indeed, steroids fluctuate significantly in preg-
nant mothers. For example, in sheep (Ovis aries) and mice (Mus musculus),
E and P levels increase slowly over the course of gestation, and then P rapidly
drops before parturition. Old World primates, including humans, show a
slightly different pattern. These species exhibit relatively high levels of E and
P for much of pregnancy, and both of these steroids precipitously decrease
before birth (reviewed by Nelson, 2000).
228 ANIMAL BEHAVIOR

To test if the temporal patterning of these steroids influences maternal behav-

ior, researchers ovariectomized (removed the ovaries of) a group of rats (Rattus
norvegicus) that had never given birth and treated them with E followed later
by P. Important to note is that P administration was stopped two days prior to
cessation of E administration. This treatment, in effect, mimicked the normal
hormone profile that expecting rat mothers experience. Females were then given
pups, and the hormone treatment induced full and appropriate maternal behav-
ior within 35 hours of pup presentation (Moltz et al., 1970). Thus, the results
not only suggest that E and P are sufficient to trigger the onset of maternal
behavior, but they also suggest that the time at which these hormones act within
the organism is similarly important for the activation of maternal care (reviewed
by Nelson, 2000). Additional evidence underscores this latter point, showing
that P interacts with E early in the pregnancy to induce maternal behavior, but
P by itself in the late stages of pregnancy acts to inhibit maternal behavior
(Sheehan & Numan, 2002). Thus, the drop in P toward the end of pregnancy
appears to be an essential component of proper female parental behavior.
Steroids and male parental behavior—Because males of most species do not
provide care for their offspring, far less is understood about interactions
between steroids and paternal behavior. For the most part, research on this
topic has centered around the effects of T on male parental behavior. This is
because, as we have seen thus far and will continue to see in this chapter,
T is an important regulator of male reproduction and can be involved in
T-mediated trade-offs between parental interactions and sex and aggression
(see sections below) (Wingfield et al., 1990).
In general, T is thought to suppress paternal behavior (reviewed by
Adkins-Regan, 2005). This point is illustrated well in male birds, as research-
ers have experimentally elevated T levels in wild males and then observed the
behavioral repercussions of this treatment. Such an experiment in dark-eyed
juncos (Junco hyemalis) showed that T-treated males spent more time calling
and less time feeding their young compared to control males. Moreover, the
females that paired and mated with the T-treated males were forced to com-
pensate for their partner’s lack of parental investment. The female mates of
the T-treated individuals fed their offspring more and spent more time brood-
ing their clutch than females that mated with control males (Ketterson &
Nolan, 1992; Ketterson et al., 1992). The precise mechanisms through which
T suppresses paternal behavior are not well understood. T does not appear to
affect other hormone systems that are integral to a male’s ability to parent his
offspring, such as the prolactin system (Schoech et al., 1998). Thus, one pos-
sibility is that T causes males to change how they spend their time (see later
sections) such that they allocate less time to paternal endeavors.
HORMONES AND ANIMAL BEHAVIOR 229

Based on the bird studies described in the paragraph above, another way to
look at the relationship between T and paternal behavior is that low T levels
facilitate paternal interactions. This idea has been supported by ample
research, such as work in male common marmosets. In this mammal, males
and females form strong pair bonds and males provide care for their young.
Studies show that socially isolated males decrease plasma T as they are exposed
to the scent of their newborn offspring. Interestingly, this effect depends on
the male’s experience as a father, with first-time fathers showing no such drop
in T after exposure to their offspring’s scent (Prudom et al., 2008).
While the negative relationship between T and paternal behavior is compel-
ling, there are also exceptions to this rule in mammals, fish, and birds (Lynn,
2008; Marler et al., 2003; Rodgers et al., 2006). The exceptions generally
appear when paternal behavior overlaps with temporal expression of aggression
and mating behavior, unlike birds, in which there is often a temporal dissoci-
ation between setting up a territory and mating, which is then often followed
by paternal care. The California mouse (Peromyscus californicus) is one of these
exceptions, as males and females mate for life (Ribble, 1991) and males pro-
vide extensive care for their offspring (Becker et al., 2010; Bester-Meredith
& Marler, 2003; Bester-Meredith et al., 1999; Frazier et al., 2006; Marler
et al., 2003). In fact, if the male is removed from his mate, the number of pups
that survive to leave the nest is dramatically reduced (Gubernick & Teferi,
2000). Experiments have uncovered that this exemplary care is influenced by
T, which is released from the gonads and then converted to E in the brain to
activate paternal behavior (Trainor & Marler, 2002). This may be one mecha-
nism for avoiding high costs of T but maintaining a positive association
between a peripheral hormone and a behavior because aromatase acts in spe-
cific brain regions to convert T to E. Males mate when the female has a post-
partum estrus (Gubernick & Alberts, 1987), and expression of paternal
behavior therefore generally overlaps with the expression of sexual and
mate-guarding behavior. Male T levels during courtship can even predict
future paternal behavior (Gleason & Marler, 2010). Thus, California mice
provide an example of the plasticity in the evolution of the links between tes-
tosterone and different reproductive behaviors, including paternal behavior.

Aggressive Behavior
The relationship between steroid hormones and aggression is as interesting
as it is complex, and one reason for this is that aggression itself has many func-
tional definitions. For example, some species use aggression to acquire re-
sources vital for survival, whereas others use it to defend their territories or
protect themselves from predators. As we indicate above, it is generally
230 ANIMAL BEHAVIOR

predicted that steroids (particularly sex steroids) mediate aggression when it

relates directly or indirectly to reproduction. Of course, additional factors,
such as behavioral syndromes or “personalities,” are thought to influence
the relationship between steroids and aggression (e.g., Carere et al., 2003),
though a detailed discussion of these interactions is beyond the scope of this
chapter.
Androgenic control of aggression—Most studies of the hormonal basis of
aggression have focused on modulation of male aggressive phenotypes. Per-
haps the most classic example comes from research in a population of red deer
(Cervus elaphus) on the island of Rum in Scotland (Lincoln et al., 1972).
Males live together peacefully during the majority of the year. During late
summer and early fall, however, males vigorously defend rutting areas
(patches of grass land) attractive to females, and in some cases this defense
requires males to fight each other with their large, bony antlers. Obviously,
winners of these fights acquire more reproductive opportunities with fecund
females. Testosterone is integral to this process; it increases in late summer
and not only activates aggressive behavior but also hardens the antlers by cut-
ting off the blood supply to the so-called velvet tissue that surrounds the ant-
lers (Lincoln et al., 1972). Researchers have castrated males and shown that
this causes males to shed their antlers prematurely before the summer and fall
precipitously in social rank.
Similar findings were discovered in Galápagos marine iguanas (Amblyrhyn-
chus cristatus) (Wikelski et al., 2005). In this species, older and larger males
court females by acquiring and defending small territories. If T’s physiological
effects and conversion to E are blocked in these males, their ability to maintain
their territories is significantly reduced and females move elsewhere to mate.
At the same time, younger and smaller males employ alternate mating strat-
egies (alternative methods of achieving reproductive success) to acquire mating
opportunities; namely, they either sneak sex from females that they encounter
on an older male’s territory or they roam among territory clusters and have sex
with females that temporarily leave an older male’s territory. Interestingly, if
T is given to either of these “sneaker” or “satellite” males, they will attempt
to establish their own territories to attract female mates (even if this means that
they incur severe injuries from fighting larger males or establish territories in
unusual locations; costs of T are discussed later in the chapter). This study
therefore indicates not only that androgens themselves are main activators of
male aggression and territoriality but also that androgen can help males switch
between different aggression-related mating strategies.
Based on the studies described above, it may seem that T universally pro-
motes aggression and male dominance. However, this idea is overly simplistic;
HORMONES AND ANIMAL BEHAVIOR 231

in fact, in some species, T neither is the hormone that exerts the most control
over aggressive phenotypes, nor is singlehandedly capable of elevating
individuals within a social hierarchy (an elegant example is provided by
Apfelbeck & Goymann, 2011). For example, when researchers pharmaco-
logically blocked song sparrows’ ability to detect circulating androgens,
males showed only a partial reduction in aggressive behavior in response to
a territorial intrusion (Wacker et al., 2010). In other words, this treatment
does not completely abolish aggression, as it does in marine iguanas. Regard-
ing T and dominance, studies in chickens (Gallus gallus domesticus) show
that socially subordinate males are unable to climb the social hierarchy even
if their otherwise low circulating T is experimentally elevated (Guhl, 1968).
This effect is termed social inertia, and described how prior social experience
plays a significant role in keeping animals in their established positions
within a given social hierarchy regardless of their T levels (Archawaranon
et al., 1991; Wiley et al., 1999).
Within California mice and other species there is also evidence for more com-
plex control of competitive behavior. Male California mice experience a transient
increase in T after a competitive encounter with another male. Transient
increases in T after a male-male encounter, however, can only elicit a maximal
ability to win a future encounter in a laboratory paradigm when specific condi-
tions are met. First, the T pulse needs to be accompanied by a winning experi-
ence; in fact, a winning experience and T contribute equally to the future
ability to win (Fuxjager, Oyegbile, et al., 2011). Second, the encounters need
to occur in the home cage (“resident advantage”) (Fuxjager & Marler, 2010;
Fuxjager et al., 2009). Context is therefore critical for the effects of T on behavior
(Gleason et al., 2009). Further support is found across Peromyscus species. While
California mice are territorial and monogamous, the closely related white-footed
mouse (P. leucopus) is less territorial and also does not normally express a T pulse
after an encounter. If the T pulse is added to the experience of winning, however,
white footed mice can reach the maximal level of winning behavior expressed by
California mice (Fuxjager, Montgomery, et al., 2011).
Estrogenic control of aggression—Many researchers have sought to elucidate
other steroidal mechanisms that guide aggression, and much of their work
focuses on estrogens. This may seem intuitively odd, given that estrogens are
considered feminizing by the standards of popular culture; however, as we will
see, estrogens have a profound effect on aggressive phenotypes. This, for
example, was elegantly demonstrated by a study that compared antagonistic
behavior between male mice that were genetically normal (wild type) and
male mice that were genetically engineered to be insensitive to E (Ogawa
et al., 1997). To do this, mice from both genetic strains were subjected to a
232 ANIMAL BEHAVIOR

“resident-intruder test,” which entails documenting the aggressive behavior of

a focal mouse (the resident) after an unfamiliar male mouse was introduced
into his home cage. When wild-type mice were used as focal animals, they
spent a great deal of time attacking the intruders. On the other hand, when
estrogen-insensitive mice were used as focal mice, they spent comparatively lit-
tle time attacking intruders. These data provide compelling support for the
idea that estrogens are capable of activating male aggression because males that
are unable to detect endogenous estrogens reduce displays of aggression in
contexts where such behavior is normal and appropriate.
In a second study, researchers expanded upon this idea by examining
whether estrogens activate certain types of aggressive behavior (Ogawa et al.,
1998). Thus, this time, aggression was measured from both wild-type and
estrogen-insensitive males that were used as intruders, and not residents.
Moreover, in these trials, resident males had been specially trained to attack
intruders (i.e., the wild-type and estrogen-insensitive males) with speed and
vigor. Wild-type intruders spent most of their time in these encounters dis-
playing defensive aggressive behavior (i.e., attacking as a way of protecting
oneself), though some of them still spent time offensively attacking the resi-
dent. Interestingly, estrogen-insensitive mice spent nearly the same amount
of time during these encounters displaying defensive aggression, but they
spent virtually no time displaying offensive aggression. These findings are con-
sistent with the notion that estrogens mediate aggression, but they also suggest
that estrogens selectively affect qualitative elements of aggressive repertoires.
The balance between estrogen and androgen control of aggression at a cellular
level is also illustrated in California mice. The effects of T on attack latency were
androgen based and not influenced by aromatase. In contrast, baseline levels of
attack latency were increased by aromatase, suggesting that estrogen receptors
may be more associated with baseline aggression (no T manipulations) and
androgen receptors with experience-induced aggression (Trainor et al., 2004).
Steroids and female aggression—Our discussion of aggression has mainly
hinged on males because far less is known about the mechanisms of female
aggression. This discrepancy is unfortunate because there is a large literature
that shows that females also behave aggressively in certain contexts. For exam-
ple, in some bird species, female-female aggression can help enforce male sex-
ual exclusivity and thus increase female fitness (Liker & Szekely, 1997). This is
the case in European starlings (Sturnus vulgaris) (Sandell, 1998; Sandell &
Smith, 1996), and researchers therefore tested the mechanisms of such interfe-
male aggression in this species by first giving wild females T implants during
the prebreeding season and then recording aggressive behavior during staged
interactions with other females (Sandell, 2007). The results showed that
HORMONES AND ANIMAL BEHAVIOR 233

T-treated females behaved more aggressively in female-female interactions

compared to control females. The degree of monogamy was also higher
between males and T-treated females than between males and control females.
Thus, these results suggest that androgens are at least sufficient to control
adaptive aggression in a female bird.
Another interesting species in which to consider the mechanisms of female
aggression is the African black coucal (Centropus grillii). This species of bird
has reversed sex roles, meaning that females actively defend territories and
attract males that care for offspring. Interestingly, studies in this species show
that female territoriality can be influenced by P (Goymann et al., 2008).
Female California mice were first found to respond to female-female encoun-
ters with changes in postencounter P levels (Davis & Marler, 2003). Recall that
this steroid is involved in regulating female sex behavior; thus, it makes sense that
selection might co-opt P to modulate other types of behavior associated with
reproductive success in sex-role-reversed species. This is not to say that sex ste-
roids, such as T and E, are not also involved in territorial aggression in African
black coucals. Even though males and females of this species show sex-steroid
profiles that are similar to birds with “traditional” sex roles, researchers have sug-
gested that the key to aggression in females might be rooted in the brain’s ability
to detect and respond to androgens. This idea has merit, as females are more sen-
sitive to androgen action in brain areas that control avian sexual territorial behav-
ior than are males (Voigt & Goymann, 2007).
To make things more interesting, there is also work that suggests that pro-
gestins and androgens are able to work together to modulate behavior. In tree
lizards (Urosaurus ornatus), males exposed to high levels of P and T during
early development will develop an orange dewlap with a blue spot and are typ-
ically characterized as territorial, whereas in the absence of these high hormone
levels, males develop a solid orange dewlap and are not only nonterritorial but
become nomadic (Moore et al., 1998).

Costs (and Trade-offs) Associated with Steroid Action

Thus far, we have considered adaptive effects of steroid action on behavior,
but it is essential to recognize that steroid action can be costly. This notion of
“cost” is rooted in the field of behavioral ecology and refers to something that
decreases fitness (ability to survive and reproduce). Moreover, the term is pur-
posefully vague in order to encompass the different ways that fitness can be
reduced. By saying that steroid action is costly, it is generally meant that one
or more properties of steroidal signaling negatively impacts an individual’s
ability to pass genes onto the next generation.
234 ANIMAL BEHAVIOR

In terms of cost, the steroid hormone given the most attention is T, and the
reason for this is grounded in studies of “honesty” in animal signaling. Recall
that T (in males) mediates the activation, maintenance, and development of
many traits used to acquire sexual opportunities and that such traits are usually
considered honest indicators of an individual’s quality. This, in theory, means
that any negative effect (cost) of T is a pathway through which sexual traits or
signals remain honest, because only high-quality males are capable of enduring
the adverse effects of T and still producing an attractive signal.
Androgens and mortality—Some of the first evidence to show that T is costly
to an organism’s fitness comes from work in mountain spiny lizards
(Sceloporus jarrovii), which inhabit the southwestern deserts of the United
States (Marler & Moore, 1988). Males of this species maintain low levels of
T from winter to early summer, when individuals aggregate and begin to form
territories. However, once fall arrives, T levels increase and induce vigorous
territoriality. In one experiment, researchers took males in the summer and
gave them T implants. This, in effect, increased the otherwise low amount
of circulating T to high levels of T normally observed in the fall. After this
treatment, males were released back onto their territories and observed for
the following weeks. Interestingly, males that had T levels experimentally
elevated were more likely to die than males that maintain normal summertime
T levels. What is therefore remarkable about this study is that it highlights an
enormous cost of T: mortality.
Of course, the follow-up question to this research in mountain spiny lizards
is, how does T actually affect whether an individual lives or not? The answer lies
in the details of the experiment, particularly with respect to the behavioral
changes that transpired in lizards treated with T (Marler & Moore, 1991).
For instance, researchers found that T simultaneously increased aggressiveness
and territoriality and decreased time spent looking for food. This, in effect,
means that T likely influences survivorship either by intensifying predation,
because males are more conspicuous during territorial interactions or by exac-
erbating energetic constraints because males forage less. To test between these
hypotheses, researchers conducted a separate experiment that was nearly iden-
tical to the study described in the paragraph above: they collected lizards,
manipulated their T levels, and observed their subsequent behavior in the
wild. However, in this case, a select number of individuals given T implants
were also given food supplementation. As it turns out, the negative effect of
T on mortality was completely rescued by providing males with extra food,
and there was no difference in conspicuousness (i.e., territorial behavior)
between T-implanted males given food supplements and those given nothing.
Thus, these results tell us that the cost of T is ultimately driven by the
HORMONES AND ANIMAL BEHAVIOR 235

energetic constraints of territorial defense rather than increased susceptibility

to predation (Marler & Moore, 1991; Marler et al., 1995); however, energy
issues may lead to increased predation.
Androgens and behavioral trade-offs—Although the research in mountain
spiny lizards helps illustrate an interesting pathway through which T nega-
tively affects fitness, it also helps lay the groundwork for a broader examina-
tion of how T can be costly because of its influence on behavioral
reallocation (i.e., the proportion of time dedicated to one behavior versus
another behavior). Like many steroids, T has pleiotropic effects (one gene
affects more than one behavior) on behavioral output that are sometimes
antagonistic; thus, while T might turn on one particular type of behavior, it
might simultaneously turn off another disparate type of behavior. This is fre-
quently observed in birds, whereby T is thought to generally increase aggres-
sion and territoriality and decrease paternal investment (see sections above).
This means that, in many species, male birds are not able to fully invest in
territory protection and offspring care at the same time. To overcome this,
selection in theory should favor mechanisms that optimize fitness and avoid
antagonistic effects of T. The so-called “challenge hypothesis” was put forth
by John Wingfield and colleagues in the 1980s to elucidate such a mechanism
(Wingfield et al., 1987; Wingfield et al., 1990). This idea posits that plasma
T increases when the social environment is unstable but decreases back to nor-
mal baseline levels when the social environment becomes stable again. Thus,
in many avian species, T levels increase in males when competition for territo-
ries intensifies and breeding occurs, but precipitously drops when males begin
to provide paternal care for their hatchlings. At the same time, in bird species
in which territoriality is relatively relaxed, T levels appear to fluctuate less.
This suggests another intrinsic prediction to the challenge hypothesis, namely
that the ability to adjust one’s endogenous T levels in response to the social
environment is related to the evolution of key social traits, such as mating sys-
tems and parental investment strategies.
Support of the challenge hypothesis has been documented in diverse taxa
beyond birds. This is remarkable because it implies a relative level of evolu-
tionary uniformity in the mechanisms of combating the costs of T. However,
it is important to know that there are many species that do not conform to the
predictions of the challenge hypothesis or that vary with regard to its predic-
tions (Knapp & Moore, 1995; Moore, 1987; Scriba & Goymann, 2010;
Wingfield et al., 1982). Why and how these exceptions occur is not always
known, but it is assumed that such species have evolved their own mechanisms
for coping with the antagonistic (and thus potentially costly) effects of
steroids, such as T. A good example is found in the dusky-footed wood rat
236 ANIMAL BEHAVIOR

(Neotoma fuscipes), whereby castrated males still exhibit seasonal changes

in aggressive behavior (i.e., more aggression during the breeding season)
(Caldwell et al., 1984). This finding is remarkable because it documents that
aggression during the breeding season can be mechanistically decoupled from
seasonal changes in gonadal hormone secretion. Given this, it seems possible
that the mechanisms that control adaptive aggression need not constrain other
suites of reproductive behavior that are mediated by T. Another example of
this is found in multiple species of tropical birds, as males appear to be season-
ally territorial but do not show seasonal changes in plasma T (Levin &
Wingfield, 1992). Some studies suggest that the androgen precursor DHEA,
which is secreted from the adrenals, mediates territorial aggression (Hau
et al., 2004). Alternatively, research suggests that synthesis of androgens de novo
in the brain can modulate territorial aggression (Pradhan et al., 2010). This is
especially intriguing because it implies that individual brain nuclei locally pro-
duce their own steroid hormones to influence behavioral output, while not
changing endogenous steroid levels throughout the entire brain and body.
Androgens and immunity—The effect of T on behavioral reallocation is not
the only reflection of this steroid’s costs; in fact, there is another mechanism
through which T is considered costly, which is encompassed by the “immuno-
competence handicap hypothesis” (ICHH). This idea proposes that T sup-
presses immune function and thus renders individuals more susceptible to
pathogen and parasite infection (Folstad & Karter, 1992). The prediction that
follows this hypothesis is that elevated T decreases disease resistance and
increases parasite load. As it turns out, many studies have found substantial
support for the predictions of the ICHH. For example, by experimentally
increasing T levels in male dark-eyed juncos, researchers successfully reduced
the male cell-mediated immune responses (Casto et al., 2001). Similarly, work
in house finches (Carpodacus mexicanus) revealed that T levels are positively
associated with levels of parasitic infection (Duckworth et al., 2001). Yet while
these studies illustrate a seemingly clear picture of the relationship between
T and the immune system, other work suggests that this picture is much more
variable. For instance, in greenfinches (Carduelis chloris), researchers gave
males either T implants or blank implants (controls) and then infected these
individuals with a virus that commonly affects this species. Results revealed
that T implantation actually decreased the concentration of virus in the blood-
stream early after infection and subsequently increased such virulence later in
the infection (Lindstrom et al., 2001). This result obviously indicates that
the relationship between T and immunity is highly complex, and this general
supposition is supported by a vast number of other studies (Roberts & Peters,
2009; Ruiz et al., 2010).
HORMONES AND ANIMAL BEHAVIOR 237

Nevertheless, the disparity among studies about how T alters immunity

and parasitism leads to the question of why so much variation exists in this
interaction. The answer to this question might lie in additional studies in
mountain spiny lizards (the same species used to assess effects of T on mortal-
ity) (Fuxjager, Foufopoulos, et al., 2011). Recall that T makes males spend
more time being territorial and less time eating. In this experiment, males in
the summer were given either a T implant, which increased circulating T to
levels normally seen in the fall breeding season, or a blank implant (controls).
All individuals were released back onto their territories and then re-captured
a few weeks later. The researchers examined the carcasses for a number of
different types of parasite, such as those that live on the outside of the body
(ectoparasites) and the inside of the body (endoparasites). Remarkably,
T-implanted males showed higher accumulation of ectoparasitic mites but
lower accumulation of gastrointestinal nematodes. There were also a number
of other intestinal and blood-borne parasites that were unaffected by T.
These opposing effects of T on parasitism are thought to be due to
T-mediated changes in behavior; that is, T likely increased mite infection
because males roamed more while being territorial, whereas T decreased intes-
tinal parasite load because males ate less. Incidentally, work on the Grant’s
gazelle (Nanger granti) showed consistent results, such that the impact of T
on parasite infection depended largely on host and parasite behavior and ecol-
ogy (Ezenwa et al., 2012). This work is important because it reveals how other
factors, in this case behavior, might confound the costs of T and lead to varia-
tion in T-immunity relationships throughout the animal kingdom.

Mechanisms of Steroid Action

In most of the examples that we describe above, we speak about the
“effects” that steroids have on behavior without addressing how these effects
are exerted. However, it is crucial to understand how and where steroids act
in order to gain a greater appreciation of the complex steroid-behavior rela-
tionships we have addressed above.
Steroid receptors—Each class of steroid binds to its own special class of ste-
roid receptor: androgen receptors (AR), estrogen receptors (ER), glucocorti-
coid receptors (GR), and progestin receptors (PR). To varying degrees,
steroid receptors are expressed in virtually all nucleated cells in the body. This,
in effect, means that most tissues are steroidal targets, and the sensitivity of
each target is determined by the abundance or amount of steroid hormone
receptors expressed there. In the classic model of steroid action, steroids cross
through cellular membrane, bind to their intracellular receptors, migrate
238 ANIMAL BEHAVIOR

(as a hormone-receptor complex) into the cell nucleus, attach as a complex to

hormone-response elements located within the DNA, and influence gene
transcription. This process is generally thought to occur more slowly (hours
to days) because effects of altered transcription can take time to accumulate
and modify phenotype.
Given the effects that steroid receptors have on the genome, we can think
of steroid receptors as transcription factors (a protein that controls the flow
of information from DNA). This is best exemplified by studies that show that
steroid receptors are autoregulatory in that they are capable of influencing
their own genetic transcription. For example, castration reduces the amount
of circulating T, which in turn alters the expression profile of AR throughout
much of the brain (Lu et al., 1998, 1999). Research also indicates that the dif-
ferent classes of steroid receptors are able to affect the transcription of other
classes of steroid receptors, Activation of PR, for instance, impacts both
the cellular production of itself and ER (Turgeon et al., 1999; Turgeon &
Waring, 2000). Taken together, this means that simple changes in the amount
of steroids that circulate in the blood can have broad and tremendous impacts on
how organisms detect steroids. Moreover, because steroid levels continually fluctu-
ate in response to extensive environmental and social stimuli (Hirschenhauser &
Oliveira, 2006; Hirschenhauser et al., 2003; Wingfield et al., 1990), we can
think of steroid detection systems as highly fluid and completely dependent
on the state of an animal at a given time.
Of course, the mechanistic complexity of steroid detection is still being
unraveled. Research shows, for example, that steroids can have rapid effects
on cellular targets, and it is thought that this occurs through nonclassic steroid
receptors that are bound to a cell’s lipid membrane. An elegant study in gold-
fish (Carassius auratus) has established the relevancy of these mechanisms to
adaptive behavioral phenotypes. During the breeding season, males of this spe-
cies preferentially associate with ovulating females and rapidly (under an hour)
produce milt (sperm and seminal fluid) in response to both female sexual cues
and the presence of male competitors. This behavior also coincides with the
rapid elevation of T. Researchers were therefore able to link these behavioral
and physiological phenomena by showing not only that T induces milt pro-
duction via its conversion to E and subsequent activation of ER, but also that
the ER being activated is a membrane-bound receptor (Mangiamele &
Thompson, 2012). This, of course, accounts for the rapidity of the adaptive
sexual response. Based on this and other work (Trainor, et al., 2007), it is ap-
parent that the nature of steroid action and the flexibility of its underlying
mechanisms are highly complex and not completely understood. How many
HORMONES AND ANIMAL BEHAVIOR 239

other types of behavior are affected by membrane-bound steroid receptors, classic

intracellular receptors, or both? The answer is yet to be determined.
Brain regions involved in social behavior—Two interconnected neural cir-
cuits, the “social behavior neural network” (also known as the social behavior
network) and the mesolimbic reward system, appear to be fundamental to
the regulation of social behavior across vertebrate taxa. Research in mammals
and birds has demonstrated the importance of brain areas in these circuits in
regulating naturally rewarding behaviors, such as sex, pair bonding, parental
care, winning a fight, and sociality and vocal communication. These findings
have since been expanded to fish, amphibians, and reptiles (O’Connell &
Hofmann, 2011a, 2011b, 2012).
The social behavior neural network is a group of reciprocally connected,
steroid-sensitive brain regions that have been implicated in multiple social
behaviors ranging from aggression to courtship and mating behavior (Good-
son, 2005; Newman, 1999). This network includes the preoptic area (POA),
the extended medial amygdala (including medial amygdala [MeA] and bed
nucleus of the stria terminalis [BST]), the lateral septum (LS), the ventrome-
dial hypothalamus (VMH), the periaqueductal or central gray (PAG), and
the anterior hypothalamus (AH). Each of these brain regions contains steroid
receptors and participates in a variety of social behaviors in many vertebrate
species, indicating that this core group of brain regions underlies steroid-
dependent social behaviors across vertebrate taxa (Goodson, 2005). Sarah
Newman (1999) proposed that the pattern of activity across this network,
rather than activity within any single region, regulates all social behaviors.
The mesolimbic reward system is characterized by large dopaminergic pro-
jections from the ventral tegmental area (VTA) to the nucleus accumbens
(NAc), and also includes the ventral pallidum (VP), striatum (Str), LS, baso-
lateral amygdala (BIA), BST, and hippocampus (HIP; Stellar & Stellar,
1985). These brain regions are steroid sensitive and highly conserved. Natural
rewards (such as food, water, and sex) and many drugs of abuse activate these
neural circuits, and animals will work to activate these circuits.
The social behavior neural network and mesolimbic reward system are
functionally connected, because both can regulate behavior, and are widely
neuroanatomically interconnected (Figure 7.2). Furthermore, these circuits
share two brain regions, the LS and BST, which are both involved in the regu-
lation of social and reward-related behaviors (O’Connell & Hofmann, 2011a).
As we will see below, this social behavior network contains nonapeptide path-
ways in addition to the steroid hormone pathways, as described in the second
primary section of the chapter.
240 ANIMAL BEHAVIOR

Figure 7.2. Side view of brain areas involved in social behavior in a representative
mammal. Brain areas that are part of the social behavior neural network are in dark
gray. Each of these areas projects to each of the others. Brain areas that are part of the
mesolimbic reward system are in light gray. Note that lateral septum (LS) and bed
nucleus of the stria terminalis (BST) are part of both the social behavior neural
network and the mesolimbic reward system.

NONAPEPTIDE HORMONES AND SOCIALITY

While steroids have a large impact on social behavior, the nonapeptides are
another set of hormones and neuromodulators that also influence a variety of
social interactions that have likely emerged with the evolution of sociality.
Nonapeptides are notable because of the developing evidence for their impor-
tance in mediating affiliative as well as aversive behavior. Less is known about
many of the effects of AVP and OT compared to the steroid hormones
described above. There is, however, exciting research involving nonapeptides
that delves more deeply into the role of the brain, in particular the social
behavior neural network and mesolimbic reward system, in mediating the
integration of biochemicals and behavior.
From an ecological perspective, we see that increased sociality occurs with
the emergence of higher levels of maternal behavior with altricial (versus pre-
cocial [relatively mature and mobile at birth or hatching]) offspring. From a
mechanistic perspective, we can speculate that the neuropeptide regulation
of social behavior began with the mother-offspring bond. This social group
expands with the evolution of paternal care or expression of cooperation
between the mother and father and may have led to the formation of a pair
bond to effectively raise altricial offspring, which are hatched or born blind
HORMONES AND ANIMAL BEHAVIOR 241

and helpless. We see a further increase in sociality and affiliative behavior

when animals live in groups and the young are more likely to mix with unre-
lated individuals (Nowak et al., 2011). The roots for hormonal correlates of
these raised levels of affiliative behavior may lie in the association between
OT release in response to vaginocervical stimulation and during parturition.
The peripheral functions of OT seem to have evolved along with related func-
tions in the brain. The bond between mother and young is one of the strong-
est social bonds formed, and there are a variety of mechanisms that contribute
to this bond.
The role of E and P in maternal behavior was described earlier, but OT
brings us a step closer to understanding the complexities of social behavior as
we see how the effect of OT are layered onto the functions of E and P. If
females are primed with E and P in a pattern mimicking the natural changes
that occur during parturition and then are administered OT, the full set of
maternal behaviors is expressed. Blocking vaginocervical stimulation and asso-
ciated increases in OT levels inhibits this process (Keverne & Kendrick, 1994;
Krehbiel et al., 1987; Levy et al., 1992). It is interesting to think about the
pairing of the natural release of OT to aid the physiological aspects of parturi-
tion with OT effects on attachment between the mother and offspring. The
speculated linkage between AVP and pair bonding is much more speculative,
but it may be associated with ejaculation (Normandin & Murphy, 2011).

Pair Bonding Behavior

A pair bond, which typically occurs in monogamous species, is an endur-
ing preferential association between two sexually mature adults and is charac-
terized by selective affiliation, contact, and copulation with the partner. An
extensive body of research on the role of nonapeptides in pair bonding has
focused on the comparison of monogamous vole species that form pair bonds
with nonmonogamous vole species that do not. Studies on monogamous prai-
rie voles (Microtus ochrogaster), in particular, have demonstrated that OT and
AVP play critical roles in pair bond formation (Young et al., 2011; Young
et al., 2005).
Although pair bonds in prairie voles can form during cohabitation in the
absence of sexual activity, mating greatly facilitates pair bond formation, par-
ticularly in males (Insel et al., 1995; Winslow, Hastings, et al., 1993).
Although both OT and AVP may facilitate pair bond formation in both sexes
(Cho et al., 1999; Winslow, Hastings, et al., 1993), OT appears to be more
important in females and AVP more important in males (Cushing & Carter,
2000; Insel & Hulihan, 1995; Winslow, Hastings, et al., 1993). For example,
central infusion (i.e., into the brain) of OT facilitates pair bond formation in
242 ANIMAL BEHAVIOR

female, but not male, prairie voles (Cho et al., 1999; Williams et al., 1994).
Likewise, central infusion of AVP facilitates pair bond formation in male,
but not female, prairie voles (Cho et al., 1999; Winslow, Hastings, et al.,
1993). Interestingly, central infusion of AVP in male montane voles (Microtus
montanus) (a nonmonogamous vole species closely related to prairie voles)
does not induce pair bond formation (Young, Winslow, et al., 1997). These
data suggest that differences in neural target tissues (those containing the
receptors) play a significant role in the regulation of pair bonding behavior
by nonapeptides.
Studies of the neuroanatomical basis of pair bonding have focused largely
on the BST, LS, NAc, and VP (Young & Wang, 2004; Young et al., 2005)
brain areas in the social behavior neural network and the mesolimbic reward
system (Figure 7.2). Monogamous vole species have different AVP and OT
receptor densities in all four of these brain regions compared to nonmonoga-
mous vole species (Insel & Shapiro, 1992; Insel et al., 1994; Smeltzer et al.,
2006; Young et al., 1996), implicating nonapeptide neurotransmission in
these areas in monogamy and its characteristic behavior of pair bonding.
Manipulations of nonapeptide neurotransmission in the BST, LS, VP, and
NAc have revealed that pair bond formation is differentially regulated in male
and female prairie voles. Specifically, in male prairie voles, the formation of a
pair bond requires BST neurons to produce and release AVP into the LS.
Three days of cohabitation with a female induces this neurotransmission in
prairie vole males (Liu et al., 2001; Wang et al., 1994). Also in male prairie
voles, enhanced AVP neurotransmission in the VP facilitates, whereas block-
ing AVP neurotransmission prevents, partner-preference formation (Lim
et al., 2004; Lim & Young, 2004; Pitkow et al., 2001). In comparison,
enhanced OT neurotransmission in the NAc facilitates, whereas blocking
OT neurotransmission prevents, partner-preference formation in female prai-
rie voles (Liu & Wang, 2003; Ross et al., 2009; Young et al., 2001). These
effects are not observed in nonmonogamous meadow voles (Ross et al., 2009).
The differences in nonapeptide receptor distributions among vole species
described above may be due to species differences in the promoter regions of
the genes for AVP and OT receptors. For example, although the coding
regions of the V1aR (a type of AVP receptor) and OTR (OT receptor) genes
are similar across vole species (Young, 1999; Young et al., 1996; Young
et al., 1999; Young, Waymire, et al., 1997), the V1aR gene of monogamous
prairie and pine voles has several polymorphic repetitive microsatellite
sequences in the promoter region that are not found in nonmonogamous
meadow or montane voles (Hammock & Young, 2002, 2004; Young, 1999;
Young et al., 1999). Furthermore, there is considerable individual variation
HORMONES AND ANIMAL BEHAVIOR 243

in the length of this microsatellite within the prairie vole species (Hammock &
Young, 2002). These microsatellite-sequence differences may underlie species
differences and individual variation in receptor expression and associated
behavior. Through selective breeding of prairie voles, Elizabeth Hammock
and Larry Young revealed that male prairie voles with a long microsatellite in
the V1aR gene promoter had higher levels of V1aR binding in the olfactory
bulb and LS than males with a short microsatellite. These males with long
microsatellites were also more likely to form pair bonds after a brief cohabita-
tion with a female and displayed higher levels of paternal behavior (Hammock
& Young, 2005). House mice, which are typically nonmonogamous, that
carry a transgene coding for the prairie vole V1aR exhibited V1aR distribu-
tions similar to prairie voles and displayed higher levels of social affiliation
when injected with AVP (Young et al., 1999). These data support the hypoth-
esis that the differences in the promoter regions of the nonapeptide receptor
genes may underlie species and individual differences in receptor distributions
and associated behavior. This story becomes more complex as we increase our
understanding of the interactions among genetics, behavior, and the environ-
ment and by examining these interactions under more natural conditions (see
the section on aggression below).
The family unit described above allows for increased sociality above that of
the mother and offspring by including the father. Within a family unit this can
be expanded into alloparenting as well, in which offspring take care of their
siblings. Here we focus on a further expansion of sociality, which is the emer-
gence of gregariousness. Within larger groups of animals we see individuals
functioning in larger, more stable groups, and we can further conceive of
changes in behavior that may allow for increases in reciprocal altruism.

Sociality and Spacing of Animal Groups; Gregariousness

Finches and waxbills (family Estrildidae) are birds that exhibit long-term
pair bonds and are biparental but vary dramatically in gregariousness.
Although the majority of estrildid species form small groups when not breed-
ing and loosely distribute in nonterritorial pairs during breeding, a small num-
ber of species either form large social flocks or are territorial (Zann, 1996).
This system has provided a useful model with which to study the role of non-
apeptides in gregariousness while holding reproductive and parental systems
relatively constant across species.
A growing body of work on finches from the James Goodson research
group suggests that nonapeptide neural circuitry (specifically AVT and MT,
the avian homologues of mammalian AVP and OT) of the medial portion of
244 ANIMAL BEHAVIOR

the BST (BSTm) and the LS mediates gregariousness. For example, two highly
gregarious finch species (zebra finches and spice finches [Lonchura punctulata])
exhibit approximately 10 times the number of AVT-containing neurons in the
BSTm than do two territorial species that live in male-female pairs year round
(Melba finches [Pytilia melba] and violet-eared waxbills [Uraeginthus granati-
nus]) and a modestly gregarious species (Angolan blue waxbill [Uraeginthus
angolensis]) (Goodson & Wang, 2006). Additionally, AVT receptors in the
LS are more abundant in the flocking species as compared to the territorial
species (Goodson et al., 2006). These data suggest that increased AVT cir-
cuitry of the BSTm and LS may promote gregariousness.
Gregarious species also have AVT neurons that are more responsive to
social group contexts, and this responsiveness is critical to gregariousness.
For example, exposure to conspecifics, but not a positive nonsocial stimulus,
increased the activity of BSTm AVT neurons in the highly gregarious zebra
finch (Goodson, Kabelik, et al., 2009; Goodson & Wang, 2006). Further-
more, blocking AVT neurotransmission between the BSTm and LS of zebra
finches eliminated their preference for contact with a larger group versus a
smaller group (Kelly et al., 2011), elegantly demonstrating that social group-
ing preferences in this gregarious species are under the direct control of AVT
neurotransmission between the BSTm and the LS.
However, AVT circuitry of the BSTm and LS does not appear to promote
sociality with all conspecifics (individuals belonging to the same species) but
specifically with individuals that elicit attraction and affiliation. For instance,
although reduction of AVT production in the BSTm of zebra finches elimi-
nated the preference for contact with a larger group versus a smaller group, it
increased preference for being closer to conspecifics in general (Kelly et al.,
2011). Additionally, although BSTm AVT neurons in territorial violet-eared
waxbills increased activity in response to exposure to their own pair-bond part-
ner, they decreased activity in response to a same-sex conspecific (Goodson &
Wang, 2006). These data in combination with those previously discussed sug-
gest that BSTm AVT cells are sensitive to the valence of social stimuli in that
they increase responses to stimuli that elicit affiliation and attraction but not
stimuli that elicit aggression and aversion. This assessment is supported by
recent data indicating that AVP neurons in the posterior BSTm of male mice
and AVT neurons in the posterior BSTm of male chickens increase activity in
response to interactions with female conspecifics but not in response to ago-
nistic encounters with male conspecifics (Ho et al., 2010; Xie et al., 2011).
OT-like circuitry seems to play a similar dual role in the regulation of
group-size preferences and preferences for familiar individuals, particularly in
females. For example, in female zebra finches, central infusions of MT
HORMONES AND ANIMAL BEHAVIOR 245

increased the preference for proximity to familiar individuals, and blocking

MT receptors had the opposite effect (Goodson, Schrock, et al., 2009). In
both sexes, central infusions of MT increased the percentage of time spent in
close proximity to a larger group while they decreased the percentage of time
spent in close proximity to a smaller group by a comparable amount. Again,
blocking MT receptors had the opposite effect. The lack of effect of MT
manipulations on total time in contact with conspecifics indicates that while
OT-like circuitry regulates group-size preferences and preferences for famili-
ars, it does not necessarily regulate preferences for all social contact.
Recent evidence in human studies sheds further light on the role of OT in
cooperation with group members and noncooperation with rival out-group
members. People given intranasal OT rather than placebo rate images of neu-
tral faces as more attractive and trustworthy (Theodoridou et al., 2009), are
more willing to interact with familiar participants again (Alvares et al.,
2010), and are more likely to transfer more money to another participant in
a financial trust game (Kosfeld et al., 2005). Interestingly, OT positively influ-
ences trust and cooperation when the protagonist is displayed as trustworthy
or familiar, but not when displayed as antagonistic or unfamiliar (De Dreu
et al., 2011; Kosfeld et al., 2005).
It is interesting to consider that as social groups become stable and there is
an increased probability of reciprocal altruism, trust could become an impor-
tant mechanism for mediating these interactions. Emotions, potentially tied
in with nonapeptides, may act as an emotional memory for past interactions.
We speculate that this may be true for affiliative behavior, as described above,
and perhaps for emotions associated with aggressive behavior, as described
below.

Aggressive Behavior
The relationship between nonapeptides and aggression is a complex one. As
seen previously with respect to pair bonding and gregarious behaviors, closely
related species often differ in their responses to nonapeptide manipulations. This
is also true with respect to aggressive behavior. For example, central infusions of
an AVP antagonist, which blocks binding to the receptor, inhibits aggression in
territorial California mice, but only in a resident-intruder paradigm. In contrast,
aggression in less territorial white-footed mice is not affected by central infusions
of AVP or AVP antagonists (Bester-Meredith et al., 2005). Similar contrasts are
observed in territorial prairie voles compared to less territorial montane voles
(Winslow, Shapiro, et al., 1993; Young, Winslow, et al., 1997). In male golden
hamsters, a solitary and aggressive species, central infusions of AVP into the
246 ANIMAL BEHAVIOR

anterior hypothalamus increases, while blocking AVP neurotransmission in the

anterior hypothalamus reduces, aggression (Caldwell & Albers, 2004; Ferris et
al., 1997; Ferris & Potegal, 1988; Potegal & Ferris, 1989). These studies and
others (Albers, 2012) indicate that AVP facilitates aggression, but not in all spe-
cies and not in all social contexts. Furthermore, prior social experience can
modify an individual’s response to AVP manipulations (Albers, 2012).
Perhaps considering the role of aggression in other social contexts can shed
light on these variations of peptidergic involvement. Animals with a monoga-
mous reproductive strategy often share many other behavioral commonalities
in support of increasing shared reproductive fitness, such as pair bonding,
biparental care, and selective aggression expressed as territory, mate, and off-
spring defense (Clutton-Brock, 1991; Kleiman, 1977). In contrast, nonmo-
nogamous animals can be more variable in their gregariousness and
aggression. Despite the similarities in the peptidergic regulation of pair bond-
ing behavior, gregarious behavior, and aggression, we should resist the tempta-
tion to think of these behaviors as existing on a single continuum (Ophir,
2011). Our most common model systems illustrate these nuances—Zebra
finches provide an example of a highly social (i.e., gregarious and nonterrito-
rial) and highly affiliative (i.e., pair bonding and parental) species with low lev-
els of aggression towards all conspecifics (Zann, 1996). In contrast, territorial,
pair bonding, and parental species such as violet-eared waxbills and prairie
voles are not antisocial as much as they are selectively social, and they display
high levels of selective aggression (Getz et al., 2003; Getz et al., 1993; Insel
et al., 1995; Jacquot & Solomon, 2004; Thomas & Birney, 1979). Meadow
and montane voles do not form pair bonds or display paternal care; they show
seasonal variation in social grouping, are nonterritorial, and display low levels
of aggression (Beery et al., 2009; Beery & Zucker, 2010; Berteaux et al.,
1999; Boonstra et al., 1993; Ferkin, 1999; Madison, 1980; Madison et al.,
1984; Vlautin et al., 2010). Golden hamsters also do not form pair bonds or
display paternal care, but they are solitary, territorial, and aggressive. Within
these variable social frameworks there exist changes in categorization of indi-
viduals as "out-group" or "in-group." Interestingly, nonapeptides are impli-
cated in directing affiliative behaviors towards in-group individuals and
aggressive behaviors towards out-group individuals. In support of this, OT
promotes cooperation with in-group but not out-group members in humans
(De Dreu, 2012).
The one-on-one interactions described above and other mechanisms may
lead to an alteration in spacing behavior, a critical component contributing
to the mating system(s) and social behaviors expressed by a species. Two in-
triguing studies found that variation in expression of V1aR and OTR was
HORMONES AND ANIMAL BEHAVIOR 247

associated with sexual fidelity and space use in male prairie voles (Ophir et al.,
2008; Ophir et al., 2012). V1aR expression in cortical brain areas associated
with spatial memory (posterior cingulate and laterodorsal thalamus) and
OTR expression in the NAc was greater in males that were better at excluding
males from their territories. In contrast, lower levels of V1aR and OTR were
expressed in males displaying less residency behavior that had either less exclu-
sive territories (and were more likely to be cuckolded) or that expressed “wan-
dering” behavior to find mates. There were no associations in brain areas
linked with pair-bond behavior in the laboratory such as reward-related brain
areas (e.g., VP and LS). As we see our understanding of the functions and
regulation of nonapeptides extended, it is likely that the well-established roles
of AVP/AVT and OT in pair-bonding and reward-related brain areas will
remain important contributors to neurohormonal regulation of social behav-
ior, but the added complexity of the natural environment reveals other layers
of interaction that may not be uncovered under laboratory conditions.

SUMMARY
We have described how hormones impact numerous aspects of social behav-
ior from less social creatures to the increasing layers of social behavior that occur
with larger stable social groups. These neurochemicals and hormones are provid-
ing communication between cells in different parts of the body and different
parts of the brain. Each brain area has different cell types characterized in part
by what receptors they express. Moreover, expression of receptors can be very
plastic and change within a tissue. We generally focus on one behavior, but as
described above, it is more complex than this because no single behavior is occur-
ring in the absence of other behaviors, some of which increase and some of which
decrease based on social and physical contexts. The pattern of communication
between cells is altered and coordinated with different groups of behaviors often
changing depending on season, courtship behavior, breeding, parental behavior,
territoriality and aggression, and the complex social interactions that occur in
social groups. Some of these can be reduced to the simple components of
approach or withdrawal, but there are clearly added layers of complexity.
Hormones and behavior is a very dynamic area of research. It can extend
into more field-related studies, a critical tool for conservation that can incor-
porate more ecological concepts such as effects of climate change on behavior.
It can extend to a more complex understanding of the relationship between
hormones and behavior, a critical tool for many of the hormonal manipula-
tions in our environment that are caused by the estrogenic-like compounds
248 ANIMAL BEHAVIOR

that we are introducing into our environment that are influencing both
humans and animals.
Our understanding of animal behavior can extend insights into many human
behavioral disorders. The tools for understanding disorders involving social inter-
actions are often significantly influenced by hormones, and animal behavior is
used to search for answers to many of these questions. At a molecular level, there
is an opportunity to explore how a huge variety of epigenetic mechanisms (see
Chapter 8) can influence sexual behavior, parental behavior, competitive and
aggressive interactions, and affiliative behaviors. The interactions will
undoubtedly be bidirectional, and resulting behavioral tendencies will express
varying degrees of stability. Finally, there is a beauty in understanding how hor-
mones are integrating and coordinating numerous aspects of the internal and
external environment that are linked in ways that influence the evolution of
behavior. Every time behaviors are tweaked through hormones there are inten-
tional and unintentional consequences that need to be explored (such as evolu-
tionary costs of behavioral levels). In these areas and more, behavioral
endocrinology will be a fascinating and fruitful area of exploration.

ACKNOWLEDGMENTS
Support was provided by a grant from the National Science Foundation
(NSF IOS 1132419).

REFERENCES AND SUGGESTED READING

Adkins-Regan, E. (2005). Hormones and Animal Social Behavior. Princeton, NJ:
Princeton University Press.
Albers, H. E. (2012). The regulation of social recognition, social communication and
aggression: Vasopressin in the social behavior neural network. Hormones and
Behavior, 61, 283–292.
Alvares, G. A., I. B. Hickie, & A. J. Guastella. (2010). Acute effects of intranasal oxy-
tocin on subjective and behavioral responses to social rejection. Experimental
and Clinical Psychopharmacology, 18, 316–321.
Apfelbeck, B. & W. Goymann. (2011). Ignoring the challenge? Male black redstarts
(Phoenicurus ochruros) do not increase testosterone levels during territorial con-
flicts but they do so in response to gonadotropin-releasing hormone.
Proceedings of the Royal Society of London, B, 278, 3233–3242.
Archawaranon, M., L. Dove, & R. H. Wiley. (1991). Social inertia and hormonal
control of aggression and dominance in white-throated sparrows. Behaviour,
118, 42–65.
Barske, J., B. A. Schlinger, M. Wikelski, & L. Fusani. (2011). Female choice for male
motor skills. Proceedings of the Royal Society of London, B, 278, 3523–3528.
HORMONES AND ANIMAL BEHAVIOR 249

Becker, E. A., B. M. Moore, C. Auger, & C. A. Marler, (2010). Paternal behavior

increases testosterone levels in offspring of the California mouse. Hormones
and Behavior, 58, 385–389.
Beery, A. K., D. M. Routman, & I. Zucker. (2009). Same-sex social behavior in
meadow voles: Multiple and rapid formation of attachments. Physiology and
Behavior, 97, 52–57.
Beery, A. K. & I. Zucker. (2010). Oxytocin and same-sex social behavior in female
meadow voles. Neuroscience, 169, 665–673.
Berteaux, D., J. Bety, E. Rengifo, & J. M. Bergeron. (1999). Multiple paternity in
meadow voles (Microtus pennsylvanicus): Investigating the role of the female.
Behavioral Ecology and Sociobiology, 45, 283–291.
Bester-Meredith, J. K. & C. A. Marler. (2003). The association between male off-
spring aggression and paternal and maternal behavior of Peromyscus mice.
Ethology, 109, 797–808.
Bester-Meredith, J. K., P. A. Martin, & C. A. Marler. (2005). Manipulations of vaso-
pressin alter aggression differently across testing conditions in monogamous
and nonmonogamous Peromyscus mice. Aggressive Behavior, 31, 189–199.
Bester-Meredith, J. K., L. J. Young, & C. A. Marler. (1999). Species differences in
paternal behavior and aggression in Peromyscus and their associations with vas-
opressin immunoreactivity and receptors. Hormones and Behavior, 36, 25–38.
Beyer, C., G. Morali, F. Naftolin, K. Larsson, & G. Perezpalacios. (1976). Effect of
some antiestrogens and aromatase inhibitors on androgen induced sexual
behavior in castrated male rats. Hormones and Behavior, 7, 353–363.
Boonstra, R., X. H. Xia, & L. Pavone. (1993). Mating system of the meadow vole,
Microtus pennsylvanicus. Behavioral Ecology, 4, 83–89.
Caldwell, G. S., S. E. Glickman, & E. R. Smith. (1984). Seasonal aggression indepen-
dent of seasonal testosterone in wood rats. Proceedings of the National Academy
of Sciences, USA, 81, 5255–5257.
Caldwell, H. K. & H. E. Albers. (2004). Effect of photoperiod on vasopressin-
induced aggression in Syrian hamsters. Hormones and Behavior, 46, 444–449.
Camazine, B., W. Garstka, R. Tokarz, & D. Crews. (1980). Effects of castration and
androgen replacement on male courtship behavior in the red-sided garter snake
(Thamnophis sirtalis perietalis). Hormones and Behavior, 14, 358–372.
Carere, C., T. G. G. Groothuis, E. Mostl, S. Daan, & J. M. Koolhaas. (2003). Fecal
corticosteroids in a territorial bird selected for different personalities: daily
rhythm and the response to social stress. Hormones and Behavior, 43, 540–548.
Casto, J. M., V. Nolan Jr., & E. D. Ketterson. (2001). Steroid hormones and
immune function: Experimental studies in wild and captive dark-eyed juncos
(Junco hyemalis). American Naturalist, 157, 408–420.
Cho, M. M., A. C. DeVries, J. R. Williams, & C. S. Carter. (1999). The effects of
oxytocin and vasopressin on partner preferences in male and female prairie
voles (Microtus ochrogaster). Behavioral Neuroscience, 113, 1071–1079.
Clutton-Brock, T. H. (1991). The Evolution of Parental Care. Princeton, NJ: Prince-
ton University Press.
250 ANIMAL BEHAVIOR

Coddington, E. & F. L. Moore. (2003). Neuroendocrinology of context-dependent

stress responses: Vasotocin alters the effect of corticosterone on amphibian
behaviors. Hormones and Behavior, 43, 222–228.
Crews, D., B. Camazine, M. Diamond, R. Mason, R. R. Tokarz, & W. R. Garstka.
(1984). Hormonal independence of courtship behavior in the male garter
snake. Hormones and Behavior, 18, 29–41.
Crews, D. & M. C. Moore. (1986). Evolution of mechanisms controlling mating
behavior. Science, 231, 121–125.
Cushing, B. S. & C. S. Carter. (2000). Peripheral pulses of oxytocin increase partner
preferences in female, but not male, prairie voles. Hormones and Behavior, 37,
49–56.
Cynx, J., N. J. Bean, & I. Rossman. (2005). Testosterone implants alter the frequency
range of zebra finch songs. Hormones and Behavior, 47, 446–451.
Davidson, J. M. (1966a). Activation of male rats sexual behavior by intracerebral
implantation of androgen. Endocrinology, 79, 783–794.
Davidson, J. M. (1966b). Characteristics of sex behavior in male rats following castra-
tion. Animal Behaviour, 14, 266–272.
Davidson, J. M. (1969). Effects of estrogen on sexual behavior of male rats. Endocri-
nology, 84, 1365–1372.
Davis, E. S. & C. A. Marler. (2003). The progesterone challenge: Steroid hormone
changes following a simulated territorial intrusion in female Peromyscus califor-
nicus. Hormones and Behavior, 44, 185–198.
Day, L. B., L. Fusani, E. Hernandez, T. J. Billo, K. S. Sheldon, P. M. Wise, et al.
(2007). Testosterone and its effects on courtship in golden-collared manakins
(Manacus vitellinus): Seasonal, sex, and age differences. Hormones and Behav-
ior, 51, 69–76.
De Dreu, C. K. W. (2012). Oxytocin modulates cooperation within and competition
between groups: An integrative review and research agenda. Hormones and
Behavior, 61, 419–428.
De Dreu, C. K. W., L. L. Greer, G. A Van Kleef, S. Shalvi, & M. J. J. Handgraaf.
(2011). Oxytocin promotes human ethnocentrism. Proceedings of the National
Academy of Sciences, USA, 108, 1262–1266.
Deviche, P. & F. L. Moore. (1988). Steroidal control of sexual behavior in the
rough-skinned newt (Taricha granulosa): Effects of testosterone, estradiol,
and dihydrotestosterone. Hormones and Behavior, 22, 26–34.
Duckworth, R. A., M. T. Mendonca, & G. E. Hill. (2001). A condition dependent
link between testosterone and disease resistance in the house finch. Proceedings
of the Royal Society of London, B, 268, 2467–2472.
DuVal, E. H. (2007). Adaptive advantages of cooperative courtship for subordinate
male lance-tailed manakins. American Naturalist, 169, 423–432.
DuVal, E. H. & W. Goymann. (2010). Hormonal correlates of social status and
courtship display in the cooperatively lekking lance-tailed manakin. Hormones
and Behavior, 59, 44–50.
HORMONES AND ANIMAL BEHAVIOR 251

DuVal, E. H. & B. Kempenaers. (2008). Sexual selection in a lekking bird: The

relative opportunity for selection by female choice and male competition.
Proceedings of the Royal Society of London, B, 275, 1995–2003.
Erskine, M. S. (1989). Solicitation behavior in the estrous female rat: A review.
Hormones and Behavior, 23, 473–503.
Ezenwa, V. O., L. S. Ekernas, & S. Creel. (2012). Unravelling complex associations
between testosterone and parasite infection in the wild. Functional Ecology,
26, 123–133.
Feng, N. Y., A. Katz, L. B. Day, J. Barske, & B. A. Schlinger. (2010). Limb muscles
are androgen targets in an acrobatic tropical bird. Endocrinology, 151,
1042–1049.
Ferkin, M. H. (1999). Meadow voles (Microtus pennsylvanicus, Arvicolidae) over-mark
and adjacent-mark the scent marks of same-sex conspecifics. Ethology, 105,
825–837.
Ferris, C. F., R. H. Melloni, G. Koppel, K. W. Perry, R. W. Fuller, & Y. Delville.
(1997). Vasopressin/serotonin interactions in the anterior hypothalamus con-
trol aggressive behavior in golden hamsters. Journal of Neuroscience, 17,
4331–4340.
Ferris, C. F. & M. Potegal. (1988). Vasopressin receptor blockade in the anterior
hypothalamus suppresses aggression in hamsters. Physiology and Behavior, 44,
235–239.
Folstad, I. & A. J. Karter. (1992). Parasites, bright males, and the immuno-
competence handicap. American Naturalist, 139, 603–622.
Frazier, C. R. M., B. C. Trainor, C. J. Cravens, T. K. Whitney, & C. A. Marler.
(2006). Paternal behavior influences development of aggression and vasopres-
sin expression in male California mouse offspring. Hormones and Behavior,
50, 699–707.
Fusani, L. (2008). Testosterone control of male courtship in birds. Hormones and
Behavior, 54, 227–233.
Fusani, L., L. Beani, & F. Dessifulgheri. (1994). Testosterone affects the acoustic
structure of the male call in the grey partridge (Perdix perdix). Behaviour,
128, 301–310.
Fusani, L., L. B. Day, V. Canoine, D. Reinemann, E. Hernandez, & B. A. Schlinger.
(2007). Androgen and the elaborate courtship behavior of a tropical lekking
bird. Hormones and Behavior, 51, 62–68.
Fuxjager, M. J., J. Foufopoulos, R. Diaz-Uriarte, & C. A. Marler. (2011).
Functionally opposing effects of testosterone on two different types of parasite:
Implications for the immunocompetence handicap hypothesis. Functional
Ecology, 25, 132–138.
Fuxjager, M. J. & C. A. Marler. (2010). How and why the winner effect forms: Influ-
ences of contest environment and species differences. Behavioral Ecology, 21,
37–45.
252 ANIMAL BEHAVIOR

Fuxjager, M. J., G. Mast, E. A. Becker, & C. A. Marler. (2009). The “home advan-
tage” is necessary for a full winner effect and changes in post-encounter testos-
terone. Hormones and Behavior, 56, 214–219.
Fuxjager, M. J., J. L. Montgomery, & C. A. Marler. (2011). Species differences in the
winner effect disappear in response to post-victory testosterone manipulations.
Proceedings of the Royal Society of London, B, 278, 3497–3503.
Fuxjager, M. J., T. O. Oyegbile, & C. A. Marler. (2011). Independent and additive
contributions of postvictory testosterone and social experience to the develop-
ment of the winner effect. Endocrinology, 152, 3422–3429.
Garstka, W. R., B. Camazine, & D. Crews. (1982). Interactions of behavior and
physiology during the annual reproductive cycle of the red-sided garter snake
(Thamnophis sirtalis perietalis). Herpetologica, 38, 104–123.
Getz, L. L., B. McGuire, & C. S. Carter. (2003). Social behavior, reproduction and
demography of the prairie vole, Microtus ochrogaster. Ethology, Ecology and
Evolution, 15, 105–118.
Getz, L. L., B. McGuire, T. Pizzuto, J. E. Hofmann, & B. Frase. (1993). Social
organization of the prairie vole (Microtus ochrogaster). Journal of Mammalogy,
74, 44–58.
Gleason, E. D., M. J. Fuxjager, T. O. Oyegbile, & C. A. Marler. (2009). Testoster-
one release and social context: When it occurs and why. Frontiers in
Neuroendocrinology, 30, 460–469.
Gleason, E. D. & C. A. Marler. (2010). Testosterone response to courtship predicts
future paternal behavior in the California mouse, Peromyscus californicus.
Hormones and Behavior, 57, 147–154.
Goodson, J. L. (2005). The vertebrate social behavior network: Evolutionary themes
and variations. Hormones and Behavior, 48, 11–22.
Goodson, J. L., A. K. Evans, & Y. Wang. (2006). Neuropeptide binding reflects con-
vergent and divergent evolution in species-typical group sizes. Hormones and
Behavior, 50, 223–236.
Goodson, J. L., D. Kabelik, A. M. Kelly, J. Rinaldi, & J. D. Klatt. (2009). Midbrain
dopamine neurons reflect affiliation phenotypes in finches and are tightly
coupled to courtship. Proceedings of the National Academy of Sciences, USA,
106, 8737–8742.
Goodson, J. L., S. E. Schrock, J. D. Klatt, D. Kabelik, & M. A. Kingsbury. (2009).
Mesotocin and nonapeptide receptors promote songbird flocking behavior.
Science, 325, 862–866.
Goodson, J. L. & Y. W. Wang. (2006). Valence-sensitive neurons exhibit divergent
functional profiles in gregarious and asocial species. Proceedings of the National
Academy of Sciences, USA, 103, 17013–17017.
Goymann, W., A. Wittenzellner, I. Schwabl, & M. Makomba. (2008). Progesterone
modulates aggression in sex-role reversed female African black coucals. Proceed-
ings of the Royal Society of London, B, 275, 1053–1060.
HORMONES AND ANIMAL BEHAVIOR 253

Gubernick, D. J. & J. R. Alberts. (1987). The biparental care system of the California
mouse, Peromyscus californicus. Journal of Comparative Psychology, A, 101,
169–177.
Gubernick, D. J. & T. Teferi. (2000). Adaptive significance of male parental care
in a monogamous mammal. Proceedings of the Royal Society of London, B,
267, 147–150.
Guhl, A. M. (1968). Social inertia and social stability in chickens. Animal Behaviour,
16, 219–232.
Hammock, E. A. D. & L. J. Young. (2002). Variation in the vasopressin V1a receptor
promoter and expression: Implications for inter- and intraspecific variation in
social behaviour. European Journal of Neuroscience, 16, 399–402.
Hammock, E. A. D. & L. J. Young. (2004). Functional microsatellite polymorphism
associated with divergent social structure in vole species. Molecular Biology and
Evolution, 21, 1057–1063.
Hammock, E. A. D. & L. J. Young. (2005). Microsatellite instability generates diver-
sity in brain and sociobehavioral traits. Science, 308, 1630–1634.
Hau, M., S. T. Stoddard, & K. K. Soma. (2004). Territorial aggression and hormones
during the non-breeding season in a tropical bird. Hormones and Behavior, 45,
40–49.
Hirschenhauser, K. & R. F. Oliveira. (2006). Social modulation of androgens in male
vertebrates: Meta-analyses of the challenge hypothesis. Animal Behaviour, 71,
265–277.
Hirschenhauser, K., H. Winkler, & R. F. Oliveira. (2003). Comparative analysis of
male androgen responsiveness to social environment in birds: The effects of
mating system and paternal incubation. Hormones and Behavior, 43, 508–519.
Ho, J. M., J. H. Murray, G. E. Demas, & J. L. Goodson. (2010). Vasopressin cell
groups exhibit strongly divergent responses to copulation and male-male inter-
actions in mice. Hormones and Behavior, 58, 368–377.
Hutchison, J. B. (1970). Differential effects of testosterone and oestradiol on male
courtship in barbary doves (Streptopelia risoria). Animal Behaviour, 18, 41–51.
Hutchison, J. B. & T. Steimer. (1983). Hormone-mediated behavioral transitions: a
role for brain aromatase. In J. Balthazart, E. Prove, & R. Gilles (eds.),
Hormones and Behaviour in Higher Vertebrates (pp. 261–274). Berlin: Springer
Verlag.
Insel, T. R. & T. J. Hulihan. (1995). A gender-specific mechanism for pair bonding:
Oxytocin and partner preference formation in monogamous voles. Behavioral
Neuroscience, 109, 782–789.
Insel, T. R., S. Preston, & J. T. Winslow. (1995). Mating in the monogamous male:
Behavioral consequences. Physiology & Behavior, 57, 615–627.
Insel, T. R. & L. E. Shapiro. (1992). Oxytocin receptor distribution reflects social
organization in monogamous and polygamous voles. Proceedings of the
National Academy of Sciences, USA, 89, 5981–5985.
254 ANIMAL BEHAVIOR

Insel, T. R., Z. X. Wang, & C. F. Ferris. (1994). Patterns of brain vasopressin

receptor distribution associated with social organization in microtine rodents.
Journal of Neuroscience, 14, 5381–5392.
Jacquot, J. J. & N. G. Solomon. (2004). Experimental manipulation of territory
occupancy: Effects on immigration of female prairie voles. Journal of Mammal-
ogy, 85, 1009–1014.
Kelly, A. M., M. A. Kingsbury, K. Hoffbuh, S. E. Schrock, B. Waxman, D. Kabelik,
et al. (2011). Vasotocin neurons and septal V-1a-like receptors potently modu-
late songbird flocking and responses to novelty. Hormones and Behavior, 60,
12–21.
Kendrick, K. M. & A. F. Dixson. (1984). Ovariectomy does not abolish proceptive
behavior cyclicity in the common marmoset (Callithrix jacchus). Journal of
Endocrinology, 101, 155–162.
Ketterson, E. D. & V. Nolan Jr. (1992). Hormones and life histories: An integrative
approach. American Naturalist, 140, S33–S62.
Ketterson, E. D., V. Nolan Jr., L. Wolf, & C. Ziegenfus. (1992). Testosterone and
avian life histories: Effects of experimentally elevated testosterone on behavior
and correlates of fitness in the dark-eyed junco (Junco hyemalis). American
Naturalist, 140, 980–999.
Keverne, E. B. & K. M. Kendrick. (1994). Maternal behavior in sheep and its
neuroendocrine regulation. Acta Paediatrica, 83, 47–56.
Kleiman, D. G. (1977). Monogamy in mammals. Quarterly Review of Biology, 52,
39–69.
Knapp, R. & M. C. Moore. (1995). Hormonal responses to aggression vary in
different types of agonistic encounters in male tree lizards, Urosaurus ornatus.
Hormones and Behavior, 29, 85–105.
Kosfeld, M., M. Heinrichs, P. J. Zak, U. Fischbacher, & E. Fehr. (2005). Oxytocin
increases trust in humans. Nature, 435, 673–676.
Krehbiel, D., P. Poindron, F. Levy, & M. J. Prudhomme. (1987). Peridural
anesthesia disturbs maternal behavior in primiparous and multiparous parturi-
ent ewes. Physiology and Behavior, 40, 463–472.
Levin, R. N. & J. C. Wingfield. (1992). The hormonal control of territorial aggres-
sion in tropical birds. Ornis Scandinavica, 23, 284–291.
Levy, F., E. B. Keverne, K. M. Kendrick, V. Piketty, & P. Poindron. (1992).
Intracerebral oxytocin is important for the onset of maternal behavior in
inexperienced ewes delivered under peridural anesthesia. Behavioral Neurosci-
ence, 106, 427–432.
Liker, A. & T. Szekely, T. (1997). Aggression among female lapwings, Vanellus vanel-
lus. Animal Behaviour, 54, 797–802.
Lim, M. M., Z. X. Wang, D. E. Olazabal, X. H. Ren, E. F. Terwilliger, & L. J.
Young. (2004). Enhanced partner preference in a promiscuous species by
manipulating the expression of a single gene. Nature, 429, 754–757.
Lim, M. M. & L. J. Young. (2004). Vasopressin-dependent neural circuits underlying
pair bond formation in the monogamous prairie vole. Neuroscience, 125, 35–45.
HORMONES AND ANIMAL BEHAVIOR 255

Lincoln, G. A., F. Guinness, & R. V. Short. (1972). The way in which testosterone
controls the social and sexual behavior of the red deer stag (Cervus elaphus).
Hormones and Behavior, 3, 375–396.
Lindstrom, K. M., D. Krakower, J. O. Lundstrom, & B. Silverin. (2001). The effects
of testosterone on a viral infection in greenfinches (Carduelis chloris): An exper-
imental test of the immunocompetence-handicap hypothesis. Proceedings of the
Royal Society of London, B, 268, 207–211.
Lisk, R. D., L. A. Ciaccio, & C. Catanzaro. (1983). Mating behavior of the golden
hamster under semi-natural conditions. Animal Behaviour, 31, 659–666.
Lisk, R. D. & L. A. Reuter. (1980). Relative contributions of estradiol and progester-
one to the maintenance of sexual receptivity in mated female hamsters. Journal
of Endocrinology, 87, 175–183.
Liu, Y., J. T. Curtis, & Z. X. Wang. (2001). Vasopressin in the lateral septum regu-
lates pair bond formation in male prairie voles (Microtus ochrogaster). Behav-
ioral Neuroscience, 115, 910–919.
Liu, Y. & Z. X. Wang. (2003). Nucleus accumbens oxytocin and dopamine interact
to regulate pair bond formation in female prairie voles. Neuroscience, 121,
537–544.
Lu, S., S. E. McKenna, A. Cologer-Clifford, E. A. Nau, & N. G. Simon. (1998).
Androgen receptor in mouse brain: Sex differences and similarities in autoregu-
lation. Endocrinology, 139, 1594–1601.
Lu, S., N. G. Simon, Y. Wang, & S. Hu. (1999). Neural androgen receptor
regulation: Effects of androgen and antiandrogen. Journal of Neurobiology,
41, 505–512.
Lynn, S. E. (2008). Behavioral insensitivity to testosterone: Why and how does testos-
terone alter paternal and aggressive behavior in some avian species but not
others? General and Comparative Endocrinology, 157, 233–240.
Madison, D. M. (1980). Space use and social structure in meadow voles, Microtus
pennsylvanicus. Behavioral Ecology and Sociobiology, 7, 65–71.
Madison, D. M., R. W. Fitzgerald, & W. J. Mcshea. (1984). Dynamics of social nest-
ing in overwintering meadow voles (Microtus pennsylvanicus): Possible conse-
quences for population cycling. Behavioral Ecology and Sociobiology, 15, 9–17.
Mangiamele, L. A. & R. R. Thompson. (2012). Testosterone rapidly increases
ejaculate volume and sperm density in competitively breeding goldfish through
an estrogenic membrane receptor mechanism. Hormones and Behavior, 62,
107–112.
Marler, C. A., J. K. Bester-Meredith, & B. C. Trainor. (2003). Paternal behavior and
aggression: Endocrine mechanisms and nongenomic transmission of behavior.
Advances in the Study of Behavior, 32, 263–323.
Marler, C. A. & M. C. Moore. (1988). Evolutionary costs of aggression revealed by
testosterone manipulations in free-living male lizards. Behavioral Ecology and
Sociobiology, 23, 21–26.
256 ANIMAL BEHAVIOR

Marler, C. A. & M. C. Moore. (1991). Supplementary feeding compensates for

testosterone-induced costs of aggression in male mountain spiny lizards, Scelo-
porus jarrovi. Animal Behaviour, 42, 209–219.
Marler, C. A., G. Walsberg, M. L. White, & M. Moore. (1995). Increased energy
expenditure due to increased territorial defense in male lizards after phenotypic
manipulation. Behavioral Ecology and Sociobiology, 37, 225–231.
McCarthy, M. M. & J. B. Becker. (2002). Neuroendocrinology of Sexual Behavior in
the Female. Cambridge, MA: MIT Press.
Mendonca, M. T., S. D. Chernetsky, K. E. Nester, & G. L. Gardner. (1996). Effects
of gonadal sex steroids on sexual behavior in the big brown bat, Eptesicus fuscus,
upon arousal from hibernation. Hormones and Behavior, 30, 153–161.
Moltz, H., M. Lubin, M. Leon, & M. Numan. (1970). Hormonal induction of
maternal behavior in ovariectomized nulliparous rats. Physiology and Behavior,
5, 1373–1377.
Moore, F. L. & L. J. Miller. (1984). Stress-induced inhibition of sexual behavior:
Corticosterone inhibits courtship behaviors of a male amphibian (Taricha
granulosa). Hormones and Behavior, 18, 400–410.
Moore, F. L. & M. Orchinik. (1994). Membrane receptors for corticosterone: A
mechanism for rapid behavioral responses in an amphibian. Hormones and
Behavior, 28, 512–519.
Moore, M. C. (1987). Circulating steroid hormones during rapid aggressive responses
of territorial male mountain spiny lizards, Sceloporus jarrovi. Hormones and
Behavior, 21, 511–521.
Moore, M. C., D. K. Hews, & R. Knapp. (1998). Hormonal control and evolution of
alternative male phenotypes: Generalizations of models for sexual differentia-
tion. American Zoologist, 38, 133–151.
Nelson, R. J. (2000). An Introduction to Behavioral Endocrinology. Second Edition.
Sunderland, MA: Sinauer Associates.
Newman, S. W. (1999). The medial extended amygdala in male reproductive behav-
ior: A node in the mammalian social behavior network. Annals of the New York
Academy of Sciences, 1999, 242–257.
Normandin, J. J. & A. Z. Murphy. (2011). Somatic genital reflexes in rats with a nod
to humans: Anatomy, physiology, and the role of the social neuropeptides.
Hormones and Behavior, 59, 656–665.
Nowak, R., M. Keller, & F. Levy. (2011). Mother-young relationships in sheep: A
model for a multidisciplinary approach of the study of attachment in mam-
mals. Journal of Neuroendocrinology, 23, 1042–1053.
O’Connell, L. A. & H. A. Hofmann. (2011a). Genes, hormones, and circuits: An
integrative approach to study the evolution of social behavior. Frontiers in
Neuroendocrinology, 32, 320–335.
O’Connell, L. A. & H. A. Hofmann. (2011b). The vertebrate mesolimbic reward sys-
tem and social behavior network: A comparative synthesis. Journal of Compara-
tive Neurology, 519, 3599–3639.
HORMONES AND ANIMAL BEHAVIOR 257

O’Connell, L. A. & H. A. Hofmann. (2012). Evolution of a vertebrate social

decision-making network. Science, 336, 1154–1157.
Ogawa, S., D. B. Lubahn, K. S. Korach, & D. W. Pfaff. (1997). Behavioral effects of
estrogen receptor gene disruption in male mice. Proceedings of the National
Academy of Sciences, USA, 94, 1476–1481.
Ogawa, S., T. F. Washburn, J. Taylor, D. B. Lubahn, K. S. Korach, & D. W. Pfaff.
(1998). Modifications of testosterone-dependent behaviors by estrogen recep-
tor alpha gene disruption in male mice. Endocrinology, 139, 5058–5069.
Ophir, A. G. (2011). Towards meeting Tinbergen’s challenge. Hormones and Behav-
ior, 60, 22–27.
Ophir, A. G., O. L. Crino, Q. C. Wilkerson, J. O. Wolff, & S. M. Phelps, S. (2008).
Female-directed aggression predicts paternal behavior, but female prairie voles
prefer affiliative males to paternal males. Brain Behavior and Evolution, 71, 32–40.
Ophir, A. G., A. Gessel, D. J. Zheng, & S. M. Phelps. (2012). Oxytocin receptor
density is associated with male mating tactics and social monogamy. Hormones
and Behavior, 61, 445–453.
Pinxten, R., E. De Ridder, J. Balthazart, & M. Eens. (2002). Context-dependent
effects of castration and testosterone treatment on song in male European star-
lings. Hormones and Behavior, 42, 307–318.
Pitkow, L. J., C. A. Sharer, X. L. Ren, T. R. Insel, E. F. Terwilliger, & L. J. Young.
(2001). Facilitation of affiliation and pair-bond formation by vasopressin
receptor gene transfer into the ventral forebrain of a monogamous vole. Journal
of Neuroscience, 21, 7392–7396.
Potegal, M. & C. F. Ferris. (1989). Intraspecific aggression in male hamsters is inhib-
ited by intrahypothalamic vasopressin receptor antagonist. Aggressive Behavior,
15, 311–320.
Powers, J. B. (1970). Hormonal control of sexual receptivity during the estrous cycle
of the rat. Physiology and Behavior, 5, 831–835.
Pradhan, D. S., A. E. M. Newman, D. W. Wacker, J. C. Wingfield, B. A. Schlinger,
& K. K. Soma. (2010). Aggressive interactions rapidly increase androgen syn-
thesis in the brain during the non-breeding season. Hormones and Behavior,
57, 381–389.
Prudom, S. L., C. A. Broz, N. Schultz-Darken, C. T. Ferris, C. Snowdon, & T. E.
Ziegler. (2008). Exposure to infant scent lowers serum testosterone in father
common marmosets (Callithrix jacchus). Biology Letters, 4, 603–605.
Remage-Healey, L. & A. H. Bass. (2004). Rapid, hierarchical modulation of vocal
patterning by steroid hormones. Journal of Neuroscience, 24, 5892–5900.
Remage-Healey, L. & A. H. Bass. (2006a). From social behavior to neural circuitry:
Steroid hormones rapidly modulate advertisement calling via a vocal pattern
generator. Hormones and Behavior, 50, 432–441.
Remage-Healey, L. & A. H. Bass. (2006b). A rapid neuromodulatory role for teroid
hormones in the control of reproductive behavior. Brain Research, 1126, 27–35.
258 ANIMAL BEHAVIOR

Ribble, D. O. (1991). The monogamous mating system of Peromyscus californicus as

revealed by DNA fingerprinting. Behavioral Ecology and Sociobiology, 29,
161–166.
Riters, L. V., M. Eens, R. Pinxten, D. L. Duffy, J. Balthazart, & G. F. Ball. (2000).
Seasonal changes in courtship song and the medial preoptic area in male
European starlings (Sturnus vulgaris). Hormones and Behavior, 38, 250–261.
Ritschard, M., S. Laucht, J. Dale, & H. Brumm. (2011). Enhanced testosterone levels
affect singing motivation but not song structure and amplitude in Bengalese
finches. Physiology and Behavior, 102, 30–35.
Roberts, M. & A. Peters. (2009). Is testosterone immunosuppressive in a condition-
dependent manner? An experimental test in blue tits. Journal of Experimental
Biology, 212, 1811–1818.
Rodgers, E. W., R. L. Earley, & M. S. Grober. (2006). Elevated 11-ketotestosterone
during paternal behavior in the bluebanded goby (Lythrypnus dalli). Hormones
and Behavior, 49, 610–614.
Ross, H. E., C. D. Cole, Y. Smith, I. D. Neumann, R. Landgraf, A. Z. Murphy, et al.
(2009). Characterization of the oxytocin system regulating affiliative behavior
in female prairie voles. Neuroscience, 162, 892–903.
Ruiz, M., S. S. French, G. E. Demas, & E. P. Martins. (2010). Food supplementation
and testosterone interact to influence reproductive behavior and immune func-
tion in Sceloporus graciosus. Hormones and Behavior, 57, 134–139.
Sandell, M. I. (1998). Female aggression and the maintenance of monogamy: Female
behaviour predicts male mating status in European starlings. Proceedings of the
Royal Society of London, B, 265, 1307–1311.
Sandell, M. I. (2007). Exogenous testosterone increases female aggression in the
European starling (Sturnus vulgaris). Behavioral Ecology and Sociobiology, 62,
255–262.
Sandell, M. I. & H. G. Smith. (1996). Already mated females constrain male mating
success in the European starling. Proceedings of the Royal Society of London, B,
263, 743–747.
Schlinger, B. A., C. Creco, & A. H. Bass. (1999). Aromatase activity in the hindbrain
vocal control region of a teleost fish: Divergence among males with alternative
reproductive tactics. Proceedings of the Royal Society of London, B, 266,
131–136.
Schoech, S. J., E. D. Ketterson, V. Nolan Jr., P. J. Sharp, & J. D. Buntin. (1998).
The effect of exogenous testosterone on parental behavior, plasma prolactin,
and prolactin binding sites in dark-eyed juncos. Hormones and Behavior, 34,
1–10.
Scriba, M. F. & W. Goymann. (2010). European robins (Erithacus rubecula) lack an
increase in testosterone during simulated territorial intrusions. Journal of Orni-
thology, 151, 607–614.
Sheehan, T. & M. Numan. (2002). Estrogen, progesterone, and pregnancy termina-
tion alter neural activity in brain regions that control maternal behavior in rats.
Neuroendocrinology, 75, 12–23.
HORMONES AND ANIMAL BEHAVIOR 259

Sisneros, J. A., P. M. Forlano, D. L. Deitcher, & A. H. Bass. (2004). Steroid-

dependent auditory plasticity leads to adaptive coupling of sender and receiver.
Science, 305, 404–407.
Sisneros, J. A., P. M. Forlano, R. Knapp, & A. H. Bass. (2004). Seasonal variation of
steroid hormone levels in an intertidal-nesting fish, the vocal plainfin midship-
man. General and Comparative Endocrinology, 136, 101–116.
Smeltzer, M. D., J. T. Curtis, B. J. Aragona, & Z. X. Wang. (2006). Dopamine, oxy-
tocin, and vasopressin receptor binding in the medial prefrontal cortex of
monogamous and promiscuous voles. Neuroscience Letters, 394, 146–151.
Stellar, J. R. & E. Stellar. (1985). The Neurobiology of Motivation and Reward. New
York: Springer-Verlag.
Theodoridou, A., A. C. Row, I. S. Penton-Voak, & P. J. Rogers. (2009). Oxytocin
and social perception: Oxytocin increases perceived facial trustworthiness and
attractiveness. Hormones and Behavior, 56, 128–132.
Thomas, J. A. & E. C. Birney. (1979). Parental care and mating system of the prairie
vole, Microtus ochrogaster. Behavioral Ecology and Sociobiology, 5, 171–186.
Trainor, B. C., I. M. Bird, & C. A. Marler. (2004). Opposing hormonal mechanisms
of aggression revealed through short-lived testosterone manipulations and
multiple winning experiences. Hormones and Behavior, 45, 115–121.
Trainor, B. C., S. Lin, M. S. Finy, M. R. Rowland, & R. J. Nelson. (2007). Photo-
period reverses the effects of estrogens on male aggression via genomic and
nongenomic pathways. Proceedings of the National Academy of Sciences, USA,
104, 9840–9845.
Trainor, B. C. & C. A. Marler. (2002). Testosterone promotes paternal behaviour in
a monogamous mammal via conversion to oestrogen. Proceedings of the Royal
Society of London, B, 269, 823–829.
Turgeon, J. L., S. M. Van Patten, G. Shyamala, & D. W. Waring. (1999). Steroid
regulation of progesterone receptor expression in cultured rat gonadotropes.
Endocrinology, 140, 2318–2325.
Turgeon, J. L. & D. W. Waring. (2000). Progesterone regulation of the progesterone
receptor in rat gonadotropes. Endocrinology, 141, 3422–3429.
Vlautin, C. T., N. J. Hobbs, & M. H. Ferkin. (2010). Male and female meadow
voles, Microtus pennsylvanicus, differ in their responses to heterospecific/con-
specific over-marks. Ethology, 116, 797–805.
Voigt, C. & W. Goymann. (2007). Sex-role reversal is reflected in the brain of African
black coucals (Centropus grillii). Developmental Neurobiology, 67, 1560–1573.
Wacker, D. W., A. J. Coverdill, C. M. Bauer, & J. C. Wingfield. (2010). Male
territorial aggression and androgen modulation in high latitude populations
of the sooty, Passerella iliaca sinuosa, and red fox sparrow, Passerella iliaca zabo-
ria. Journal of Ornithology, 151, 79–86.
Wang, Z. X., W. Smith, D. E. Major, & G. J. Devries. (1994). Sex and species-
differences in the effects of cohabitation on vasopressin messenger-RNA
expression in the bed nucleus of the stria terminalis in prairie voles (Microtus
260 ANIMAL BEHAVIOR

ochrogaster) and meadow voles (Microtus pennsylvanicus). Brain Research, 650,

212–218.
Whalen, R. E. (1974). Estrogen-progesterone induction of mating in female rats.
Hormones and Behavior, 5, 157–162.
Wikelski, M., S. S. Steiger, B. Gall, & K. N. Nelson. (2005). Sex, drugs and mating
role: Testosterone-induced phenotype-switching in Galapagos marine iguanas.
Behavioral Ecology, 16, 260–268.
Wiley, R. H., L. Steadman, L. Chadwick, & L. Wollerman. (1999). Social inertia in
white-throated sparrows results from recognition of opponents. Animal Behav-
iour, 57, 453–463.
Williams, J. R., T. R. Insel, C. R. Harbaugh, & C. S. Carter. (1994). Oxytocin
administered centrally facilitates formation of a partner preference in female
prairie voles (Microtus ochrogaster). Journal of Neuroendocrinology, 6, 247–250.
Wingfield, J. C., G. F. Ball, A. M. Dufty, R. E. Hegner, & M. Ramenofsky. (1987).
Testosterone and aggression in birds. American Scientist, 75, 602–608.
Wingfield, J. C., R. E. Hegner, A. M. Dufty, & G. F. Ball. (1990). The “challenge
hypothesis”: Theoretical implications for patterns of testosterone secretion,
mating systems, and breeding strategies. American Naturalist, 136, 829–846.
Wingfield, J. C., D. L. Maney, C. W. Breuner, J. D. Jacobs, S. Lynn, M. Ramenof-
sky, et al. (1998). Ecological bases of hormone-behavior interactions: The
“emergency life history stage.” American Zoologist, 38, 191–206.
Wingfield, J. C., A. L. Newman, G. J. Hunt, & D. S. Farner. (1982). Endocrine
aspects of the female-female pairing in the western gull (Larus occidentalis
wymani). Animal Behaviour, 30, 9–22.
Wingfield, J. C. & R. M. Sapolsky. (2003). Reproduction and resistance to stress:
When and how. Journal of Neuroendocrinology, 15, 711–724.
Winslow, J. T., N. Hastings, C. S. Carter, C. R. Harbaugh, & T. R. Insel. (1993).
A role for central vasopressin in pair bonding in monogamous prairie voles.
Nature, 365, 545–548.
Winslow, J. T., L. Shapiro, C. S. Carter, & T. R. Insel. (1993). Oxytocin and com-
plex social behavior: Species comparisons. Psychopharmacology Bulletin, 29,
409–414.
Woolley, S. C., J. T. Sakata, & D. Crews. (2004). Evolutionary insights into the regu-
lation of courtship behavior in male amphibians and reptiles. Physiology and
Behavior, 83, 347–360.
Xie, J., W. J. Kuenzel, P. J. Sharp, & A. Jurkevich. (2011). Appetitive and consum-
matory sexual and agonistic behaviour elicits FOS expression in aromatase
and vasotocin neurons within the preoptic area and bed nucleus of the stria
terminalis of male domestic chickens. Journal of Neuroendocrinology, 23,
232–243.
Young, K. A., K. L. Gobrogge, Y. Liu, & Z. X. Wang. (2011). The neurobiology of
pair bonding: Insights from a socially monogamous rodent. Frontiers in
Neuroendocrinology, 32, 53–69.
HORMONES AND ANIMAL BEHAVIOR 261

Young, L. J. (1999). Oxytocin and vasopressin receptors and species-typical social

behaviors. Hormones and Behavior, 36, 212–221.
Young, L. J., B. Huot, R. Nilsen, Z. X. Wang, & T. R. Insel. (1996). Species differ-
ences in central oxytocin receptor gene expression: comparative analysis of pro-
moter sequences. Journal of Neuroendocrinology, 8, 777–783.
Young, L. J., M. M. Lim, B. Gingrich, & T. R. Insel. (2001). Cellular mechanisms of
social attachment. Hormones and Behavior, 40, 133–138.
Young, L. J., R. Nilsen, K. G. Waymire, G. R. MacGregor, & T. R. Insel. (1999).
Increased affiliative response to vasopressin in mice expressing the V-1a recep-
tor from a monogamous vole. Nature, 400, 766–768.
Young, L. J. & Z. Wang. (2004). The neurobiology of pair bonding. Nature Reviews
Neuroscience, 7, 1048–1054.
Young, L. J., K. G. Waymire, R. Nilsen, G. R. MacGregor, Z. Wang, & T. R. Insel.
(1997). The 5’ flanking region of the monogamous prairie vole oxytocin recep-
tor gene directs tissue-specific expression in transgenic mice. Integrative Neuro-
biology of Affiliation, 807, 514–517.
Young, L. J., J. T. Winslow, R. Nilsen, & T. R. Insel. (1997). Species differences in
V(1)a receptor gene expression in monogamous and nonmonogamous voles:
Behavioral consequences. Behavioral Neuroscience, 111, 599–605.
Young, L. J., A. Z. M. Young, & E. A. D. Hammock. (2005). Anatomy and neuro-
chemistry of the pair bond. Journal of Comparative Neurology, 493, 51–57.
Zann, R. (1996). The Zebra Finch: A Synthesis of Field and Laboratory Studies. Oxford,
UK: Oxford University Press.
This page intentionally left blank
 8

Epigenetics and Animal Behavior

David Crews

INTRODUCTION
The tendency to succumb to the seduction of dichotomies in lieu of grappling
with the reality of complexity is human. This is best seen in the nature/nurture
debate that dates at least to the ancient Greeks and is reprised every generation
in science under various guises (e.g., innate versus acquired, instinct versus
learned). And with every generation there is a refutation of this false
dichotomy as being sterile, yielding no useful offspring in our knowledge base
(e.g., Lehrman, 1970; Gottleib, 2002; Bateson & Gluckman, 2011). This
seemingly endless rediscovery is perhaps a good thing as it forces scientists to
develop new ways of investigating and illustrating that nothing in biology
and psychology is simple. In many ways, the rediscovery of epigenetics
refreshes the debate and, perhaps for the first time, offers a solution.
Put simply, epigenetics refers to traits that are not determined by traditional
molecular bases for inheritance. A more precise definition would be that epige-
netic effects are changes in the phenotype and/or specific traits that result from
the environmental modification of the molecular factors and processes around
DNA that regulate genome activity yet are independent of the DNA sequence.
Note that the term environment is inclusive of all stimuli that may impinge on
the organism during its life cycle. How researchers have interpreted epige-
netics reflects its dual origins and the nature of the question being posed. At
a basic level it is useful to differentiate molecular epigenetics and molar epi-
genetics, or bottom-up versus top-down epigenetics. The former perspective
264 ANIMAL BEHAVIOR

has emerged within the last 25 years from modern genetics and molecular
biology and focuses on molecular levels of analysis. The latter perspective has
a deeper history, preceding the rediscovery of Mendel’s studies, and focuses
on questions of evolution and adaptive significance as evident in psychobiol-
ogy and evolutionary biology. Thus, the object of study in molecular epige-
netics is transcriptional and translational control during embryonic
development, while in molar epigenetics it is the individual’s interactions with
its biotic and physical environment through time. More will be said about this
distinction later.
Another distinction that must be made is that between environmentally
induced epigenetic modifications and parental genomic imprinting, forms
of epigenetic marks that are carried in the male and female germlines. Parental
genomic imprints refer to genes that are expressed in a parent-of-origin fash-
ion, that is, from the mother (maternal imprint) or from the father (paternal
imprint). Both environmentally induced epigenetic modifications and paren-
tal genomic imprinting involve methylation and histone modifications, but
genomic imprints are sex specific, although most of these epigenetic control
regions are matrilineal. Regardless of the parental origin, the imprinted gene is
expressed or silenced in the same way in both male and female offspring. Some
imprints are tissue specific but always show monoallelic expression (only one
allele of a gene is actively transcribed). At this stage there is no evidence that
the imprint is individually specific or that each father (or mother) has an individ-
ual “signature,” but this is due more to the species studied to date (e.g., inbred
laboratory rodents) than demonstrated as a general principle. It is known that if
strains are crossed the imprint changes, but as yet there is no information on nat-
urally occurring species and whether it is possible that the imprint may vary
depending upon the population or perhaps lineage. This is a particularly impor-
tant question in light of the role of sexual selection in the evolution of traits in
outbred (versus inbred) species. For example, if male A mates with two females
(E and G), would the paternal imprint of the offspring of the litters (or singleton) pro-
duced from those matings be different or the same? The converse question would
apply to a female that produces two litters by different males (let us further
assume that each litter has but one father of all the young in the litter rather than
being a result of multiple paternity). Is her maternal imprint identical in the respec-
tive litters? While DNA methylation is clearly involved in genomic imprinting,
the signal for the imprint is not yet known.
This chapter will not deal with genomic imprinting further as there are
excellent reviews relating genomic imprinting to brain and behavior available
(e.g., Keverne & Curley, 2008; Keverne, 2009). Instead, I will focus herein-
after entirely on environmentally induced epigenetic modifications.
EPIGENETICS AND ANIMAL BEHAVIOR 265

Finally, it is important to consider the issue of life stages. Individuals are

particularly sensitive and vulnerable to environmentally induced epigenetic
modifications during early life stages or in the period of transition from one
stage to another. The time of maximal neuronal plasticity is in the earliest
stages of life, beginning before birth and, in mammals, up to weaning.
Although the individual’s capacity to respond to environmental change or
insult with heritable phenotypic variation at a later stage is possible, it is
during this early period that hormones and genotype predispose an individ-
ual’s responses to future experiences throughout the life cycle as well as its sus-
ceptibility to developing disorders (Gilbert & Epel, 2008; Bateson &
Gluckman, 2011). Although most research has focused on the earliest life
stages (fetus and neonate), another period of extreme vulnerability is the
period surrounding adrenarche (the increase in activity of the adrenal glands
just before puberty) and pubarche (the onset of puberty). It is during adoles-
cence that the body (including the brain) is reshaped by hormones and the
individual graduates from dependence to independence, assuming the proper-
ties of maturity. Stressors experienced during this period also have enduring
effects, including neural remodeling, impaired learning and memory, and
altered emotional behaviors in adulthood.

MOLECULAR VERSUS MOLAR EPIGENETICS

Investigators in the field of epigenetics come from one of two distinct line-
ages. This split history is similar to the origins of the modern study of animal
behavior where European ethologists and American comparative psychologists
differed in their approach to behavior, both in perspective and substance (see
Chapter 2). Both molecular epigenetics and molar epigenetics share a
common history, namely the sixteenth- and seventeenth-century debates of
preformationism versus epigenesis. The central question then was, and con-
tinues to be, how a fully integrated multicellular organism develops from a sin-
gle cell (the fertilized egg). Preformationists believed that adult features were
present fully formed in the egg and simply unfolded during growth; August
Weissman belonged to this group and asserted that the eggs contained all of
the elements (later known as genes) to determine the phenotype that would
develop. Those believing in epigenesis held that traits emerge as a consequence
of the progressive interaction of the constituent parts of the zygote with the
environment in which it develops. Although others such as Charles Darwin
and Jean-Baptiste Lamarck were believers in epigenesis, the pivotal role of
the environment in the developmental process was first demonstrated empiri-
cally by Oscar Hertwig (1894) and subsequently by Richard Woltereck
266 ANIMAL BEHAVIOR

(1909), whose early work on Daphnia, an organism that can reproduce asex-
ually by cloning, demonstrated that genetically identical individuals would
develop very different morphs depending upon their environment.

Molecular Epigenetics
Prior to the 1940s, the gene as the unit of heritable material was a theoreti-
cal concept without a physical identity. In 1942 Conrad Waddington pro-
posed the term epigenetics as a conceptual model of how genes might interact
with their environment and give rise to the phenotype (Waddington, 1942).
It is in this sense that the term epigenetics is commonly used in molecular
and developmental genetics today, namely, “the study of the mechanisms of
temporal and spatial control of gene activity during the development of com-
plex organisms” (Holliday, 1990, p. 329). This relatively recent area of
research focuses on processes such as DNA methylation (addition of a methyl
group at specific positions on two of the four DNA bases) and histone modi-
fications (changes to the proteins that package DNA) that are heritable in
the short term but do not change the DNA or create mutations. Because arti-
ficial manipulation of methylation patterns is often lethal, or at the least results
in maladaptive traits or monsters, this method of research illuminates normal
development by creating abnormalities or anomalies. It also helps us to under-
stand the processes that occur when environmental factors affect methylation
of DNA during the normal development of organisms.
Waddington continues to dominate the way we think of epigenetics, and
his image of an epigenetic landscape as an emergent process is the defining
concept of how epigenetics operates. His structural depiction envisions how
the environment shapes phenotypic outcomes and also the importance of tim-
ing as well of genes and environment. Although Waddington’s formulation
conveys the idea that development is irreversible and results in discrete out-
comes rather than continuous outcomes, these aspects have now been refuted.
Attendant concepts advocated by Waddington (1942) were canalization
and genetic assimilation. Canalization connotes the differentiation of the
gene, cell type, and embryo as development progresses and pathways becomes
more entrenched, thereby making it harder for the canalized development to
be dislodged and moved into another pathway. The concept of genetic assimi-
lation emerged in part from his work with Drosophila. Waddington speculated
that environmentally induced changes in phenotype could become incorpo-
rated into the genome, as evidenced by the persistence of the phenotype even
after the original selection pressure is relaxed. It is in this manner that natural
selection acts on developmental pathways leading to adaptive change in the
genome rather than relying on genetic mutation. This dynamic view of
EPIGENETICS AND ANIMAL BEHAVIOR 267

development incorporates both homeostasis (the stability of a final steady

state) and homeorhesis (the stability of the process of development itself).

Molar Epigenetics
There are two types of molar epigenetics. The first arose from early evolu-
tionists who asked how different phenotypes within a species were shaped by
different environments. This area of study fell out of favor for about 60 to
70 years in European and American science. Interestingly, it continued as a
major field of study in Russia and was represented in small part in this country
in the work of Theodosius Dobzhansky and his students, most notably
Richard C. Lewontin (2000). Today, it has reemerged as a vigorous area of
research among evolutionary biologists and behavioral ecologists. New
research on the origins of polymorphisms (multiple phenotypes in a single spe-
cies) and polyphenisms (multiple phenotypes from a single genotype) has led
to a concept now commonly referred to as phenotypic plasticity, which is
considered one of the driving forces in the relatively new union of develop-
mental biologists with evolutionary biologists (evo-devo).
The other type of molar epigenetics has an equally old history. In psychol-
ogy there has long been an interest in behavioral development or behavioral
organization. Zing-Yang Kuo, who worked principally in the 1920s and
1930s, created much of the theory. Unfortunately, Kuo returned to China
where the political strife interrupted research, and as a consequence his contri-
butions were marginalized (Kuo, 1967; Greenberg, 2000). The other major
figure in the field was Karl S. Lashley and his students, most notable for the
purposes of this review Frank A. Beach (regarded as one of the founders of
neuroendocrinology) and Theodore C. Schneirla (whose approach to the
development and display of species-typical behaviors focused on the interac-
tion of the genetic and the environmental levels of biological organization).
Examples of this integrative approach are now numerous, but two classic
efforts were those of Daniel S. Lehrman and Jay S. Rosenblatt. Lehrman con-
ducted elegant work on the elaborate interaction of parent and offspring that
results in ring doves (Streptopelia risoria) learning to care for their young, and
Rosenblatt carried out exquisite research on the physiological and behavioral
events that underlie the development of maternal behavior in cats and later
rats. Both were students of Schneirla and emphasized the dynamic nature of
a process that involves the interaction of the internal milieu and the organism
and the interaction of the organism and its environment. Moreover, they
defined the environment broadly to include the behavior and physiology of
socially important species members. In so doing they laid the foundation for
psychobiology, a vibrant field that focuses on how experiences accumulate
268 ANIMAL BEHAVIOR

throughout life to shape the way in which the individual interacts with its
social and physical environment (Gottleib, 2002).
It is not my purpose to venture into the relatively unexplored frontier that
lies in uniting the two subdisciplines of molar epigenetics (namely that of evo-
lutionary and developmental biology and psychobiology) and behavioral
neuroendocrinology. However, it is useful to be reminded of Ernst Mayr’s
constant refrain that behavior is at the leading edge of evolution and the obser-
vation of Michel and Moore (1995, p. 178) that “mechanisms that underlie
much of behavioral evolution may reside in the processes studied by develop-
mental psychobiologists.”
However, it is necessary to emphasize before going further that the individ-
ual is the unit of selection and that an approach that integrates both molecular
and molar epigenetics will be necessary to reveal the mechanisms that underlie
behavioral evolution (Bateson & Gluckman, 2011). That is, the continuity
between molecular and molar epigenetics is revealed as the constituent ele-
ments interact both positively and negatively in a temporal, spatial, and condi-
tional (internal as well as in the social and physical environments) context
(Nijhout, 2004). As adaptive responses emerge, they, in turn, set the stage
for future variation. Thus, evolution is a tandem process involving first devel-
opment, with its built-in flexible responsiveness to both gene products and
environment, followed by selection, which dictates which variants are spread
and maintained (Stearns, 1989; Lewontin, 2000; West-Eberhard, 2003). In
this sense the “genome learns from its experience” (Jaenisch & Bird, 2003).
Obviously, suites of genes underlie the fundamental plasticity of an organism,
particularly during development or life-stage transitions. How do these gene net-
works interact with the experiences that accumulate during an individual’s life his-
tory? An important interface between the environment (either internal or
external) and the genotype is that of epigenetic modifications. Exactly how these
modifications come about is still relatively unknown, but recent studies at both
the molecular and molar levels indicate that the origin of such effects may occur
in previous generations. That is, experiences of earlier generations can modify
regulatory factors affecting gene expression such that the DNA sequence itself
is not changed but the individual’s physiology and behavior are substantially
influenced. Understanding how such modifications actually occur will increase
our understanding of how the environment influences the relationship between
genotype and behavior during sensitive developmental periods.
Before reviewing this literature, it is important to distinguish between
mitotic versus meiotic epigenetic modifications, or what I have termed
context-dependent versus germline-dependent epigenetic modifications
(Crews, 2008, 2010).
EPIGENETICS AND ANIMAL BEHAVIOR 269

CONTEXT-DEPENDENT VERSUS GERMLINE-DEPENDENT EPIGENETIC MODIFICATIONS

The scope of environmental effects that influence patterns of gene expres-
sion in the brain and subsequently behavior is virtually limitless. The extent
to which environmentally induced epigenetic modifications can become
inherited traits depends both upon the nature of the stimulus and the mecha-
nism of its action. At a molecular level, CpG sites (regions of DNA where a
cytosine nucleotide and a guanine nucleotide are adjacent) are often associated
with 5’ promoter regions of genes and have a higher probability of undergoing
mutation than other regions of the genome. Consequent changes in DNA
methylation patterns at CpG islands (DNA regions that contain a high fre-
quency of CpG sites) would persist and, if imprinted in the germline, have
the potential of becoming heritable.

Context-dependent Epigenetic Modification

Best studied are the epigenetic modifications that either have an effect early
in life, such as exposure to endocrine disruptor chemicals (EDCs) in utero or
smoking during childhood and adolescence. In the first instance the onset of
disease manifests later during the individual’s lifetime, while in the latter in-
stance, the deleterious effects of smoking decline with time only if the individ-
ual is no longer exposed to the stimulus. Similarly, an epigenetic modification
can be perpetuated across generations by simple persistence of the causal envi-
ronmental factor such that each generation is exposed to the same conditions.
For example, if the diet (Dolinoy, 2008; Faulk & Dolinoy, 2011) or environ-
mental toxicant such as lead continues to be present in the environment, then
the epigenetic modification will be manifested in each generation. This type of
epigenetic modification lends itself to relatively straightforward therapeutic
venues such as providing methyl donors in the diet (Dolinoy, 2008), and
removing the environmental toxicant, whether smoking or lead. Hence, the
environmental exposure will induce epialleles (genes that differ in the extent
of methylation but otherwise are identical), but this environmentally induced
epigenetic state can be reversed by a different environmental factor. This
mitotically based effect can be termed “context-dependent” epigenetic change
(Crews, 2008).
The best example of context-dependent epigenetic modification and behav-
ior is that of Meaney and colleagues (Kappeler & Meaney, 2010). In a long
series of elegant studies this group has demonstrated that the nature and
amount of care a pup receives from the mother modulates its reaction to stress
later in life, largely through effects on the glucocorticoid receptor (GR) in the
hippocampus. This maternal effect can cross generations, but its heritability
270 ANIMAL BEHAVIOR

depends upon the pup’s experience in the first week of life. Recently this
group has documented that being reared by a high-quality mother results in
the expression of the transcription factor A (NGFI-A), a nerve growth fac-
tor–inducible protein, that binds to the first exon of the GR gene, resulting
in increased expression of GR. High-quality maternal care during this critical
period demethylates NGFI-A and the acetylation of histones. Just as cross-
fostering can reverse these molecular and behavioral changes, infusion of
methionine, a histone deacetylase inhibitor, into the hippocampus can also
reverse these events. It is important to point out, however, that the effect of
high- and low-quality mothering cannot be selected for and eventually disap-
pears after five generations. That is, it is neither possible to selectively breed
for quality of maternal behavior or to pass the effect nongenomically
indefinitely.

Germline-dependent Epigenetic Modification

Germline-dependent epigenetic modifications are fundamentally different
than context-dependent epigenetic modification in that the epigenetic imprint
has become independent of the original causative agent. That is, the epigenetic
modification is transferred to subsequent generations because the change in
the epigenome has been incorporated into the germline. Thus, the effect is
manifested each generation, even in the absence of the causative agent. In such
instances the DNA methylation of heritable epialleles is passed through to
subsequent generations rather than being erased as occurs normally during
gametogenesis and shortly after fertilization. It is important to note that
because germline-dependent epigenetic modifications are mediated through
the germline, they tend to be sex linked. Examples of this type of epigenetic
modification are still relatively rare. Nonetheless, the work of Skinner and col-
laborators on certain pesticides and fungicides demonstrates that such effects
can occur (see below).

Transgenerational Inheritance
The defining distinction between context- and germline-dependent
epigenetic modifications lies in the timing and persistence of the exposure.
Exposure to environmental or psychological stressors will bring about change
in the epigenome, but the transmission of the effects of that exposure can
occur in two basic ways. Context-dependent epigenetic modifications are in
direct response to the stimulus. Thus, an endocrine disruptor in the environ-
ment will induce changes in all individuals that are exposed to it and, as long
as the environment stays contaminated, further generations will also exhibit
EPIGENETICS AND ANIMAL BEHAVIOR 271

the modification (unless individuals undergo adaptive molecular changes that

buffer them from the toxicant). All of the exposed individuals will have a body
burden of the chemical and can pass that to their offspring (males can transmit
the effect to their biological children or F1, whereas females can pass it to both
their offspring and their grandchildren, or F2, but not to the F3 generation)
(Skinner, 2008). On the other hand, germline-dependent epigenetic modifica-
tions can be transmitted to future generations without the requirement of
additional exposure. In such instances removal of the contaminant will not
result in resumption of the original, nonmodified state because the modifica-
tion has become part of the germline and will pass to all future generations.
Thus, only germline-dependent epigenetic modifications are truly transge-
nerational in nature.
Some might argue that using the term epigenetics without referring to a spe-
cific epigenetic mechanism is unacceptable. However, as detailed above, a
proper definition of epigenetics extends beyond that used by molecular biolo-
gists, incorporating in addition functional outcomes. Thus, transcription,
physiological, brain, and behavior changes all fall within the proper definition
of epigenetics.

EPIGENETICS IS A PERSPECTIVE, NOT A TECHNIQUE

By now it should be obvious to the reader that the most important aspect of
how one goes about studying behavioral epigenetics is to realize that it is more
an issue of perspective or question and less of the tools and techniques to be
implemented. For this reason virtually any aspect of behavior is open to the
investigator; tissue differentiation, developmental psychobiology, cognitive
development, psychopathology, life-history strategies, and phenotypic plastic-
ity are just a few examples. After the question is formed comes the choice of
organism to study. The two basic choices are naturally occurring species and
conventional animal models. It is important to know the advantages and dis-
advantages of each. By studying diversity (naturally occurring species) we gain
insights into evolutionary and ecological principles that can then be applied to
vertebrates, including mammals (Crews & Moore, 1986). In my own work,
discoveries made with lizards, such as hormone independence of sexual behav-
ior and the role of progesterone and its synergy with testosterone in sexual
behavior of males, have also been demonstrated in mammals and have become
important fields of investigation. However, naturally occurring species have
certain drawbacks, not the least of which is that they require special environ-
mentally relevant cues not easily simulated in the laboratory. Another impor-
tant disadvantage is the fact that usually only adult individuals are available
272 ANIMAL BEHAVIOR

for study. Moreover, the investigator should always keep in mind that, com-
pared with the original population, these individuals are only the ones that have
managed to survive, and that the experiences they may have encountered as they
grew are often lost to us. Also, many of the molecular tools that are routine to
those working with rats and mice are not readily applied to the unconventional
animals. Thus, those interested in the development of behavior, particularly
how events early in life influence later behavior or how the epigenetic changes
that occur from particular experiences may alter future behavior, may find it nec-
essary to use conventional animal models. These animals have been stripped of
their ecologically relevant traits, are well studied, and have been the template
on which molecular tools were forged. Again the investigator must keep in mind
the limitations of the animal. In general, model systems are analogous to a drag-
ster. Basically, in drag racing the machine is an engine on a chassis guided by a
driver, with the goal of getting to the quarter-mile mark as fast as possible. This
is no different from an inbred strain of rodent, bird, amphibian, fly, nematode,
or other organism in that inbred strains have maximized fitness (reproduction
and growth rate) in an artificial and basically barren environment. That is, the
conventional animal model organism is basically a gonad guided by a brain, with
the “goal” of reproducing as fast as possible. In both the model organism and the
dragster, there are no “bells and whistles” that may be demanded by the average
customer, or in the case of a C57 mouse, a day in the life in the wild.
Next comes the question of what phenotype to study. A phenotype consists
of multiple traits; each trait is defined as any measurable aspect of the individ-
ual. In general, our understanding of a particular phenotype increases propor-
tionally with the number of traits that are measured in the same individual.
Selection of the particular morphological, physiological, behavioral, and brain
nucleus traits should be predicated on the literature and demonstrated to be
important to the question at hand. The same principle applies to genes in that
individual genes only have meaning in the context of other genes within and
outside their functional categories.

EXAMPLES OF ENVIRONMENTAL FACTORS BRINGING ABOUT EPIGENETIC CHANGES

It is important to note that study of epigenetic modifications need not delve
into the molecular underpinnings. The nature of the question is what is impor-
tant, not the techniques employed. Several examples will illustrate this point.

Temperature and Sexual Experience as Agents of Epigenetic Modification

In the leopard gecko (Eublepharis macularius), temperature rather than
sex chromosomes determine gonadal sex. Low (26°C) and high (34°C)
EPIGENETICS AND ANIMAL BEHAVIOR 273

incubation temperatures produce only females, while intermediate incubation

temperatures produce different sex ratios; 30°C (Tf) produces a female-biased
sex ratio (25:75, or Tf), and 32.5°C a male-biased sex ratio (75:25, or Tm).
Incubation temperature not only establishes the gonadal sex of the individual
but also accounts for much of the within-sex variation observed in the mor-
phology, growth, endocrine physiology, and aggressive and sexual behavior
of the adult (Sakata & Crews, 2004). For example, males in general grow more
rapidly and are larger than females from the same incubation temperature; Tm
males, however, grow more rapidly and to a larger size than do Tf males.
Hatchling, young, and adult Tm and Tf males do not differ in circulating con-
centrations of androgens. Estrogen levels do differ significantly, however, with
Tf males having higher levels than do Tm males. Despite this similarity in cir-
culating androgen levels in adulthood, males from the two temperature
morphs differ significantly in their scent-marking response to exogenous hor-
mones in adulthood, indicating neuroendocrine differences between the Tf
and Tm males. There are also between-sex as well as within-sex differences
in glucocorticoid levels in response to stress. Females have higher circulating
levels of corticosterone than males, but, for both females and males, Tm indi-
viduals have significantly lower levels than do Tf individuals. Brain neuro-
chemistry is also influenced by incubation temperature. For example, a
significantly higher number of TH-ir (tyrosine hydroxylase immunoreactive)
cells are found in the ventral tegmental area (VTA) of sexually inexperienced
Tf versus Tm males that had been castrated and androgen-implanted, sug-
gesting that embryonic temperature plays a role in differentially organizing
dopaminergic (DA) systems of the temperature morphs. This is supported
by the finding of significantly higher DA levels in the nucleus accumbens of
Tf males compared to Tm males that have interacted with a receptive female
across a barrier. Finally, sexually experienced Tf and Tm males both show
strong preferences in a Y-maze apparatus to females or their odors, but the
type of female they choose depends upon their incubation history. For exam-
ple, given the simultaneous choice between two females from different incuba-
tion temperatures, Tf males prefer females from eggs incubated at high
temperatures (34°C), while Tm males prefer the Tf females. Among females,
Tm females are less attractive to males than are Tf females and will even attack
males, a typically male pattern of aggression.
The long-term effects on the brain of significant life-history events are best
revealed using cytochrome oxidase (CO) histochemistry. This is one of many
measures of metabolic activity, but it has the advantage that CO is a rate-
limiting enzyme in oxidative phosphorylation, the major pathway in brain
metabolism. Consequently, the abundance and activity of CO activity in a
274 ANIMAL BEHAVIOR

brain area is a measure of the metabolic capacity of that brain region over time.
In other words, CO abundance not only reflects the metabolic history of an
area but, because it determines the amount of ATP available in a neuron, it
constrains the amount of activity a neuron can sustain (Sakata et al., 2005).
It differs from other types of brain activity measures such as 2-deoxyglucose,
immediate early gene expression, magnetic resonance imaging, and so forth
in that it does not assess the current activity of the brain area so much as its
past history of activation.
As in other vertebrates, the septum (SEP), ventromedial hypothalamus
(VMH), anterior hypothalamus (AH), nucleus sphericus (SA) (homolog of
mammalian medial amygdala), preoptic area (POA), and periventricular pre-
optic area (PP) are major integrative areas for hormonal effects on sexual and
agonistic behavior in the leopard gecko. Incubation temperature influences
the metabolic capacity of forebrain nuclei in adult leopard geckos, and, fur-
ther, these differences correlate with the differences exhibited in their sexual
and agonistic behaviors as adults. Sexual experience also influences the organi-
zation of the neural circuits underlying social and sexual behavior. As illus-
trated in Figure 8.1, the functional landscape changes significantly according
to incubation temperature of the embryo but not so much according to the
sexual experience of the individual in adulthood. That is, if one examines the
effect of incubation temperature (comparison within columns), both inexperi-
enced and experienced Tm females show greater activity in the AH, NS, and
SEP (but not in the POA, VMH, or PP, which are unchanged). The POA
and AH particular nuclei are centrally involved in maturation of the
hypothalamus-pituitary-gonadal axis and the NS and SEP in the maturation
of the hypothalamus-pituitary-adrenal axis. A different picture emerges when
comparing inexperienced and experienced females from within each incuba-
tion temperature (comparison within rows). In this instance adult sexual expe-
rience modifies the effect of embryonic incubation temperature in the POA in
the all-female incubation temperature and in the POA and VMH in Tm
females, but to a lesser degree (note difference in scale). This suggests that in
the leopard gecko incubation temperature has a more profound effect on brain
organization than does adult sexual experience.

Litter Composition Shapes the Development of Brain and Behavior

For those interested in behavioral development in mammals, it is not neces-
sary to go further than the litter environment for the context in which epige-
netic effects can occur. Specifically, it is precisely this period of postnatal
development, during which the individual is nurtured within the litter, that
most influences its behavior as an adult and, as such, the activity of the neural
Figure 8.1. Incubation temperature modifies the abundance of cytochrome oxidase in
limbic nuclei subserving sociosexual behavior in the adult female leopard gecko
(Eublepharis macularius). Illustrated are means of cytochrome oxidase abundance
relative to background in each nucleus of each of four groups of geckos. Eggs were
incubated at one of two temperatures (all-female, or 26°C, and male-niased, or
32.5°C), but the hatchlings were raised at identical temperatures. At one year of age
females were allowed to breed for one reproductive season (experienced) or remained
inexperienced. Note that in both sexually inexperienced and experienced groups
(columns), comparison of the two different incubation temperatures results in a sig-
nificant percentage increase in most, but not all, nuclei relative to overall brain activ-
ity. This in turn results in significant differences in the overall circuit. Within each
incubation temperature (rows), however, adult sexual experience modifies the effect of
embryonic incubation temperature in only the POA (all-female) or the POA and
VMH (male-biased) nuclei. The effect of experience on the overall circuit is not sig-
nificant at the all-female temperature and only marginally significant at the male-
biased temperature. Values are average cytochrome oxidase abundance in identified
cell nuclei relative to background. Brain nuclei: ventromedial hypothalamus (VMH);
anterior hypothalamus (AH); nucleus sphericus (NS); preoptic area (POA); periven-
tricular preoptic area (PP); septum (SEP). The bottom row reveals the effect of
embryonic temperature; peaks above the plane indicate values that are greater at the
male-biased incubation temperature. The right column reveals the effect of adult
experience; peaks above the plane indicate values that are greater in sexually experi-
enced individuals. An asterisk indicates significant differences in particular nuclei.
276 ANIMAL BEHAVIOR

circuitry that underlies these behaviors. Consideration of such epigenetic

effects is particularly needed in studies of genetically modified mice, on which
much of the genotype × environment (interaction of genotype and environ-
ment) research is conducted. Although the mouse is ideal for genetic work,
too often the investigators expediently will assign animals to experimental
groups without considering family-of-origin issues such as sex ratio, genotype
ratio, size of litter, maternal care, or unanticipated stresses that occur in any
colony. These issues are not trivial, as any behavioral neuroscientist working
with rats knows, but may not be considered important by molecular neuro-
scientists who work principally with mice. It might be argued, as a justification
for this research flaw, that in the colony as a whole there exists a Mendelian
genotype ratio and an even (50:50) sex ratio. However, this is not a valid argu-
ment because the experimental animals actually used in the study do not come
from a population but from a litter that has a particular sex and genotype ratio.
In model systems that are the result of the mating of heterozygotes (HTZ) to
yield litters of varying numbers of wild-type (WT), HTZ, and knockout (KO)
(genetically engineered mouse with an inactivated or “knocked-out” gene)
young of both sexes, the ratio of the various genotypes is as important as the
sex ratio of the litter. Appropriately designed studies need to control for the
distribution of individuals within the groups representing all litter types.
Indeed, until the investigator can show that both sex ratio and genotype ratios
are equally distributed and litters are equally represented in all of the experi-
mental groups, any conclusions are suspect.
This is a bold statement, but evidence backs this contention. In collaboration
with Sonoko Ogawa, I have examined how the sex and genotype ratios of a litter
might contribute to the development of behavior in mice having a null mutation
of the estrogen receptor ERα (Crews et al., 2004, 2009). By mating mice hetero-
zygous for a null mutation for this gene it is possible to reconstitute litters shortly
after birth to control for both sex (,/<) ratio and KO and WT genotype ratios.
Thus, the possible combinations are (1) same-sex, same-genotype litters
(e.g., ,WT/,WT); (2) same-sex, mixed-genotype litters (e.g., ,WT/,KO); (3)
mixed-sex, same-genotype litters (e.g., ,WT/<WT); or (4) mixed-sex, mixed-
genotype litters(e.g., <KO/,WT). By reconstituting the litters into one of the
16 possible combinations, the effect of genotype can be examined without the
potential confound of the presence of the opposite sex in the litter, and the effect
of siblings of the opposite sex can be studied without the potential confound of
littermates with a different genotype.
The results of such studies indicate clearly that both factors are important
in shaping the behavior when the individual is an adult. For example, one of
the behavioral diagnostics of ERαKO female mice is that as adults they are
EPIGENETICS AND ANIMAL BEHAVIOR 277

very aggressive (Ogawa et al., 1998). However, this occurs only if ,KO are
raised in litters containing other ,KO; if raised with ,WT or <WT they fail
to show aggressive behavior and are comparable to ,WT mice in social contact
time. Indeed, litter composition influences the development of sociosexual
behaviors in ERKO mice of both sexes. Extending this work to the brain,
I examined the pattern of metabolic activity in various brain nuclei of the mice
raised in these controlled litter groups (Crews et al., 2009).
It is of interest that WT females raised in same-sex, same-genotype groups
spend significantly more time in social contact in a resident-intruder test com-
pared to KO females raised in same-sex, same-genotype groups (Figure 8.2).
Further, it appears that female WT siblings are able to compensate for this
deficit, just as KO siblings cause a deficit in WT females. The neural network

Figure 8.2. Social contact time in genetically modified mice raised in reconstituted litters
arising from mating of mice heterozygous for a null mutation of the estrogen receptor a
gene. Litters consisted of equal numbers of wild-type (WT) or knockout (KO) female (F)
or male (M) mice. Shown are results of individuals raised in single-sex, single-genotype
litters (FW/FW and FK/FK); single-sex, mixed-genotype litters (FW/FK and FK/FW);
mixed-sex, single-genotype litters (FW/MW); and mixed-sex, mixed-genotype litters
(FK/MW); The tested animals were the sex and genotype of the first symbol (e.g., a
female WT individual raised in the single-sex, mixed-genotype litter (FK/FW). Signifi-
cant differences occurred between FW/FW and FK/FW, FW/- and FK/FK, and FK/FK
and FK/MW.
278 ANIMAL BEHAVIOR

that underlies sociosexual behavior varies in different ways. The relative effects
of sex independent of genotype, and of genotype independent of sex, on the
neural network are striking (Figure 8.3). Taken together these findings indi-
cate that in studies with genetically modified mice, litter composition during
the preweaning period must be considered because it can affect the develop-
ment of behavior and the neural network responsible for the regulation of
emotional behaviors.

Figure 8.3. Effect of genotype of sisters on the metabolic activity in a social behavior
network in female wild-type (WT) and estrogen receptor α knockout (KO) mice.
Groups are presented according to the type of sibling with which the WT (top row) or
KO female (middle row) was raised, with ,KO or ,WT sisters, respectively. The
limbic functional landscape map on the upper right represents the difference in ,WT
raised with a ,KO sister. Below that is the complement, that is, the effect of a WT
sister on the metabolic activity in ,KO females. The bottom row indicates genotype
differences between ,WT and ,KO females raised in same-sex, same-genotype litters.
The far right column shows the effect of having a sister having the opposite genotype.
The nuclei are presented in a clockwise fashion reflecting a rostral-caudal dimension:
main bed nucleus of the stria terminalis (BNSTma); anteroventral periventricular
nucleus (AVPe); medial preoptic area (MPOA); anterior hypothalamus, anterior
(AHA); medial amygdaloid nucleus, posterodorsal (MeAPD); medial amygdaloid
nucleus, posteroventral (MeAPV); and ventromedial hypothalamic nucleus, ventro-
lateral (VMHVL).
EPIGENETICS AND ANIMAL BEHAVIOR 279

Endocrine Disrupting Chemicals and Epigenetics

Whether intentional or not, an array of chemicals that mimic or block the
action of endogenous hormones are now a permanent part of our environ-
ment. Classified as EDCs, they behave as biological signals and activate the
parts of the endocrine system associated with the steroid/retinoid/thyroid
superfamily of receptors (McLachlan, 2001). Exposure can be limited to a very
restricted period, exist throughout the individual’s lifetime, or have occurred
only in previous generations. Any or all of these exposures can influence all
aspects of an individual’s life history.
It is becoming evident that the mechanism of action of EDCs is probably
epigenetic—in other words, they cause heritable changes in gene function
without changing the DNA sequence, that is, without causing mutations
(these are sometimes referred to as epimutations) (Crews & McLachlan,
2006). Michael Skinner and his group (Skinner & Guerrero-Bosagna, 2009;
Skinner, 2011) provided conclusive evidence that EDCs can reprogram
methylation patterns in the germline, and hence their effects can be transmit-
ted to future generations and expressed without further exposure. They dem-
onstrated that exposure of gestating female rats to the pesticide
methoxychlor or the fungicide vinclozolin during the period of embryonic
sex determination induces an epigenetic transgenerational phenotype through
reprogramming the germline in a sex-specific manner. Specifically, in each
generation males whose ancestor had been treated underwent progressive sper-
matogonial apoptosis (cell death), decreased sperm count and motility, and, as
the animals aged, an accelerated development of adult-onset disease including
cancer, prostate disease, kidney disease, and immune-cell defects. A series of
new imprinted-like genes that transgenerationally transmits this altered epige-
nome to promote disease phenotypes appears not only in the sperm epige-
nome but also in the brain epigenome (Skinner et al., 2008).
The behavior of these individuals is also epigenetically modified. Females
discriminate and prefer male descendants of the line that was not exposed to
the chemical, whereas similarly epigenetically imprinted males do not exhibit
such a preference (Crews et al., 2007) (Figure 8.4).
Specifically, in a partner-preference test, F3-generation females of both the
vinclozolin and control lineages discriminate and prefer males who do not
have a history of exposure; males do not exhibit such a preference. Odor-
preference tests rule out possible differences in the odor-discrimination ability
of epigenetically modified animals; males and females of both lineages explore
odors of the opposite sex much more than familiar (self) odors or novel odors
of the same sex, and all animals explore novel odors of the same sex more than
their own odors.
280 ANIMAL BEHAVIOR

Figure 8.4. Female rats whose great-grandmothers were exposed to the EDC vinclozo-
lin, a commonly used fungicide with endocrine-disrupting properties, and hence were
epigenetically altered, prefer males from the unexposed vehicle lineage (right panel).
Males do not show this preference (left panel). Both females and males from control
and EDC lineages were tested with pairs of vehicle- and EDC-lineage stimulus part-
ners. Presented are the mean (þ1 standard error) differences in the time spent in each
behavior. Right panel: behaviors exhibited by females from vehicle and EDC lineages
towards males from vehicle lineage (positive, right side) and EDC lineage (negative,
left side). Left panel: behaviors exhibited by males from vehicle and EDC lineages
towards females from vehicle lineage (positive, right side) and EDC lineage (negative,
left side). Data from Crews and colleagues (2007).

Stress Induces Context-dependent Epigenetic Modifications

Stress, particularly if sustained, can lead to impaired immunity, disease, and
neurological changes characteristic of major depressive illness and particularly
chronic anxiety disorders (Romeo et al., 2009; McEwen, 2010). Chronic
restraint stress (CRS) in rats has been a standard paradigm for studying such
effects on physiology, brain, and behavior. For example, six hours daily of
immobilization restraint for three weeks results initially in elevated corticoster-
one levels, but after 21 days, the hypothalamic-pituitary-adrenal axis shows
adaptation and levels are back to normal. This, however, underlies a
progressive atrophy of the dendrite length and branching of pyramidal neu-
rons in the CA3 region of the hippocampus, a process mediated by corticoster-
one potentiating postsynaptic activity and the release of excitatory amino acids
from adjacent mossy fiber terminals arising from the granule neurons in the
dentate gyrus and acting via NMDA (glutamate) receptors. Conversely, there
is an increase in dendritic spine density of neurons in the basolateral amygdala
EPIGENETICS AND ANIMAL BEHAVIOR 281

and medial prefrontal cortex (mPFC) and decreased neurogenesis in the den-
tate gyrus. In addition to these structural changes, stressed rats exhibit a variety
of specific cognitive deficits in spatial learning and memory, as well as
increased anxiety-like and agonistic behavior.
The effects of stress, however, appear to vary depending upon the sex of in
the individual and when the stress occur (Shors, 2006; Romeo et al., 2009).
For the purposes of this chapter, I will only consider the literature on male
rats, as that is the epigenetic model best studied. In male rats the effects of
chronic stress early in development tend to be irreversible, resulting in perma-
nent structural changes in the hippocampus and altered adult sociosexual and
anxiety-related behaviors, while those experienced as an adult can be reversed.
If the stress occurs during the peripubertal-juvenile transition, the effects are
similar to early effects, if not exaggerated. In rats, CRS influences serotonin
and dopamine activity in CA3 of the hippocampus, dopamine and its metab-
olites in CA1 of the hippocampus as well as the mPFC, and dopamine and its
metabolites in the basolateral amygdala.

Integration of Germline- and Context-dependent Epigenetic Modifications

It is clear that individuals modify, or even recreate, their environment via
behavior. Two challenges that must be considered are (1) environmental stres-
sors and (2) psychological stressors. Both are forced upon the individual,
because they either are part of the environment the organism is born into or
are visited upon it during its life. Both induce epigenetic modifications, but
of a different sort. The former is dependent upon exposure while the latter
can result in alterations to the genome of future generations independently
of changes to the DNA sequence through mechanisms that include DNA
methylation. There is now clear evidence that an individual’s likelihood of
developing health problems involves a combination of that individual’s own
exposures as well as exposures of ancestors in generations past. Another factor
that must be considered is (3) the hormonally induced differentiation of body
and brain triggered by genes, resulting in sex differences in physiological and
behavioral responsiveness to environmental stimuli. In rodents and humans,
males and females differ substantially in reactions to environmental challenges
and their propensity to develop illness, disease, and affective disorders.
Recent studies with rats have combined both the germline-dependent
(transgenerational epigenetic modification) effects with a context-dependent
(CRS during adolescence) effect. In this case only males were studied, but
the results clearly confirmed that a single exposure to vinclozolin three gener-
ations removed alters the brain epigenome, transcriptome, physiology, behav-
ior, and metabolic activity in discrete brain nuclei in F3 descendant males.
282 ANIMAL BEHAVIOR

What is also evident is that this transgenerational epigenetic modification

causes the F3 males to respond differently to stress. The following are a few
examples of such changes at the different levels of biological organization
(Crews, 2011).
The pattern of body weight (BW) gain differs according to lineage and
stress, with vinclozolin-lineage males gaining weight more rapidly and becom-
ing heavier than control-lineage males; CRS abolishes this difference. In males,
CRS results in lower corticosterone (CORT) levels in both the control- and
vinclozolin-lineage groups. Lineage, but not stress, influences circulating tes-
tosterone (TESTO) levels. Among stressed animals, TESTO levels are signifi-
cantly higher in vinclozolin-lineage males relative to control-lineage males.
At the level of behavior, a number of differences are evident (Table 8.1).
The open-field (OF) test measures anxiety and emotionality. We find that
control-lineage nonstressed (CL-NS) males spend more time in the corners
of the OF than do vinclozolin-lineage nonstressed (VL-NS) males (Table 8.1).
Exposure to CRS during adolescence has opposite effects in the two line-
ages; CL males move out of corners and into the center, indicating greater
exploration, whereas VL males move from the center into corners, indicating
greater anxiety (interaction between lineage and stress). There is also an effect
of stress independent of lineage: stressed males move faster through the center
than do nonstressed males, indicating that CRS during adolescence increases
anxiety later in adulthood. Sociability 1 measures social approach, anxiety,
and exploration. We find that lineage effects are restricted to the stress condi-
tion, with VL males traveling further and faster than the CL males and choos-
ing to associate with the stimulus animal more than nonstressed individuals.
In general, CRS during adolescence affects line crossing and latency to first
entry into the chamber containing the stimulus animal. In the nonstress con-
dition, VL males visit the stimulus animal for longer periods and move
between chambers less than do CL males. Sociability 2 measures social novelty
and working memory. We find that CL-NS males spend more time with the
stimulus male that is novel than with the familiar male. Only VL males show
effects of stress, traveling further and faster than VL-NS males; they also spend
less time in the center compartment and more time with the familiar and
novel stimulus males. Comparison of the two tests reveal that in Sociability
1, VL-S males tend to spend less time in the center compartment than do
VL-NS males, a difference that becomes significant in Sociability 2, suggesting
that following CRS, VL males display greater affiliation behavior with the
familiar individual. In CL males, there is no effect of CRS, but mean center
time decreases in Sociability 2, a difference significant only in the stress condi-
tion. Similarly, VL-S males tend to spend more time in the animal chamber in
EPIGENETICS AND ANIMAL BEHAVIOR 283

Sociability 1 than do VL-NS males; in Sociability 2, this difference becomes sig-

nificant, again suggesting formation of a social bond with the familiar animal.
At the level of brain nuclei, we find effects of both lineage and stress. Table 8.1
summarizes these findings, and Table 8.2 shows the nature of the effects.
For example, amygdaloid nuclei are differentially affected by lineage; CO activity
in posteromedial cortical amygdala (PMCo) is higher in CL males regardless

Table 8.1. Summary of data obtained at three different levels of biological organization
in pilot experiment on the interaction of transgenerational epigenetic modifications three
generations removed and exposure of chronic restraint stress during adolescence on adult
behavior of male rats. VL ¼ vinclozolin lineage; CL ¼ control lineage; S (stress) ¼
restraint stress during adolescence; NS ¼ nonstress; L × S Interaction ¼ interaction
between ancestral and proximate exposures; Yes ¼ an interaction exists; No ¼ no inter-
action exists. Symbols: – ¼ no effect; > ¼ greater in one group compared to the other;
þþ ¼ statistically significant difference; – ¼ no statistically significant effects.
Trait Lineage Stress L × S interation

Body weight VL > CL þþ (VL) Yes

ASI VL > CL þþ (VL) Yes
Corticosterone – þþ (VL) Yes
GSI – þþ (VL) No
Testosterone – þþ (VL) Yes
Leptin – – No
Behavior Lineage Stress L × S interaction

Forced Swim – – –
Open Field þþ (CL) þþ (VL CL) Yes
Sociability 1 – trend (VL) NS
Sociability 2 þþ (VL) þþ (VL) Yes
Nucleus Lineage Stress L × S interaction

BLA þþ (S) – Yes

CA1 þþ (S) þþ (VL) Yes
CA3 þþ (S) þþ (VL) Yes
MeAmy þþ (NS) þþ (CL) Yes
MePD þþ (S) þþ (VL CL) Yes
PMCo þþ (NS S) – No
Stria – (S) þþ (VL) Yes
284 ANIMAL BEHAVIOR

Table 8.2. Significant changes in cytochrome oxidase abundance in brain nuclei as a

consequence of Lineage and Stress.

of stress condition. The medial posterior dorsal amygdala (MePD) shows opposite
effects following CRS (pronounced increase in CL and decrease in VL males). In
the medial amygdala (MeAmy), CO activity is opposite in the lineages depending
upon stress. We also see that the stria terminalis (ST), and not the bed nucleus of
the stria terminalis (BnST), shows marked changes, indicating that activity in this
major pathway is being modified by lineage. In the CA1 and CA3 of the hippo-
campus, stress decreases metabolic activity in the VL males but has no effect in
CL males.
Thus, the alteration of baseline brain development brought about by trans-
generational epigenetic modification promotes a change in neural genomic
activity that correlates with changes in physiology and behavior, revealing
the interaction of genetics, environmental epigenetics, and epigenetic transge-
nerational inheritance in shaping the adult phenotype. This is the first empiri-
cal evidence in an animal that ancestral exposure to a known EDC modifies
how descendants of these progenitor individuals perceive and respond to a
stress challenge experienced during their own life history.

SUMMARY
We are at the very beginning of studies of the epigenetics of behavior.
Behavioral phenotypes are affected by multiple factors, some beginning in
generations past and others originating during sensitive periods or life stages.
We now know that genes do not cause behavior, and there is little evidence
that, outside of disease and pathology, genotypes predispose individuals to
behave in particular ways. Understanding the development of behavior has
yielded more information on the causes of behavior. For example, different
experiences during sensitive life stages produce variation among individuals
that markedly influence how the individual responds to social and sexual cues
later in adulthood. This variation is the substrate on which evolution can act.
EPIGENETICS AND ANIMAL BEHAVIOR 285

Several studies that deconstruct various confounds inherent in research in

developmental psychobiology have been presented. For example, in the
ERαKO mouse, we have seen how the sex and genotype ratios of the litter
have separate and distinct effects on the nature and quality of the individual’s
behavior later in adulthood, as well as on the metabolic activity within net-
works of brain nuclei that underlie these behaviors. The finding that func-
tional neural systems can be reorganized depending upon the composition of
the litter in which the individual develops is startling yet yields a deeper under-
standing of how neural systems are organized early in life.
The recent discovery that the environment can affect the genome of future
generations without changing the DNA sequence has particular relevance to
understanding the brain and behavior. However, it remains to be seen if our
increasing understanding of the causes and effects of epigenetic modifications
will produce generalizable insights into the causes and functions of behavior.
This chapter describes some of the mechanisms by which factors influence
adult behavioral responsiveness and their underlying neural substrates; of par-
ticular interest in this regard is how the environment can produce significant
individual variation in social behaviors. Two distinct epigenetic modifications
are described: context-dependent modifications are similar to proximate envi-
ronmental effects, while germline-dependent modifications are equivalent to
ultimate environmental effects in shaping brain and behavior.
It is clear that ultimate and proximate events interact to influence how an
individual responds to events in its own life history, and the study of epige-
netics may be the method by which these issues can be addressed. To date,
only one study has demonstrated that germline-dependent epigenetic modifi-
cations laid down in previous generations alter how individuals respond to a
stressor (CRS) experienced during adolescence (Crews et al., 2012). This
result suggests that different types of experiences can result in different epige-
netic modifications that are independent of one another but that together
influence the phenotype in novel ways.

REFERENCES AND SUGGESTED READING

Bateson, P. & P. Gluckman. (2011). Plasticity, Robustness, Development and Evolu-
tion. Cambridge, UK: Cambridge University Press.
Crews, D. (2008). Epigenetics and its implications for behavioral neuroendocrinol-
ogy. Frontiers in Neuroendocrinology, 29, 344–357.
Crews, D. (2010). Epigenetics, brain, behavior and the environment. Hormones, 9,
41–50.
Crews, D. (2011). Epigenetic modifications of brain and behavior: Theory and prac-
tice. Hormones and Behavior, 59, 393–398.
286 ANIMAL BEHAVIOR

Crews, D., T. Fuller, E. G. Mirasol, D. W. Pfaff, & S. Ogawa. (2004). Postnatal envi-
ronment affects behavior in adult transgenic mice. Experimental Biology and
Medicine, 229, 935–939.
Crews, D., R. Gillette, S. V. Scarpino, M. Manikkam, M. I. Savenkova, & M. K.
Skinner. (2012). Epigenetic transgenerational alterations to stress response in
brain gene networks and behavior. Proceedings of the National Academy of Sci-
ences, USA, 109, 9143–9148.
Crews, D., A. C. Gore, T. S. Hsu, N. L. Dangleben, M. Spinetta, T. Schallert, M. D.
Anway, & M. K. Skinner. (2007). Transgenerational epigenetic imprints on mate
preference. Proceedings of the National Academy of Sciences, USA, 104, 5942−5946.
Crews, D. & J. A. McLachlan. (2006). Epigenetics, evolution, endocrine disruptors,
health and disease. Endocrinology, 147, (Suppl.), S4–S10.
Crews, D. & J. A. Moore. (1986). Evolution of mechanisms controlling mating
behavior. Science, 103, 121–125.
Crews, D., D. Rushworth, F. Gonzalez-Lima, & S. Ogawa. (2009). Litter environ-
ment affects agonistic behavior and brain metabolic activity of adult estrogen
receptor α knockout mice. Frontiers in Behavioral Neuroscience, 3, 12.
Dolinoy, D. C. (2008). The agouti mouse model: An epigenetic biosensor for nutri-
tional and environmental alterations on the fetal epigenome. Nutritional
Reviews, 66 (Suppl 1), S7–11.
Faulk, C. & D. C. Dolinoy. (2011). Timing is everything: The when and how of environ-
mentally induced changes in the epigenome of animals. Epigenetics, 6, 791–797.
Gilbert S. F. & D. Epel. (2008). Ecological Developmental Biology: Integrating Epige-
netics, Medicine, and Evolution. Sunderland, MA: Sinauer Associates.
Gottlieb, G. (2002). Individual Development and Evolution: The Genesis of Novel
Behavior. Mahwah, NJ: Lawrence Erlbaum Associates.
Greenberg, G. (2000). Lessons from Zing-Yang Kuo. Past to Future, 2, 13–37.
Hertwig, O. (1894). Präformation oder Epigenese? Grundzüge einer Entwicklungstheorie
der Organismen. Jena, GER: G. Fischer.
Holliday, R. (1990). DNA methylation and epigenetic inheritance. Philosophical
Transactions Royal Society of London, B, 326, 329–338.
Jaenisch, R. & B. Bird. (2003). Epigenetic regulation of gene expression: How the
genome integrates intrinsic and environmental signals. Nature Genetics, 33,
245–254.
Jessen, H. M. & A. P. Auger. (2011). Sex differences in epigenetic mechanisms may
underlie risk and resilience for mental health disorders. Epigenetics, 6, 857–861.
Kappeler, L. & M. J. Meaney. (2010). Epigenetics and parental effects. BioEssays, 32,
818–827.
Keverne, E. B. (2009). Monoallelic gene expression and mammalian evolution.
BioEssays, 31, 1318–1326.
Keverne, E. B. & J, P. Curley. (2008). Epigenetics, brain evolution and behaviour.
Frontiers in Neuroendocrinology. 29, 398–412.
Kuo, Z.-Y. (1967). The Dynamics of Behavior Development. New York: Random
House Press.
EPIGENETICS AND ANIMAL BEHAVIOR 287

Lehrman, D. S. (1970). Semantic and conceptual issues in the nature-nurture problem.

In L.R. Aronson, E. Tobach, D. S. Lehrman, & J. S. Rosenblatt (eds.), Develop-
ment and Evolution of Behavior (pp. 17–52). San Francisco: W. H. Freeman.
Lewontin, R. C. (2000). The Triple Helix: Gene, Organism and Environment. Cam-
bridge, MA: Harvard University Press.
McEwen, B. S. (2010). Stress, sex, and neural adaptation to a changing environment:
mechanisms of neuronal remodeling. Annals of the New York Academy of
Sciences, 1204, 38–59.
McLachlan, J. A. (2001). Environmental signaling: What embryos and evolution teach
us about endocrine disrupting chemicals. Endocrine Reviews, 22, 319–341.
Michel, G. F. & C. L. Moore. (1995). Developmental Psychobiology: An Interdiscipli-
nary Science. Cambridge, MA: MIT Press.
Nijhout, H. F. (2004). The importance of context in genetics. American Scientist, 91,
416–418.
Ogawa, S., V. Eng, J. Taylor, D. B. Lubahn, K. S. Korach, & D. W. Pfaff. (1998).
Roles of estrogen receptor-alpha gene expression in reproduction-related
behaviors in female mice. Endocrinology, 139, 5070–5081.
Romeo, R. D., A. C. Tang, & R. M. Sullivan. (2009). Early life experiences: Endur-
ing behavioral, neurological, and endocrinological consequences. In D. Pfaff,
A. Arnold, A. Etgen, R. Rubin, & S. Fahrbach (eds.), Hormones, Brain and
Behavior. Second Edition (pp. 1975–2004). New York: Elsevier.
Sakata, J. T. & D. Crews. (2004). Developmental sculpting of social phenotype and
plasticity. Neuroscience and Biobehavioral Reviews, 28, 95–112.
Sakata, J. T., D. Crews, & F. Gonzalez-Lima. (2005). Behavioral correlates of differ-
ences in neural metabolic activity. Brain Research Reviews, 48, 1–15.
Shors, T. J. (2006). Significant life events and the shape of memories to come: A
hypothesis. Neurobiology of Learning and Memory, 85, 103–115.
Skinner, M. K. (2008). What is an epigenetic transgenerational phenotype? F3 or F2.
Reproductive Toxicology, 25, 2–6.
Skinner, M. K. (2011). Environmental epigenetic transgenerational inheritance and
somatic epigenetic mitotic stability. Epigenetics, 6, 838–842.
Skinner, M. K., M. D. Anway, M. I. Savenkova, A. C. Gore, & D. Crews. (2008).
Transgenerational epigenetic programming of the brain transcriptome and
anxiety behavior. PLoS One, 3, e3745.
Skinner, M. K. & C. Guerrero-Bosagna. (2009). Environmental signals and transge-
nerational epigenetics. Epigenomics, 1, 111–117.
Stearns, S. C. (1989). The evolutionary significance of phenotypic plasticity.
BioScience, 39, 436–445.
Waddington, C. H. (1942). The epigenotype. Endeavour, 1, 18–20.
West-Eberhard, M. J. (2003). Developmental Plasticity and Evolution. New York:
Oxford University Press.
Woltereck, R. (1909). Weitere experimentelle Untersuchungen über Artveränderung,
speziel über das Wesen quantitativer Artunterschiede bei Daphniden.
Verhandlungen der deutschen zoologischen Gesellschaft 19, 110–173.
This page intentionally left blank
 9

Behavioral Development and

Ontogenetic Adaptation
Jeffrey R. Alberts and Christopher Harshaw

INTRODUCTION
Our goal is for this chapter to alter the way you look at life. More specifically,
we think we can open your eyes to new, interesting, and important ways to
see, to understand, and to appreciate development. Recognition that develop-
ment is an important part of the scientific study of animal behavior has a very
respectable history (Chapter 2). Niko Tinbergen, one of three ethologists rec-
ognized with a Nobel Prize in 1973, said that in order to understand fully an
animal’s behavior, it was essential to understand how it develops (Tinbergen,
1963). Tinbergen saw development as one of the four basic aims of a science
of ethology or animal behavior (Box 9.1). We agree with Tinbergen that
development is an essential component in a biological understanding of
behavior. In fact, we think that understanding development not only addresses
many important aspects of behavior but links the other components of a com-
plete analysis of behavior.

“BEING” AND “BECOMING”: TWO VIEWS OF DEVELOPMENT

Much of the field of animal behavior involves discovering and studying the
marvelous and diverse modes by which animals “earn their living” in the natu-
ral world. For example, we can learn about kinds of navigation and migration
290 ANIMAL BEHAVIOR

Box 9.1
Tinbergen’s four questions

In 1963, Niko Tinbergen, a Dutch ethologist working at Oxford Univer-

sity, in England, published a paper that has become essential reading for stu-
dents of animal behavior. In that the article on “The Aims of Ethology,”
Tinbergen enumerated four kinds of questions that must be answered in
order to achieve a full understanding of behavior in the context of ethol-
ogy—the science of animal behavior. It is useful to think about each of these
questions in relation to a particular behavior that might be studied, such as a
vocal emission, a species-typical postural display, or a pattern of exploratory
behavior, just to cite a few possibilities. According to Tinbergen, the four
key questions address the following topics.

1. Causation. What are the stimuli that elicit the behavior?

2. Function. How does the behavior affect the animal’s survival and reproduc-
tive success?
3. Development. How is the behavior formed and expressed during develop-
ment, and what are the contributions of experience and learning?
4. Evolution. How does the behavior compare with the similar behavior in related
species, and what was the evolutionary process that shaped the behavior?

(Chapter 12), hunting and feeding strategies (Volume 2, Chapter 1), courtship
and reproductive patterns (Volume 2, Chapter 6), and many other such phe-
nomena. When we learn about such stunning examples and then think about
the study of development, it is natural to conclude that the study of develop-
ment, or ontogeny, is the study of how an infant becomes the adult form
that is capable of such stunning feats of behavioral life. John Tyler
Bonner (1958, p.1) bluntly expressed this viewpoint when he stated, “the goal
of development is the final form and function of the adult.” This is a view that
emphasizes development as a process of “becoming.” There is a focus on an end-
point—reproduction, territorial defense, nest building—and development is the
process that prepares the offspring to achieve the endpoint. Viewed this way, the
developmental process includes growth—with increasing strength, expanded and
improved sensory function, and acquisition of motor patterns including complex
behavioral displays and signaling. Special body features and coloration also
develop, often as part of sexual maturation or with the attainment of dominance
status, and these physical features are often used in behavioral displays. The view
of development as the process of becoming is popular, and it is very likely that
this is basically the way that you look at development.
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 291

Yet there are other, contrasting views of development. One of the alterna-
tives is to see development as a series of ways of “being.” In this framework,
the characteristics of the newborn are seen as being complete and well
formed—for the newborn. The infant is recognized in terms of what it is at
the time rather than in terms of what it is not (an adult, for example). As the
infant grows and changes, it is seen similarly as a well-formed youngster
(pup, cub, fledgling, etc., depending on the species) rather than as an incom-
plete or unfinished version of the adult. It is important to recognize that a state
of being well formed and complete can only make sense in relation to a con-
text. A feature of an animal is not good, complete, or functional in its own
right—its qualities can only be judged in relation to its immediate setting. In
this way, the simple grasp reflex of an infant primate can be seen as a behavior
that ensures a strong, reliable grip on the parent’s fur or hair while it is being
carried, even during vigorous movements by the parent. The term reflex often
implies a primitive behavior or a stray element of a more complex, voluntary
behavior that is yet to form. But when we consider development as a series
of ways of being, the same reflexive behavior can be seen in another light—
one of functional integrity in relation to a particular, immediate environment.
Because being and becoming are two contrasting ways of defining develop-
ment, should we consider which is the “proper” or “correct” definition? How do
we choose one versus the other? The answer is not to choose! Being and becoming
are both valid ways of viewing development because the developing animal is,
at the same time, both being and becoming. In the words of a notable develop-
mental physiologist, every developing animal has a “dual vocation” (Adolph,
1968). Every infant is, at the same time, in the business of being a functional,
successful, adapted infant—as well as growing and changing into an adult.
The challenge is for us to see and to appreciate how both of these roles are
accomplished. In this chapter, we will focus on the less familiar and more spe-
cialized perspective of development as a sequence of states of being, for this is
the one that connects most richly to the field of animal behavior. The concept
of ontogenetic adaptation will be the vehicle that will carry us on a journey
through various general considerations and specific examples of development
as a sequence of states of being.

TWO WORTHY WORDS: ONTOGENY AND ADAPTATION

Ontogeny
Ontogeny is synonymous with development. Ontogeny and ontogenetic share
the prefix onto-, which comes from a Greek word meaning “to be.” Onto- is used
to form words that give the sense “of or pertaining to being.” The root -geny is
292 ANIMAL BEHAVIOR

the same one we see in genesis, denoting “modes of formation or generation.” Put
them together (onto + geny) and you have modes of generating a state of being:
development. Admit it, it is an important and lovely word.

Adaptation
The word adaptation comes from the Greek root aptus, “to fit.” To that
root is added the prefix ad-, meaning “toward,” so adaptation denotes a thing
or condition involving some kind of fit, fitness, or fitting. This is true enough,
but the usage of the term adaptation is remarkably important and diverse. The
diversity of usage the word adaptation has been troublesome, as have the
various criteria for what qualifies as an adaptation within the framework of a
particular definition.
Let us consider the different uses of the term adaptation. You will see that
they all have something in common, which is that they recognize the fit or fit-
ness of something or, in some cases, the identification of a specific, immediate
functional role in an animal’s life. Such immediate roles are often called proxi-
mate functions. Other important uses of the term adaptation apply to the evo-
lutionary history of a trait and its functions for survival and reproduction.
Such historical processes and functions that affect the shaping and mainte-
nance of species are often call ultimate functions.
As we noted earlier, the field of animal behavior involves recognition of
many marvelous and diverse phenomena displayed by animals in the natural
world. But, as a science, animal behavior is more than a collection of wonder-
ful animal stories. What elevates animal behavior from a collection of anec-
dotes and makes it into a complete and coherent life science is evolutionary
theory. Evolution unites all of the life sciences, from biochemistry to ecology.
It is far beyond the scope of this chapter to review evolutionary theory in rela-
tion to animal behavior, but we can take a key element of evolutionary think-
ing and discuss it here. The concept of adaptation gets to the heart of the
matter. If we establish a common basis for using the term adaptation, we will
be able to appreciate within an evolutionary context an extraordinary range
of phenomena and see connections among them that will help make more
sense out of the extraordinary complexity of the natural world.

PUTTING IT TOGETHER: ONTOGENETIC ADAPTATION

Now let us combine the two terms to create a new one, ontogenetic adap-
tation. We will discuss the term in a preliminary way and then delve into its
meaning in more depth before considering some examples.
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 293

To those who only see development as a process of becoming, ontogenetic

adaptation is a radical idea. After adopting the idea of ontogenetic adaptation,
each stage of development becomes a complete and full form. An infant is no
longer just an incomplete adult. It is a complete infant!
Of course, an infant is not yet an adult, and in most cases it is very different
from the adult. We will not forget this, and we will come back to it later. But,
for now, let us concentrate on the ideas surrounding ontogenetic adaptation.
A major point here is to recognize that natural selection operates on the organism
at each stage of its development, from the earliest phases of life and continuously
throughout. Thus, we are applying the powerful idea of adaptation as a histori-
cal process that produces functional fit to an environment, usually some fea-
ture in the environment that serves a proximate end (Box 9.2). In short, we
are discussing ontogeny as adaptation or, more accurately, as a series of adap-
tations to a sequence of environments.

Box 9.2
Uses of the term adaptation

There are a variety of uses of the crucially important term adaptation that
are pertinent to biological and behavioral analyses. The meanings can vary
considerably, and there is some overlap among them. Although they differ
in meaning, each is a legitimate and correct meaning. Thus, it is always
important to keep in mind the different uses and define clearly which kind
of meaning is intended when using the term.

1. Relation between the organism and its environment. Adaptation can be used to
describe the overall relation between an organism and its environment. This
is its most general usage and may point the way to further, more focused
analyses of adaptation.
After looking a picture of a tadpole, such as the one in Figure 9.2, it is rea-
sonable to say, “the shape its body, the gills and fins indicate that tadpoles
are adapted to an aquatic environment.”
2. A feature of an organism that serves some proximate end. A popular use of
adaptation is to identify some behavioral, physiological, or anatomical fea-
ture of an organism that provides an obvious, functional benefit—often
one that can be recognized or inferred by direct observation.
After observing some of the animals in Figure 9.4, we might well state that
“the sharp canine teeth of the lion are adaptations to piercing and tearing the
flesh of prey” and “the whale’s pectoral fins, the membranous wings of the fruit
bat, and the articulated limbs of the primate are all adaptations serving locomo-
tion in the species-typical habitats of each of these mammals.”
294 ANIMAL BEHAVIOR

3. The process of adapting within the individual’s lifetime. Part of this usage is
distinct because it explicitly refers to a process that changes within the life-
time of an individual, independent of any necessary genetic change. Because
it implies a change in an animal’s body—including changes in the brain that
mediate learning—this type of adaptation is also called somatic adaptation.
Somatic adaptations include processes such as developing a thicker coat of
fur and layers of fat insulation during winter months, adjusting levels of
digestive enzymes in response to changes in diet, and increasing vigilance
when signs of nearby predators are detected.
The ability to adapt somatically is an adaptation of the type denoted in
usage 2.
4. A historical process spanning multiple generations that alters a population’s char-
acteristics, usually implying a genomic change. This use of adaptation refers to
the process by which adaptive features (usage 2) are obtained.
Beginning in the Cretaceous period, there began a diversification of mam-
malian forms, including an explosion of rodent species, displaying adapta-
tions to a wide range of habitats.

THE ROLES OF BEHAVIOR IN THE ONTOGENETIC NICHE

We have already stated that adaptations do not exist in an absolute way.
Adaptation refers to the organization and function of an animal’s body and
behavior; such organization and function are not absolute states or conditions
that exist on their own. Organization and function are relative terms; their
meanings are seen in relation to an end. Such relations make the difference
between organization and order, and the difference between function and
operation, per se. A dentist’s office is organized in relation to examining
patients’ teeth and performing procedures such as tooth cleaning, drilling, fill-
ing, and perhaps some oral surgeries. An assistant might prepare a tray of
instruments, organized specifically in relation to one of these procedures. A
tray of dental tools cannot be organized in relation to nothing, though it is
possible to imagine a tray with various tools placed in some order but without
a functional purpose.
Adaptation, then, is the organization of the animal—its total organization—
and such adaptation is in relation to the animal’s environment. Just like the
example of the dentist’s office, animal adaptation exists in relation to a func-
tional end. For an adult animal, the functional end may be hunting for food,
winning a mate, maintaining a territory, building a nest, or nursing a new-
born. What about for the newborn? For an infant, the functional end may be
finding a nipple, eliciting parental attention, being held or carried, or soliciting
more food than the others in the same nest.
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 295

It is possible to apply to our discussion of ontogeny and adaptation some

terms borrowed from the language of ecology. Habitat refers to the location
and physical characteristic (temperature, light levels, chemical profile, nutri-
tive characteristics, predator populations, etc.) of that location. Think of hab-
itat as an address at which an animal might live. Niche (pronounced “nich,”
not “nish”) is a related term, but one that is used to include the kinds of activ-
ities an animal uses to live successfully in a particular habitat. Niche-related
activities involve modes of procuring food, staying warm (or cool), marking
and defending territory, and avoiding predators, to name just a few. In short,
an animal’s niche involves its occupation, or how it “earns a living” at a particu-
lar address.
For most species, there are developmental habitats and, consequently,
developmental niches. As you will soon see, an early developmental habitat
of all mammals is the mother’s womb. All mammalian embryos reside at a ute-
rine address during gestation. The uterus, as a developmental niche, involves
how the embryo and fetus use an umbilical cord to breathe and feed. It is
how the fetus earns its living in the uterine niche.
Ontogenetic adaptations are the specializations that shape life in develop-
mental niches. The concept of ontogenetic adaptation shares the power and
importance of the traditional concept of adaptation as a unifying principle that
can help organize an otherwise bewildering array of phenomena and facts.
One way of thinking about development is as a stack of ontogenetic adap-
tations, with the height of the stack representing developmental time, and
each layer being a point in the flow of developmental time. As we look at each
layer in the stack we can see a different set of adaptations to a different envi-
ronment. This is because animals move through different niches as they
develop. Thus, development can be described as a sequence of adaptations to
a sequence of developmental habitats (Galef, 1981; Alberts & Cramer,
1988). Let us consider a few representative examples, using different kinds of
animals. We will consider amphibian, avian, and mammalian development
and see how ontogenetic adaptation can be used a general framework.

THE DIRECTIONS OF DEVELOPMENT

Whether we see development as being or becoming, there is a shared view
that infants differ from their adult counterparts. This is true in terms of body
size, body shape, body composition, hormone systems, brain organization,
and behavior. Comparison of an infant with its adult counterpart provides
an opportunity to contrast the two. Comparisons can reveal differences, and
in mammals the differences between infant and adult are often vast, but a
296 ANIMAL BEHAVIOR

comparison cannot inform us about the transformation itself. Development is

the transformation process.
Again, when we think of developmental transformations, especially involv-
ing cells and systems, we tend to think that they are mostly smooth and that
they proceed in a direction: infants grow to larger adult size; hormonal activity
generally increases and does so dramatically at times such as puberty, preg-
nancy, and lactation; brain cells increase in number and size, and there arises
greater complexity in their connections.
As you might guess by now, there are so many exceptions to the basic
notion that development proceeds smoothly and in one direction that it is
advisable to examine all developmental pathways for discontinuities as well
as for continuities. One kind of discontinuity is seen clearly when a behavior,
body part, circuit of nerves, or hormone function disappears after a period of
continuous presence. Such discontinuities can be created in a variety of ways,
which we will discuss later, but we can see that discontinuities can highlight a
boundary between the feature that ends and the many related features that do
not. A developmental discontinuity can identify and define a component of
behavior or even a brain component, or a component that is shaped by natural
selection. There are other, similar surprises—development can be destructive
as well as constructive; some developmental transformations follow increasing
simplicity rather than complexity and loss of ability rather than gains.
The concept of ontogenetic adaptation can help demystify and make sense
out of such unexpected differences. It is particularly helpful to focus on the
individual organism, the integrated individual, because this is where evolu-
tionary pressure (natural selection) makes developmental change. We can
now anticipate that ontogenetic adaptation will sometimes require the crea-
tion of adaptive specializations at one stage of life that are irrelevant or inter-
fere with adaptions to a later developmental niche.

LIFESTYLES OF AMPHIBIANS, BIRDS, AND MAMMALS:

VIEWS OF SOME ONTOGENETIC NICHES
We have just surveyed a lot of ideas, many of which may be new to you.
Looking back, we placed the study of development among the top issues in
understanding an animal’s behavior (Box 9.1). We have seen that the impor-
tant term adaptation can be used in at least four different yet partially overlap-
ping ways (Box 9.2). We combined ontogeny with adaptation and introduced
ontogenetic adaptation. Then we put all of these ideas within the conceptual
framework of an ontogenetic niche and discussed how development is a series
of adaptations to a series of ontogenetic niches. According to this view,
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 297

development can take many forms and many directions because, as with the
process of biological evolution, there is no single path or final goal.
With so many new and different ideas, it may be difficult for you to judge
them when they are presented in relatively pure form rather than as part of
some specific examples. Looking forward to the next sections, we will describe
three distinct forms of developmental process prevalent in the natural world.
They do not comprise a complete sample, but they are a good start. As you
will see, we illustrate the form of development called metamorphosis by
describing the style of development used by many amphibians, such as frogs.
We then describe two more developmental styles, one shared by all mammals
and another used by birds. Your job, in each case, is to see how an animal’s
body and behavior at each stage of development is in adaptive relations with
the environment that is typical for each stage.

AMPHIBIAN DEVELOPMENT BY METAMORPHOSIS

Metamorphosis involves conspicuous and dramatic changes in body struc-
ture and behavior based on cellular growth and differentiation. Most species
of frogs develop by metamorphosis, so we will focus on them and then note
some of the exceptions, namely species of frogs that use the alternative
of direct development. Frogs are not the only animals that undergo meta-
morphosis as part of their development. There are insects, mollusks,
crustaceans, and other amphibians, as well as some fish, that develop by
metamorphosis.
In frogs, development by metamorphosis always starts with eggs that hatch
and release tadpoles (or polliwogs in some vocabularies). Tadpoles hatch from
externally fertilized eggs and begin their independent behavioral life in water.
Figure 9.1 is a drawing of tadpole, and, as can be seen by the animal’s body
plan, with its absence of appendages and with its fin-like tail, this is an organ-
ism adapted to an aquatic habitat. As adults, frogs are amphibious—living
both (ambi-) in water and on land, whereas the premetamorphic frogs are tad-
poles and are completely aquatic.
Tadpoles emerge from eggs (you can see many examples on Wikimedia
Commons). These eggs are fertilized externally. A representative reproductive
sequence might consist of a male frog emitting its species-typical vocalization,
perhaps situated at the edge of pond. Females of this species are attracted to the
call, which brings them into the male’s vicinity. Mating occurs while the male
is mounted behind the female so that he can deposit his sperm onto the eggs after
they are released from the female. The typical mating posture for many frogs is
called amplexus (which means “embrace”). It typically involves a male mounting
298 ANIMAL BEHAVIOR

Figure 9.1. The morphology of a tadpole indicates that this is an organism that earns
its living in an aquatic habitat. Before it gives up herbivory and shifts to terrestrial
life as a frog, it will reduce the size of its swimming muscles, vascularize for respiration
the skin that serves as its fin, shrink its coiled gut, grow and differentiate the miniscule
hindlimb that is visible on the lateral surface, and remove the operculum over the gills
to expose a growing forelimb. Clearly, the tadpole is more than an incomplete frog.

a female and clasping her body with his forelegs, positioning himself to fertilize
her eggs soon after they leave her body.
In some species, the female frog lays her eggs in a sticky mass that she
attaches to rocks or vegetation under water. In other species, the eggs are laid
and fertilized on the leaves of plants that hang over water so that when the tad-
poles hatch, they immediately fall into an aqueous habitat, which becomes
their next niche in a species-typical sequence.
The amphibian egg itself is a niche, and the embryo must earn its living in
the egg, for this is its early environment in which its brain, gut, sensory sys-
tems, and tadpole body take shape. Indeed, behavior begins in the amphibian
egg, as it does in the avian egg and in the mammalian uterus.
Looking again at Figure 9.1 it should be immediately obvious that the
organism that hatches from a frog egg is adapted to an aquatic lifestyle, as evi-
denced by its proportionately large fin-like tail, gills, and legless body. The
notes with the figure provide an indication of the total drama of metamorpho-
sis. The changes are conspicuous, indeed. Through growth, deletion, and
other processes of remodeling, the tadpole is transformed, inside and out, into
an organism adapted to a more terrestrial life, which earns its living in an
entirely different manner. In fact, there is great diversity among frog species.
We have described their reproduction and development in very general terms
and, for the record, we note that there are some frog species that never become
tadpoles. Such species undergo direct development. From the eggs of such
species hatch tiny frogs—they do not produce tadpoles. These frogs have
abandoned metamorphosis! Direct development is a derived strategy; all direct
developers evolved from predecessors that used metamorphosis (e.g., Elinson,
2001).
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 299

ONTOGENETIC NICHES OF MAMMALIAN DEVELOPMENT

Mammalian development provides many examples of ontogenetic adaptation.
Earlier in this chapter, we introduced the distinction between habitat and niche
because this provides a foundation for seeing behavioral development as an
important component in a sequence of adaptations within a sequence of ontoge-
netic niches. Not only is development understandable within a sequence of
niches, but there is a sequence of developmental niches shared by all mammals.
To appreciate the enormity of the idea of a common sequence of develop-
mental niches for all mammals, first consider the diversity of niches in which
adult mammals are found. The lion (Panthera leo) in Figure 9.2 represents
the many mammalian species that live their lives on the ground, hunting ter-
restrial prey, chasing it down, and killing and consuming it. In contrast, ani-
mals such as the shrew and the mole occupy a different niche, living most of
their lives below the surface of the ground in burrows, eating insects and plant
material. Other mammals live underwater. The fully aquatic mammals
include all the whales and dolphins. Again, these are air-breathing, live-
bearing, lactating mammals—not fish—though they are specialized to live
an aquatic life amid the fish. Finally, Figure 9.2 shows a big brown bat
(Eptesicus fuscus), which is a type of insect-eating bat that lives in a variety of

Figure 9.2. All of the animals shown here are mammals, but they live dramatically dif-
ferent lives. The spider monkeys (Ateles sp.) shown in the upper left are largely arbo-
real, whereas the lion (Panthera leo) is terrestial. The big brown bat (Epesticus fuscus),
on the upper right, flies and roosts for a living, as do all the other bats, whereas the
dolphin (Grampus grisieus) in the lower left represents the various aquatic mammals.
Moles, such as the Talipidae sp. shown on the lower right, earn a living underground.
300 ANIMAL BEHAVIOR

habitats across North, Central, and northern South America and the Carib-
bean. The entire order of Chiroptera (the bats) occupies a niche above the
ground, roosting in trees, caves, and attics, often in giant colonies consisting
of tens of thousands of individuals. The bats are a large and diverse order, con-
sisting of more than 900 species. All of them fly to earn their living. Some feed
on insects, some on fish, others on plant material. Vampire bats, of course,
ingest blood for their subsistence. The bats comprise many diverse specialists.
And, of course, there are many other mammals that live above the terrestrial
substrate, including a variety of primate species that are essentially arboreal.
Ontogenetic adaptation helps us recognize unity and commonality within
the astonishing diversity of mammalian life. The developmental process is uni-
fied by a common sequence of early developmental niches. Whether the spe-
cies produces young that differentiate into aquatic, arboreal, terrestrial, or
subterranean forms, and regardless of whether it subsists as a carnivore, herbi-
vore, insectivore, or omnivore, all are utterly similar in their developmental
origins. These features are universal among, and unique to, mammals.

Ontogenetic Niches of the Norway Rat

For this part of our discussion, we will focus on one species—the Norway
rat (Rattus norvegicus)—to exemplify the basic principles under consideration.
Like some other mammals, Norway rats reside all over the world in a wide
range of habitats, where they can subsist on a variety of foods.
After a gestation period of just 22 days, females give birth to a litter of 10 or
11 pups. This rat is an altricial species, meaning that it produces immature
offspring, which is typical of most mammals. The young are born furless, with
sealed eyes and ears, quite incapable of standing on four feet. For at least three
weeks, they receive intense maternal care in a specially prepared nest, and after
about one month they are prepared for weaning to more independent life in
the colony.
Figure 9.3 depicts four ontogenetic niches in the early life of Norway rats.
Each image is intended to convey the relations between an organism and its
immediate environment (habitat). Importantly, the sequence of images is
meant to convey how organismal development can be seen as a sequence of
adaptations to a sequence of niches. The sequence shown in Figure 9.3 also
characterizes a general pattern for mammalian development.
Figure 9.3a features a fetus in utero. In this case it is a rat fetus, but for all
mammalian fetuses the uterine habitat is remarkably similar. The uterus is
an aquatic environment. For all mammals, behavioral life begins in a fluid-
filled amniotic sac. The fetus eats, breathes, and excretes via the umbilical
cord, a unique adaptation to uterine life. Fetal behavior is also a vital aspect
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 301

Figure 9.3. Four ontogenetic niches of Rattus norvegicus. The Norway rat begins life in
(a) a uterine habitat where breathing and feeding are via the placenta and umbilical cord.
Like all other mammals, a next habitat is (b), the exterior of the mother’s body, where it
can find a nipple and obtain mother’s milk. The mother gradually spends more time
away from the nest and young and (c) the huddle of littermates increasingly becomes
habitat. The drawing in the lower right depicts (d) a social group, which next becomes a
predominant habitat for the weanling pup. Each ontogenetic niche shown here requires
a distinct set of adaptive specializations for successful life in it.

of adaptation to the uterine environment. Fetal behavior includes limb move-

ments, active orientation in relation to gravity, “breathing” movements that
shift amniotic fluid in and out of the lungs, swallowing, and state changes
between sleep and activity.
Figure 9.3b shows a mother rat nursing a litter of infants. This image rep-
resents the postnatal habitat for the newborn mammal—the outer surface of
the mother’s body. Of course, this is where the mammalian mother’s nipples
are located, and her nipples signify one of the great, universal features of
302 ANIMAL BEHAVIOR

mammalian life—the provision of mother’s milk by lactation and the behav-

ioral interactions that support the milk transfer.
Figure 9.3c depicts a clump or huddle of rat pups, which represents a new
niche for the developing rat. In the first days after giving birth, the mother
rat spends nearly all of her time in the nest and in contact with the newborns.
She then starts to withdraw her maternal behavior, very gradually at first, but
as she does the pups begin to interact with each other, and these behavioral
interactions can be intricate (Cramer et al., 1990). The mother returns to
the nest periodically and provides milk, warmth, licking, and many forms of
maternal stimulation (Rosenblatt & Lehrman, 1963). But, as young mammals
develop, they soon begin to reside in other, nonmaternal habitats—Figure 9.3c
represents this kind of early nonmaternal niche.
Finally, Figure 9.3d shows a few weaning-age rat pups playing with a bit of
food, connoting their entrance into a habitat outside of the nest and the early
formation of a social niche that is part of their life in the larger world of the
colony. Numerous, species-typical adaptive features enable the weanling to
move from mother’s milk to the larger and more complex world of foraging
or hunting for food. A variety of sensory, navigational, learning, and memory
mechanisms can be involved. In addition, this is the niche in which social rela-
tions including sexual relations are usually formed, and this of course is the
threshold of reproductive behavior and the evolutionary fulcrum of reproduc-
tive success (see Chapter 4). Again, there are many differences in the details of
this niche in different mammalian species, but the general stage is well repre-
sented. Figure 9.3 shows only four habitats. Many others could be added,
but the goal was to illustrate a perspective—a sequence of adaptations to a
sequence of habitats.

ONTOGENETIC NICHES THAT DEFINE AVIAN DEVELOPMENT

All birds begin life as embryos, enclosed in a rigid shell—their first develop-
mental niche. The location of the egg and conditions under which incubation
and hatching take place vary tremendously among species and orders. Most
birds are incubated by a parent or parents, in a nest, hole, burrow or nest site
constructed or at least occupied and defended by their parent(s). In contrast to
mammals, about 75 to 81 percent of birds exhibit biparental care, with both
parents and sometimes nonparent “helpers” (in 9% of species) investing in
incubation or rearing (Cockburn, 2006). This is the seemingly natural
arrangement that most of you are likely to know.
Some species of birds, however, hatch in nests that do not belong to their
parent (i.e., in the case of intraspecific, or within-species, brood parasitism,
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 303

as occurs in many ducks) or even a bird not of their own species (in the case of
heterospecific, or cross-species, brood parasitism, as occurs, e.g., in cuckoos
and cowbirds; see Chapter 10). Cases of heterospecific parasitism provide
some especially dramatic examples of ontogenetic adaptations in which the
hatchlings of one species display adaptation—necessarily—to the constraints
and conditions imposed by another, host species’ niche. To earn a living suc-
cessfully, the young parasite must manage not just to survive but also to grow
and thrive under conditions set by another species’ evolutionary history.
The story begins with the egg itself. Although characteristics of the egg are
not properly part of the young (being determined by maternal genotype, diet,
etc.), they are part of the package necessary for success. In cases where hosts
exhibit egg rejection behavior (destroying or ejecting foreign eggs), the parasite’s
egg must camouflage, in size and coloration, with the eggs of the host. As an
additional defense against rejection, adult brown-headed cowbirds (Molothrus
ater) have been observed to bully hosts into accepting their eggs and to even pun-
ish hosts severely for egg rejection—destroying the host’s eggs, ensuring that the
genotypes underlying their clever (or insubordinate) behavior are not passed on
(Hoover & Robinson, 2007). Similar retaliatory ‘mafia-like’ behavior has been
observed in at least one species of cuckoo (Soler et al., 1995).
Once the egg is accepted, incubated, and hatched, many parasites
exhibit gape coloration and begging calls that resemble those of host species’
young—often with characteristics that present the parents with especially attrac-
tive cues, ensuring that parasites receive more than their share of food and paren-
tal care. For example, in one study of indigo bunting (Passerina cyanea) nests
parasitized by cowbirds, provisioning rate was higher in parasitized than nonpar-
asitized nests, but young buntings from parasitized nests were nonetheless smaller
than those from nonparasitized nests (e.g., Dearborn et al., 1998).
Some species of cuckoo, such as the European cuckoo (Cuculus canorus), addi-
tionally, actively eliminate competitors upon hatch—a behavior first described
by Edward Jenner, of smallpox vaccine fame. These cuckoo chicks often hatch
earlier than the eggs of their host. Within hours of hatching, these cuckoo chicks
display egg eviction behavior, consisting of a backward extension of the feather-
less wings, which effectively cups an unhatched egg of the host in an unusual
but species-typical depression behind the neck. The cuckoo chick then thrusts
its legs, backing up to the edge of the nest, and the host’s own egg is “evicted”
from the nest! Initially, this apparently reflexive behavior is expressed most vigo-
rously when the cuckoo chick senses an object on its back or behind the wings.
As you might anticipate, the behavior subsides and is not displayed by older birds
(Jenner, 1788; Gill, 1983). You can witness some examples of this extraordinary
behavior at http://youtu.be/lgIL8dbSMUA and http://youtu.be/RDettPoRuNU.
304 ANIMAL BEHAVIOR

Principles of adaptation to the posthatching niche also apply under more typ-
ical, nonparasitic circumstances. Hatchlings must be adapted to, and adapt to,
the niche they encounter (see Box 9.2 to remind yourself about the uses of the
term adaptation), which may or may not include a nest, broodmates, or even
parental care. In the extreme case of “super-precocial” species, such as mega-
podes (e.g., the Australian brush turkey, Alectura lathami, and the maleo, Macro-
cephalon maleo), young hatch as miniature adults. These young display
remarkable completeness and even fly in some cases within a short time after
hatching—without ever having been incubated by an adult (Starck & Ricklefs,
1998). Other, more typically precocial birds (e.g., fowl, ducks, quail) hatch with
precocious sensory and motor systems and can locomote within a short time after
hatching but still rely on parents for limited care and supervision for several days
or weeks after hatching. In highly altricial species, as in most song birds, includ-
ing the brood parasites described above, young hatch in a highly vulnerable,
immature state and require a great deal of further parental investment before they
are capable of making a living as independent adults (Starck & Ricklefs, 1998).
There are numerous factors that influence the nature of the postfledging
niche, including whether the species is capable of flight or swimming, the
nature and presence of predators, climate, whether or not the species is migra-
tory or exhibits complex vocal behavior, and the predominant mating system,
as well as the size, complexity, and durability of social groups and other
factors. Although the general sequence of embryo-in-an-egg ! hatchling !
independent, breeding adult can be said to be universal in birds, the rate of
development and progression through these niches, as well as the specific
nature of the final niche(s) inhabited by the adult, vary considerably. Each of
the ontogenetic niches traversed by a given species of bird—and often the
transitions between them—involve distinct ontogenetic adaptations.

ADAPTATIONS OF EARLY LIFE

Now we will turn to some topics and examples to help illustrate the ideas
that we have been discussing in this chapter. We begin with the transition
from pre- to postnatal life, for this represents a major niche shift for the off-
spring, and thus it is a time for discovering ontogenetic adaptations. First we
will consider hatching, and then we will discuss mammalian birth.

Adaptations Surrounding Hatching

Hatching involves a combination of morphological and behavioral special-
izations—a hatching muscle (M. complexus) and egg tooth—as well as sensory
receptors in the neck and a coordinated pattern of motor movements, all of
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 305

which are adaptations to the challenge of emerging from the confines of a rigid
shell. Hatching is a distinct challenge for an embryo because its growth in the
days leading up to hatching completely fills the space within the egg, necessi-
tating a specific prehatch posture (e.g., head tucked under wing), with few
degrees of freedom to execute the movements vital for successful hatching.
For most species, the hatching muscle, located on the back of the neck, grows
rapidly in size, swelling with lymph in the days prior to hatch. The greatly
swollen muscle functions as a cushion and brace as well as a part of the muscu-
lature used for “pipping” (creating an initial break or indentation in the egg-
shell) and chipping out of the egg (Fisher, 1958).
The egg tooth and the hatching muscle together provide the means of pen-
etrating the egg membranes and eggshell. In some species the egg tooth is used
to rupture and cut or tear the egg, but in others the egg tooth appears to func-
tion as a guide, sliding along the inner surface of the shell as the embryo
rotates within the egg, keeping the beak positioned so that it can repeatedly
exert pressure on the same location(s), progressively weakening the shell
(Brooks & Garrett, 1970). Examples of hatching behavior can be viewed at
http://youtu.be/QY6me0BqZMQ and http://youtu.be/aFi7KxHJu-0. To use
its egg tooth, an embryo must process sensory information (tactile, vestibular,
proprioception) and combine these perceptions with a series of coordinated,
full-body contractions involving the hatching muscle (Hamburger & Oppen-
heim, 1967) as well as pressure-sensitive receptors in the back of the neck. By
combining perception, orientation, and implementation of its specialized
anatomy, embryonic behavior becomes hatching behavior and the egg is
pipped open and cracked.
How does the chick escape the confines of its cracked egg? Clearly, the simulta-
neous pushing of both legs and coordinated lifting of the head to force open
the egg and free the legs, wings, and head is an important tactic. Note that
once the chick has hatched, such leg movements disappear and are replaced
by alternating leg movements. In other words, hatching involves one organiza-
tion of leg movements, while walking involves a distinctly different organiza-
tion. In precocious, ground-nesting species, such as most ducks and fowl, for
example, the switch from one organization takes place right after hatching
and lasts for the rest of the animal’s life. How is this switch accomplished?
Anne Bekoff and colleagues discovered that the switch from synchronous to
alternating leg movements is determined by nerve endings in the chick’s neck!
In the prehatching posture mentioned earlier, the chick’s neck muscles are
stretched when its head is folded under a wing, and this stretching is eased
when the neck is straightened as the head comes free of the closed shell. Bend-
ing the chick’s neck, by placing a chick into an artificial glass egg, reinstates the
306 ANIMAL BEHAVIOR

synchronous hatching movements (Bekoff & Kauer, 1982), and injection of

an anesthetic into the skin of the neck inhibits such reinstatement (Bekoff &
Sabichi, 1987).
The timing of hatching also presents some important social components
with fitness benefits. Although there is an extensive literature on the advan-
tages of hatching asynchrony in altricial species (e.g., production of pheno-
typic variability and reduction of brood size in times of food shortage;
Magrath, 1990), many species of precocial birds and reptiles derive survival
benefits from synchronous hatching (e.g., dilution of predation risk). Conse-
quently, ontogenetic adaptations promoting hatching synchrony have been
identified in several species. Quail and chicken embryos, for example, produce
a clicking sound during the final 24 to 36 hours prior to hatching, which
appears to synchronize the hatching of the clutch (Vince, 1964). In a series
of studies, Margaret Vince found that hatching could be accelerated or delayed
by up to 24 hours depending upon the stimulation an embryo received from
other embryos in the clutch. The benefits of this adaptation are particularly
evident in ducks, which leave the nest as a group shortly after hatching. A
straggler has little chance of survival without its mother and brood: 40 to
85 percent of ducklings (depending on species) die, due to predation and
other factors, even under good circumstances (e.g., Ball et al., 1975).
Box 9.3 describes a case study of the red-eyed tree frog (Agalychnis calli-
dryas), which illustrates other important aspects of the timing of hatching.
The embryos and tadpoles of this species provide a view of how ontogenetic
adaptations associated with hatching can fit into the overall behavioral ecology
of a species—and that of its predators.

Box 9.3
Niche shift in red-eyed tree frog development

Life can be doubly treacherous for the red-eyed tree frog (Agalychnis calli-
dryas). Inhabiting wetlands from the Yucatán to Panama, this amphibian
begins its behavioral life in an egg, which is part of a large clutch laid in a
gelatinous mass on vegetation overhanging temporary ponds. The embryos
are capable of hatching just five days after fertilization, by which time they
are already recognizable as tadpoles, with functioning eyes, well-developed
gill circulation, muscular responses, and opercular development. Left undis-
turbed, however, they will hatch later, by day 10, when they are much larger
and have developed their external mouthparts. Upon hatching the tadpole
immediately drops into the water. This represents a major niche shift from
embryonic life in an egg habitat to a tadpole’s aquatic life in a pond.
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 307

What makes life treacherous for the red-eyed tree frog? While living in its
egg on a leaf, the embryo is sought by a variety of aerial and terrestrial pred-
ators, including the cat-eyed snake (Leptodeira septentrionalis). If this were
not tough enough, as a tadpole, A. callidryasis is attacked by predators in
the water, such as freshwater shrimp, Macrobrachium americanum, that
spends most of their time on the bottom of ponds, a microhabitat in which
their attacks are most effective.
Karen Warkentin found that these tadpoles exhibit a behavioral strategy
that reflects a trade-off between the two types of predation risk. Over several
years, she studied more than 1,000 egg clutches in Costa Rica. When a
clutch of eggs was attacked by a snake, the young instantly shifted into
hatching behavior, quickly emerged, and dropped into the water. If not
exposed to such attacks, hatching is delayed and the young emerge as larger,
more differentiated tadpoles (Warkentin, 1995). There are measurable fit-
ness advantages to the later hatching, when it is possible, for the larger tad-
poles are better equipped to avoid predation from the shrimp waiting for
them on the bottom of the pond. Indeed, the older hatchlings avoid the
bottom more and are more effective in evasive fleeing from aquatic preda-
tors (Warkentin, 1999). These investigations included field experiments
showing that the embryos respond specifically to predatory attacks but not
mere mechanical stimuli, such as movements or vibrations.
The general principle illustrated by these field studies is that growth and
mortality, the major fitness components of prereproductive life, increase
with size in A. callidryasis. Such fitness components can also vary with hab-
itat. Warkentin (1995, 1999) showed a detailed set of relations in which
timing of hatching behavior and shifting to a new niche was determined
by the embryos’ perception of immediate predation activity. The tadpoles’
niche shift was accompanied by regulated repertoire of age-related behav-
ioral strategies that improved their chances of survival in their new, aquatic
niche.

Adaptations Surrounding Mammalian Birth

The process of mammalian birth presents numerous mechanical, physio-
logical, and behavioral challenges. As we have noted, mammalian birth marks
a major niche shift for offspring. It is a transition from an adapted state of fetal
life, comprising features such as umbilical feeding and respiration, to an
adapted state of neonatal life with the onsets of suckling and pulmonary
breathing. Fetal behavior is an essential component to such development.
Without regular, vigorous fetal behaviors such as limb movements, respiratory
308 ANIMAL BEHAVIOR

movements, and swallowing, functional joints cannot develop, and lung and
gastrointestinal development are compromised (Drachman & Coulombre,
1962; Moessinger, 1983; Alberts & Ronca, 1993).
In most mammals, it is mechanically necessary for the fetus to pass head-
first through the birth canal. As we have already noted, fetuses move through-
out gestation and consequently are usually oriented randomly within the
womb. As birth approaches, it is essential for a fetus to actively orient itself
for the journey to the outside world. In humans and other primates, this
involves adoption of a head-downward posture. Such active positioning
depends on integrated sensory and motor actions by the fetus. Fetal behavior
is an often-overlooked aspect of early development, but it is important to
many aspects of postnatal life and to the birth process itself.
The rich and complex behavioral repertoire of mammals is enabled by the
evolution of large, experience-sensitive brains. Not surprisingly, this is accom-
panied by relatively large heads, which themselves create a mechanical chal-
lenge for birth. Passage of a large human head through the birth canal of a
human female is made possible by the flexibility provided by the incomplete
bone and skull development. Figure 9.4 shows the state and structure of the
infant human skull. Note the incomplete bone formation of the developing
skull. The separate plates will eventually come together and seal as the bone
tissue itself hardens.
In its immature state, however, the skull is very flexible, and the baby’s
head can be squeezed to pass through the birth canal. Is this kind of fit an adap-
tation? With reference to the kinds of uses of the term shown earlier in
Box 9.2, it seems clear that a soft, malleable skull meets criteria 1 and 3 because
a soft skull serves a proximate function (1) and contributes to the overall match
of the neonate to its shifting niches (3). But—and this is a very important
distinction—we cannot make a good case that the malleable skull is an adapta-
tion that arose from natural selection for this purpose because immature bird
skulls also show the same soft and malleable characteristics, and bird heads
are protected by a hard shell. All known fossil evidence indicates that mam-
mals and birds share some common, distant ancestors, organisms that grew
their skulls according to the process we now see in both birds and mammals.
This tells us, then, that the soft, separate plates of the immature skull existed
historically before there were mammals and that mammalian development
took advantage of—co-opted—the preexisting characteristic and made use of
it in the process of mammalian birth, but natural selection did not produce
the characteristic. In this way, the malleable skull of a neonatal mammal does
not meet criterion 4 for an adaptation because it did not arise historically for
its current function. Evolutionary biologists sometimes use the term
Figure 9.4. Two views of the skull of a newborn human. Note that the cranial plates
are incompletely formed. They join at “sutures,” which remain incomplete at birth,
leaving a diamond-shaped space or “soft spot” at the top of the skull and another at
the back. The sutures are part of the flexibility of the skull, which can safely comply to
the pressures of squeezing through the birth canal.
310 ANIMAL BEHAVIOR

exaptation (ex ¼ by reason of + aptus ¼ fit) to recognize the proximate utility

of a feature that did not arise for its current function (Gould & Vrba, 1982).

ADAPTIVE SPECIALIZATIONS FOR EARLY FEEDING

Tadpole Feeding
If tadpole feeding were just juvenile frog feeding, we might expect to find
that tadpoles, for instance, snap at tiny insects or other specks of food and gulp
them down because their adult counterpart frogs are known to snap at insects
and gulp them down. In fact, this simplistic view would ignore the species of
frogs that are carnivorous and miss other species of insect-eating frogs that cap-
ture their prey by shooting from their mouth a specialized, sticky tongue of
remarkable length, which they can aim at a hapless fly and use to yank the
fly out of the air and into their mouth and then gulp it down. Across the meta-
morphic species, whether they nab their prey by hopping after it, leaping upon
it, snapping, or shooting out a sticky tongue, their premetamorphic state is a
tadpole that feeds like a tadpole.
Tadpoles are specialized to be suspension feeders (Griffiths, 1961; Wasser-
sug, 1972). Suspension feeding involves taking up from water food particles
that are too small to be seen or even sensed individually (Jørgensen, 1966).
Even as suspension feeders, tadpoles of different species exhibit species-
typical and ecologically matched specializations. For example, tadpoles of
some species only eat plankton. These obligate planktivores are largely
restricted to phytoplankton (Wassersug, 1975). Other tadpoles feed with spe-
cialized, keratinized mouthparts that enable them to handle mixtures of algae,
microbes, bacteria, and other materials. Basically, tadpoles are adapted to
extract fine particulate matter from suspension (Kenny, 1969).
Tadpole feeding thus involves behavior and associated mouth structures
that actively pump water and suspended materials over appropriate surfaces.
Tadpoles accomplish this with a specialized “buccal pump,” which is a kind
of biological piston. The tadpole mouth and skull are equipped with muscles
and cartilage that enable them to move the floor of the mouth up and down,
creating the pump. Other portions of the tadpole mouth form a valve that pro-
duces one-way flow that directs water over filter-like structures and eventually
into the gut. Indeed, the specialized buccal pumping behavior and associated
anatomy of the tadpole separates this group from all other amphibian larvae
(Gradwell & Pasztor, 1968; Wassersug, 1975).
Once the suspended particles are captured and ingested, tadpoles utilize
their distinctly long and slender gut with its large surface area and small vol-
ume, which is shaped and proportioned for efficient chemical, as opposed to
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 311

mechanical, breakdown of the cell walls of the algae and plankton (Strath-
mann, 1971).
Tadpoles display patterns of utilization that are seen in invertebrate suspen-
sion feeders. When food is not abundant in the environment, they retain par-
ticles in their gut and process the material more thoroughly. When food is
abundant, no effort is made to digest it thoroughly, and it is allowed to pass
rapidly through the tadpole gut. Overall, the rate of moving water and partic-
ulates by suspension feeders depends on the concentration of suspended mat-
ter (Wassersug, 1975).

Early Adaptations in Avian Feeding

A number of ontogenetic adaptations underlie the success of early feeding
in altricial birds. Begging behaviors, for example, provide hatchlings a means
both of signaling parents that they are ready for food and of competing with
siblings for limited resources. You can view numerous examples of young
birds’ begging behavior at sites such as http://youtu.be/ZEaoUEMM1CE.
Begging behaviors reoccur later in life in many species as part of courtship rit-
uals or while the female is on the nest. Although there are instances of elabo-
rate exchanges of food midflight, in most cases females adopt behaviors
similar to those displayed by young birds (e.g., wing fluttering) and males
place food directly in females’ mouths (Lack, 1940). Interestingly, begging-
like behavior can also be induced in adult pigeons (Columba livia) in the lab
if a noncontingent food reward is delivered from a wall rather than the center
of an enclosure (Timberlake & Lucas, 1985).
Pecking behavior, often directed at the parents’ bill, is also critical for pro-
curing food from the parents in many species. Tinbergen famously explored
the range of stimuli that can elicit pecking in herring gull (Larus argentatus)
chicks, finding that movement of the parents’ beaks and the mew call emitted
by parents were critical releasers of chick pecking. In addition, the bright red
spot on parents’ bills provided a critical “target” for chicks to peck at, signifi-
cantly enhancing pecking (Tinbergen & Perdeck, 1950). Although Tinbergen
found that placing the spot someplace else on a model (e.g., on the forehead)
reduced responsivity in chicks, Jack Hailman (1969) showed that if the height
and movement of the forehead and bill spot were equated, the superiority of
the latter disappeared. Hailman moreover suggested that the clumsiness of
chicks’ initial pecks could be viewed as adaptive, as chicks invariably miss the
parents’ bill on occasion, instead striking the food the parent is grasping,
whereupon they begin to feed.
In some birds, particularly pigeons and doves, young are fed from a secreted
milk-like substance (crop milk), regurgitated through a process similar to
312 ANIMAL BEHAVIOR

normal vomiting. As documented by Daniel Lehrman (1955), prolactin secre-

tion by the pituitary gland—which can be triggered in response to stimulation
during incubation or even by the mere presence of squabs—causes the over-
proliferation of cells in the epithelium of the crop sac of the pigeon. The
sloughing off of these cells is the basis for crop milk. The mechanical stimula-
tion of the parents’ breast (above the crop) by the squab plays a key role in
inducing the regurgitation of crop milk (Lehrman, 1955).

Adaptations for Early Ingestion by Mammals

Just the name mammals tells us a lot. Among the features that are shared by
all mammals are mammary glands in the females—and in both sexes there is
endothermy and body hair. We have already noted the wide variety of species
in this class and recognized that they inhabit a wide variety of niches—aquatic,
terrestrial, subterranean, and arboreal, in many different climates and habitats.
Remarkably, the females in every species in this diverse class possess mammary
glands—anatomically and endocrinologically specialized structures that pro-
duce and secrete milk, which is itself a remarkable substance. Mother’s milk
shows species-specific adaptive relations to the species-typical niches, particu-
larly in terms of fat content and quantities produced (Oftedal, 1984, 1993).
There are important behaviors that go along with the presence of the mam-
mary gland and the hormones associated with it. Maternal behavior and mam-
mary glands are functionally united. Thus we see that lactation, the
production and transfer of mother’s milk, involves an integration of anatomy,
physiology, and behavior.
Mammalian mothers lactate, and mammalian infants suckle. We will focus
on the infants’ side of the relationship and examine it for examples of ontoge-
netic adaptation. Suckling behavior involves orally grasping a nipple, exerting
mechanical pressure to it, and, when milk becomes available through the teat,
transferring it from nipple to mouth and then swallowing it (Blass & Teicher,
1980).
How is it that a newborn mammal, an organism that has never encountered a
nipple, is capable of responding to the presence of this novel stimulus with the
vitally adaptive behavior of nipple attachment and suckling? To many observers,
this has been clear example of a kind of innate or “hard-wired” behavior. In
contrast to this seemingly clear logic, we have learned from research with a
variety of species that the newborn mammal is not naïve about suckling and
that there are a host of precursors—experienced, practiced, and learned in
the womb—that contribute to the developmental assembly of a complex,
adaptive behavior that is present at birth (Smotherman & Robinson, 1988).
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 313

For the newborn of many species, odor cues are essential for getting to the
nipple and for the first suckle. If the odor is removed from the mother’s body
or if the infant has impaired olfactory function, suckling may never begin
(Teicher & Blass, 1976)! The key odor cue for suckling is present in the amni-
otic fluid, so the mammalian fetus is exposed to it during gestation and
throughout the birth process. Experimenters have changed the odor of amni-
otic fluid by adding distinct flavors to the mother’s diet or by injecting small
amounts of a novel odorant into the amniotic fluid toward the end of gesta-
tion. Newborns exposed to a novel odor in amniotic fluid attach to nipples
scented with the experimental odor and may even not respond to the natural
cue! These results were a good indication that the fetus’ experience in utero
is part of an important learning process through which the newborn acquires
its odor-guided responses to the mother’s body (Pedersen & Blass, 1982;
Alberts & Ronca, 2012).
Other evidence indicates that the learned olfactory cue generally activates
the infant and that the enhanced activity includes head movements and prob-
ing and nuzzling with the facial areas (Pedersen & Blass, 1982). At this point,
the newborn’s sensitive skin, especially around the mouth (the so-called perio-
ral area), which is richly innervated by early-developing nerves, triggers a sim-
ple rooting reflex when the tactile receptors encounter a protruding nipple.
The infant opens its mouth, turns toward the side of stimulation, and grasps
the nipple.
Mammalian infants have a complete secondary palate in the roof of the
mouth, separating the nasal airway and the oral cavity, so that an infant’s
mouth can be dedicated to holding a nipple yet breathe nasally. Other ani-
mals, specifically nonmammalian predators, have such structures, and these
have been retained by mammalian infants, providing vital adaptation to suck-
ling and breathing (cf. Oftedal, 2002; Alberts & Pickler, 2012).

ONTOGENETIC ADAPTATIONS IN REGULATION AND LEARNING

Compared to humans, infant rats and mice are born in a highly immature
state. Blind, deaf, furless, and lacking both subcutaneous fat and the capacity
to warm themselves through shivering, researchers long characterized them
as ectotherms—organisms dependent upon external sources of heat, like so-
called cold-blooded organisms such as reptiles and amphibians (e.g., Pembrey,
1897). The early postnatal niche for infant rats and mice nonetheless contains
a number of features that render this characterization incomplete. Their post-
natal niche contains a nest, a mother that builds and maintains the nest and
cares for the pups, as well as, critically, littermates.
314 ANIMAL BEHAVIOR

Situated among littermates, and thus behaving as a group or huddle, infant

rats and mice display sophisticated regulatory capabilities; that is, they form a
dynamically writhing mass, and as individual pups cool they dive to its center,
forcing warmer pups to surface, thus resulting in behavior that resembles the pro-
cess of convection. Huddles also expand and contract horizontally, adjusting the
size of the huddles in concert with changes in air temperature (Alberts, 1978).
Employing simultaneous measures of behavior and infrared thermography, we
demonstrated that mouse pups derive significant thermoregulatory benefit when
huddling in response to cold (Harshaw & Alberts, 2012). Pups thus display
sophisticated “group regulatory behavior” when studied in their natal niche that
is unobservable if they are removed from that niche and studied in isolation.
Like many infant mammals, infant mice and rats also have large deposits of
brown adipose tissue (“brown fat,” or BAT), particularly around the neck and
between the shoulder blades. Originally considered a “hibernating gland”
because the size of BAT deposits was observed to increase and decrease season-
ally in hibernating mammals, BAT was long thought to serve a nutritive func-
tion for hibernators. It was nonetheless eventually found to serve primarily
thermogenic function, helping hibernators maintain core body temperature.
Brown fat provides a particularly interesting example of adaptation, as its func-
tion in hibernators may represent an exaptation (evolutionary co-opting) of a
system originally evolved to aid infants in coping with the challenges of their
early developmental niche—in particular, with the challenges of small size,
large surface-to-mass ratio, furlessness, immature capacity for shivering
thermogenesis, and consequent proneness to rapid heat loss.
Brown fat has been characterized as an ontogenetic adaptation to early post-
natal life (Blumberg, 2001), being the sole source of endogenous (internal,
self-generated) heat production available to young mammals. When the
mother leaves or pups are isolated from the nest, brown fat becomes active,
preventing pups from sustaining a debilitating loss of body heat.
Thermotactic responses thus combine with endogenous heat production to
produce sophisticated group regulatory behavior in young rodents. Indeed,
agent-based models of huddling in rats have shown that prior to 10 days of
age, models based entirely on a few simple rules of individual behavior can
account for patterns of contact displayed by groups of infant rats (Schank &
Alberts, 1997). There is some evidence that the thermotactic responses of young
rodents may develop asymmetrically, in that specific responses to heat may
emerge ontogenetically prior to specific responses to cold. For example, 1-, 5-,
and 11-day-old rats were trained to perform an operant head-turning to receive
either warmth in a cool environment or cooling in a hot environment. Pups of
all ages were able to learn to turn their heads to receive warmth, whereas only
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 315

5- and 11-day-old pups learned to turn their heads to cool off in a hot environ-
ment. These results suggest that asymmetry of thermotaxis may be an ontoge-
netic adaptation in day-old rats, as the potential benefits of avoiding
overheating normally do not outweigh the risks of rapid heat loss due to cooling
for these young pups (Hoffman et al., 1999).

ONTOGENETIC ADAPTATIONS IN PARENT-OFFSPRING RELATIONS

For the young of species exhibiting parental care, the absence of such care,
particularly at very young ages, entails almost certain death. As Konrad Lorenz
(1937) and John Bowlby (1969) recognized, systems that promote and main-
tain proximity to caregivers are thus critical in such species. Although Lorenz
and Bowlby characterized such systems as highly specialized or modularized
adaptations, they can also be viewed as systems properties (that is, not residing
wholly in the neonate)—the product of selection for an entire developmental
system, containing numerous recurrent genetic and extragenetic elements
(Lickliter & Harshaw, 2010).
Importantly, the exigencies of the ontogenetic niche persist even when care-
givers are inexperienced, subpar, or abusive. That is, a negligent or abusive
caregiver is often preferable to an almost certain death given the complete
absence of care (Bowlby, 1969). Resistance to the acquisition of aversions to
caregivers is thus evident in a great many species, including humans, as well
as many species of monkeys and birds. Attachment or “imprinting” systems
and resistance to the acquisition of aversions to caregivers can thus be charac-
terized as interrelated ontogenetic adaptations to early postnatal life.
For species lacking well-developed motor control at birth (e.g., rats and
mice) the maintenance of proximity is largely under control of the dam, which
will quickly retrieve pups if they wander. Hatchlings of precocial birds (e.g.,
ducks, fowl, quail), in contrast, can locomote freely shortly after hatching
and thus occupy a niche with quite different requirements. Without a system
to keep them close to their parent(s), hatchlings could easily wander and find
themselves in danger or in the jaws of a predator. Filial imprinting, the pro-
cess by which avian hatchlings form preferences for parents or surrogates (and
come to fear other social stimuli), can thus be viewed as an ontogenetic adap-
tation to these circumstances (Bolhuis, 1999).
Filial imprinting reveals a profound and perhaps surprising lack of species
identification (the sense of who is and who is not a member of their own spe-
cies); as hatchlings, many birds will readily form an attachment to the first
conspicuous object they encounter upon hatching, whether it be their mother,
a human, or an object such as a rotating red box (Lorenz, 1937). At first
316 ANIMAL BEHAVIOR

glance, such a system may appear dangerously risky—a solution that evolution
ought to have avoided. A mother is nonetheless a normally occurring and nec-
essary feature of the developmental system for such species (i.e., the chicken
and egg), and embryos are unlikely to hatch and even less likely to survive in
the absence of a broody adult (Lickliter & Harshaw, 2010). Species that
have evolved developmental systems not requiring parental care, such as
the Australian brush turkey—which hatch in piles of vegetation, warmed by
the decay of plant matter, and live fairly solitary lives prior to breeding—are
exceptions. In all other cases, hatchlings do best to form an attachment to
whatever adult happens to be present, even if a member of an entirely different
species (many ducks, for example, practice both intra- and interspecific
parasitism—laying some portion of their eggs in the nests of other birds).
Hatchlings moreover do well to maintain such attachments, even in the
face of suboptimal care or abuse. Exposing chicks to aversive stimulation in
the presence of an imprinting object—for example, sounding a loud clapper
whenever a chick closely approaches the object, or shocking chicks in the pres-
ence of the object—results in an increased preference for the imprinting stimu-
lus (e.g., Hess, 1964)! A number of studies have also shown that the following
of an imprinted object is enhanced by increasing the arousal levels of hatch-
lings, either by associating the following context with shock prior to testing
or by administering adrenergic agonists. Although the mechanisms underlying
this ontogenetic adaptation in birds are relatively poorly understood relative to
rodents (discussed below), its consequences can be observed in the wild and in
certain cases can lead to the death of offspring. For example, in coots (e.g.,
Fulica atra) and moorhens (e.g., Gallinula chloropus), parents are frequently
aggressive with chicks, tousling them—grasping the chick’s head, picking it
up, and shaking it vigorously—a behavior that can have fatal consequences
for chicks (Horsfall, 1984; Leonard et al., 1988).
Similar adaptations against acquiring aversions to caregivers have been doc-
umented in dogs (Canis lupus familiaris) and nonhuman primates, but the
phenomenon has been most intensively studied in rodents, which are gener-
ally altricial and far more dependent on the mother than are precocial birds
and their parents. The functions of this adaptation are clearer in the ontoge-
netic niche occupied by the newborn rodent, in that not all stimulation from
mother is guaranteed to be positive or pleasant (Hofer & Sullivan, 2001).
For example, a mother may step on pups as she enters and exits the nest or
when she adjusts her posture while on the nest. She may bite or even nick
pups accidentally as she handles or transports them. Many such experiences
would have the potential to induce an aversion to mother were there not an
adaptation against the formation of such aversions. The mechanism
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 317

underlying this adaptation in rats appears to involve high levels of activity of a

central noradrenergic system involving the locus coeruleus, and correspond-
ingly low levels of activity in the amygdala, relative to later stages of life (Mor-
iceau & Sullivan, 2005). It is noteworthy that young rats are capable of
learning negative associations—but they are protected from learning aversive
associations to parental stimuli at ages when they fully depend on the parent
and such learning would interfere with close relations to parental resources
(Alberts & Gubernick, 1984; Alberts, 1994).

THE APPEARANCE AND DISAPPEARANCE (AND REAPPEARANCE) OF INFANTILE BEHAVIOR:

HELPFUL CLUES TO IDENTIFYING ONTOGENETIC ADAPTATION
Many of the examples of ontogenetic adaptation that we have discussed in
this chapter have used the proximate meanings of adaptation (see Box 9.2),
which emphasize the general fit of organism to environment and the immedi-
ate or proximate function of a feature. These are the more general ways to use
the term adaptation, and it is often not clear when an infant’s behavior is really
a complete, specific, functional action that has been shaped by natural selec-
tion versus some incomplete, general precursor to some other behavior. In
the context of understanding an animal’s behavioral biology, we are compelled
to recognize features—behavioral, anatomical, and physiological—that are the
result of evolutionary shaping by natural selection.
There is a class of infantile behaviors that present a surprising developmen-
tal course. Rather than growing stronger, becoming more refined, or expand-
ing in the animal’s repertoire, these behaviors disappear. Earlier in the
chapter, we noted this phenomenon (see “The Directions of Development”),
but here we are recognizing this as a distinct type of developmental trajectory.
As you will see, careful observations and some clever experiments can provide
useful lessons in the appearance, disappearance, and reappearance of some, but
not all, infantile behaviors.

Transport Reflexes
The young of many species are incapable of sustained, general locomotion.
In such species, it is common to see adults carrying infants. Such transport
behavior takes many forms. The human parent can, of course, carry an infant
in its arms. Quadrupeds that are constrained to ambulate on their forelegs
more typically carry their young by orally grasping them. In such species, the
young often display a transport reflex (Brewster & Leon, 1980). In a variety
of species, including rodents and cats, the parent will be seen to grasp the
318 ANIMAL BEHAVIOR

Figure 9.5. Demonstrating the grasp reflex in a human infant. This classic photograph
shows the surprising strength of the grasp reflex in an infant, who is holding its own
weight with one hand.

offspring by the skin behind the neck. The reflex is thus triggered in the off-
spring: limbs are retracted, and the tail may be brought to the ventral midline.
In this position, the kitten facilitates the efficiency of the mother’s behavior. In
studies of how the reflex is elicited, this is the body area from which the reflex
can be triggered most efficiently. Moreover, the reflex appears developmentally
—when the offspring have grown to a size that their limbs will interfere with
the parents’ carrying. Prior to that age, the reflex is not expressed. The decline
of the reflex corresponds to the age when the offspring become independent
and will move on their own (Brewster & Leon, 1980). You can view some exam-
ples of transport reflexes at http://youtu.be/hzw6x8bClR0.
There are other reflexes displayed by infants that may operate similarly. The
grasp reflex is seen in many primate infants, including humans, and is easily elic-
ited by the sensation of hair on the palm, which elicits a strong, reflexive grasp
response. Figure 9.5 shows a classic demonstration of an infant human sustaining
its own weight with this simple reflex. It is believed that this stunning feat by a
human infant is a vestige of a reflexive grasping onto the fur of a primate mother,
who may be running and jumping while carrying her offspring.
The grasp reflex, like many other such reflexes, disappears in human infants
by three to six months of age, as it does in the nonhuman primates that show
such reflexes. In a variety of experimental studies and in some clinical conditions
involving neural damage, these reflexes reappear, suggesting that when they ini-
tially were lost it was through active suppression, or inhibition. In this way, the
reflex circuits remain present in the nervous system but are actively inhibited
by later-developing systems.
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 319

The Rooting Reflex

Lightly stroking the skin near an infant’s mouth can evoke the rooting
reflex, which consists of an orienting head movement turning toward the side
of stimulation while opening the mouth; then, if an object such as a nipple is
encountered, the reflex is completed by an oral grasp of the nipple (or experi-
menter’s finger). The rooting reflex is present early in life, and then it disap-
pears. That is, as newborns progress into infancy, suckling behavior appears
to become increasingly more controlled and less “reflexive.”
Simple responses such as the rooting reflex, the grasp reflex, and a urination
(“micturition”) reflex all diminish and then disappear, meaning that they are
no longer elicited by the stimuli that evoke the responses in early life. The dis-
appearance of an adaptive behavior can be a helpful signal that the original
behavior was a functional entity on which selection pressure has eased because
it is no longer needed or because other behaviors with fitness value may be
compromised if the original behavior is not replaced. Disappearance of a
behavior is not, however, a necessary feature of an ontogenetic adaptation,
but a discrete period of existence almost always highlights the ways in which
it may be a complete, well-formed behavioral entity in its own right.

AVOIDING THE ADAPTATIONIST TRAP

We have been quite liberal in our use of the term adaptation. This was per-
missible because we carefully defined the different uses and attempted to clarify
or specify the type of adaptation we intended to denote each time. There is a risk
in being too liberal in the uses or not defining the intended meaning of adapta-
tion. The risk is to fall into the adaptationist trap, in which everything is adaptive,
or, similarly, functional significance is assumed for every behavior or for each fea-
ture of an organism. We try not to commit these errors. We seek to discover the
biological significance of a behavior during development, and, whether or not it
defines itself by disappearing, our task is to determine its functional role in an
animal’s life. If a strong and compelling role appears, we can seek evidence that
this is a feature for which natural selection has operated.

CONCLUSIONS
We have surveyed a wide range of behavioral phenomena in a diverse set of
animals. Hopefully, we have provided information, analyses, and ideas that
have changed the way you look at development, for this was our goal when
we began the chapter. To us, there is great insight and understanding to be
gained from a developmental perspective of behavior, particularly one that
320 ANIMAL BEHAVIOR

integrates basic biological concepts such as adaptation. Here we have seen that
the behavior of young, immature animals is more than a hodgepodge of simple
reflexes that will mature or coalesce into the full and functional behavior of an
adult. Instead, we have seen that embryos behave adaptively—in ways that fit
to their immediate surroundings—often with specializations lacking in the
adult form. This view showed us evidence that natural selection works on each
stage of the life cycle and helped us to appreciate that components of early life
contribute importantly to fitness in the ultimate, evolutionary sense of adapta-
tion. The multidimensional concept of ontogenetic adaptation can be a help-
ful beacon, illuminating features of developing animals and leading to
rigorous analyses of adaptive fitness of behavior.

ACKNOWLEDGMENTS
The writing of this chapter was supported, in part, by NIMH grant RO1-
MH082019 to JRA.

REFERENCES AND SUGGESTED READING

Adolph, E. F. (1968). Origins of Physiological Regulations. New York: Academic Press.
Alberts, J. R. (1978). Huddling by rat pups: Group behavioral mechanisms of tem-
perature regulation and energy conservation. Journal of Comparative and
Physiological Psychology, 92, 231–245.
Alberts, J. R. (1994). Learning as adaptation of the infant. Acta Paediatrica, 83, 77–85.
Alberts, J. R. & C. P. Cramer. (1988). Ecology and experience: Sources of means and
meaning of developmental change. In E. M. Blass (ed.), Handbook of Behav-
ioral Neurobiology, Vol. 9 (pp. 1–39). New York: Plenum.
Alberts, J. R. & D. J. Gubernick. (1984). Early learning as ontogenetic adaptation for
ingestion by rats. Learning and Motivation, 15, 334–359.
Alberts, J. R. & R. H. Pickler. (2012). Evolution and development of dual ingestion
systems in mammals: Notes on a new thesis and its clinical implications.
International Journal of Pediatrics, 2012. doi:10.1155/2012/730673.
Alberts, J. R. & A. E. Ronca. (1993). Fetal experience revealed by rats: Psychobiolog-
ical insights. Early Human Development, 35, 153–166.
Alberts, J. R. & A. E. Ronca. (2012). The experience of being born: A natural context
for learning to suckle. International Journal of Pediatrics, 2012. doi:10.1155/
2012/129328.
Ball, I. J., D. S. Gilmer, L. M. Cowardin, & J. H. Riechmann. (1975). Survival of
wood duck and mallard broods in north-central Minnesota. Journal of Wildlife
Management, 39, 776–780.
Bekoff, A. & J. A. Kauer. (1982). Neural control of hatching: Role of neck position in
turning on hatching leg movements in post-hatching chicks. Journal of Compa-
rative Physiology, 145, 497–504.
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 321

Bekoff, A. & A. L. Sabichi. (1987). Sensory control of the initiation of hatching in

chicks: Effects of a local anesthetic injected into the neck. Developmental
Psychobiology, 20, 489–495.
Blass, E. M. & M. H. Teicher. (1980). Suckling. Science, 210, 15–22.
Blumberg, M. S. (2001). The developmental context of thermal homeostasis. In
E. Blass (ed.), Handbook of Behavioral Neurobiology, Vol. 13 (pp. 199–228).
New York: Kluwer Academic/Plenum Publishers.
Bolhuis, J. J. (1999). The development of animal behavior: From Lorenz to neural
nets. Naturwissenschaften, 86, 101–111.
Bonner, J. T. (1958). The Evolution of Development. Cambridge, UK: Cambridge
University Press.
Bowlby, J. (1969). Attachment and Loss, Vol. 1. New York: Basic Books.
Brewster, J. & M. Leon. (1980). Facilitation of maternal transport by Norway rat
pups. Journal of Comparative and Physiological Psychology, 94, 80–88.
Brooks, W. S. & S. E. Garrett. (1970). The mechanism of pipping in birds. Auk, 87,
458–466.
Cockburn, A. (2006). Prevalence of different modes of parental care in birds. Proceed-
ings of the Royal Society of London, B, 273, 1375–1383.
Cramer, C. P., Thiels, E., & J. R. Alberts. (1990). Weaning in rats: I. Maternal
behavior. Developmental Psychobiology, 23, 479–493.
Dearborn, D. C., A. D. Anders, F. R. Thompson III, & J. Faaborg. (1998). Effects of
cowbird parasitism on parental provisioning and nestling food acquisition and
growth. Condor,100, 326–334.
Drachman, D. & A. Coulombre. (1962). Experimental clubfoot and arthrogryposis
multiplex congenita. The Lancet, 280, 523–526.
Elinson, R. P. (2001). Direct development: An alternative way to make a frog.
Genesis, 29, 91–95.
Fisher, H. I. (1958). The “hatching muscle” in the chick. Auk, 75, 391–399.
Galef, B. G., Jr. (1981). The ecology of weaning: Parasitism and the achievement of
independence by altricial mammals. In P. Klopfer & D. J. Gubernick (eds.),
Parental Care in Mammals (pp. 211–239). New York: Plenum Press.
Gill, B. J. (1983). Brood-parasitism by the shining cuckoo Chrysococcyx lucidus at
Kaikoura, New Zealand. Ibis, 125, 40–55.
Gottlieb, G. (1968). Prenatal behavior of birds. Quarterly Review of Biology, 43,
148–174
Gottlieb, G. (1971). Development of Species Identification in Birds: An Enquiry into the
Prenatal Determinants of Perception. Chicago: University of Chicago Press.
Gould, S. J. & E. S. Vrba. (1982). Exaptation—a missing term in the science of form.
Paleobiology, 8, 4–15.
Gradwell, N. & V. M. Pasztor. (1968). Hydrostatic pressures during normal ventila-
tion in the bullfrog tadpole. Canadian Journal of Zoology, 46, 1169–1174.
Griffiths, I. (1961). The form and function of the fore-gut in anuran larvae
(Amphibia, Salientia) with particular reference to the manicotto glandulare. Pro-
ceedings of the Zoological Society of London, 137, 249–283.
322 ANIMAL BEHAVIOR

Hailman, J. P. (1969). How an instinct is learned. Scientific American, 221, 98–108.

Hamburger, V. & R. Oppenheim. (1967). Prehatching motility and hatching behav-
ior in the chick. Journal of Experimental Zoology, 166, 171–204.
Harshaw, C. & J. R. Alberts. (2012). Group and individual regulation of physiology
and behavior: A behavioral, thermographic, and acoustic study of mouse devel-
opment. Physiology and Behavior, 106, 670–682.
Hess, E. H. (1964). Imprinting in birds. Science, 146, 1128–1139.
Hofer, M. A. & R. M. Sullivan. (2001). Toward a neurobiology of attachment. In
C. A. Nelson & M. Luciana (eds.), Handbook of Developmental Cognitive
Neuroscience (pp. 599–616). Cambridge, MA: MIT Press.
Hoffman, C. M., G. S. Flory, & J. R. Alberts. (1999). Ontogenetic adaptation and
learning: A developmental constraint in learning for a thermal reinforcer.
Developmental Psychobiology, 34, 73–86.
Hoover, J. P. & S. K. Robinson. (2007). Retaliatory mafia behavior by a parasitic
cowbird favors host acceptance of parasitic eggs. Proceedings of the National
Academy of Sciences, USA, 104, 4479–4483.
Horsfall, J. A. (1984). Brood reduction and brood division in coots. Animal Behav-
iour, 32, 216–225.
Grim, T., J. Rutila, P. Cassey, & M. E. Hauber. (2009). The cost of virulence: An
experimental study of egg eviction by brood parasitic chicks. Behavioral
Ecology, 20, 1138–1146.
Jenner, E. (1788). Observations on the natural history of the cuckoo. Philosophical
Transactions of the Royal Society of London, B, 78, 219–237.
Jørgensen, C. B. (1966). Biology of Suspension Feeding. Oxford, UK: Pergamon Press.
Kenny, J. S. (1969). Feeding mechanisms in anuran larvae. Journal of Zoology, 157,
225–246.
Lack, D. (1940). Courtship feeding in birds. Auk, 57, 169–178.
Lehrman, D. S. (1955). The physiological basis of parental feeding behavior in the
ring dove (Streptopelia risoria). Behaviour, 7, 241–286.
Leonard, M. L., A. G. Horn, & S. F. Eden. (1988). Parent-offspring aggression in
moorhens. Behavioral Ecology and Sociobiology, 23, 265–270.
Lickliter, R. & C. Harshaw. (2010). Canalization and malleability reconsidered: The
developmental basis of phenotypic stability and variability. In K. Hood,
C. Halpern, G. Greenberg, & R. Lerner (eds.), Handbook of Develop-
mental Science, Behavior, and Genetics (pp. 491–525). Chichester, UK:
Wiley-Blackwell.
Lickliter, R. & T. Hellewell. (1992). Contextual determinants of auditory learning
in bobwhite quail embryos and hatchlings. Developmental Psychobiology, 25,
17–31.
Lorenz, K. (1937). The companion in the bird’s world. Auk, 54, 245–273.
Magrath, R. D. (1990). Hatching asynchrony in altricial birds. Biological Reviews, 65,
587–622.
Moessinger, A. C. (1983). Fetal akinesia deformation sequence: An animal model.
Pediatrics, 72, 857–863.
BEHAVIORAL DEVELOPMENT AND ONTOGENETIC ADAPTATION 323

Moriceau, S. & R. M. Sullivan. (2005). Neurobiology of infant attachment. Develop-

mental Psychobiology, 47, 230–242.
Oftedal, O. T. (1984). Milk composition, milk yield and energy output at peak lacta-
tion: A comparative review. Symposia of the Zoological Society of London, 51,
33–85.
Oftedal, O. T. (1993). The adaptation of milk secretion to the constraints of fasting
in bears, seals, and baleen whales. Journal of Dairy Science, 76, 3234–3246.
Oftedal, O. T. (2002). The mammary gland and its origin during synapsid evolution.
Journal of Mammary Gland Biology and Neoplasia, 7, 225–252.
Pedersen, P. E. & E. M. Blass. (1982). Prenatal and postnatal determinants of the 1st
suckling episode in albino rats. Developmental Psychobiology, 15, 349–355.
Pembrey, M. S. (1897). The effect of variations in external temperature upon the out-
put of carbonic acid and the temperature of young animals. Journal of Physiol-
ogy, 18, 363–379.
Rosenblatt, J. S. & D. S. Lehrman (1963). Maternal behavior in the laboratory rat. In
H. L. Rheingold (ed.), Maternal Behavior in Mammals (pp. 8–57). New York:
John Wiley and Sons, Inc.
Schank, J. C. & J. R. Alberts. (1997). Self-organized huddles of rat pups modeled
by simple rules of individual behavior. Journal of Theoretical Biology, 189,
11–25.
Smotherman, W. P. & S. R. Robinson. (1988). The uterus as environment: The
ecology of fetal behavior. In E. M. Blass (ed.), Handbook of Behavioral
Neurobiology, Vol. 9. Developmental Psychobiology and Behavioral Ecology
(pp. 149–196). New York: Plenum Press.
Soler, M., J. J. Soler, J. G. Martinez, & A. P. Møller. (1995). Magpie host manipula-
tion by great spotted cuckoos: Evidence for an avian mafia? Evolution, 49,
770–775.
Starck, J. M. & R. E. Ricklefs. (1998). Patterns of development: The altricial-
precocial spectrum. In J. M. Starck & R. E. Ricklefs (eds.), Avian Growth
and Development: Evolution within the Altricial-Precocial Spectrum (pp. 3–30).
Oxford, UK: Oxford University Press.
Strathmann, R. R. (1971). The feeding behavior of planktotrophic echinoderm
larvae: Mechanisms, regulation, and rates of suspension feeding. Journal of
Experimental Marine Biology and Ecology, 6, 109–160.
Teicher, M. H. & E. M. Blass. (1976). Suckling in newborn rats: Eliminated by
nipple lavage, reinstated by pup saliva. Science, 193, 422–425.
Timberlake, W. & G. A. Lucas. (1985). The basis of superstitious behavior: Chance
contingency, stimulus substitution, or appetitive behavior? Journal of the
Experimental Analysis of Behavior, 44, 279–299.
Tinbergen, N. (1963). On aims and methods of ethology. Zeitschrift für Tierpsycholo-
gie, 20, 410–433.
Tinbergen, N. & A. C. Perdeck. (1950). On the stimulus situation releasing the beg-
ging response in the newly hatched herring gull chick (Larus argentatus argen-
tatus Pont.). Behaviour, 3, 1–39.
324 ANIMAL BEHAVIOR

Vince, M. A. (1964). Synchronization of hatching in American bobwhite quail

(Colinus virginianus). Nature, 203, 1192–1193.
Warkentin, K. M. (1995). Adaptive plasticity in hatching age: A response to preda-
tion risk trade-offs. Proceedings of the National Academy of Sciences, USA, 92,
3507–3510.
Warkentin, K. M. (1999). The development of behavioral defenses: A mechanistic
analysis of vulnerability in red-eyed tree frog hatchlings. Behavioral Ecology,
10, 251–262.
Wassersug, R. (1972). The mechanism of ultraplanktonic entrapment in anuran
larvae. Journal of Morphology, 137, 279–287.
Wassersug, R. J. (1975). The adaptive significance of the tadpole stage with
comments on the maintenance of complex life cycles in anurans. American
Zoologist, 15, 405–417.
 10

The Individual in Context: A Social Network

Approach to the Study of Behavior
David J. White, Andrew P. King, and Meredith J. West

INTRODUCTION
The study of sociality has a venerable history in animal behavior (Allee, 1927;
Alexander, 1974; Wilson, 1975; Hinde, 1976; Trivers, 1985; Hamilton,
1995; de Waal & Tyack, 2003). It is an area of pronounced interest, perhaps
owing to our own species’ social proclivities. Research in the area has provided
ample evidence that social environments have important influences over
physiology, behavior, health, and even evolution (West-Eberhard, 1983;
House et al., 1988; Altmann et al., 1995; Heyes & Galef, 1996; Hoffman
et al., 1999; Fernald & White, 2000; Adkins-Regan, 2005; Dunbar, 2007).
With that said, there are several distinct challenges in conceptualizing and
experimenting on social animals that have limited our progress in producing
a unified theory of the biological basis of sociality. These challenges all relate
to the complex interaction that exists between individuals and their social
environment. Social individuals are intimately linked to their groups in a feed-
back relationship: individuals influence groups, and, at the same time, groups
influence individuals. As a consequence of this interrelationship, it becomes
very difficult to study, or even think about, individuals’ traits—those charac-
teristics or behaviors of an individual that are produced through specific neural
circuits and are the features upon which evolution has acted—because these
traits are not independent of the social environment. Traits depend on the
326 ANIMAL BEHAVIOR

social environment in several ways: (1) they have evolved to function in social
contexts, (2) their development reflects a history of social influences across the
individual’s lifetime, and (3) their expression can depend on the immediate
social context. This chapter will discuss some of the problems involved in
studying social traits and highlight some network methods for experimenting
on and measuring the learning, development, and function of social behavior.
These methods, we argue, hold promise to change the way we understand the
connection between individuals and their social world.

CONCEPTUAL CHALLENGES UNDERLYING SOCIALITY

The study of sociality requires a shift in focus away from individuals and
instead toward networks of individuals. Such an orientation is not common
in proximate or ultimate studies of animal behavior. For example, studies of
mechanistic control of behavior typically center on an individual’s brain and
neuroendocrine system. Focusing solely on the individual, however, can
neglect those proximate pathways that extend beyond the individual’s brain
into the social system. These pathways exist and are as biologically reliable
and important as the physiological pathways that exist within individuals
(West & King, 1987).
The mechanistic study of bird song, for example, has been highly successful
in outlining the neural circuitry and endocrine substrates that allow young
male songbirds to develop an effective song in adulthood (Zeigler & Marler,
2008). This process involves hearing a “tutor” and practicing singing in order
to create a signal that matches the memory of the tutor’s song. The typical
conception of this pathway considers one early social link—hearing the song
of a conspecific—but it neglects to consider the myriad other social links that
may be critical in influencing the ultimate development of the mature vocali-
zation (Baptista, 1990). Beyond hearing song early in life, other social links
can affect not only song development but also gene expression, hormone activ-
ity, and ultimately reproductive success (Snowdon & Hausberger, 1997;
Jarvis et al., 2000; West et al., 2003; Adkins-Regan, 2005). In brown-
headed cowbirds (Molothrus ater), for example, social experiences throughout
life are important influences on song development and reproductive out-
comes. Males change their song in response to feedback gained from interac-
tions with conspecifics (Figure 10.1). Females provide brief wing movements
(“wingstrokes”) in response to males’ singing overtures that are reliable indica-
tors of the females’ song preferences (West & King, 1988), and males can use
this feedback to modify their song such that it matches females’ preferences. In
addition, males can also change their song in response to the reactions of other
THE INDIVIDUAL IN CONTEXT 327

males, who will respond aggressively to the songs that females find attractive.
Thus, the development of an effective communication signal in a male cow-
bird requires an individual having properly functioning neural circuitry but
also depends on the neural pathways and behavior of receivers (rival males,
potential mates) to operate effectively. This view of development is not
individual-based but instead integrates the neural function and behavior of
interconnected individuals within the social environment. In order for behav-
ior in one individual to function effectively, the behavior of others must also
function effectively. This social view of song development opens up many pos-
sibilities to study pathways that might otherwise be ignored under an
individual-centric view of song. For example, the visual system in male cow-
birds must play a role in song development in order for males to take in infor-
mation from the gestures of females, as might tactile information from the
aggressive interactions with other males.
Similar to research on mechanism, work on the evolution of behavior has
also predominantly maintained a focus on individuals over social networks.
Because evolution acts on gene frequencies within gene pools, and the target
of selection is the individual, selection acting on groups of unrelated

Figure 10.1. Social interactions among male cowbirds. The male on the left is singing
and giving a wing-spread display to the male in the center, who is responding with a
neck stretch, while the male on the right observes the interaction.
328 ANIMAL BEHAVIOR

individuals has been considered to be weak at most (Williams, 1974;

Dawkins, 1976). Social environments, however, can have numerous influen-
ces over the process of evolution (Frank, 1998; Keller, 1999). Social environ-
ments, for example, can be a source of selection pressure; they can influence
the traits that relate to reproductive success, influence the direction and inten-
sity of selection, and ultimately influence the fitness of individuals. Living in
social groups can also create selection pressures for new types of traits that do
not exist in asocial individuals, such as favoring skills that allow individuals
to attend to, track, and manipulate social relationships (Byrne & Whiten,
1988; Dunbar, 2007). Social environments can also obscure relationships
between traits and reproductive success. For example, social living can produce
a system of inheritance of behavioral traditions across generations that need
not rely on genetic inheritance (Galef, 1976; West & King, 1987; Freeberg,
1996). Yet social environments contain individuals and thus genes, so the
environments themselves can evolve (Wolf et al., 1999; Wade, 2001; Moore
et al., 2002). In sum, the biological basis of sociality, both from a proximate
and ultimate perspective, cannot be conceptualized only at the individual level
but instead must take into account the social contexts that surround individ-
uals as they develop, behave, and reproduce.

METHODOLOGICAL CHALLENGES UNDERLYING THE STUDY OF SOCIALITY

Research paradigms that study social individuals must overcome a host of
methodological challenges. Social behavior is inherently highly variable and
multifaceted; numerous individuals with competing interests and varying
experiences and skills create the social environment and influence one another.
Social contexts are a moving target. The environment changes not only as
individuals come and go but also as the individuals themselves change (as they
learn and mature). The many interconnections between individuals and their
groups that make the examination of sociality so compelling are exactly those
characteristics that make its scientific study so difficult. An individual’s mature
behavior can be the product of a bewildering number of social interactions
that have occurred throughout a lifetime. Each individual has different experi-
ences and thus different opportunities for social learning. It quickly becomes
intractable for an experimental approach to deal with this level of complexity
and identify the suite of social interactions that are important for effective
development.
In addition to dealing with the complexities of social conditions, there is
also the issue of data collection. Because it is difficult even to conceptualize
an individual’s traits, what then should be measured? The amounts or types of
THE INDIVIDUAL IN CONTEXT 329

behavior produced by a social individual can be related to the activity (or inac-
tivity) of other individuals in the group, and thus measures of an individual
may relate to idiosyncrasies of immediate or past social contexts. For example,
the “aggressiveness” of a territorial male could be measured by the number of
fights in which the individual engages. However, this number may depend on
the number and quality of the potential competitors present or on the number
and quality of contests the male experienced across his lifetime (Chase et al.,
1994; Hsu & Wolf, 1999). Similarly, the success of a male’s courting behavior
will be influenced by the presence and preferences of females as well as the
skills of the rival males present. In addition, traits can be highly context-
based. Adaptive behavior in one context may be maladaptive in another.
For example, an adult male may benefit by being highly aggressive in the
presence of young males but may be at a disadvantage among other, highly
aggressive adult males. Thus selection may favor highly malleable behavioral
strategies that change with the social context. Measuring individuals’ charac-
teristics across numerous social contexts and over time would provide
important information about the reliability and repeatability of indivi-
duals’ behavior, but such an undertaking is rarely possible (but see Gersick
et al., 2012).
Too often the actual measures collected from social individuals are tallies of
some easy-to-measure behavior, for example (ones that we have been guilty of
using) the number of songs a songbird sings, the number of times a male fights
with other males, or the number of wingstrokes a female produces. Implicit in
these measures are several assumptions: (1) that more is better—a ranking of
birds based on the numbers of these behaviors they produce will reveal some-
thing important about the birds’ relative quality; and (2) that these behaviors
are characteristics of the individuals that are not maintained by peculiarities
of their immediate or developmental environments. These assumptions, how-
ever, may not be met. It is often the case that those behaviors that are most
frequent and most easily measured by observers are not in fact the ones that
are most biologically important (West & King, 1988), and when the same
individual is examined across different environments, the patterns of its behav-
ior can be dramatically different (White et al., 2012).
Another methodological challenge associated with studying social behavior is
in statistical analysis. The most common statistical methodologies used in
studies of behavior contain explicit assumptions about the independence of
individuals’ data. Yet if individuals’ data were independent, there would be little
reason for studying social behavior in the first place. Thus, pseudoreplication
(Kroodsma, 1989) becomes an important issue for experimental design and
analysis: there is some level of dependence among individuals, and without
330 ANIMAL BEHAVIOR

the ability to study vast numbers of independent groups, traditional statistics

prove woefully ineffective for studies of interdependent individuals living in
groups.

THE CLASSIC TRADE-OFF—LAB TRAITS AND WILD TRAITS

Traditionally, one technique used to overcome these methodological chal-
lenges has been to simplify the context by reducing the number of individuals
and the complexity of the environment under study. Control and replication,
the hallmarks of experimental design, become easier as the number of interact-
ing individuals is reduced. Behavioral neuroscience and animal learning
psychology have enjoyed tremendous success using such a focus on traits that
can be studied in highly controlled contexts. One problem with this simplify-
ing approach, however, is that the traits studied in the lab do not always reflect
the traits as they operate in the natural world because the lab environment
explicitly removes from the equation the interactions of effects that occur
across a lifetime when multiple individuals engage with one another.
Researchers interested in how behavior has evolved require an understand-
ing of how behavior functions within organisms’ natural ecology. Researchers
in fields such as ethology and behavioral ecology typically eschew the lab envi-
ronment and its lack of ecological validity and attempt to study and even run
experiments on animals in their natural environments. Work in the field,
studying groups of animals in great detail across years and even lifetimes, can
amass datasets on behavior and reproductive success in the wild. While these
researchers have the ability to better examine the biological significance of
traits in the wild, they generally lack experimental control and thus the ability
to remove the many confounding variables that interfere with drawing causal
conclusions.
Recently, there have been a number of technological and methodological
advances that have helped moderate these challenges and have allowed for
new means of running controlled experiments on groups of individuals. Exam-
ples include data-acquisition devices that allow observers to collect more data
more quickly than has been possible in the past (White et al., 2002a); robotic
“animats” that can interact with subjects in the wild (Patricelli et al., 2002;
Taylor et al., 2008); microphone and camera arrays, along with GPS devices,
that can collect a vast amount of data across space and time without the need
for observers to be present (Fitzsimmons et al., 2008); and, most recently,
new genetic and neurophysiological techniques that can manipulate the behav-
ioral phenotypes of target individuals living in social conditions (Maguire et al.,
2013). Complementing these new methods is the ever-increasing speed,
THE INDIVIDUAL IN CONTEXT 331

power, and storage of modern computers that can produce datasets of a size,
breadth, and detail not possible previously. This increase in computer power
has allowed for new statistical techniques to be developed that can examine
dependence and independence of social individuals (Whitehead & Dufault,
2008). Taken together, these methodological and technological advances pro-
vide the foundation for new ways to study sociality.

A NEW METHODOLOGY: EXPERIMENTING ON GROUPS

Here we highlight the methods we have employed over the past decade to
study social development in songbirds. Songbirds offer a powerful model for
the study of social behavior because so much is known about the proximate
control and evolutionary function of their central social trait: song. Work in
behavioral neuroscience has clearly detailed the neural architecture underlying
song learning and song production (Nottebohm 1980; Zeigler & Marler,
2008). From an evolutionary perspective, observations in the field and lab
have revealed how song relates to reproductive success (Catchpole & Slater,
2008). In most temperate species, males use song to compete with males and
to court females. Females have preferences for songs, which allow them
to select as mates those males who have high heritable fitness (Searcy &
Andersson, 1986). The song signal, however, is only one component of a
communication network, and the relationship between sociality and song
can be complex. Interactions among numerous individuals can change both
the form and function of song. Social factors have been shown to play a critical
role in songbird behavior and physiology and in affecting how selection acts
on signals and behavior (Snowdon & Hausberger, 1997; Jarvis et al., 2000;
White et al., 2010a).
Cowbirds provide a good example of the complexities that emerge as indi-
viduals interact. Cowbirds are gregarious obligate brood parasites; thus they
do not build nests or raise their young—instead they lay eggs in nests of other
bird species. Once the young cowbirds leave their host nests, they join flocks
of other cowbirds and spend the rest of their lives in these flocks, competing,
cooperating, communicating, and breeding (Friedmann, 1929; Ortega,
1998). Like other songbirds, song is central to the cowbirds’ social system.
Male cowbirds use song to establish dominance relationships with males
and to court females (Dufty, 1986). Females are also active—but much more
subtle—participants in assortative mating (King et al., 2003). Females do not
sing, but they do have preferences for songs (King & West, 1990). Females
can exercise their choice of males by regulating their social distance, providing
or removing opportunities for males to sing to them. They also will compete
332 ANIMAL BEHAVIOR

with other females for access to mates (Freed-Brown et al., 2006). In addition,
through wingstroking, females can modify males’ song development to suit their
individual preferences (West & King, 1988; King et al., 2005). Taken together,
dominance, courtship, song preferences, and competition combine in complex
ways to produce the social interactions necessary for effective breeding.
The first step in designing a new paradigm to study individuals as they live
and interact in groups was to set up a facility where groups could be housed,
individuals could be measured, and experimental control could be imposed
(King et al., 2002). Using large outdoor aviaries at field stations at Indiana
University and the University of Pennsylvania, we housed cowbirds in large
groups (approximately 20–70 individuals each). While these conditions can-
not be considered directly analogous to cowbird social conditions in the wild,
the aviaries do provide the birds with environments rich in ecological and
social complexity. In these groups, we could manipulate the social experiences
available by changing the composition of the flocks. This preparation brought
with it new challenges for data collection and statistical analysis. We refined
the preparation by devising a automated speech-recognition data-collection
technology (White et al., 2002a), using agent-based simulations to test prop-
erties of data-collection procedures (White & Smith, 2007), incorporating
social network analyses to measure social structure (Miller et al., 2008), and
creating new techniques for experimentally modifying social behavior (West &
King, 2008). These tools and techniques have allowed us to maintain complexity
in the environment while not sacrificing experimental control. The speech-
recognition data-collection system has been especially valuable in allowing us to
take extensive and detailed measures of birds’ interactions and to collect vast
libraries of social behavior on numerous individuals across their lifetimes.
Results that emerged from these experiments have consistently challenged
our preconceived assumptions about the relationship between individuals
and groups. Traits that we had always considered as invariant (such as sexual
preferences, aggressiveness, reproductive output, and song quality) were
remarkably plastic in response to the developmental and immediate social
environments. Results of this work have prompted us to think more about
socially distributed behavior, where tendencies toward social learning and
social interactions are the variables that have critical connections to fitness. It
has changed how we think of the development and proximate causes of behav-
ior, as well as the targets upon which evolution acts.
Conducting experiments in large, confined settings was a critical compo-
nent of the research paradigm. In these conditions, birds always have access
to social information. By varying the demographic composition of the flocks,
we could change the type of social information available within different
THE INDIVIDUAL IN CONTEXT 333

groups. For example, in one experiment we exposed juvenile male cowbirds to

two different types of social learning environments (White et al, 2002b). Some
juvenile males were housed only with female cowbirds (juvenile-only, or J,
condition), while other juvenile males were housed with females and with
adult males (juvenile-adult, or JA, condition). Thus the JA-condition juveniles
had access to social information about song and behavior from sexually expe-
rienced adult males, while the J juveniles had to generate their breeding behav-
iors on their own. What started as a song-development experiment quickly
broadened to a study of a wide variety of effects on other behavior. There were
indeed effects on juvenile males’ song development and song quality, but there
were also widespread effects on other aspects of their behavior that we did not
anticipate. For example, while JA juveniles developed social and breeding
behaviors typical of normal cowbirds (aggressive interactions with other males,
adherence to a linear dominance hierarchy, pair bonding with females, and
monogamous mating patterns; Friedmann, 1929), J juveniles developed atypi-
cal behavior: they rarely engaged in aggressive countersinging bouts or fights
with other males, they did not pair bond with females, and instead they copu-
lated promiscuously with multiple mates. Furthermore, we have found that
these atypical patterns of behavior persisted into adulthood and could be cul-
turally transmitted to a new generation of juvenile males (White et al., 2007).
This was our first discovery of the latent plasticity underlying species-
typical breeding behavior. Since then we have experimented on a wide array
of different social variables in the same manner: changing the social composi-
tion of flocks and measuring the effects on the individuals living within those
flocks. Our manipulations have influenced just about every variable we have
thought to measure, from song development through mating skills, sexual
preferences, and reproductive success. In addition, we have found that social
factors can influence cowbirds throughout their lifetimes. For example, we
conducted an experiment on adult males that showed that adults, similar to
juveniles, were extremely malleable in response to aspects of the social envi-
ronment (White et al., 2010a). In this experiment, adult males in one (static)
condition lived in stable groups that did not change across one year. These
birds thus were very familiar with their group-mates and only their group-
mates. In another group (the dynamic condition), birds were exposed to new
individuals as other males and females briefly immigrated and emigrated
through the groups. This manipulation had pronounced effects on the repro-
ductive success of the birds in the two conditions with dynamic-condition
birds having better skills related to dominating other males and courting
females quickly and effectively. We tested the same groups of males in a second
year with half of each group staying in the same conditions as the year before
334 ANIMAL BEHAVIOR

and half switching from one condition to the other (Gersick et al., 2012).
There were two important findings. First, the effect replicated across years—
the dynamic condition enhanced males’ social skills related to reproducing—
which meant that we had reversed the effects on the birds that switched condi-
tions from one year to the next. Second, evaluations of courting behavior and
reproductive success revealed that only the males that had been housed in the
static condition for both years showed any consistency in their behavior; that is,
they showed stability in how they sang to females and males and also in their mat-
ing success. All other groups showed no consistency in their behavior. Thus the
repeatability of many aspects of the males’ behavior was extremely low across years
for adult males who changed social conditions. Only when the groups remained
stable did their behavior and mating success remain stable (White et al., 2012).
Of the many surprises that have come from these experiments, perhaps the
most dramatic have come from experiments testing female behavior. Like
males, females have been shown to be plastic in response to changes in their
social environment. Through modifications to their groups, we have been able
to modify females’ sexual preferences for male song. We have been able to
change preferences for individual males. We have even been able to remove
preferences for local song types over subspecies song types. In addition, the
reproductive output of females—how many eggs they laid—turned out to
be affected by the social environment in which they found themselves
(Gros-Louis et al., 2003; King et al., 2003; West et al., 2006).

NEWEST ADVANCES
This social approach to studying behavior discovered a wealth of new fac-
tors important for the development of effective communication and mating.
For example, decades of work that used song-tutoring techniques in an
attempt to create malleability in females’ song preferences all failed (King &
West, 1983, 1990). Only when these techniques were used in social contexts
were we able to influence females’ preferences.
Studying birds in groups, however, highlighted many of the challenges that
were discussed above—most dramatically, how difficult it was to conceptual-
ize individuals’ traits. Many of the birds’ characteristics—their aggressiveness,
their song quality, their courtship skills, and their reproductive success—were
determined to a large extent by the developmental and immediate social ecol-
ogy the birds experienced. Thus we were left with the problem of trying to
characterize these highly malleable individuals. What characteristics could selec-
tion be acting on? What was it about individuals that allowed some to succeed and
gain reproductive success and others to fail? The answer, we suggest, is that those
THE INDIVIDUAL IN CONTEXT 335

individuals who were successful were those who gained the most effective
social skills: the abilities to read and react to social conditions quickly. Those
who could detect accessible mates, or evaluate relative dominance levels of
other males present and then quickly adjust their behavior in response, had
advantages in terms of competition and courtship, thereby gaining more repro-
ductive opportunities under various social conditions. If this is true, then selec-
tion should favor social learning and attentional abilities—those tendencies of
individuals to seek out others and learn from them during development.
This conception of individuals’ social learning proclivities provided us with
an answer to the problem of how to study individuals’ traits: one should not
study individuals’ behavioral outputs but instead study the specific patterns
of how individuals interact with their social environment. Our history of try-
ing to study behavioral outputs (e.g., total songs sung, total fights, total chatter
calls or wingstrokes by females) as they relate to song development or repro-
ductive success have usually met with failure. These tallies, which were aggre-
gated across weeks or months, measured gross amounts of behavior but could
not reveal the specifics of how individuals interacted with one another. Said
another way, the total number of times a behavior occurred says nothing about
whether the behavior was used in the appropriate context or was functionally
effective. The touchstone of effective communication in birds or humans is
not just the presence of a signal, be it a song or a word, but rather how it is used.
We required pragmatic measures that were sensitive to function to know
whether a behavior was being used in an appropriate context (Gros-Louis
et al., 2010). Thus, we needed a way to categorize different types of social inter-
actions in order to measure the potential functional effectiveness of a behavior.

SEQUENCES AND CONTINGENCIES

Determining how social learning tendencies and social learning opportuni-
ties could shape pragmatic development was our new challenge. It required
measuring more than just individuals’ behavior, but also the social interactions
that surrounded those behaviors on a precise time scale. We drew from what
had been discovered by lab researchers studying the mechanisms of learning.
One of the seminal contributions of animal learning psychology has been in
revealing the importance of contingencies (Skinner, 1938; Rescorla & Solo-
mon, 1967). Both classical and associative learning depend not on the num-
ber of times events occur but on their relationship: that is, learning occurs
when reactions are contingent on actions. This suggests that it is not neces-
sarily important to know how much a male sang directed song, for example,
but instead what happened during each directed singing interaction that may have
336 ANIMAL BEHAVIOR

provided the singer with feedback (or lack of feedback) about his use of the song.
Perhaps he received an approach or wingstroke reaction from a female, or
the song immediately led to a fight from another male. These are the contin-
gent interactions that provide opportunities for learning and that can shape
the social behavior repertoire across life.
Our first rudimentary investigation of sequences of behavior derived from a
meta-analysis of a number of groups we had studied across many years
(White et al., 2010b). We had been collecting eggs laid during breeding sea-
sons in our captive conditions and then, by subjecting them to microsatellite
parentage analyses, we were able to determine each egg’s parentage. We had
known for some time that egg production was highly variable within and
across groups; some groups would yield only a handful of eggs in a breeding
season while other groups would lay many hundreds. Until we amassed data
from all of the groups, we had been unable to determine what it was that con-
tributed to this variation. In the meta-analysis, we looked for variables that
explained reproductive success: both copulation success and the number of
offspring produced. We found that the best predictor of females’ copulation
success was the amount of female-directed song she received from her pair-
mate. This was not surprising, as we had seen this relationship in the past.
What was surprising was that there was only one variable that related to off-
spring production. It was the amount of male-male countersinging that
occurred in the groups. Males engage in countersinging as a form of competi-
tion where two or more males sing in close proximity (within approximately
30 cm). Males take turns singing back and forth with little time elapsing
between songs (usually no more than 10 to 15 seconds). There were two
notable aspects of the relationship between countersinging and egg produc-
tion. First, countersinging is a social variable and could not be understood at
an individual, or even a dyadic, level. It required the coordinated behavior of
many males. Second, examining countersinging required a sequential analysis.
Even though countersinging was comprised of male-directed song, the total
amount of male-directed singing was not related to egg production. Only
when male-directed song is used in a specific social context did it relate to
female reproductive stimulation.
Clearly the approach used for the meta-analysis—studying patterns of
social interactions across many groups—was not practical. It required exten-
sive numbers of groups to be run across many years in order to have the num-
ber of independent groups necessary to avoid pseudoreplication. In addition,
while we happened to find this one type of sequential social interaction that
was important for reproductive success, we needed a method to examine many
different types of social interactions simultaneously across development.
THE INDIVIDUAL IN CONTEXT 337

SOCIAL NETWORKS
Recently, we have looked to social network analysis to provide a more sensi-
tive means for studying social behavior. This is a suite of statistical tools that
evaluate and quantify social structure and social behavior that has recently
become popular in areas of animal behavior (Sih et al., 2009). These tools have
proven to be sensitive to aspects of social structure and patterns of social interac-
tions that we could not see with traditional parametric statistical analysis (Miller
et al., 2008). For example, we have incorporated some social network metrics to
examine how cowbird social networks change in response to the introduction of
new individuals (Gersick et al., 2012; White et al., 2010a, 2012). We found
that these metrics revealed properties of social structure that were not apparent
with other analyses. Yet the social network methods still lacked in important
ways, leaving the techniques only marginally useful in studying social develop-
ment at a detailed level. For example, the majority of the extant social network
methods dichotomize data to evaluate whether a link exists between two indi-
viduals or not. This effectively removes or at least greatly reduces the informa-
tion available in valued datasets. Another critical weakness was that there were
only a few methods that took into account sequences of actions and reactions.
We therefore attempted to develop our own social network method for meas-
uring the important aspects of social exchanges: the clusters of multivariate, con-
tingent interactions that change across time and that can involve many different
individuals. Thus the new method had to be able to take into account (1) multi-
variate data (that is, different types of actions and reactions), (2) directed data
(tracking the direction of the social exchange), (3) weighted data (maintaining
information about the number of interactions that exist), and (4) sequences
(to detect the reactions that are contingent on actions) from multiple individuals
(not just single subjects or even dyads). Finally, (5) we needed to take into
account how the network structures change across time because change is inte-
gral to the study of learning and development. Thus our methods had to be able
to measure interactions across different time series ranging from contingent
reactions in milliseconds to developmental milestones in months or years.
At present there are no social network methods that can account for all of
these aspects of social datasets. The closest technique that exists is the triadic
census that was first developed by Paul Holland and Samuel Leinhardt (Holland
& Leinhardt, 1970; Davis & Leinhardt, 1972). The triadic approach has been
used in animal behavior to describe a variety of systems including social interac-
tions, formation of coalitions, dominance hierarchies, grooming networks, and
alloparental care (Caplow, 1959; Chase, 1980; Hogue et al., 1996; Sinha,
1998; Chase et al., 2002; Faust & Skvoretz, 2002; Wittig et al., 2008; Schubert
et al., 2009). Recently, triads have been used in a meta-analysis of network
338 ANIMAL BEHAVIOR

relations in different species of animals to compare social structures (Faust &

Skvoretz, 2002). Stanley Wasserman and Katherine Faust (Wasserman &
Faust, 1994) define the triadic census as describing all the possible combina-
tions of interactions among three individuals (Figure 10.2). There are

Figure 10.2. All possible combinations of triadic interactions. The first number repre-
sents the number of reciprocal dyads in a triad (both individuals interact with one
another), the second number represents the number of asymmetric dyads in a triad (one
individual does not reciprocate an interaction), and the last number represents the
number of null triads (two individuals do not interact). If a letter is present, it represents
one of four triadic states: transitive (T), cyclic (C), up (U), or down (D) triad. A tran-
sitive triad occurs when individual A is connected to individual B, individual B is con-
nected to individual C, and individual A is also connected to individual C. Cyclicality
occurs when individual A is connected to individual B, individual B is connected to
individual C, and individual C is connected to individual A. An up triad occurs when
individual A is connected to individual C and individual B is connected to individual C.
A down triad occurs when individual C is connected to individual A and individual B.
THE INDIVIDUAL IN CONTEXT 339

15 possible ways three individuals can interact, including a simple sequence

where one individual interacts with another up to reciprocal interactions among
all three individuals. Characterizing social interactions like this is an advance
over tallying individuals’ behavior, but its inability to use valued and multivari-
ate relations in its calculation limits its application to learning and development.
We modified the triadic census to use valued, multivariate, directional rela-
tions in a specified time frame to define triad formation. We assembled triads
from a time-coded series of events creating temporal triads. The events can be
any directed or undirected behaviors that are then transformed into a tabula-
tion of triadic frequencies if certain behavioral and timing criteria are met.
The criteria are specific to the type of behavior involved and the type of rela-
tions that are the focus of the particular triad tabulation. For example, after a
male sings a directed song, all reactions to the vocalization by other individuals
within a specific timeframe would be documented as part of a triadic
sequence. Knowing the specific behaviors that constitute a type of triad, a
sequence of action and reaction is uncovered. By tabulating the frequency of
triadic interactions, a frequency distribution of triad types is created. This dis-
tribution constitutes a social signature of recurring behavior sequences that
define the consequences of behaviors and ultimately what is available to be
learned. The triadic analysis provides a more complete description of the
nature of variation in social interactions than does an individual or even dyadic
analysis.

TWO EXAMPLES OF THE UTILITY OF THIS SOCIAL SEQUENCE ANALYSIS

Our earlier investigation into sequences of behavior revealed that counter-
singing among males was a particularly important type of social interaction
because it related to egg production of female group-mates (White et al.,
2010b). Finding this pattern, however, required examining 17 groups across
numerous years just to achieve sufficient statistical power. Here we used our
new sequence-analysis methods on only one group of 24 male and 28 female
cowbirds in an attempt to determine whether it could detect the biologically
relevant patterns of social interactions discovered in the meta-analysis. We
focused on two of the triad types. The first did not measure triads at all but
instead measured the dyadic interactions between a male and his pair-mate.
Thus this triad was comprised of events where a single male simply sang to a
female (triad 012 from Figure 10.2). The second triad type did take into
account triadic interactions; here it involved detecting sequences where a male
sang to another male, the recipient sang back, and then the instigating male
sang to a female (triad 111U). Thus, the male engaged in countersinging
340 ANIMAL BEHAVIOR

and then, no more than 15 seconds later, courted his female. We found that
the number of 012 triads in which males engaged was the only triad type that
correlated with their copulation success (r ¼ 0.85). The only triad type
that correlated with offspring production was the 111U triad (r ¼ 0.66).
Thus, this analysis supported the past meta-analysis indicating that the dyadic
singing interaction between males and pair-mates was the best predictor of
copulation success, but understanding egg production required taking into
account the social singing context. It required looking beyond the dyadic
interaction and revealed a specific pattern of social singing characterization.
The sequence analysis served as an advance over the meta-analysis in two ways.
First, it was much more sensitive to finding effects: one group in one breeding
season versus 17 groups across three years. Second, it revealed that the rela-
tionship between countersinging and egg production was linked to a specific
singing strategy: countersing, then court.
We have also used temporal triads to study the development of reproduc-
tive skills in groups of juvenile male birds that were raised with either sexually
inexperienced juvenile females or experienced adult females (Freed-Brown
et al., 2006). These housing conditions had pronounced effects on the males’
courtship competence and their reproductive success as measured by the
increased egg production from the adult females over that of the juvenile
females. We knew from previous work that juvenile females were physiologi-
cally capable of laying as many eggs as adult females, so the greater egg produc-
tion from the adults must have been a result of their males using more effective
courtship behavior to stimulate egg production. Traditional analyses were
unable to detect what happened during development that caused these func-
tional effects; these analyses revealed overlapping total amounts of singing
and social affiliations among males and females. Analysis of triadic distribu-
tions, however, found that the males with the adult females engaged in more
context-specific male-directed song than the males raised with the juvenile
females. Female cowbirds use a “chatter” vocalization to attract males to court
them, and that also incites male competition. Figure 10.3 shows a distribution
of triads containing male-male directed song after a female has chattered and a
male has approached her. We can see from Figure 10.3 that the males with the
adult females managed to sing many more songs to males in the context of
reacting to a female chatter than did the males with juvenile females. Thus, a
tabulation of triadic interactions picked up a contextual difference in song
use. Male countersinging increases female egg production, thus explaining
the reproductive advantage of the males raised with the adult females.
Given the differences in the courtship behavior of males raised with adult or
juvenile females, how might the adult females be different from the juveniles?
THE INDIVIDUAL IN CONTEXT 341

Figure 10.3. Triadic episodes by triad type for four flocks of brown-headed cowbirds.
Two flocks contained adult and two contained juvenile females. Thus, each triad
category has four bars, each representing one flock. The heights of the vertical bars
represent the frequencies of triadic episodes created by male-male directed songs
immediately after a male has approached a chattering female.

Figure 10.4 shows an example of triadic distributions for an adult and a juve-
nile female. The bars represent triads created by a female approach of another
individual. Each female has an equal number of triads, but approaches are dis-
tributed very differently across triad types. The adult female is much more
stereotyped in her interactions than the juvenile female. Thus, the adult

Figure 10.4. The distribution of triadic categories created by the approach behavior of
an individual adult female and an individual juvenile female brown-headed cowbird.
The heights of the bars represent the frequencies of triad types initiated by the females
approaching other individuals.
342 ANIMAL BEHAVIOR

female is more predictable in her interactions and more likely to give system-
atic feedback to males who approach her. This is likely the reason the males
with the adults learned more effective courtship behaviors. These differences
give us clues to variables affecting female reproductive behavior as we have
found that the patterns of adult female approach behavior are stable over sea-
sons and even years and that female approach behaviors can be predictive of
reproductive output (Kohn et al., 2011, 2013, 2013).
Given the poor courtship skills of the males raised with juvenile females, it
would seem that each new generation of females must learn how to interact with
males, and exactly what is learned will influence what can be taught to young
males. This was a surprising finding because we had studied song development
during the first year of life in male cowbirds housed in flocks composed of a
mix of juvenile and adult females, and we had not been able to detect a difference
in female social behavior. In retrospect, it is clear to us that in those flocks com-
posed of males and adult and juvenile females, the juvenile females adopted adult
female behavior rapidly after their first contact with adult females. The social
organization can thus act as a highly flexible learning-based developmental
mechanism. The ability of the adult females in this study to coordinate their
behavior to foster the development of male courtship skills reveals that the under-
lying pattern of what is learned is adaptively coordinated behavior. Thus, the evo-
lutionary success of this parasitic species seems reliant on cultural transmission
for the acquisition of some of the most basic species-typical behaviors (see also
Freeberg et al., 1999, 2001, White et al., 2007).

DISCUSSION
Taken together, this chapter has laid out the case for the need to create a
new way to study social animals, focusing less on the individual and more on
the network of individuals in order to deal with the problem inherent in social
systems: that the social individual is not an independent entity; its neurobiol-
ogy, genes, and behavior are all influenced in dramatic ways by the social con-
text that surrounds the individual throughout its life. This means that
mechanistic and functional investigations focused solely on the individual
obscure (1) important social inputs and outputs in neural pathways and (2)
important forces that affect the direction and intensity of selection.
Social network analysts have argued for an understanding of social behavior
at different levels of analysis (James et al., 2009; Krause et al., 2009). While
the most common level of analysis has been at the individual or group/popu-
lation level (Sih et al., 2009), the triadic level has received little attention in
studies of animal behavior (Faust & Skvoretz, 2002). However, this may be
THE INDIVIDUAL IN CONTEXT 343

a constraint of using methods designed in other fields that without modifica-

tion are not readily applied to animal behavior data (see James et al. [2009]
for a discussion). Conducting research at different levels of analyses does not
always provide new insights as each level of analysis can reveal similar results
(Faust & Skvoretz, 2002). By conducting an analysis using temporal triads,
contextual information is collected about functional use of behavior that is
not measured at the individual or group level. The triadic perspective is espe-
cially important because it begins to clarify what is meant by behaviors labeled
as social interaction. Few procedures look at the intercalation of male and/or
female behaviors during the encounters that constitute the daily lives of group
members. Here we can begin to see what cowbirds are doing when we say they
are interacting, and as we have shown, the outcome reveals diverse behaviors
in multiple contexts. In the broadest terms, analyses such as those shown here
uncover the “experience” of experience, a fundamental attribute of an animal’s
developmental ecology. Through an understanding of the types of interactions
at the individual and triadic level we can see how social activity can create an
ecology that fosters the transfer of biologically salient information.
A unified theory of the biological basis of sociality must account for social
behavior across multiple levels of analysis (Tinbergen, 1963): proximate mech-
anisms underlying the physiological basis of social interactions, the development
of behavioral repertoires used in social interactions, the functional consequen-
ces resulting from social interactions, and finally the evolutionary history of
social life across species. Investigating patterns of social interactions by individ-
uals across their lifetimes and across varying social environments has provided a
window into understanding social behavior at many of these levels of analysis.
We are able to see how individuals acquire social information and what contin-
gencies exist as a consequence of their actions. It is these contingent learning
experiences that serve as the building blocks of behavior across lifetimes. These
early tendencies to engage in social interactions and acquire social information
are intimately tied to reproductive success. These early propensities—
approaching, learning, and engaging with others (Schneirla, 1965; Hood
et al., 1995)—and their underlying physiological mechanisms (Goodson
et al., 2009; Goodson & Kingsbury, 2011)—allow individuals to develop
self-created social niches and thus may be the actual site of social evolution.

ACKNOWLEDGMENTS
We thank V. Anne Smith and Jennifer L Miller for their feedback in devel-
oping some of the triadic concepts presented in this chapter. Ken Yasukawa
and Zuleyma Tang-Martínez provided valuable comments on an earlier draft
of this chapter.
344 ANIMAL BEHAVIOR

REFERENCES AND SUGGESTED READING

Adkins-Regan, E. (2005). Hormones and Animal Social Behavior. Princeton, NJ:
Princeton University Press.
Alexander, R. D. (1974). The evolution of social behavior. Annual Review of Ecology
and Systematics, 5, 325–383.
Allee, W. C. (1927). Animal aggregations. Quarterly Review of Biology, 2, 367–398.
Altmann, J., R. Sapolsky, & P. Licht. (1995). Baboon fertility and social status.
Nature, 377, 688–689.
Baptista, L. F. (1990). Song learning in the white-crowned sparrow (Zonotrichia leu-
cophrys): Sensitive phases and stimulus filtering revisited. In R. van den Elzen,
K.-L. Schuchmann, & K. Schmidt-Koenig (eds), Proceedings of the
International Meeting Centennial of the DO-G, Current Topics in Avian Biology
(pp. 143–152). Bonn, Germany: Deutsche Ornithologen-Gesellschaft.
Byrne, R. & A. Whiten. (1988). Machiavellian Intelligence: Social Expertise and the
Evolution of Intellect in Monkeys, Apes, and Humans. New York: Oxford Uni-
versity Press.
Caplow, T. (1959). Further development of a thoery of coalitions in the triad. Ameri-
can Journal of Sociology, 5, 488–493.
Catchpole, C. K. & P. J. B. Slater. (2008). Bird Song: Biological Themes and Varia-
tions. Second Edition. Cambridge, UK: Cambridge University Press.
Chase, I. D. (1980). Social processes and hierarchy formation in small groups: A com-
parative perspective. American Sociological Review, 45, 905–924.
Chase, I. D., C. Bartolomeo, & L. A. Dugatkin. (1994). Aggressive interactions and
inter-contest interval: How long do winners keep winning? Animal Behaviour,
48, 393–400.
Chase, I. D., C. Tovey, D. Spangler-Martin, & M. Manfredonia. (2002). Individual
differences versus social dynamics in the formation of animal dominance hier-
archies. Proceedings of the National Academy of Sciences, USA, 99, 5744–5749.
Davis, J. A. & S. Leinhardt. (1972). The structure of positive interpersonal relations
in small groups. In J. Berger (ed.), Sociological Theories in Progress (pp. 218–
251). Boston: Houghton Mifflin.
Dawkins, R. (1976). The Selfish Gene. New York: Oxford University Press.
de Waal, F. B. M. & P. L. Tyack (eds.). (2003). Animal Social Complexity: Intelligence,
Culture, and Individualized Societies. Cambridge, MA: Harvard University Press.
Dufty, A. M., Jr. (1986). Singing and the establishment and maintenance of domi-
nance Hierarchies in captive brown-headed cowbirds. Behavioral Ecology and
Sociobiology, 19, 49–55.
Dunbar, R. I. M. (2007). Evolution in the social brain. Science, 317, 1344–1347.
Faust, K. & J. Skvoretz. (2002). Comparing networks across space and time, size and
species. Sociological Methodology, 32, 267–299.
Fernald, R. D. & S. A. White. (2000). Social control of brains: From behavior to
genes. In M. S. Gazzaniga (ed.), The Cognitive Neurosciences (pp. 1193–
1208). Cambridge, MA: MIT Press.
THE INDIVIDUAL IN CONTEXT 345

Fitzsimmons, L. P., J. R. Foote, L. M. Ratcliffe, & D. J. Mennill. (2008) Frequency

matching, overlapping and movement behaviour in diurnal countersinging
interactions of black-capped chickadees. Animal Behaviour, 75, 1913–1920.
Frank, S. A. (1998). Foundations of Social Evolution. Princeton, NJ: Princeton Univer-
sity Press.
Freeberg, T. M. (1996). Assortative mating in captive cowbirds is predicted by social
experience. Animal Behaviour, 52, 1129–1142.
Freeberg, T. M., S. D. Duncan, T. L. Kast, & D. A. Enstrom. (1999). Cultural influ-
ences on female mate choice: An experimental test in cowbirds, Molothrus ater.
Animal Behaviour, 57, 421–426.
Freeberg, T. M., A. P. King, & M. J. West. (2001). Cultural transmission of vocal tra-
ditions in cowbirds (Molothrus ater) influences courtship patterns and mate
preferences. Journal of Comparative Psychology, 115, 201–211.
Freed-Brown, G., A. P. King, J. L. Miller, & M. J. West. (2006). Uncovering sources
of variation in female sociality: Implications for the development of social pref-
erences in female cowbirds (Molothrus ater). Behaviour, 143, 1293–1315.
Friedmann, H. (1929). The Cowbirds: A Study in the Biology of Social Parasitism.
Springfield, IL: C. C. Thomas.
Galef, B. G., Jr. (1976). Social transmisson of acquired behavior: A discussion of tra-
dition and social learning in vertebrates. Advances in the Study of Behavior, 6,
77–100.
Gersick, A. S., N. Snyder-Mackler, & D. J. White. (2012). Ontogeny of social skills:
Social complexity improves mating and competitive strategies in male brown-
headed cowbirds. Animal Behaviour, 83, 1171–1177.
Goodson, J. L. & M. A. Kingsbury, (2011). Nonapeptides and the evolution of social
group size in birds. Frontiers in Neuroanatomy, 5, 13.
Goodson, J. L., S. E. Schrock, J. D. Klatt, D. Kabelik, & M. A. Kingsbury. (2009).
Mesotocin and nonapeptide receptors promote estrildid flocking behavior.
Science, 325, 862–866.
Gros-Louis, J. M., M. J. West, & A. P. King. (2010). Comparative perspectives on
the missing link: Communicative pragmatics. In M. S. Blumberg, J. H. Free-
man, & S. R. Robinson (eds.), The Oxford Handbook of Developmental and
Comparative Neuroscience (pp. 684–707). Oxford, UK: Oxford University
Press.
Gros-Louis, J. M., D. J. White, A. P. King, & M. J. West. (2003). Female brown-
headed cowbirds’ (Molothrus ater) social assortment changes in response to
male song: A potential source of public information. Behavioral Ecology and
Sociobiology, 56, 163–173.
Hamilton, W. D. (1995). Narrow Roads of Gene Land: The Collected Papers of W. D.
Hamilton, Vol, 1. Evolution of Social Behaviour. New York: W. H. Freeman.
Heyes, C. M. & B. G. Galef (eds.). (1996). Social Learning in Animals: The Roots of
Culture. San Diego, CA: Academic Press.
Hinde, R. A. (1976). Interactions, relationships and social structure. Man, 11, 1–17.
346 ANIMAL BEHAVIOR

Hoffman, H. A., M. E. Benson, & R. D. Fernald. (1999). Social status regulates

growth rate: Consequences for life history strategies. Proceedings of the National
Academy of Sciences, USA, 96, 14171–14176.
Hogue, M. E., J. P. Beaugrand, & P. C. Lague. (1996). Coherent use of information
by hens observing their former dominant defeating or being defeated by a
stranger. Behavioral Processes, 38, 241–252.
Holland, W. & S. Leinhardt. (1970). A method for detecting structure in sociometric
data. American Journal of Sociology, 76, 492–513.
Hood, K. E., G. Greenberg, & E. Tobach (eds.). (1995). Behavioral Development:
Concepts of Approach/Withdrawal and Integrative Levels. New York: Garland.
House, J. S., K. R. Landis, & D. Umberson. (1988). Social relationships and health.
Science, 241, 540–545.
Hsu, Y. & L. L. Wolf. (1999). The winner and loser effect: Integrating multiple expe-
riences. Animal Behaviour, 57, 903–910.
James, R., D. P. Croft, & J. Krause. (2009). Potential banana skins in animal social
network analysis. Behavioral Ecology and Sociobiology, 63, 989–997.
Jarvis, E. D., S. Ribeiro, M. L. da Silva, D. Ventura, J. Vielliard, & C. V. Mello.
(2000). Behaviourally driven gene expression reveals song nuclei in humming-
bird brain. Nature, 406, 628–632.
Keller, L. (ed.). (1999). Levels of Selection in Evolution. Princeton, NJ: Princeton Uni-
versity Press.
King, A. P. & M. J. West (1983). Female perception of cowbird song: A closed devel-
opmental program. Developmental Psychobiology, 16, 335–342.
King, A. P. & M. J. West. (1990). Variation in species-typical behavior: A contempo-
rary theme for comparative psychology. In D. A. Dewsbury (ed.), Contempo-
rary Issues in Comparative Psychology (pp. 331–339). Sunderland, MA: Sinauer.
King, A. P., M. J. West, & M. Goldstein. (2005). Nonvocal shaping of avian
song development: Parallels to human speech development. Ethology, 111,
101–117.
King, A. P., M. J. West, & D. J. White. (2002). The presumption of sociality: Social
learning in diverse contexts in brown-headed cowbirds (Molothrus ater). Jour-
nal of Comparative Psychology, 116, 173–188.
King, A. P., D. J. White, & M. J. West. (2003). Female proximity stimulates develop-
ment of male competition in juvenile brown-headed cowbirds, Molothrus ater.
Animal Behaviour, 66, 817–828.
Kohn, G. M., A. P. King, R. Dohme, G. R Meredith, & M. J. West. (2013). In the
company of cowbirds, Molothrus ater ater: Robust patterns of sociability pre-
dict reproductive performance. Journal of Comparative Psychology, 127, 40–48.
Kohn, G. M., A. P. King, U. Pott, & M. J. West. (2013). Robust fall social displays
predict spring reproductive behavior in brown-headed cowbirds (Molothrus
ater ater). Ethology, 119, 1–11.
Kohn, G. M., A. P. King, L. L. Scherschel, & M. J. West. (2011). Social niches and
sex assortment: Uncovering the developmental ecology of brown-headed cow-
birds, Molothrus ater. Animal Behavior, 82, 1015–1022.
THE INDIVIDUAL IN CONTEXT 347

Krause, J., D. Lusseau, & R. James. (2009). Animal social networks: An introduction.
Behavioral Ecology and Sociobiology, 63, 967–973.
Kroodsma, D. E. (1989). Suggested experimental designs for song playbacks. Animal
Behaviour, 37, 600–609.
Maguire, S., M. F. Schmidt, & D. J. White. (2013). Social brains in context: Lesions
targeted to the neural song control system of female cowbirds affects their social
network. PloS ONE 8(5): e63239. doi:10.1371/journal.pone.0063239.
Miller, J. L., A. P. King, & M. J. West. (2008). Female social networks influence male
vocal development in brown-headed cowbirds (Molothrus ater). Animal Behav-
iour, 76, 931–941.
Moore, A. J., K. F. Haynes, R. F. Preziosi, & P. J. Moore. (2002). The evolution of
interacting phenotypes: Genetics and evolution of social dominance. American
Naturalist, 160, S186–S197.
Nottebohm, F. (1980). Brain pathways for vocal learning in birds: A review of the
first 10 years. Progress in Psychobiology and Physiological Psychology, 9, 85–124.
Ortega, C. P. (1998). Cowbirds and Other Brood Parasites. Tucson: University of Ari-
zona Press.
Patricelli, G. L., J. A. Uy, G. Walsh, & G. Borgia. (2002). Male displays adjusted to a
female’s response. Nature, 415, 279–280.
Rescorla, R. A. & R. L. Solomon. (1967). Two-process learning theory: Relationships
between Pavlovian conditioning and instrumental learning. Psychological
Review, 74, 151–182.
Schneirla, T. C. (1965). Aspects of stimulation and organization in approach/with-
drawal processes underlying vertebrate behavioral development. Advances in
the Study of Behavior, 1, 1–74.
Schubert, M., N. Pillay, & C. Schradin. (2009). Parental and alloparental care in a
polygynous mammal. Journal of Mammalogy, 90, 724–731.
Searcy, W. A. & M. Andersson. (1986). Sexual selection and the evolution of song.
Annual Review of Ecology and Systematics, 17, 507–533.
Sih, A., S. F. Hanser, & K. A. McHugh. (2009). Social network theory: New insights
and issues for behavioral ecologists. Behavioral Ecology and Sociobiology, 63,
975–988.
Sinha, A. (1998). Knowledge acquired and decisions made: Triadic interactions
during allogrooming in wild bonnet macaques, Macaca radiata. Philosophical
Transactions of the Royal Society of London, B, 353, 619–631.
Skinner, B. F. (1938). The Behavior of Organisms: An Experimental Analysis. Oxford,
UK: Appleton-Century.
Snowdon, C. T. & M. Hausberger (eds.). (1997). Social Influences on Vocal Develop-
ment. Cambridge, UK: Cambridge University Press.
Taylor, R. C., B. A. Klein, J. Stein, & M. J. Ryan. (2008). Faux frogs: Multimodal
signalling and the value of robotics in animal behaviour. Animal Behaviour,
76, 1089–1097.
Tinbergen, N. (1963). On aims and methods of ethology. Zeitschrift für Tierpsychol-
gie, 20, 410–433.
348 ANIMAL BEHAVIOR

Trivers, R. (1985). Social Evolution. Menlo Park, CA: Benjamin Cummings.

Wade, M. J. (2001). Maternal effect genes and the evolution of sociality in haplo-
diploid organisms. Evolution, 55, 453–458.
Wasserman, S. & K. Faust. (1994). Social Network Analysis. Cambridge, UK:
Cambridge University Press.
West, M. J. & A. P. King. (1987). Settling nature and nurture into an ontogenetic
niche. Developmental Psychobiology, 20, 549–562.
West, M. J. & A. P. King. (1988). Female visual displays affect the development of
male song in the cowbird. Nature, 334, 244–246.
West, M. J. & A. P. King. (2008). Deconstructing innate illusions: Reflections on
nature-nurture-niche from an unlikely source. Philosophical Psychology, 21,
383–395.
West, M. J., A. P. King, & D. J. White. (2003). The case for developmental ecology.
Animal Behavior, 66, 617–622.
West, M. J., A. P. King, D. J. White, J. Gros-Louis, & G. Freed-Brown. (2006). The
development of local song preferences in female cowbirds (Molothrus ater):
Flock living stimulates learning. Ethology, 112, 1095–1107.
West-Eberhard, M. J. (1983). Sexual selection, social competition, and speciation.
Quarterly Review of Biology, 58, 155–183.
White, D. J., A. S. Gersick, G. Freed-Brown, & N. Snyder-Mackler. (2010a). The
ontogeny of social skills: Increasing social complexity enhances reproductive
success in adult cowbirds. Animal Behaviour, 79, 385–390.
White, D. J., A. S. Gersick, & N. Snyder-Mackler. (2012) Social networks and the
development of social skills in cowbirds. Philosophical Transactions of the Royal
Society of London, B, 367, 1892–1900.
White, D. J., J. Gros-Louis, A. P. King, M. A. Papakhian, & M. J. West. (2007).
Constructing culture in cowbirds (Molothrus ater). Journal of Comparative
Psychology, 121, 113–122.
White, D. J., A. P. King, & S. D. Duncan. (2002a). Voice recognition technology as
a tool for behaviour research. Behavior Research Methods, Instruments and Com-
puters, 34, 1–5.
White, D. J., A. P. King, & M. J. West. (2002b). Facultative development of court-
ship and communication in juvenile male cowbirds (Molothrus ater). Behav-
ioral Ecology, 13, 487–496.
White, D. J., A. P. King, M. J. West, J. Gros-Louis, & E. M. Tuttle. (2010b). Effects
of singing on copulation success and egg production in brown-headed cow-
birds, Molothrus ater. Behavioral Ecology, 21, 211–218.
White, D. J. & V. A. Smith. (2007). Testing measures of animal social association by
computer simulation. Behaviour, 144,1447–1468.
Whitehead, H. & S. Dufault. (2008). Techniques for analyzing vertebrate social
structure using identified individuals: Review and recommendations. Advances
in the Study of Behavior, 28, 33–74.
Williams, G. C. (1974). Adaptation and Natural Selection. Princeton, NJ: Princeton
University Press.
THE INDIVIDUAL IN CONTEXT 349

Wilson, E.O. (1975). Sociobiology: The New Synthesis. Cambridge, MA: Harvard
University Press.
Wittig, R. M., C. Crockford, J. Lehmann, P. L. Whitten, R. M. Seyfarth, & D. L.
Cheney. (2008). Focused grooming networks and stress alleviation in wild
female baboons. Hormones and Behavior, 54, 170–177.
Wolf, J. B., E. D. Brodie, III, & A. J. Moore. (1999). Interacting phenotypes and the
evolutionary process. II. Selection resulting from social interactions. American
Naturalist, 153, 254–266.
Zeigler, H. P. & P. Marler (2008). Neuroscience of Birdsong. Cambridge, UK:
Cambridge University Press.
This page intentionally left blank
 11

A Time for Every Purpose: Biological Clocks

and Animal Behavior
Megan Hastings Hagenauer and Sean P. Bradley

INTRODUCTION
What comes to mind when you think of time? All too frequently, it seems the
response is: “I wish I had more of it.” In the modern world, time is a nuisance,
a limited quantity that constantly has to be divided by schedules and fought
against to complete deadlines.
Now, in your mind, take a moment to step out of your busy world and into
the summer woods near my hometown in southern Wisconsin. Here there are
no clocks, no deadlines. Time is the rhythm of life.
At sunrise, you wake to a chorus of birds singing. If you go for a walk, you
might spot a deer or rabbit in the fresh dew and mist or hear a splash in a pond
as fish catch insects. By midmorning the birds have quieted, squirrels run up
and down tree trunks, and bees and hummingbirds buzz between flowers. As
the day gets hotter, the animals slow down, and cicadas drone in the trees. A
hawk soars lazily overhead. In the evening, the air cools and you find yourself
swatting away more mosquitoes. The rabbit and deer reemerge. The chirping
of crickets and the croaking of bullfrogs fill the air. At nightfall, the dry leaves
on the ground rustle with nightcrawlers, and every once in a while you spot
glowing eyes in the darkness—a raccoon slinking into the bushes. You may
hear the howling of coyotes or the hoot of an owl.
352 ANIMAL BEHAVIOR

In the woods, the seasons have more meaning than they do in a town or
city. Winter is a time of short days, frigid temperatures, and little food,
whereas summer is a time of relative abundance. For each species, there is an
optimal time to nest and reproduce, to store up food or body fat, to grow or
shed fur and feathers, to hibernate or migrate south.
In this chapter, we will examine these rhythms in animal behavior using
Tinbergen’s (1963) four great questions. Overall, we will focus on daily
rhythms in animal behavior, as there is a greater body of literature regarding
their mechanisms, but we will briefly touch on seasonal and lunar rhythms
near the end of the chapter.

ULTIMATE MECHANISMS (PHYLOGENY): HOW DID THESE RHYTHMS IN ANIMAL

BEHAVIOR EVOLVE?
Daily and seasonal rhythms in animal behavior reflect a species’ temporal
niche. A niche is a description of a species’ position within an ecosystem—
the particular set of biological and physical conditions to which the species
has adapted in order to survive and reproduce. Therefore, a temporal niche
is the species’ typical daily or seasonal pattern of activity, including the timing
of foraging, reproduction, and rest. Examples of temporal niches include
diurnality (day-active species, such as hummingbirds and squirrels),
nocturnality (night-active species, such as earthworms, raccoons, and owls),
and crepuscularity (species that are active during the transition times of dawn
and dusk, such as mosquitos and rabbits; Figure 11.1). It is rare for an organ-
ism to be truly cathemeral and have activity distributed approximately evenly
across the 24-hour cycle, although some species, such as the lemur, exhibit this
temporal phenotype.
Why would species evolve to be active for only part of the day or part of the year?
Wouldn’t it be advantageous to forage or hunt all of the time? As stated nicely by
Noga Kronfeld-Shor and Tamar Dayan (2008), being active in the same envi-
ronment at different times of the day or the year exposes animals to very differ-
ent environmental challenges, including not only the obvious difference
between light and dark but also differences in ambient temperatures, humid-
ity, food availability, predation risk, prey availability, competitors, mates,
and more.
By evolving a specific daily and seasonal activity pattern, species become
specialists: efficient competitors under a particular environmental condition.
At the times when they are not active, they spend time on quiet recuperation:
sleep, digestion, lactation, growth, recovery, and repair. For this reason, in
most species you can see physical adaptations related to a particular temporal
Figure 11.1. Temporal niche (adapted from Schmitz & Motani, 2011). The frequency
of diurnal (white), nocturnal (black), and crepuscular/cathemeral (gray) activity pat-
terns among avian species, reptiles, and mammals (as estimated from a sample of
1,401 species).
354 ANIMAL BEHAVIOR

niche, including color vision and heat tolerance in diurnal species, and low-
light vision and auditory acuity in nocturnal species (Kronfeld-Shor & Dayan,
2008).
Daily rhythms are thought to have emerged 3 billion years ago with the
evolution of cyanobacteria, one of the first life forms on earth. These rhythms
allowed the cyanobacteria to optimize photosynthetic processes and growth in
relationship to the solar cycle. As the solar cycle provides a strong selective
pressure for many other organisms, daily rhythms were highly evolutionarily
conserved and are also thought to have evolved independently multiple times
(Yerushalmi & Green, 2009). These rhythms are now found in almost every
organism on earth, including fungi, plants, fish, reptiles, birds, and mammals
(DeCoursey, 2004). Thus, it is likely that there have been daily rhythms in
behavior since the evolution of species capable of behavior!
Nevertheless, up until recently it was thought that the full diversity of
rhythmic patterns in animal behavior did not emerge until mammals evolved
200–250 million years ago. The dinosaurs, which were the dominant terres-
trial vertebrates of the Triassic and Jurassic periods, were considered to be
cold-blooded (ectothermic) and therefore likely to use solar radiation for
behavioral temperature regulation like many modern lizards. Because of this
limitation, it was thought that warm-blooded (endothermic) mammalian spe-
cies were able to compete by making use of the nocturnal niche (Smale et al.,
2003). However, new research suggests that dinosaurs may actually have been
warm-blooded, which would allow them to easily occupy a variety of temporal
niches. This hypothesis was bolstered by a recent study showing that the eye-
bone structure of dinosaurs indicates that they occupied a similar distribution
of temporal niches as modern terrestrial vertebrates, with an almost equivalent
number of nocturnal (large-eyed) species, diurnal (small-eyed) species, and
crepuscular/cathemeral species. Like modern vertebrates, this study found that
the temporal niche of dinosaurs was related to their ecological niche: flying
species were primarily diurnal, similar to modern birds, and predators were
primarily nocturnal, similar to modern mammalian carnivores. Large herbi-
vores appeared to be entirely cathemeral, most likely due to their need to for-
age for many hours each day to satisfy metabolic requirements and to avoid
overheating during the middle of the day (Schmitz & Ryosuke, 2011). Thus
the imagery in the movie Jurassic Park is likely to have been accurate: velocir-
aptors did hunt at night, even while other species, such as pterodactyls, were
active during the day.
Despite this new interpretation, the common ancestor of mammals is still
likely to have been nocturnal, and the majority of modern mammals exhibit
this activity pattern. Diurnality, on the other hand, appears to have evolved
A TIME FOR EVERY PURPOSE 355

multiple times in separate independent lineages. For example, a literature sur-

vey of the activity patterns of 700 modern rodent species found that the tax-
onomy of a rodent species (family, subfamily, tribe) largely predicted the
species’ activity pattern. The majority of species exhibited a nocturnal activity
pattern, but over the course of evolution this activity pattern would have
needed to switch (invert) a minimum of seven times to produce the full tax-
onomy of activity patterns observed today. These evolutionary shifts in activ-
ity pattern could have been driven by extreme selective pressures, such as
cold climates (which encourage diurnality because it is warmer during the
day) and subterranean habitat (which does not strongly select for either diur-
nal or nocturnal behavior because subterranean rodents have limited exposure
to sunlight). However, in general it seems that temporal niche was highly con-
served on an evolutionary scale, and even closely related species inhabiting
similar habitats did not develop different activity patterns to reduce competi-
tion, most likely due to constraints such as eye structure and thermoregulation
(Roll et al., 2006).
There are exceptions to this rule—despite evolutionary constraint, occa-
sionally the activity pattern of a species will shift in response to both intra-
and interspecies competition, a process called temporal partitioning. For
example, many bullfrog species are cannibalistic. In order to avoid being eaten,
small, vulnerable frogs are most active during the day, whereas the older,
threatening adults are active at night (Pizatto et al., 2008). Hummingbirds
also divide up the day so that dominant individuals are active at a different
time than are subordinates. In this case, dominant hummingbirds feed earlier
in the day than other birds, essentially claiming “first dibs” on good sources of
nectar. By becoming active later in the day, subordinate hummingbirds are
able to reduce the probability of aggressive encounter (Craig & Douglas,
1984). This form of temporal partitioning, in which individuals divide up
the normal active period (in this case, daytime) into two niches (dominant
and subordinate) is far more common than the extreme temporal partitioning
(diurnal/nocturnal) observed in bullfrogs, most likely because of evolutionary
constraints (Kronfeld-Shor & Dayan, 2008).
Under extreme circumstances, competition between species can also drive
individuals or a species to switch their time of activity. One example of this
comes from Kronfeld-Shor’s research on two rodent species in a rocky Israeli
desert near the Dead Sea: the common spiny mouse (Acomys cahirinus) and
golden spiny mouse (A. russatus). In the wild, these closely related species over-
lap in everything from microhabitat to food sources. In order to facilitate coex-
istence, they maintain activity at different times of the day: the common spiny
mouse is active at night (like most desert rodents), and the golden spiny mouse
356 ANIMAL BEHAVIOR

is active during the day. However, the golden spiny mouse still shows physio-
logical signs of a nocturnal legacy in its retina and thermoregulation. If the
common spiny mouse is systematically removed from the habitat using trap-
ping, the golden spiny mouse quickly regains nocturnal activity. The golden
spiny mouse also develops nocturnal behavior immediately when brought into
a laboratory, away from the competitive constraints of a natural habitat.
Kronfeld-Shor and Dayan (2008) suggest that their species may represent an
intermediate evolutionary step in the development of diurnal behavior.

PROXIMATE MECHANISMS (CAUSATION): HOW ARE RHYTHMS IN ANIMAL

BEHAVIOR GENERATED?
Logically, we can consider two general hypotheses for the origin of behav-
ioral rhythms: (1) Biological rhythms are a response to a rhythmic natural
environment (“I wake in the morning because the sun shines through my win-
dow”). (2) Biological rhythms are generated by an internal clock (“I wake in
the morning because my body knows that it is morning”).
Although daily and seasonal rhythms in animal behavior were clearly an
important focus for early human hunters and agriculturalists, the earliest
known scientific attempt to determine the origin of a biological rhythm was
by Jean-Jacques d’Ortous de Mairan in 1729. He observed that the leaves of
a mimosa plant opened to the sun during the daytime and closed at night even
when the plant was continually enclosed in a darkened closet (Figure 11.2).
This meant that the rhythm of the leaves was unlikely to be a direct response
to the sun. Another French scientist, Henri-Louis Duhamel du Monceau, fol-
lowed up on his work in 1758 and went to great lengths to show that the daily
rhythms of the mimosa plant were not due to environmental influences, such
as daily fluctuations in environmental temperature. At one point, in an effort
to create a consistently dark and temperature-controlled environment, he
wrapped a mimosa plant in a blanket, placed it in a leather trunk, and stashed
the trunk in a dark wine cellar! The mimosa plant still continued to open and
close its leaves rhythmically, suggesting that the rhythms were generated by an
endogenous (internal) clock. Despite these intensive efforts, whether biologi-
cal rhythms could be truly endogenous or were simply driven by subtler envi-
ronmental time cues (such as “atmospheric electricity”!) was still vigorously
debated into the nineteenth and early twentieth centuries (Zivcovic, 2008).
During the twentieth century, it was found that animals also maintained
daily rhythms when isolated from daily environmental time cues in a labora-
tory. Interestingly, under constant conditions, these rhythms began to drift
(free-running rhythms), such that for any particular individual, the rhythm
A TIME FOR EVERY PURPOSE 357

Figure 11.2. Circadian rhythms in plants (adapted from http://www.hhmi.org/biointer

active/museum/exhibit00/02_1.html). An illustration of a plant opening its leaves
during the day and closing its leaves at night (illustration originally by Charles Darwin
in The Power of Movement in Plants).

would start either progressively earlier each day or progressively later. This
meant that the internal clock of the animals had a period, or cycle length, that
was not exactly 24 hours (Figure 11.3; Aschoff, 1965).
In the 1960s, Jurgen Aschoff demonstrated that this was true for the inter-
nal clock of humans as well. To do this experiment, he made clever use of a
soundproof underground bunker built in Germany during World War II.
Volunteers, many of whom were students who wanted to cram for an exam,
agreed to isolate themselves in the bunker for up to 3 or 4 weeks. During this
time, they had no access to clocks or sunlight, although they could turn on the
358 ANIMAL BEHAVIOR

Figure 11.3. Evidence for an internal clock: free-running period (adapted from Aschoff,
1965). The free-running activity rhythms of a chaffinch (Fringilla coelebs; left) and
human (right) from two separate experiments. At the top, the gray bars indicate the
original laboratory daily light cycle (gray ¼ imposed darkness). At the beginning of
the experiment, both the chaffinch and human subject were active during the lighted
part of the day, as illustrated by the thin horizontal lines. During this part of the
experiment, their activity is said to have been entrained by the light cycle. Then, after
several days, the finch was placed into constant conditions (dim light) and the human
subject was allowed to choose his or her own daily light cycle without the help of any
external time cues (no access to clocks or the outside world). At this point, the activity
rhythms of both the chaffinch and human subject started to free-run and drifted
progressively later each day relative to the outside world. This drift means that their
internal clocks were overestimating the length of a day and therefore had a free-
running period of greater than 24 hours. (The drift was so great over the full course of
the experiment that the x axis had to include more than 24 hours of clock time to
depict the drift!) At the end of the experiment, a laboratory light cycle was reimposed,
and the chaffinch and human subject returned to having an entrained activity rhythm
with a 24-hour cycle again.

lights in the bunker as they wished. Their only connection to the outside
world was an antechamber through which food and supplies were secretly
delivered (including a requisite daily bottle of German beer). During their
time in the bunker, the subjects were monitored for time perception and for
a variety of physiological rhythms, including daily rhythms in body tempera-
ture, sleep, and urinary metabolites. Aschoff himself lived in the bunker for
10 days and wrote about the feeling of timelessness.
After a great curiosity about “true” time during the first 2 days of bunker life,
I lost all interest in this matter and felt perfectly comfortable to live “timeless.”
. . . I had difficulty in deciding whether I had slept long enough. On day 8, I got
A TIME FOR EVERY PURPOSE 359

up after only 3 hours of sleep. Shortly after breakfast I wrote in my diary:

“Something must be wrong. I feel as if I am on dogwatch.” I went to bed again
and started the day anew after three more hours of sleep (Aschoff, 1965,
p. 1429).

As in earlier animal experiments, the circadian rhythms of the human sub-

jects drifted, with subjects tending to wake up slightly later each day. Aschoff
(1965) himself was surprised to learn that in 10 days his waking-up time had
drifted to 3 p.m. This drift meant that Aschoff’s circadian system was overesti-
mating the length of a day—it had a free-running period close to 25 hours.
Other subjects had free-running periods as short as 23.8 hours or as long as
25.9 hours. The presence of an individualized free-running period was one
of the strongest pieces of evidence in favor of an internal clock because if the
subjects’ physiological rhythms were simply a response to subtle daily environ-
mental time cues, then the rhythms universally would have had a 24-hour
period (Aschoff, 1965).
If the body’s internal clock drifts by a little bit each day, how can it tell time
accurately? Patricia DeCoursey, Colin Pittendrigh, and Serge Daan conducted
a series of experiments in the 1960s and 1970s that indicated that the internal
clock of animals is most sensitive to time cues (also called zeitgebers) at times
of day when the cues are not normally expected to be present. Their experi-
ments focused on daily light exposure because the solar day has been the most
predictable daily time cue for all of evolution. To do these experiments, they
placed animals into constant darkness so that their rhythms would begin to
free-run. The animal’s free-running rhythm was then used to determine the
animal’s “subjective day” (i.e., when the animal assumed daytime to be, based
on its own internal clock) and “subjective night.” They found that animals
kept in constant darkness would shift their rhythms later in response to brief
light exposure during the animals’ subjective evening (“there is light, so it
must still be day”), whereas the animals would shift their rhythms earlier in
response to light exposure during the subjective morning (“there is light, so
the day must already be starting”). In contrast, light exposure during the
middle of the subjective day (i.e., when the animals would already expect there
to be daylight) produced no shift at all (Figures 11.4 and 11.5). Through this
mechanism, the endogenous rhythm of the animals could be synchronized
(entrained) by daily time cues to follow a 24-hour rhythm. This process was
called photic entrainment (DeCoursey, 1960; Pittendrigh & Daan, 1976b).
These endogenously generated rhythms that could be entrained to a 24-
hour period by the solar day were granted an official name: circadian
rhythms. This name came from the Latin words circa, meaning “around,”
and diem, meaning “day”—literally, “around a day.”
360 ANIMAL BEHAVIOR

Figure 11.4. An illustration of commonly used circadian terminology. Depicted on top

is a simple sinusoidal model of a subject’s behavioral or physiological rhythm (e.g.,
rest and activity), and depicted below is the environmental cycle to which the subject
was exposed (e.g., a laboratory light-dark cycle).

Why wouldn’t animals just evolve an internal clock with a period of exactly
24 hours? Pittendrigh and Daan’s (1976a) experiments suggested that it was
adaptive for animals to have an endogenous rhythm with a period that was
not exactly 24 hours because the constant process of correction allowed the
rhythms to better track the changes in photoperiod (the duration of light expo-
sure each day) that occurred as the seasons changed.
Once it was clearly demonstrated that circadian rhythms were generated by
an internal clock, the hunt began to locate this clock with the body. In order
to locate the internal clock in the mammalian brain, Irving Zucker’s labora-
tory reasoned that because the solar day is the primary time cue that entrains
circadian rhythms, a daily clock would require light input to function prop-
erly. Therefore, using rats (Rattus norvegicus), his laboratory searched for
regions in the brain that received direct projections from the retina and
lesioned (damaged) them. To determine if there was an effect on circadian
function, they examined daily behavioral rhythms in water consumption and
wheel-running activity. They expected that if the lesions disrupted circadian
entrainment, these behavioral rhythms might free-run even when the animals
were housed in a normal daily light cycle because the internal clock would no
longer be reset by light. If the lesions disrupted circadian rhythm generation
A TIME FOR EVERY PURPOSE 361

Figure 11.5. Exposure to light has the largest effect on circadian timekeeping at night
(adapted from Pittendrigh & Daan, 1976b). (a) Depicted is the free-running
rhythm of a golden hamster (Mesocricetus auratus) kept in constant darkness. The
hamster had a free-running period that was less than 24 hours long, so that each day
the hamster’s activity rhythm drifted a little bit earlier relative to normal clock time
(depicted on the x axis). On the eighth day of the experiment, the hamster was
exposed to a brief 15-minute pulse of light in the subjective evening (i.e., when the
hamster expected evening to be based on its own internal clock, as indicated by its
free-running activity). This caused the hamster’s activity rhythm to shift several hours
later (a phase delay). (b) After repeating this same experiment many times, the
researchers were able to construct a graph called a phase response curve that illustrates
the effect of brief light exposure at different times of the subjective day. This subjec-
tive time is also called circadian time and is determined by examining the animals’
free-running activity rhythm. For a nocturnal animal, a circadian time of 12 hours
would be when the animal begins its daily activity (the subjective evening), whereas a
circadian time of 20 hours would be early subjective morning. The hamster responded
to light exposure during the subjective evening by shifting its activity later (phase
delay, depicted as a negative phase shift on the y axis) and responded to light exposure
during the subjective morning by shifting its activity earlier (phase advance, depicted
as a positive phase shift on the y axis). Light exposure during the middle of the sub-
jective day had little effect, so this time period is referred to as the dead zone.

itself, then the daily behavioral rhythms would disappear completely (arrhyth-
mia), especially under constant conditions.
After several failed attempts to locate the circadian clock by destroying parts
of the visual pathway, they decided to lesion the suprachiasmatic nucleus
(SCN), which is a tiny region in the hypothalamus at the base of the brain.
The SCN was named due to its location above (“supra”) where the left and
right optic nerves cross (“chiasm”) to send visual information to the opposite
hemisphere. Robert Moore’s laboratory had recently demonstrated that the
SCN was not only located above the optic chiasm but actually received its
own direct neural projections from the retina (called the retinohypothalamic
362 ANIMAL BEHAVIOR

tract; Moore & Lenn, 1971). Curt Richter (1965) had also demonstrated that
damage in the hypothalamus could influence behavioral rhythms. Therefore,
it was possible that the SCN was the brain’s internal clock, or circadian pace-
maker. When Zucker’s laboratory lesioned the SCN, they found that the rats
did indeed lose daily behavioral rhythms (Stephen & Zucker, 1972). How-
ever, at first they did not believe the results—in the laboratory, the experiment
was actually nicknamed “Fred’s Folly” after Friedrich Stephen, the graduate
student performing the experiments (Weaver, 1998). In the end, Fred was
vindicated—further replications of the experiment showed that daily behav-
ioral rhythms disappeared following lesioning of the SCN even while the
behavioral drive was preserved. The rats still maintained a similar overall
amount of activity, food, and water consumption. Zucker and Stephen had
located the circadian pacemaker! However, unbeknownst to them, Moore’s
laboratory had also followed up on their discovery of the retinohypothalamic
tract by performing a series of knife-cut and lesion studies aimed at the SCN
(Moore & Eichler, 1972). They discovered that daily rhythms in adrenal hor-
mone levels in the blood disappeared following lesioning and decided to pub-
lish the results at the same time as the Zucker laboratory (Moore & Eichler,
1972, Stephen & Zucker, 1972). Both groups are now credited for having dis-
covered that the SCN is the location of the main circadian pacemaker in mam-
mals (Weaver, 1998).
While eliminating rhythmicity by lesioning the SCN demonstrated that the
SCN was necessary for normal circadian function, further evidence was
needed to determine whether the SCN was sufficient to produce rhythms in
behavior. It remained possible that daily rhythms were produced by several
brain structures, including the SCN, or that there were pathways traveling
through the SCN that were necessary for maintaining daily rhythmicity. This
question could not be fully addressed until several years later following the dis-
covery of a mutant hamster (Mesocricetus auratus) with an exceptionally short
free-running period of 20 hours. This mutant hamster was named the “tau
mutant hamster” because free-running period is sometimes denoted with the
Greek letter  (tau). The tau mutant was then bred to produce an entire line
of hamsters with short free-running periods. In a bold move, Martin Ralph
and Michael Menaker used these mutant hamsters to demonstrate that they
could restore behavioral rhythmicity following SCN lesions by transplanting
healthy fetal SCN tissue into the brain. Even more exciting, they found that
the free-running period of the host hamster matched that of the tissue donor
—an SCN transplant from a tau mutant hamster caused the normal host ham-
ster to develop a free-running period of 20 hours, whereas an SCN transplant
from a normal hamster caused a tau mutant host hamster to develop a more
A TIME FOR EVERY PURPOSE 363

Figure 11.6. SCN transplants restore circadian rhythmicity in SCN-lesioned hamsters

(Mesocricetus auratus) (adapted from Silver et al., 1996). Shown on the left is a pho-
tograph of an SCN transplant encapsulated in a membrane and placed into the third
ventricle of the brain. Shown on the right is an activity record for a tau mutant ham-
ster. First, the hamster was housed in constant darkness to confirm that he had a very
short free-running period of 21.4 hours due to his mutation (top). To make the
activity patterns easier to see, the activity record is double-plotted, which means that
each day is plotted twice (which is why the x-axis includes 48 hours of clock time).
Then the hamster received an SCN lesion (SCN-X) and became arrhythmic until it
received an SCN transplant from a wild-type tissue donor. Several weeks after the
transplant it re-developed a normal circadian rhythm with a free-running period of
roughly 24 hours.

typical free-running period of close to 24 hours (Ralph et al., 1990; Figure

11.6). Thus, the 24-hour periodicity of circadian rhythms was a property of
the SCN itself.
How are circadian rhythms generated within the SCN? To answer this ques-
tion, David Welsh and Steven Reppert dissected out SCN neurons and kept
364 ANIMAL BEHAVIOR

them alive in a dish (cell culture) so that they could record the neurons’ elec-
trical activity (electrophysiology). They found that most SCN neurons could
maintain daily rhythms in electrical activity even when isolated from other
neurons and from the rest of the brain (Welsh et al., 1995). The cells them-
selves were circadian pacemakers! When isolated, each cell had its own free-
running period (“danced to the beat of its own drummer”). In contrast, when
cultured with other SCN neurons, the rhythms in the different cells were able
to synchronize by reciprocal communication into a single, strong daily
rhythm. This process of synchronization was called oscillator coupling.
To determine how circadian rhythms were generated within individual
cells, Joseph Takahashi and other chronobiologists (biologists who study clock
phenomena) used a technique called mutagenesis to cause genetic mutations
in a variety of species, including fruit flies, bread mold, plants, and mice. Then
they screened the progeny (offspring) for individuals with disrupted circadian
rhythms. If an individual was found to be active at an odd time of day or to
develop arrhythmia under constant conditions instead of a normal free-
running rhythm, they would sequence the individual’s genome for the identity
of the responsible gene. Using this technique, they developed a long list of
potential clock genes (Weaver, 1998).
It turned out that these clock genes generated circadian rhythms within
individual pacemaker cells in the SCN by means of a molecular cycle. As
you may know, in order for genes to be expressed in the body, a polymerase
enzyme binds to the promoter region of the gene and transcribes the gene
to form messenger RNA (mRNA). This mRNA is then translated to con-
struct proteins. Chronobiologists found that a 24-hour cycle develops within
circadian pacemaker cells because proteins produced by particular clock genes
inhibit their own transcription, leading to daily cellular oscillations. The core
feedback loop consists of a positive arm, which contains proteins that drive tran-
scription (BMAL1 and CLOCK), and a negative arm, which contains proteins
that inhibit transcription (PER, CRY, REVERBα). To initiate the cycle,
BMAL1 and CLOCK proteins drive the transcription of Per, Cry, and Reverbα.
The protein REVERBα then feeds back to inhibit the transcription of Bmal1,
whereas the PER and CRY proteins (PER1, PER2, PER3, CRY1, CRY2) inhibit
their own transcription and the transcription of Reverbα (Figure 11.7; Bae et al.,
2001; Zheng et al., 2001; for review see Hastings & Herzog, 2004).
The identification of the clock-gene feedback loop in the SCN helped
explain circadian entrainment to the solar day. As discussed earlier, chronobi-
ologists had discovered that circadian entrainment occurs because the internal
clock is most sensitive to time cues at times of day when the cues are not
expected to normally be present. For example, light exposure during the
A TIME FOR EVERY PURPOSE 365

Figure 11.7. Circadian rhythms are generated by clock genes. Depicted is an illustration
of the transcriptional-translational feedback loop that generates circadian rhythms in
the SCN. Arrows with a plus sign (þ) indicate that an element drives the transcription
of another element, whereas arrows with a negative sign (−) indicate transcriptional
inhibition.

evening shifts daily rhythms later, and light exposure during the early morning
shifts rhythms earlier, whereas light exposure during the middle of the day has
little effect. Thus, the phase of the circadian system is constantly being “cor-
rected” so that it is aligned properly with the solar day.
Yasufumi Shigeyoshi and colleagues (1997) discovered that this enhanced
sensitivity of circadian rhythms to light at night was related to the clock gene
Per1. To perform this experiment, Shigeyoshi used a technique called in situ
hybridization, which is a manner of visualizing the amount of mRNA present
in a brain region using a radioactive probe. The way that in situ hybridization
works is due to mRNA being single-stranded: you can construct probe mRNA
that is complementary to the mRNA of interest (Per1) using radioactive
nucleotides. Then, when you apply the probe to brain tissue, the probe binds
to the Per1 mRNA in the tissue to make double-stranded mRNA. Because the
probe is radioactive, wherever it binds in the tissue also becomes radioactive.
This means that you can quantify radioactivity to measure Per1 mRNA levels.
Using this technique, Shigeyoshi and colleagues (1997) measured how Per1
in the SCN respond to light exposure in mice (Mus musculus). They placed
the mice into constant dark conditions so that their activity rhythms started
to free run, then they sacrificed them at different times of the day so that they
could measure Per1 mRNA in their SCNs. As expected (Figure 11.8), they
found that Per1 levels had a beautiful daily rhythm in the SCN, with a
366 ANIMAL BEHAVIOR

Figure 11.8. Clock-gene rhythms are affected by environmental time cues (adapted
from Shigeyoshi et al., 1997). Shown at the top is the daily rhythm of Per1 expres-
sion in the SCN. The SCN is indicated by the two arrows. The time points with the
brightest labeling have the most Per1 mRNA. The highest Per1 expression occurred at
circadian time 4 (CT4; subjective midday), and the lowest from CT12 to CT20
(subjective night). The picture with the star is tissue from a mouse that had been
briefly exposed to 30 minutes of light in the middle of the subjective night. This brief
light exposure caused the amount of Per1 in the SCN to increase dramatically.
Depicted below it are graphs that illustrate the daily rhythm of Per1 under constant
darkness and the induction of Per1 in response to brief light exposure during the
night. After this brief increase in response to light, the rhythm in Per1 shifts so that it
now peaks at a later time.

peak in the middle of the day (after transcription had been driven by Bmal1
and Clock) and low levels during the night (when transcription was suppressed
by Per and Cry protein feedback). However, if the mice were exposed to a brief
(30-minute) pulse of light during the night, Per1 levels suddenly increased,
and this increase caused the overall rhythm in Per1 to shift. For example, if
the light exposure occurred in the early morning, then Per1 would peak early
that day, and the circadian rhythm would shift earlier. If the light exposure
occurred in the evening, then Per1 would peak later that day, and circadian
rhythms would shift later. In contrast, in the middle of the subjective day
(keep in mind, these mice were free-running under constant dark conditions!)
A TIME FOR EVERY PURPOSE 367

light had no effect: the Per1 rhythm was already peaking at this time, so the
circadian rhythm in Per1 was not altered. In this manner, light information
from the retinohypothalamic tract could correct the phasing (timing) of the
clock-gene feedback loop so that it was properly aligned with the solar day.
The identification of clock genes also paved the way for another exciting
discovery: the SCN is not the only circadian pacemaker in the body. To make
this discovery, Hajime Tei and Michael Menaker’s laboratory cleverly created
a transgenic rat that had tissue that glowed (fluoresced) whenever Per1 levels
were elevated. This transgenic rat was created by inserting a gene for a fluores-
cent protein from fireflies (luciferase) after the promoter region for the gene
Per1 (Yamazaki et al., 2000). This meant that whenever Per1 transcription
was initiated, the luciferase gene was also transcribed and then translated into
fluorescent protein. The researchers dissected SCN tissue from this transgenic
rat and placed it into a cell culture dish. They were then able to measure
rhythms in “firefly” light emitted by the tissue as a way of determining when
Per1 was likely to be elevated each day. Astoundingly, they found that daily
rhythms in Per1-luciferase in the SCN persisted for up to 32 days in a dish!
More surprising, though, was the discovery that many other tissues fluoresced
in a rhythmic fashion, including the lungs, liver, and muscle. Nevertheless, the
rhythm in these tissues was not as strong as in the SCN, and petered out
(dampened) after only a couple of days in a dish (Yamazaki et al., 2000). This
dampening turned out to be due to individual pacemaker cells in the tissue
becoming uncoupled (desynchronized) so that they no longer produced a
strong, united rhythm.
These results suggested that the circadian system might actually consist of a
hierarchy of oscillators, with the SCN at the top serving as the “master pace-
maker,” which receives direct information about the solar day from the retino-
hypothalamic tract. The SCN could then send daily signals to local “slave
oscillators” in the brain and periphery to drive and coordinate their daily
rhythms. In this sense the SCN would function more as the conductor of an
orchestra rather than the sole source of circadian rhythm generation for the
body.
To test this hypothesis, their laboratory exposed their transgenic rats to a
“jet lag” scenario: they shifted the time that the lights were on in the animal
housing rooms either six hours earlier (similar to the jet lag someone would
experience if they traveled east from New York to France) or six hours later
(similar to traveling west from France to New York). Then they dissected vari-
ous tissues from the transgenic rats and observed how rapidly these tissues
responded to the time shift. As expected based on Shigeyoshi’s experiment
(Shigeyoshi et al., 1997), the SCN responded to the change in the light
368 ANIMAL BEHAVIOR

schedule almost immediately with a shift in the timing of Per1-luciferase

rhythms. However, the rest of the body lagged behind: rhythms in the skeletal
muscle and lungs took several more days to shift, and the liver was not com-
pletely shifted even after six days on the new lighting schedule (i.e., during
your first week after traveling to France, your liver may still think that you
are in New York!). They concluded that the disruption is likely to be the gen-
eral explanation of symptoms of jet lag, and, more importantly, the malaise
associated with shift-work schedules experienced by more than 20 percent of
the U.S. work force (Yamazaki et al., 2000).
In the end, they made several prophetic statements about the role of hierar-
chical oscillators in the circadian system:
The existence of distributed oscillators, whatever their properties and connec-
tions, is a common attribute of all circadian systems so far described in multicel-
lular organisms. This may be the simplest way to construct a system which
requires that many constituent rhythms have different phases, because the
phase of each rhythm can be adjusted by modifying the response properties of
the tissue’s own peripheral oscillator to an invariant signal from the central
oscillator [SCN] or from the environment. Such organization provides great
flexibility, both phylogenetically and ontogenetically, and as more is learned
we can expect to find that major adjustments of relative phase relationships
have been brought about by natural selection (e.g. in diurnal vs. nocturnal
organisms) and also occur in the course of development. (Yamazaki et al.,
2000, p. 684)

Each of their predictions has turned out to be true. In addition, it is impor-

tant to note that more recent studies have not only replicated the findings
indicating that there are circadian oscillators throughout the body but also
have shown that their influence is surprisingly profound. Using microarray
analysis, which is a method of determining the amount of mRNA present
for thousands of genes at a time, these studies indicated that 2 to 10% of all
mRNA in any given tissue is rhythmically expressed. These rhythms influence
everything from metabolism and hormone secretion to neurotransmitter sig-
naling and cellular repair (Akhtar et al., 2002; Duffield et al., 2002; Storch
et al., 2002).
From these results, it has been suggested that circadian rhythms play a role
in survival that is far more fundamental than simply driving overt rhythms in
behavior—they may serve to time complex physiological events within dispa-
rate tissues in the body. Like a group of bank robbers in a movie who syn-
chronize their watches before entering a bank so that all robbers do their
jobs in the correct sequence (distracting customers, disarming the security sys-
tem, opening the safe), our bodies may use the circadian system to coordinate
A TIME FOR EVERY PURPOSE 369

Figure 11.9. Disrupted circadian rhythms can lead to death (adapted from Davidson et
al., 2006). Elderly mice (Mus musculus) exposed to frequent “jet lag” have increased
mortality. Jet lag was imitated by shifting the laboratory light cycle every four days so
that it was either six hours earlier (phase advancing-like what you would experience if
you traveled east from New York to France) or six hours later (phase delaying). After a
month of this procedure, almost half of the mice subjected to phase-advancing jet lag
had died. In contrast, over 90 percent of the mice survived that had experienced no jet
lag or phase-delaying jet lag. In general the circadian system has an easier time
adapting to phase delays than phase advances.

such processes as ovulation, childbirth, immune response, and growth.

Indeed, there is now evidence from both humans and laboratory rodents that
when circadian rhythms become chronically disrupted, such as during shift
work or repetitive jet lag, there can be severe consequences: immune response
is deficient, fertility problems and cancer are more likely to develop, and
mortality rates increase (Davidson et al., 2006; Navara & Nelson, 2007;
Mahoney, 2010; Figure 11.9).

ULTIMATE MECHANISMS (ADAPTATION): ARE CIRCADIAN RHYTHMS NECESSARY FOR

SURVIVAL IN THE WILD?
Despite early evidence in favor of an internal circadian clock, it still took a
long time for the endogenous nature of circadian rhythms to be accepted
because, as summarized nicely by Menaker (1996), it seemed unnecessarily
complex. Why would natural selection favor an internal oscillator when nature
is replete in temporal cues?
370 ANIMAL BEHAVIOR

Some of the first experiments regarding the importance of endogenously

generated circadian rhythms came from one of the fathers of animal behavior
research, Karl von Frisch. Von Frisch demonstrated that honey bees used a
time-compensated solar compass to navigate to flowers. After discovering a
particularly good food source, the honey bees would return to the hive and
perform intricate motions called waggle dances that communicated to their
hive mates the direction of the food source in terms of the sun. Within the
dance, vertical movement up the comb represented flying in the direction of
the sun, and the duration of the “waggle” portion of the dance represented
the distance to the food source. Von Frish’s results were striking because he
found that over the course of a day, the angle of these dances would change
to compensate for the movement of the sun even when the dancer had not left
the darkness of the hive for hours. Even more impressively, during the night
the dancing took into account the movement of the sun on the other side of
the earth (Brady, 1982).
Gustav Kramer, an avian biologist, demonstrated a similar form of time-
compensated celestial navigation in birds. Kramer noticed that at the time
of year when migration would normally occur in the wild, captive birds
housed in an outside aviary developed an intense restlessness, accompanied
by increased attempts to move in the normal direction of migration. By plac-
ing European starlings (Sturnus vulgaris) in circular cages that blocked out
other landmarks, Kramer demonstrated that the birds were likely able to
derive this directionality from the location of the sun. However, it remained
possible that the birds were somehow navigating by sensing the direction of
the earth’s magnetic field, similar to a compass. To rule out this possibility,
Kramer used a set of mirrors to change the apparent location of the sun and
found that the captive starlings changed their behavior accordingly (Brady,
1982).
It was clear that the birds must not simply be navigating using the location
of the sun in the sky because the sun moves as the day progresses. They would
have to compare the location of the sun to an internal clock. Kramer illus-
trated this time compensation by training starlings at the same time each day
to look for food in a particular direction within the testing cage (for example,
looking for food on the east side of the cage in the morning). Then he brought
the starlings back to the room at a different time of day, when the sun was in a
different location in the sky, and found that the birds took into account how
the sun would have moved when they searched for their food (i.e., in the after-
noon they continued to look for food on the east side of the cage, even though
the sun was now located in the west; Brady, 1982). Unfortunately, Kramer
was unable to follow up on these experiments because in 1958 he slipped
A TIME FOR EVERY PURPOSE 371

and fell to his death while trying to catch wild rock pigeons (Columba livia;
Globig, 2007). Since then, other researchers have applied his methods to
examine avian navigation at night. Using a planetarium, they showed that
songbirds could navigate using the movement of star constellations across
the night sky in reference to their own internal clocks (Brady, 1982).
Perhaps the best study indicating the importance of endogenous circadian
rhythms for ensuring a species’ survival in the wild came from the DeCoursey
laboratory (DeCoursey et al., 2000). DeCoursey’s five-member team spent
4,200 hours tracking the daily activity rhythms of wild chipmunks (Tamias
striatus) in a 4-hectare area of oak forest in the Allegheny Mountains of
Virginia. They chose chipmunks as their species of interest because chip-
munks show a high degree of den fidelity. This meant that the chipmunks
tended to live in the same burrow system for a large part of their lives, allowing
DeCoursey’s team to track their daily activity rhythms and survival rates.
Using live traps baited with peanut butter, DeCoursey’s team caught each
of the 126 adult chipmunks who lived in the project site and marked them
using toe clipping or ear tagging. Fifty-four of these individuals were then tem-
porarily brought back to the laboratory for circadian rhythm monitoring and
SCN surgery. Half of the animals received an SCN lesion, whereas the other
half received a similar control surgery in which the SCN was not destroyed
(called a “sham surgery”). Following surgery, DeCoursey’s lab confirmed that
the animals that received SCN lesions were arrhythmic but that they did not
have any unwanted side effects of surgery, such as visual impairment. Each
of these individuals was then fitted with a little radio collar with a 2-cm
antenna projecting over its head and returned to the wild (repatriated).
The process of repatriation was more difficult than expected: due to over-
crowding in the territory, the project chipmunks’ burrows had been taken
over during their absence. The new residents resisted their entry, chasing the
original owners away and fighting with them. This restlessness attracted local
predators. One of the project chipmunks was eaten almost immediately:
within an hour of repatriation, DeCoursey’s team observed the radio collar sig-
nal moving rapidly away from the study site, and the next day the empty collar
was found on the ground with tooth marks on the antenna. Distressingly,
during the next two weeks, over 40 percent of the original project chipmunks
were killed as territories were reestablished.
Despite this setback, after the initial disturbance period, the subtle impor-
tance of a functional circadian clock became quite clear. To examine the chip-
munks’ activity rhythms in the wild, DeCoursey’s team performed 28-hour
censuses of all above-ground activity. During these marathon data-collection
sessions, the team would set up live traps near burrow entrances at the study
372 ANIMAL BEHAVIOR

site and monitor them every two hours. They also canvassed the study site sys-
tematically for evidence of chipmunk activity using radiotelemetry. What the
team found was surprising—the chipmunks never left their burrows during
the night. Even without a functioning circadian clock, the chipmunks knew
to avoid nighttime activity on the surface. However, this aversion did not pre-
vent them from engaging in nighttime activity in their burrows. DeCoursey’s
team noted that the SCN-lesioned animals seemed to be particularly restless at
night. Apparently nocturnal predators noticed this activity as well because by
the end of the study period more than 80 percent of the SCN-lesioned chip-
munks had been lost. In contrast, the chipmunks that had received sham sur-
geries had much higher survival rates, with very few deaths following territory
reestablishment (DeCoursey et al., 2000).

SIDE NOTE: THE DIFFICULTY AND IMPORTANCE OF STUDYING CIRCADIAN RHYTHMS

IN THE WILD
DeCoursey’s study clearly indicated that endogenous circadian rhythms
were important for survival in the wild. How could you examine the cause of a
rhythm in the wild without lesioning the SCN? To begin with, you would need
to demonstrate that the rhythm was truly endogenous and not just a direct
response to a rhythmic environment (masking). A provocative example was
published recently (Fernandez-Duque et al., 2010) regarding the activity
rhythms of South American owl monkeys (Aotus azarai). As might be pre-
dicted by their name, owl or night monkeys are the only anthropoids (mon-
keys, apes, and humans) that exhibit nocturnal behavior. They also exhibit
activity around the crepuscular periods of dawn and dusk. Earlier researchers
had noticed that this crepuscular activity seemed to be present at some times
but not others, potentially due to weather patterns, food availability, preda-
tors, and competition. Eduardo Fernandez-Duque and colleagues addressed
this question by catching several owl monkeys in Argentina and fitting them
with collars that recorded their activity for several months. What the results
showed was rather remarkable—the activity patterns precisely followed the
lunar cycle. When the moon rose early in the night, the monkeys were active
early in the night, whereas when the moon rose late at night, the monkeys
were active late. The crepuscular activity happened most prominently when
there was a new moon and the night had been continuously dark, so the mon-
keys were most active at dawn and dusk.
This activity pattern raised a question: did the owl monkeys genuinely have an
endogenous circalunar clock, or was their nocturnal activity simply inhibited
(masked) by complete darkness? To test this hypothesis, the research team made
A TIME FOR EVERY PURPOSE 373

clever use of data during three nights in which there were lunar eclipses.
During these nights, they found that the owl monkeys were quite active up
until the point that the moon disappeared. Then their activity dropped to
almost nothing (Figure 11.10). Fernandez-Duque suggests that the ability of

Figure 11.10. The activity rhythms of owl monkeys (Aotus azarai) are masked by moon-
light (adapted from Fernandez-Duque et al., 2010). (a) A graph depicting the daily
activity of an owl monkey over several months. The owl monkey displayed prominent
crepuscular activity around the time of dawn and dusk, the timing of which changed
with the seasons (time is given in clock time on the x axis). The owl monkey also
exhibited nocturnal activity except on nights when there was no moonlight (these
nights are marked with black spots). The activity record appears to have diagonal
“stripes” of activity because the monkey was only active after the moon had risen each
night, and the timing of moonrise grows increasingly later during the lunar cycle.
(b) On a night when there was a lunar eclipse the owl monkeys’ activity decreased to
almost nothing. This meant that the monkeys’ activity was not controlled by a lunar
clock but instead was simply passively responding to overall moonlight levels
(masking).
374 ANIMAL BEHAVIOR

the monkeys to make use of moonlight may have been a key adaptation for
them to exploit the nocturnal niche (Fernandez-Duque et al., 2010).
Previous studies performed in the laboratory on owl monkeys had sug-
gested that the monkeys were simply nocturnal with crepuscular activity
(Fernandez-Duque et al., 2010). In the wild, the story was clearly more com-
plicated. This appears to be true for many species: circadian behavior in the
laboratory may not reflect the function of activity patterns in the field. For
example, in my own research species, a small Chilean rodent called the degu
(Octodon degus), individuals in the wild are strictly diurnal—they never exit
their burrows at night. Nevertheless, their rhythms do shift in response to
many environmental factors. When it is hot during the summer, they restrict
their activity to the early morning and late evening when temperatures are
likely to be cooler. This shift in activity pattern is less likely to happen if the
animals have access to shade or a good food source (Hagenauer & Lee, 2008).
To the frustration of many researchers, when housed in a laboratory, degus
shift their time of activity dramatically in response to seemingly subtle changes
in housing conditions. For example, one of the easiest mechanisms for a circa-
dian biologist to measure the daily activity rhythms of laboratory animals is by
adding a running wheel to their cage and counting the number of wheel turns
occurring during any particular 10-minute interval. However, for the degu,
the addition of a running wheel to the cage can cause them to become almost
completely nocturnal. This huge switch in behavior puzzled researchers until
they considered the unnatural behavior of the animals in response to the run-
ning wheel. The degus found the wheels to be so compelling, they would run
the equivalent of a marathon each day (sometimes up to 700 wheel turns every
10 minutes)! Normally in the wild degus spend less than 2 percent of their day
engaging in running behavior. Thus, despite the fact that the degus were
housed under “normal” room temperatures, they were overheating from the
exercise. This caused their activity rhythms to take on an extreme “summer”
pattern, in which they shifted their activity to a time of day when their body
temperature was naturally lower (Hagenauer & Lee, 2008). Clearly it is
important to run both controlled laboratory studies and field studies to truly
understand a rhythm in animal behavior!

PROXIMATE MECHANISMS (ONTOGENY): HOW DO RHYTHMS IN ANIMAL

BEHAVIOR DEVELOP?
Circadian and seasonal rhythms play an important role in reproduction.
Because of energetic demands, in many species it is essential that pregnancy
and early development occur during a season that is rich in food availability.
A TIME FOR EVERY PURPOSE 375

For fast developers (such as rodents) this means that mating, pregnancy, and
birth occur during the spring and summer months, whereas for slow develop-
ers (such as ruminants), mating occurs in the fall and the young are born in the
spring.
The timing of reproductive behaviors during the day also matters. In the
wild, mating is a relatively risky enterprise (imagine how vulnerable an animal
is to predators when engaging in a mating display!). To make the process as
efficient as possible, some species, such as nocturnal rodents, have a daily
rhythm that controls the time of day that females ovulate in response to high
estrogen. This rhythm ensures that the timing of ovulation is carefully coordi-
nated with the normal active period and mating opportunities. Likewise, birth
and lactation are also carefully timed (Mahoney, 2010). Therefore, even by
the first days of life an animal’s existence is already dictated by rhythms.
In order for the offspring to be prepared for this rhythmic world, circadian
rhythms develop during late fetal development. However, at the time of birth
in altricial species (a species that is born at a relatively early developmental
stage, such as humans and laboratory rodents) the SCN has not fully devel-
oped input connections from the retina or output connections to slave oscilla-
tors throughout the body (Weinert, 2005). For example, human infants show
low-amplitude temperature rhythms soon after birth, even when they are born
prematurely and maintained under the constant lighting of the intensive care
unit. However, their rhythms in sleeping and waking are not well established
until two to four months of age (McGraw et al., 1999; Mirmiran et al.,
2003). Young infants also need to eat frequently (every few hours) in order
to maintain rapid growth, so their sleep/wake cycle is not as consolidated as
it is in adults (i.e., sleeping is scattered in bouts across the day and night). This
lack of a consolidated, entrained rhythm of sleeping and waking in infants is
part of the reason why young parents get so little sleep.
This weak rhythmicity is a property of young laboratory rodents as well.
The SCN exhibits clock-gene rhythms several days before birth, but at birth
daily rhythms are still only weakly expressed in behavior and hormone secre-
tion. The retinohypothalamic tract is not fully developed, so daily rhythms
are not entrained by the solar day (the solar day is also questionably relevant
to the pups’ existence in a dark burrow!). Instead, the rhythms are entrained
by the pups’ interaction with the mother (dam), a process called maternal
entrainment (Weinert, 2005).
An extreme example of maternal entrainment comes from the rabbit (Oryc-
tolagus cuniculus). In the wild, female rabbits give birth to their pups in a sep-
arate nursery burrow away from the communal warren to keep them hidden.
In order to reduce the risk of predation, the mother only visits the litter once
376 ANIMAL BEHAVIOR

per day for three to five minutes to allow the pups to nurse. During these few
minutes, the pups need to drink up to 20 or 30 percent of their body weight in
milk in order to survive. In order to start suckling immediately, the rabbit
pups within the dark burrow depend on their circadian clock to anticipate
their mother’s daily arrival. Twenty to thirty minutes before she arrives, they
arouse and uncover themselves from the nesting material in order to be acces-
sible (Jilge, 1993).
By the age of weaning, daily rhythms need to transition from maternal
entrainment to photic entrainment (entrainment by the solar day). For noc-
turnal rats and mice, this means a large transition in activity pattern. Up until
the point of weaning, the pups have been entrained by maternal rhythms to be
active during the day because the dam primarily nurses during the day and for-
ages at night. When the pups begin to leave the burrow to seek their own food
and water, the pups need to develop an adult-typical nocturnal activity pattern
to effectively forage and avoid predators (Thiels et al., 1990). As their growth
slows, daily rhythms in sleep and activity also start to consolidate into a more
solid rhythm, and the rhythms of slave oscillators throughout the body
develop an adult organization (Weinert, 2005).
In contrast, human children do not need to completely change their tem-
poral niche or chronotype (the time of day that they are active) at the time
of weaning, but they do undergo large changes in chronotype around the time
of adolescence. At this point, daily rhythms shift later so that teenagers have a
tendency to go to bed later and wake up later than children or adults (Figure
11.11). In our laboratory, we hypothesized that these changes were due to
pubertal hormones because they correlated with the timing of puberty (girls
showed changes in chronotype at a younger age than boys). Furthermore,
there was a sex difference in the magnitude of the changes: by the end of ado-
lescence, men were substantially more evening type than women. This sex dif-
ference diminished as chronotype drifted progressively earlier during the
rest of adulthood (Roenneberg et al., 2003; Carskadon et al., 2004). In sup-
port of the hypothesis that pubertal hormones drive these big changes in the
time of sleep and activity during adolescence, we showed that laboratory
rodents also show large changes in chronotype during puberty. By manipulat-
ing hormone levels using castration, hormone capsule implants, and chemicals
that inhibit hormone production, we were able to show that these adolescent
chronotype changes were due to the sex hormones testosterone and estrogen
(Hagenauer et al., 2011a, 2011b). Because of this growing evidence that ado-
lescent chronotype changes are a natural part of pubertal development, many
high schools in the United States are considering shifting their schedules later
to fit the later chronotype of adolescents. Policy changes of this sort are likely
A TIME FOR EVERY PURPOSE 377

Figure 11.11. Human chronotype changes with age (adapted from Roenneberg et al.,
2004). Individuals showed a transition into a more evening chronotype during ado-
lescence (ages 10–20 years), as determined by self-reported midsleep phase. This trend
then reversed for the rest of adult life. Males were more evening type than females
from the age of late puberty through menopause.

to decrease teenage sleep deprivation, improve psychological well-being, and

enhance learning (Carskadon et al., 2004).

SEASONAL RHYTHMS
Thus far, we have discussed a variety of strategies adopted by animals to
meet the challenges of the daily cycle of day and night brought about by the
Earth’s rotation. Presenting perhaps an even greater challenge are the drastic
changes in the environment that occur across an entire year. Species living in
temperate climates must contend with seasonal variations in their environ-
ment, such as resource availability and ambient temperature. Many animals
378 ANIMAL BEHAVIOR

facing such conditions demonstrate seasonal rhythms to mitigate the hard-

ships, or exploit the opportunities, of their changing environment.
You are likely familiar with—or have even directly observed—a number of
seasonal adaptations observed across the animal kingdom, such as the prehi-
bernation hoarding of squirrels or the migration of birds. Indeed, the arrival
of some birds is sometimes called the first sign of spring. But how do these
squirrels know when to begin preparing for hibernation, and what cue sends
migratory birds on their way? Different species use different mechanisms to
adapt to the changing seasons.
Some seasonal rhythms are strictly endogenous and are generated by inter-
nal mechanisms even in the absence of external information. Because these
rhythms have a period of approximately one year, they are referred to as
circannual rhythms. The most well-described example of a circannual
rhythm is that of hibernation in golden-mantled ground squirrels (Callosper-
mophilus lateralis). Golden-mantled ground squirrels begin hibernation in
the late summer and emerge in the early spring, leaving only limited time to
reproduce and forage to build sufficient fat to last through hibernation.
Researchers wanted to know if some facet of the environment, such as day
length or temperature, caused this hibernation, so they brought a number of
these squirrels into the laboratory, where light and temperature conditions
remained the same throughout the year (Pengelley et al., 1976). Despite the
constant conditions, the squirrels still entered hibernation. After monitoring
these squirrels for several years, researchers observed that the time of year that
the squirrels entered hibernation appeared to be a free-running rhythm, sug-
gesting that this behavior is the result of a circannual biological clock. Lesion-
ing the SCN does not abolish the circannual rhythm of hibernation (Zucker et
al., 1983), which means that the circadian and circannual clocks are likely to
be anatomically distinct. To date, the anatomical site of the circannual clock
remains an open question.
In contrast with true endogenous circannual rhythms are seasonal rhythms
that require exogenous input. Perhaps the most common form of exogenous
seasonal rhythmicity is photoperiodism, the measurement of day length to
time biological rhythms. Although the number of hours of light each day
may not directly influence the survival of an animal, day length is a strong pre-
dictor of the state of the environment. In the summer, long day lengths
accompany higher temperatures and relative resource abundance, whereas
shorter day lengths in the winter coincide with low temperatures and generally
scarcer resources. Because light is such a powerful predictor of environmental
conditions, some species use changes in lighting conditions alone to drive
important changes in physiology. Siberian hamsters (Phodopus sungorus)
A TIME FOR EVERY PURPOSE 379

attempt to survive the winter by reducing their body mass and suppressing
reproductive function so as to decrease the amount of food they require to sur-
vive (Hoffman, 1973). At the same time, they undergo a fur molt, changing
from dark gray in the summer to gray-white during the fall and winter. Unlike
the hibernation of ground squirrels, this process does not occur automatically
in constant laboratory conditions. Rather, the process of adopting the winter
phenotype only occurs when hamsters are exposed to short, winter-like day
lengths. Lighting conditions alone can induce members of this species into
shutting down their reproductive system and greatly reducing their body mass.
In order for animals to react to a photoperiod, they must have a means of
measuring the number of hours of light each day. Animals use the hormone
melatonin as a biological signal of the presence or absence of light. Melatonin
is produced in the pineal gland when an animal is exposed to periods of sus-
tained darkness and ceases during exposure to light. Thus, animals would
experience longer melatonin production each night during the long nights of
winter than the long days of summer. Animals that exhibit photoperiodism
use these melatonin signals to approximate the season and alter their physiol-
ogy accordingly. A unique aspect of this mechanism is that the duration of
melatonin production, rather than the amount of melatonin produced, is what
affects photoperiodic changes. This was demonstrated in a clever experiment
(Goldman et al., 1984) in which Siberian hamsters were infused with melato-
nin for several hours either in the middle of the night, when melatonin would
already be present, or a few hours before the onset of darkness, so as to artifi-
cially extend the length of time that melatonin would constantly be present.
While both of these infusions contained the same amount of melatonin, the
infusion delivered a few hours before the onset of darkness added to the overall
duration of the normal melatonin signal and caused the hamsters to suppress
their reproductive systems, but the infusion during the darkness had no such
effect because it only increased the amount of melatonin present but not the
duration of its presence in the body. Interestingly, even though the seasonal
clock is not located in the SCN, light information reaches the pineal gland
via a pathway that depends on retinal input to the SCN (Arendt, 2005).
Human beings in most of the world are insulated from much of the sea-
sonal variability in the environment. We have invented climate-control sys-
tems to keep ourselves cool in the summer and warm in the winter,
supermarket shelves are stocked with abundant food regardless of time of year,
and electric lights can allow us as many hours of light each day as we choose.
Despite this insulation, seasonal variability is still observed in a number of
traits in humans. In autumn months, people tend to eat larger, more calorie-
rich meals yet feel less satisfied than they would after eating a smaller meal
380 ANIMAL BEHAVIOR

during the spring (de Castro, 1990). Although human beings are by no means
seasonal breeders, there is seasonal variability in human sexual behavior. Stud-
ies in the United States have observed that people are more likely to have their
first sexual experience (Rodgers et al., 1992) or engage in later sexual activity
(Udry & Morris, 1967) during the summer months. Seasonal affective disor-
der (SAD) is a mood disorder that causes depression-like symptoms during
one season (most commonly winter) of each year. Patients with the disorder
sometimes find relief in light therapy, in which exposure to artificial sunlight
is used to reduce the production of melatonin. These observational studies
cannot exclude the role of factors beyond the changing environment (such
as cultural influences) in driving these patterns of behavior, but it is nonethe-
less the case that human behavior is variable over the course of the year.

CONCLUSION
As you have seen, there is tremendous evidence that internal clocks are
essential for the functioning of ecological communities, internal physiology,
and human health. Therefore, it is ironic that now, at the beginning of the
twenty-first century, our economic and environmental policies are becoming
increasingly indifferent to the importance of daily rhythms. At this point in
history, the ubiquity of electrical lighting has made it so that more than 80 per-
cent of the population of the United States and 66 percent of the population
of the European Union are exposed to such severe light pollution that true
night is never really experienced. Under these conditions, the light pollution
mimics twilight and can suppress normal nighttime secretion of the hormone
melatonin (Navara & Nelson, 2007).
This excess of artificial lighting interferes with the partitioning of temporal
niches, such that visual predators develop an advantage at night, and the activ-
ity of nocturnal prey is suppressed. Light pollution allows crepuscular species
to extend their active period into the night. It can also interfere with the sea-
sonal timing of breeding because the artificially extended photoperiod mimics
summer-like conditions. Because of the importance of celestial light for navi-
gation, there is growing evidence that light pollution can disrupt the migration
of birds, butterflies, and eels (Navara & Nelson, 2007). Overall, though, we
are only just beginning to understand the devastating ecological consequences
of light pollution, and a great deal more research needs to be done.
Humans are not immune to the dangers presented by our own modern
lighting. Even while we seek medical cures for cancer, we blatantly disregard
the importance of our ancient internal rhythms for human health. There is
growing evidence that circadian disruption in the form of shift work or regular
A TIME FOR EVERY PURPOSE 381

transatlantic travel places individuals at much higher (3×) risk for various
forms of cancer (Navara & Nelson, 2007). Due to this evidence, the World
Health Organization has now listed shift work as a probable carcinogen, but
recent studies suggest that urban light pollution alone may be sufficient for
increasing the risk of breast cancer (Kloog et al., 2011). Many researchers
believe this increased risk of cancer is related to suppression of nighttime mela-
tonin secretion (Kloog et al., 2011). Melatonin is important for controlling
seasonal rhythms, as discussed earlier, but is also a potent antioxidant and
regulator of growth, metabolism, and immunity (Navara & Nelson, 2007).
Disrupted rhythms have also been linked to the development of depressive ill-
ness (McClung, 2007) and fertility problems (Mahoney, 2010). Despite these
known risks, businesses have been moving sales and manufacturing into the
evening hours to increase productivity and profit, such that now over 20 per-
cent of the urban workforce is exposed to shift-work conditions (Navara &
Nelson, 2007).
What can you do to help? The first step is to limit your use of light at night to
only that which is absolutely necessary. If you need to expose yourself to light
at night, try to make it a warm-colored light because both the circadian time-
keeping system and melatonin secretion are most sensitive to blue-green light
(this is because a large percentage of the retinal input to the SCN is via a spe-
cial photoreceptor called the melanopsin ganglion cell) (Morin & Allen,
2006). To protect yourself from blue-green light at night, you can download
software that automatically changes the intensity and color of your computer
monitor depending on the time of day at http://stereopsis.com/flux/. The sec-
ond step is to reduce light pollution from outdoor light fixtures by shielding
them so that the light emitted is focused downward, away from the sky. If
the lighting is meant to primarily serve security purposes, consider using
motion-sensitive lighting. If you want to learn more about how to reduce light
pollution and its ecological consequences, you can join the Dark Skies move-
ment at http://www.darksky.org/ or take part in a large global study of light
pollution at http://www.globeatnight.org/.

ACKNOWLEDGMENTS
The authors would like to acknowledge Chelsea Anderson and Chelsea
Fournier for the construction of the figures. We are also deeply indebted to
Mary Ellen Hagenauer, John Basler, Maggie Elliott, and Jessica Koren for
their editing expertise. Finally, we would like to thank our funding sources,
including a trainee fellowship in Pediatric Endocrinology granted to the Uni-
versity of Michigan via the National Institute of Health (M.H.: #T32-
382 ANIMAL BEHAVIOR

DK071212) and a laboratory grant from the National Science Foundation

(M.H.: #IBN-0952046).

REFERENCES AND SUGGESTED READING

Akhtar, R. A., A. B. Reddy, E. S. Maywood, J. D. Clayton, V. M. King, A. G. Smith,
T. W. Gant, M. H. Hastings, & C. P. Kyriacou. (2002). Circadian cycling of
the mouse liver transcriptome, as revealed by cDNA microarray, is driven by
the suprachiasmatic nucleus. Current Biology, 12, 540–550.
Arendt, J. (2005). Melatonin: Characteristics, concerns, and prospects. Journal of Bio-
logical Rhythms, 20, 291–303.
Aschoff, J. (1965). Circadian rhythms in man—a self-sustained oscillator with an
inherent frequency underlies human 24-hour periodicity. Science, 148,
1427–1432.
Bae, K., X. Jin, E. S. Maywood, M. H. Hastings, S. M. Reppert, & D. R. Weaver.
(2001). Differential functions of mPer1, mPer2, and mPer3 in the SCN circa-
dian clock. Neuron, 30, 525–536.
Brady, J. (1982). Biological Timekeeping, Society for Experimental Biology Seminar
Series 14. Cambridge, UK: Cambridge University Press.
Carskadon, M. A., C. Acebo, & O. G. Jenni. (2004). Regulation of adolescent sleep:
Implications for behavior. Annals of the New York Academy of Sciences, 1021,
276–291.
Craig, J. L. & M. E. Douglas. (1984). Temporal partitioning of a nectar resource in
relation to competitive asymmetries. Animal Behavior, 32, 624–629.
Davidson, A. J., M. T. Sellix, J. Daniel, S. Yamazaki, M. Menaker, & G. D. Block.
(2006). Chronic jet-lag increases mortality in aged mice. Current Biology, 16,
R914–R916.
de Castro, J. M. (1990). Seasonal rhythms of human nutrient intake and meal pat-
tern. Physiology and Behavior, 50, 243–248.
DeCoursey, P. J. (1960). Daily light sensitivity rhythm in a rodent. Science, 131,
33–35.
DeCoursey, P. J. (2004). The behavioral ecology and evolution of biological timing
systems. In J. C. Dunlap, J. J. Loros, & P. J. DeCoursey (eds.), Chronobiology:
Biological Timekeeping (pp. 27–66). Sunderland, MA: Sinauer Associates.
DeCoursey, P. J., J. K. Walker, & S. A. Smith. (2000). A circadian pacemaker in free-
living chipmunks: Essential for survival? Journal of Comparative Physiology A,
186, 169–180.
de Mairan, J. J. O. (1729). Observation botanique. Histoire de l’Academie Royale des
Sciences, 35–36.
Duffield, G. E., J. D. Best, G. H. Meurers, A. Bittner, J. J. Loros, & J. C. Dunlap.
(2002). Circadian programs of transcriptional activation, signaling, and pro-
tein turnover revealed by microarray analysis of mammalian cells. Current Biol-
ogy, 12, 551–557.
A TIME FOR EVERY PURPOSE 383

du Monceau, H.-L. D. (1758). La Physique des Arbres. Paris, FR: Chez H. L. Guerin
& L. F. Delatour.
Fernandez-Duque, E., H. de la Iglesia, & H. G. Erkert. (2010). Moonstruck pri-
mates: Owl monkeys (Aotus) need moonlight for nocturnal activity in their
natural environment. PLoS ONE, 5, e12572.
Globig, M. (2007). A world without day or night. Max Planck Research, 2, 60–61.
Goldman, B. D., J. M. Darrow, & L. Y. Yogev. (1984). Effects of timed melatonin
infusions on reproductive development in the Djungarian hamster (Phodopus
sungorus). Endocrinology, 114, 2074–2083.
Hagenauer, M. H., A. F King, B. Possidente, M. Y. McGinnis, A. R. Lumia, E. M.
Peckham, & T. M. Lee. (2011a). Changes in circadian rhythms during
puberty in Rattus norvegicus: Developmental time course and gonadal depen-
dency. Hormones and Behavior, 60, 46–57.
Hagenauer, M. H., J. H. Ku, & T. M. Lee. (2011b). Chronotype changes during
puberty depend on gonadal hormones in the slow-developing rodent, Octodon
degus. Hormones and Behavior, 60, 37–45.
Hagenauer, M. H. & T. M. Lee. (2008). Circadian organization of the diurnal cavio-
morph rodent, Octodon degus. Biological Rhythm Research, 39, 269–289.
Hastings, M. H. & E. D. Herzog (2004). Clock genes, oscillators, and cellular net-
works in the suprachiasmatic nuclei. Journal of Biological Rhythms, 19,
400–413.
Hoffmann, K. (1973). The influence of photoperiod and melatonin on testis size,
body weight, and pelage colour in the Djungarian hamster (Phodopus sungo-
rus). Journal of Comparative Physiology, 85, 267–282.
Jilge, B. (1993). The ontogeny of circadian rhythms in the rabbit. Journal of Biological
Rhythms, 8, 247–260.
Kloog, I., B. A. Portnov, H. S. Rennert, & A. Haim. (2011). Does the modern urban-
ized sleeping habitat pose a breast cancer risk? Chronobiology International, 28,
76–80.
Kronfeld-Schor, N. & T. Dayan, (2008). Activity patterns in rodents: The physio-
logical ecology of biological rhythms. Biological Rhythm Research, 39, 193–211.
Mahoney, M. M. (2010). Shift work, jet lag, and female reproduction. International
Journal of Endocrinology, 2010, 1–9.
McClung, C. A. (2007). Circadian rhythms, genes, and the biology of mood disor-
ders. Pharmacology and Therapeutics, 114, 222–232.
McGraw, K., R. Hoffmann, C. Harker, J. H. Herman. (1999). The development of
circadian rhythms in a human infant. Sleep, 22, 303–310.
Menaker, M. (1996). Colin Pittendrigh: 1918–1996. Nature, 381, 24.
Mirmiran, M., Y. G. H. Maas, & R. L Ariagno, (2003). Development of fetal and
neonatal sleep and circadian rhythms. Sleep Medicine Reviews, 7, 321–334.
Moore, R. Y. & V. B. Eichler (1972). Loss of a circadian adrenal corticosterone rhythm
following suprachiasmatic lesions in the rat. Brain Research, 42, 201–206.
Moore, R. Y. & N. J. Lenn (1971). A retinohypothalamic projection in the rat. Jour-
nal of Comparative Neurology, 146, 1–14.
384 ANIMAL BEHAVIOR

Morin, L. P. & C. N. Allen. (2006). The circadian visual system, 2005. Brain
Research Reviews, 51, 1–60.
Navara, K. J. & R. J. Nelson. (2007). The dark side of light at night: Physiological,
epidemiological, and ecological consequences. Journal of Pineal Research, 43,
215–224.
Pengelley, E. T., S. J. Amundson, B. Barnes, & R. C. Aloia. (1976). Relationship of light
intensity and photoperiod to circannual rhythmicity in the hibernating ground
squirrel, Citellis lateralis. Comparative Biochemistry and Physiology, 53a, 273–277.
Pittendrigh, C. S. & S. Daan. (1976a). Functional analysis of circadian pacemakers in
nocturnal rodents II: The variability of phase response curves. Journal of Com-
parative Physiology, 106, 253–266.
Pittendrigh, C. S. & S. Daan. (1976b). Functional analysis of circadian pacemakers in
nocturnal rodents IV: Entrainment: Pacemaker as clock. Journal of Compara-
tive Physiology, 106, 291–331.
Pizatto, L., T. Child, & R. Shine. (2008). Why be diurnal? Shifts in activity time en-
able young cane toads to evade cannibalistic conspecifics. Behavioral Ecology,
19, 990–997.
Ralph, M. R., R. G. Foster, F. C. Davis, & M. Menaker. (1990). Transplanted supra-
chiasmatic nucleus determines circadian period. Science, 247, 975–978.
Richter, C. P. (1965). Biological Clocks in Medicine and Psychiatry. Springfield, IL: C.
C. Thomas.
Rodgers, J. L., D. F. Harris, & K. R. Vickers. (1992). Seasonality of first coitus in the
United States. Social Biology, 39, 1–14.
Roenneberg, T., T. Kuehnle, P. P. Pramstaller, J. Ricken, M. Havel, A. Guth, & M.
Merrow. (2004). A marker for the end of adolescence. Current Biology, 14,
R1038–1039.
Roll, U., T. Dayan, & N. Kronfeld-Schor. (2006). On the role of phylogeny in deter-
mining activity patterns of rodents. Evolutionary Ecology, 20, 479–490.
Schmitz, L. & M. Ryosuke. (2011). Nocturnality in dinosaurs inferred from scleral
ring and orbit morphology. Science, 332, 705–708.
Schulkin, J. (2005). Curt Richter: A Life in the Laboratory. Baltimore, MD: Johns
Hopkins University Press.
Shigeyoshi, Y., K. Taguchi, S. Yamamoto, S. Takekida, L. Yan, H. Tei, T. Moriya, S.
Shibata, J. J. Loros, J. C. Dunlap, & H. Okamura. (1997). Light-induced
resetting of a mammalian circadian clock is associated with rapid induction
of the mPer1 transcript. Cell, 91, 1043–1053.
Silver, R., J. LeSauter, P. A. Tresco, & M. N. Lehman. (1996). A diffusible coupling
signal from the transplanted suprachiasmatic nucleus controlling circadian
locomotor rhythms. Nature, 382, 810–813.
Smale, L., T. Lee, & A. A. Nunez. (2003). Mammalian diurnality: Some facts and
gaps. Journal of Biological Rhythms, 18, 356–366.
Stephan, F. K. & I. Zucker. (1972). Circadian rhythms in drinking behavior and
locomotor activity of rats are eliminated by hypothalamic lesions. Proceedings
of the National Academy of Sciences, USA, 69, 1583–1586.
A TIME FOR EVERY PURPOSE 385

Storch, K. F., O. Lipan, I. Leykin, N. Viswanathan, F. C. Davis, W. H. Wong, & C.

J. Weitz. (2002). Extensive and divergent circadian gene expression in liver
and heart. Nature, 417, 78–83.
Thiels, E., J. R. Alberts, & C. P. Cramer. (1990). Weaning in rats (2): Pup behavior
patterns. Developmental Psychobiology, 23, 495–510.
Tinbergen, N. (1963). On aims and methods of ethology. Zeitschrift für Tierpsycholo-
gie, 20, 410–433.
Udry, R. J. & N. M. Morris. (1967). Seasonality of coitus and seasonality of birth.
Demography, 4, 673–679.
Weaver, D. R. (1998). The suprachiasmatic nucleus: A 25-year retrospective. Journal
of Biological Rhythms, 13, 100–112.
Weinert, D. (2005). The ontogenetic development of the mammalian circadian sys-
tem. Chronobiology International, 22, 179–205.
Welsh, D. K., D. E. Logothetis, M. Meister, & S. M. Reppert. (1995). Individual
neurons dissociated from rat suprachiasmatic nucleus express independently
phased circadian firing rhythms. Neuron, 14, 697–706.
Yamazaki, S., R. Numano, M. Abe, H. Akiko, R. Takahashi, M. Ueda, G. D. Block,
Y. Sakaki, M. Menaker, & H. Tei. (2000). Resetting central and peripheral
circadian oscillators in transgenic rats. Science, 288, 682–685.
Yerushalmi, S. & R. M. Green. (2009). Evidence for the adaptive significance of cir-
cadian rhythms. Ecology Letters, 12, 970–981.
Zheng, B., U. Albrecht, K. Kaasik, M. Sage, W. Lu, S. Vaishnav, Q. Li, Z. S. Sun, G.
Eichele, A. Bradley, & C. C. Lee. (2001). Nonredundant roles of the mPer1
and mPer2 genes in the mammalian circadian clock. Cell, 105, 683–694.
Zucker, I., M. Boshes, & J. Dark. (1983). Suprachiasmatic nuclei influence circan-
nual and circadian rhythms of ground squirrels. American Journal of Physiology,
244, R472–R480.
This page intentionally left blank
 12

Getting There: Animal Orientation,

Navigation, and Migration
James L. Gould

CHALLENGES AND STRATEGIES

All animals must orient to their environment to survive. Without technology
humans are outperformed handily by many species, implying the existence
of sensory channels and processing strategies unavailable to us. On a local
scale, for instance, individual forager honey bees learn the sun’s complex pat-
tern of movement through the sky, map their home range, utilize a variety
of cues we cannot even sense (UV light, polarization, magnetic fields, locale
odors, and so on), and then draw miniature maps for one another in the hive
specifying the location of food, water, and potential nesting sites (Gould &
Gould, 1995). On a global scale, the newly fledged young of many seabird
species launch themselves alone onto a journey of thousands of miles across
open water to relatively precise targets—a small set of islands in the South
Pacific, for instance—regardless of wind and weather (Gill et al., 2005;
McCaffery, 2008) (Figure 12.1).
Whether the trips are epic or mundane, selection operates strongly against
the navigationally incompetent. The result is a plethora of finely tuned strate-
gies and backup systems incorporating elaborate mixtures of innate instruc-
tions and programmed learning. These species-specific solutions nevertheless
fall into a relatively consistent set of operational categories (Gould & Gould,
2012) (Table 12.1). This chapter will look at each class of navigational
388 ANIMAL BEHAVIOR

Figure 12.1. Migration tracks of bar-tailed godwits (redrawn after Gill et al., 2009).
Birds, both experienced and first-time migrants, depart south-southwest from the
Yukon Delta National Wildlife Refuge in Alaska and fly solo as much as 7,000 miles
nonstop to wintering grounds in the South Pacific. Some arrive in New Zealand
directly from Alaska, while others make brief stops in Melanesia. Nonstop flights are
shown as solid lines; dotted lines indicate routes after first landing.

solution, from simplest to most complex. But these categories are human con-
structs: animals in the field move smoothly between strategies as chance dic-
tates, blending two or more when appropriate and in some cases switching
their approach entirely with age and experience.

TAXIS AND KINESIS

Over its history, the study of behavior has often been the subject of debates
more about belief than fact (Gould & Gould, 1999). A classic example is the
seeming dichotomy between behaviorists on the one hand, who held that
instinct does not exist and that everything (beyond unconditioned stimuli
and reflexes) is learned, and ethologists on the other, who maintain that
instinct dominates behavior and even guides much of learning. A less dramatic
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 389

Table 12.1. Common orientation and navigation strategies (Gould & Gould, 2012).
Strategy Description

Taxis and kinesis Orienting directly toward or away from a cue

Compass orientation Maintaining a constant bearing relative to a cue (or a
constant absolute direction if compensating for cue
movement)
Vector navigation Using a sequence of compass bearings to steer a course;
generally (but not inevitably) independent of
landmarks
Piloting Navigating relative to familiar landmarks; may or may
not involve compasses
Inertial navigation Dead reckoning: keeping track of each leg of a journey
to compute location later; generally independent of
landmarks
True navigation Navigating knowing absolute or relative location of a
distant goal; generally independent of landmarks

tension is seen between reductionists, who insist that all behavior is based on
simple stimuli and responses, and antireductionists, who see the programs
and processes of behavior as being so completely integrated that dissection
into constituent elements destroys the very thing being studied.
Reductionists and behaviorists alike were drawn to very simple accounts of
animal navigation (Fraenkel & Gunn, 1940). If a homing pigeon (Columba
livia) found its loft after being taken away 100 km and released, it was either
because it got lucky and stumbled across it after a random search or had zeroed
in on some mysterious cue emanating from home and tracked up the putative
gradient. In their view, animals mostly oriented directly to cues (a taxis) or at a
fixed angle to them (a compass response). Thus for taxis-based behavior, a
trapped honey bee will fly toward the light (positive phototaxis) while a star-
tled roach will run away (negative phototaxis). Direct orientation toward or
away from gravity, the wind, temperature, odors, and the like provided a sim-
ple machine-like account of animal actions. A moth tracking a pheromone
trail upwind is a classic example of a taxis-based homing response (Kennedy
et al., 1980).
In contrast to simple taxis-dependent behaviors, in which the organisms
move directly toward or away from a stimulus, however, animals can also
respond by altering their speed or the straightness of their path depending
390 ANIMAL BEHAVIOR

on the strength of the cue. Bacteria, for instance, change their orientation
more often when a desirable stimulus is getting weaker, sampling new direc-
tions as conditions deteriorate. When things are improving, the organism
persists longer along its current route (Adler, 1976). This glorified getting-
warmer strategy, known as kinesis, produces directed movement only statisti-
cally; at any given moment the animal is in fact traveling at random.
While taxis- and kinesis-based movement sound simple, the reality is more
complex. The orientation of many animals, from microscopic copepods and
other zooplankton performing their daily vertical migration in the ocean all
the way up through birds and mammals, usually begins in advance of the sim-
ple cue that is supposed to trigger and guide the behavior (Haney, 1998). The
more we learn about animal navigation, the less simple taxis and kinesis seem
to account for the phenomena.

ANIMAL COMPASSES
A simple adumbration of the taxis-based strategy is for the creature to adopt
a fixed bearing relative to a cue—the compass response alluded to above.
Thus, for instance, a migrating bird might be programmed to fly southeast
in the autumn (and then northwest in the spring) as opposed to a simple north
or south track along the corresponding latitudinal gradient. Bees orient their
communication dances on a vertical comb to gravity in the darkness of the
hive but not simply to up or down (as would be the case for a taxis). Instead
they signal the direction of the food source with respect to the sun outside,
with the azimuth from the food to the sun translated into the angle between
the dance and vertical—a gravity-based compass orientation.
But there is more to it than this. Honey bees will orient their dances to the
sun directly if forced to dance on a horizontal surface; a bright light will do as
well. In these cases we might naïvely suppose that the animal treats the sun as
the dominant cue and maintains a fixed angle relative to this beacon as it trav-
els. The problem with this perspective is that the sun moves from east to west
over the course of the day. To fly a true compass direction, both birds and bees
must compensate for the sun’s changing azimuth direction and adjust their
bearings relative to it. Bees and homing pigeons alike can be clock-shifted—
forced to live for a time on a light/dark cycle out of phase with the outside
world, they eventually become entrained to an artificial dawn and dusk. When
tested, they act as though they believe the sun should lie in a predictable differ-
ent direction and thus depart reproducibly off course—90° for a six-hour
clock shift, typically. Most evidence suggests that these and other animals
learn the pattern of solar movement, interpolating and extrapolating as
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 391

Figure 12.2. The sun’s arc at 42° north latitude (redrawn after Gould, 1982a). (a) The
path through the sky is shown for the summer and winter solstices. On all other dates
the arc is parallel to these two trajectories but somewhere in between. The time of
dawn and the maximum elevation of the sun are different each day. The consistent
features are that the sun is in the south and at its maximum daily elevation at solar
noon; noon occurs exactly halfway between dawn and dusk; and the angle between
the rising sun and the noon sun is identical to the angle between the noon sun and the
setting sun. (b) The rate of movement of the sun varies over the course of the day,
being highest at noon and slowest at dawn and dusk. The graph here is for the
summer solstice at 42° north; the curve is different for other dates and latitudes.

necessary, and recalibrate frequently to take into account the changing path as
the season wears on (Figure 12.2; Gould & Gould, 2012). Clearly compass
orientation need not be simple.
Most diurnal animals use the sun as their primary cue. When the sun is
hidden many are able to infer the sun’s location from the pattern of polarized
light in the sky—a pattern created by the interaction between sunlight and the
earth’s atmosphere (Brines & Gould, 1982). The light from the sun is unpo-
larized, but a photon scattering off an air molecule tends to become polarized
perpendicularly to the plane it is traveling in. Thus when an animal looks at a
point in the sky—say the zenith, directly overhead, and at noon, when the sun
is due south—a plane is defined by three points: the sun (the source of the
photon), the zenith (where the photon was scattered), and the eye of the
observer. The light reaching the animal from the zenith has a net polarization
orthogonal to this plane (oriented east-west in this example). Moreover, the
degree of polarization (ranging from 0 to nearly 100% in the UV) varies with
392 ANIMAL BEHAVIOR

the angle of scatter. It is minimal at 0° and 180° and maximum at 90°. In

theory, the resulting bulls-eye pattern of light and dark would permit an
organism to infer the position of the sun even when it is hidden (Figure 12.3).

Figure 12.3. Pattern of polarized skylight (redrawn after Gould, 1982a). Allowing for
the difficulty of representing a hemisphere on a flat piece of paper, at any point
overhead the pattern of polarized light in a clear sky is perpendicular to the plane also
containing the sun and the observer. The circular bands of intensity (indicated here by
the thickness of the polarization lines) reflect the fact that the degree of polarization
depends on the angle of scatter, being strongest at 90°. This illustration represents the
sky at noon on the equinox at latitude 42°. The small circle in the lower right repre-
sents the minimum 10° circle of sky needed by honey bees to orient themselves.
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 393

Reality is less wonderful. For one thing, a photon can scatter more than
once en route to the eye, particularly at low elevations; only polarization
well above the horizon is sufficiently reliable for orientation. Atmospheric
conditions—humidity, heat, and pollution—all increase this problem of sec-
ondary scattering. Organic compounds evaporated from leaves even tend to
rotate the polarization (again, a problem most evident at low elevations)
(Brines & Gould, 1982; Gould, 1982). Nevertheless, honey bees forced to
communicate on a horizontal surface and shown as little as 10° of blue sky
can accurately orient their dances to celestial cues, indicating that they are
inferring the azimuth of the sun with considerable precision (Gould & Gould,
1995). Birds too can orient to polarization patterns, and some nocturnal
migrants (for reasons of predator avoidance and the need to minimize over-
heating, most birds migrate at night) regularly use twilight polarization as their
initial cue before setting off on the evening’s journey (Gould & Gould, 2012).
Another widely used nocturnal cue is the pattern of stars overhead (Emlen,
1967a, 1967b, 1970). Many species appear to imprint as fledglings on the
stars, somehow ignoring the bright but fickle planets. The location of any
given star appears to change as the night progresses and the constellations
rotate about the pole point (marked by Polaris at the moment). Moreover,
each star rises ever later as the season slips by. Nevertheless, even without
learning to recognize Polaris, a bird can triangulate the pole point (and thus
true north—and, for that matter, compute its current latitude, which corre-
sponds to the elevation of the pole point). But because their initial imprinting
is in the spring, whereas the first migration occurs in the autumn when many
new constellations populate the sky, birds must recalibrate their mental pic-
tures of the heavens. Moreover, as a bird travels south, new constellations
appear above the southern horizon while some familiar ones disappear from
the northern sky. For transequatorial migrants, there may be a 100% replace-
ment of celestial markers. Evidence strongly suggests that recalibration occurs
frequently en route (Able & Able, 1990, 1995; Gould, 1995; Gould &
Gould, 2012).
Although any number of marine creatures use the lunar-based tidal cycle to
organize their activity, very few appear to treat the moon as a compass. Apart
from sandhoppers and distinctly nonmarine dung beetles, the moon itself
seems to be ignored. Alternatively, it may be researchers rather than navigating
animals guilty of overlooking this cue.
Celestial cues, whether diurnal or nocturnal, are useless under overcast con-
ditions. That bees and homing pigeons continue to navigate with eerie preci-
sion when the sky is obscured led some early researchers to suppose they
must be able to see through the clouds, perhaps in the mysterious ultraviolet
394 ANIMAL BEHAVIOR

range. But the clock-shift tests mentioned earlier disprove this potential
explanation: while clock-shifted pigeons are reoriented during sunny condi-
tions, when released under overcast conditions they depart along the correct
(unshifted) bearing (Keeton, 1969). Clearly there is a backup system. For
nearly all animals, this secondary compass is magnetic.
There are two common types of magnetic compass. One, found in many
birds and some insects, depends on a paramagnetic interaction between light
and the pigment cryptochrome in the eyes (Gould & Gould, 2012). Crypto-
chrome absorbs blue-green light and is used by many organisms—plants
included—to set their circadian rhythms. This compass requires a minimum
level of light (about that of a quarter moon in a clear night sky) in the blue-
green range of the spectrum; under starlight or in red light, these animals are
disoriented. A curiosity of this paramagnetic phenomenon is that the response
is independent of polarity—that is, the organism cannot distinguish north
from south. It must instead depend on the inclination of the field (the angle
between the field lines and the surface of the earth) to identify the direction
of the pole. In the northern hemisphere, the inclination line is taken as signi-
fying north.
A different mechanism is found in fish, mammals, and certain insects; it
relies on magnetite crystals (Gould & Gould, 2012). (Magnetite-based organs
have another role in birds, described later.) In salmon, where the organ has
been characterized in detail, the magnetic grains are arranged in chains and
arrayed in a structure between the nasal cavity and the eyes. The same location
has high densities of magnetite in homing pigeons, dolphins, sea turtles, and a
variety of other migrants. A possible second detector is found in the lagena of
the inner ear (Wu & Dickman, 2011). An organ based on permanent magnets
does not fail in darkness and correctly distinguishes north from south. It is dis-
rupted by a high-intensity pulse of magnetic energy, a manipulation that
leaves cryptochrome-based systems unaffected.
For migrants, particularly those many species nesting in higher latitudes or
even the Arctic, a magnetic compass (whatever its physiological basis) is prob-
lematic. At most points on Earth, there is a discrepancy between magnetic and
true north—a difference that can exceed 45° in the Arctic. Moreover, this
declination generally changes with latitude. If the Earth’s magnetic field is
to be used as a compass, it must first be calibrated against true north. Humans
do this by observing the pole point (the location in the northern sky about
which the sun, polarization pattern, and stars appear to rotate), by taking the
sun’s azimuth at its highest elevation (solar noon) as true south, or by averag-
ing the sun’s azimuth at dawn and dusk to infer the noon azimuth. Whatever
strategy it uses, an animal traveling any significant distance will need to
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 395

recalibrate its magnetic compass against celestial information to maintain

accuracy, and the best evidence suggests this is done on a daily basis when
sky conditions permit (Gould & Gould, 2012).
Most of the more complex navigation strategies discussed below incorpo-
rate one or more of these compass measurements as an essential component.

VECTOR NAVIGATION
Many migration paths (but not those of waterbirds, as we will see) seem to
involve two or more separate legs of compass orientation—the strategy known
as vector navigation. For example, a first-year red-eyed vireo (Vireo olivaceus)
flies southeast from its birthplace in northern British Columbia alone and in
the dark to (approximately) Tennessee or Kentucky, whereupon it steers south
to Central America, alters course to the southeast again, and continues on to
northern Brazil (Figure 12.4). Vireos from a population in southern Quebec
instead fly southwest to Tennessee or Kentucky and then pick up the trail
south. In the autumn, individuals in one population of garden warblers (Sylvia
borin) in northern Europe fly southwest, then southeast to northern Africa,
while those in another depart southwest, then turn southeast to the same des-
tination. A direct route would have taken them over the Alps. This pattern of
population-specific orientation is regularly seen in species employing just a
single-step compass strategy: members of one population of blackcap warblers
in Germany fly southwest to Spain in the fall whereas birds in another nearby
group set out (again, solo and at night) west to Wales (Gould & Gould,
2012).
Tests with birds raised in the lab from eggs taken from nests in the wild
show that these directional preferences are innate. In the case of the
Germany-to-Wales group mentioned above, the route is only about 50 years
old, the result it seems of one or more genetic mutations followed by intense
selection (Berthold et al., 1992). Also innate seems to be the birds’ knowledge
of when to turn, switching from one vector to another. Birds in cages during
the migratory season attempt to escape in the appropriate direction in the
night when their peers are flying in the wild. (Hybrids fly an intermediate
direction and distance [Helbig, 1991, 1996]). Experiments with vector migra-
tors indicate that they have two systems operating. The primary strategy
involves measuring latitude and comparing it with the inherited value in the
navigational program. This can be done by measuring the elevation of the pole
point, or (if the sky is cloudy) the magnetic inclination, which changes from
90° to 0° as one moves from the pole to the equator. A caged bird obligingly
reorients its fruitless departure attempts as the experimenter changes the
396 ANIMAL BEHAVIOR

Figure 12.4. A dogleg route south (Gould & Gould, 2012). Many populations of red-
eyed vireos migrating to their wintering grounds from western Canada fly southeast
into the United States, then south to Central America, then southeast again into the
Amazon. Most populations from eastern Canada set off southwest, and then join the
main route in the south-central United States. (The Mercator projection used here
shows true compass bearings but progressively exaggerates distances and land areas in
more northerly regions.)

latitudinal parameter gradually to the appropriate natural value. A backup sys-

tem measures how many days the bird has been flying (or, in the case of the
caged birds, attempting to fly) and sets course on the new heading when more
than enough time has passed. The trip is ended at the appropriate latitude or
after the innately specified duration (Gwinner & Wiltschko, 1978; Gould &
Gould, 2012).
The ontogeny of the initial bout of vector navigation has been roughly
worked out. Despite its obvious shortcomings, for most species the magnetic
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 397

compass is the one the first-year birds seem to trust the most. If experimenters
prevent the nestlings from calibrating their magnetic compass to celestial bea-
cons prior to that first autumn journey, the birds simply fly south; this seems
to be the default fail-safe strategy for northern-hemisphere migrants. Another
change is seen when the birds return in the spring and is maintained in sub-
sequent trips: the vector strategy of first-years leads to relatively inefficient
curved paths on the globe as the animals persist along fixed bearings. (The
routes look straight on the well-known Mercator projection but deviate from
the “great circle” route, which is the actual shortest path, taking into account
the realities of spherical geometry.) During later redeployments the migrants
do indeed fly great circles. Apparently the birds learn (or, more likely, cali-
brate) the spherical optimization of the initial compass vectors. Quite possibly
they alter the turning and stopping points based on experience as well
(Wiltschko, 1991; Wiltschko & Wiltschko, 1995; Gould & Gould, 2012).
Tests with hand-reared nestlings show that while each population has its
unique set of vectors and latitude preferences (as well as departure dates), there
is a substantial degree of initial variation among individuals. Learning and
selection, each in their very different ways, shape the details of the routes
and timings in both the short and long terms (Davis et al., 2006; Pulido et
al., 2001; Pulido, 2007; Gould & Gould, 2012).

PILOTING
Many animals, humans included, rely on familiar landmarks or memorized
routes to move about in their home range. This local-area navigation is known
as piloting. It appears to go through two steps. At the outset individuals com-
mit to memory specific routes—from home to a particular food source, say.
The creature uses some mixture of visual, olfactory, auditory, and tactile cues,
depending on species-specific biases and the immediate contingencies of the
task. These various landmarks are remembered like pages in a multimedia
photo album, along with turning angles and intervening distances (though
perhaps not encoded in exactly this way) (Etienne et al., 1996; Gallistel &
Cramer, 1996; Wills et al., 2010). With sufficient information from land-
marks an animal can navigate locally without reference to a compass, though
it does seems to keep silent track of that information (Dyer & Gould, 1981).
In many species this route-specific navigation develops into a map-like repre-
sentation as the animal discovers the points where routes cross and begins to
create a mental representation of this web of paths. These creations are often
called cognitive maps and in vertebrates are stored in a specific structure in
the hypothalamus (McNaughton et al., 1996; Jacobs & Schenk, 2003).
398 ANIMAL BEHAVIOR

Figure 12.5. Kidnapped-bee experiment (redrawn after Gould, 1986). Left: Foragers
visiting food station A were captured departing the hive and carried in darkness to
release site B. Right: When set free they circled and flew toward station A, hidden in
the woods. Also shown is the direction predicted if the foragers had left the release site
on the same bearing they adopted leaving the hive, and the orientation expected if
they had simply flown back to the hive to get their bearings.

Their creation, once thought of as a superior intellectual feat largely confined

to humans, is now understood to be a routine and rather mundane piece of
orientational programming. The more modern view of the light cognitive load
involved is perhaps best illustrated by the work showing that spiders and
honey bees can “imagine” novel routes to goals (Figure 12.5). Such an ability
is considered the best test for the existence of cognitive mapping, and therefore
of the ability to make plans—a trait usually considered a function of higher
intelligence (Gould, 1986; Tarsitano & Jackson, 1994; Menzel et al., 2005).
Memorizing landmarks along routes need not be confined to local areas.
Geese, swans, ducks, cranes, and a number of other day-migrating birds fly
in family groups along a traditional route from breeding to wintering grounds.
The young learn the landmarks en route and recapitulate the journey in sub-
sequent years. This strategy accounts for two quite remarkable facts. First-
year birds whose parents do not migrate (as, for example, the large flocks of
resident geese in the northeast United States) do not themselves leave for the
winter, there being no one to show them the way (nor will they be able to
guide their own young when the time comes, dooming their offspring to a
year-round life in New Jersey and other less-than-optimal habitats). The other
surprising spinoff of this route-based piloting system is that the young of these
species can be imprinted on visual cues and then, led by an ultralight aircraft
displaying the imprinting cue, be taught a novel route and goal (Gould &
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 399

Gould, 2012). This is a trick with enormous conservation potential for a select
few species.

INERTIAL NAVIGATION (DEAD RECKONING)

In the absence of familiar landmarks, an animal may keep track of its loca-
tion by measuring the bearing and length of each leg of its journey and infer-
ring its current location geometrically. In at least some species this is an initial
step in building up cognitive maps. Thus when honey bees are moved to a
new area and trained along an indirect three-leg, 240-m route around a tall
building to a food source, returning foragers are able to signal by means of
their dance the correct as-the-crow-flies location to within 3 m of actual dis-
tance (90 m) and 5° of its direction (Frisch, 1967). Clearly they have accu-
rately reconstructed a route they have never actually flown. This strategy of
inertial navigation is known to sailors as dead reckoning.
The widespread human use of dead reckoning on both land and sea has col-
ored the study of animal navigation. Researchers have often assumed that iner-
tial guidance by birds and bees, for instance, makes use of the same cues and
computational techniques that humans have relied on. Before the age of radios
and GPS, blue-water sailors kept track of position by taking compass readings
for each leg of their journey while measuring the distance traveled on each leg
by reckoning their speed (by tossing a log on a rope over the side and counting
the number of knots played out over a fixed interval such as a minute). They
could then multiply this value for velocity by the amount of time traveled in
that particular direction and come up with a dead reckoning.
The cues for navigating humans are thus compass bearing, speed, and time.
In the honey bee example above, the initial supposition was that the foragers
take a compass bearing relative to the sun on each of the three legs, measure
flight speed along each, and time each of the three components of the trip.
After computing distance for each leg a bee calculates its net displacement
(distance and direction from the starting point) geometrically.
There are two clear problems with this model of dead reckoning. The first
is a general difficulty with any strategy of inertial navigation: drift. While
measuring speed though the water or in the air may seem easy enough, in
the absence of clear visual cues it is nearly impossible for a bird or bee to detect
crosswinds, headwinds, or tailwinds, or for an aquatic creature like a salmon or
sea turtle to judge the equivalent currents in the water. The result is that the
measurements for each of the legs will be systematically wrong, and errors will
accumulate with time (Figure 12.6; Gould & Gould, 2012). (Errors in meas-
uring bearings and distances also build up, but because they are typically ran-
dom, increase only as the square root of the duration.)
400 ANIMAL BEHAVIOR

Figure 12.6. Inertial navigation. In this example a sailor or animal sets off on a circui-
tous path involving a mix of three directions and three speeds. The navigator must
then determine the net displacement by integrating the apparent legs of the journey
(dark track) and then computing the homeward bearing and distance: 215° and 70 km
in the absence of errors. However, there is an undetected southward drift of 4.2 km/
hour, which gives rise to an actual route (gray line—“track made good”), for which the
correct homeward bearing and distance are quite different: 252° and 53 km.

The other general difficulty is in measuring the passage of time. Humans

regularly misjudge time by 20 percent or (usually) more, and animals in lab
tests are no better (Gallistel & Gibbon, 2000). And yet the measured accuracy
of animals performing dead reckoning implies a time sense good to within at
least 5 percent, and honey bees are better than this. Because the error being
measured has many sources, any judgment of time would need to be still more
accurate. In case after case (humans included), careful tests reveal that animals
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 401

largely (though not entirely) measure distance directly rather than as speed
multiplied by time. On the ground, ants (Wittlinger et al., 2006), crabs (Walls
& Layne, 2009), and humans (Mittelstaedt & Mittelstaedt, 2001; Durgin et
al., 2008) appear to count steps, having already calibrated their strides. In
the air and water, we expect by analogy that animals sum up wing beats and
tail strokes, though at present there is no direct evidence that they do so.
A basic lesson from work on inertial navigation is that animals employ the
strategy only when nothing better is available. On long-distance migratory
trips systematic and other errors accumulate to the point that geographical
precision is impossible. For these animals, midcourse and end-of-journey cor-
rections by means of a map sense (discussed next) solves the problem.

TRUE NAVIGATION
The ultimate goal for human navigators was true navigation—an ability to
determine global position. For our species the task is broken down into two
components for which we have no inborn gauge: longitude (east/west position)
and latitude (north/south location). Latitude is the lesser challenge (Gould &
Gould, 2012): the elevation of the pole point gives this value directly, so all
one needs is a stable base and a clear night sky (a surprisingly rare combination
of conditions for sailors), plus correction for the unexpectedly large parallax
effects in determining the horizon, which may be hazy or worse and is affected
by the observer’s height above the water. Latitude can also be inferred from
the sun’s elevation, though this requires correcting for the date (the sun’s arc
changes over a range of 47° over the course of the year), accurately judging solar
noon, and (of course) having a stable base and clear skies. As Edmund Halley
showed hundreds of years ago, the inclination of the earth’s magnetic field,
corrected at least roughly for location, can yield latitude regardless of weather.
Longitude is more difficult (Gould, 2008). Humans measure it by keeping
track of the time difference between a point of known longitude (ultimately,
Greenwich, England) and the current location (typically measured at solar noon
and correcting for the sun’s ephemeris—an irregularity caused by the earth’s ellip-
tical orbit around the sun). Clocks accurate enough for this strategy were not
affordably available about 200 years ago. Alternative methods used earlier involved
observing the eclipses of the moons of Jupiter or measuring the angle between the
moon and background stars (correcting for refraction effects at lower celestial eleva-
tions). Prior to that, the aptly named strategy of dead reckoning was the standard
approach. The problem of longitude for humans is summarized by the common
mantra “location is time.” Now, of course, atomic clocks on orbiting satellites deal
with this challenge for humans. Animals take one or more alternative approaches.
402 ANIMAL BEHAVIOR

Evidence that animals have a map sense came initially from homing pigeons,
where the accuracy of return in the absence of visual cues is an astonishing 2 to
4 km (Schmidt-Koenig & Walcott, 1978). It is clear now, however, that a sim-
ilar or identical ability is present in a variety of migrating or wide-ranging species
when subjected to similar displacement experiments. The animals are typically
captured during migration (when highly motivated to orient accurately) and
taken in sensory isolation to a new location often thousands of kilometers away,
generally well to the east or west of what is typically a north-south route
(Åkesson et al., 1995; Chernetsov et al., 2008; Thorup et al., 2007). Imagine a
test in which the birds were migrating south and were 1,000 km from their win-
tering grounds when intercepted (Figure 12.7). They are moved east 1,000 km.

Figure 12.7. Migrational displacement experiment. In this idealized example summarizing

numerous tests, birds are migrating south from their breeding area. They are intercepted
1,000 km from the wintering grounds and taken in sensory isolation to a release point 1,000
km east. From this point they might (1) continue south oblivious to the displacement, (2)
return west to the capture site, or (3) set course along a novel heading to the original target,
now to the southwest. First-time migrants depart south; experienced birds fly southwest.
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 403

When released, the birds might (1) continue along the vector they were flying
when captured (south), or (2) compute their present location and head back to
the capture point to pick up the journey where it was interrupted (west), or (3)
compute their current position and fly toward their original destination (south-
west). First-year migrants typically continue on their former bearing, basically
flying into oblivion (alternative 1—south in the example used here); experienced
birds set a novel course for their original target (alternative 3—southwest).
These results show that measuring local time within a few seconds is simply
not part of the animal solution to the problem. (And, indeed, the map sense is
unaffected by clock-shifting, an even more dramatic test.) Moreover, some-
thing important is learned (or calibrated) during the first autumn flight—
information that allows older birds to behave as though they have a map sense.
Many species will not orient correctly except during migration, so research has
tended to focus on those that are, for reasons best known to themselves, highly
motivated all year—notably homing pigeons, sea turtles, spiny lobsters, and
even newts.
It is important to realize that displacement tests cannot easily distinguish
between a true map sense—global position—and an ability to judge relative
location (distance and direction from home or the goal). In practice this dis-
tinction is not relevant to animals, but it has great theoretical import to
researchers. In my view, there is no convincing reason to believe animals know
global position, and several lines of evidence suggest that the relative-location
strategy is the approach they are employing (Gould & Gould, 2012).
The key questions are what cues are being used to judge position, how is
that information being processed, and what experience does to calibrate the
behavior. Already we have seen that a failure to distinguish between first-year
and experienced fall migrants will produce scattered and misleading data
because the two groups deduce quite different answers to the puzzle presented
by large-scale displacement. A similar pattern bedeviled homing-pigeon
research for years: younger birds gather orientational information during dis-
placement and rely mainly on their magnetic compass when set free; older
birds, on the other hand, deduce position once they are at the release site
and depend mostly on the sun as a navigational beacon (Wiltschko &
Wiltschko, 1985; Wiltschko, 1991). Using a mix of ages in a single release cre-
ated enormous confusion.
This change in the homing strategy is a result of experience. Birds that have
been more restricted in their flights around the home loft take longer to switch
over; those given extra training (a kind of environmental enrichment) mature
sooner; those not permitted to range outside the loft never grow up naviga-
tionally. This last highly unnatural protocol for rearing, which allows no
404 ANIMAL BEHAVIOR

free-flight experience during the sensitive period for calibrating and switching
strategies, is the basis for a large group of otherwise well-controlled experi-
ments implicating odors blown in from distant sources as the basis of map
behavior (Wallraff, 2005).
For normally reared pigeons—those allowed to nest in colonies (aka lofts)
and fly about locally to forage—odors appear to be irrelevant under most con-
ditions; the surgical trauma involved in some tests seeking to obliterate the
birds’ sense of smell does have dramatic effects, and odors do have some sort
of activational effect on the true map system, however (Jorge et al., 2009).
While it is interesting that pigeons have a short-range olfactory backup system,
this seems largely unimportant in their normal lives after growing up under
natural conditions. Odor seems an unlikely cue to be relevant or useful to
long-distant migrants in determining position.
The most compelling hypothesis to account for the map-like ability of so
many animals imagines that the travelers use magnetic field cues. A long list
of indirect experiments implicated magnetic information in homing pigeons
as early as 1980 (Gould, 1980; Moore, 1980; Walcott, 1980); much more
has been written to fill out the story since (Gould, 1982b; Phillips, 1996;
Walker, 1998; Freake et al., 2006; Lohmann et al., 2007). Errors in initial
departure directions, for instance, are related to the gradient of the local field
(Keeton, 1973; Windsor, 1975) and change predictably during magnetic field
storms, when the earth’s field strength varies by 1 to 5 percent as a result of
sunspot activity (Keeton et al., 1974; Kowalski et al., 1988). These storms also
delay return times (Dornfeldt, 1996). Even a 5 percent change has no visible
effect on a magnetic compass, however. Most dramatically, pigeons released
on sunny days from sites of pronounced anomalies in the magnetic field fly
about almost at random until they are well away from the disturbance (usually
generated by large iron deposits underground), whereas at magnetically nor-
mal sites the birds leave well oriented toward home (Figure 12.8; Walcott,
1978).
If pigeons and long-distance migrants are using the earth’s field to place
themselves, how would this system work? Nearly all hypotheses converge on
the same two parameters and ontogeny of calibration (Gould & Gould,
2012). There are gradients in inclination (or vertical field strength, which
measures the same thing) and total field strength that run from pole to pole
but are not aligned with each other (Figure 12.9). In most parts of the world
they form a skewed grid. For a homing pigeon, measuring the direction and
slope of these two gradients would be crucial. This is probably part of what
they are doing in those several weeks of “play flights” that precede the shift
from using map cues gathered en route to a release site, as opposed to the
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 405

Figure 12.8. Effects of magnetic topography. Total magnetic field strength is repre-
sented here as contour lines. The normal pattern in the northeastern United States
is a gradual increase in strength to the north-northwest. Left: At the iron mine the
pigeons are initially disoriented. Right: At Wooster, Massachussets, where the field is
relatively normal, departure bearings are far more accurate and consistent. The back-
ground field at both sites is approximately 50,000 gamma.

ability to deduce location once there. Thus an older bird at a release site would
measure inclination and total strength; extrapolating from the gradients at
home, the pigeon would infer its current location and thus the compass bear-
ing back to the loft. To the extent that the direction and slope near the loft are
not completely typical of the general pattern in the region, there should be sys-
tematic errors in initial orientation at the release site, and this is exactly what is
observed. As they fly back, the older birds should be able to take new map
readings and correct for these initial errors—again, pretty much what tracking
records show (e.g., Michener & Walcott, 1967). Irregularities at the specific
release site should introduce consistent departure errors—mistakes that will
be affected by field changes induced by magnetic storms (again, a pattern seen
in the literature). A variety of tests implicate the magnetite organ as the loca-
tion sensor (e.g., Munro et al., 1997).
For many years the debate about the map sense turned on these correlations
and several experiments that sought to interfere with magnetic-field detection
at the release site. More recently, however, direct tests have pretty much set-
tled the issue for a variety of animals (though not, for technical reasons,
406 ANIMAL BEHAVIOR

Figure 12.9. Magnetic gradients (Gould & Gould, 2012). The gradients of total and
vertical magnetic-field intensity intersect at about a 30° angle in the northeastern
United States

homing pigeons). The experiments involve collecting migrants or homing spe-

cies and testing them immediately at or near the capture site. The animals are
placed in cueless chambers inside magnetic coils. The experimenters then set
the coils for an arbitrary set of magnetic coordinates—inclination and total
strength. With nothing more than an altered magnetic field to go by, the crea-
tures take up reliable headings that would bring them back to the capture site
had they actually been physically moved anywhere (Lohmann et al., 1995,
2001, 2004; Fischer et al., 2001; Phillips et al., 2002; Boles & Lohmann,
2003; Lohmann, 2007; Fuxjager et al., 2011; Putman et al., 2011). (Pigeons
refuse to cooperate in testing chambers; researchers in general assume that
the pigeon map is not a special creation and probably operates along the same
lines as the position-sense system of other homers. Pigeons can be conditioned
to isolated magnetic-field components, however [Dennis et al., 2007].)
As should be clear from the putative ontogeny of the pigeon “map,” the
birds do not have a global system; instead they know only their position rela-
tive to home. The situation for long-distance migrants on their initial trip,
before they have an opportunity to gather map information about the destina-
tion, is probably different. While the innate information used in that first fall
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 407

includes vectors (bearings and distances), this will not normally be sufficient
for finding a localized target given the inevitable measurement errors and drift.
Could magnetic coordinates be innately encoded? This seems to be the case for
sea turtles in the Atlantic. Although they imprint on the magnetic parameters
of the home beach (enabling them to return to nearly the same exact spot years
later), their task for the first three years of life is to feed and mature within the
North Atlantic gyre in the Sargasso Sea.
Given the speed of the currents relative to the swimming ability of the
hatchlings and the fatal consequences of being swept along by any of the mas-
sive currents that diverge from the gyre, their survival strategy is to aim them-
selves into the Sargasso regardless of position. But the appropriate compass
angle varies over 360° as they move from Florida across the Atlantic, then
south along the north African coast, then west to the Caribbean, then north
to Florida. And yet the hatchlings, only a few hours old and in a featureless test
chamber surrounded by coils, reliably aim themselves into the gyre from any
plausible virtual location the researcher chooses to provide in the coils
(Fuxjager et al., 2011). This suggests that sea turtles have a set of innate mag-
netic coordinates which they use until they are old enough to exit the current
and take up coastal feeding as adults. For long-distance migrants flying their
first year to localized targets, a set of magnetic coordinates would seem a criti-
cal piece of information for arriving at their particular destination.
The possibility of a human map sense received a boost when Baker (1980)
reported that blindfolded students could point back to their college after trans-
port wearing blindfolds. Moreover, placing magnets on the forehead suppos-
edly interfered with this ability. Numerous subsequent tests (e.g., Gould &
Able, 1981; Fides et al., 1984) showed that this phenomenon disappears when
better controls are used. Unaided humans seem to be limited to local-area cog-
nitive maps and an ability to maintain a constant sun-compass bearing; some
recent work suggests an unconscious ability to compensate for the sun’s east-
to-west movement (Souman et al., 2009). Despite our many cultural and sci-
entific accomplishments, without a GPS or other elaborate navigational gear
our species is basically clueless when it comes to getting from here to there.

REFERENCES AND SUGGESTED READING

Able, K. P. & M. A. Able. (1990). Calibration of the magnetic compass of a migratory
bird by celestial rotation. Nature, 347, 378–389.
Able, K. P. & M. A. Able. (1995). Interactions in the flexible orientation system of a
migratory bird. Nature, 375, 230–232.
Adler, J. (1976). Chemotaxis in bacteria. Annual Review of Biochemistry, 44,
341–356.
408 ANIMAL BEHAVIOR

Åkesson, S., U. Ottoson, & R. Sandberg. (1995). Bird migration: Displacement

experiments with young autumn migrating wheatears along the Arctic coast
of Russia. Proceedings of the Royal Society of London, B, 262, 189–195.
Baker, R. R. (1980). Goal orientation by blindfolded humans after long-distance dis-
placement: Possible involvement of a magnetic sense. Science, 210, 555–557.
Berthold, P., A. J. Helbig, G. Mohr, & U. Querner. (1992). Rapid microevolution of
migratory behavior in a wild bird species. Nature, 360, 668–670.
Boles, L. C. & K. J. Lohmann. (2003). True navigation and magnetic maps in spiny
lobsters. Nature, 421, 60–63.
Brines, M. L. & J. L. Gould. (1982). Skylight polarization patterns and animal orien-
tation. Journal of Experimental Biology, 96, 69–91.
Chernetsov, N., D. Kishkinev, & H. Mouritsen. (2008). A long-distance avian
migrant compensates for longitudinal displacement during spring migration.
Current Biology, 18, 188–190.
Davis, L. A., E. H. Roalson, K. L. Cornell, K. D. McClanahan, & M. S. Webster.
(2006). Genetic divergence and migration patterns in a North American pass-
erine bird: Implications for evolution and conservation. Molecular Ecology, 15,
2141–2152.
Dennis, T. E., M. J. Rayner, & M. M. Walker. (2007). Orientation to geomagnetic
intensity in homing pigeons. Proceedings of the Royal Society of London, B,
274, 1153–1158.
Dornfeldt, K. (1996). Pigeon homing in the meteorological and solar-geomagnetic
environment: What pigeon race data say. Ethology, 102, 413–435.
Durgin, F. H., M. Akagi, C. R. Gallistel. (2008). The precision of locomotor
odometry in humans. Experimental Brain Research, 193, 429–436.
Dyer, F. C. & J. L. Gould. (1981). Honey bee orientation: A backup system for
cloudy days. Science, 214, 1041–1042.
Emlen, S. T. (1967a). Migratory orientation in the indigo bunting. Part I: Evidence
for the use of celestial cues. Auk, 84, 309–342.
Emlen, S. T. (1967b). Migratory orientation in the indigo bunting. Part II: Mecha-
nisms of celestial orientation. Auk, 84, 463–489.
Emlen, S. T. (1970). Celestial rotation: Its importance in the development of migra-
tory orientation. Science, 170, 1198–1201.
Etienne, A. S., J. Maurer, & V. Séguinot. (1996). Path integration in mammals and
its interaction with visual landmarks. Journal of Experimental Biology, 199,
201–209.
Fides, B. N., B. J. O’Loughlin, J. L. Bradshaw, & W. J. Evans. (1984). Human ori-
entation with restricted sensory information: No evidence of for magnetic sen-
sitivity. Perception, 13, 229–236.
Fischer, J. H., M. J. Freake, S. C. Borland, & J. B. Phillips. (2001). Evidence for the
use of magnetic map information by an amphibian. Animal Behaviour, 62,
1–10.
Fraenkel, G. S. & D. L. Gunn. (1940). The Orientation of Animals: Kineses, Taxes,
and Compass Reactions. Oxford, UK: Oxford University Press.
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 409

Freake, M. J., R. Muheim, & J. B. Phillips. (2006). Magnetic maps in animals: A

theory comes of age? Quarterly Review of Biology, 81, 327–347.
Frisch, K. V. (1967). The Dance Language and Orientation of Bees. Cambridge, MA:
Harvard University Press.
Fuxjager, M. J., S. Brian, B. S. Eastwood, & K. J. Lohmann. (2011). Orientation
of hatchling loggerhead sea turtles to regional magnetic fields along a trans-
oceanic migratory pathway. Journal of Experimental Biology, 214,
2504–2508.
Gallistel, C. R. & A. E. Cramer. (1996). Computations on metric maps in mammals:
Getting oriented and choosing a multi-destination route. Journal of Experimen-
tal Biology, 199, 211–217.
Gallistel, C. R. & J. Gibbon. (2000). Time, rate, and conditioning. Psychological
Review, 107, 289–344.
Gill, R. E., T. Piersma, G. Hufford, et al. (2005). Crossing the ultimate ecological
barrier: Evidence for an 11,000-km-long nonstop flight from Alaska to New
Zealand and Eastern Australia by bar-tailed godwits. Condor, 107, 1–20.
Gill, R. E., T. L. Tibbitts, D. C. Douglas, C. M. Handel, D. M. Mulcahy, J. C.
Gottschalck, et al. (2008). Extreme endurance flights by landbirds crossing
the Pacific Ocean: Ecological corridor rather than barrier? Proceedings of the
Royal Society of London, B, 276, 447–457.
Gould, J. L. (1980). The case for magnetic sensitivity in birds and bees. American Sci-
entist, 68, 256–267.
Gould, J. L. (1982a). Ethology: The Mechanisms and Evolution of Behavior. New York:
W. W. Norton.
Gould, J. L. (1982b). The map sense of pigeons. Nature, 296, 205–211.
Gould, J. L. (1986). The locale map of honey bees: Do insects have cognitive maps?
Science, 232, 861–863.
Gould, J. L. (1995). Constant compass calibration. Nature, 375, 184.
Gould, J. L. (2008). Animal navigation: The longitude problem. Current Biology, 18,
R214–R216.
Gould, J. L. & K. P. Able. (1981). Human homing: An elusive phenomenon. Science,
212, 1061–1063.
Gould, J. L. & C. G. Gould. (1995). The Honey Bee. Second Edition. New York: W.
H. Freeman.
Gould, J. L. & C. G. Gould. (1999). The Animal Mind. Second Edition. New York:
W. H. Freeman.
Gould, J. L. & C. G. Gould. (2012). Nature’s Compass: The Mystery of Animal Migra-
tion. Princeton, NJ: Princeton University Press.
Gwinner, E. & W. Wiltschko. (1978). Endogenously controlled changes in migratory
direction of the garden warbler. Journal of Comparative Physiology, 125, 267–
273.
Haney, J. F. (1988). Diel patterns of zooplankton behavior. Bulletin of Marine
Science, 43, 583–603.
410 ANIMAL BEHAVIOR

Helbig, A. J. (1991). Inheritance of migratory direction in a bird species: A cross-

breeding experiment with SE- and SW-migrating blackcaps. Behavioral Ecology
and Sociobiology, 28, 9–12.
Helbig, A. J. (1996). Genetic basis, mode of inheritance, and evolutionary changes of
migratory directions in palearctic warblers. Journal of Experimental Biology,
199, 49–55.
Jacobs, L. F. & F. Schenk. (2003). Unpacking the cognitive map: The parallel map
theory of hippocampal function. Psychological Review, 110, 285–315.
Jorge, P. E., A. E. Marques, & J. B. Philips. (2009). Activational rather than naviga-
tional effects of odors on pigeon homing. Current Biology, 19, 650–654.
Keeton, W. T. (1969). Orientation by pigeons: Is the sun necessary? Science, 165,
922–928.
Keeton, W. T. (1973). Release site bias as a possible guide to the “map” component in
pigeon homing. Journal of Comparative Physiology, 86, 1–16.
Keeton, W. T., T. S. Larkin, & D. T. Windsor. (1974). Normal fluctuations in the
earth’s magnetic field influence pigeon orientation. Journal of Comparative
Physiology, 95, 95–103.
Kennedy, J. S., A. R. Ludlow, & C. J. Sanders. (1980). Guidance system used in
moth sex attraction. Nature, 288, 475–477.
Kowalski, U., R. Wiltschko, & E. Füller. (1988). Normal fluctuations of the geomag-
netic field may affect initial orientation in pigeons. Journal of Comparative
Physiology, A, 163, 593–600.
Lohmann, K. J. (2007). Sea turtles: Navigating with magnetism. Current Biology, 17,
R102–104.
Lohmann, K. J., S. D. Cain, S. A. Dodge, & C. M. F. Lohmann. (2001). Regional
magnetic fields as navigational markers for sea turtles. Science, 294, 364–366.
Lohmann, K. J., C. M. F. Lohmann, L. M. Ehrhart, D. A. Bagley, & K. Swing.
(2004). Geomagnetic map used in sea turtle navigation. Nature, 428, 909.
Lohmann, K. J., C. M. F. Lohmann, & N. F. Putman. (2007). Magnetic maps in ani-
mals: Nature’s GPS. Journal of Experimental Biology, 210, 3697–3705.
Lohmann, K. J., N. D. Pentcheff, G. A. Nevitt, G. Stetten, R. Zimmer-Faust, H.
Jarrard, et al. (1995). Magnetic orientation of spiny lobsters in the ocean:
Experiments with undersea coil systems. Journal of Experimental Biology, 198,
2041–2048.
McCaffery, B. J. (2008). On scimitar wings. Birding, 40(5), 50–59.
McNaughton, B. L., C. A. Barnes, J. L. Gerrard, K. Gothard, M. W. Jung, J. J.
Knierim, et al. (1996). Deciphering the hippocampal polyglot: The hippocam-
pus as a path-integration system. Journal of Experimental Biology, 199, 173–185.
Menzel, R., U. Greggers, A. Smith, et al. (2005). Honey bees navigate according to a
map-like spatial memory. Proceedings of the National Academy of Sciences,
USA, 102, 3040–3045.
Michener, M. & C. Walcott. (1967). Homing of single pigeons: Analysis of tracks.
Journal of Experimental Biology, 47, 99–131.
GETTING THERE : ANIMAL ORIENTATION, NAVIGATION , AND MIGRATION 411

Mittelstaedt, M. & H. Mittelstaedt. (2001). Idiothetic navigation in humans: Estima-

tion of path lengths. Experimental Brain Research, 139, 318–332.
Moore, B. (1980). Is the homing pigeon’s map magnetic? Nature, 285, 69–70.
Munro, U., J. Munro, & J. B. Phillips (1997). Evidence for a magnetite-based naviga-
tional “map” in birds. Naturwissenschaften, 84, 26–28.
Phillips, J. B. (1996). Magnetic navigation. Journal of Theoretical Biology, 180,
309–319.
Phillips, J. B., M. J. Freake, J. H. Fischer, & S. C. Borland. (2002). Behavioral titra-
tion of a magnetic map coordinate. Journal of Comparative Physiology A, 188,
157-160.
Pulido, F. (2007). The genetics and evolution of avian migration. BioScience, 57,
165–174.
Pulido, F., P. Berthold, G. Mohr, & U. Querner. (2001). Heritability of the timing of
autumn migration in a natural bird population. Proceedings of the Royal Society
of London, B, 268, 953–959.
Putman, N. F., C. S. Endres, C. M. F. Lohmann, & K. J. Lohmann. (2011). Longi-
tude perception and bicoordinate magnetic maps in sea turtles. Current Biol-
ogy, 21, 463–466.
Schmidt-Koenig, K. & W. T. Keeton (eds.). (1978). Animal Migration, Navigation,
and Homing. Berlin: Springer-Verlag.
Schmidt-Koenig, K. & C. Walcott. (1978). Tracks of pigeons homing with frosted
lenses. Animal Behaviour, 26, 480–486.
Souman, J. L., I. Frissen, M. N. Sreenlvasa, & M. O. Ernst. (2009). Walking straight
into circles. Current Biology, 19, 1538–1542.
Tarsitano, M. S. & R. R. Jackson. (1994). Jumping spiders make predatory detours
requiring movement away from prey. Behaviour, 131, 65–73.
Thorup K., I. A. Bisson, M. S. Bowlin, R. A. Holland, J. C. Wingfield, M. Ramenofsky,
et al. (2007). Evidence for a navigational map stretching across the continental
U.S. in a migratory songbird. Proceedings of the National Academy of Sciences,
USA, 104, 18115–181159.
Walcott, C. (1978). Anomalies in the Earth’s magnetic field increase the scatter of
pigeons’ vanishing bearings. In K. Schmidt-Koenig & W. T. Keeton (eds.),
Animal Migration, Navigation, and Homing (pp. 143–151). Berlin: Springer-
Verlag.
Walcott, C. (1980). Magnetic orientation in homing pigeons. IEEE Transactions on
Magnetics, 16, 1008–1013.
Walker, M. M. (1998). On a wing and a vector: A model for magnetic navigation by
homing pigeons. Journal of Theoretical Biology, 192, 341–349.
Wallraff, H. G. (2005). Avian Navigation: Pigeon Homing as a Paradigm. Berlin:
Springer-Verlag.
Walls, M. L. & J. E. Layne. (2009). Direct evidence for distance measurement via
flexible stride integration in the fiddler crab. Current Biology, 19, 25–29.
Wills, T. J., F. Cacucci, N. Burgess, & J. O’Keefe. (2010). Development of the
hippocampal cognitive map in preweanling rats. Science, 328, 1573–1576.
412 ANIMAL BEHAVIOR

Wiltschko, R. (1991). The role of experience in avian navigation and homing. In P.

Berthold (ed.), Orientation in Birds (pp. 250–269). Basel, Germany:
Birkhäuser Verlan.
Wiltschko, R. & W. Wiltschko. (1985). Pigeon homing: Change in navigational
strategy during ontogeny. Animal Behaviour, 33, 583–590.
Wiltschko, R. & W. Wiltschko (eds.). (1995). Magnetic Orientation in Animals.
Berlin: Springer-Verlag.
Windsor, D. M. (1975). Regional expression of directional preferences by experi-
enced homing pigeons. Animal Behaviour, 23, 335–343.
Wittlinger, M., R. Wehner, & H. Wolf. (2006). The ant odometer: Stepping on stilts
and stumps. Science, 312, 1965–1967.
Wu, L.-Q. & D. Dickman. (2011). Magnetoreception in an avian brain in part medi-
ated by inner ear lagena. Current Biology, 21, 418–423.
 Glossary

Abnormal psychology—The study of unusual patterns of behavior, emotion, and

thought.
Accuracy—The closeness of a measurement to the actual value.
Action potential—A brief change in the electrical potential of the cell membrane of
a neuron. Action potentials move along the axons of neurons to convey signals from
one part of the nervous system to another.
Activity—Awake behavior of the animal in studies of circadian rhythm, commonly
measured as number of revolutions for wheel running in mice and locomotion in
flies.
Adaptation—The process of adjustment of an organism to its environment. The pro-
cess of evolutionary modification that improves an organism’s survival and reproduc-
tive success. A trait that evolved by natural selection for its current functional role.
Adaptive—A trait that enhances the survival and reproductive success of its bearer.
Additive effect—The quantitative change in a trait that is associated with substitut-
ing alleles for that trait.
Additive genetic variation—The portion of the genetic variance of a quantitative
character attributed to the average effects of substituting one allele for another at a
given locus or at the multiple loci governing a polygenic trait.
Adrenal glands—Endocrine glands of mammals that sit on top of the kidneys.
Adrenarche—The increase in activity of the adrenal glands just before puberty.
Adrenocorticotropic hormone (ACTH)—A peptide hormone produced and
secreted by the anterior pituitary gland that stimulates the release of corticosteroids.
414 GLOSSARY

Agent-based simulation—A simulation in which the modeler identifies a collection

of autonomous decision-making entities (agents), defines their behavior, puts them
in a certain environment, and establishes connections.
Alarm call—A signal produced by an animal in response to danger.
Alloparental care—Care given to young that are not offspring.
Alternative mating strategies—Distinct behaviors, sizes, colorations, and body
forms of one species that have evolved to secure mates.
Altricial—Hatched or born blind and helpless.
Altruism—Behavior that benefits other organisms at a cost to the individual per-
forming the behavior.
Amplexus—A form of mating in which a male amphibian grasps a female with his
front legs to enable him to fertilize the female’s eggs.
Androgen—A steroid hormone that controls the development and maintenance of
masculine characteristics.
Animal behavior—The study of the causation, development, function, and evolu-
tion of behavior.
Animal husbandry—The controlled cultivation, management, and production of
domestic animals by means of breeding to improve the qualities considered desirable
by humans.
Animal learning psychology—The study of learning in animals from a psychologi-
cal perspective.
Animal models—Living, nonhuman animals used in research to investigate human
biology and thus avoiding harm to humans.
Animal psychology—The study of animal behavior from a psychological
perspective.
Antagonist—A chemical that blocks binding to a particular receptor.
Ant dipping—A method of foraging for ants in which chimpanzees push sticks into
nests.
Anthropoids—Monkeys, apes, and humans.
Anthropomorphism—Attributing human characteristics (or characteristics assumed
to belong only to humans) to nonhuman animals.
Antireductionists—Those who see the programs and processes of behavior as being
so completely integrated that dissection into constituent elements destroys the very
thing being studied.
Appetitive behavior—Activity that increases the likelihood of satisfying a specific
motivation.
Arrhythmic—Lacking a pattern of activity.
Associative learning—The process by which an association between two stimuli or
a behavior and a stimulus is learned.
GLOSSARY 415

Assortative mating—A mating preference in which similar individuals mate with

one another.
Azimuth—An angular measurement in a spherical coordinate system.
Balance—In statistics, use of equal numbers of replicates or methods to mitigate
potential effects of order or other confounds.
Baldwin effect—The interaction of evolution with learning by individual animals
over their lifetimes.
Begging—Behavior by which depending young solicit care from their parents and
other caregivers.
Behavioral ecology—An approach to investigations of how different ecological
circumstances affect animal behavior and the evolution of traits with functions that
fit these different environments.
Behavioral neuroscience—The study of the mechanisms of behavior from a neuro-
science perspective.
Behavioral syndrome—Correlation of rank-order differences among individuals
through time or across situations.
Behaviorism—All animal actions can be explained purely by objective observations,
with no recourse to or need for involving introspection or other internal mecha-
nisms.
Behaviorists—Psychologists who held that instinct does not exist and that every-
thing beyond unconditioned stimuli and reflexes is learned.
Bias—An improper design that produces a systematic difference between a measure-
ment and its actual value.
Biogeography—The study of the distribution of species, organisms, and ecosystems
in space and through geological time.
Biparental care—When both parents provide care that increases the fitness of their
offspring.
Blocking (matching)—Arranging experimental units into groups (blocks) that are
similar to one another.
Bower—A structure built by a male bowerbird to attract mates.
Breeding designs—Use of information about who is related to whom to distinguish
between additive, dominance, and epistatic effects of genes.
Brood parasite—A species of animal that uses a host of the same species (intraspe-
cific brood parasite) or a different species (interspecific brood parasite) to care for its
own offspring.
Brown adipose tissue (brown fat)—One of two types of fat found in mammals,
especially in newborns and hibernators, that functions to generate body heat in ani-
mals that do not shiver.
416 GLOSSARY

Canalization—Differentiation of the gene, cell type, or embryo during develop-

ment as pathways become more entrenched, thereby making it harder to move into
another pathway.
Cathemeral—The behavior of animals with activity distributed approximately
evenly across the 24-hour cycle.
Celestial navigation—A global positioning mechanism that uses the sun or other
stars to advance to a destination.
Cell culture—The complex process by which cells are grown under controlled
conditions.
Central pattern generator—A neural network that produces rhythmic patterned
outputs (e.g., walking) without sensory feedback.
Chronobiologist—A biologist who studies clock phenomena.
Chronotype—An attribute reflecting the time of the day an animal is active.
Circadian rhythm—An endogenous activity rhythm with a period of about
24 hours.
Circalunar—A period of about one lunar cycle (29.5 days).
Circannual rhythm—An endogenous activity rhythm with a period of about a
year.
Classical learning—A learning process that occurs through associations between an
environmental stimulus and a naturally occurring stimulus.
Classification—A method to group and categorize organisms by biological type,
such as genus or species.
Clock—Circadian locomotor output cycles kaput is a gene encoding a transcription
factor (CLOCK) that affects both the persistence and period of circadian rhythms.
Clock genes—Genes that affect the persistence and period of circadian rhythms.
Clock-shifted—A light/dark cycle out of phase with the earth’s 24-hour light/dark
cycle.
Cognitive ethology—The study of animal cognition to understand the evolution,
function, causation, and development of the species-specific behavioral repertoire.
Cognitive map—A map-like mental representation of a web of paths.
Communication—The use of signals to change or modify the behavior of other
individuals.
Comparative method—A means of study that is based on comparisons of different
species.
Comparative psychology—The scientific study of the behavior and mental processes
of nonhuman animals from a psychological perspective.
Compass response—Orientation at a fixed angle.
GLOSSARY 417

Completely randomized design—An experimental design that allows the study of

the effects of one or more factors while controlling for confounding factors.
Confounded—An improper experimental design that prevents the proper analysis of
treatment effects.
Confounding variable—An unforeseen and uncontrolled variable that prevents
proper analysis of a treatment effect.
Conspecific—A member of the same species.
Consummatory behavior—Activity that satisfies a specific motivation.
Context-dependent epigenetic modifications—A change in the epigenome that is
in direct response to the stimulus.
Contingency—Dependence of an event on the fulfillment of a condition.
Control—An experimental method that eliminates extraneous effects on a process
of interest.
Convection—Current caused by uneven distribution of heat. For example, warm
rat pups at the center of a huddle flow upward as cooler pups at the periphery flow
towards the center.
Correlation—A statistical method that attempts to identify associations among vari-
ables that are not functionally dependent.
Corticotropin-releasing hormone (CRH)—A peptide hormone that stimulates the
pituitary synthesis of ACTH.
Countersinging—Singing in response to the songs of another individual or other
individuals.
Crepuscularity—The behavior of animals that are active during the transition times
of dawn and dusk.
Crop milk—A secretion from the lining of the crop of pigeons and doves with
which the parents feed their young by regurgitation.
Cryptochrome—A class of flavoproteins sensitive to blue light and involved in circa-
dian rhythms and sensing magnetic fields in many animals.
Culture—Sharing of behavior by a group of animals through social transmission
among peers and between generations.
Cycle—A gene that encodes a transcription factor (CYCLE) that affects the persis-
tence and period of circadian rhythms.
Dead reckoning (inertial navigation)—Use of known or estimated speeds and
directions to advance from the current position.
Declination—The angle between magnetic north and true north.
Deductive reasoning—Construction or evaluation of arguments that attempt to
show that a conclusion necessarily follows from a set of premises or hypotheses.
Defender—In game theory, a strategy of defending an important resource such as a
bower.
418 GLOSSARY

Dependent variable—A measured quantity that is thought to vary in response to

an independent variable.
Design of layout—How experimental treatments are assigned to experimental units
(what treatment a subject receives).
Design of treatments—How experimental treatments relate to one another (e.g.,
factorial, nested, graded).
Destroyer—In game theory, a strategy of defending an important resource such as a
bower and visiting other bowers to destroy them.
Development—The sequence of changes that occur as an individual grows and
matures throughout its life.
Dewlap—A longitudinal flap of skin that hangs beneath the lower jaw or neck of
many vertebrates.
Diffusion-chain experiment—An iterated procedure to test for social learning in
which the observer in a two-action experiment is used as a demonstrator for a new
naïve observer, which then becomes a new demonstrator, and so forth.
Direct development—A life cycle without a larval stage and without metamorpho-
sis, in which individuals hatch or are born with only minor differences from adults.
Displacement activity—A seemingly inappropriate behavior thought to result from
two conflicting motivations in a particular situation.
Dissociation—Changes in the time elapsed between (or relationship among) ele-
ments of the circadian cycle including the light/dark cue, the initiation of gene
expression, and the awake behavior of the animal.
Diurnality—The behavior of animals that are active in the daytime.
Divination—The art of foretelling future events or discovering knowledge by inter-
preting omens or signs.
Dominance—The effect of one allele at a locus being partially or entirely hidden by
another allele at the same locus. A social relationship in which one individual has
power or priority of access over another.
Dominance hierarchy—A ranking of individuals in a social group based on their
dominance relationships.
Dominance relationship—Social relationship in which one individual has power or
priority of access over the other.
Do-this experiment—A procedure to test for imitation in which subjects are first
trained to repeat arbitrary actions performed by a human demonstrator when the
demonstrator says, “do this” and then are presented with new arbitrary actions to
see if these are also copied when the human demonstrator says, “do this.”
Doubletime—A gene that encodes a protein (DBT) that alters the period of circa-
dian rhythms in Drosophila.
Dove strategy—In game theory, a strategy of displaying but never fighting (escalating).
GLOSSARY 419

Duration—How long a behavior is performed.

Ecological crossover—Different genotypes are superior in environments that are
separated either in space or time.
Ectoparasite—A parasite that lives on the surface of the host.
Ectothermic—An organism that controls body temperature through external means.
Egg eviction—Ejection of one or more host eggs by a newly hatched brood parasite.
Egg rejection—Removal of an egg from the nest.
Electrophysiology—The study of the electrical properties of cells and tissues.
Embryological development—The events that occur during early development of
an organism.
Endocrine disruptor chemicals (EDCs)—Chemicals that interfere with the hor-
mone system in animals, including humans.
Endocrine glands—Glands of the endocrine system that secrete hormones directly
into the bloodstream.
Endocrine system—The system of glands, each of which secretes a type of hormone
directly into the bloodstream to regulate the body.
Endogenous clock—A roughly 24-hour activity cycle that does not rely on environ-
mental cues.
Endoparasite—A parasite that lives within another organism.
Endoplasmic reticulum—A cellular organelle in eukaryotic organisms that forms an
interconnected network of tubules, vesicles, and cisternae and functions as a packag-
ing system.
Endothermic—An organism that generates heat to maintain its body temperature.
Entrainment—A process by which a biological rhythm is synchronized by an envi-
ronmental periodicity.
Enumeration methods—Statistical methods to analyze counts (frequencies of
occurrence).
Environmentally induced epigenetic modifications—Changes in the epigenome in
response to environmental factors.
Epaulets—Conspicuous feather patches on the wings of birds such as the red-
winged blackbird.
Epialleles—Genes that differ in the extent of methylation but are otherwise identical.
Epigenesis—The development in an organism through a sequence of steps in which
cells differentiate and organs form; the belief that animals develop through a
sequence of steps (cf. preformationism).
Epigenetic effects—Changes in the phenotype or specific traits that result from the
environmental modification of the molecular factors and processes around DNA
that regulate genome activity yet are independent of the DNA sequence.
420 GLOSSARY

Epigenetics—Traits determined by more than traditional molecular bases for

inheritance.
Epimutation—Heritable change in gene function without changing the DNA
sequence.
Epistasis—The effects of one gene being modified by one or several other (modi-
fier) genes.
Equipotentiality—In response to brain damage in certain areas, other parts of the
brain being able to assume some of the roles normally carried out by that damaged
area.
Estrogen—A steroid hormone formed by the ovary, placenta, and testis that stimu-
lates female secondary sex characteristics.
Ethogram—A record of all of the behavior patterns and activities of an organism.
Ethologists—Behavioral biologists who maintain that instinct dominates behavior
and even guides much of learning.
Ethology—The scientific study of animal behavior from a biological perspective.
Evo-devo (evolutionary developmental biology)—A field of biology that compares
the developmental processes of different organisms to discover how those processes
evolved.
Evolution—A change across successive generations in the heritable characteristics of
biological populations (descent with modification).
Evolutionarily stable strategy (ESS)—In game theory, a strategy that cannot be
invaded by any other strategy.
Evolutionary history—The lineage of a behavior from its initial appearance in
evolution.
Exaptation—A process by which a character acquires a function for which it was
not originally selected. A character having a function for which it was not originally
selected.
Exocytosis—The process by which a cell sends the contents of secretory vesicles
across the cell membrane and into the extracellular space.
Experimental study—A type of research that uses manipulation and controlled test-
ing to understand causal processes.
Experimental variable—A variable whose values are independent of changes in the
values of other variables (see independent variable).
External validity—The ability to generalize from one procedure or set of subjects
to other conditions or populations.
Filial imprinting—A form of learning in which a very young animal fixes its atten-
tion on the first object with which it has visual, auditory, or tactile experience and
thereafter follows that object.
Fitness—Ability to survive and reproduce.
GLOSSARY 421

Fixed effects—Effects of a treatment that are assumed to be the same (fixed) in dif-
ferent studies and measured without error.
Focal animal sampling— Observations are limited to specific (focal) individuals (or
groups) for a specified period.
Follicle-stimulating hormone (FSH)—A hormone synthesized and secreted by the
anterior pituitary gland that regulates the development, growth, pubertal matura-
tion, and reproductive processes of the body.
Food chain—A linear sequence of links in a food web starting from primary pro-
ducers (e.g., plants) and ending at a species that is eaten by no other species in the
web.
Forager—A gene identified in Drosophila that encodes a cGMP-dependent protein
kinase (PKG) that affects food-searching behavior.
Forebrain—The most anterior of the three primary regions of the vertebrate brain
(telencephalon and diencephalon) that includes the cerebral hemispheres, the thala-
mus, and the hypothalamus; in higher vertebrates it is the main control center for
sensory and associative information processing, visceral functions, and voluntary
motor functions.
Free-running rhythm—A drift in activity rhythms under constant conditions such
that for a particular individual the activity cycle would start either progressively ear-
lier each day or progressively later.
Frequency—The number of occurrences of an event or the number per unit time
(rate).
Function—The way in which a behavior improves the survival and reproductive
success of the individual performing it.
Functional magnetic resonance imaging (fMRI)—An imaging procedure that mea-
sures brain activity by detecting associated changes in blood flow.
Functional relationship—A particular mathematical relationship between a depen-
dent (Y) variable and one or more independent (X) variables (Y is a function of X).
Game theory—A mathematical method for calculating success based on the choices
of others.
Genetic architecture—The underlying genetic basis of a phenotypic trait.
Genetic assimilation—The incorporation of environmentally induced changes in
phenotype into the genome.
Genomic imprinting—Only the maternal or the paternal genes are expressed.
Genotype by environment interactions (G×E)—The phenotypic effect that results
when the same environmental influences produce different responses from different
genotypes.
Germline-dependent epigenetic modifications—Changes in the epigenome that
have been incorporated into the germline.
422 GLOSSARY

Glial cell—A nonneuronal cell that maintains homeostasis, forms myelin, and pro-
vides support and protection for neurons in the brain and other parts of the nervous
system such as in the autonomic nervous system.
Glucocorticoid—A steroid hormone that regulates the metabolism of glucose.
Gonadotropin-releasing hormone (GnRH)—A peptide hormone that stimulates
the release of FSH and LH from the anterior pituitary gland.
Goodness of fit—How well a set of data conforms to an expected distribution of
values.
Gradient—A type of treatment in which values vary continuously over a range of
possible values.
Grasp reflex—Clenching of the fingers or toes on stimulation of the palm or sole;
normal only in infancy.
Group selection—A mechanism of evolution that relies on presumed advantages to
a group that are brought about by the differential extinction and proliferation of
groups.
Guided transmission—An increase in fidelity of copying of the behavior of a model
when the model provides explicit pedagogical cues.
Habitat—The area or environment where an organism or ecological community
normally occurs.
Habituation—The decline of a behavior with the repeated presentation of an elicit-
ing stimulus.
Hatching asynchrony—A pattern in which eggs in a brood do not hatch together.
Hawk strategy—In game theory, a strategy of always fighting (escalating).
Heritability (heritable variation)—The proportion of total variance in a trait that
can be attributed to additive genetic variance in that trait.
Heterospecific (interspecific) brood parasitism—A reproductive system in which a
species of animal (the parasite) uses a different species (the host) to raise its young.
Hindbrain—The most posterior of the three primary regions of the vertebrate brain
(rhombencephalon) that includes the cerebellum and medulla oblongata; controls
autonomic functions and equilibrium.
Homeorhesis—The stability of the process of development.
Homeostasis—The stability of a final steady state.
Hormone—A chemical released by a cell or a gland in one part of the body that
sends out messages that affect cells in other parts of the organism.
Huddle—A densely packed group of individuals.
Hypothalamic-pituitary-adrenal (HPA) axis—A complex set of direct influences
and feedback interactions among the hypothalamus, the pituitary gland, and the
adrenal glands.
GLOSSARY 423

Hypothalamic-pituitary-gonadal (HPG) axis—A complex set of direct influences

and feedback interactions among the hypothalamus, the pituitary gland, and the
gonads.
Hypothalamus—A small portion of the brain that links the nervous and endocrine
systems via the pituitary gland.
Hypothesis testing—The process by which proposed explanations are examined by
testing their unique predictions.
Immediate early gene—A gene that is activated transiently and rapidly in response
to a wide variety of cellular stimuli.
Imprinting—Learning that occurs at a particular age or a particular life stage that is
rapid and apparently independent of the consequences of behavior (e.g., filial
imprinting, sexual imprinting).
Inclination—The angle between –90° (up) to 90° (down) made by a compass nee-
dle with the horizontal at any point on the Earth's surface.
Independent—One event whose occurrence makes it neither more nor less probable
that another occurs.
Independent variable—The variable being manipulated or changed to investigate
its effect on the dependent variable.
Inertial navigation (dead reckoning)—Use of known or estimated speeds and
direction to advance from the current position.
Innate—Present from birth.
Innate releasing mechanism—The mechanism triggered by the releaser (sign
stimulus) that produces a stereotyped action of an animal.
In situ hybridization—A method of visualizing the amount of mRNA present in a
brain region using a radioactive probe.
Interacting phenotypes—Traits whose expression is affected by interactions with
members of the same species.
Interaction—A lack of independence between variables that makes it more difficult
to predict the effects of those variables on a dependent variable.
Internal validity—The ability to identify a causal relationship.
Interval—The length of time between events. A type of variable in which values
denote the numeric difference between two points but lack a nonarbitrary zero
point.
Intraspecific brood parasitism—A reproductive system in which an animal (the
parasite) uses another member of the same species (the host) to raise its young.
Introspection—The self-observation and reporting of conscious inner thoughts,
desires, and sensations.
Isogenic—Genetically identical.
424 GLOSSARY

Kinesis—Orientation by altering speed or turning rate depending on the strength of

a cue.
Kin selection—A mechanism that proposes that because close relatives share signifi-
cant proportions of their genes they can, in effect, pass along traits to subsequent
generations, even without reproducing themselves.
Knockout—The inactivation of a specific gene. An organism in which a specific
gene has been inactivated.
Kymograph—An instrument for recording variations in pressure (e.g., blood pres-
sure), or tension (e.g., muscle tension) using a pen or stylus to mark a rotating
drum.
Lagena—Part of the vestibular system of the inner ear.
Latency—A measure of time delay.
Latitude—A geographic coordinate that specifies north-south position on the earth.
Lek—An assembly area where animals display and court potential mates.
Lesion—An experimental method in which a brain region is naturally or intention-
ally destroyed to observe any resulting changes such as degraded or enhanced perfor-
mance on some behavior.
Life zones—A means of describing areas with similar plant and animal communities.
Lipophilic—Having an affinity for, and tending to combine with or dissolve in, lip-
ids and therefore being able to pass through cellular membranes with relative ease.
Longitude—A geographic coordinate that specifies east-west position on the earth.
Luciferase—A class of oxidative enzymes used in bioluminescence.
Luteinizing hormone (LH)—A glycoprotein hormone produced by the anterior
pituitary gland.
Maladaptive—A trait that reduces survival and reproductive success.
Map sense—A mechanism that enables an individual to determine its location.
Masking—Direct effects of light on behavior overriding an animal’s internal clock
and at other times reinforcing the drive produced by that clock.
Mass action—The cerebral cortex acting as a single entity for many types of
learning.
Matching (blocking)—A matched subject design that uses separate experimental
groups for each treatment but relies upon matching every subject in one group with
one and only one equivalent in another.
Maternal entrainment—A process by which an offspring’s biological rhythm is
synchronized by interaction with the mother.
Maternal imprint—Expression of genes inherited from the mother.
Melanopsin ganglion cell—A type of nerve cell in the retina of the mammalian eye.
GLOSSARY 425

Melatonin—A naturally occurring compound (N-acetyl-5-methoxytryptamine) that

varies in a daily cycle, thereby allowing the entrainment of circadian rhythms.
Meta-analysis—A statistical method that combines results of several studies that
address related research hypotheses.
Metamorphosis—Conspicuous and dramatic developmental changes in body struc-
ture and behavior based on cellular growth and differentiation.
Method of exclusion—An approach to testing for social learning that depends upon
exclusion of alternative explanations of interpopulation differences in behavior.
Microarray analysis—A technique used to measure the amount of expression of
large numbers of genes simultaneously, or to genotype multiple regions of a genome,
that uses a collection of microscopic DNA spots attached to a solid surface.
Microsatellite parentage analysis—A technique used to identify parents based on a
specific sequence of DNA nucleotides made up of short, repeated sequences.
Midbrain—The middle of the three primary regions of the vertebrate brain (mesen-
cephalon) that includes the tectum, tegmentum, and substantia nigra; involved in
functions such as vision, hearing, eye movement, and body movement.
Mixed ESS—An equilibrium mix of individuals each playing a pure strategy, or an
individual playing different strategies at equilibrium probabilities.
Mobbing—An antipredator behavior in which prey individuals collectively attack or
harass a predator.
Mobbing call—A vocalization that elicits mobbing behavior or recruits individuals
to mob a predator.
Model assumptions—Conditions that are assumed to exist when constructing a
theoretical model.
Modeling—Developing physical, conceptual, or computer-based representations to
study real systems.
Molar epigenetics—A focus on epigenetic questions of evolution and adaptive
significance.
Molecular epigenetics—A focus on molecular levels of epigenetic analysis.
Monoallelic expression—Only one allele of a gene being actively transcribed.
Monogamy—A form of social behavior in which an individual has only one mate at
any one time.
Monte Carlo model—A method of predicting results that relies on repeated random
sampling.
Morgan’s canon—The belief that animal behavior should be explained as simply as
possible.
Motor nervous system—The part of the nervous system that controls muscles and
therefore movement and behavior.
426 GLOSSARY

Multiple working hypotheses—Several proposed explanations, each of which makes

at least one unique prediction, for a phenomenon of interest.
Mutagenesis—The process by which the genetic information of an organism is
changed (mutated) in a stable manner.
Natural experiment—An observational study in which the design of treatments has
been made “by nature” rather than by experimenters.
Natural selection—The nonrandom process by which traits become either more or
less common in a population as a function of differential survival or reproduction of
their bearers.
Nearest-neighbor solution—A solution based on the rule of choosing the location
closest to the current location.
Necessary—A condition of a statement that must be satisfied for the statement to
be true.
Negative feedback control—A control system in which the production of a product
reduces its production.
Neuroethology—The evolutionary and comparative approach to the study of animal
behavior and its underlying mechanistic control by the nervous system.
Neuromodulator—A substance that alters nerve impulse transmission.
Neuron—A nerve cell.
Neurotransmitter—Endogenous chemicals that transmit signals from a neuron to a
target cell across a synapse.
Niche—The particular set of biological and physical conditions to which the species
has adapted in order to survive and reproduce.
Nocturnality—The behavior of animals that are active in the nighttime.
Nominal—A type of variable comprised of categories (names) rather than
quantities.
Nonadditive effects—The quantitative change in a trait caused by interactions
among alleles both within and across genetic loci.
Nonapeptide—A peptide made of nine amino acid subunits.
Obligate brood parasite—A species that has lost the ability to raise its own young
and relies on a host species to care for its offspring.
Observational study—A type of research in which no manipulation or control is
attempted.
Ontogenetic adaptation—Development as a series of adaptations to a sequence of
environments.
Ontogenetic niche—An ecological role that is specific for a size category or life cycle
stage.
Ontogeny—The developmental history of an organism (or trait such as behavior)
within one lifetime.
GLOSSARY 427

Operant chamber—A laboratory apparatus (also called a Skinner box) used in

experimental psychology to study animal behavior.
Operant conditioning—Learning in which an individual modifies the occurrence
and form of its own behavior in response to the association of the behavior with a
stimulus.
Optimality theory—The view that behavior has evolved to maximize net fitness
benefit (benefit minus cost).
Optimal model—A method of predicting results in which the best of a set of pos-
sible solutions is chosen.
Ordinal—A type of variable in which values are ranked but do not denote the
numeric difference between points (e.g., first, second, third).
Organic selection—The development of variable phenotypes not based on genetics.
Ovariectomy—Removal of the ovaries.
Pacemaker—An organism’s internal clock mechanism.
Pair bond—An enduring preferential association between two sexually mature adults
that is characterized by selective affiliation, contact, and copulation of the partners.
Parental genomic imprinting—Genes being expressed in a parent-of-origin
fashion.
Parturition—The act of giving birth.
Paternal care—Parental care provided by the father.
Paternal imprint—Expression of genes inherited from the father.
Peer review—The practice of using peers (other experts) to evaluate research pro-
posals and results.
Peptide hormone—A peptide that is secreted into the bloodstream and has endo-
crine functions in animals.
Period—The time required to complete one cycle (cycle length).
Period (per)—A gene that encodes a protein (PER) whose oscillations in transcrip-
tion play a central role in the molecular mechanism of Drosophila circadian rhythm.
Petroglyph—A recognizable form or symbol created by chipping, pecking, or scrap-
ing a rock surface.
Phenotypic plasticity—The ability of a phenotype to vary as a result of environ-
mental influences on its genetic makeup.
Pheromone—A secreted or excreted substance that elicits a social response in mem-
bers of the same species.
Photic entrainment—A process by which a biological rhythm is synchronized by
an environment light cue.
Photoperiodism—A form of exogenous seasonal rhythmicity that uses day length to
time biological rhythms.
428 GLOSSARY

Phototaxis—Movement in response to light.

Physiological ecology—The study of biophysical, biochemical, and physiological
processes used by animals to cope with factors of their physical environment or
employed during ecological interactions with other organisms.
Pictograph—Picture painted onto rocks, cave walls, or other generally protected
surfaces.
Piloting—Local-area navigation that relies on familiar landmarks or memorized
routes.
Pituitary gland—An endocrine gland with two components, the anterior pituitary
and the posterior pituitary, that is functionally linked to the hypothalamus.
Plasticity—The capacity of animals with the same genotype to vary in developmen-
tal pattern or behavioral phenotype according to varying environmental conditions.
Pleiotropic effects—One gene affecting more than one phenotypic trait.
Polarized—The orientation of oscillations in the plane perpendicular to a transverse
wave’s direction of travel.
Pollen-hoarding syndrome—A suite of behaviors in honeybees that includes storing
large quantities of pollen (pollen hoarding) in the hive.
Polygyny—A form of social behavior in which males have more than one mate;
males mate multiply, resulting in stronger sexual selection on males.
Polymorphic repetitive microsatellite sequences—Short, repeating sequences of
DNA that vary in the number of times the short sequence is repeated.
Polyphenisms—Multiple phenotypes from a single genotype.
Positron emission tomography (PET)—An imaging procedure that detects pairs of
gamma rays emitted indirectly by a positron-emitting tracer that is introduced into
the body on a biologically active molecule.
Posterior pituitary—The nonglandular collection of axonal projections from the
hypothalamus that terminate in the posterior portion of the pituitary gland and
release peptide hormones.
Power—The probability that we will correctly reject a statistical null hypothesis (1 – ).
Precision—The ability of a measurement to be consistently reproduced.
Precocial— Relatively mature and mobile at birth or hatching.
Prediction—A statement of the expected results of a study (observation, experi-
ment, modeling) that is based on a particular hypothesis (a prediction of the
hypothesis).
Preformationism—The belief that an organism is fully formed at conception.
Progeny—The immediate descendant or descendants of an individual.
Promoter—A region of DNA that facilitates the transcription of a particular gene.
GLOSSARY 429

Proximate causation—The reason a behavior is performed, in terms of its immedi-

ate trigger or mechanisms that control its performance.
Pseudorandom—A sequence of deterministic numbers that approximates the prop-
erties of random numbers.
Pseudoreplication—Improper use of measurements as independent replicates.
Pubarche—The onset of puberty.
Quantitative genetics—The study of continuously varying traits and their underly-
ing mechanisms.
Quantitative trait loci (QTL)—Segments of DNA containing or linked to the
genes that underlie a quantitative (continuously varying) trait.
Quasi-experiment—An experimental study in which not all confounding variables
are controlled.
Random effects—Effects that are assumed to represent a random sample of all pos-
sible values in a population.
Randomize—To assign treatments or gather samples without any recognizable pat-
terns or regularities.
Randomized block design—Arranging experimental units in groups (blocks) that
are similar to one another; the blocking factor is a source of variability that is not of
primary interest to the experimenter.
Random variation—Variability caused by chance factors so that its value cannot be
predicted.
Ratcheting—A mechanism by which knowledge accumulates incrementally across
cultural generations.
Ratio—A type of variable with set numeric intervals and a nonarbitrary zero value.
Reaction norm—The pattern of phenotypic expression of a single genotype across a
range of environments.
Receptor—A cell-surface molecule that binds with chemical signals originating out-
side the cell.
Reciprocal altruism—A behavior whereby an organism acts in a manner that tem-
porarily reduces its fitness while increasing another organism’s fitness, with the
expectation that the other organism will act in a similar manner at a later time.
Reductionists—Those who insist that all behavior is based on simple stimuli and
responses.
Regression—A statistical method that attempts to identify a functional relationship
between a dependent variable and one or more independent variables.
Rejection—A conclusion that the hypothesis being tested is incorrect.
Releaser (sign stimulus)—The stimulus that releases the stereotyped actions of
animals.
430 GLOSSARY

Repeated measures design—Use of the same subjects with every condition of the
research, including the control.
Replication—Using more than one subject per observation or experimental treat-
ment group.
Reproductive success—The relative production of fertile offspring by a genotype, or
the number of surviving offspring produced by an individual.
Retaliator strategy—In game theory, a strategy that displays against the dove strat-
egy but attacks (retaliates against) the hawk strategy.
Retinohypothalamic tract—A photic input pathway involved in the circadian
rhythms of mammals.
Rooting reflex—Head-turning and sucking movements elicited in a normal infant
by gently stroking the side of the mouth or cheek.
Rule-based solution—Use of one or more simple rules to reach a solution.
Rule of one variable—An experimental design in which the control and experimen-
tal treatments are identical except for the experimental manipulation, so that any
effect must be caused by the manipulation.
Rutting area—Location where male ruminant animals rub their antlers or horns on
trees or shrubs, fight with each other, wallow in mud or dust, and herd receptive
females together.
Scan sampling—A group of individuals being scanned at specified intervals and the
behavior of each individual recorded at that instant.
Seasonal affective disorder—A mood disorder that causes depression-like symptoms
during one season (most commonly winter) of each year.
Seasonal rhythm—Changes in activity that correspond with the seasons.
Sensitive period—A period during development when learning occurs.
Sensory nervous system—The part of the nervous system responsible for processing
information from sense organs.
Sexual selection—Differential reproductive success caused by competition for mates.
Signal—A trait (structure or behavior) that has evolved specifically because it
changes the behavior of a receiver in ways that benefit the signaler.
Sign stimulus (releaser)—The stimulus that releases the stereotyped actions of
animals.
Silencing—The switching off of a gene by a mechanism in the cell.
Simulation—A representation of key characteristics or behaviors of a selected
system, often requiring a computer (or network of computers) to run.
Social behavior—The interactions between two or more individual animals, usually
of the same species.
Social isolation—Lack of contact with other individuals of a social species.
GLOSSARY 431

Sociality—The tendency of individuals to develop social links and live in groups.

Social learning—Facilitation of the acquisition of adaptive patterns of behavior by
observing the behavior of others.
Socially distributed behavior—Behavior involving multiple interacting individuals.
Social network analysis—The study of social relationships in terms of individual
members of a social group and the connections or relationships among them.
Sociobiology—The study of the social systems of animals in relation to their ecology
and biology.
Somatic adaptation—The effect on the fitness of cells as a result of the accumula-
tion of mutations in those the cells during a lifetime.
Specific action potential—A proposed motivation energy that accumulates over
time if a specific behavior is not performed.
Stealer—In game theory, a strategy of defending an important resource such as a
bower and stealing decorations from other bowers.
Steroid hormone—A hormone with a steroid chemical structure that helps to con-
trol metabolism, inflammation, immune functions, salt and water balance, develop-
ment of sexual characteristics, and the ability to withstand illness and injury.
Steroidogenic glands—Endocrine glands that produce steroid hormones.
Strategy—In game theory, a fixed and predictable way of behaving in a contest. A
proposed way of behaving that is thought to affect fitness (does not assume con-
scious decision making).
Strong inference—A model of scientific inquiry that emphasizes the need for alter-
native hypotheses, rather than a single hypothesis, to avoid confirmation bias.
Sufficient—A condition that, if satisfied, ensures a statement’s truth.
Suprachiasmatic nucleus (SCN)—A tiny region in the hypothalamus at the base of
the brain.
Synapse—A structure that permits a nerve cell to pass an electrical or chemical
signal to another cell.
Taxis—Orientation directly to a cue.
Teaching—Behavior that facilitates learning of a naïve individual at some cost to
the teaching individual.
Temporal niche—An animal’s daily and seasonal rhythms.
Temporal partitioning—A shift in the activity pattern of a species in response to
intra- and interspecies competition.
Termite fishing—A method of foraging for termites in which chimpanzees push
sticks into nests.
Territoriality—A behavior pattern in animals consisting of the occupation and
defense of a territory.
432 GLOSSARY

Territory—A fixed area containing specific resources such as food, nest sites, or
mates from which one or more individuals exclude other members of the same
species.
Test of independence—A statistical method to determine if two categorical variables
are independent of each other.
Theriomancy—The use of animals in divination.
Timeless—A gene in Drosophila that encodes an essential protein (TIM) that regu-
lates circadian rhythms.
Tousling—When moorhen parents grasping a chick’s head, picking it up, and shak-
ing it vigorously.
Transcription—The process by which information on DNA is used to form mes-
senger RNA (mRNA).
Transcription factor—A protein that controls the flow of information from DNA
to mRNA.
Transcriptome—The set of all RNA molecules, including mRNA, rRNA, tRNA,
and other noncoding RNA produced in one or a population of cells.
Transgene coding—A gene or genetic material that has been transferred from one
organism to another.
Transgenic mice—Mice containing additional, artificially introduced genetic
material in every cell.
Translation—The process by which information on messenger RNA is used to
form protein.
Transport reflex—A young animal that is incapable of sustained locomotion adopt-
ing a posture that facilitates the mother’s ability to transport it.
Treatment effect—The amount of change caused by membership in a treatment
group compared to a control group.
Tropism—Growth or turning movement of a biological organism.
True navigation—Ability to use current global position to advance to a destination.
Two-action method—A procedure to test for social learning in which a demonstra-
tor individual performs a task in only one of two (or more) distinct ways.
Two-factor designs—An experimental design to assess effects of two treatments
(factors).
Type I error—The probability of incorrectly rejecting a statistical null hypothesis
(α).
Type II error—The probability of incorrectly accepting a statistical null hypothesis
().
Ultimate causation—The reason a particular behavior evolved, in terms of its effect
on reproductive success.
Umwelt—The biological foundations of an animal’s world.
GLOSSARY 433

Uniformitarianism—The assumption that the same natural laws and processes that
operate now have always operated in the past and apply everywhere.
Validity—The extent to which a measurement or method is accurate or correct.
Variable—A feature or factor that takes on different values.
Vector navigation—Migration paths that involve two or more separate legs of com-
pass orientation.
Verifiable—Ability to obtain identical or nearly identical results using the same
methods.
Waggle dance—A figure-eight dance of the honeybee that is thought to provide
information about the direction and distance to food, water, or potential hive
locations.
Wild type—The phenotype of the typical form of a species as it occurs in nature.
Working hypothesis—A provisional explanation of a biological phenomenon that
makes testable predictions.
Zeitgeber—An external cue that synchronizes an organism's endogenous clock to
the Earth’s 24-hour light/dark cycle.
This page intentionally left blank
 About the Editors and Contributors

EDITORS
KEN YASUKAWA is Mead Family Professor of Biology at Beloit College and
Honorary Fellow in Zoology at the University of Wisconsin at Madison.
Using Niko Tinbergen as his model, he has studied the behavior and ecology
of birds and has focused primarily on the red-winged blackbird since 1973.

ZULEYMA TANG-MARTÍNEZ is Founders Professor of Biology at the

University of Missouri–St. Louis. She is interested in animal communication
and focuses on chemical communication.

CONTRIBUTORS
JEFFREY R. ALBERTS is Professor of Psychological and Brain Sciences at
Indiana University in Bloomington. His research focuses on the ontogeny of
species-typical behavior, maturation of sensory and physiological processes,
and the development of learning and memory.

SARAH JANE ALGER is an Honorary Fellow in the Department of Psychol-

ogy at the University of Wisconsin, Madison. Her research addresses questions
about how hormones and the brain regulate vocal communication and sexual
behavior in songbirds and rodents.
436 ABOUT THE EDITORS AND CONTRIBUTORS

BRONWYN H. BLEAKLEY is an Assistant Professor of Biology at Stonehill

College in Easton, Massachusetts. Her research integrates evolutionary biol-
ogy, genetics, and behavioral ecology to understand the role the social environ-
ment plays in the evolution of complex social behavior, such as cooperation,
social learning, and cannibalism.

SEAN P. BRADLEY is a doctoral student and fellow in the Institute for Mind
and Biology at the University of Chicago. His research is focused on describ-
ing interactions between circadian rhythms, physiology, and behavior, with
special emphasis on the role of food in the entrainment of biological rhythms.

DAVID CREWS is Ashbel Smith Professor in the Section of Integrative Biol-

ogy and Department of Psychology at the University of Texas. He is interested
in the diversity in reproductive controlling mechanisms and the evolution of
brain mechanisms controlling behavior. His research is conducted in both
the laboratory and the field because the causal mechanisms and functional
outcomes of reproductive processes are revealed at each level while illuminat-
ing the relations among the levels.

ANNE DANIELSON-FRANÇOIS is an Assistant Professor of Biology in the

Department of Natural Sciences at the University of Michigan-Dearborn in
Dearborn, Michigan. Her research bridges evolutionary biology and animal
behavior to explore how variation in mating behavior evolves and is main-
tained in natural populations.

LEE C. DRICKAMER is Emeritus Professor of Biological Science at

Northern Arizona University. He has studied social behavior and behavior
development in a variety of animals. He is the author of a major textbook in
animal behavior and has served both as the President of the Animal Behavior
Society and as Secretary-General of the International Council of Ethologists.

MATTHEW J. FUXJAGER is Assistant Professor in the Department of

Biology at Wake Forest University. His research focuses on endocrine mecha-
nisms of complex social behavior and how selection interacts with these mech-
anisms to drive behavioral diversification.

BENNETT G. GALEF is Emeritus Professor in the Department of Psychol-

ogy, Neuroscience and Behaviour at McMaster University in Hamilton,
Ontario. He and his students have spent more than 40 years studying the role
of social information in the development of adaptive patterns of behavior from
ABOUT THE EDITORS AND CONTRIBUTORS 437

food choice to mate choice in animals as diverse as Norway rats and Japanese
quail.

BRETT M. GIBSON is Associate Professor of Psychology at the University of

New Hampshire. He studies spatial cognition in humans and nonhuman animals.

JAMES L. GOULD is a professor in the Department of Ecology and Evolu-

tionary Biology at Princeton University. He has worked on many aspects of
orientation, navigation, communication, mate choice, learning, and cognition
in a variety of insects, fish, and birds.

CHRISTOPHER HARSHAW is a Postdoctoral Researcher in the Depart-

ment of Psychological and Brain Sciences at Indiana University in Blooming-
ton. His research interests are in the development of social behavior,
comparative cognition, and developmental psychopathology.

MEGAN HASTINGS HAGENAUER is a research fellow in the Molecular

Behavioral Neuroscience Institute at the University of Michigan. She
researches biological timekeeping in relationship to reproductive hormones,
teenage sleep patterns, temporal niche, and mood disorder.

ANDREW P. KING is a Senior Scientist in the Department of Psychological

and Brain Sciences at Indiana University, a founding member of the Center
for the Integrative Study of Animal Behavior, and co-director of the Animal
Behavior Farm at Indiana University.

CATHERINE A. MARLER is Professor of Psychology at the University of

Wisconsin, Madison. Her research focuses on plasticity in social behavior
and the underlying neuroendocrine mechanisms.

LAURA SMALE is Professor of Psychology and Zoology at Michigan State

University in East Lansing. She is interested in the neural mechanisms con-
trolling 24-hour (circadian) rhythms in diurnal animals and how these mech-
anisms are different in animals that are active during nighttime hours.

MEREDITH J. WEST is a Chancellor’s Professor in the Department of

Psychological and Brain Sciences at Indiana University, a founding member
of the Center for the Integrative Study of Animal Behavior, and co-director
of the Animal Behavior Farm at IU.

DAVID J. WHITE is an Assistant Professor in the Department of Psychology

at Wilfrid Laurier University.
This page intentionally left blank
 About the Editorial Board

Daniel T. Blumstein, Department of Ecology and Evolutionary Biology,

University of California–Los Angeles, Los Angeles, California
Dorothy L. Cheney, Department of Biology, University of Pennsylvania,
Philadelphia, Pennsylvania
David L. Clark, Department of Biology, Alma College, Alma, Michigan
Ellen S. Davis, Department of Biological Sciences, University of Wisconsin–
Whitewater, Whitewater, Wisconsin
Bennett G. Galef, Jr., Department of Psychology, Neuroscience, and Behav-
iour, McMaster University, Hamilton, Ontario, Canada
James C. Ha, Department of Psychology, University of Washington, Seattle,
Washington
Kim L. Hoke, Department of Biology, Colorado State University, Ft. Collins,
Colorado
Robert L. Jeanne, Department of Entomology, University of Wisconsin–
Madison, Madison, Wisconsin
Patricia B. McConnell, Department of Zoology, University of Wisconsin–
Madison, Madison, Wisconsin
Douglas W. Mock, Department of Biology, University of Oklahoma,
Norman, Oklahoma
440 ABOUT THE EDITORIAL BOARD

Michael Noonan, Department of Biology, Canisius College, Buffalo, New

York
Sarah R. Partan, School of Cognitive Science, Hampshire College, Amherst,
Massachusetts
Robert M. Seyfarth, Department of Psychology, University of Pennsylvania,
Philadelphia, Pennsylvania
 Index

Abnormal psychology, 53 nonapeptide hormones, 245–47; steroid

ABS (Animal Behavior Society), 9–10, 57, hormones and, 229–33; steroids and
59, 61, 64 female aggression, 232–33; testosterone
ABS Handbook, 9–10 and, 41, 230–31
Accuracy, 18 AH (Anterior hypothalamus), 239, 274, 275
ACTH (Adrenocorticotropic hormone), 220 “Aims and Methods of Ethology”
Action potentials, 187, 203–4 (Tinbergen), 55, 290
Activity levels, 169 Alarm call, 66–67
Adaptation: biological clocks, 369–72; Allee, Warder Clyde, 57, 58
definition of, 292; disappearance and Alloparental care, 243, 337
reappearance of, 317; of early life, 304– Alternate mating strategies, 230
10; Lamarck and, 45; use of term, 293–94 Alternative mating strategy gene, 161
Adaptive behavior, 329 Altmann, Jeanne, 63
Adaptive specializations for early feeding, Altricial offspring, 240–41
310–13 Altricial species, 300, 304, 306
Additive effects, 161–62 Altruism, 66
Additive genetic variation, 168 American crows, 123
Adrenal glands, 218 The American Journal of Psychology, 51
Adrenarche, 265 American Mammals (Stone and Cram), 57
Adrenocorticotropic hormone (ACTH), 220 American Museum of Natural History, 58
African black coucal, 233 American Psychological Society, 61
African gray parrots, 131 American Society of Zoologists, 57
Africanized honey bees, 164–65 American zoology and physiology, 56–60
Agent-based simulations, 332 Amphibians, development by
Aggressive behavior: androgenic control of, metamorphosis, 297–98
230–31; estrogenic control of, 231–32; Amplexus, 222, 297–98
442 INDEX

Anatomy, 39 Aschoff, Jurgen, 357–59

Androgen receptors (AR), 237 Asia, early civilizations in, 36–38
Androgens: aggressive behavior, 230–31; Association for the Study of Animal
behavioral trade-offs, 235–36; hormones Behaviour, 9, 61, 64
included in, 218; immunity, 236–37; Association for the Study of Comparative
mortality, 234–35 Psychology, 51
Androstenedione, 218, 221 Associative learning, 335–36
Anecdotal examples of imitation, 132 Assortative mating, 331
Animal Aggregations (Allee), 58 Astrology, 38
Animal behavior, discipline of, 33 Audubon, John J., 50
Animal behaviorists: experimental studies, Australian brush turkey, 316
18–20; hypotheses testing, 4, 5–6; Avians: development of, 302–4; feeding, early
introduction to, 1–2; observational adaptations in, 311–12; ontogenetic
studies, 10–18; training, 3–4; use of niches, 302–4
models, 20–24, 74–75 AVP (Arginine vasopressin), 158, 187,
Animal Behavior Society (ABS), 9–10, 57, 195–97, 206, 220, 241–46
59, 61, 64 AVP antagonist, 245–46
Animal Behaviour (journal): animal treatment
guidelines, 9–10; on domestic animals, Baerends, Gerard, 55
39; early issues of, 44; funding sources, Bait avoidance, 117–18
2–3; shifts in focus, 64, 65 Balance, 11
Animal cognition, 70–71 Baldwin, James M., 50
Animal Communities in Temperate America Baldwin effect, 50
(Shelford), 57 Banks, Edwin M., 58
Animal compasses, 390–95 Barlow, George, 62
Animal culture debate, 139, 140–42 Bartram, John, 50
Animal husbandry, 36 Bartram, William, 50
Animal Intelligence (Romanes), 49 BAT (Brown adipose tissue), 314
Animal learning psychology, 330 Bateson, Patrick, 3, 62–63
The Animal Mind—A Textbook of Bats, 121, 223, 299–300
Comparative Psychology (Washburn), 53 Beach, Frank A., Jr., 54, 59, 68, 267
Animal models, 53 Bed nucleus of the stria terminalis (BST),
Animal psychology, 50 239–40, 242, 244–45
Animal Studies (Jordan, Kellogg, Heath), 57 Begging behaviors, 311
Ant dipping, 137, 138 Behavioral development. See Ontogenetic
Anterior hypothalamus (AH), 239, 274, 275 adaptation and behavior development
Anthony, Harold E., 57 Behavioral ecologists, 168
Anthropomorphism, 9, 37 Behavioral ecology, 233, 330
Ape culture, 136 Behavioral ecology and sociobiology, 65–67
Appetitive behavior, 56 Behavioral Ecology: An Evolutionary
AR (Androgen receptors), 237 Approach, 65–66
Arc configurations, 108 Behavioral endocrinology, 68–71
Arginine vasopressin (AVP), 158, 187, Behavioral epigenetics, 174–76
195–97, 206, 220, 241–46 Behavioral genetics: introduction to, 151–54;
Aristotle, 39, 152 single-gene effects on behavior, 154–59;
Arrhythmia, 361 summary, 176. See also Complex inherit-
Artifacts, 34–35 ance and quantitative traits
INDEX 443

Behavioral Neuroendocrinology Brood parasitism, 116, 302–3

(organization), 68 Brooks, Robert, 168
Behavioral neuroscience, 187, 330 Brown, Jerram, 63, 65
Behavioral syndromes, 230 Brown, Michael, 110
Behavioral trade-offs, and androgens, 235–36 Brown adipose tissue (BAT), 314
Behaviorism, 52 Brown bats, 223, 299–300
Behaviour, 43, 61 Brown-headed cowbirds, 116–17, 193,
Bekoff, Anne, 305 326–27. See also Cowbirds
Belyaev, Dmitri, 166 BST (Bed nucleus of the stria terminalis),
Benzer, Seymour, 155 239–40, 242, 244–45
Bernard, Claude, 47 Buccal pump, 310
Bias, 18, 42 Buffon, George Louis Leclerc Comte de, 44
Bio-Ecology (Clements and Shelford), 57 Burroughs, John, 50
Biogeography, 44 BW (Body weight), 282
Biological clocks: adaptation, 369–72;
difficulty and importance of studying in California mice, 229, 231, 232, 233, 246
the wild, 372–74; introduction to, Canalization, 266
351–52; ontogeny, 374–77; phylogeny, Cane toads, 121
352–56; proximate mechanisms Caregivers, resistance to the acquisition of
(causation), 356–69; seasonal rhythms, aversions to, 315–17
377–80; summary, 380–81. See also Carpenter, C. Ray, 53
Circadian rhythms Carter, Sue, 194
Biparental care, 302 Castro, Lauraeno, 141
Birds. See Avians Cathemeral, 352–54
Bird song, 326–28, 331–34, 335–36. See also Causation, 71–72, 290, 356–69
Countersinging Cave paintings, 35
Birth of mammals, adaptations surrounding, Center-near vs. peripheral-far
307–10 configurations, 105
Blackbirds, 123 Central pattern generators, 187
Blaser, Rachel, 98 Chaffinch, 358
Bleakley, Bronwyn, 173 Challenge hypothesis, 235–36
Blocking, 11 Champagne, Frances, 175
Blue tits, 121–22 Chatter, 340
Boake, Chris, 168 Chickens, 231
Body weight (BW), 282 Chimpanzees: ant dipping, 137, 138; culture
Bois-Reymond, Emil du, 47 in, 140–42; development of traditional
Bonner, John Tyler, 290 behaviors of, 139–40; do-this experi-
Bowerbirds, 24 ments, 131; ecological effects on tradi-
Bowers, 24 tional behaviors, 137–38; genetic effects
Bowlby, John, 315 on traditional behaviors, 138–39; method
Bowman, William, 47 of exclusion, 136, 140; nut cracking, 140;
Brain, 187–89, 239–40 solving TSP-like problems, 80–81;
Breadcrumb sponge isopods, 161 teaching, 127; use of efficient routes, 108
Breeding designs, 168–69 Chiroptera bats, 300
British birds, 133–35 Christman, Gene, 6
British Journal of Animal Behaviour, 43, 61 Chronic restraint stress (CRS), 280–83
Brodie, Butch, 173 Circadian oscillator, 203
444 INDEX

Circadian pacemaker, 362 Context-dependent epigentics modifications,

Circadian rhythms: action potentials, 187, 268, 269–70, 281–84
203–5; changes in cells associated with Contingencies, 335–36
SCN, 205–7; definition of, 359; early Continuously distributed traits. See
study of, 200–202; fruit flies, 155–56; Quantitative traits
light/dark cycle and oscillator, 202–3; Control, 18
maladaptive valley, 200, 208–10; Convection, 314
necessity of, 369–72; overview of, Cook, Robert, 110–11
199–200; in plants, 357; plasticity in, 208; Cooperation, evolution of, 173
QTL analysis and, 169; studying in the Cooperation among Animals, with Human
wild, 372–74. See also Biological clocks Implications (Allee), 58
Circadian timekeeping, 361 Cornell, Heather, 123
Clark, Mertice, 118 Corps of Discovery, 46
Clark, William, 46 Correlation, 20
Classical learning, 335–36 Corsica, 120
Classification, 39, 45 Corticosterone, 219, 221–22, 282
Clements, Frederic E., 57 Corticotropin-releasing hormone
Clock, 156, 169 (CRH), 220
Cluster configurations, 107 Coues, Elliott, 50
CO (Cytochrome oxidase) histochemistry, Countersinging, 333, 336, 339–40
273–74 Counting, 116–17
Cognition, 70–71 Courtship behavior: competence, 340–42;
Cognitive ethology, 62, 70–71 consistency of, 334; learning, 116–17;
Cognitive maps, 81, 99–101 steroid hormones and, 224–27; steroids
Collias, Nicholas, 58 and elaborate displays, 226–27; steroids
Communal breeding grounds, 124–25 and vocalizations, 224–26
Communication, definition of, 67 Cowbirds, 116–17, 303, 331–34, 339–42.
Comparative method, 55 See also Brown-headed cowbirds
Comparative psychology, 33, 52–54, 59 Cows, significance of in India, 37
Compasses, 390–95 Cram, William E., 57
Competence, 340–42 Cramer, Audrey, 81, 99
Completely randomized design, 19 Crepuscular, 204–5
Complex inheritance and quantitative traits: Crepuscularity, 352–54
Drosophila as model of QTL analysis, Crews, David, 223–24
163–64; effects for quantitative traits, CRH (Corticotropin-releasing
174; environment’s genes, 172–74; hormone), 220
epigenetics, 174–76; introduction to, Crop milk, 311–12
159–61; linking QTL for behavior to CRS (Chronic restraint stress), 280–83
evolution using nonmodel systems, Csibra, Gergely, 141, 142–43
164–67; phenotypic plasticity, genotype Cuckoos, 303
and environmental interactions, 169–72; Culture, 130, 136. See also Animal culture
quantitative genetics, 167–69; debate; Human culture; Traditions and
quantitative trait loci, 161–63 culture
Computer simulation, 24 Culture debate, 139, 140–42
Confounding variables, 7, 10–11 Cumulative culture, 126
Conspecifics, 244, 326 Cuticular hydrocarbons, 164
Consummatory behavior, 56 Cuvier, Georges, 46
INDEX 445

Cyanobacteria, 354 Diffusion-chain experiments, 130

Cycle, 156 5α-dihydrotestosterone (DHT), 218, 221
Cycle length, 357 Dinosaurs, 354
Cyclical behavior patterns, 209–10 Direct development, 298
Cytochrome oxidase (CO) histochemistry, Discontinuities, in development, 296
273–74 Displacement activity, 56
Dissociation, 169
Daan, Serge, 359, 360–61 Diurnality, 352–55
Daily rhythms, 354 Diurnal mammals: definition of, 199; early
Dark Ages, 40–41 study of circadian rhythms in behavior,
Dark-eyed juncos, 228, 236 200–202; light/dark cycle and oscillator,
Darwin, Charles, 45, 46, 48–49, 153, 265 202–3; maladaptive valley, 208–10;
Darwin, Erasmus, 45 outwards from the oscillator, 203–5;
Data analysis, for observational responsiveness to signals from SCN,
studies, 13–15 203–5. See also Biological clocks;
Data sheet, example of, 14 Circadian rhythms
Davenport, Charles Benedict, 57 Divination, 38
Davies, Nicholas, 65 DNA methylation, 266, 270, 281
Da Vinci, Leonardo, 42 Dobzhansky, Theodosius, 267
Dawkins, Marian Stamp, 6, 63–64 Dogs, 316
Dawkins, Richard, 62 Dolphins, 299
Dayan, Tamar, 352, 356 Domestication of plants and animals, 35–36
Dead reckoning, 399–401 Dominance, 161–62
De Anima (Aristotle), 152 Dominance hierarchies, 74, 333, 337
DeCoursey, Patricia, 202, 359 Dominance relationships, 189, 331
Deductive reasoning, 41 Dopamine, 187
Defenders, 24 Do-this experiments, 130–31
Degus, 204–5 Doubletime, 156
Dehydroepiandrosterone (DHEA), 218 Doves. See Hawk-dove game
Dependent variables, 18 Drew, Michael, 110
Descartes, René, 41–42, 152 Drickamer, Lee C., 63, 65
The Descent of Man and Selection in Relation Drosophila fruit flies: circadian rhythms,
to Sex (Darwin), 48 155–56; ecological crossover, 171–72;
Design of layout, 19 epistasis, 169; foraging behavior, 156–57;
Design of treatments, 19–20 genetic assimilation, 266; as a model for
Destroyers, 24 QTL analysis, 163–64
Development, behavioral, 289. See also Dugatkin, Lee, 123–24
Ontogenetic adaptation and behavior Duration, 12
development Dusky-footed wood rat, 235–36
Development, Tinbergen on, 290 Dyadic interactions, 339–40
Developmental discontinuities, 296
Developmental habitats, 295 Early civilizations, in Asia and the Middle
Developmental niche, 295, 299 East, 36–38
Dewlap, 233 Early feeding, adaptive specializations,
Dewsbury, Donald A., 63 310–13
DHEA (Dehydroepiandrosterone), 218 Early life, adaptations of, 304–10
DHT (5α-dihydrotestosterone), 218, 221 Eating, learning how to, 117–20
446 INDEX

Ecological crossover, 171–72 Epigenesis, 45, 265

Ecology, development of, 47, 57 Epigenetic effects, 263
Ectoparasites, 237 Epigenetic mechanisms, 248
Ectothermic, 354 Epigenetic modifications, 268–71, 281–84
Ectotherms, 313 Epigenetics: context-dependent vs.
EDCs (Endocrine disruptor chemicals), 269, germline-dependent modifications,
270–71, 279–80 269–71; environmental factors bringing
Efficient traveling behavior, 81–83, 108 changes in, 272–84; in history of animal
Egg eviction, 303 behavior, 72; introduction to, 174–76,
Egg rejection, 303 263–65; manipulation of, 72–73;
Egg tooth, 304–5 molecular vs. molar epigenetics, 265–68;
Ehrman, Lee, 63 as a perspective vs. technique, 271–72;
Eibl-Eibesfeldt, Irenaus, 59–60, 63 summary, 284–85
Elaborate displays, 226–27 Epimutations, 279
Elements of Zoology (Davenport), 57 Epistasis, 161–62, 169
Elephants, 37–38 Equipotentiality, 54
Embryological development, 49 ER (Estrogen receptors), 237, 238
Emerson, Alfred E., 57 ESS (Evolutionarily stable strategy), 22–23,
Emlen, Stephen T., 62 62, 67
Emulation, as social learning, 125, 126 Estradiol, 219, 221, 227–28, 231–33, 241
Endler, John, 168 Estriol, 219
Endocrine disruptor chemicals (EDCs), 269, Estrogen receptors (ER), 237, 238
270–71, 279–80 Estrogens, 218, 219, 231–32
Endocrine glands, 218 Estrone, 219
Endocrine system, 217 Ethograms, 39, 46, 56
Endogenous clock, 356, 358 Ethologists, 89
Endoparasites, 237 Ethology: definition of, 55; discipline of, 33;
Endoplasmic reticulum, 220 emergence of, 48–52; European, 54–56;
Endothermic, 354 in history of animal behavior, 51; and lab
Entrainment, 199, 359 environment, 330; North American vs.
Enumeration methods, 20 European distinctions, 56–57
Environmental factors as agents of epigenetic Ethology, The Biology of Behavior
changes: endocrine disrupting chemicals (Eibl-Eibesfeldt), 60
and epigenetics, 279–80; integration of European blackbirds, 123
germline- and context-dependent European cuckoo, 303
modifications, 281–84; introduction to, European ethology, 54–56
272–74; litter composition, 274–78; European starlings, 80, 232–33
stress, 280–81; temperature and sexual Evo-devo, 267
experience, 272–74 Evolution, 48, 290, 292
Environmental interactions: environment’s Evolutionarily stable strategy (ESS), 22–23,
genes, 172–73; phenotypic plasticity and 62, 67
genotype, 169–72 Evolutionary history, 210, 303
Environmentally induced epigenetic Evolutionary pressure, 296
modifications, 264 The Evolution of Behavior (Brown), 65
Environmental stressors, 281 Exaptation, 310
Epaulets, 4, 6 Exclusion, method of, 136
Epialleles, 269 Exocytosis, 220
INDEX 447

Experimental designs, 19–20 Frequency, 12

Experimental studies, 6, 18–20 Fringe-lipped bats, 121
Experimental variable, 7 Frisch, Karl von, 54, 55
Experiments, vs. observations, 6–8 Frogs, 121, 297, 355
Exposure, timing and persistence of, 270–71 Fruit flies: circadian rhythms, 155–56;
The Expression of the Emotions in Man and ecological crossover, 171–72; epistasis,
Animals (Darwin), 48 169; foraging behavior, 156–57; genetic
Extended medial amygdala, 239 assimilation, 266; as a model for QTL
External validity, 7–8 analysis, 163–64
FSH (Follicle-stimulating hormone)
Faust, Katherine, 338 (copy), 220
Feeding, adaptive specializations for, 310–13 Fujita, Kazuo, 99–100
Female aggression, 232–33 Function, Tinbergen on, 290
Female mammals, 221 Functional magnetic resonance imaging
Fetal positioning, 308 (fMRI), 68
Field Book of North American Mammals Functional relationships, 20
(Anthony), 57 Functions, in animals, 39
Field Guide of Wild Birds and Their Music
(Mathews), 57 Galápagos finches, 135–36
A Field Guide to the Birds (Peterson), 57 Galápagos marine iguanas, 230
Filial imprinting, 315–16 Galdikas, Birute, 132–33
Finches, 243–45 Gallistel, Randy, 81, 99
Fisher, Ronald A., 66, 153 Galton, Francis, 153
Fitness, 233, 292, 328, 331 Galvani, Luigi, 47
FitzGerald, Randall, 191 Game parks, 44
Fixed effects, 19 Game theory, 21, 23, 63, 67
Flying squirrels, 202 Gaulin, Steven, 191
fMRI (Functional magnetic resonance Genetically modified mice, 276–78
imaging), 68 Genetic architecture, 161–63
Focal animal sampling, 13 Genetic assimilation, 266
Follicle-stimulating hormone (FSH), 220 Genetics, 66–67, 72–73
Food, finding, 156–57 Genomic imprinting, 264
Food chains, 42 Genomic sequencing, 162
Forager gene, 156–57 Genotype, and environmental
Foraging behavior: academic focus on, 66; interactions, 169–72
food selection, 119; honey bees, 157; Genotype by environment interactions
learning, 117–19; learning how to eat, (G x E interactions), 170–72, 174–76
120; learning to attack “toxic” prey, Gergely, György, 141, 142–43
120–21; selecting a place to eat, 118; Germline-dependent epigenetics modifica-
social influences, 117–18 tions, 268, 269–70, 281–84
Forced Movements, Tropisms, and Animal Ginchansky, Rachel, 98
Conduct (Loeb), 58 Glial cells, 187
Forebrain, 188–89 Global strategy configurations, 105–7
Forsman, Jukka, 122 Glucocorticoid receptors (GR), 237, 269–70
Four questions of Tinbergen, 8, 290, 352 Glucocorticoids, 218, 219–20
Free-running period, 358–59 GnRH (Gonadotropin-releasing hormone),
Free-running rhythms, 356–57, 358, 359, 361 219–20
448 INDEX

Golden hamsters, 221, 246, 361 Hatching asynchrony, 306

Golden spiny mice, 208, 355–56 Hatching muscle, 304–5
Goldfish, 238 Hawk-dove game, 22–23
Gonadotropin-releasing hormone (GnRH), Heath, Harold, 57
219–20 Hebb, Donald O., 54
Goodness of fit, 20 Hegner, Robert William, 57
Goodson, James, 244 Heinroth, Oskar, 54
Gowaty, Patricia A., 63 Hepper, Peter, 119
GR (Glucocorticoid receptors), 237, 269–70 Hereditary Genius (Galton), 153
Gradients, 20 Heritable fitness, 331
Grant’s gazelle, 237 Heritable variation, 209–10
Grasp reflex, 291, 318 Hertwig, Oscar, 265
Grass rats, 206–7, 209–10 Heterospecific brood
Great tits, 121–22 parasitism, 303
Greek and Roman civilizations, 38–40 Hindbrain, 188
Greenfinches, 236 Hinde, Robert A., 59–60, 63
Gregariousness and nonapeptide hormones, History of animal behavior: American
243–45 zoology and physiology, 56–60;
Grey partridge, 224 behavioral ecology and sociobiology,
Griffin, Donald, 62, 70 65–67; comparative psychology, 33,
Grooming networks, 337 52–54, 59; current state of, 71–75; Dark
Group nesting, 120–21 Ages, 40–41; early civilizations in Asia and
Group regulatory behavior, 314 the Middle East, 36–38; early human
Groups, experimentation on, 331–34 history, 34–36; eighteenth-century
Group selection, 21–22, 66, 67 studies, 44–46; emergence of ethology,
Guided transmission, 141 48–52; epigenetics, 72; European
Guidelines for the Treatment of Animals in ethology, 54–56; Greek and Roman
Research and Teaching, 9 civilizations, 38–40; initial studies of
Gulf toadfish, 225 animal behavior, 41–44; integration
Guppies, 124, 173 of approaches, 60–65; integration of
G x E interactions (Genotype by environment proximate and ultimate causation, 71–72;
interactions), 170–72, 174–76 introduction to, 33–34; manipulation of
genetics and epigenetics, 72–73;
Habitat, 295 modeling, 20–24, 74–75; neuroscience
Habituation, 9 and behavioral endocrinology, 68–71;
Hailman, Jack, 311 nineteenth century studies, 46–52;
Hale, Edgar B., 58 Renaissance, 41–44; rise of natural
Hall, G. Stanley, 51 history, 41–44; strong inference, 73–74
Hamilton, William D., 64 Holland, Paul, 337
Hamilton, William J., 57, 60, 63 Holst, Erich von, 55
Hammock, Elizabeth, 243 Homeorhesis, 267
Hamsters, 41 Homeostasis, 68, 70, 267
Handbook of Physiology (Muller), 47 Honey bees, 128, 157, 164–65
Hansen, Andrew, 6 Hormones: definition of, 218; introduction
Harlow, Harry, 53 to, 217; nonapeptide hormones, 220,
Harvey, William, 42 240–47; overview of, 218–20; steroid
Hatching, adaptations of, 304–6 hormones, 218–20; steroid hormones
INDEX 449

and social behavior, 220–40; summary, Infantile behavior: identification of

247–48. See also Steroid hormones ontogenetics adaptations, 317–19; rooting
Hormones and Behavior (Beach), 54 reflex, 319; transport reflexes, 317–18
House mice, 243 Inferential statistics, 27–31
HPA (Hypothalamic-pituitary-adrenal) axis, Innate behavior, 312
220, 222–23, 280 Innate releasing mechanism, 56
HPG (Hypothalamic-pituitary-gonadal) axis, Inoue-Nakamura, Noriko, 140
219, 222–23 Interacting phenotypes, 172
Hrdy, Sarah, 63 Interactions, 19
Huddle, 314–15 Internal clock, 356, 358
Huddle as habitat, 301, 302 Internal validity, 7–8
Human culture, 141 International Ethological Conferences, 61
Human history, early, 34–36 International Society for Behavioral
Humans: assessment of routes chosen by Ecology, 64
people vs. pigeons, 87–90; determination International Society for Neuroethology, 64
of routes chosen by, 85–87; teaching in Intervals, 12
preindustrial societies, 127; as travelers in Intraspecific brood parasitism, 302–3
2-D space, 83–84 An Introduction to Behavioral Ecology, 65–66
Humle, Tatyana, 138 An Introduction to Comparative Psychology
Hummingbirds, 355 (Morgan), 49
Huxley, T. H., 47 Introspection, 52
Hypothalamic-pituitary-adrenal (HPA) axis, Isogenic lines, 163
220, 222–23, 280 Israel, 120
Hypothalamic-pituitary-gonadal (HPG) axis,
219, 222–23 Jacobs, Lucia, 192–93
Hypothalamus, 188–89, 219 James, William, 51
Hypotheses, rejection of, 5 Janik, Vincent, 139
Hypothesis testing, 4, 5–6 Janson, Charles, 99
Japanese macaques, 133–35
ICHH (Immuno-competence handicap Jordan, David Starr, 57
hypothesis), 236 Journal of Animal Behaviour, 43
Imitation, as social learning, 125, Journal of Insect Behavior, 64
126, 129–33 Jurassic Park, 354
Immediate early gene, 207
Immunity, and androgens, 236–37 Kacelnik, Alex, 80
Immuno-competence handicap hypothesis Kamil, Alan, 6–7
(ICHH), 236 Kawamura, Hiroshi, 205
Immunology, 71 Kellogg, Vernon Lyman, 57
Imprinting, 55, 315–16 Kenward, Ben, 135
Incubation temperature, 273–74, 275 Kinesis, 388–90
Independence, test of, 20 King, Andrew, 116
Independent replication, 10 Kin selection, 66
Independent variables, 18 Kline, Linus W., 50
Indiana University, 332 Knock out function of particular
Indigo bunting, 303 genes, 157–58
Indirect genetic effects, 173 Knockout mice, 158, 276–78
Inertial navigation, 399–401 Kolliker, Albert von, 47
450 INDEX

Konopka, Ronald, 155 MacGregor, James, 97–98

Krebs, John, 65 Mackay, Trudy, 163–64, 169
Kronfeld-Shor, Noga, 352, Magendie, Francois, 47
355–56 Magic coin, 15–18
Kuo, Zing-Yang, 267 Mairan, Jean-Jacques d’Ortous de, 356
Kymograph, 47 Maladaptive behavior, 329
Maladaptive traits, 266
Lab traits, vs. wild traits, 330–31 Maladaptive valley, 200, 208–10
Lactation, 312–13 Male parenting behavior, and
Laland, Kevin, 139, 142 steroids, 228–29
Lamarck, Jean Baptiste de, Malthus, Thomas, 45
44, 45, 265 Mammals: adaptations surrounding birth,
Lashley, Karl, 54, 267 307–10; developmental niches of,
Latency, 12 299–300; development of, 299–302; early
Latent plasticity, 333 ingestion, adaptations for, 312–13
Lateral septum, 239–40, 242, 244 Mammals of the Eastern United States
Leaders in Animal Behavior (Drickamer and (Hamilton), 57
Dewsbury), 63, 65 Mammary glands, 312–13
Learning, 70 Manakins, 226–27
LeConte, John E., 50 Manhattan Project, 23
Leeuwenhoek, Anton von, 43 Manual of the Anatomy of the Invertebrate
Lehner, Philip, 15 Animals (Huxley), 47
Lehrman, Daniel, 59, 69, 267, 312 Maps, as models, 21
Leinhardt, Samuel, 337 Marler, Peter, 60, 63
Leks, 124–25, 226 Marmosets, 223
Lenn, Nicholss, 201 Martin, Paul, 3
Leopard gecko, 272–74, 275 Masking, 199
Leroy, Charles, 45–46 Mass action, 54
Lesions, 201, 360, 362–63 Matching, 11
Lewis, Meriwether, 46 Maternal effects, 174, 175, 176
Lewontin, Richard C., 267 Maternal imprint, 264
LH (Luteinizing hormone), 220 Mathematical modeling, 24
Life stages, 265 Mathews, Ferdinand S., 57
Life zones, 50 Mating behavior, learning, 123–25. See also
Lim, Miranda, 197–98 Courtship behavior
Linnaeus, Carolus, 44, 45 Matsuzawa, Tetsuro, 138, 140
Lions, 299 Maynard Smith, John, 21–23
Lipophilic hormones, 218 Mayr, Ernst, 268
Litter composition, as agent of epigenetic Meadow voles: overview of, 190–91; pair
modification, 274–78 bonds and parental care, 246; partner
Littermates, 313 preferences and pair bonds, 193–99;
Loeb, Jacques, 57–58 space, 191–93
Long-term pair bonding, 158–59 Meaney, Michael, 175
Lonsdorf, Elizabeth, 139 Meerkats, 127–28
Lorenz, Konrad, 54–55, 62, 153–54, 315 Memory, 70
Luteinizing hormone (LH), 220 Menagerie du Parc at Versailles, 44
Lyell, Charles, 48 Mendel, Gregor, 153
INDEX 451

Mendelian genetics, 49 Monogamous mating systems, 158–59, 189,

Mental Evolution in Animals (Romanes), 49 193–94, 196, 198
Menzel, Emil, 80–81, 99 Montane voles, 246
Merian, Maria Sibylla, 43 Monte Carlo (random) model, 85–90,
Merriam, C. Hart, 50 92–93, 95–96, 98, 100, 104
Mesolimbic reward system, 239–40 Moore, Bruce, 131
Mesotocin (MT), 220, 245 Moore, Michael, 223–24
Meta-analysis, 336, 339–40 Moore, Robert, 201, 361
Metamorphosis, 297–98 More Than Kin and Less Than
Method of exclusion, 136, 140 Kind (Mock), 25
Methodological advances in studying animal Morgan, Conway Lloyd, 49
behavior, 330–31 Morgan’s canon, 49
Methods, as theme in writings, 67 Morris water maze, 192
Methylation, 176 Mortality, and androgens, 234–35
Mice: aggressive behavior, 231–32; California Mother as habitat, 301–2
mice, 231, 232, 233, 246; golden spiny Moths, 172
mice, 208, 355–56; house mice, 243; Motivation, 70
knockout mice, 158, 276–78; ontogenetic Motor nervous system, 47
adaptations in regulation and learning, Mountain spiny lizards, 234, 237
313–15; parenting behavior, 158; Mount St. Helens eruption, 8
transgenic mice, 197–98; white-footed MT (Mesotocin), 220, 245
mice, 246; wild type mice, 231–32 Muir, John, 50
Microsatellite parentage analyses, 336 Muller, Hermann, 155
Microsatellite-sequence differences, 243 Muller, Johannes, 47
Midbrain, 188 Multiple working hypotheses, 4
Middle East, early civilizations in, 36–38 Mutagenesis, 155
Migration. See Orientation, navigation, and Mythology, 37
migration
Mills, T. Wesley, 51 N (Nitrogen) cycle, 47
Mixed ESS, 23 NAc (Nucleus accumbens), 239, 242
Miyata, Hiromitsu, 99–100 Nadel, Lynn, 192
Mobbing, 122–23 Naïve individuals, 122–23, 126
Mobbing call, 123 Nash, John, 21
Mock, Douglas, 25 National Institutes of Health (NIH), 61
Model assumptions, 21 National Science Foundation, 61
Modeling, 20–24, 74–75 Natural experiments, 7, 8
Models, as theme in writings, 67 Natural history, 41–44
Molar epigenetics: definition of, 263–64; vs. Natural History (Pliny the Elder), 40
molecular epigenetics, 265–68; overview The Natural History of Selbourne (White), 44
of, 267–68 Naturalists, 42–43, 50
Molecular epigenetics: definition of, 263–64; Natural selection: cyclical behavior patterns,
vs. molar epigenetics, 265–68; overview 209–10; at each stage of development,
of, 266–67 293; evolutionary pressure, 296;
Moles, 299 formulation of, 48; hormones and, 217;
Monceau, Henri-Louis Duhamel du, 356 Malthus contribution, 45; vs. sex-specific
Monkeys, 81, 108 genetic differences, 164; soft skulls
Monoallelic expression, 264 and, 308–10
452 INDEX

Navigation. See Orientation, navigation, and Odor cues, 313

migration Offspring nurturing, single-gene effects on,
Nearest-neighbor model, 85–90, 92–93, 157–59
95–96, 98–99, 100 Ogawa, Sonoko, 276
Nearest-neighbor solution, 82–83 O’Keefe, John, 192
Negative feedback control, 220 Okichoro, 131
Nero, Robert, 7 “On Aims and Methods of Ethology”
Nesting, learning where to, 121–22 (Tinbergen), 55, 290
Neuroethology, 55 One-way/round-trip global
Neuromodulators, 187, 220 configurations, 105
Neurons, 47, 185 On the Origin of Species by Means of Natural
Neuroscience, 68–71 Selection (Darwin), 48
Neurotransmitters, 187 Ontogenetic adaptation and behavior
New Caledonian crow, 135 development: adaptation, definition of,
Newman, Sarah, 239 292; adaptations of early life, 304–10;
Niche, 295, 352 amphibian development by
Niche shift, 306–7 metamorphosis, 297–98; appearance
NIH (National Institutes of Health), 61 and disappearance of infantile behavior,
Nile grass rats, 206–7, 209–10 317–19; avian development, 302–4;
Nitrogen (N) cycle, 47 avoidance of adaptationist trap, 319;
Nobel Prize, 89, 155 being vs. becoming, 289–91; definition
Noble, Gladwyn K., 58 of, 291, 292–93; development, directions
Nocturnal animals, 199. See also Circadian of, 295–96; early feeding, adaptive
rhythms specializations for, 310–13; introduction
Nocturnality, 352–55 to, 289; mammalian development,
Nominal scale measurements, 12 299–302; ontogenetics, definition of,
Nonadditive effects, 161 291–92; overview of, 292–93; in parent-
Nonapeptide hormones: aggressive behavior, offspring relations, 315–17; in regulation
245–47; gregariousness, 243–45; and learning, 313–15; roles of behavior in,
introduction to, 220, 240–47; pair- 294–95; summary, 319–20; views of
bonding behavior, 241–43; sociality and some ontogenetic niches, 296–97
spacing of animal groups, 243–45 Ontogenetic niche: avian development,
Nonapeptide neural circuitry, 244 302–4; introduction to, 296–97,
Norway rat, 117–18, 300–302 299–300; Norway rat, 300–302
Nucleus accumbens (NAc), 239, 242 Ontogeny, 290, 291–92, 374–77
Nucleus sphericus (SA), 274, 275 Operant chamber, 83
Nut cracking, 140 Operant conditioning, 53
Optic chiasm (OX), 201
Objectivity, 42 Optimality theory, 67
Obligate brood parasites, 331 Optimal model, 85–86, 87–90, 92–93, 104
Observational sampling, 12 Orangutans, 132–33
Observational studies: data analysis, 13–15; Ordinal measurements, 12
definition of, 6; introduction to, 10–12; Organic selection, 50
magic coin, 15–18; recording medium, The Organization of Behavior (Hebb), 54
13; recording method, 12–13; Orians, Gordon, 6, 21
sampling, 12 Orientation, navigation, and migration:
Observations, vs. experiments, 6–8 animal compasses, 390–95; challenges
INDEX 453

and strategies, 387–88; inertial navigation Pfaff, Donald W., 64

(dead reckoning), 399–401; piloting, Phase response curve (PRC), 202, 361
397–99; taxis and kinesis, 388–90; true Phenotypic plasticity, 72, 169–72, 267
navigation, 401–7; vector navigation, Phenotypic variation, 175
395–97 Pheromones, 164
Ormerod, Tom, 97–98 Photic entrainment, 359
Oscillator, 202–5 Photoperiod, 360
Outdoor aviaries, 332 Phototaxis, 155
Ovariectomized, 228 Phylogeny, 352–56
Owls, 123 Physiological ecology, 58
OX (Optic chiasm), 201 Physiology, American, 56–60
Oxytocin, 158, 187, 195–96, 220, 241–42 Pictographs, 35
Pied flycatchers, 121–22
Pacemaker, 362 Pigeons: 3-D test, 101–3; ability to plan
PAG (Periaqueductal or central gray), 239 ahead, 101–8; arc configurations, 108;
Page, Rachel, 121 assessment of routes chosen by, 87–90,
Pair bonds: in cowbirds, 333; definition 104–8; center-near vs. peripheral-far
of, 241; neural mechanisms needed for, configurations, 105; cluster configura-
189; nonapeptide hormones, 241–43; tions, 107; cognitive maps, 81, 99–101;
pair-bond behavior, 241; partner determination of routes chosen by,
preferences and, 193–99 85–87, 104; feeding, early adaptations in,
Parental care, 246 311–12; global strategy configurations,
Parental genomic imprinting, 264 105–7; one-way/round-trip global
Parenting behavior: gene regulating, 158; role configurations, 105; as travelers in 2-D
of E and P, 241; steroid hormones and, space, 84–85
227–29 Piloting, 397–99
Parent-offspring relations, ontogenetics Pine seeds, 120
adaptations in, 315–17 Pittendrigh, Colin, 359, 360–61
Park, Orlando, 57 Pituitary gland, 220
Park, Thomas, 57 Plainfin midshipman fish, 225–26
Parker, Geoffrey, 63 Planning ahead: animals and, 108–11;
Partner preferences, 193–99 pigeons and, 101–8. See also Route
Parturition, 241 planning
Paternal care, 191 Plasticity, 208, 217, 268, 333
Paternal imprint, 264 Plato, 39
Pecking behaviors, 311 Pleiotropic effects, 165, 235
Peer-review, 24–25 Pliny the Elder, 40
Peer-reviewed journals, 24–25 Ploger, Bonnie, 13
Peptide hormones, 217 POA (Preoptic area), 239, 274, 275
Peptidergic involvement, 246 Pollen-foraging behavior, 165
Periaqueductal or central gray (PAG), 239 Pollen-hoarding syndrome, 165
Period, 156, 169, 357 Polygynous mating systems,
Periventricular preoptic area (PP), 274, 275 190, 193–94, 196
Perspectives in Ethology, 63 Polymorophisms, 267
PET (Positron emission tomography), 68 Polymorphic repetitive microsatellite
Peterson, Roger Troy, 57 sequences, 243
Petroglyphs, 35 Polyphenisms, 267
454 INDEX

Positron emission tomography (PET), 68 Quantitative traits. See Complex inheritance

Posterior pituitary, 220 and quantitative traits
Posthatching niche, 304 Quasi-experiments, 7, 8
Power, statistical, 17
PP (Periventricular preoptic area), 274, 275 Random effects, 19
PR (Progestin receptors), 237, 238 Randomization, 11
Practical Zoology (Hegner), 57 Randomized blocks, 19
Prairie voles, 158–59, 241–43, 246 Random sampling, 11
PRC (Phase response curve), 202, 361 Random variation, 18
Precision, 18 Ratcheting, 141
Precocial offspring, 240–41, 304 Ratio, 12
Predictions, 4 Rats: ability to discriminate between familiar
Preformationism, 265 and unfamiliar locations, 110; circadian
Preoptic area (POA), 239, 274, 275 function and brain lesions, 360; endocrine
Primates, 53, 316 disrupting chemicals and epigenetics,
Principle of Animal Ecology (Allee, et al), 57 279–80; foraging behavior, 117–19;
The Principles of Psychology (James), 51 germline- and context-dependent epige-
Progesterone, 221, 227–28, 233, 241 netic mutations, 281–82; grass rats,
Progestin receptors (PR), 237, 238 206–7, 209–10; ontogenetic adaptations
Promoter region, 243 in regulation and learning, 313–15; resis-
Proximate and ultimate analyses of behavior, tance to the acquisition of aversions to
integration of: brain basics, 187–89; caregivers, 316–17; roof rats, 120
diurnal mammals, 199–210; in history of Ray, John, 42–43
animal behavior, 71–72; introduction to, Reaction norm, 170
185–87; social systems, 189–99; Reamur, Antoine, 43
summary, 210–11 Receptors, and hormones, 218
Proximate level of analysis. See Proximate Reciprocal altruism, 67, 245
and ultimate analyses of behavior, Recording mediums, 13
integration of Recording methods, 12–13
Proximate mechanisms, 356–69 Red deer, 230
Proximate studies of animal behavior, 326 Red-eyed tree frogs, 306–7
Pruett-Jones, Melinda, 24 Red-sided garter snake, 223
Pruett-Jones, Steve, 24 Red-winged blackbirds, 3–4, 7–8
Pseudorandom, 11 Reflexes, 291
Pseudoreplication, 10, 329–30, 336 Regression, 20
Psychological stressors, 281 Regulation and learning, ontogenetics
Pubarche, 265 adaptations in, 313–15
Reindeer, 208
QTL. See Quantitative trait loci (QTL) Reinhold, Klaus, 168
Quantitative geneticists, 168 Rejection of hypotheses, 5
Quantitative genetics, 153–54, 167–69 Releasers, 56, 311
Quantitative trait loci (QTL): Drosophila as Renaissance, 41–44
model of QTL analysis, 163–64; honey Repeated measures design, 19
bees, 164–65; introduction to, 161–63; Replication, 10, 18
linking QTL for behavior to evolution Reproduction, 66, 68–69, 297–98
using nonmodel systems, 164–67; Reproductive success, 22, 73, 224,
Russian foxes, 166–67 326, 334, 340
INDEX 455

Research funding, 2–3 behaviorist training, 3–4; experimental

Retaliator, 23 studies, 18–20; experiments vs.
Retinohypothalamic tract (RHT), observations, 6–8; hypotheses testing,
201, 361–62 5–6; introduction to, Scientific study of
Reverse genetics, 157 animal behavior; models, 20–24;
Richter, Curt, 362 observational studies, 10–18; peer-review,
Rodents, 316–17. See also Rats 24–25; preliminary considerations in test-
Roeder, Kenneth, 64 ing hypotheses, 8–10; research funding,
Roman civilization. See Greek and Roman 2–3. See also History of animal behavior
civilizations SCN (Suprachiasmatic nucleus), 189, 201–7,
Romanes, George John, 49, 129, 132 361–68
Roof rats, 120 Scorpions, 128
Rooting reflex, 313, 319 Scott, John Paul, 53, 59
Rosenblatt, Jay S., 267 Searcy, William, 7
Rough-skinned newts, 221–22 Seasonal rhythms, 377–80
Route planning: assessment of routes chosen Selective breeding, 36
by people vs. pigeons, 87–90; cost of Sensitive period, 116
inefficient routes, 90–97; human travelers Sensory nervous system, 47
in 2-D space, 83–84; interpretation of Seppanen, Janne-Tuomas, 122
results, 97–101; introduction to, 79; Septum, 274, 275
people vs. pigeons as travelers in 2-D Sequence-analysis methods, 339–40
space, 83–97; pigeons planning ahead, Sequences, 335–36
101–8; pigeon travelers in 2-D space, Sexual behavior: activation of, 220–21;
84–85; planning ahead in animals, evolution of steroid medicated behavior,
108–11; routes chosen by humans vs. 223–24; independent of sex steroids, 223;
pigeons, 85–87; traveling salesperson quantitative genetics, 168–69; steroid
problem, 80–83 hormones and, 220–24; suppression of,
Rowher, Sievert, 6 221–22
Rule-based solutions, 82 Sexual experience, as agent of epigenetic
Rule of one variable, 7 modification, 272–74
Russian foxes, 166–67 Sexual selection, 48–49, 164, 217, 264
Russon, Anne, 132–33 Shaw, Evelyn, 63
Rutting areas, 230 Shelford, Victor E., 57, 58
Ryan, Michael, 121 Shimomura, Kazuhiro, 169
Signals, 335
SA (Nucleus sphericus), 274, 275 Sign stimulus, 56
Saint-Hilaire, Etienne Geoffrey, 46 Silenced genes, 264
Sampling, 12 Silver foxes, 166–67
Sato, Tetsu, 205 Simplicity, 18
Say, Thomas, 50 Simulations, 23–24
Scan sampling, 13 Single-gene effects on behavior: finding
Schmidt, Karl P., 57 food, 156–57; introduction to, 154–55;
Schneirla, Theodore C., 267 keeping time, 155–56; nurturing
Schwartz, William, 203 offspring, 157–59
Science, 49 Sinusoidal model, 360
Scientific study of animal behavior: animal Skinner, B. F., 52–53, 153
behaviorists, 1–6, 10–24, 74–75; animal Skinner, Michael, 279
456 INDEX

Skinner box. See Operant chamber Society for Neuroscience, 64

Smith, Douglas, 6 Society for the Study of Reproduction, 68
“The Snark Was a Boojum” (Beach), 59 Sociobiology, 53
Social behavior, 48–49, 239–40 Soft skulls, 308–10
Social behavior network, 239 Sokolowski, Marla, 156–57
Social behavior neural network, 239–40 Somatic adaptation, 294
Social group as habitat, 301, 303 Songbirds, 331
Social isolation, 53 Song sparrows, 231
Sociality: bird song, 326–28, 331–34, Space, 191–93
335–36; conceptual challenges Spaulding, Douglas, 49
underlying, 326–28; definition of, 325; Specific action potential, 56
methodological challenges underlying the Spiny mice, 355–56
study of, 328–30 Statistical analysis, 329–30
Sociality and spacing of animal Stealers, 24
groups, 243–45 Steiniger, von Fritz, 117–18
Social learning: attacking toxic Stephan, Fred, 200–201
prey, 120–21; counting, 116–17; eating, Stephen, Friedrich, 362
117–20; examples of, 115–25; imitation, Steroid hormones: aggressive behavior,
125, 126, 129–33; introduction to, 115; 229–33; androgens and behavioral trade-
summary, 142–43; teaching, 127–28; offs, 235–36; androgens and immunity,
traditions and culture, 133–42; types of, 236–37; androgens and mortality,
125–33; what to fear, 122–23; 234–35; biosynthesis pathway, 219; costs
where to nest, 121–22; with whom to associated with steroid action, 233–37;
mate, 123–25 courtship behavior, 224–27; mechanisms
The Social Life of Animals (Allee), 58 of steroid action, 237–40; overview
Socially distributed behavior, 332 of, 218–20; parenting behavior,
Social network analyses, 332 227–29; roles of, 217;
Social network analyses sex behavior, 220–24
methodology, 337–39 Steroidogenic glands, 218
Social network approach to study of behavior: Steroid receptors, 237–39
conceptual challenges underlying sociality, Stinging behavior, 164–65
326–28; experimenting on groups, Stone, Witmer, 57
331–34; introduction to, 325–26; lab Strategy, 22
traits vs. wild traits, 330–31; Stress, as agent of epigenetic modification,
methodological challenges underlying 280–81
study of sociality, 328–30; newest Striatum, 239
advances in, 334–35; sequences and Strong inference, 73–74
contingencies, 335–36; social networks, Studying Animal Behavior (Dewsbury), 63
337–39; summary, 342–43; utility of The Study of Instinct (Tinbergen), 55
social sequence analysis, 339–42 “Subjective day,” 359
Social sequential analysis, utility of, 339–42 “Subjective night,” 359
Social systems: academic focus on, 66; Super-precocial species, 304
introduction to, 189–91; partner Suprachiasmatic nucleus (SCN),
preferences and pair bonds, 193–99; 189, 201–7, 361–68
space, 191–93 Suspension feeders, 310–11
Society for Behavioral Swammerdam, Jan, 43
Neuroendocrinology, 64 Synapses, 185
INDEX 457

Tadpoles, 297–98, 307, 310–11 Traditions and culture: culture in

Takahashi, Joseph, 169 chimpanzees and humans, 140–42;
Teaching, in social learning, 126, 127–28 development of chimpanzee traditional
Tebbich, Sabine, 135 behaviors, 139–40; ecological effects on
Technological advances in studying animal traditional chimpanzee behaviors,
behavior, 330–31 137–38; evidence of in apes, 136;
Temperature, as agent of epigenetic exclusion, method of, 136; Galápagos
modification, 272–74 finches, 135–36; genetic effects on
Temporal niche, 352–55 traditional chimpanzee behaviors,
Temporal partitioning, 355 138–39; historical examples of animal
Temporal triads, 340 tradition, 133–35; Japanese macaques and
Termite fishing, 139–40 British birds, 133–35; method of
Territoriality, 191 exclusion, 140; recent examples of animal
Territory, 4, 6, 70, 222, 294 tradition, 135–36
Testing hypotheses, 5–6, 8–10 Training animal behaviorists, 3–4
Test of independence, 20 Transcription factors, 238
Testosterone: activation of sex behavior, 221; Transcriptome, 167
aggressive behavior, 41, 230–31; as an Transgene coding, 243
androgen, 218; androgens and mortality, Transgenerational inheritance, 270–71
234–35; costs of, 234; courtship behavior, Transgenic mice, 197–98
224–25, 226; evolution of steroid Transgenic voles, 197–98
mediated sex behavior, 223; female Transport reflexes, 317–18
aggression, 232–33; in goldfish, 238; Traveling salesperson problem, 80–83
immunity, 236–37; male parenting Treatment effects, 18
behavior, 228–29; pleiotropic effects, Treatment of animals, guidelines for, 9–10
235; stress and, 282; suppression of sex Triadic approach, 337–39
behavior, 222 Triadic distributions, 340–42
Theriomancy, 38 Triadic interactions, 338–39
Thermotactic responses, 314 Triadic level of analysis, 342–43
Thorndike, Edward L., 52, 129 Trivers, Robert, 21, 64
Thorpe, W. H., 55 Tropisms, 58
3-D test, 101–3 True navigation, 401–7
Time, keeping, single-gene Trust, 245
effects on, 155–56 Túngara frogs, 121
Time cues, 359 Two-action method, 130
Timeless, 156, 169 Type I error, 16–17
Timing of hatching, 306–7 Type II error, 16–17
Tinbergen, Niko: drawings, 2; four questions,
8, 290, 352; in history of animal behavior, Uexkill, Jakob von, 49–50, 55
54–55, 62, 153–54; Nobel Prize, 89; Ultimate causation, 71–72
pecking behavior, 311; ultimate and Ultimate level of analysis. See Proximate
proximate levels of analysis, 186, 210 and ultimate analyses of behavior,
Tobach, Ethel, 63 integration of
Toro, Miguel, 141 Ultimate studies of animal behavior, 326
Tousling, 316 Umwelt, 49–50
Toxic prey, attacking, 120–21 Underground bunker experiment, 357–59
Trade-offs, 235–36 Uniformitarianism, 48
458 INDEX

University of Pennsylvania, 332 Watson, James B., 52

Urination reflex, 319 Waxbills, 243–44
Uterus as habitat, 300–301 Weissman, August, 265
West, Meredith, 63–64, 116
Validity, 6–8 West-Eberhard, Mary Jane, 63
Vampire bats, 300 Wheeler, William Morton, 49
Van Schaik, Carel, 136 White, Gilbert, 44
Variables, 18 White-crowned sparrows, 222
Vasopressin (AVP). See Arginine White-footed mice, 246
vasopressin (AVP) Whiten, Andy, 136, 137, 140
Vasotocin, 220, 244–45 Whitman, Charles O., 49, 54
Vector navigation, 395–97 Wide applicability, 18
Ventral palladium (VP), 239, 242 Wild traits, vs. lab traits, 330–31
Ventral tegmental area (VTA), 188, 239 Wild type mice, 231–32
Ventromedial hypothalamus (VMH), 239, Williams George C., 64
274, 275 Willoughby, Francis, 42–43
Verifiability, 42 Wilson, Edward O., 64
Vervet monkeys, 99 Wingfield, John, 235
Vesalius, Andreas, 42 Woltereck, Richard, 265–66
Vince, Margaret, 306 Working hypotheses, 4
VMH (Ventromedial hypothalamus), 239, Wright, Sewell, 153
274, 275
Vocalizations, 224–26 Yerkes, Robert M., 53
VP (Ventral palladium), 239, 242 Young, Larry, 197, 243
VTA (Ventral tegmental area), 188, 239 Young, William C., 64

Waddington, Conrad, 266 Zeitgebers, 359

Wallace, Alfred Russell, 48 Zeitschrift für Tierpsychologie, 43
Warkentin, Karen, 307 Zoology, American, 56–60
Washburn, Margaret Floy, 53 Zoos/menageries, 43–46
Wasserman, Stanley, 338 Zucker, Irving, 200–201, 360, 362
 Animal Behavior
How and Why Animals Do the Things They Do

Volume 2: Function and Evolution

Ken Yasukawa, Editor

Copyright 2014 by Ken Yasukawa
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted, in any form or by any means, electronic, mechanical,
photocopying, recording, or otherwise, except for the inclusion of brief quotations in a
review, without prior permission in writing from the publisher.
Library of Congress Cataloging-in-Publication Data
Animal behavior : how and why animals do the things they do / Ken Yasukawa, editor.
pages cm
Includes bibliographical references and index.
ISBN 978–0–313–39870–4 (hard copy : alk. paper) — ISBN 978–0–313–39871–1
(ebook) 1. Animal behavior. I. Yasukawa, Ken, 1949–
QL751.A6498 2013
591.5—dc23 2013012228

ISBN: 978–0–313–39870–4
EISBN: 978–0–313–39871–1
18 17 16 15 14 1 2 3 4 5
This book is also available on the World Wide Web as an eBook.
Visit www.abc-clio.com for details.
Praeger
An Imprint of ABC-CLIO, LLC
ABC-CLIO, LLC
130 Cremona Drive, P.O. Box 1911
Santa Barbara, California 93116-1911
This book is printed on acid-free paper
Manufactured in the United States of America
 Contents

1. Finding Food: Foraging Affects All Aspects of an Animal’s Life 1

Anna Dornhaus

2. Predation and Antipredator Behavior 23

Theodore Stankowich

3. Animal Communication: How and Why Animals

Communicate—or Do They? 51
Zuleyma Tang-Martínez

4. Mating Systems and the Measurement of Sexual Selection 99

Kenyon B. Mobley

5. The Evolution of Ornaments and Armaments 145

Geoffrey E. Hill

6. Sexual Conflict: All’s Not Fair in Love—It’s War! 173

Zenobia Lewis

7. A Nest of Vipers: Conflict and Cooperation in Families 195

Michelle Pellissier Scott
iv CONTENTS

8. Make Space Enough between You: Intraspecific Variation

in Animal Spacing 221
Nancy G. Solomon and Brian Keane

9. Let’s Get Together: The Evolution of Social Behavior 257

Walter D. Koenig and Janis L. Dickinson

10. Ecological and Evolutionary Feedbacks in the Evolution

of Aggression 295
Renée A. Duckworth

11. Altruism and Kinship 327

Lee Alan Dugatkin

12. Evolutionary History of Behavior 339

Terry J. Ord
Glossary 359
About the Editor and Contributors 377
About the Editorial Board 381
Index 383
 1

Finding Food: Foraging Affects All Aspects

of an Animal’s Life
Anna Dornhaus

INTRODUCTION
Why do horses sleep only 3 hours a day, while lions may sleep 18 hours? Why
do birds commonly sing at dawn rather than in the afternoon? Why are
chimpanzees violent and patriarchal in their social life, while the closely related
bonobos are peaceful and females dominate males? Behavioral ecology is the
science that seeks to explain the diversity in lifestyle we find in animals,
including the origins of our own human habits. Foraging plays a central role
in this: all of the above questions have been answered by referring to the forag-
ing needs of these respective species, demonstrating the central role that
finding food plays in the evolution of animal behavior.
Foraging in animal behavior research refers to finding and collecting re-
sources. Foraging for most animals is one of the most energetically costly
and risky activities, particularly if the animal has to leave its home or shelter
and cover great distances. Because of this, the requirements of foraging affect
many other aspects of an animal’s life. For example, plant matter has a much
lower energy density than meat, which means that herbivores, such as horses
(Equus ferus), may need to spend a lot more time collecting food than carni-
vores such as lions (Panthera leo), who can afford to spend a lot of time resting
(Lesku et al., 2009). Some animals have so little reserves that they have to
ration energetically expensive activities depending on the amount of food
2 ANIMAL BEHAVIOR

found each day—this is thought to be the case for small birds in cold climates,
who can only afford to sing (to defend territory or attract mates) when they
have survived the night and find that they still have fuel left over, leading to
a dawn chorus (Hutchinson et al., 1993; Thomas, 1999). In the longer term,
adaptations to the necessities of foraging change a species’ overall life style. For
example, bonobos (Pan paniscus) eat leaves and shoots of plants, which are
common and easy to find. This allows groups of related females to stay together,
which led to the evolution of strong female-female bonds. Chimpanzees (Pan
troglodytes), on the other hand, mostly eat fruits and other foods that are patchy
in distribution and require travel—which disadvantages females with infants,
who travel slowly and thus have to search on their own if they want to avoid
arriving at a resource already depleted by others in the group. This may prevent
intense bonding and thus lead to the more patriarchal, aggressive societies we see
in chimpanzees compared to bonobos (Wrangham, 1993). Even our unique
human traits, such as large brains, cooperation, and trading, are thought to
depend in large part on novel foraging behaviors by early hominids (Pennisi,
1999; Boehm, 2004).
Foraging is among the best-studied animal behaviors, due to the fact that
foraging animals typically have to leave their shelter and move around, making
them more visible and susceptible not only to predation but also to research
(Kamil and Roitblat, 1985). The second advantage of foraging behavior, from
a research perspective, is that success or failure in foraging is more easily
defined and measured than in many other activities: a foraging animal collects
a food of a certain caloric value in a certain amount of time. This led to forag-
ing behavior being one of the first topics in which quantitative and predictive
models were used intensively. The field of modeling foraging behavior came to
be known as optimal foraging (Pyke et al., 1977; Stephens et al., 2007). The
fact that measuring foraging success seemed so straightforward allowed models
to actually predict optimal behavioral strategies in extraordinary detail. This
led to some criticism, as some felt that such precise predictions were unrealistic
or unhelpful (reviewed in Stearns & Schmid-Hempel, 1987; Raine et al.,
2006). However, all fields of studying animal behavior, and perhaps of organ-
ismal biology, have benefited from the idea that the “target” of selection,
or the optimal behavior, is not necessarily obvious to a casual observer and
can be derived quantitatively and objectively within given constraints. Much
of biology is now emulating similar approaches, trying to clarify precisely
which factors will drive evolution of organismal traits and in what directions.
These can be nonintuitive, particularly in cases where subtle differences in
assumptions can actually lead to differences in the direction of selection
(Kacelnik, 1993).
FINDING FOOD : FORAGING AFFECTS ALL ASPECTS OF AN ANIMAL ’ S LIFE 3

One behavior thus studied is the phenomenon of “partial loads” in central-

place foragers, particularly bees (Raine et al., 2006). Honey bees visiting a
plentiful nectar source will often not carry as much back to their hive as they
could, despite the fact that the same forager will make multiple subsequent
trips to the same source to collect more nectar. This seems counterintuitive
and indeed inefficient: why not carry a full crop with every trip? Mathematical
models have helped us to understand the conditions under which partial,
instead of full, loads may be advantageous to bees. First, if a flower offers
nectar at decreasing returns, that is, the last bit of nectar is harder to extract,
then it may be worth leaving before a full load has been collected in order to
visit more other sources (Charnov, 1976). Second, it may be that if heavy
loads impose high transport costs, and if a bee’s lifespan is affected by how
hard it has to work, the total nectar collected in the life of a foraging bee is
in fact higher if only partial loads are collected (Schmid-Hempel et al., 1985;
Wolf & Schmid-Hempel, 1989). Third, bees can recruit their nestmates to a
rich food source; it may be worth limiting the amount of time the first forager
spends there in favor of recruiting more bees to the resource before it is discov-
ered by competitors, thus increasing the success of the group as a whole at the
expense of the success of that first individual (Núñez, 1982; Varju & Núñez,
1991). Finally, since information about new, superior sources could arrive at
any time at the hive, a forager may prefer to return to the hive frequently to
check for such information in order to prevent wasting time at an inferior re-
source (Dornhaus et al., 2006). Mathematical models have allowed researchers
to derive clear, quantitative predictions not only about how early a forager
should leave a resource but also about which factors should affect this decision
and how—for example, according to the third hypothesis, bees should collect
smaller loads if the resource is of high quality, whereas the fourth hypothesis
predicts that they should collect larger loads in this case. Empirical tests have
revealed that the former is the case, supporting the idea that honey bee forag-
ers sacrifice their own foraging success in order to improve that of the hive as a
whole (Varju & Núñez, 1993). Without mathematical descriptions of the fac-
tors involved, behavioral scientists would not have been able to distinguish
between these hypotheses. Moreover, the models pointed to what had to be
investigated empirically in order to understand this aspect of foraging
behavior.
In the example above, models helped explain a trait that had already been
observed in bees, namely partial loads. But what if the animals do not perform
the behavior the model predicts? Animals may not be optimally adapted, and
any studies trying to identify the adaptive function of a trait or behavior are
sometimes criticized as being adaptationist, that is, naïvely assuming that
4 ANIMAL BEHAVIOR

evolution will be able to lead to perfection (Gould & Lewontin, 1979). How-
ever, this is misunderstanding the aim of such models. First, without quantita-
tive models, we would not be able to determine whether a particular behavior
in fact is the one that, given certain constraints, is most likely to allow the ani-
mal to reach its goal. Second, models help in defining the problem (the
constraints and the goal) more sharply (Stephens et al., 2007). What is the
“goal” of any animal behavior? The behaviors that are most likely to persist in
a population are those that allow individuals to contribute maximally to the
genetic makeup of the next generation, largely by maximizing the number
of offspring produced. This is what we call evolution. There are many com-
ponents to this—finding and selecting a mate, avoiding predators and
diseases, finding shelter or otherwise dealing with the elements, and in many
cases establishing a place in a social community (Davies et al., 2012). But for
many animals, the amount of food collected is a limiting factor in how
many offspring can be produced. This means that behaviors that allow an
animal to collect more food often have a direct effect on fitness, enabling
researchers to employ the shorthand of using “food collected” or “food
collected per time” as so-called fitness proxies, that is, as the currencies that
animal behavior is adapted to maximize (Ydenberg & Hurd, 1998). Indeed
animals perform behaviors that we can only understand as strategies to
increase foraging success by just a few percent points; the existence of such
behaviors shows that evolution can lead to subtle adaptations and optimiza-
tion. It is important, however, to recognize the limitations of such fitness
proxies—primarily, that other aspects of the animal’s life are ignored for the
purpose of a particular study.
If an animal appears to behave in a way very different from the predictions
of a model, this may mean that the animal has not yet evolved the optimal
behavior. However, in many cases, it is likely to mean that we have not yet
properly understood which problem the animal is trying to solve: we may
not understand the constraints under which it is operating, or we may not
know the currency being maximized—for example, perhaps the animal needs
to avoid predators or attempts to find mates while foraging (McArthur et al.,
2012). Researchers now also study how the computational mechanisms that
enable animals to choose behaviors themselves evolve (McNamara & Houston,
2009; Fawcett et al., 2012a, 2012b). For example, this allows us to understand
predictable mistakes that result from applying a rule of thumb that provides the
best results when averaging over the situations the animal commonly finds
itself in but that may not appear optimal in a particular situation, or laboratory
experiment (Kamil & Roitblat, 1985; Stephens et al., 2004; Seth, 2007;
Houston, 2009). Interestingly, behaviors that are identified with such mental
FINDING FOOD : FORAGING AFFECTS ALL ASPECTS OF AN ANIMAL ’ S LIFE 5

states as optimism, disappointment, and impulsivity may be interpreted as

resulting from such rules of thumb—more on this below.

THE FORAGING TRIP

A honey bee leaves the hive to search for flowers that may provide nectar.
The bee starts this trip because she is old enough, has a genetic predisposition
to nectar foraging, and has sensed some commotion near the hive entrance, a
likely indication that some of her nestmates are already foraging successfully.
She (all honey bee workers are female) flies to a stand of cherry trees about a
mile from the hive, using her knowledge of local landmarks to find them.
Although she foraged successfully from these trees a few days ago, their bloom
has faded and the bee starts searching elsewhere. Relying on her innate prefer-
ences, she picks some blue wildflowers, but they either do not contain nectar
or it is too deep for her to reach. After each unsuccessful attempt, she travels
a bit further until she can taste the sweet sugary nectar in some yellow flowers
on a meadow. After first locating nectar, she changes her travel path to a more
meandering one, testing every yellow flower but bypassing flowers of other
colors or shapes. Many of the yellow flowers are empty, but some contain
nectar—never more than a tiny drop, often less than a few microliters. After
couple hundred flower visits and having collected about a third of her body
weight (although she could carry more than two times as much), the bee
returns to her hive. There, she passes the collected nectar on to hive-workers;
but since she had to queue to complete this task, she decides not to communi-
cate her discovery but to return immediately to the meadow for another forag-
ing trip.
Foraging behavior has many facets. In an animal as well studied as the
honey bee, we know many of the decisions made and the factors that will
affect them. In bees, early naturalists as well as modern behavioral scientists
have studied the factors that prompt bees to start and stop foraging trips,
choose or reject certain flowers, navigate away from and back to their nest,
and communicate with their nestmates to affect their foraging behavior in
turn. I will discuss each of these aspects of foraging separately below.

TO FORAGE OR NOT TO FORAGE

At each trip or bout of foraging activity, an animal has to make several deci-
sions, starting with whether to attempt to collect food at all, or when to do so.
This, as any other decision, depends on the benefits and costs of foraging com-
pared to another activity the animal could engage in instead. Benefits of
6 ANIMAL BEHAVIOR

foraging of course depend on the prospects for success, and many animals
innately prefer to forage when cues such as weather indicate that resources will
be plentiful. For example, desert ants are sensitive to overheating and will for-
age only under cool or moist conditions (Cerda et al., 1998). On the other
hand, ants can also learn to associate temperature with food (Kleineidam
et al., 2007). Many animals learn cues associated with resources, good foraging
conditions, or even sophisticated combinations; for example, bees learn not
only which areas are most profitable but also at what times of day which loca-
tions are most likely to provide food. They can even learn how likely it is that
resources will change in profitability, that is, learn when it is worth learning
(Dornhaus & Franks, 2008).
Whether a particular animal attends to environmental cues in its decision
to forage should depend on how much these cues predict foraging success.
For example, the availability of edible leaf material will not change in the short
term for a caterpillar sitting on a plant, but the availability of nectar in flowers
changes quickly and drastically with the time of day and weather, making it
useful for bees to attend to such cues. Whether a particular cue provokes an
innate response or is only used after the animal has learned to associate it with
foraging success depends on how predictable its effects are over the long term
(Dunlap & Stephens, 2009). The life cycle and circadian rhythm of specialist
cactus bees, which forage on the flowers of a particular cactus, are adapted to
the timing of nectar availability in this plant (McIntosh, 2005). The timing
of foraging in generalist honey bee foragers, on the other hand, is flexible
and is shaped by the bees’ experience with the nectar availability of plants in
the current season and site. This difference between specialist and generalist
foragers has wide-ranging consequences for species resilience under environ-
mental change.
The benefits of foraging depend not only on the expected returns but also
on how the food is used. Is food necessary to satisfy the immediate energy needs
of the animal, is it stored for future need, or is it fed directly to offspring or other
individuals? If food collected is used primarily for provisioning, that is, stored
or provided to other individuals, then foragers may attempt to minimize ener-
getic expenditure of foraging because any food used up for the foraging activity
itself will not be available to provision. When just foraging for themselves,
however, animals may behave so as to maximize daily gain without regard
for how much is spent attaining it (Ydenberg et al., 1994). For example, the
sooty shearwater (Puffinus griseus), a seabird, may travel 1,500 km from its nest
to catch fish in productive waters to maximize its own energy gain, but when
feeding chicks will often perform short trips, which lead to less expensive travel
but are less productive (Weimerskirch, 1998).
FINDING FOOD : FORAGING AFFECTS ALL ASPECTS OF AN ANIMAL ’ S LIFE 7

Do animals ever have “enough food”? The idea that foraging behavior may
not so much evolve to maximize food collected as ensure that a certain mini-
mum food requirement is met has been called satisficing in foraging research
and has been somewhat controversial (Nonacs & Dill, 1993; Ydenberg &
Hurd, 1998). It seems obvious from observation that many animals spend a
lot of time in inactivity when they could be collecting more food—remember
the sleeping lions mentioned earlier in this chapter. However, evolutionary
considerations suggest that only if additional food can be converted into off-
spring or stored as a buffer against lean times, it should be collected regardless
of current energy requirements. Indeed, many cases of inactivity are explained
by the fact that either another factor limits reproduction in the long term or
that current energetic costs of foraging exceed its benefits (Herbers, 1981).
For example, there may be limits to the speed with which food can be digested
or processed (Burd & Howard, 2005). Alternatively, if there is temporal varia-
tion in food availability and food cannot be stored, how many offspring can be
raised may be limited by the minimum amount of food that can always be
collected; in brief periods of food abundance, an animal would then appear
to forage less than expected compared to food availability simply because not
enough offspring are present to consume this food. Overall, however, the fact
remains that animals vary in the degree to which they display inactivity
(Careau et al., 2008) or sleep (Lesku et al., 2009), and that the function of
neither is completely understood (Cirelli & Tononi, 2008).

FINDING FOOD
Once an animal has decided to forage, it has to actually locate its prey or
food source. The pattern in which animals search is thought to resemble a
random walk, in which each new step is in a new, random direction. “Corre-
lated” random walks imply that the new direction is affected by the previous
direction, usually such that the animal is more likely to keep a straight heading
rather than choosing a particular different direction. Even within these
assumptions of a seemingly mindless search algorithm, a search path can be
optimized for the presumed density and clumping of food sources (Nathan
et al., 2008; Sims et al., 2008). However, many animals usually forage in a
known area and may return to previously profitable locations or travel along
known paths (called traplines). For example, bumble bees tend to visit several
known profitable flowers in a fixed sequence (Thomson, 1996). These bees are
doing multiple foraging trips in a row, and the traplining strategy evens out
the time intervals after each flower is revisited. It turns out that this is the opti-
mal strategy for resources that slowly but at a constant rate renew themselves,
8 ANIMAL BEHAVIOR

such as flowers that typically produce nectar continuously (Ohashi et al.,

2007). If a food item has been located by a forager, the forager will often
concentrate further search in the same area, a phenomenon loosely termed
“win-stay, lose-shift” (implying that if you find something, stay close; if not,
shift to another area), a strategy that is optimal if food distribution is patchy
in nature (Hodges, 1985).

WHAT NOT TO EAT

Once a forager comes near a prey item, it has to both detect it as a stimulus
and decide to attack it. Detection will depend on the foragers’ physiological
sensitivity to the color, pattern, smell, or sounds produced by the prey but also
on the foragers’ experience and current state of mind. Detecting a stimulus
from a noisy scene is often hard; much research in computer science is devoted
to trying to replicate the skills of even simple animals in visually recognizing
particular items under varying and often patchy lighting, with a noisy back-
ground, and from different angles (Smeulders et al., 2000; Galleguillos &
Belongie, 2010). In addition, although some food items, such as flowers for
bees, actively signal their presence (because flowers are pollinated in exchange
for the nectar they offer), others, such as prey animals to be eaten, will be try-
ing to avoid detection. Also, foragers will likely encounter many objects that
are either harmful or a waste of time but that resemble the sought prey items.
For example, orchids have no nectar but mimic other flowers to fool bees
(Dafni, 1984; Gigord et al., 2002), yellow or white crab spiders will camou-
flage on flowers of the same color (Chittka, 2001), and noxious butterflies
may look similar to tasty ones to a bird (Skelhorn & Rowe, 2010; Stoddard,
2012). Foragers thus need to be discriminating and will often evolve particular
sensory sensitivity to the most useful stimulus associated with their preferred
food (Sabbah et al., 2010). In addition, foragers will often reject items that
appear too similar to unwanted items, even to the point of rejecting many suit-
able ones; this is called the peak shift phenomenon in signal detection theory
(Chittka et al., 2009; Lynn et al., 2005).
If a forager is searching for a particular prey type, it tends to be able to
detect items of this type more easily than other potential prey; the forager is
said to have a search image (Dukas & Ellner, 1993; Dukas, 2004). Why this
is the case is not entirely clear, but it may be that even if foragers can recognize
several prey types they have difficulty holding more than one in their working
memory, where it is readily accessible for comparison with the scene encoun-
tered. For example, bees are known to show “flower constancy,” where profit-
able flowers are ignored in favor of visiting a series of flowers of the same type,
FINDING FOOD : FORAGING AFFECTS ALL ASPECTS OF AN ANIMAL ’ S LIFE 9

sometimes for an extended period. If a bee switches to a different flower type,

it appears to take longer to recognize it. This may reflect the neural costs of
accessing information that is not currently in the working memory of the
bee (Waser et al., 1996). Several other explanations for flower constancy have
also been proposed, however, from the difficulty of learning to handle multiple
flower types to avoiding competition with other bees. Indeed, specialization
on particular food sources is thought to confer several advantages—ability to
specialize on finding, handling, and digesting food as well as savings in cogni-
tive complexity of the forager (Bernays, 2001).
Once a food item has been identified by the forager, the forager is still faced
with choices: sometimes among several foods, sometimes between choosing to
attack a particular food item now or to ignore it in favor of searching for a bet-
ter one (Stephens et al., 2007). These choices are referred to as diet selection
and may be based simply on how long it takes to handle particular food items
or on a multitude of factors affecting item quality, including specific nutrients
required by the forager (Sih & Christensen, 2001). Animals use a variety of
direct and indirect ways to assess the quality of a resource. One of the interest-
ing aspects of this is that the quality assessment appears to be strongly influ-
enced by recent experience: animals exhibit behavior that appears to show
disappointment or even optimism and pessimism (Morgan et al., 2012; Nettle
& Bateson, 2012). Together, these phenomena have been termed contrast
effects because they appear when animals experience a contrast between what
they have experienced in the past and what they are experiencing now. It is
not entirely clear why they occur—after all, if a particular food item is worth
more than the effort to handle it, the animal should be using it regardless of
whether it is better or worse than expected. However, this is not how animals
act. Similarly unexpected, apparently nonoptimal behavior has been shown in
other contexts, such as when animals choose a small reward now over a much
bigger one later (“impulsivity” [Stevens et al., 2005]), when risky alternatives
are preferred or rejected compared to stable ones with the same average yield
(“risk sensitivity” [Caraco et al., 1990; Cartar & Abrahams, 1996]), or when
irrelevant options bias choice (“irrationality” [Shafir et al., 2002]). New
research is now uncovering the actual cognitive mechanisms, or decision
rules, that animals use and how they have evolved. Often, particular decision
rules evolve because they usually, across environments or stimuli typically
encountered by a particular type of animal, function well; this does not mean
that they provide optimal solutions in all contexts (Houston et al., 2007;
Fawcett et al., 2012b). Contrast effects, impulsivity, risk sensitivity, and irra-
tionality may be examples of such rules—for example, animals adapted to
catching flighty insects may be more impulsive than those using stable
10 ANIMAL BEHAVIOR

resources such as plants (Stevens et al., 2005). In addition, previous experience

allows the animal to assess its own state as well as that of its environment,
allowing better prediction of what is likely to be experienced in the near future
(Schuck-Paim et al., 2004). This provides an intriguing angle to view human
emotions: are our irrational-seeming reactions in some contexts the result of
evolved decision rules (Bateson et al., 2011; Nettle & Bateson, 2012)? Many
of the insights on diet selection, or how animals judge whether to use a par-
ticular resource, also apply to contexts other than foraging, for example when
selecting mates; very similar models are also used for humans. For example,
models in sequential choice theory calculate whether a prey item of a particu-
lar quality and taking a particular time to handle and digest should be chosen
depending on what other items the forager is likely to encounter later
(Stephens et al., 2007); the same models can be used to predict whether an
animal should choose a mating partner of a particular quality (given the distri-
bution of potential mating partners) (Castellano et al., 2012) and can also be
used to calculate whether a person should take a job offered based on what
he or she estimates will be offered in the future and other problems of this
nature.
Choices while foraging are also affected by other aspects of an animal’s life,
such as the risks of encountering predators, mates, or competitors (McArthur
et al., 2012; Wajnberg, 2012). Animals are almost always vigilant to avoid a
predator attack and often switch between looking for food and surveying the
area for possible threats. If the probability of being attacked is higher in
particular patches, animals demand “hazardous duty pay” (Brown & Kotler,
2004); in other words, such patches have to be particularly profitable to be vis-
ited by a forager. Another aspect is that some areas may be safer, for example if
they provide more cover, and will be preferred in the presence of predators;
this can lead to changes in overall food use, as in deer who avoid open aspen
stands when wolves are present in the area (Fortin et al., 2005). Similarly, ani-
mals often avoid competition with individuals of the same or other species.
This may be achieved, for example, by specializing on a subset of the suitable
foods or by restricting activity to particular times. Dominant individuals or
species will then often forage during the most profitable times and on the most
profitable sources, while others may be displaced to suboptimal ones (Alanara
et al., 2001). For example, in desert ants foraging on cacti for nectar (from
extrafloral nectaries; see below), the less dominant species forage later in
the day, at higher temperatures, putting them at higher risk of dehydration
and overheating (Morris et al., 2005).
When a forager ends its foraging bout or leaves a patch of resources, it may
then return to its nest to unload and start a new foraging bout or turn to other
FINDING FOOD : FORAGING AFFECTS ALL ASPECTS OF AN ANIMAL ’ S LIFE 11

activities. If it does this after every trip, it is central-place foraging (Orians &
Pearson, 1979); other animals have no fixed home base and instead roam
widely within a home range or over large distances (Hays et al., 2006). When
to stop foraging in a patch is a decision that depends largely on how the rate of
acquiring food from the patch changes over time and how long the forager has
to travel between the food patch and its home base. If the food source provides
decreasing marginal returns, that is, it becomes harder and harder to extract
food, the forager will eventually leave it to find a new, fresh resource, possibly
well before it has collected all it can (see discussion on partial loads above).
How foragers navigate the route back to their nest or among foraging patches
is another interesting research area that is discussed in Volume 1, Chapter 12.

EXTRACTING REWARD
So far, we have mostly discussed how to find and recognize food and how
to decide whether to attack it—whether it is a prey animal or a plant or any-
thing else. However, actually capturing the prey and extracting the food
reward can be a challenge in itself and may be a stereotyped behavior or exhibit
a high degree of innovation; it may also involve sophisticated behavioral
sequences, whether innate or learned. Consider the emerald cockroach wasp
(Ampulex compressa [Fouad et al., 1996; Libersat, 2003]). The wasp needs live
but subdued insect prey to feed to her developing offspring. The female will
hunt for cockroaches and, once they are found, sting into the brain of the
roach precisely in the suboesophagal ganglion. This, apparently, turns the
roach into the wasp’s puppet zombie! The roach appears to lose all will of its
own and is guided—live and able to walk—to its doom in the form of the
wasp’s nest and her larvae (Libersat et al., 2009). This foraging strategy on
the part of the wasp is not only surprising and unique, it is also fairly complex
but very likely completely genetically determined. Many other animals, on the
other hand, have been shown to learn, and even pass on to others in a type of
culture, particular foraging strategies useful in a local area (Volume 1, Chapter
4). For example, dolphins off the coast of western Australia have been shown
to harvest cup-shaped sponges to use as a “glove” over their sensitive snouts
when digging on the sea floor (Krutzen et al., 2005); great tits (Parus major)
famously learned to open milk bottles with aluminum caps left on British
doorsteps (Fisher & Hinde, 1949; Lefebvre, 1995); and apes such as chimpan-
zees and orangutans have been shown to master a variety of tool-using tech-
niques in foraging, from cracking nuts with wooden hammers and anvils to
fashioning spears for hunting monkeys (Whiten et al., 1999; van Schaik et
al., 2003). Why do some animals have such sophisticated, flexible foraging
12 ANIMAL BEHAVIOR

strategies? It was long thought that carnivores, hunting for live animal prey,
needed to be particularly clever and perhaps had evolved more cognitive and sen-
sory abilities as a result. However, the evidence for this is not clear (Garamszegi
et al., 2002). On the other hand, it seems that innovative foraging behavior
enables some animals to invade new habitats more easily (Sol et al., 2002).

SOCIAL FORAGING
Many animals do not just rely on their own skills and experience to find
and identify palatable and profitable food sources (Galef & Giraldeau, 2001;
Laland, 2004; see also Volume 1, Chapter 4). For example, many birds will
closely watch others in the flock searching for seeds on the ground; if another
bird appears to have found something, they will then quickly join it to exploit
that same resource. This strategy has been termed “scrounging” (as opposed to
“producing,” which refers to birds searching for food themselves) (Giraldeau
et al., 1994). Bees may also watch conspecifics choose flowers and instead of
attempting to use the exact same flower may use this public information to
pick other flowers of that same color (instead of using private information,
or sampling many flower types themselves to determine which species provide
the most nectar) (Leadbeater & Chittka, 2007). Rats are known to learn the
scent of palatable food from the smell at the mouth of other rats, enabling
them to quickly spread information about novel food types in their group
(Galef, 2009). Cliff swallows, and other birds that nest in large groups, will
watch for other birds returning with a mouthful of insects; if they spot one,
they will follow it out to the presumably profitable hunting grounds (Brown,
1988). If information is exchanged at a nest site or colony as in this case, the
colony is referred to as information center. In all these examples, individuals
are using cues gained simply by observing behavior of other foragers; the pro-
ducers of that information do not behave in such a way as to improve the
information exchange. However, many cooperative group-living animals also
purposefully exchange information in the form of signals (Seeley, 1998):
honey bees signal the direction and distance of flower patches in the “waggle
dance” (Gruter & Farina, 2009), some ants lay pheromone trails to lead nest-
mates to food sources (Hölldobler & Wilson, 1990), other ant species teach
(facilitate learning by others at a cost to themselves) their nestmates the route
to food by leading them in a tandem run to it (Richardson et al., 2007;
Hoppitt et al., 2008). Highly cooperative groups may also improve their for-
aging efficiency in other ways, such as by employing division of labor (Powell
& Franks, 2006; Burns & Dyer, 2008); sometimes this even occurs in non-
cooperating groups (Giraldeau & Lefebvre, 1986).
FINDING FOOD : FORAGING AFFECTS ALL ASPECTS OF AN ANIMAL ’ S LIFE 13

Interestingly, both the explicit signals and the cues generated by foragers
may also be exploited by animals from other species (Coolen et al., 2003).
In addition, many organisms rely partially or completely on other species to
aid them in food acquisition—such interactions are termed mutualisms (Sachs
& Simms, 2006; Aanen & Hoekstra, 2007). For example, plants provide all
food for bees and many other pollinators (nectar and pollen) and gain pollina-
tion services in return. Plants also provide food in return for protection from
herbivores, such as to many ants, which feed on extrafloral nectaries. Corals
are colonies of polyps (cnidarian animals) that rely on their algal symbionts
to generate nutrients by photosynthesis. Such algal symbionts are also present
in several other animals and even fungi: lichens are the result of such a fungal-
algae cooperation. Cleaner fish rely on larger fish, called clients, to come to
their “cleaning station,” where the cleaner fish eats both parasites on the skin
as well as sometimes the skin of the client—a very interesting interaction with
subtle choices made by both clients and cleaners about whom to pick and
whom to pick on (Barbu et al., 2011). Such mutualisms are quite common,
and in fact all of us carry many microorganisms, which are relying on us for
food and shelter. Some of them help us digest complex foods, some crowd
out harmful bacteria (e.g., Li et al., 2008). And each animal cell contains mito-
chondria, which are often called the power plants of the cell but used to be
free-living bacteria. These were engulfed by an early eukaryotic cell, starting
a cooperation based on mutual exchange of nutrients and initiating the evolu-
tion of complex life on earth (Dyall et al., 2004).

INDIVIDUAL TASTES
We sometimes like to think that each human is particularly unique because
we all come with our own personalities, preferences, and behavioral idiosyn-
crasies. This is so; but it turns out that this is also true for probably all other
animals (Dall et al., 2012; Sih et al., 2004). Some individual fish are bolder
than others of the same species, seeming to taunt predators (Ward et al.,
2004); some spiders are more aggressive to both their mates and enemies than
others (Pruitt et al., 2008); some bumble bees will only forage on purple flow-
ers while their own sisters prefer other flower types (Raine & Chittka, 2007).
These “animal personalities” at first seemed strange to scientists: after all, there
must be one optimal strategy, and why would any individuals stray from it?
However, we now understand that personalities, stable behavioral differences
among individuals of the same species, can evolve for many reasons. For exam-
ple, herbivores need to not only ingest enough of all necessary nutrients but
also avoid accumulating toxins (practically all plants produce toxins to deter
14 ANIMAL BEHAVIOR

herbivory). This can lead to individual preferences or “tastes” being affected by

individual experience, if grazing animals need to learn which foods are palatable
or develop resistances to particular toxins (Provenza et al., 2003). Individuals
may also differ in their body size or growth rate, leading to differences in risk
taking and thus food and food-patch preferences (Stamps, 2007); this can lead
to a linkage between multiple behavioral traits, called a behavioral syndrome
(Sih et al., 2004). In addition, different individuals may specialize on particular
foods to avoid competition with others of the same species (Dall et al., 2012).

FOOD AND LIFESTYLE

Food types differ enormously in their spatial and temporal distribution.
Some, such as plant leaves (e.g., grass) are available continuously over wide
areas; others are hard to find and extinguish quickly, such as flowers that
may only offer pollen within an hour of sunrise; and yet others must be sub-
dued before they can be consumed, such as dangerous prey animals. In the
first case, the main concern of a forager will be selecting the right type of food,
such as a plant with few toxins; in the second case, a forager will focus on how
to collect information about food availability as efficiently as possible; and in
the third, a forager may need to acquire sophisticated handling or hunting
strategies. Although all of these foragers have to locate, select, and handle food,
their resulting foraging strategies will be different. They will be likely to have
sophisticated and subtle adaptations in those parts of the foraging process that
make the most difference to foraging success.
Such differences in priorities can also occur within species. For example, the
reproductive success of males in many species is limited more by how many mat-
ing partners they can acquire than by how much food they can collect. Conversely,
the reproductive success of females is rarely limited by the number of mating part-
ners because males typically can provide more sperm in a single mating than a
female can produce offspring. This has the effect that female distribution in space
is often determined by resource availability, whereas male distribution in space
may mostly depend on where females are. The different types of overlap between
male and female ranges can lead to different mating systems, such as pair bonds,
promiscuity, or harems defended by a single male (Davies & Lundberg, 1984).

EXPLAINING DIVERSITY IN ANIMAL BEHAVIOR

Honey bee foragers are parts of a highly coordinated collective foraging sys-
tem and visit hundreds of individual flowers to which they may travel half a
million times their body length (this equals 1,000 km, or 625 miles, for
FINDING FOOD : FORAGING AFFECTS ALL ASPECTS OF AN ANIMAL ’ S LIFE 15

humans) in a single trip but return home after each excursion to share their
food (Seeley, 1985). Chimpanzees may or may not share food but will copy
each other’s sophisticated techniques for extracting it. Birds may not search
for food at all but follow others to take it from them when they have located
any. Snakes and spiders often gain enough nutrients in a single meal that they
can afford to be inactive for weeks or months afterwards; hummingbirds and
bats must find food every day or risk starvation; and sloths must spend practi-
cally all their waking time eating to be able to extract enough nutrients from
their hard-to-digest food. These are examples of the amazing diversity of life-
styles and diets in animals. Behavioral ecology, as a science, attempts to
explain why this diversity exists, that is, which factors led to the evolution of
different foraging strategies in different animals (Davies et al., 2012). Experi-
ments in which the behavior of the animal or its food source is manipulated
can show which foraging strategy performs best under particular conditions.
Comparative studies, in which the foraging strategies of several species of ani-
mal are compared and related to other factors such as habitat type, body size,
and so forth, can show which of these factors appear to have affected the evo-
lution of particular foraging behaviors historically. Modeling studies can give
precise predictions of how subtle adaptations may affect foraging success.
The insights gained from this research have direct and practical applications,
for example in the management of grazing animals (Provenza et al., 2003) or
pollination of crops (Weinberg & Plowright, 2006). But we have also learned
that the beautiful variety seen in nature of animal lifestyles, cognition, and
social behavior can only be understood if we understand the foraging habits
of these species.

ACKNOWLEDGMENTS
I would like thank the NSF (grants IOS-1045239 and IOS-0841756) for
support.

REFERENCES AND SUGGESTED READING

Aanen, D. K. & R. F. Hoekstra. (2007). The evolution of obligate mutualism: If you
can’t beat ‘em, join ‘em. Trends in Ecology and Evolution, 22, 506-509.
Alanara, A., M. D. Burns, & N. B. Metcalfe. (2001). Intraspecific resource partition-
ing in brown trout: The temporal distribution of foraging is determined by
social rank. Journal of Animal Ecology, 70, 980–986.
Barbu, L., C. Guinand, R. Bergmuller, N. Alvarez, & R. Bshary. (2011). Cleaning
wrasse species vary with respect to dependency on the mutualism and behav-
ioural adaptations in interactions. Animal Behaviour, 82, 1067–1074.
16 ANIMAL BEHAVIOR

Bateson, M., B. Brilot, & D. Nettle. (2011). Anxiety: An evolutionary approach.

Canadian Journal of Psychiatry—Revue Canadienne De Psychiatrie, 56,
707–715.
Bernays, E. A. (2001). Neural limitations in phytophagous insects: Implications for
diet breadth and evolution of host affiliation. Annual Review of Entomology,
46, 703–727.
Boehm, C. (2004). What makes humans economically distinctive? A three-species
evolutionary comparison and historical analysis. Journal of Bioeconomics, 6,
109–135.
Brown, C. R. (1988). Enhanced foraging efficiency through information centers:
A benefit of coloniality in cliff swallows. Ecology, 69, 602–613.
Brown, J. S. & B. P. Kotler. (2004). Hazardous duty pay and the foraging cost of pre-
dation. Ecology Letters, 7, 999–1014.
Burd, M. & J. J. Howard. (2005). Central-place foraging continues beyond the nest
entrance: The underground performance of leaf-cutting ants. Animal Behav-
iour, 70, 737–744.
Burns, J. G. & A. G. Dyer. (2008). Diversity of speed-accuracy strategies benefits
social insects. Current Biology, 18, R953–R954.
Caraco, T., W. U. Blanckenhorn, G. M. Gregory, J. A. Newman, G. M. Recer, &
S. M. Zwicker. (1990). Risk-sensitivity: Ambient temperature affects foraging
choice. Animal Behaviour, 39, 338–345.
Careau, V., D. Thomas, M. M. Humphries, & D. Reale. (2008). Energy metabolism
and animal personality. Oikos, 117, 641–653.
Cartar, R. V. & M. V. Abrahams. (1996). Risk-sensitive foraging in a patch departure
context: A test with worker bumble bees. American Zoologist, 36, 447–458.
Castellano, S., G. Cadeddu, & P. Cermelli. (2012). Computational mate choice:
Theory and empirical evidence. Behavioural Processes, 90, 261–277.
Cerda, X., J. Retana, & S. Cros. (1998). Critical thermal limits in Mediterranean ant
species: Trade-off between mortality risk and foraging performance. Functional
Ecology, 12, 45–55.
Charnov, E. (1976). Optimal foraging: The marginal value theorem. Theoretical Pop-
ulation Biology, 9, 129–136.
Chittka, L. (2001). Camouflage of predatory crab spiders on flowers and the colour
perception of bees. Entomologia Generalis, 25, 181–187.
Chittka, L., P. Skorupski, & N. E. Raine. (2009). Speed-accuracy tradeoffs in animal
decision making. Trends in Ecology and Evolution, 24, 400–407.
Cirelli, C. & G. Tononi. (2008). Is sleep essential? PLoS Biology, 6, e216. doi:
10.1371/journal.pbio.0060216.
Coolen, I., Y. van Bergen, R. L. Day, & K. N. Laland. (2003). Species difference in
adaptive use of public information in sticklebacks. Proceedings of the Royal Soci-
ety of London, B, 270, 2413–2419.
Dafni, A. (1984). Mimicry and deception in pollination. Annual Review of Ecology
and Systematics, 15, 259–278.
FINDING FOOD : FORAGING AFFECTS ALL ASPECTS OF AN ANIMAL ’ S LIFE 17

Dall, S. R. X., A. M. Bell, D. I. Bolnick, & F. L. W. Ratnieks. (2012). An evolution-

ary ecology of individual differences. Ecology Letters, 15, 1189–1198.
Davies, N. B., J. R. Krebs, & S. A. West. (2012). An Introduction to Behavioural Ecol-
ogy. Fourth Edition. Hoboken, NJ: Wiley-Blackwell.
Davies, N. B. & A. Lundberg. (1984). Food distribution and a variable mating system
in the dunnock, Prunella modularis. Journal of Animal Ecology, 53, 895–912.
Dornhaus, A., E. J. Collins, F. X. Dechaume-Moncharmont, A. I. Houston, N. R.
Franks, & J. M. McNamara. (2006). Paying for information: Partial loads in
central place foragers. Behavioral Ecology and Sociobiology, 61, 151–161.
Dornhaus, A. & N. R. Franks. (2008). Individual and collective cognition in ants and
other insects (hymenoptera: Formicidae). Myrmecological News, 11, 215–226.
Dukas, R. (2004). Causes and consequences of limited attention. Brain Behavior and
Evolution, 63, 197–210.
Dukas, R. & S. Ellner. (1993). Information-processing and prey detection. Ecology,
74, 1337–1346.
Dunlap, A. S. & D. W. Stephens. (2009). Components of change in the evolution of
learning and unlearned preference. Proceedings of the Royal Society of London, B,
276, 3201–3208.
Dyall, S. D., M. T. Brown, & P. J. Johnson. (2004). Ancient invasions: From endo-
symbionts to organelles. Science, 304, 253–257.
Fawcett, T., S. Hamblin, & L.-A. Giraldeau. (2012a). We can study how mechanisms
evolve without knowing the rules of chess or the workings of the brain. Behav-
ioral Ecology, 24, 14–15.
Fawcett, T. W., S. Hamblin, & L.-A. Giraldeau. (2012b). Exposing the behavioral gam-
bit: The evolution of learning and decision rules. Behavioral Ecology, 24, 2–11.
Fisher, J. & R. A. Hinde. (1949). The opening of milk bottles by birds. British Birds,
42, 347–357.
Fortin, D., H. L. Beyer, M. S. Boyce, D. W. Smith, T. Duchesne, & J. S. Mao.
(2005). Wolves influence elk movements: Behavior shapes a trophic cascade
in Yellowstone National Park. Ecology, 86, 1320–1330.
Fouad, K., F. Libersat, & W. Rathmayer. (1996). Neuromodulation of the escape
behavior of the cockroach Periplaneta americana by the venom of the parasitic
wasp Ampulex compressa. Journal of Comparative Physiology A: Sensory Neural
and Behavioral Physiology, 178, 91–100.
Galef, B. G. (2009). Strategies for social learning: Testing predictions from formal
theory. Advances in the Study of Behavior, 39, 117–151.
Galef, B. G. & L.-A. Giraldeau. (2001). Social influences on foraging in vertebrates:
Causal mechanisms and adaptive functions. Animal Behaviour, 61, 3–15.
Galleguillos, C. & S. Belongie. (2010). Context based object categorization: A critical
survey. Computer Vision and Image Understanding, 114, 712–722.
Garamszegi, L. Z., A. P. Møller, & J. Erritzoe. (2002). Coevolving avian eye size and
brain size in relation to prey capture and nocturnality. Proceedings of the Royal
Society of London, B, 269, 961–967.
18 ANIMAL BEHAVIOR

Gigord, L. D. B., M. R. Macnair, M. Stritesky, & A. Smithson. (2002). The potential

for floral mimicry in rewardless orchids: An experimental study. Proceedings of
the Royal Society of London, B, 269, 1389–1395.
Giraldeau, L.-A. & L. Lefebvre. (1986). Exchangeable producer and scrounger roles
in a captive flock of feral pigeons: A case for the skill pool effect. Animal Behav-
iour, 34, 797–803.
Giraldeau, L.-A., C. Soos, & G. Beauchamp. (1994). A test of the producer-scrounger
foraging game in captive flocks of spice finches, Lonchura punctulata. Behav-
ioral Ecology and Sociobiology, 34, 251–256.
Gould, S. J. & R. C. Lewontin. (1979). The spandrels of San Marco and the Pan-
glossian paradigm: A critique of the adaptationist programme. Proceedings of
the Royal Society of London, B, 205, 581–598.
Gruter, C. & W. M. Farina. (2009). The honeybee waggle dance: Can we follow the
steps? Trends in Ecology and Evolution, 24, 242–247.
Hays, G. C., V. J. Hobson, J. D. Metcalfe, D. Righton, & D. W. Sims. (2006). Flex-
ible foraging movements of leatherback turtles across the North Atlantic
Ocean. Ecology, 87, 2647–2656.
Herbers, J. M. (1981). Time resources and laziness in animals. Oecologia, 49, 252–262.
Hodges, C. M. (1985). Bumble bee foraging: Energetic consequences of using a
threshold departure rule. Ecology, 66, 188–197.
Hölldobler, B. & E. O. Wilson. (1990). The Ants. Cambridge, MA: Harvard Univer-
sity Press.
Hoppitt, W. J. E., G. R. Brown, R. Kendal, L. Rendell, A. Thornton, M. M. Web-
ster, & K. N. Laland.. (2008). Lessons from animal teaching. Trends in Ecology
and Evolution, 23, 486–493.
Houston, A. I. (2009). Flying in the face of nature. Behavioural Processes, 80,
295–305.
Houston, A. I., J. M. McNamara, & M. D. Steer. (2007). Do we expect natural selec-
tion to produce rational behaviour? Philosophical Transactions of the Royal Soci-
ety of London, B, 362, 1531–1543.
Hutchinson, J. M. C., J. M. McNamara, & I. C. Cuthill. (1993). Song, sexual selec-
tion, starvation and strategic handicaps. Animal Behaviour, 45, 1153–1177.
Kacelnik, A. (1993). Leaf-cutting ants tease optimal foraging theorists. Trends in Ecol-
ogy and Evolution, 8, 346–348.
Kamil, A. C. & H. L. Roitblat. (1985). The ecology of foraging behavior: Implica-
tions for animal learning and memory. Annual Review of Psychology, 36,
141–169.
Kleineidam, C. J., M. Ruchty, Z. A. Casero-Montes, & F. Roces. (2007). Thermal
radiation as a learned orientation cue in leaf-cutting ants (Atta vollenweideri).
Journal of Insect Physiology, 53, 478–487.
Krutzen, M., J. Mann, M. R. Heithaus, R. C. Connor, L. Bejder, & W. B. Sherwin.
(2005). Cultural transmission of tool use in bottlenose dolphins. Proceedings of
the National Academy of Sciences, USA, 102, 8939–8943.
Laland, K. N. (2004). Social learning strategies. Learning and Behavior, 32, 4–14.
FINDING FOOD : FORAGING AFFECTS ALL ASPECTS OF AN ANIMAL ’ S LIFE 19

Leadbeater, E. & L. Chittka. (2007). The dynamics of social learning in an insect

model, the bumblebee (Bombus terrestris). Behavioral Ecology and Sociobiology,
61, 1789–1796.
Lefebvre, L. (1995). The opening of milk bottles by birds: Evidence for accelerating
learning rates, but against the wave-of-advance model of cultural transmission.
Behavioural Processes, 34, 43–53.
Lesku, J. A., T. C. Roth II, N. C. Rattenborg, C. J. Amlaner, & S. L. Lima. (2009).
History and future of comparative analyses in sleep research. Neuroscience and
Biobehavioral Reviews, 33, 1024–1036.
Li, M., B. Wang, M. Zhang, M. Rantalainen, S. Wang, H. Zhou, Y. Zhang, J.
Shen, X. Pang, M. Zhang, H. Wei, Y. Chen, H. Lu, J. Zuo, M. Su, Y.
Qiu, W. Jia, C. Xiao, L. M. Smith, S. Yang, E. Holmes, H. Tang, G. Zhao,
J. K. Nicholson, L. Li, & L. Zhao. (2008). Symbiotic gut microbes modulate
human metabolic phenotypes. Proceedings of the National Academy of Sciences,
USA, 105, 2117–2122.
Libersat, F. (2003). Wasp uses venom cocktail to manipulate the behavior of its cock-
roach prey. Journal of Comparative Physiology A: Neuroethology Sensory Neural
and Behavioral Physiology, 189, 497–508.
Libersat, F., A. Delago, & R. Gal. (2009). Manipulation of host behavior by parasitic
insects and insect parasites. Annual Review of Entomology, 54, 189–207.
Lynn, S. K., J. Cnaani, & D. R. Papaj. (2005). Peak shift discrimination learning as a
mechanism of signal evolution. Evolution, 59, 1300–1305.
McArthur, C., P. Orlando, P. B. Banks, J. S. Brown. (2012). The foraging tightrope
between predation risk and plant toxins: A matter of concentration. Functional
Ecology, 26, 74–83.
McIntosh, M. E. (2005). Pollination of two species of ferocactus: Interactions
between cactus-specialist bees and their host plants. Functional Ecology, 19,
727–734.
McNamara, J. M. & A. I. Houston. (2009). Integrating function and mechanism.
Trends in Ecology and Evolution, 24, 670–675.
Morgan, K. V., T. A. Hurly, M. Bateson, L. Asher, & S. D. Healy. (2012). Context-
dependent decisions among options varying in a single dimension. Behavioural
Processes, 89, 115–120.
Morris, W. F., W. G. Wilson, J. L. Bronstein, & J. H. Ness. (2005). Environmental
forcing and the competitive dynamics of a guild of cactus-tending ant mutual-
ists. Ecology, 86, 3190–3199.
Nathan, R., W. M. Getz, E. Revilla, M. Holyoak, R. Kadmon, D. Saltz, & P. E.
Smouse. (2008). A movement ecology paradigm for unifying organismal
movement research. Proceedings of the National Academy of Sciences, USA,
105, 19052–19059.
Nettle, D. & M. Bateson. (2012). The evolutionary origins of mood and its disorders.
Current Biology, 22, R712–R721.
Nonacs, P. & L. Dill. (1993). Is satisficing an alternative to optimal foraging theory?
Oikos, 67, 371–375.
20 ANIMAL BEHAVIOR

Núñez, J. A. (1982). Honeybee foraging strategies at a food source in relation to its

distance from the hive and the rate of sugar flow. Journal of Apicultural
Research, 21, 139–150.
Ohashi, K., J. D. Thomson, & D. D’Souza. (2007). Trapline foraging by bumble
bees: IV. Optimization of route geometry in the absence of competition.
Behavioral Ecology, 18, 1–11.
Orians, G. H. & N. E. Pearson. (1979). On the theory of central place foraging. In
D. J. Horn, R. D. Mitchell, & G. R. Stairs (eds.), Analyses of Ecological Systems
(pp. 154–177). Columbus, OH: Ohio State University Press.
Pennisi, E. (1999). Human evolution: Did cooked tubers spur evolution of big
brains? Science, 283, 2004–2005.
Powell, S. & N. R. Franks. (2006). Ecology and the evolution of worker morphologi-
cal diversity: A comparative analysis with Eciton army ants. Functional Ecology,
20, 1105–1114.
Provenza, F. D., J. J. Villalba, L. E. Dziba, S. B. Atwood, & R. E. Banner. (2003).
Linking herbivore experience, varied diets, and plant biochemical diversity.
Small Ruminant Research, 49, 257–274.
Pruitt, J. N., S. E. Riechert, & T. C. Jones. (2008). Behavioural syndromes and their
fitness consequences in a socially polymorphic spider, Anelosimus studiosus.
Animal Behaviour, 76, 871–879.
Pyke, G. H., H. R. Pulliam, & E. L. Charnov. (1977). Optimal foraging: A selective
review of theory and tests. Quarterly Review of Biology, 52, 137–154.
Raine, N. E. & L. Chittka. (2007). Flower constancy and memory dynamics in bum-
blebees (hymenoptera: Apidae: Bombus). Entomologia Generalis, 29, 179–199.
Raine, N. E., T. C. Ings, A. Dornhaus, N. Saleh, & L. Chittka. (2006). Adaptation,
genetic drift, pleiotropy, and history in the evolution of bee foraging behavior.
Advances in the Study of Behavior, 36, 305–354.
Richardson, T. O., P. A. Sleeman, J. M. McNamara, A. I. Houston, & N. R. Franks.
(2007). Teaching with evaluation in ants. Current Biology, 17, 1520–1526.
Sabbah, S., R. L. Laria, S. M. Gray, C. W. Hawryshyn. (2010). Functional diversity
in the color vision of cichlid fishes. BMC Biology, 8, 133. doi: 10.1186/
1741-7007-8-133
Sachs, J. L. & E. L. Simms. (2006). Pathways to mutualism breakdown. Trends in
Ecology and Evolution, 21, 585–592.
Schmid-Hempel, P., A. Kacelnik, & A. I. Houston. (1985). Honeybees maximize
efficiency by not filling their crop. Behavioral Ecology and Sociobiology, 17,
61–66.
Schuck-Paim, C., L. Pompilio, & A. Kacelnik. (2004). State-dependent decisions
cause apparent violations of rationality in animal choice. PLoS Biology, 2,
e402. doi: 10.1371/journal.pbio.0020402.
Seeley, T. D. (1985). Honeybee Ecology: A Study of Adaptation in Social Life. Prince-
ton, NJ: Princeton University Press.
Seeley, T. D. (1998). Thoughts on information and integration in honey bee colo-
nies. Apidologie, 29, 67–80.
FINDING FOOD : FORAGING AFFECTS ALL ASPECTS OF AN ANIMAL ’ S LIFE 21

Seth, A. K. (2007). The ecology of action selection: Insights from artificial life. Philo-
sophical Transactions of the Royal Society of London, B, 362, 1545–1558.
Shafir, S., T. A. Waite, & B. H. Smith. (2002). Context-dependent violations of
rational choice in honeybees (Apis mellifera) and gray jays (Perisoreus canaden-
sis). Behavioral Ecology and Sociobiology, 51, 180–187.
Sih, A., A. Bell, & J. C. Johnson. (2004). Behavioral syndromes: An ecological and
evolutionary overview. Trends in Ecology and Evolution, 19, 372–377.
Sih, A. & B. Christensen. (2001). Optimal diet theory: When does it work, and when
and why does it fail? Animal Behaviour, 61, 379–390.
Sims, D. W., E. J. Southall, N. E. Humphries, G. C. Hays, C. J. A. Bradshaw, J. W.
Pitchford, A. James, M. Z. Ahmed, A. S. Brierley, M. A. Hindell, D. Morritt,
M. K. Musyl, D. Righton, E. L. C. Shepard, V. J. Wearmouth, R. P. Wilson,
M. J. Witt, & J. D. Metcalfe. (2008). Scaling laws of marine predator search
behaviour. Nature, 451, 1098–1102.
Skelhorn, J. & C. Rowe. (2010). Birds learn to use distastefulness as a signal of toxic-
ity. Proceedings of the Royal Society of London, B, 277, 1729–1734.
Smeulders, A. W. M., M. Worring, S. Santini, A. Gupta, & R. Jain. (2000). Content-
based image retrieval at the end of the early years. IEEE Transactions on Pattern
Analysis and Machine Intelligence, 22, 1349–1380.
Sol, D., S. Timmermans, & L. Lefebvre. (2002). Behavioural flexibility and invasion
success in birds. Animal Behaviour, 63, 495–502.
Stamps, J. A. (2007). Growth-mortality tradeoffs and “personality traits” in animals.
Ecology Letters, 10, 355–363.
Stearns, S. C. & P. Schmid-Hempel. (1987). Evolutionary insights should not be
wasted. Oikos, 49, 118–125.
Stephens, D., J. S. Brown, & R. C. Ydenberg. (2007). Foraging: Behavior and Ecology.
Chicago: University of Chicago Press.
Stephens, D. W., B. Kerr, & E. Fernández-Juricic. (2004). Impulsiveness without dis-
counting: The ecological rationality hypothesis. Proceedings of the Royal Society
of London, B, 271, 2459–2465.
Stevens, J. R., E. V. Hallinan, & M. D. Hauser. (2005). The ecology and evolution of
patience in two new world monkeys. Biology Letters, 1, 223–226.
Stoddard, M. C. (2012). Mimicry and masquerade from the avian visual perspective.
Current Zoology, 58, 630–648.
Thomas, R. J. (1999). Two tests of a stochastic dynamic programming model of daily
singing routines in birds. Animal Behaviour, 57, 277–284.
Thomson, J. D. (1996). Trapline foraging by bumblebees: I. Persistence of flight-path
geometry. Behavioral Ecology, 7, 158–164.
van Schaik, C. P., M. Ancrenaz, G. Borgen, B. Galdikas, C. D. Knott, I. Singleton, A.
Suzuki, S. S. Utami, & M. Merrill. (2003). Orangutan cultures and the evolu-
tion of material culture. Science, 299, 102–105.
Varju, D. & J. Núñez. (1991). What do foraging honeybees optimize? Journal of
Comparative Physiology A: Sensory, Neural, and Behavioral Physiology, 169,
729–736.
22 ANIMAL BEHAVIOR

Varju, D. & J. Núñez. (1993). Energy balance versus information exchange in forag-
ing honeybees. Journal of Comparative Physiology A: Sensory, Neural, and
Behavioral Physiology, 172, 257–261.
Wajnberg, E. (2012). Multi-objective behavioural mechanisms are adopted by forag-
ing animals to achieve several optimality goals simultaneously. Journal of Ani-
mal Ecology, 81, 503–511.
Ward, A. J. W., P. Thomas, P. J. B. Hart, & J. Krause. (2004). Correlates of boldness
in three-spined sticklebacks (Gasterosteus aculeatus). Behavioral Ecology and
Sociobiology, 55, 561–568.
Waser, N. M., L. Chittka, M. V. Price, N. Williams, & J. Ollerton. (1996). Generali-
zation in pollination systems, and why it matters. Ecology, 77, 1043–1060.
Weimerskirch, H. (1998). How can a pelagic seabird provision its chick when relying
on a distant food resource? Cyclic attendance at the colony, foraging decision
and body condition in sooty shearwaters. Journal of Animal Ecology, 67,
99–109.
Weinberg, D. & C. M. S. Plowright. (2006). Pollen collection by bumblebees
(Bombus impatiens): The effects of resource manipulation, foraging experience
and colony size. Journal of Apicultural Research, 45, 22–27.
Whiten, A., J. Goodall, W. C. McGrew, T. Nishida, V. Reynolds, Y. Sugiyama, C. E.
G. Tutin, R. W, Wrangham, & C. Boesch. (1999). Cultures in chimpanzees.
Nature, 399, 682–685.
Wolf, T. J. & P. Schmid-Hempel. (1989). Extra loads and foraging life span in
honeybee workers. Journal of Animal Ecology, 58, 943–954.
Wrangham, R. W. (1993). The evolution of sexuality in chimpanzees and bonobos.
Human Nature: An Interdisciplinary Biosocial Perspective, 4, 47–79.
Ydenberg, R. & P. Hurd. (1998). Simple models of feeding with time and energy
constraints. Behavioral Ecology, 9, 49–53.
Ydenberg, R. C., C. V. J. Welham, R. Schmid-Hempel, P. Schmid-Hempel, & G.
Beauchamp. (1994). Time and energy constraints and the relationships
between currencies in foraging theory. Behavioral Ecology, 5, 28–34.
 2

Predation and Antipredator Behavior

Theodore Stankowich

INTRODUCTION: THE DANGER LURKING

It is dusk on the coast of northern California, and a five-year-old adult female
black-tailed deer quietly drops her head in a clearing of short coastal grass to feed
on small, soft, green shoots. The wind is blowing gently from the west off the
ocean, and the air is calm and quiet. As the dark of night approaches, she gradu-
ally diverts more attention to her surroundings, aware of the increasing danger.
Meanwhile a lone cougar picks up a faint scent of deer and quietly begins to
make his way through the brush towards its sources. He steps quietly between
two rows of bushes that line a grassy clearing when the deer comes into view about
40 meters away—he immediately freezes. The cougar stares at the deer with inter-
est, but he’s a bit too far away to attack just yet. He needs to get closer—much
closer. He tries to survey the clearing for a better spot to launch an attack—it
has to be downwind of her so she won’t be able to smell him coming but close
enough to capture her quickly. When he spots a large clump of brush about
10 meters from her, he starts to creep his way quietly through the brush. Suddenly,
a crackle of dead sticks and leaves under his paw pierces the night air and catches
the attention of the deer, causing her to raise her head to investigate. He freezes.
She sniffs the air for any signs of danger. Staring into the bushes, bits and
pieces of a lurking figure are visible—a patch of tawny fur, a tail tip, then a
pair of eyes stare in her direction from a gap in the leaves. It’s unclear what
kind of animal it is. If it’s a cougar it’s far enough away that she would be easily
able to flee an attack. A coyote, on the other hand, could be one of many in
24 ANIMAL BEHAVIOR

the area, making escape a difficult proposition. Becoming visibly agitated, she
expels air through her nose loudly, stomping her right foreleg on the earth.
Her tail flags back and forth. The heads of four other deer in the vicinity
pop up from the vegetation—two adult females and their fawns. Cuing in to
what this deer is staring at, they join in, staring into the bushes, snorting,
and footstamping. She takes a few steps toward the bushes to try to get a better
look, stomping and snorting frequently.
The noise and movement of the deer have made it clear to the cougar that
he has been spotted and has lost a possible meal. With his quarry on high alert,
there is no way to use stealth to get close enough to have any chance of a suc-
cessful attack. The shadowy figure moves away and disappears deep into the
brush.
After several minutes, the collective snorts become softer and the deer
begins to move off to forage in a nearby wooded area that might offer more
concealment. She has survived another evening but will probably avoid this
area for a while.
This vignette of a dangerous encounter between a black-tailed deer (Odo-
coileus hemionus hemionus) and a cougar (Puma concolor), while somewhat
anthropomorphic and dramatized, illustrates the dynamic interplay of behav-
ioral decisions that both predators and prey must make when they encounter
each other. Very few encounters result in attacks and captures by predators,
so a large volume of literature has examined the roles of behavioral decision
making and morphological adaptations that help prey reduce the probability
of being captured by predators and help predators increase the probability of
successfully acquiring a meal.
Several behavioral scientists have developed frameworks for conceptualizing
the diversity and functions of antipredator behaviors. Steve Lima and Law-
rence Dill (1990) developed an attack sequence for predator-prey encounters
that compartmentalizes behavioral decisions that each party must make at
each step, resulting in a probability of death given an encounter situation.
A modified form of this flow chart appears in Figure 2.1a. Briefly, they
divide the interaction into four main steps: encounter, detection, attack, and
capture/escape. Because there are many possible nonlethal outcomes of the
interaction, often only a small proportion of all encounters lead to prey death.
This is a useful framework to build from because it nicely illustrates the differ-
ent opportunities each party has to either halt the sequence from leading to
death (in the case of the prey) or push the sequence forward toward prey death
(in the case of the predator), and there are many behavioral decisions and
morphological adaptations that help both parties in their cause. To this end,
R. Brian Langerhans (2007) presented a different type framework for studying
PREDATION AND ANTIPREDATOR BEHAVIOR 25

Figure 2.1a. Frameworks for understanding predator-prey interactions. Redrawn and

modified substantially from Lima and Dill (1990), this flow chart shows the possible
outcomes of an “encounter situation” between a prey and its predator, emphasizing
that only a proportion of all encounters lead to the death of the prey through con-
sumption or fatal injury. (Copyright 1990 Canadian Science Publishing. Used by
permission.)

such antipredator traits in terms of how they affect prey fitness (Table 2.1).
His brief review of antipredator adaptations examines the myriad of ways ani-
mals seek to avoid encountering predators, avoid a predator’s detection, and
deter attacks and consumption by predators.
In this chapter, I modify the Lima and Dill (1990) sequence, adding several
new intermediate steps, and integrate Langerhans’s (2007) trait-based
approach to develop a comprehensive framework that demonstrates the
opportunities for natural selection to shape both prey and predator traits that
enhance fitness over the course of the predator-prey encounter sequence. In
Figure 2.1b, recognition has been added to the detection phase to emphasize
the fact that often an animal may detect the presence of another animal
26 ANIMAL BEHAVIOR

Figure 2.1b. Framework used in this chapter to demonstrate the myriad of ways that
natural selection can shape assessment, behavioral decisions, and morphology in
predators and prey.

in their area but still lack enough information to determine if it a dangerous or

innocuous species. I have also added an assessment/communication phase
before the attack/flight phase to account for incidences where the prey judges
how much of a risk this particular individual poses to its safety, when the
predator weighs the costs and benefits of attack and the likelihood of success,
and when there is overt exchange of signals or cues between the parties that
allow them to assess each other more rapidly. Finally, I have added the option
of defense to the final stage to capture the cases where prey have formidable
behavioral or morphological defenses with which to repel predators during
physical combat or to deter them prior.
Using this general framework, my goal in this chapter is to give a brief over-
view of the many directions that predator-prey interactions can take given a
PREDATION AND ANTIPREDATOR BEHAVIOR 27

Table 2.1. Redrawn and edited from Langerhans (2007), this is a framework for
understanding how antipredator traits affect direct fitness and survivorship,
organized by the chronological stage of the predatory encounter in which they are
employed. Some prey traits listed in the table are not discussed in this chapter.
ASF: avoiding a predator’s sensory field; DSF: avoiding detection within a preda-
tor’s sensory field; ATD: attack deterrence; CPD: capture deterrence; CND:
consumption deterrence.
How the antipredator trait directly enhances
individual survivorship and fitness

Prey trait Avoidance Escape

Activity level ASF

Crypsis DSF
Development time ASF
Use predator-free habitat ASF
Active defense CPD, CND
Aposematism ATD
Attack diversion CPD
Autotomy CPD
Chemical defense ATD, CPD, CND
Death feigning ATD
Deimatic behavior ATD, CPD
Mimicry ATD
Protective morphologies ATD, CPD, CND
Rapid retreat, protean CPD
behavior
Grouping ASF, DSF ATD, CPD, CND
Use protective habitat ASF, DSF ATD, CPD, CND
Vigilance ASF, DSF ATD, CPD, CND

single interaction. From the perspective of each party, the opponent is not a
static, predictable automaton, but instead each party alters its behavior based
on its opponent’s behavior, the environmental situation, its own physical sta-
tus, and its past experience with its opponent’s species. New and important
model systems of predator-prey study will be highlighted throughout the
chapter to illustrate the past, present, and future of this exceptionally vibrant
field of study. Finally, it should become clear by the end of this chapter that
future studies of predator-prey interaction should focus on the behavioral
28 ANIMAL BEHAVIOR

decisions of both parties either simultaneously in natural interactions when

possible or in dual studies of each party responding to dynamic controllable
models of their opponent.

ENCOUNTER
In order for the attack sequence to even begin and for there to be any pos-
sibility of death for an animal, the predator and prey must come near enough
to each other for the possibility of detection to occur. From the perspective of
prey, they behave to minimize the probability that they will encounter poten-
tial predators, while hungry predators will behave to maximize this possibility.
Strategies vary from selecting permanent home ranges or habitat types with
more prey or fewer predators, to coordinated hunting or foraging trips by
predators, to, as is detailed below, daily or seasonal migrational patterns.
Many species move around in their habitat to change the probability of
coming within a detectable distance of predators or prey, and these cyclic
migrations may occur daily or seasonally. Eduardo Arraut and colleagues
(2010) showed that male Amazonian manatees (Trichechus inunguis) spend
the summer high-water season in deeper várzea lakes to take advantage of
abundant foraging opportunities but migrate to rivers when the water level
in lakes falls during the low-water season. Low water levels in lakes increase
the risk of predation by jaguars (Panthera onca), caimans, and humans, and
rivers offer fewer predators but also fewer foraging opportunities. Diel (daily)
vertical migration (DVM) in pelagic environments is a common pattern of
migration that has been associated with increased foraging opportunities and
decreased probability of predation. Many studies have shown that zooplank-
ton move to deeper waters during the daylight hours and return to shallower
depths at night in order to avoid predatory fish that rely on vision to detect
prey: capture rates in shallow illuminated waters are far greater than in darker
environments (Zaret & Suffern, 1976). DVM, however, does occur in meso-
predatory fish (midlevel predators) as well. Mark Scheuerell and Daniel
Schindler (2003) found that juvenile sockeye salmon (Oncorhynchus nerka)
showed similar DVM patterns, but there was a lack of DVM in their zoo-
plankton prey. Zooplankton in the Alaskan lakes they studied remained at
the well-lit surface of the lake throughout the day, while the salmon dove to
greater depths during the day (Figure 2.2), maintaining a constant-light envi-
ronment, thereby allowing them to reduce the odds of visual detection by
larger piscivorous (fish-eating) fish, in which DVM were not detected.
Simply changing activity rates and schedules may affect the likelihood that
predators and prey will encounter each other in a meaningful way. For
PREDATION AND ANTIPREDATOR BEHAVIOR 29

Figure 2.2. Box plot indicates the depth distribution of predatory fish of Nerka Lake
(from Scheurell & Schnidler, 2003). Black dots show the time series of depth of
juvenile sockeye salmon (Oncorhynchus nerka). The solid black line shows the log of
light intensity at the surface of the water. Salmon exhibit DVM patterns, rising to the
surface during the night when incident light is minimal to maintain a constant low-
light environment when in the presence of potential predators, which stay near the
surface. (Copyright 2003 Ecological Society of America)

terrestrial mammals and birds, nocturnality and diurnality offer different

advantages and disadvantages. For prey, most large mammalian carnivores
are nocturnal hunters, so being active during the day minimizes the likelihood
of encountering them. However, most birds of prey are diurnal, and if you are
small enough to be captured and killed by a large predatory bird, the daytime
offers tremendous danger. Shifts in activity can be paired with migration pat-
terns to add further control of one’s susceptibility to predation or likelihood of
encountering prey (e.g., DVM). Other physiological and foraging considera-
tions (e.g., thermoregulation, availability of food) contribute to the decision
of when to be active, and many species hedge their bets by staying close to
sources of refuge that can offer protection should a predator enter the area.
30 ANIMAL BEHAVIOR

Burrows, dense vegetation, shells, rocks, and so forth all offer outstanding pro-
tection from predators and reduce a prey animal’s detectability.
Finally, living or hunting in groups may affect an individual’s probability of
encountering a predator or prey animal. For prey, if aggregations of prey are
no more detectable than singletons, then predators may suffer increased search
times between encounters with prey (Krause & Ruxton, 2002). Often groups
of predators can increase their odds of detecting faint cues of distant prey ani-
mals. If one member of the group picks up on the scent of potential prey,
others that did not detect the scent may benefit by following that trail and,
in general, increasing the probability of encountering prey.

DETECTION AND RECOGNITION

Once prey and predator have encountered each other in time and space so
that they are potentially detectable to each other, actual detection of the other
party is not a foregone conclusion. There are multiple morphological (e.g.,
camouflage body shape and coloration) and behavioral (e.g., freezing) adapta-
tions used to minimize the likelihood of detection, and there are sensory
modalities both prey and predators use (e.g., binocular versus lateral vision,
olfactory cues, electroreception, echolocation, vibrations) to maximize the
likelihood of detecting the other party. From the above discussion of sockeye
salmon (Oncorhynchus nerka) maintaining a constant-light environment to
minimize detection by larger predators, it is clear that the distinction between
adaptations that influence either the encounter or detection phase are “fuzzy”
at best, but it is helpful to make a distinction between being in the same place
at the same time so detection of another organism is even possible and actually
detecting that other organism.

Detection of Other Organisms

Animals use every sensory modality to detect other organisms in their envi-
ronment. While they are typically more useful for long-distance detection of
organisms in the surrounding area, olfactory cues from urine, scent markings,
and wind-blown odors of hidden organisms are helpful for detecting the pres-
ence of potential predators or prey in the immediate vicinity as well. For ter-
restrial organisms, the most commonly relied-on modality for detection of
predators and prey is vision, and visual systems tend to evolve to suit both
predator detection and foraging needs (e.g., predators often form search
images, giving them the heightened ability to detect a prey item; see Chapter 1).
Esteban Fernández-Juricic and his collaborators study the components of
the avian visual system that aid in detection and how predatory species and
PREDATION AND ANTIPREDATOR BEHAVIOR 31

prey species differ in their visual abilities. Typically predatory species have
high acuity and larger binocular fields while prey have wider lateral areas to
detect predators from all around them and lower visual acuity. Recent research
shows that, for example, white-crowned sparrows (Zonotrichia leucophrys) and
California towhees (Pipilo crissalis) are both prey species that have wider lateral
visual fields and faster head-movement rates (a proxy for scanning behavior)
than predatory species to enhance detection ability (Fernández-Juricic et al.,
2011). However, these prey species show wider binocular fields than their aer-
ial avian predators, which may be related to their ability to detect prey (e.g.,
seeds and small insects) at close distances from the beak. Further, studying vis-
ual morphology in concert with behavioral responses to predators and conspe-
cifics allows these ornithologists to understand (1) trade-offs between
opposing sensory needs (distant predator detection from all directions versus
binocular vision for detecting and manipulating food on the ground) and (2)
how animals can behaviorally compensate for differences in visual ability due
to morphology (towhees have lower spatial resolution on the periphery of
the retina than white-crowned sparrows and, therefore, have faster sideways
head movements to compensate). Similarly, visual-field configuration and
scanning behavior in birds of prey is matched to prey-searching strategies
and environmental obstructions: red-tailed hawks (Buteo jamaicensis), which
scan for prey in open habitats from high perches, have smaller binocular areas,
a lower degree of eye movement, and make fewer head movements than
Cooper’s hawks (Accipiter cooperi), which ambush prey through brush in for-
ested areas (O’Rourke, Hall, et al., 2010; O’Rourke, Pitlik, et al., 2010).
Finally, increased prey group size can enhance detection speed (Krause &
Ruxton, 2002; Pulliam, 1973; Treherne & Foster, 1980). This is most evident
in species that respond behaviorally immediately upon detection of a predator,
and individuals in these larger groups typically have reduced individual rates of
vigilance (Cresswell, 1994).

Camouflage: Avoiding Detection

While both predator and prey have an interest in detecting the other as
quickly as possible, both parties also seek to avoid being detected by the other,
and there are multiple behavioral and morphological adaptations that have
evolved to make organisms less detectable to other species. There are multiple
forms of camouflage (Endler, 1981; Ruxton et al., 2004). Background match-
ing is simply having a coloration that blends in with the background habitat
(e.g., vegetation, soil) and is both common and effective. Masquerade, a vari-
ant of background matching, occurs when the organism is detectable but
resembles an inedible object, such as a leaf swaying in the wind or a stick. This
32 ANIMAL BEHAVIOR

strategy is particularly common in terrestrial and aquatic arthropods that have

exoskeletons that can be easily altered in shape and color by natural selection.
Disruptive coloration is a phenomenon similar to background matching
where animals use bold blocks of color or projections of the body to break
up their body outline against the background environment. Cuttlefish provide
a striking example of an animal capable of changing its coloration to either
match the background or be disruptive. Another camouflage strategy particu-
larly common in vertebrates is countershading, where the animal’s back is
darker in color than its belly. Because light usually comes from above the ani-
mal, the darker shadow beneath the animal is compensated for by the lighter
coloration, making the animal appear less tubular in shape if viewed from
the side. This strategy is particularly effective in aquatic animals because, when
viewed from above, their darker back helps blend in with the dark depths of
the water, and when viewed from below, the lighter belly helps blend in with
the light coming from above the water’s surface. Transparency, the dominant
form of camouflage in invertebrates that live in the open water where there are
no sources of refuge and no background surfaces to blend into, can be found
in some jellyfish, comb jellies, and larval fish.
One very specialized alternative to traditional camouflage is called aggres-
sive mimicry, where the predator mimics some signal that either is benign to
the prey so they do not avoid it or is attractive to the prey (i.e., a lure) so the
prey approaches the predator. Deep-sea anglerfish live in dark ocean environ-
ments and use a bioluminescent fleshy appendage that hangs in front of their
head to attract unsuspecting fish toward their mouth. Therefore, with aggres-
sive mimicry, the predator avoids detection by diverting the target’s attention
towards some attractive object, allowing it to close the distance between itself
and the prey without ever moving.
In addition to these morphological adaptations, many species have behav-
ioral tactics to help avoid detection. Probably the most basic example is still-
ness. Motion against the background environment is typically very easy to
detect. Many predators remain motionless for long periods to avoid detec-
tion by their prey, and many species will remain motionless when they are
not required to feed or move through their environment in order to mini-
mize the chances of predators detecting their movements. Further, as an
alternative to fleeing, many species will freeze if they have been moving
when they detect a predator or prey (David, 2005). Young white-tailed deer
(Odocoileus virginianus), for example, will freeze more often than adults, and
freezing is more common in vegetative cover (Smith, 1991). Finally, mas-
querading species often accentuate their ruse by mimicking the natural
movements of the objects they resemble. Mantids and leaf-like insects often
PREDATION AND ANTIPREDATOR BEHAVIOR 33

sway their bodies from side to side to resemble foliage moving naturally in
the wind.
The vast majority of both predator and prey species adopt some form of the
above coloration to avoid detection, but some have evolved morphologies that
make them more detectable. Often, greater detectability results from colora-
tion that makes them more attractive to potential mates; many others are
boldly colored to advertise their defensive abilities to predators, a phenome-
non known as aposematism. I will discuss this topic in detail later.

Recognition
Once detection has occurred, each party must also recognize the other as
potential prey, irrelevant, a conspecific, or a potential threat. Recognition after
detection is not a foregone conclusion: I have personally witnessed a black-
tailed deer stare at a model leopard (extinct in the study site for more than
600,000 years) from 10 meters away without concern or recognition of danger
as she chewed her food (Stankowich & Coss, 2007). There is a great literature
on neural mechanisms governing recognition of objects as dangerous or poten-
tially edible. Both processes are the product of inherited recognition patterns
that evoke adaptive alert responses and plastic learning capabilities that allow
for changes in local predator composition and temporal or spatial variation
in prey availability. Richard Coss (Coss, 1999, 2010; Coss & Ramakrishnan,
2000; Stankowich & Coss, 2007) and Eberhard Curio (Curio, 1975, 1993)
have studied the evolution and persistence of predator recognition in a variety
of mammals (ground squirrels, black-tailed deer, and bonnet macaques
[Macaca radiata]) and in birds (pied flycatchers [Ficedula hypoleuca]), respec-
tively. In some prey species, perceptual capabilities are integrated with
higher-order inferences of predator hunting strategies in different environ-
ments (Coss, 2010). Predators and prey use several different sensory modal-
ities to recognize unique biologically important patterns and features of their
opponent above background noise. Olfactory cues in the form of alarm odors
of depredated conspecifics (e.g., Schreckstoff) can be a potent cue for many
aquatic species (Kats & Dill, 1998). Many terrestrial species are able to
winnow out extraneous sounds and remain sensitive to relevant sounds of
predators and prey, and countless studies across the animal kingdom have
shown that common biological patterns (e.g., rosettes or spots, forward-
facing eyes of predators) are potent recognition cues (Coss & Goldthwaite,
1995; Coss & Ramakrishnan, 2000; Coss et al., 2005; Curio, 1975). Recogni-
tion of some cues can be heritable so that recognition of evolutionarily con-
stant predators or prey happens innately, or selection may also favor the
ability to learn to recognize morphological characteristics in particular settings
34 ANIMAL BEHAVIOR

when encounters occur in unpredictable settings (see “Discussion” section for

remarks on the role of experience and learning in predator-prey interactions).
For sit-and-wait predators, prey may learn to avoid particular locations where
threatening encounters or attacks have occurred previously, much like the deer
in the introductory vignette. It is noteworthy, however, that sometimes recog-
nition can be difficult, even to the point that animals must assess the organism
for some time before deciding whether it is a potential predator or potential
prey (Cooper & Stankowich, 2010).

ASSESSMENT AND COMMUNICATION

Assessment
Once one or both parties have recognized the other as either a potential
threat or a potential meal, there may be a phase of assessment where the organ-
isms gather information regarding the likelihood of attack or capture.
Recently, there has been a growing interest in the factors that influence risk
assessment during encounters with predators. Dan Blumstein and I (Stanko-
wich & Blumstein, 2005) analyzed more than 100 previous studies of flight
responses by prey to determine what types of factors influence how prey assess
risk and to what degree they perceive fear. Animals perceive greater threat
when predators behave as if they intend to attack or are targeting the subject,
when they are more familiar with the predator due to experience (naïve indi-
viduals may be relatively fearless), when they are farther from refuge, and
when they lack cryptic or defensive morphologies like armor or chemical
defenses (Figure 2.3). Black-tailed deer (Odocoileus hemionus columbianus),
like many other species, are particularly attentive to the behavior of the
approaching predator, perceiving greater risk when the predator approaches
more directly and at greater speeds (Stankowich & Coss, 2006), but they are
not attuned to fine-scale changes in a predator’s gaze or posture during
approach that might further inform their decision to flee. A similar suite of
factors (e.g., size, handling time, probability of success, hunger state) likely
influences predators during their assessment of potential prey. David Scheel
(1993), for example, found that lions (Panthera leo) spent longer durations
stalking larger groups of prey and prey groups with more individuals in the
head-up position than they did with smaller prey groups or those with fewer
vigilant individuals, suggesting that lions assess more vigilant groups for longer
because they are more difficult to attack. Ultimately, many models of assess-
ment during foraging are applicable to predators assessing prey (see “Attack
and Flight” section below and Chapter 1 in this Volume; Brown et al.,
2011). In general, theory suggests that animals take in much more
PREDATION AND ANTIPREDATOR BEHAVIOR 35

Figure 2.3. The diagram summarizes the results of a review and meta-analysis of the
many factors that potentially influence flight-initiation distance in animals (from
Stankowich & Blumstein, 2005). Thicker lines and bolder, larger fonts indicate that
greater statistical and theoretical support exists for the influence of the factor on flight
decisions. Among others, threatening predator behavior and longer distances signifi-
cantly and consistently increase perceptions of fear in animals, while others like group
size and physical condition had weaker or less consistent effects. (Copyright 2005
Royal Society and Highwire Press)

information than they are capable of dealing with and likely weigh and deem-
phasize certain factors during assessment: natural selection should favor indi-
viduals that pay attention to only one or a few factors to streamline decision
making, something very important during fleeting interactions with potential
predators and potential prey (Bernays & Wcislo, 1994).

Communicating with Predators

In many predator-prey interactions, the prey assesses that it would be suc-
cessful in escaping to safety or defending itself should the predator choose to
attack. However, any such escalation raises the risk of injury to the prey ani-
mal and, at the very least, prolonged fleeing behavior or combative interac-
tions are certainly energetically costly. Natural selection, therefore, has
36 ANIMAL BEHAVIOR

favored individuals that can successfully communicate to a potential predator

that any such attack would be fruitless and would only cause similar unneces-
sary risk and energy expenditure on the part of the predator.
When predators hunt using stealth, often simply advertising to the predator
that it has been detected is enough to deter the predator from further pursuit.
In the introductory vignette, the snorting and footstamping of the deer, in
addition to potentially warning other deer in the area of a potential threat, is
also detectable by the predator and is available as a signal that its cover has
been blown. Often these behaviors become more overt and even risky adver-
tisements of prey quality (Stankowich, 2010), whereby prey behave or harass
the predator in a way signaling that any attack would likely be unsuccessful
given the escape or defensive ability of the prey animal. Don Owings and Matt
Rowe studied the tail flagging and snake-harassment behavior of California
ground squirrels (Spermophilus beecheyi) for more than 30 years, showing that
squirrels may sequentially approach and retreat from snakes while kicking sub-
strate at them, flagging their tails back and forth, and even biting in order to
deter further attack by the snake and indicate the defensive quality of the
squirrel (Owings & Coss, 1977; Owings & Morton, 1997; Rowe & Owings,
1990, 1996; Swaisgood, Owings, et al., 1999; Swaisgood, Rowe, et al., 1999;
Swaisgood et al., 2003). Recently, Aaron Rundus and colleagues (2007) dem-
onstrated that these signals are tuned to the sensory modalities of the species of
snake with which the squirrels are confronted. California ground squirrels will
flag their tails when accosted by both rattlesnakes (Crotalus oreganus) and
gopher snakes (Pituophis melanoleucus) but will increase blood flow to their
tails to raise the temperature in the tail to emit radiant heat when flagging
towards a rattlesnake, which can detect infrared radiation using pit organs,
but not towards a gopher snake, which lacks pit organs (Figure 2.4; Rundus
et al., 2007). Further, rattlesnakes were more likely shift to defensive postures
than predatory postures when presented with a robotic squirrel with a heated
flagging tail compared to one with an unheated flagging tail.

Aposematism
Visual communication in the form of body coloration is also a common
way to advertise to and warn potential predators that they should not attempt
an attack. Bold, contrasting color patterns are very often found on species that
bear some sort of defensive weapon or trait that would make them an unpleas-
ant meal for predators. Aposematism, or warning coloration, is found in a vari-
ety of animal groups but most commonly in insects, marine gastropods, and
amphibians. Several teams of researchers have examined the roles of predator
PREDATION AND ANTIPREDATOR BEHAVIOR 37

Figure 2.4. Mean body surface temperature in each condition across six body regions
in California ground squirrels (from Rundus et al., 2007). Testing conditions are
coded as: & ¼ rattlesnake, 4 ¼ gopher snake, ^ ¼ conspecific,  ¼ baseline.
Squirrels showed increased emission of infrared radiation over baseline and control
trials when exposed to rattlesnakes, which can detect infrared radiation, but not when
exposed to gopher snakes, which cannot. This result suggests that squirrels assess risk
and adjust their response both behaviorally and physiologically to the type of predator
with which they are confronted. * P < 0.01 compared with conspecific and gopher
snake; † P < 0.01 compared with baseline, conspecific, and gopher snake. (Copyright
2007 National Academy of Sciences, USA)

and sexual selection on coloration and toxicity in dendrobatid poison dart

frogs. Poison dart frogs are brightly colored species (reds, blues, and yellows,
often contrasted against black) with variably toxic secretions in their skin,
and bright coloration has been shown to be an honest signal of toxicity, in at
least some species (Maan & Cummings, 2011). Brice Noonan and Aaron
Comeault (2009) placed clay frog models in the wild that varied in coloration:
black body with yellow dorsum to represent the conspicuous local species,
blue-bodied with black-and-yellow mottled dorsum to represent a novel
38 ANIMAL BEHAVIOR

conspicuous phenotype from a different location, and brown to represent a

cryptic phenotype. By examining the markings in the clay of the attacked
models, they were able to show that birds attacked the novel conspicuous models
more often than both the familiar local conspicuous and cryptic models (Figure
2.5), suggesting that avian predators learn to associate specific color patterns with
toxicity rather than simply avoiding all boldly colored frogs. This suggests that
predator selection has a purifying effect on coloration in these populations so that
most individuals have similar patterns. There is evidence, however, of trade-offs

Figure 2.5. Bars indicate the frequency of cryptic brown frog models, novel blue frog
models, and local yellow frog models attacked by visual (black) and nonvisual
(white) predators (modified from Noonan & Comeault, 2009). While both novel and
local frog models were conspicuously colored, birds preferentially attacked the novel
blue models and avoided local yellow models, suggesting that lack of bird predation
on yellow frogs is due to a learned association between yellow and toxic prey rather
than a general avoidance of brightly colored prey. (Copyright 2007 Royal Society and
Highwire Press)
PREDATION AND ANTIPREDATOR BEHAVIOR 39

between investing energy in toxicity versus conspicuousness, where degree of

toxicity and conspicuousness can become decoupled, and investing in alterna-
tive strategies (either more conspicuous or more toxic) increases diversity in
the warning signal and provides some level of protection for nontoxic mimic
species (Darst et al., 2006). Michael Speed, Graeme Ruxton, and their col-
leagues have developed several models of the evolution of aposematism
(Blount et al., 2009; Broom et al., 2006; Speed & Ruxton, 2005, 2007),
one of which suggests that the great variation in the relationship between con-
spicuousness and toxicity in dendrobatid frogs can likely be linked to differ-
ences in ecological conditions where high costs to conspicuousness might
favor heavier investment in toxicity as a defensive strategy, favoring further
decoupling of the two traits (Speed & Ruxton, 2007).
Finally, some mammals display bold, conspicuous color patterns (e.g.,
skunks, polecats, honey badgers [Mellivora capensis]), which may also signal
warnings to predators. A recent study by Stankowich, Caro, and Matt Cox
found that bold, salient, black-and-white coloration in mammalian carnivores
was found in stockier species, possessing noxious anal-gland secretions or pug-
nacious defensive dispositions, and living in more open habitats where they
might be more exposed to predators and less able to rely on vegetation or water
as a refuge (Stankowich et al., 2011). Contrasting coloration in other types of
mammals (e.g., ungulates and whales), however, likely has camouflage or
social communicative functions (Caro, 2005a; Caro, 2009; Caro et al.,
2011; Caro & Stankowich, 2010).

ATTACK AND FLIGHT

Decision to Flee
The studies of risk assessment and decisions to flee are significantly inter-
twined: researchers typically study both using flight-initiation distance, the
distance between the predator and prey at which the prey flees, as the primary
metric. Ron Ydenberg and Larry Dill were the first to develop a theoretical
economic model predicting when animals would flee from an approaching
predator based on the costs of fleeing and the costs of remaining. When the
costs of remaining exceeded the costs of fleeing, escape is predicted (Ydenberg
& Dill, 1986). Hundreds of studies have examined the factors influencing the
decision to flee (see “Assessment” above; Stankowich & Blumstein, 2005).
While the original economic model is intuitive and accurately predicts most
flight decisions, more recent models have focused on optimizing prey fitness
as opposed to simply looking for a break-even point in the costs of remaining
versus fleeing. William Cooper and William Frederick have developed general
40 ANIMAL BEHAVIOR

optimality models that maximize prey fitness and allow benefits to increase
during the encounter (Cooper & Frederick, 2007). They have applied and
refined these models to examine the effect of variable lethality and other fac-
tors on optimal flight decisions. When lizards lose their tails during nonlethal
attacks by predators (i.e., tail autotomy), they suffer a reduction in locomotor
performance and are thus more vulnerable to predators in future encounters
(i.e., predatory attacks increase in lethality), and they may also suffer reduced
fitness due to reduced social status and foraging ability. The net result of these
reductions in initial fitness and increase in lethality is that optimal flight dis-
tance increases (Cooper & Frederick, 2010). The ability of these optimality
models to account for changes in both initial fitness and performance during
predatory encounters makes them straightforward to modify to examine
changes in other factors (e.g., sex, age, health). Clearly, a wide variety of
potential factors are weighed during an animal’s decision of when to flee from
an approaching predator (Figure 2.3).

Decision to Attack
There is a great deal of theory but few reliable empirical data on how pred-
ators select which type of prey to hunt and which individuals to attack (Lima,
1998). Granted, the great majority of acts of predation occur either (1) on
prey that cannot defend themselves or flee, and thus the decision to attack is
simply a matter of selecting the most energetically beneficial individual(s) to
attack; or (2) with predators that will attack any individual prey item that it
recognizes in its vicinity. In both cases, the decision of whether or not to attack
is a foregone conclusion, and there are many empirical studies of prey selection
on immobile or defenseless prey (see Chapter 1 for discussion). On the other
hand there are many predator-prey interactions in nature where a predator
must assess a potential prey item and decide whether to attack it, and if so,
when and how. Very few empirical studies, however, are available that explore
this decision in predators, probably because predators are difficult to study in
the wild and witnessing natural acts of predation is rare. John Quinn and Will
Cresswell have had some success in testing attack decisions of sparrowhawks
(Accipiter nisus) in natural settings. Using intensive observational sampling of
sparrowhawks hunting redshanks (Tringa tetanus), they found that sparrow-
hawks were more likely to attack redshanks in larger groups that were closer
to cover and were spaced farther apart within those groups (Quinn & Cress-
well, 2004). A subsequent study showed that the targeted redshanks were
more widely spaced apart than their nearest neighbors that were not targeted
(Quinn & Cresswell, 2006). These findings suggest that predators that hunt
prey using costly attack styles and low probability of success pay close
PREDATION AND ANTIPREDATOR BEHAVIOR 41

attention to the behavior of potential targets, target prey that appear more vul-
nerable, and decide to attack only when the circumstances are favorable.

PURSUIT, ESCAPE, AND DEFENSE

Pursuit and Capture
The means by which predators have evolved to pursue and capture prey are
too numerous to cover here in any comprehensive way. Many predators pur-
sue their prey over some distance, similar to a cheetah (Acinonyx jubatus) chas-
ing down a gazelle at high speeds. Some predators pursue prey so as to keep
the moving target in front of them, which results in the predator following a
curved path (assuming a straight trajectory for the prey) toward the prey. In
contrast, some predators move to intercept the prey at some point in the dis-
tance. Dragonflies, for example, aim their pursuit at some point in front of
the prey, keeping it in a fixed position on its retina (Olberg et al., 2000). As
opposed to long chases, many predators are stealth hunters, moving to within
very short distances under cover before attacking. Large predatory cats and
jumping spiders in the genus Portia among many others follow this strategy;
most of these stealth hunters prefer sites that offer good cover, and some take
advantage of environmental disturbances to use as a smokescreen to further
cover their approach. Finally, some predators use traps to capture prey without
an active pursuit. As Frodo Baggins can attest, the most well-known examples
of predatory traps are spider webs. Orb-weaving spiders create the familiar cir-
cular webs between objects in their environment, and the sticky webs ensnare
flying insects that do not detect them until it is too late. While most prey cap-
tured in webs are small (<2 mm), Samuel Venner and Jérôme Casas (2005)
showed that spiders cannot survive without catching larger (>10 mm) prey
that are rare and difficult to ensnare in the web.

Defense
Once a predator has attacked, an organism has two choices—attempt to
flee to safety or attempt to hold its ground and defend itself physically—but
sometimes these options bleed together. We can classify prey defenses into
several different anthropomorphic categories: armor, combat weapons, and
chemical weapons (Caro, 2005b; Emlen, 2008; Stankowich, 2012). Defensive
armor is one of the most common morphological adaptations that prey have
evolved for protecting themselves against attacking predators. Armor may take
the form of spines or quills, as in sticklebacks, hedgehogs, and porcupines or of
thick protective plates, as in turtles, shelled gastropods and arthropods,
42 ANIMAL BEHAVIOR

armadillos, and pangolins. When an organism only has one or a few spines,
they often function to simply make it harder to swallow the animal (Hoogland
et al., 1956). When spines or quills cover the entire body, on the other hand,
they also make the prey harder to handle without risking injury, and in some
cases may even cushion the animal if dropped or knocked off of tree branches
or rocks (Vincent & Owers, 1986). Combat weapons include horns or antlers;
large teeth, fangs, or tusks; or long claws, paired with physical strength.
Chemical weapons include sprayed noxious or malodorous chemicals, venom,
and secreted toxins, among others. Mobbing by prey is also an effective
defense against single predators and is found in a wide variety of animal groups
(e.g., muskoxen [Ovibos moschatus] protecting young offspring, bees protect-
ing their hive, shorebirds protecting their nests). While there is a broad array
of morphological and behavioral adaptations used in defense against predators,
some may have evolved primarily in response to natural selection by predators,
while others may be secondarily co-opted for use against predators but evolved
primarily for some other function (e.g., combat with conspecifics or feeding).
In mammals, weaponry that evolved primarily for defensive purposes tends to
be found in intermediate-sized species that live in more exposed environments
with fewer opportunities for refuge, whereas weaponry that evolved primarily
for within-species combat (e.g., sexual weaponry in males) tends to be found
in larger species where size influences the outcomes of competitive interactions
(Stankowich, 2012; Stankowich & Caro, 2009).

Escape
If an animal is too small to defend itself against a predator or when its
defensive tactics fail, it is forced to attempt an escape to safety. The specific
escape strategy depends upon the medium and habitat but may include run-
ning, swimming, or flying away; dropping out of a tree; jumping out of the
way; burrowing; or diving into water. When prolonged escape is used, species
often use fast, erratic turns (i.e., protean behavior) that make it difficult for
heavier predators to adjust. If a group of animals scatters and flees erratically,
it may confuse the predator, making it difficult to follow any one individual.
In fact, if individuals all have different innate escape patterns, then it would
be impossible for predators to learn the pattern to the escape path; therefore,
the production and maintenance of such variation in escape pattern via genetic
recombination may actually promote the evolution of sexual reproduction
over asexual reproduction (the “red tooth hypothesis”; French, 2010). Paolo
Domenici and colleagues (2008) directed puffs of air directly towards the
heads of cockroaches (Periplaneta americana) to elicit an escape response and
measured the angle of the escape trajectory: cockroaches turned their bodies
PREDATION AND ANTIPREDATOR BEHAVIOR 43

and fled in four preferred trajectories (90º, 120º, 150º, and 180º). These four
paths probably result in variation in escape response sufficient to keep the
predator guessing. Further, protean behaviors, if performed at a distance,
may also help communicate agility to the pursuing predator: Thomson’s
gazelles (Eudorcas thomsonii) often bound up and down as they flee (termed
stotting) to advertise their quality to a predator and encourage it to stop the
pursuit.
Species differences in escape strategies typically depend on a wide range of
ecological and morphological factors. Steve Lima, for example, was able to
group 43 emberizine finch species into five types based on sociality and vegeta-
tion density (Lima, 1993). Species may also vary their escape strategy based on
the type of predator they are faced with. Redshanks suffer the lowest capture
rates by flying away from sparrowhawks and merlins (Falco columbarius), but
when faced with peregrine falcons (F. peregrinus), diving into creeks and
crouching are far more effective modes of escape (Cresswell, 1996).

DISCUSSION
Using the attack sequence provides a useful framework for reviewing the
many morphological and behavioral adaptations of predators and prey.
Detailed coverage of all of these adaptations would fill several volumes (but
see Caro, 2005b, for a good start), and it is clear from even this cursory over-
view that a huge array of factors besides predation and the need for food drive
the evolution of these adaptations. Some recent studies have focused on the
persistence of antipredator adaptations in the absence of predators (i.e., relax-
ation in selection). Some costly morphological traits may be lost rapidly if they
are coded for by only one or a few genes, while many behavioral traits can take
hundreds of thousands of years to degrade (Lahti et al., 2009). Predator recog-
nition, for example, is highly variable in its tendency to degrade in the absence
of predators. Moose (Alces alces) showed reduced recognition of wolves (Canis
lupus) after only 50 to 130 years of absence (Berger et al., 2001), and tammar
wallabies (Macropus eugenii) suffered a breakdown in visual predator recogni-
tion after 130 years of isolation from mammals (Blumstein et al., 2004). At
longer time scales, Columbian black-tailed deer showed loss of recognition
of spotted cats after 600,000 years (Stankowich & Coss, 2007), and California
ground squirrels retained snake recognition after 300,000 years of allopatry
but lost it after 3 to 5 million years (Coss, 1991; Goldthwaite et al., 1990).
A few central factors influence predator-prey evolution up and down the
attack sequence. Group size has powerful effects on both predator hunting
behavior and prey antipredator responses (reviewed at length in Krause &
44 ANIMAL BEHAVIOR

Ruxton, 2002). Larger prey group sizes result in increased vigilance ability
(many-eyes effect), greater speed of information transfer between individuals,
dilution of risk, and the ability to confuse or mob predators. Spatial position
within groups influences predation risk but also influences feeding rates
by predators in foraging groups. Large prey groups also make for potentially
larger meals for predators. Larger predator groups can allow for capture of prey
that are larger than they would otherwise be able to bring down if hunting
alone and improve the likelihood of prey detection. There are many benefits
and costs to grouping behavior for all organisms, and their effects on predation
risk and ability to locate prey are strong influences on optimal group size.
Lifetime experience also significantly influences predator-prey interactions
up and down the attack sequence. For many organisms that are more develop-
mentally plastic, exposure to certain predator cues at an early age can influence
both morphological development and adult behavior, including predator rec-
ognition and risk assessment. Maud Ferrari, Doug Chivers, Grant Brown,
and their colleagues have, in recent years, made great strides in understanding
the effect of exposure to predator cues and conspecific alarm odors in early
developmental environments on learning ability and predator recognition.
They tested the ability of fathead minnows (Pimephales promelas) to generalize
between a known predator species and other species that varied in evolution-
ary relatedness. They found that after teaching minnows to fear lake trout
(Salvenilus namaycush) by pairing the lake trout odor with conspecific alarm
odors, the minnows showed strong antipredator responses to the lake trout,
brook trout (S. fontinalis, a member of the same genus as lake trout), and rain-
bow trout (Oncorhynchus mykiss, a member of a different genus but the same
family), but they did not recognize odors of predatory northern pike (Esox
Lucius, a member of a different order) or nonpredatory white suckers (Catosto-
mus commersoni, a member of a different superorder) (Figure 2.6: Ferrari et al.,
2007). In a similar study they found nearly identical generalization patterns in
rainbow trout responding to predators of varying degrees of relatedness
(Brown et al., 2011).
Clearly, the evolution of adaptive predatory and antipredatory behaviors
and morphologies is influenced by many environmental factors and sources
of selection, including the need to locate and attract potential mates. In fact,
reproductive and defensive strategies often evolve to complement each other
insomuch as some entire mating systems may be disadvantageous in light of
morphological constraints based on predator avoidance or the need to hunt.
In this respect, broad syntheses of sexual and natural selection are needed
to look for general patterns and correlations between reproductive, feeding,
and antipredator adaptations. As this chapter demonstrates, there many ways
Figure 2.6. The mean change in shoaling index (top) and line crosses (bottom) for fat-
head minnows conditioned to fear lake trout odor paired with either water (black
bars) or conspecific alarm cues (white bars) (redrawn from Ferrari et al., 2007). Min-
nows were able to generalize their conditioned fear of lake trout to brook trout (same
genus as lake trout) and rainbow trout (different genus but same family as lake trout)
but not to northern pike (different order from lake trout) or nonpredatory white
suckers (different superorder from lake trout). Lifetime experience with predatory
stimuli has significant effects on future antipredator responses. (Copyright 2007
Royal Society and Highwire Press)
46 ANIMAL BEHAVIOR

that organisms can adapt to new threats or new foods in their environment,
and these adaptations complement or trade off with other fitness-enhancing
traits already present in the species’ repertoire.

ACKNOWLEDGMENTS
Many thanks to E. Fernández-Juricic, M. Ferrari, and R. Coss for advice on
particular references, access to original data or images, and comments on por-
tions of this chapter.

REFERENCES AND SUGGESTED READING

Arraut, E. M., M. Marmontel, J. E. Mantovani, E. M. L. M. Novo, D. W. Macdon-
ald, & R. E. Kenward. (2010). The lesser of two evils: Seasonal migrations of
Amazonian manatees in the Western Amazon. Journal of Zoology, 280,
247–256.
Berger, J., J. E. Swenson, & I. L. Persson. (2001). Recolonizing carnivores and naive
prey: Conservation lessons from Pleistocene extinctions. Science, 291,
1036–1039.
Bernays, E. A. & W. T. Wcislo. (1994). Sensory capabilities, information processing,
and resource specialization. Quarterly Review of Biology, 69, 187–204.
Blount, J. D., M. P. Speed, G. D. Ruxton, & P. A. Stephens. (2009). Warning dis-
plays may function as honest signals of toxicity. Proceedings of the Royal Society
of London, B, 276, 871–877.
Blumstein, D. T., J. C. Daniel, & B. P. Springett. (2004). A test of the multi-predator
hypothesis: Rapid loss of antipredator behavior after 130 years of isolation.
Ethology, 110, 919–934.
Broom, M., M. P. Speed, & G. D. Ruxton. (2006). Evolutionarily stable defence and
signalling of that defence. Journal of Theoretical Biology, 242, 32–43.
Brown, G. E., M. C. O. Ferrari, P. H. Malka, S. Russo, M. Tressider, & D. P. Chiv-
ers. (2011). Generalization of predators and nonpredators by juvenile rainbow
trout: Learning what is and is not a threat. Animal Behaviour, 81, 1249–1256.
Caro, T. (2005a). The adaptive significance of coloration in mammals. BioScience, 55,
125–136.
Caro, T. (2005b). Antipredator Defenses in Birds and Mammals. Chicago: University
of Chicago Press.
Caro, T. (2009). Contrasting coloration in terrestrial mammals. Philosophical Transac-
tions of the Royal Society of London, B, 364, 537–548.
Caro, T., K. Beeman, T. Stankowich, & H. Whitehead. (2011). The functional sig-
nificance of colouration in cetaceans. Evolutionary Ecology, 25, 1231–1245.
Caro, T. & T. Stankowich. (2010). The function of contrasting pelage markings in
artiodactyls. Behavioral Ecology, 21, 78–84.
PREDATION AND ANTIPREDATOR BEHAVIOR 47

Cooper, W. E. & W. G. Frederick. (2007). Optimal flight initiation distance. Journal

of Theoretical Biology, 244, 59–67.
Cooper, W. E. & W. G. Frederick. (2010). Predator lethality, optimal escape behav-
ior, and autotomy. Behavioral Ecology, 21, 91–96.
Cooper, W. E. & T. Stankowich. (2010). Prey or predator? Body size of an approaching
animal affects decisions to attack or escape. Behavioral Ecology, 21, 1278–1284.
Coss, R. G. (1991). Context and animal behavior III: The relationship between early
development and evolutionary persistence of ground squirrel antisnake behav-
ior. Ecological Psychology, 3, 277–315.
Coss, R. G. (1999). Effects of relaxed selection on the evolution of behavior. In S. A.
Foster & J. A. Endler (eds.), Geographic Variation in Behavior: Perspectives on
Evolutionary Mechanisms (pp. 180–208). New York: Oxford University Press.
Coss, R. G. (2010). Predator avoidance: Mechanisms. In M. D. Breed & J. Moore
(eds.), Encyclopedia of Animal Behavior, Vol. 2 (pp. 757–764). Oxford, UK:
Academic Press.
Coss, R. G. & R. O. Goldthwaite. (1995). The persistence of old designs for percep-
tions. In N. S. Thompson (ed.), Perspectives in Ethology, Vol. 11 (pp. 83–148).
New York: Plenum Press.
Coss, R. G. & U. Ramakrishnan. (2000). Perceptual aspects of leopard recognition by
wild bonnet macaques (Macaca radiata). Behaviour, 137, 315–335.
Coss, R. G., U. Ramakrishnan, & J. Schank. (2005). Recognition of partially con-
cealed leopards by wild bonnet macaques (Macaca radiata). The role of the
spotted coat. Behavioural Processes, 68, 145–163.
Cresswell, W. (1994). Flocking is an effective anti-predation strategy in redshanks,
Tringa tetanus. Animal Behaviour, 47, 433–442.
Cresswell, W. (1996). Surprise as a winter hunting strategy in sparrowhawks Accipiter
nisus, peregrines Falco peregrinus and merlins F. columbarius. Ibis, 138,
684–692.
Curio, E. (1975). The functional organization of anti-predator behaviour in pied fly-
catcher: A study of avian visual perception. Animal Behaviour, 23, 1–115.
Curio, E. (1993). Proximate and developmental aspects of antipredator behavior.
Advances in the Study of Behavior, 22, 135–238.
Darst, C. R., M. E. Cummings, & D. C. Cannatella. (2006). A mechanism for diver-
sity in warning signals: Conspicuousness versus toxicity in poison frogs. Pro-
ceedings of the National Academy of Sciences, USA, 103, 5852–5857.
David, E. (2005). Die hard: A blend of freezing and fleeing as a dynamic defense—
implications for the control of defensive behavior. Neuroscience and Biobehavio-
ral Reviews, 29, 1181–1191.
Domenici, P., D. Booth, J. M. Blagburn, & J. P. Bacon. (2008). Cockroaches keep
predators guessing by using preferred escape trajectories. Current Biology, 18,
1792–1796.
Emlen, D. J. (2008). The evolution of animal weapons. Annual Review of Ecology,
Evolution, and Systematics, 39, 387–413.
48 ANIMAL BEHAVIOR

Endler, J. A. (1981). An overview of the relationships between mimicry and crypsis.

Biological Journal of the Linnean Society, 16, 25–31.
Fernández-Juricic, E., M. D. Gall, T. Dolan, C. O’Rourke, S. Thomas, & J. R.
Lynch. (2011). Visual systems and vigilance behaviour of two ground-
foraging avian prey species: White-crowned sparrows and California towhees.
Animal Behaviour, 81, 705–713.
Ferrari, M. C. O., A. Gonzalo, F. Messier, & D. P. Chivers. (2007). Generalization of
learned predator recognition: an experimental test and framework for future
studies. Proceedings of the Royal Society of London, B, 274, 1853–1859.
French, R. M. (2010). The red tooth hypothesis: A computational model of predator-
prey relations, protean escape behavior and sexual reproduction. Journal of
Theoretical Biology, 262, 165–176.
Goldthwaite, R. O., R. G. Coss, & D. H. Owings. (1990). Evolutionary dissipation
of an antisnake system: Differential behavior by California and Arctic ground
squirrels in above- and below-ground contexts. Behaviour, 112, 246–269.
Hoogland, R., D. Morris, & N. Tinbergen. (1956). The spines of sticklebacks (Gas-
terosteus and Pygosteus) as means of defence against predators (Perca and Esox).
Behaviour, 10, 205–236.
Kats, L. B.& L. M. Dill. (1998). The scent of death: Chemosensory assessment of
predation risk by prey animals. [Review]. Ecoscience, 5, 361–394.
Krause, J. & G. D. Ruxton. (2002). Living in Groups. Oxford, UK: Oxford Univer-
sity Press.
Lahti, D. C., N. A. Johnson, B. C. Ajie, S. P. Otto, A. P. Hendry, D. T. Blumstein, et
al. (2009). Relaxed selection in the wild. Trends in Ecology and Evolution, 24,
487–496.
Langerhans, R. B. (2007). Evolutionary consequences of predation: Avoidance,
escape, reproduction, and diversification. In A. M. T. Elewa (ed.), Predation
in Organisms: A Distinct Phenomenon (pp. 177–220). Berlin: Springer.
Lima, S. L. (1993). Ecological and evolutionary perspectives on escape from predatory
attack: A survey of North American birds. Wilson Bulletin, 105, 1–47.
Lima, S. L. (1998). Stress and decision making under the risk of predation: Recent
developments from behavioral, reproductive, and ecological perspectives.
Advances in the Study of Behavior, 27, 215–290.
Lima, S. L. & L. M. Dill. (1990). Behavioral decisions made under the risk of preda-
tion: A review and prospectus. Canadian Journal of Zoology, 68, 619–640.
Maan, M. E. & M. E. Cummings. (2012). Poison frog colors are honest signals of
toxicity, particularly for bird predators. American Naturalist, 179, E1–E14.
http://www.jstor.org/stable/10.1086/663197.
Noonan, B. P. & A. A. Comeault. (2009). The role of predator selection on polymor-
phic aposematic poison frogs. Biology Letters, 5, 51–54.
Olberg, R. M., A. H. Worthington, & K. R. Venator. (2000). Prey pursuit and inter-
ception in dragonflies. Journal of Comparative Physiology A: Neuroethology,
Sensory, Neural, and Behavioral Physiology, 186, 155-162.
PREDATION AND ANTIPREDATOR BEHAVIOR 49

O’Rourke, C. T., M. I. Hall, T. Pitlik, & E. Fernández-Juricic. (2010). Hawk eyes I:

Diurnal raptors differ in visual fields and degree of eye movement. PLoS ONE,
5, e12802.
O’Rourke, C. T., T. Pitlik, M. Hoover, & E. Fernández-Juricic. (2010). Hawk eyes
II: Diurnal raptors differ in head movement strategies when scanning from
perches. PLoS ONE, 5, e12169.
Owings, D. H. & R. G. Coss. (1977). Snake mobbing by California ground squirrels:
Adaptive variation and ontogeny. Behaviour, 62, 50–69.
Owings, D. H. & E. S. Morton. (1997). The role of information in communication:
An assessment/management approach. In D. H. Owings, M. D. Beecher, &
N. Thompson (eds.), Perspectives in Ethology, Vol. 12 (pp. 359–390). New
York: Plenum Press.
Pulliam, H. R. (1973). On the advantages of flocking. Journal of Theoretical Biology,
38, 419–422.
Quinn, J. L. & W. Cresswell. (2004). Predator hunting behaviour and prey vulner-
ability. Journal of Animal Ecology, 73, 143–154.
Quinn, J. L. & W. Cresswell. (2006). Testing domains of danger in the selfish herd:
Sparrowhawks target widely spaced redshanks in flocks. Proceedings of the Royal
Society of London, B, 273, 2521–2526.
Rowe, M. P. & D. H. Owings. (1990). Probing, assessment, and management during
interactions between ground squirrels and rattlesnakes. Part 1: Risks related to
rattlesnake size and body temperature. Ethology, 86, 237–249.
Rowe, M. P. & D. H. Owings. (1996). Probing, assessment and management during
interactions between ground squirrels (Rodentia: Sciuridae) and rattlesnakes
(Squamata: Viperidae). 2: Cues afforded by rattlesnake rattling. Ethology,
102, 856–874.
Rundus, A. S., D. H. Owings, S. S. Joshi, E. Chinn, & N. Giannini. (2007). Ground
squirrels use an infrared signal to deter rattlesnake predation. Proceedings of the
National Academy of Sciences, USA, 104, 14372–14376.
Ruxton, G. D., T. N. Sherratt, & M. P. Speed. (2004). Avoiding Attack: The Evolu-
tionary Ecology of Crypsis, Warning Signals, and Mimicry. New York: Oxford
University Press.
Scheel, D. (1993). Watching for lions in the grass: The usefulness of scanning and its
effects during hunts. Animal Behaviour, 46, 695–704.
Scheuerell, M. D. & D. E. Schindler. (2003). Diel vertical migration by juvenile sock-
eye salmon: Empirical evidence for the antipredation window. Ecology, 84,
1713–1720.
Smith, W. P. (1991). Ontogeny and adaptiveness of tail-flagging behavior in white-
tailed deer. American Naturalist, 138, 190–200.
Speed, M. P. & G. D. Ruxton. (2005). Aposematism: What should our starting point
be? Proceedings of the Royal Society of London, B, 272, 431–438.
Speed, M. P. & G. D. Ruxton. (2007). How bright and how nasty: Explaining diver-
sity in warning signal strength. Evolution, 61, 623–635.
50 ANIMAL BEHAVIOR

Stankowich, T. (2010). Risk-taking in self-defense. In M. D. Breed & J. Moore

(eds.), Encyclopedia of Animal Behavior, Vol. 3 (pp. 79–86). Oxford, UK:
Academic Press.
Stankowich, T. (2012). Armed and dangerous: Predicting the presence and function
of defensive weaponry in mammals. Adaptive Behavior, 20, 34–45.
Stankowich, T. & D. T. Blumstein. (2005). Fear in animals: A meta-analysis and
review of risk assessment. Proceedings of the Royal Society of London, B, 272,
2627–2634.
Stankowich, T. & T. Caro. (2009). Evolution of weaponry in female bovids. Proceed-
ings of the Royal Society of London, B, 276, 4329–4334.
Stankowich, T., T. Caro, & M. Cox. (2011). Bold coloration and the evolution of
aposematism in terrestrial carnivores. Evolution, 65, 3090–3099.
Stankowich, T. & R. G. Coss. (2006). Effects of predator behavior and proximity on
risk assessment by Columbian black-tailed deer. Behavioral Ecology, 17,
246–254.
Stankowich, T. & R. G. Coss. (2007). The re-emergence of felid camouflage with the
decay of predator recognition in deer under relaxed selection. Proceedings of the
Royal Society of London, B, 274, 175–182.
Swaisgood, R. R., D. H. Owings, & M. P. Rowe. (1999). Conflict and assessment in
a predator-prey system: Ground squirrels versus rattlesnakes. Animal Behav-
iour, 57, 1033–1044.
Swaisgood, R. R., M. P. Rowe, & D. H. Owings. (1999). Assessment of rattlesnake
dangerousness by California ground squirrels: Exploitation of cues from rat-
tling sounds. Animal Behaviour, 57, 1301–1310.
Swaisgood, R. R., M. P. Rowe, & D. H. Owings. (2003). Antipredator responses of
California ground squirrels to rattlesnakes and rattling sounds: The roles of
sex, reproductive parity, and offspring age in assessment and decision-making
rules. Behavioral Ecology and Sociobiology, 55, 22–31.
Treherne, J. E. & W. A. Foster. (1980). The effects of group size on predator avoid-
ance in a marine insect. Animal Behaviour, 28, 1119–1122.
Venner, S. & J. Casas. (2005). Spider webs designed for rare but life-saving catches.
Proceedings of the Royal Society of London, B, 272, 1587–1592.
Vincent, J. F. V. & P. Owers. (1986). Mechanical design of hedgehog spines and
porcupine quills. Journal of Zoology, 210, 55–75.
Ydenberg, R. C. & L. M. Dill. (1986). The economics of fleeing from predators.
Advances in the Study of Behavior, 16, 229–249.
Zaret, T. M. & J. S. Suffern. (1976). Vertical migration in zooplankton as a predator
avoidance mechanism. Limnology and Oceanography, 21, 804–813.
 3

Animal Communication: How and Why

Animals Communicate—or Do They?
Zuleyma Tang-Martı́nez

INTRODUCTION AND TERMS

As a child growing up in Venezuela, I loved to climb trees and then, from that
vantage point, watch iguanas (which are very large lizards) scurrying around in
neighboring trees and interacting with one another. One behavior, which
involved the iguanas bobbing their heads rapidly up and down, was particu-
larly noticeable, and I often wondered what this head bobbing could possibly
mean. Years later, when I became an animal behaviorist, I learned that head
bobbing is one of the most common behaviors that many lizards use to com-
municate with one another. Each species has its own pattern of head bobs,
and they use these behaviors during courtship or aggressive interactions.
In the first part of this chapter, I will address only intraspecific communi-
cation—communication that occurs between or among members of the same
species—initially focusing specifically on more traditional views of communi-
cation. Later I also will discuss the possibility that communication can occur
between members of different species—interspecific communication. Also
later in the chapter, I will attend to recent debates and different perspectives
(e.g., Dawkins & Krebs, 1978; Owren et al., 2010; Rendall et al., 2009) that
question whether communication is honest and whether it is best described as
the exchange of information or as a manipulative interaction; related to this
issue is the question of whether animal signals can be said to encode
52 ANIMAL BEHAVIOR

information or are best understood by focusing on their effects or functions,

independently of whether signals encode accurate information (see “Is Com-
munication Honest?” and “Do Signals Really Convey Information?” below).
As you will see, throughout the history of the study of communication, many
debates have arisen, and as we work through these debates and consider differ-
ent perspectives, new insights can be gained. Science, at its core, is about
debating issues and testing alternative hypotheses; the study of communica-
tion is an excellent example of this approach.
At this point you might ask, what is communication? and how do we know
that animals can communicate with one another? In animal behavior we say that
communication has occurred when the behavior of one individual (A) affects
the behavior of another individual (B). Additionally, communication is usu-
ally assumed to result in “reduction of uncertainty” (e.g., Seyfarth et al.,
2010). Some definitions (Slater, 1983) additionally require that animal A ben-
efits (in terms of survival or reproductive success) from the response of animal
B. Generally, the effect of communication is one of probabilities. For example,
using an analogy in humans, if individual A is speaking and individual B does
something (e.g., putting his finger to his lips) and individual A immediately
stops talking, we can say that communication occurred. Likewise, in the liz-
ards, if one individual head bobs and another individual that was approaching
stops, then we can say that communication has occurred. In both cases the
behavior of one individual affected the probability that the second individual
would perform a certain behavior. In other words, the behavior of A decreased
the probability that B would continue speaking, or in the case of the lizard,
that B would continue approaching. However, the effects can also be an
increase in the probability of a behavior occurring in B. For example, head
bobbing by A could result in B suddenly beginning to also head bob—an
increase in the frequency of head bobbing by B.
From a traditional perspective, all communication systems have a sender, a
signal, and a receiver. In the lizard example, individual A was the sender and
individual B was the receiver. The head bob was the signal. Consider that you
are listening to the radio. In this case, the broadcasting station is the sender,
the sound coming from the radio is the signal, and you are the receiver. But
what comes from the radio is not just random noise; rather, it contains infor-
mation, so we say that all signals have information, sometimes also called a
message content (but see “But Do Signals Really Convey Information?” below
for an alternative perspective). Now imagine that you and a friend are listening
to the radio when a tornado warning is broadcast. Your friend is fascinated by
storms, so she runs to the window to see if she can see the tornado approach-
ing. You are frightened, so you immediately head down to the safety of your
ANIMAL COMMUNICATION 53

basement. Both you and your friend are receivers, but you have responded
differently to the same message. We can say that your interpretation of the
message is different.
The same thing can happen in animal communication. Consider a male
bird singing in its territory. The information content of the song (signal)
may be that he is a male in reproductive condition, he is looking for a mate,
and he already has a territory that he will defend against any intruders. A
female bird (receiver) listens to the song and interprets it as an invitation.
She approaches the male as a potential mating partner. A male bird listening
to the same song, and receiving the same messages, interprets the song as a
warning and flies away to avoid getting into a fight with the territorial male.
Thus, any time that we study communication we must be aware that the sig-
nal, information content, and interpretation are all different and that different
receivers can have different interpretations—and therefore different responses
to the same signal and message (e.g., Marler, 1961; Smith, 1977; but also see
discussions in Ruxton & Schaefer, 2011; Searcy & Nowicki, 2005).
W. John Smith (1977), who first proposed the concept of “messages,” also
distinguished among different types of messages. Behavioral messages are
those that provide information on what the sender is most likely to do next
(what behavior, the probability that it will occur, and how intensely the behav-
ior may be performed). Nonbehavioral messages provide information about
the identity of the sender (i.e., they are “identifiers”) or about location (of
the sender or other relevant item such as food). This framework is the one that
has dominated the study of animal communication, but as alluded to earlier, it
has recently been challenged. There are other interesting aspects of communi-
cations systems. For example, some signals may be directed specifically at a
particular individual. A baby bird that makes shrill noises when it is hungry
is directing those calls specifically to its parent. On the other hand, a honeybee
(Apis mellifera) that detects a danger (e.g., a predator) will produce an alarm
odor (alarm pheromone) that alerts all bees in the vicinity that there is a dan-
ger (Boch & Shearer, 1971; Wager & Breed, 2000). In the latter case, when
the sender produces a signal that is propagated throughout the environment
without having one specific intended receiver, we refer to the signal as a
broadcast signal. Note also that any time there is a broadcast signal, and even
in some cases where the signal is intended only for one receiver, eavesdrop-
ping can happen. For example, in the case of the baby bird’s feeding calls,
a nest predator may eavesdrop on the calls and use them to locate and eat
the baby bird. Individuals in social species may give calls to indicate to group
members that they have found food; other individuals who are not group
members can eavesdrop and also take advantage of the food. Another
54 ANIMAL BEHAVIOR

well-known example involves the túngara frog (Physalaemus pustulosus). In this

species males vocalize to attract females. However, there are frog-eating bats
(Trachops cirrhosus) that can use the males’ mating calls to locate the calling
males and then eat them (Ryan et al., 1982). Other examples (e.g., Templeton
& Green, 2007) include birds of one species, such as nuthatches (Sitta cana-
densis) eavesdropping and responding to the alarm calls of a different species,
the black-capped chickadee (Poecile atricapillus). And, for an even more com-
plex example involving several species of birds, see Rob Magrath and col-
leagues (2009). In fact, in some cases, birds and mammals may even respond
to one another’s alarm calls; hornbill birds (Ceratogymna elata), for example,
respond to alarm calls of Diana monkeys (Cercopithecus diana) (Rainey et al.,
2004). Obviously, broadcast signals and eavesdropping have recently become
important and very active areas for study in animal communication. Some of
these examples are also discussed in more detail in the section “Interspecific
Communication” (below).
In addition to signals, animal behaviorists frequently also use the word dis-
play. Technically, a display is a stereotyped behavior that has evolved to func-
tion specifically as a signal (or a series of signals) in communication. Some
displays are considered “derived” behaviors because, presumably, they were
originally used in a different context (i.e., not in communication) but over
time became modified specifically for use in communication. This issue will
be discussed in more detail below, under “Ritualization and Origins of Dis-
plays.” However, for the purposes of this chapter, I consider signal and display
as interchangeable terms unless indicated otherwise.

MODALITIES OF COMMUNICATION
All animal communication involves the use of senses. Specifically, signals
may be visual (e.g., postures, displays, coloration), acoustic (e.g., songs, calls),
chemical (e.g., glandular secretions, saliva, urine), tactile (e.g., touching, strok-
ing, plucking), electrical (e.g., electric pulses), or vibrational (e.g., waves trans-
mitted through substrates such as the ground, plant stems, or water). I suspect
the first four modalities did not surprise you because they are senses that
humans are aware of and use routinely. On the other hand, you may not have
realized the importance of electrical communication in some species. Weakly
electric fishes, for example, communicate with a series of electric pulses pro-
duced by specialized organs (Hagedorn, 1986). These pulses can encode infor-
mation by varying characteristics such as their pattern, repetition rate, or
frequency. Likewise, many species are now known to communicate by send-
ing waves through substrates. Sometimes it is difficult to distinguish these
ANIMAL COMMUNICATION 55

types of signals from sounds (after all, sounds are also waves that travel
through the air). Certain insects (e.g., tree hoppers) communicate by produc-
ing waves that travel through the stems of plants, and they use different types
of waves depending on what is being communicated (Cocroft, 2005; Cocroft
& Rodriguez, 2005). Some spiders use parts of their bodies to generate waves
that travel through the ground and are perceived by receivers (e.g., Gibson &
Uetz, 2008; Sivalinghem et al., 2010), and some underground rodents either
butt their heads against the roofs of their burrows (Heth et al., 1987; Rado
et al., 1998), or footdrum on the ground (Narins et al., 1997; Randall &
Lewis, 1997), generating waves that travel through the ground to other con-
specifics. The term seismic communication is often used to describe signals
that are propagated through the ground. In all cases of vibrational communi-
cation, humans cannot hear the “sounds” that are generated unless they use
specialized equipment. However, using such equipment, for example in your
backyard, can result in a cacophony of amazing sounds: whines, grunts,
knocks, whistles, chirps, and so on—sort of the insect equivalent of “heavy
metal”!

Visual Communication
Visual signals are used by many different species of animals and can be either
dynamic or static. A static signal, such as the color pattern of the peacock’s tail,
is one that does not change. A dynamic signal can change while the display is
taking place. For example, some fish have dark stripes on their sides, but they
can make the stripes very dark or so light they are almost invisible (and every-
thing in between) during an aggressive encounter (e.g., Baylis, 1974). The
darkness of the stripes communicates motivation—such as whether the fish
at that instant in time is feeling more aggressive or more scared. Visual signals
may be morphological, consisting of special structures (e.g., crests or elongated
tails in birds) or of colors or patches of colors (the black chest bib in house
sparrows (Passer domesticus), or behavioral, consisting of movements, postures,
or displays produced by the animal (e.g., head bobbing of lizards—but note
here that this use of behavioral differs somewhat from that of Smith [1977],
who uses this term specifically with regard to messages or information content
of signals). Frequently, the morphological and behavioral signals are combined
in one display. Many lizards have dewlaps—a fold of skin that can be extended
into a semicircular structure—under their chins. Dewlaps can be very colorful
and are often displayed during head bobbing, thereby combining a morpho-
logical signal with a behavioral one. In some species there is evidence that
females can recognize males of their own species by the color of the dewlap
and/or the pattern of head bobbing and push-ups, and that they may use these
56 ANIMAL BEHAVIOR

signals to avoid mating with the wrong species (e.g., Hunsaker, 1962; Jenssen,
1970). Some animals (e.g., some birds and insects) even produce and perceive
ultraviolet signals that humans cannot see.
A special type of visual signal involves structures that are produced by the
sender and then used as part of a display. The best know example of this is
found in bowerbirds (Borgia, 1986). In this group, males build elaborate
structures—called bowers—out of twigs and other vegetation and then deco-
rate them with colorful fruit, flower petals, string, shiny pieces of paper, and
many other types of objects. Once the bower is completed the male courts
passing females by bowing, posturing, and calling near the entrance to the
bower. Females apparently choose males based on their bowers, as well as on
the male’s behaviors (Borgia, 1986).

Acoustic Communication
Acoustic signals are also widespread throughout the animal kingdom, being
most commonly associated with birds, mammals, anuran amphibians (frogs
and toads), and insects. Many sounds, such as the chirping of crickets, the
songs of birds, the calling of frogs, and the roar or growling of mammals, are
audible to the human ear. Most humans can only hear sounds that fall into
the range of 20 to 20,000 hertz (Hz). However, in some cases animals produce
acoustic signals that are above or below this audible range and cannot be heard
by humans without the use of specialized equipment. For example, many
rodents produce very high-frequency (ultrasonic—inaudible to us) calls that
are important in parental care and sexual behavior (e.g., Ehret, 2005). In
another interesting case, Katy Payne, studying elephants in a zoo, noticed
behavior in elephants that suggested they were making sounds that she could
not hear. She went on to demonstrate that elephants communicate over long
distances by using very low-frequency (infrasonic) rumbling sounds that also
are inaudible to humans (Payne et al., 1986; Payne, 1998). In fact, by some
estimates, most of the sounds that elephants make are at such low frequencies
that they cannot be heard by our ears. Because some of these sounds are also
transmitted through the ground, they are sometimes classified as seismic
signals rather than as true sounds.
In some cases, animals may use objects to generate sounds. A well-know
example occurs in chimpanzees (Pan troglodytes). A variety of sounds used
during communication are produced by chimps slapping or stomping on the
ground or vigorously shaking or dragging tree branches (Muller & Mitani,
2005).
Acoustic signals tend to be extremely complex, and many different charac-
teristics of the sound may be important in communication. In addition to
ANIMAL COMMUNICATION 57

loudness, duration, and frequency (expressed in hertz), other parameters of the

sound, such the repetition rate, the number of pulses per second, the time
intervals between pulses, and the patterns of the different sounds that make
up the signal, can be critically relevant in animal communication. Different
species may pay attention to different acoustic parameters. Not surprisingly,
the study of the characteristics of sounds that are most important is an active
and fertile area of research (especially in frogs and birds) in animal
communication.

Chemical Communication
Chemical communication, which includes both olfactory and gustatory sig-
nals, is nearly ubiquitous, occurring throughout the animal kingdom, with the
possible exception of among birds (but see below). Because olfaction and taste
can be so difficult to distinguish, this chapter concentrates on olfaction and
the use of odors or scents as signals. Frequently, odors are produced by special-
ized glands, but urine, feces, and saliva can also contain relevant information.
An interesting characteristic of chemical signals is that, in many cases, they can
be deposited and broadcast throughout the environment even when the
sender is no longer present. One example you are probably familiar with is uri-
nation in dogs (Canis familiaris). Dogs will sniff and then repeatedly “mark”
certain trees—or the proverbial fire hydrants—and then move on. It appears
that each dog may be leaving a sort of “calling card” that keeps everyone
informed of who has been there and, possibly, how long ago (Wyatt, 2003).
An alternative explanation for this common dog behavior is that each dog
may be “erasing” the previous dog’s scent mark by urinating on top of it
(Wyatt, 2003). Also in dogs, everyone is likely familiar with the sex phero-
mone produced by female dogs when they are in heat. This odor is also broad-
cast and is highly attractive to any male dogs in the general vicinity (Beach &
Gilmore, 1949; Dunbar, 1977). Sex pheromones are important not only in
mammals but also in many insects, crustaceans, fishes, salamanders, reptiles,
and other taxa.
There often is a great deal of misunderstanding about the definition of
pheromone, and many people think that all pheromones function in sexual
behavior, but this is not correct. A pheromone is any chemical substance that
is secreted to the outside by an individual of one species and has an effect on
the behavior or physiology of a conspecific—that is, another individual of
the same species. While it is true that many pheromones are used in sexual
contexts, they can also function in other contexts, including in communicat-
ing alarm (as in the case of the honey bees mentioned earlier), social status,
and individual or group identity; in eliciting aggregations; or in creating scent
58 ANIMAL BEHAVIOR

trails. Moreover, the definition also makes it clear that chemical cues that act
between different species are not pheromones.
Bark beetles of the genus Ips are a good example of species that produce
pheromones used specifically for facilitating aggregations of conspecifics
(Young et al., 1973; Bakke, 1978; Byers, 1983). Ips attacks on pine trees can
result in terrible infestations that may kill hundreds or thousands of living
trees. When the first beetle finds a suitable tree (usually an older tree or one
that has been weakened by fire or disease), it immediately begins producing
an aggregation pheromone that attracts large numbers of conspecific beetles.
These beetles produce additional pheromone, resulting in the entire tree
becoming riddled with beetles burrowing under the bark and eating the
phloem, ultimately destroying the tree. Because of the commercial damage
that is wreaked by Ips infestations, as well as by other species of beetles that
also attack various species of trees, an enormous amount of research has been
done on their aggregation pheromones in the hopes of being able to control
outbreaks of these infestations.
Some experts on chemical communication limit the use of the word phero-
mone to cases in which all receivers respond in the same way and in a stereotyped
manner to the chemical signal. For example, all honey bees in a hive respond in
the same manner to the bee alarm pheromone, by increasing movement, search-
ing for the threat, and immediately attacking it. This is why a person who is stung
by a bee is likely to get multiple stings—the first bee that attacks releases an alarm
pheromone that alerts all the other bees, and they then descend on the unfortu-
nate person and continue to attack by stinging (however, the alarm pheromone
is quite complex, with different components eliciting different responses—see
Wager & Breed, 2000). According to this more restrictive definition, odors that
communicate individual identity in mammals are not pheromones because dif-
ferent receivers can respond differently to the same odor based on their previous
experience with the sender (friendly versus hostile, or familiar versus unfamiliar).
For the sake of simplicity, in this chapter I define a pheromone as an intraspecific
chemical signal, regardless of the type of response that it elicits.
Animals that use chemical communication may mark the substrate, objects
in their environment (e.g., trees, rocks, twigs), or their conspecifics or even
spread odors on themselves by self-grooming. As an example, during social
encounters, male neotropical singing mice (Scotynomys teguina) rub their front
paws on their abdominal scent gland and then allogroom females, spreading
the male secretions all over the females’ fur (Fernández Vargas et al., 2011).
The function of this marking of females is not well understood. Interestingly,
as their name implies, singing mice also produce acoustic signals in both
audible and ultrasonic ranges.
ANIMAL COMMUNICATION 59

Birds have always been considered to be a rather conspicuous exception to

the wide use of chemical communication because they are believed to have
poor olfactory abilities and are not know to use chemical signals. However,
recent work in Ellen Ketterson’s lab at Indiana University may change this
perception. Danielle Whittaker, a postdoctoral student of Ketterson’s, has
demonstrated that the preen oil of dark-eyed juncos (Junco hyemalis) varies
with population, individual identity, and sex (Whittaker et al., 2010). More-
over, both female and male juncos can distinguish the sex of an individual
from its preen oil and prefer the preen oil odors of males (Whittaker et al.,
2011). This ability to discriminate odors in this bird species may have impor-
tant effects on their parental behavior (including potentially on reproductive
success) because when preen oil was applied to the nests of females, they spent
less time incubating their eggs if the odors were from unfamiliar individuals
but not when they were their own odors (Whittaker et al., 2009). These are
very exciting findings that may help to change our assumptions about the
ability of some birds to communicate by means of chemical signals.

Tactile and Vibrational Communication

Tactile communication is very familiar to humans because we use it very
frequently. When you hug or put your arm around a friend’s shoulder to com-
fort him, you are using tactile communication—your gesture communicates
your sympathy and concern without a need for words. In nonhuman animals,
tactile communication can take many forms. Primates (monkeys and apes) use
many tactile signals that are similar to those used by humans. Additionally,
allogrooming in primates appears to serve dual functions—it removes parasites
from the skin and fur of the individual being groomed, and it also strengthens
and reinforces existing social bonds between the two individuals involved in
this interaction (Nakamichi & Shizawa, 2003; Majolo et al., 2005; Dunbar,
2010). In most primate species, subordinate individuals groom more domi-
nant ones, and the behavior also serves to acknowledge that the groomer
recognizes the higher social status of the partner (Seyfarth, 1980; Schino,
2001). Other taxa also may use touch, including grooming, licking, or strok-
ing, as part of displays that function in different contexts. Although there do
not appear to be any rigorous scientific studies, it is a common belief that can-
ids, including domestic dogs, use licking to reinforce friendly interactions and
social bonds; however, at present this is only a popular conjecture. On the
other hand, some insects, such as fruit flies (Drosophila), use licking as part
of their courtship displays (Bastock & Manning, 1955; Spieth, 1974). Some
insects, crustaceans, and spiders use tapping and stroking behaviors during
courtship and mating. One interesting example is found in the brown recluse
60 ANIMAL BEHAVIOR

(Loxosceles reclusa), a venomous spider found in parts of the southern Midwest

of the United States. Courting males and females approach one another very
slowly and then initiate a prolonged period of stroking motions; the male in
particular extends his long front legs over the female and gently touches and
strokes her prior to mating (Tang-Martínez, personal observation).
In some cases it is difficult to distinguish clearly between tactile and vibra-
tional signals. For example, in web-building spiders, males entering a female’s
web may be mistaken for prey and get eaten by the female (Robinson, 1982;
Elgar, 1991). Complicating matters even more, females in these species tend
to be huge in comparison to the males, and it is very easy for a female to over-
power a male. Consequently, courtship in these species is a very risky business,
and males must be very careful in communicating their intentions to the
female. Typically, males will stand at the edge of the web and begin slowly
plucking the web, using certain species-specific patterns that indicate court-
ship (Robinson, 1982; Maklakov et al., 2003). Depending on the female’s
response, the male may very slowly approach the female, continuing to pluck
as he does so. If all goes well, he will end up as a successful suitor rather than
becoming a tasty snack for the female! Traditionally, the plucking signal of
the male was considered a form of tactile communication. However, it can just
as easily be interpreted as a vibrational signal because the plucking creates
vibrations that are propagated through the web (Maklakov et al., 2003).

Multiple Modalities
I have discussed each modality of communication separately. However, in
real life, an animal may often use more than one modality simultaneously. I
have already mentioned bowerbirds using both acoustic and visual signals in
their courtship display. Grouse, among many other species of birds, also com-
bine visual and acoustic signals during courting behaviors—they strut or
stomp while conspicuously displaying bright colors on their heads, body, or
tails, while simultaneously making “booming” or drumming calls that can be
heard over surprisingly long distances (Dantzker et al., 1999). In some mam-
mals, scent communication is accompanied by conspicuous visual signals.
For example, South American bush dog (Speothos venaticus) females perform
a handstand as they urine-mark objects in their environment (Kleiman,
1972). Likewise, ring-tailed lemurs (Lemur catta) undulate their tails conspic-
uously at the same time that they release anal-gland scents; this seems to con-
stitute a visual signal, as well as perhaps helping to waft the odor towards the
intended receiver. Invertebrates, such as some jumping spiders, also combine
modalities, sometimes using tactile and chemical cues simultaneously or both
visual and vibrational cues (e.g., Girard et al., 2011).
ANIMAL COMMUNICATION 61

CONTEXTS OF COMMUNICATION
From the information provided above, you probably already have a pretty
good grasp of the types of situations in which animals use communication.
Communication is a sine qua non for sociality—that is, all animals that are
social by necessity must have some form of communication. However, even
solitary (not living in groups) animals must communicate for the purpose of
mating and reproductive behavior. (The only possible exception to this gener-
alization might be species that reproduce by cloning, are hermaphroditic, and
are not likely to encounter other individuals of their own species.)
In general, the contexts in which communication is used include, but are
not limited to, sexual behavior, agonistic behavior, parental care, group co-
ordination, and food and predators. I discuss these uses below.

Sexual Behavior
Signals used for reproduction include sex attractants (such as sex phero-
mones), sexual excitants (sometimes called aphrodisiacs), and displays that
help the male and female to coordinate their behaviors so that mating can
occur. In many species of crickets, for example, males produce two distinctly
different songs during courtship (Zuk et al., 2008). The first song, called the
calling song, attracts the female and is very different from one species to
another (e.g., Gray & Cade, 2000; Fitzpatrick & Gray, 2001); once the
female approaches, the male switches to a new song, called the courtship song,
which seems to excite the female, making it more likely that she will accept
him as a mate (but many courtship attempts, whether in crickets or other taxa,
are unsuccessful because females may refuse to mate even after assiduous
courtship on the part of males). Keep in mind also that, in some species, it is
the females that produce signals to attract males. For example, it is the female
in most moth species (e.g., silkworm moth [Bombyx mori], gypsy moth
[Portheria dispar]) that produces the sex pheromone that attracts males, some-
times from distances of 2 km or more (Jacobson & Beroza, 1963; Wilson &
Bossert, 1963). Likewise, in dogs it is the female that produces a powerful
sex pheromone to attract males (Beach & Gilmore, 1949; Dunbar, 1977).
As we saw in the case of the orb-web spiders, courtship behavior can be tricky,
and, in most species, both the male and female may risk injury or attack if they
proceed too quickly, before their potential mate is ready. Thus, courtship
often involves complex sequences of signaling behaviors by both the male
and the female. Sometimes these complex sequences are referred to as court-
ship dances, and these have been described in spiders, birds, fishes, and even
in cephalopods, such as squids (e.g., Sauer et al., 1997; Von Hippel, 2000;
62 ANIMAL BEHAVIOR

Trainer et al., 2002; Girard et al., 2011). In species that stay together for an
extended period after mating, courtship signals and activities may also help
to create and reinforce a pair bond—the two individuals will stay together
and interact in an amicable and nonaggressive manner and will cooperate in
activities that help to increase their reproductive success (e.g., in parental
care—see below).

Agonistic Behavior
In animal behavior we use the term agonistic to refer to interactions that
involve both aggression (offensive signals and behaviors) and submission
(defensive signals and behavior). The use of this specialized term makes sense
if we consider what is really happening when two individuals are involved in
what most nonspecialists would call an aggressive encounter. Imagine that
you are physically fighting with another person—most likely your feelings
(as well as the feelings of your opponent) will fluctuate between aggression
and fear because you cannot be sure that opponent will not be a better fighter
and beat you up. Your behavior is likely to go from advancing and threatening
your opponent one instant to backing off and trying to protect yourself the next.
Thus, you and the opponent are continuously going back and forth between
aggression and submission. Thus, aggression only addresses one aspect of what
happens in these types of interactions, and agonism becomes a more accurate
word. Agonism can occur in many different contexts. Individuals, whether male
or female, may compete for mates, food, territories, resources, or shelters,
among other reasons. Sometimes competition can involve actual physical fights,
but in many cases animals use only ritualized displays (see below) and signal
until one individual backs down and the other gains access to the resource.
In almost all species, signals and displays that indicate threat or aggression
have similar characteristics. They make the animal look bigger, and weapons
are exposed. A direct stare also is considered a signal of aggression in many spe-
cies. Think of a dog that is behaving very aggressively—it will stand tall with
head held high, ears will be perked straight up, hackles on the back will be
raised, tail will be up, teeth will be bared in a snarl, and the dog will likely stare
directly at its opponent. By contrast, a dog showing submission will cower
(body pressed close to the floor), ears will be pressed back against the head, tail
will be hidden between the legs, the teeth will be hidden behind a grimace,
and the head (and eyes) will be turned directly away from the opponent. In
an extreme case, the submissive dog might even fall down to the ground com-
pletely and roll over on its back. Both of these submissive displays make the
dog look as small and as nonthreatening as possible. The aggressive and sub-
missive displays are virtually mirror images of each other. We call such displays
ANIMAL COMMUNICATION 63

that are exact opposites of one another antithetical displays because the pos-
ture in one case is the antithesis (exact opposite) of the other (Darwin,
1872). Images for antithetical aggressive and submissive displays in dogs and
discussion by Darwin (including also other images of dog and cat postures)
can be found at http://www.brocku.ca/MeadProject/Darwin/Darwin
_1872_02.html.
I used the example of dogs because it is probably familiar to many of you.
However, the same types of behaviors are seen in many other species across
most of the animal kingdom. Ungulates, such as deer or antelope, may stand
very tall, stare at their opponents, and swing their antlers or horns (weapons)
threateningly in the air. Fishes expand their gill covers, making them look
much bigger from the front; toads may puff up their bodies, making them-
selves appear alarmingly larger than they really are. Invertebrates also make
themselves look larger and may display their weapons. For example, mantis
shrimp (marine stomatopod crustaceans) rear up and perform “meral spreads”
by opening and displaying their brightly colored raptorial appendages (power-
ful weapons that can smash down on an opponent, causing serious injury or
death) in a very conspicuous threat display (Caldwell & Dingle, 1975). And,
of course, the submissive displays tend to be opposites of the threat displays.
See also section “Typical- and Variable-Intensity Displays,” below, for more
information on agonistic signals and displays.

Parental Care
Not all species of animals show parental care, but among those that do,
communication looms big. Signals relevant to parental care include the beg-
ging or squeaking calls of baby birds and mammals, which generally garner
the attention of the parents, who may then feed or protect their young. Larvae
of some social insects (e.g., social bees and wasps) use special signals to beg
food from the adults (Kaptein et al., 2005). Recognition between parents
and their offspring can be critical to the survival of the young, and it is not sur-
prising that many species in many different taxa have evolved signals that aid
in this recognition. Young who become lost or isolated from their parent(s)
typically produce signals that attract the parent and lead to the retrieval of
the offspring (Smotherman et al., 1974). In birds, calls are often used for
parent-offspring identification (reviewed in Halpin, 1991). In many other
taxa, including some fishes, salamanders, and mammals, chemical signals play
a similar role (reviewed in Halpin, 1991). Parents may also warn their off-
spring of danger. A case in point is a group of cichlid fishes (family Cichlidae)
called mouthbrooders. In most of these species, the female carries the fertilized
eggs in her mouth until the young hatch. (During this time, swallowing,
64 ANIMAL BEHAVIOR

coughing, and yawning are mostly suppressed so that the female does not acciden-
tally swallow or disgorge the developing young.) After the fry (baby fish) hatch
they continue to associate with the mother and return to her mouth to sleep or
when they are in danger. In fact, the mother and offspring communicate with vari-
ous signals to ensure that the fry return the mother’s mouth when a threat is per-
ceived (Baerends & Baerends-Von Roon, 1950; Balshine-Earn & Earn, 1998).

Group Coordination
Among social species, signals frequently serve to keep the group together or
to coordinate behaviors. For example, in many species of primates, as mem-
bers of the troop spread out to forage (whether in savannahs or in trees), indi-
viduals will emit contact calls to keep in touch with the group and allow
others to determine where all group members are (e.g., Sugiura, 1998). Special
signals that may be visual, acoustic, or chemical are also used by some species
to initiate particular behaviors, such as group movement from one place to
another or a coordinated attack on prey (e.g., Arnold & Zuberbühler, 2006).
Among the truly social insects (termites, ants, social bees and wasps) chemical
communication controls much of the group’s behavior, ranging from parental
care to alarm behavior to (in some cases) coordinated attacks on other groups.
Among the army ants (e.g., genus Eciton), daily raids away from the nest or
nomadic movement from one nest to another are exquisitely coordinated, pri-
marily by chemical cues (Deneubourg et al., 1989). Thousands of ants literally
march in files over the landscape (ground, branches, etc.), attacking and eating
everything in their path, while simultaneously maintaining extraordinary
group cohesion and order. I was lucky to witness such a raid on a visit to Costa
Rica in 2012; this was an amazing experience because (at the risk of being
anthropomorphic) the ants gave the impression of being unstoppable and
moved as if they had one mind and one common goal.

Food and Predators

Many species use signals to communicate the existence or location of food
and to warn conspecifics against predators (but note that there is some debate
as to the function of so-called alarm signals in some species; see the section
“Interspecific Communication” for a few alternative possibilities).
Signals related to the presence of food most frequently occur in species that
are social or live in aggregations. We have already discussed the aggregation
pheromone of bark beetles, which alerts other beetles that a suitable tree (food)
has been located. Treehoppers generate vibrational signals that propagate
along plant stems and announce that a good source of food has been
ANIMAL COMMUNICATION 65

encountered (Cocroft, 2005). Tent caterpillars lay chemical trails that lead
other nestmates to food and then back to the nest (Fitzgerald & Webster,
1993). And when ants invade your home, think pheromones! Foraging ants
leave scent trails that guide other ants to food—for example, on your kitchen
counter (Morgan, 2008). As a result, the best way to get rid of the ants is to
destroy their pheromone trails (unfortunately, this is not always easy). Birds
and mammals will often use special vocalizations to alert group members to
the presence of food (e.g., Caine et al., 1995; Elgar, 1986; Elowson et al., 1991).
Perhaps the most amazing example of communication about food is the
waggle dance of the honey bee (Von Frisch, 1967). Worker bees (all females)
belong to different castes, each one specialized for different tasks. Some work-
ers are “scouts” whose job is to leave the hive, find food, and then come back
to the hive and signal to the foragers where the food is. Foragers then follow
the “instructions” of the scouts, find the food, and bring it back to the hive.
After initially finding food, the scouts perform a waggle dance on a vertical
surface inside the hive, and the foragers follow the dance by maintaining con-
tact with the scouts as they dance (the inside of the hive is dark, so tactile cues
are used). The waggle dance resembles a flattened figure eight with the scout
bee going around one circle and then the other, waggling its abdomen as it
moves through the straight portion (called the straight run) between the two
circles. As the scout dances, the angle of the straight run in relation to straight
up on the vertical indicates the exact direction of the food in relation to the
position of the sun outside. Note that the straight run does not directly point
to the food but rather indicates where the food is in relation to the sun; if the
sun is not visible outside (as in a completely overcast sky) the foragers will not
be able to use the waggle dance to determine direction. In addition to indicat-
ing the direction of the food, scouts also indicate the distance to the food by
how many waggles they perform during the straight run (the more waggles,
the longer it takes to do the straight run and the farther away the food is)
and the quality or richness of the source by how long they continue perform-
ing the waggle dance (see Volume 1, Chapter 12). Interestingly, if the food is
very close, scouts do not perform the waggle dance; instead, they perform a
“round dance” that consists of moving around and around in a simple circle.
Foragers leave the hive and fly in concentric circles in the vicinity of the hive
until they locate the food. Images of the honey bee waggle dance and round
dance can be found in these links.

http://media.wiley.com/mrw_images/els/articles/a0002940/image_n/
nfg002.gifhttp://images.tutorvista.com/content/biotic-community/honey
-bee-worker-waggle-dance.jpeg
66 ANIMAL BEHAVIOR

http://1.bp.blogspot.com/_pOW4pcoh2Z0/Sun4XSUM4PI/AAAAAAAAA
FE/0roWi1owCSc/s320/honeybees.jpg
Interestingly, the importance of the waggle dance has recently been chal-
lenged (Grüter & Farina, 2008, 2009). These authors claim that forager bees
that follow the dances often disregard the direction and location information
encoded by the dances. They suggest that the information on the location of
food sources (i.e., spatial or navigational information) is used primarily as a
back-up if other sources of information are not available or are inconsistent.
In response, Axel Brockmann and Moushumi Sen Sarma (2009) argue that
just because honey bees may use a variety of information (e.g., food odors
brought back by scouts or their own previous experience), it does not mean
that the waggle dance is not important or that the information it conveys is
not relevant. Specifically, they suggest that the location information encoded
in the waggle dance may be particularly important when food is patchily dis-
tributed and difficult to find, as is often the case in the tropics and subtropics
where eight of the nine species of honey bees are found and where the waggle
dance likely evolved.
Interestingly, this is not the first time that the conventional understanding
of the waggle dance has been challenged. As far back as the 1960s, several sci-
entists questioned whether the waggle dance really serves as a “language” that
communicates location and/or distance to the food sources; instead they sug-
gested that bees were using odors (rather than the information in the waggle
dance) to find food (e.g., Johnson, 1967; Wenner, 1967). Von Frisch
(1967) countered this criticism by pointing to a number of his own experi-
ments that seemed to show very clearly that bees do, in fact, use the dances
to locate food. Subsequently, James Gould (1975) conducted a series of
elegant and exquisitely controlled experiments to test the assertions of the crit-
ics. He concluded that under different conditions (e.g., stress level of the bees
and the abundance of flower sources the bees are using), honey bees could use
either the information in the waggle dance or odors. However, he also found
excellent evidence that bees are able to gain information from the dance and
that, even when odors are used, the waggle dance is necessary to recruit the
foragers. He also suggested that Wenner (1967) and Von Frisch (1967)
obtained different result and reached different conclusions because of differ-
ences in methodology that resulted in their concentrating on two different
stages of the recruitment process using the dance. Recently, Tania Munz
(2005) has summarized the history of this early challenge to the waggle dance.
The recurring debates over the waggle dance of honey bees is a good exam-
ple of how science works. Karl Von Frisch’s discovery of the waggle dance
ANIMAL COMMUNICATION 67

more than 60 years ago won him the 1973 Nobel Prize in Physiology or Medi-
cine. Moreover, the waggle dance has remained one of the most celebrated
examples of complex communication in animals and appears in every textbook
on animal behavior. Yet even the most widely accepted ideas in science can be
challenged, tested, and revised—and this is an integral and necessary part of
the practice of science. My own opinion of the current debate is that Brock-
mann and Sen Sarma (2009) are correct in their criticism of Grüter and Farina
(2009), but it also is likely that honey bees use a variety of cues when foraging
and are not limited only to the information encoded in the waggle dance. This
should not be surprising because most animals do show flexibility in their
behaviors—and, of course, this is the same conclusion that Gould reached
in 1975.
The alarm pheromones of honey bees already have been discussed. Many
other species produce alarm signals in the presence of potential predators. Best
known are the alarm calls of birds and of mammals. These calls may elicit
alertness (e.g., meerkats stand on their hind legs and look around for the
predator [Manser, 1999]), evasive behaviors (e.g., ground squirrels immedi-
ately run and hide in their burrows [Leger & Owings, 1978]), or attack behav-
iors (e.g., birds may gather and gang up on the predator, a behavior know as
mobbing [Grieser, 2009]). Fishes in some species, when attacked by a preda-
tor, release an alarm substance that causes other fishes nearby to quickly swim
away from the area of the injured fish (Chivers & Smith, 1994).
These are certainly not the only contexts in which animals use signals and
displays for communication. However, they are likely the most common.

INFORMATION CONTENT
Peter Marler (1961), one of the pioneers in the study of animal communi-
cation, proposed that most signals carry information about five different
aspects of the sender: (1) species identity, (2) individual identity, (3) sexual
identity, (4) physiological condition/motivation, and (5) information about
the environment.

Species Identity
Most communication is intraspecific, and most species produce signals that
differ from those of even other closely related species (e.g., Roelofs & Comeau,
1969; Muroyama & Thierry, 1998; Fitzpatrick & Gray, 2001; Guillete et al.,
2010). Thus, almost by definition, most signals will carry information about
the species of the sender. Depending on the situation, receivers may respond
to this information in different ways. For example, courting males and females
68 ANIMAL BEHAVIOR

may use these signals to determine whether their prospective mate is a member
of their own species (e.g., Blair, 1964; Roelofs & Comeau, 1969; Gray &
Cade, 2000). If it is not, then typically the courtship is aborted and the ani-
mals continue searching for an appropriate conspecific partner. Signals that
help prevent hybridization mistakes by identifying the species of the sender
are known as behavioral (or ethological) isolating mechanisms.

Individual Identity
Many signals (but not all) also contain information about the individual
identity of the sender (reviewed in Halpin, 1986; Vaché et al., 2001). Individ-
ual identity can be used in many different contexts. In many mammals, the
individual odors of nest mates are learned during early life, and later used to
discriminate between siblings and non-siblings (Halpin, 1991; Tang-
Martínez, 2001). Likewise, parents may learn the individual odors of their off-
spring and then use the information to restrict parental care to their own
young. Both of these cases are considered examples of kin recognition.
Among social species it is often important to discriminate between neighbors
and strangers, between friend and foe, and between individuals of higher and
lower social rank. Ground squirrels have been found to respond very differ-
ently to the odors of their neighbors as compared to those of conspecifics from
other social groups (Harris & Murie, 1982). Using playback experiments, it
has been demonstrated that birds can distinguish between the songs of their
territorial neighbors and unfamiliar intruders (e.g., Wunderlee, 1978; Lovell
& Lein, 2004). Many ungulates (e.g., sheep [Ovis aries] and cattle [Bos taurus])
seem to use a variety of different cues, including visual, acoustic, and chemical,
to determine the individual identity of conspecifics (e.g., Alexander & Shillito-
Walser, 1978; Barfield et al., 1994; Ligout & Porter, 2004; Ligout et al.,
2004). Even among invertebrates, mantis shrimp respond differently when
presented with the odor of an individual they have previously defeated, as
compared to that from an individual that has defeated them (Caldwell,
1985). In the first instance they approach the odor and perform meral dis-
plays; in the second, they immediately flee and hide. All of these examples
(and there are many others) imply that animal can extract information on
individual identity from the signals produced by conspecifics. Although some-
what different in nature, signals that convey information on individual iden-
tity may also allow for recognizing group members (e.g., Keil et al., 2012).
These types of signals are of particular importance in certain social species.
For example, social insect, such as some bees, are known to distinguish the
odors of hive members from those of intruders (Breed, 1998). In this case
the cue appears to hive-specific (a hive odor) and not an individual odor per se.
ANIMAL COMMUNICATION 69

Sexual Identity
Some signals contain information on the sex of the sender. In some cases,
the signal is produced by only one sex, which automatically provides informa-
tion on the sender’s sexual identity. For example, although in many tropical
bird species both males and females sing, in most North American songbirds
only males sing (Marler & Slabbekoorn, 2004; Slater & Mann, 2004).
In other species, specific signals indicate sex of the sender. Morphological
structures or colors may vary greatly between males and females in some spe-
cies (e.g., birds and insects), thus providing a clear visual signal that indicates
sex (e.g., Badyaev & Hill, 2003; Emlen et al., 2005; Ribak et al., 2009). Neo-
tropical singing mice are able to distinguish the sex of senders by their odors
(Fernández-Vargas et al., 2008). Many other examples exist involving the
various modalities of communication.

Physiological Condition/Motivation
The physiological condition of an individual will affect its motivation and,
consequently, its behavior. When we speak of physiological condition, most
frequently we are referring to the animal’s hormonal profile and neural pro-
cesses at any point in time. As examples, among vertebrates, high levels of tes-
tosterone in males are correlated with sexual and aggressive motivations and
behaviors; in females an increase in certain hormones and a concomitant
decrease in others bring about estrus and lead to the production of particular
signals and displays that indicate the female is in reproductive condition
(e.g., Vandenbergh, 1969; Dunbar, 1977; Fernández-Vargas et al., 2008).
The songs and visual displays performed by the males of many bird species
during courtship are signals that the male is physiologically ready and moti-
vated to mate (Fusani, 2008). The same signals, conveying the same informa-
tion (reproductive condition and motivation), may also signify threats when
directed at other males. Likewise, females use signals that indicate their will-
ingness to mate: female birds perform a conspicuous copulation solicitation
display in which the female crouches and elevates the tail, exposing the cloaca
(Harju, 1971; Yasukawa et al., 1987), and female rodents perform lordosis (a pos-
ture in which the female bends the body, raises its rump in the air, wiggles the ears,
moves the tail to one side exposing the vagina, and produces squeak-like vocaliza-
tions [Beach, 1976]). In both instances these displays signal sexual readiness and
are invitations for the male to mount. Sex pheromones produced by female mam-
mals and insects also are examples of signals that indicate readiness to mate.
Threat and submissive displays, respectively, carry the message that the
sender is motivated to attack or is more likely to behave submissively. During
70 ANIMAL BEHAVIOR

any agonistic encounter, the motivation of the two participants may vary rap-
idly, going from being very aggressive to being frightened and submissive.
These motivational changes can be signaled almost from second to second
by the displays of the opponents (see the section “Typical- and Variable-
Intensity Displays”).
Other motivational states, such as parental care or hunger (e.g., in offspring
that depend on parents for food) also fall into this category (Redondo & Arias
de Reyna, 1988). Of course, in general, every species will have its own signals
and displays to communicate various physiological and motivational states.
The previous discussions under the contexts of communication include many
other relevant examples.

Information about the Environment

This category primarily includes signals about food and predators. The
alarm signals of insects, fishes, birds, and mammals, some of which were dis-
cussed previously, are excellent examples. Both birds and mammals may also
encode information about the type of predator detected, specifically if the
predator is an aerial predator or a terrestrial predator (e.g., ground squirrels;
Sherman, 1977, 1985). The responses of receivers differ depending on the type
of predator. Some mammals, particularly ungulates, perform visual alarm dis-
plays (Bildstein, 1983). These include tail-flagging in deer and spronking or
stotting in various antelope species (Caro, 1986). When deer detect a potential
predator, they run while holding the tail straight up (with the underside show-
ing) and moving it quickly from side to side; typically the contrasting coloration
of the underside of the tail makes this tail-flagging particularly conspicuous.
During spronking or stotting, antelope that have detected danger run, then stop
and leap straight up into the air, come back down, run, and repeat the behavior.
Deer and antelope that observe these behaviors scan the surroundings, presum-
ably for predators, and then also take evasive actions (Caro, 1995).
The waggle dance of honey bees, the scent trails of ants, and the aggrega-
tion pheromones of bark beetles illustrate the complexity and variety of some
signals relating to presence and location of food. Other food-relevant signals,
for example in social vertebrates, may be much simpler and consist of calls
and other vocalizations.

Information about Status

In addition to the messages identified by Marler (1961). signals also may
convey information on status or quality of the sender. According to this view,
male signals used during courtship convey information about the quality of
ANIMAL COMMUNICATION 71

the displaying male (e.g., his health, survival potential, strength, or stamina).
and females use this information in selecting which male to mate with (Hill,
1991; Kayser & Hill, 2000). Likewise, signals used during agonistic encoun-
ters may not simply say, “I am feeling aggressive”; rather, they may communi-
cate the status or fighting ability (i.e., the quality) of the sender (Pryke et al.,
2001; Wyman et al., 2008). Note that in both cases, such signals would do
more than simply communicate physiological condition or motivation.

Metacommunication
Another interesting category includes signals that serve to modify the meaning
of other signals. The best-known example is the “play bow” of dogs (http://4
hpetpals.osu.edu/images/dogPlayful.jpg). When dogs signal that they want to
play they assume a particular posture: head down, forelegs stretched out in front,
rump in the air, tail wagging (Bekoff, 1995). The message is not only that the dog
wants to play but also that all behavior that follows will be done in play and is not
to be interpreted as real aggression. Thus, behaviors such as growling, barking,
lunging, or jumping on the play partner are modified into play behaviors rather
than maintaining their original threat function. Signals or displays that modify
the meaning of other signals or displays are considered an example of metacom-
munication—that is, communication about communication (Bekoff, 1995).

RITUALIZATION AND ORIGINS OF DISPLAYS

One question that early animal behaviorists struggled with was the evolu-
tionary origin of displays. Many displays are very exaggerated and stereotyped,
and one cannot help but wonder where they came from and what affected
their evolution. Because we cannot go back in time and actually see what dis-
plays were like in the past and how they changed over time, we can only infer
where different displays may have originated. Nonetheless, there is general
agreement that many displays originally arose from behaviors that were not
used for communication but rather had a different function in a different con-
text (Daanje, 1950; Tinbergen, 1952; Morris, 1956; Lorenz, 1972).
Researchers suggest that there are several possible noncommunicative sour-
ces for many of the displays that we are most familiar with: intention move-
ments, autonomic behaviors, displacement activities, ambivalent
behaviors, and redirected behaviors.

Intention Movements
Most animals, just before they perform a behavior, will engage in move-
ments and activities that are predictive of what the animal is about to do.
72 ANIMAL BEHAVIOR

A typical example is flight-intention movements in birds. Before a bird flies

off, it will bend its legs, extend it head and body forward, and open its wings.
This behavior has nothing to do with communication; the animal simply must
assume this posture in preparation for flight. Many courtship displays in birds
involve the male bending the legs, leaning forward, with head outstretched
and wings open. However, in these cases, the male simply holds this posture
in a very exaggerated manner while positioning himself in front of the female.
This display is believed to be derived from the intention movement for flight
(Daanje, 1950). However, the behavior has now become a stereotyped signal
used in communication.

Autonomic Behaviors
When animals are cold or scared, they may fluff up their fur or feathers
either to keep warm or as an automatic response to something frightening (this
is where the expression “I was so scared that my hair was standing on end”
comes from). Such behaviors are automatic physiological responses that are
not under the direct control of the animal; they have nothing to do with com-
munication. Now consider a dog that is signaling that it is highly motivated to
behave aggressively—chances are that its hackles (the fur on its shoulders and
upper back) will be raised. This “piloerection” is believed to be derived evolu-
tionarily from the piloerection that happens as an autonomic response
(Morris, 1956). Some scent marking behaviors (e.g., with urine or feces) are
likewise believed to be derived from normal urination or defecation (Morris,
1956). Similarly, erection of the penis, sometimes made more conspicuous
by bright coloration, is used by some primates as a threat display (Wickler,
1967). It has been proposed this signaling behavior is derived from automatic
erection of the penis during copulation.

Displacement Activities
During conflict situations, animals will sometimes suddenly stop what they
are doing and briefly perform another, apparently totally unrelated, behavior.
For example, when two birds are fighting, one of them might suddenly stop
and start preening itself for a few seconds before returning to the fight. These
behaviors are called displacement activities and are thought to function to
relieve high levels of tension that are likely to occur during conflict situations
(Zeigler, 1964). In some birds (e.g., some ducks) courting males include a dis-
play that consists of very conspicuously preening their wing feathers; other
related species simply point to the edge of the wing in an exaggerated and very
stereotyped display (Lorenz, 1972). The area that is preened or pointed to is
ANIMAL COMMUNICATION 73

likely to be very colorful, which makes the display even more striking. This is
considered an example of a courtship display that comes from a displacement
activity.

Ambivalent Behaviors
In conflict situations, when animals are undecided and their motivation
changes quickly from one instant to the next, they may perform ambivalent
behaviors. For example, a bird attempting to defend her eggs from an
approaching predator may charge the predator but then immediately run back
towards the nest; an individual in an agonistic encounter with a conspecific
may alternate between approaching its opponent and running away from it.
In all such cases it appears that the animals cannot decide what to do, so it
alternates between two mutually contradictory behaviors. A number of dis-
plays seem to have evolved from ambivalent behaviors. Niko Tinbergen
(1951) described the courtship and breeding behavior of the stickleback
(Gasterosteun aculeatus). Sticklebacks are small fish in which the male defends
a territory, builds a nest, and courts passing females. If a female is interested,
she will enter the territory, lay eggs in the male’s nest, and leave; the male stays
behind and takes care of the eggs. The courtship display of the male stickle-
back is called the “zig-zag dance” because the male will repeatedly alternate
between approaching the female and then swimming away from her, doing a
zig-zag towards and away from the female (Tinbergen, 1951). In gulls, there
is a “facing away” or “head flagging” display (Tinbergen, 1959). Again during
courtship, the male and the female rapidly alternate between looking at each
other and then quickly looking away, moving the head repeatedly back and
forth in an exaggerated and stereotyped manner. These displays are considered
to be modified ambivalent behaviors that have evolved into displays.

Redirected Behaviors
Again, in conflict situations, one or both participants may suddenly attack
objects in the environment (e.g., rocks or tree trunks) or even subordinate
bystanders. In such cases, it is assumed that the animal is directing aggression
at a “safe” target that is not capable of retaliating (Moynihan, 1955; Tinbergen,
1959). This in turn may relieve some of the aggression in the animal. During
both agonistic and courtship displays, some species use behaviors reminiscent
of redirected aggression. The previously discussed example of chimps slapping
the ground and vigorously shaking branches (Muller & Mitani, 2005) might
be another example of redirected aggression, although the exact motivation of
such displays is not really known.
74 ANIMAL BEHAVIOR

Ritualization occurs when behaviors that originally were used in a noncom-

municative context evolve into behaviors (displays) used specifically for commu-
nication. When this happens, we say that the behavior has been emancipated
from its original function and motivation (Huxley, 1923; Tinbergen, 1952;
Morris, 1957; Blest, 1961). For example, a displacement activity, originally used
to relieve tension during conflicts, evolves into a sexually motivated display that
functions to promote mating. During the process of ritualization, displays tend
to become increasingly exaggerated, stereotyped, and conspicuous. In some cases
anatomical changes, for example exaggerated structures or colors, also evolve,
making the display even more striking (Tinbergen, 1952; Morris, 1957).

TYPICAL- AND VARIABLE-INTENSITY DISPLAYS

Displays have three components: form, frequency, and duration. The form
of the display is its “appearance”—in the case of visual displays, it is what the
display actually looks like. The frequency refers to how many times the display
is performed in a given unit of time, and the duration is how long the display
lasts when it is performed.
Animal behaviorists distinguish between typical-intensity and variable
(“graded”)-intensity displays (Leyhausen, 1956; Morris, 1957). In a typical-
intensity display, the form of the display does not change with subtle changes
in the level of motivation of the sender. Images showing several typical-
intensity courtship displays (in three different species of birds) can be found
in the following links.

http://d3bdn46jsbygsl.cloudfront.net/144_snowy_egret.jpg
http://2.bp.blogspot.com/_bOKmjbY7wEo/S5gTdchiR4I/AAAAAAAACC0/
iJqC_i4OfdI/s320/Greater+Sage+Grouse.jpg
http://standingoutinmyfield.files.wordpress.com/2012/06/blue-bird-of
-paradise.jpg?w=529
For example, a courtship display consisting of the legs-bent, head-forward,
open-wings posture (described above) is displayed in exactly the same, very
stereotyped manner regardless of whether the sender is very highly motivated
or only moderately motivated to mate. And if, during the display, the sender
suddenly becomes more highly motivated, the appearance of the display will
not change. Levels of motivation can, nonetheless, be communicated by the
frequency or duration of the display. Consider an analogy of a doorbell. The
sound of the doorbell never changes, regardless of whether the bell-ringer is
very anxious to get you to come to the door or is very patient. The sound of
ANIMAL COMMUNICATION 75

the doorbell is the form. For example, a laid-back bell-ringer will ring the
doorbell once for a relatively short duration and then wait. But a bell-ringer
who is very anxious to get your attention can ring the doorbell repeatedly in
a short period (increase in frequency) or can press the doorbell button and
keep it pressed so that the doorbell rings for an extended period (increase in
duration). This is exactly what happens in a typical-intensity display: the form
is constant regardless of the exact motivational level of the sender, but level of
motivation can still be communicated by the frequency or duration of the dis-
play. Although there are no set rules, many courtship displays tend to be of
typical intensity, with the form communicating only a motivation to mate
(examples in Morris, 1957; Hinde, 1970).
In contrast, in variable- or graded-intensity displays, the sender’s level of
motivation is indicated by continuous changes in the form of the display
(e.g., Leyhausen, 1956). Images of variable- or graded-intensity displays can
be found in the following links.

http://rint.rechten.rug.nl/rth/dennen/agr_fear1.jpg shows the continuum

from extreme fear to extreme aggression in the facial expression in dogs.
http://rint.rechten.rug.nl/rth/dennen/agr_fear2a.jpg shows the continuum
from extreme fear to extreme aggression in the body postures of cats.
For example, consider antithetical displays in a dog involved in an agonistic
interaction. When the dog is highly aggressive it stands tall, ears up and tail up;
when it is highly submissive, it cringes with the body lowered, the ears laid back
flat against the head, and the tail tucked between its legs. But as the dog’s motiva-
tion fluctuates between high levels of aggression and high levels of fear (submis-
sion), these postures will form a continuum with, for example, the tail going
back and forth between being held high to an intermediate position, to down near
the rump, to completely tucked between the legs. Therefore, even subtle changes
in the sender’s motivation can be communicated by changes in the form of the
display. Changes in levels of motivation can still also be indicated by frequency
and duration. Going back to the doorbell analogy, this would be similar to having
a bank of doorbell buttons that vary gradually in the tone (form) depending on
how anxious the ringer happens to be. In this case, the bell-ringer can select which
button to use depending on her level of motivation and can go back and forth
among the buttons as her motivation changes. Again, she can also indicate her
level of motivation by varying the frequency or duration of ringing, regardless of
which button she happens to be pushing. Many agonistic displays are of variable
intensity with the form of the display changing rapidly as the motivation of the
sender changes (examples in Leyhausen, 1956; Hinde, 1970).
76 ANIMAL BEHAVIOR

In general it can be argued that typical-intensity displays convey clear and

unambiguous information about the overall motivation of the sender (e.g.,
“I want to mate”). On the other hand, their form does not convey information
about second-to-second changes in the motivation of the sender (but remem-
ber that such changes can still be conveyed by duration and frequency of the
display). Variable-intensity displays are very rich in the information they com-
municate, indicating even very subtle changes in the motivation of the sender
(e.g., “I am really aggressive—oops, I am a bit more scared now—and now I
am really terrified and about to run away”). Some models suggest that agonis-
tic displays are variable precisely because aggressive motivation varies so much
during agonistic encounters (discussions in Wilson, 1975; Enquist, 1985;
Bradbury & Vehrencamp, 1998). However, the down side of such a continu-
ous and detailed flow of information is that the form of the display can change
so quickly that the meaning may be more likely to be misunderstood.

IS COMMUNICATION HONEST?
So far, all the information I have provided is considered the traditional or
classical perspective of animal communication. Richard Dawkins and John
Krebs (1978) argued that this perspective misses the point because, in their
view, what we have discussed as communication is really all about manipula-
tion and deception—animals will “lie” (give out misinformation) to attain their
goals. For example, they suggest that when an animal is engaged in an agonistic
interaction it will do whatever is necessary to defeat the opponent. A subordi-
nate individual has no interest in accurately communicating to the opponent
that it is frightened and submissive; instead it should “bluff” the opponent by
performing threat displays that signal high levels of aggression—the animal lies
to obtain its goal of outcompeting the opponent. Similarly, according to
Dawkins and Krebs, a male courting a female should not “tell the truth” about
his social status or quality—instead he should pretend to be of very high status
and quality even if, in reality, he is of inferior quality. The goal of the male is to
mate with the female, and he will lie to manipulate the female into accepting
him as a mate.
These ideas provoked a strong reaction. Robert Hinde (1981), a highly
respected British animal behaviorist, suggested that Dawkins and Krebs had
misrepresented the traditional view of communication and erected a ‘straw
man.’ Both Hinde (1981) and W. John Smith, another student of communi-
cation (Smith, 1977, 1986) argued that Dawkins and Krebs’s perspective did
not make sense from an evolutionary perspective. For example, lying and
deception could only evolve if there was little or no fitness cost to these
ANIMAL COMMUNICATION 77

behaviors. These ideas are illustrated by an example. Imagine an agonistic

interaction, and suppose that the subordinate individual lies and pretends to
be highly aggressive. If the truly dominant individual calls his bluff and
attacks, the costs to the sender may be extreme—he may be badly injured or
even killed. Thus the cost of lying should prevent its evolution. A number of
other researchers, writing both prior to and following Dawkins and Krebs’s
1978 paper, argued that the cost of signals makes them necessarily honest
(Zahavi 1975, 1977; Enquist ,1985; Grafen, 1990). Others pointed out that
if all communication truly is about lying and misinformation, then everyone
“knows” that this is the case and this strategy would not be effective—lies
would routinely be ignored (Wallace, 1973). Additionally, if some individuals
are successful at lying, then populations should evolve better mechanisms of
detecting lies such that lying could not become established in populations
(Wallace 1973; Smith, 1986). These and other arguments culminated with
an emphasis on the evolution of “honest signals.”
Among the scientists mentioned above, Amotz Zahavi, an Israeli scientist,
was arguably the most important because he proposed the handicap principle
(Zahavi, 1975, 1977), which essentially contends that signals are, by necessity,
honest. Consider the case of a courting male—assume a peacock displays his
very long and colorful tail as a signal of his superior quality, as compared to
males of inferior quality. Zahavi argued that the cost of producing the long,
colorful tail is very high—the male had to use energy to produce the tail, he
is more susceptible to predation because his tail makes him more conspicuous
to predators, and he is less able to maneuver and to escape predators or survive
storms. In other words, having such a long and colorful tail is a handicap, and
only males of the most superior quality can afford to have such a handicap.
Inferior males simply do not have what it takes to produce the tail and still
survive, and therefore it is impossible for them to lie about their quality.
Therefore, argued Zahavi, the display is an honest indicator of male quality.
Although initially Zahavi’s ideas were criticized, in more recent times they
have gained wide support (e.g., Pomiankowski, 1987; Grafen, 1990).
Is there any evidence that animals do sometimes lie? This is a very difficult
question that can lead to philosophical, semantic, and anthropomorphic argu-
ments. For example, how do we define lying? How do we know if an animal
intends to lie? Is there a difference between lying in humans and in nonhuman
animals? Specifically, from a biological and evolutionary perspective, an animal
does not have to intentionally provide misinformation. Instead we can recon-
ceptualize the concepts of honesty and dishonesty of signals by thinking in
terms of reliable versus unreliable information. Thus an animal may provide
unreliable information without necessarily intending to do so. (See also the
78 ANIMAL BEHAVIOR

next section, “Do Signals Really Convey Information?”) In reality, we cannot

know if an animal “intends” to lie or deceive (because we have no way of deter-
mining what goes on in the mind of the animal, this question is not within the
realm of science). Thus, although many researchers continue to use the words
lying and deception, to avoid anthropomorphism and unwarranted assump-
tions, it is best to use the terms reliable versus unreliable information.
The traditional view of animal communication had always posited that sig-
nals usually provide only probabilistic and conditional information about the
motivation of the sender (Smith 1977, 1986; Hinde, 1981). Both of these
authors emphasize that communication is probabilistic (in the sense that sig-
nals indicate only that its probable—not certain—that a particular behavior
or event will follow) and conditional (signals indicate that senders’ behaviors
depends on how the receiver responds). Thus, in the example of the two ani-
mals involved in an agonistic encounter, if animal A performs a highly aggres-
sive threat display, it is not indicating to animal B that “no matter what you
do, I am definitely going to attack you and beat you up.” Instead, the message
is “if you do X (condition), there is a high probability that I will attack.” If B
does X, animal A may attack 70 percent of the time, but it may retreat the
other 30 percent; and if B does Y, animal A is much more likely not to attack
but may still do so (Hinde, 1981; Smith, 1986). Using this probabilistic and
conditional perspective, does it make any sense to say that animal A lied to or
deceived animal B because 30 percent of the time it does not attack even if B does
X? Hinde (1981) and Smith (1986) do not think so.
What is the empirical evidence? Most studies have yielded ambiguous results
that are open to interpretation (e.g., Rohwer, 1977; Rohwer & Rohwer,
1978). However, a few studies do suggest that, at least in some cases, animals
may misinform. In great tits (Parus major, a small European bird similar to
chickadees), dominant individuals have priority of access to feeding sites and
chase away subordinates. Anders Møller (1988) observed that when dominant
individuals were monopolizing the food, some subordinates would give alarm
calls, causing the dominants to fly away. The subordinates would then
immediately land at the feeding platform and eat until the dominants
returned. Although it is impossible to know what the subordinates were
“thinking,” these results can certainly be interpreted as a fairly clear case of
“crying wolf” to gain access to a food resource. But note that this strategy
should work only as long as it is not used too frequently. If subordinates con-
tinuously cry wolf, the dominants should quickly learn to ignore them because
they are not giving honest or reliable signals that a predator has been seen.
Interestingly, this behavior of giving false alarm calls to gain access to food or
other resources seems to occur in a number of avian and primate species.
ANIMAL COMMUNICATION 79

Some studies that have directly examined misinformation in situations

where it might be expected to occur have instead found that signals tend to
be honest. Hermit crabs (Pagurus bernhardus) use the shells of snails as a home
and shelter. Because availability of shells is limited, hermit crabs in poor-
quality shells (e.g.. shells that are too small for them) often try to evict other
crabs from better-quality shells. When a challenger approaches a resident crab
with the better-quality shell, the resident may perform a threat display to dis-
courage the challenger from approaching closer. But some residents do not
perform the threat display. Mark Laidre (2009) conducted an experiment in
which model crabs (dead crabs in a challenge posture and placed inside a shell)
challenged residents by approaching them. If threatened by the resident, the
model would either flee or continue to approach until it was very close to
the resident (of course the behavior of the challenger was remotely controlled
by the experimenter). In almost all cases, the resident’s response was consistent
with its behavior on first approach (i.e., the response indicated by the resi-
dent). That is, if a resident threatened and the challenger continued to
approach, the resident would almost always attack. If the resident did not
threaten when first approached, it would almost always back down and run
away if the challenger came closer. If the threatened residents had been “lying”
about their intention and willingness to attack, then they should have backed
down as soon as the challenger continued to approach. Moreover, crabs that
subsequently ran away should have also threatened to deceive the challengers
into thinking that they would be attacked. Therefore, this suggests that the
threat display in this species is an honest signal. Unfortunately, few other stud-
ies on honesty versus dishonesty have been so carefully designed and
controlled.

DO SIGNALS REALLY CONVEY INFORMATION?

In recent years a new controversy has arisen about whether responses to sig-
nals are due to information contained in the signals or are simply byproducts
of sensory physiology or other biological processes. For example, in the túng-
ara frog, males vocalize to attract females. It is well known that females in most
species in the genus Physalemus prefer males that produce elaborations of the
normal male vocalization (these can be added chucks, whines, or repetitions
of parts of the normal call) (Ryan, 1990; Ryan & Rand, 1993). What is most
interesting is that the males of some of these species do not normally produce
such elaborations; yet, when the elaborations are added to the normal species
calls, the females prefer these artificially modified calls to the normal calls. In
some species, males have, in fact, evolved more complex calls that include,
80 ANIMAL BEHAVIOR

for example, multiple whines or chucks. Because these elaborations do not

appear to correlate with any obvious male qualities, and females respond to
them for apparently arbitrary reasons (a preexisting sensory bias), it would
seem to make little sense to say that these vocalizations are signals that convey
any meaningful information about the sender (e.g., Ryan & Rand, 1993).
In another species, the water mite (Neumania papillator), courting males
attract females by producing vibrations in the water that seem to mimic the
vibrations produced by prey (Proctor, 1991), suggesting that they are exploit-
ing the females’ predatory response. This, and some other examples from vari-
ous species, have been considered examples of sensory traps (Christy, 1995)
by which males attract females (for mating) by producing signals that mimic
other stimuli occurring in different contexts to which females are adapted to
respond (e.g., responses to food or to predators). As in the case of the Physale-
mus frogs, such signals do not seem to provide accurate information to the
females.
Based on such examples, Michael Ryan and collaborators (Rendall et al.,
2009; Owren et al., 2010) subsequently offered a critique to communication
in general, claiming that there is little clear-cut evidence that supports the idea
that most signals convey information (as asserted by the more traditional views
of communication that I presented above). Instead they posit that the issue is
not whether senders signal information per se but rather that the signals influ-
ence or manage the behavior of the receiver in predictable ways (see also
Owings & Morton, 1997). In their view, the critical function of signaling is
for the sender to influence the behavior of the receiver and not the transmis-
sion of encoded information. There have been several responses to these ideas.
Robert Seyfarth and Dorothy Cheney (animal behaviorists who have done
extensive work on alarm vocalizations in monkeys), as well as several collabo-
rators (Seyfarth et al., 2010) offer a different perspective. They point out that
information is best defined as the “reduction of uncertainty in the receiver”—
that is, by perceiving the signal the receiver obtains more accurate knowledge
about its environment. In other words, if the signal has a reliable relationship
to a subsequent event or situation (e.g., alarm call equals “a predator is
nearby”), then it is accurate to say the signal encodes information and that
the receiver responds based on that information. It is irrelevant whether the
sender gave the alarm call with the “intention” of warning others because even
a signal that is produced as a result of arousal of the caller (e.g.. as a result of
fear when it detects a predator) can still contain information that receivers
can use. Specifically, they point out that in many species of birds and mam-
mals, callers give different alarm calls when they detect different types of pred-
ators, such as terrestrial versus aerial predators (see discussion above and
ANIMAL COMMUNICATION 81

examples in Seyfarth et al., 2010). Moreover, the receivers respond differently

to each type of call, and their avoidance responses are consistent with the type
of predator. Thus, Seyfarth and colleagues (2010) assert that alarm calls do in
fact contain information about the presence of a predator, the type of preda-
tor, and (in some cases) the level of threat posed by the predator. Thus, calls
inform listeners about the state of their world. These authors conclude that,
in general, animal signals encode real information and that the informational
approach to communication has been, and continues to be, valuable in
advancing our understanding.
Two other papers (Carazo & Font, 2010; Font & Carazo, 2010) reaffirm
that signals contain information but also offer a novel definition, contending
that information must be defined in terms of its adaptive significance to the
receiver. Thus, Pau Carazo and Enrique Font emphasize the evolution of sig-
naling and the selection pressures (costs and benefits) affecting both senders
and receivers; signaling systems evolve precisely because receivers are able to
make beneficial decisions based on the information they have obtained. They
also argue that signals always have meaningful information but that this does
not mean that they contain messages deliberately encoded by senders.
More recently, a theoretical paper (Ruxton & Schaeffer, 2011) attempts to
resolve some of these controversies by offering a series of new definitions that
the authors hope will clarify some of the ambiguities. Specifically, they suggest
that sender (“informer” in their terminology) and receiver (“perceiver” in their
terminology) engage in a trade; the informer influences the perceiver and, as a
result, the perceiver undergoes a change in what it knows about its world—
that is, it gains information as the result of the communication exchange.
The authors then go on to offer new definitions for “influence” and “informa-
tion” that they hope will resolve the debate.
In my opinion, it seems clear that in some cases (e.g., exploitation of preexist-
ing sensory biases, as in the case of the Physalemus frogs; sensory traps) it is diffi-
cult to argue that any meaningful information is being provided by the sender.
On the other hand, there are many cases where it is equally difficult to argue that
signals do not encode information and that no information is exchanged. I agree
that more careful and explicit definitions are necessary in order to avoid confu-
sion. However, perhaps it is also important to realize that, in some ways, there
is nothing special about animals being able to obtain information from conspe-
cific cues and signals. In reality, information is all around us at all times, whether
or not it is encoded in conspecific signals. For example, if we get up in the morn-
ing and the sky is heavily overcast and we hear thunder, we have information that
the probability of rain is high (a reduction of uncertainty about the kind of
weather that we can expect); certainly, no one would suggest that there is any
82 ANIMAL BEHAVIOR

intentionality involved. Likewise, a dog can urinate with no intention of encod-

ing information or communicating anything to another dog. But if a second
dog perceives the odor and can determine the sex and reproductive condition
of the urinator, then it has gleaned meaningful information from the odor. Like-
wise, an escaping prisoner has no intention of leaving behind any information,
but a bloodhound can obtain enough information from the odor left behind
and follow the prisoner’s trail. Although this discussion goes beyond animal sig-
nals and communication, I hope that it reinforces the point that intentionality
is not necessary for information transfer and that signals and cues produced by
conspecifics are only one way in which animals can obtain relevant information
from their environment. Perhaps placing animal signaling systems in a broader
perspective that emphasizes the acquisition and processing of information would
avoid some of the conceptual and semantic debates that have arisen. For addi-
tional discussions and perspectives on the issues of signal reliability and decep-
tion, see also an excellent book by William Searcy and Steve Nowicki (2005),

ENVIRONMENT AND THE EVOLUTION OF COMMUNICATION

Traditionally, animal behaviorists have always been aware of the impor-
tance of the environment to the evolution of signals and the sensory modalities
involved (Marler, 1961; Smith, 1977). For example, if we consider a fish spe-
cies that evolved and lives in very muddy water, we would not expect it to have
evolved visual signals because it is too difficult to see such signals in turbid
water. Instead, these fish would be predicted to use acoustic, chemical, or elec-
trical signals. The same is true of terrestrial species that inhabit areas of very
heavy undergrowth where vision is impeded. On the other hand, aquatic species
that live in bodies of water with rapid and unpredictable currents may be at a
disadvantage if they were to use chemical signals because these would quickly
be diluted and carried away by currents. Heavy vegetation may, under some
circumstances, absorb and change acoustic signals, thereby attenuating them;
in such cases one would expect species to use other modalities or to modify their
acoustic signals to increase their effectiveness. Background noise, such as from
heavy winds or even human traffic, can also interfere with acoustic signals;
Henrik Brumm and Hans Slabbekoorn (2005) have reviewed the many ways that
various taxa change their vocalizations to counteract the effects of noisy environ-
ments. Likewise, lizard species (genus Anolis) that live in windy areas with lots
of rapidly moving vegetation have been found to speed up their territorial
displays, which makes their displays more conspicuous (Ord et al., 2007).
John Endler (1992) proposed the concept of sensory drive, a complex
model that includes some of the ideas discussed above about effects of the
ANIMAL COMMUNICATION 83

environment on signal evolution. However, Endler defines the environment as

including both the physical environment and the presence of other animals,
whether conspecifics or not. Specifically, Endler argues that the evolution of sig-
nals, the sensory systems (including both sensory organs and brain areas needed
for processing signals) of receivers, and signaling behavior have coevolved. That
is, changes in one set of traits have affected the evolution of the other sets of traits.
Additionally, these traits also will be affected by the physical characteristics of the
environment and by ecological interspecies processes such as predation.
Among many examples, in the túngara frog mentioned previously, the sen-
sory organs of females appear to respond most strongly to certain frequencies
of sound; males have evolved calls in those frequencies, and females therefore
respond very strongly to these calls (Ryan & Rand, 1990). Phylogenetic analy-
ses suggest that the sensory sensitivity of the females existed first and that it
helped to shape the evolution of vocalizations in males (Ryan & Rand,
1993). Also in this system, bat-eating frogs use the vocalizations of males to
locate them (Ryan et al., 1982). According to the sensory drive model, over
time we should see modifications in male calls to counteract the ability of bats
to find them. The final result should be a compromise between being able to
effectively attract females and being able to avoid predation. Moreover, both
females and bats can be expected to respond evolutionarily to any changes in
male calls, setting up the potential for an “evolutionary arms race” between
male frogs and their bat predators. The physical environment can also affect
not only the structure of signals themselves but also signaling behavior. For
example, there are certain bird species in which males perform courtship dis-
plays in specialized courtship arenas within tropical forests (e.g., Anciães &
Prum, 2008). In these species, the males will choose to display only in micro-
habitats that have certain light conditions and at times of the day when their
displays will be most visible with regards to the light conditions. Increased
use of such specialized microhabitats for signaling may, in turn, affect the evo-
lution of sensory systems to make communication more effective.
In summary, sensory drive, which is currently an important area of study,
emphasizes that signals, signaling behaviors, and sensory systems do not evolve
in a vacuum. Rather they evolve jointly, and their evolution is further shaped
by the physical environment. The end result is signals, sensory systems, and
signaling behaviors that are highly effective in a given environment.

INTERSPECIFIC COMMUNICATION
The discussion up to this point has concentrated on intraspecific communi-
cation. The obvious question is whether communication can also occur
84 ANIMAL BEHAVIOR

between members of different species. The answer is yes, although most com-
munication does appear to be intraspecific. Nonetheless, interspecific commu-
nication has been demonstrated in several different contexts.
One context involves alarm calls. Several studies have shown that birds of dif-
ferent species respond to the calls of heterospecifics. For example, Magrath and
colleagues (2009) recorded normal alarm calls from three Australian species and
then conducted a playback experiments to determine the responses of the three
species to heterospecific calls. Two species, the superb fairy wren (Malurus cya-
neus) and the white-browed scrubwren (Sericornis frontalis), often flock together
and have common predators. Each species recognized the “hawk alarm calls” of
the other species and immediately sought cover. The situation with the third spe-
cies, a honeyeater (Phylidonyris novaehollandiae), was more complicated. Honey-
eaters only respond to scrubwren alarm calls, while fairy wrens and scrubwrens
both respond to honeyeater calls. Primates also are known to respond to hetero-
specific alarm calls. In one experiment, Klaus Zuberbühler (2000) demonstrated
that Diana and Campbell’s monkeys (Cercopithecus campbelli) reacted with alarm
to one another’s calls. Even more interesting is the case of a bird, the yellow
casqued hornbill (Ceratogymna elata), which responds to the “eagle alarm calls”
of the Diana monkey but not to “leopard alarm calls”; both crowned eagles
(Stephanoaetus coronatus) and leopards (Panthera pardus) prey on Diana mon-
keys, but only the eagle preys on hornbills (Rainey et al., 2004). Thus, the horn-
bill not only recognizes the alarm calls of a mammal but also discriminates
appropriately between calls that indicate two different categories of predator.
Another area in which interspecific communication may be important is
prey-predator interactions. In fact, some researchers have suggested that alarm
calls in birds and mammals, and behaviors such as stotting and spronking in
ungulates, actually communicate information to the predator. Alarm calling
or spronking may be pursuit-deterrence signals, although this notion remains
relatively controversial (Caro, 1995). The idea is that the sender is “saying” to
the predator one of two things: (1) “I have seen you and am taking evasive
action—you can’t surprise me, so you are wasting your time pursuing me”
or (2) “I am of such high quality that I can afford the risk of letting you know
that I am here; I am of such good quality that I can escape, so you are wasting
your time pursuing me.”
Although there has been a great deal of discussion about pursuit-deterrence
signals and behaviors, few studies have demonstrated that this type of commu-
nication actually occurs. One exception is a series of experiments conducted
by Manuel Leal (1999) on a Puerto Rican lizard (Anolis cristatellus) and its
snake predator (Alsophis portaricencis). Leal used carefully controlled experi-
ments to demonstrate that the lizard displays when approached by the snake
ANIMAL COMMUNICATION 85

and that the display frequency of the lizard is positively correlated with its
stamina in a running apparatus. This result suggests that lizards that display
at higher frequencies are also more likely to be able to outrun the snake. More-
over, other observations and experiments indicate that when Alsophis observe a
lizard displaying, they may be less likely to attack (Leal & Rodriguez-Robles,
1995). There are, however, some problems with this latter study because of
small sample sizes and the inability to adequately control the hunger level
and motivation of the snakes (Manuel S. Leal, personal communication).
Nonetheless, this is one of the most convincing studies of possible prey-
predator communication. Additionally, it provides evidence that the
frequency of the lizard display may be an honest signal, accurately indicating
running ability and the likelihood of escaping from a snake that attacks.
Another interesting example of what appears to be a pursuit-deterrence sig-
nal is described by Matthew Barbour and Rulon Clark (2012), in a ground
squirrel–rattlesnake system. Under natural field conditions, California ground
squirrels (Otospermophilus beechei) tail-flag (conspicuously wave their tail from
side to side) when they detect a snake. Free-ranging rattlesnakes (Crotalus ore-
ganus oreganus), which normally prey on these squirrels, are less likely to strike
and more likely to leave the area when an adult squirrel approaches them and
tail-flags. The authors conclude that tail-flagging possibly communicates that
the squirrel is vigilant; this might indicate to the snake that the squirrel will
not be an easy target; the snake may then go to another area where it might
have better luck catching a less vigilant squirrel. Although many questions
remain, it is also possible that tail-flagging is simply a “nervous” gesture by
squirrels that have seen a snake; however, if both snakes and squirrels are able
to extract relevant information, the behavior could have two effects: deterring
snake attacks and alerting other squirrels to the presence of danger. Not all
responses given by prey in response to predators always have such clear-cut
beneficial results. For example, Jan Randall and Marjorie Matocq (1997)
found that footdrums produced by banner-tailed kangaroo rats (Dipodomys
spectabilis) frequently deter predator gopher snakes (Pituophis melanoleucus
affinis) from stalking and attacking the footdrumming individual; higher
amounts of footdrumming resulted in decreased rates of stalking. The conclu-
sion is again that footdrumming acts as a deterrent by informing the snake that
its intended prey is alert and vigilant. However, the exact results depended on
the hunger level of the snakes; nonhungry snakes were more likely to avoid the
footdrumming, while hungry snakes investigated and came closer. Moreover,
because kangaroo rats also use footdrumming in territorial displays, it appears
that snakes that are very hungry may actually eavesdrop on the territorial dis-
play and use the footdrums to locate and capture prey. Thus, this is a case in
86 ANIMAL BEHAVIOR

which there are not only important benefits but also potential costs even if the
signal in some contexts serves to deter predators.
Finally, mutualistic interactions often require interspecific communication.
Here I discuss only two examples among many. Honeyguides (e.g., Indicator
indicator) are African birds that are known to guide honey badgers (Mellivora
capensis) and humans to bee colonies (Dean & MacDonald, 1981; Isack &
Reyer, 1989). The honeyguide, by making certain calls, communicates that
it has found a beehive and recruits honey badgers or humans to follow it.
Once at the hive, the badger or human tears into the hive and takes the honey;
the honeyguide, which by itself is not capable of breaking open a beehive, also
benefits because it obtains food in the form of wax and exposed larvae.
One of the most amazing mutualisms involves the cleaner fish (e.g., Lab-
roides dimidiatus). These fishes, which are very small and live in coral reefs,
set up “cleaning stations” attended by many different species of larger fish,
including many large and voracious predators. “Customer fish” line up and
wait their turn to be cleaned. Labroides starts by picking off dead cells and par-
asites from the body of the “customer” but eventually enters the mouth to
remove debris and decaying food from between the teeth, entrance to the gills,
and throat of the customer (Feder, 1966; Côté, 2000). During this time, the
large predator fish rests in a trance-like state, with its mouth wide open while
the cleaner goes in and out. Clearly, such a potentially dangerous interaction
(especially for the cleaner) requires unequivocal communication between
cleaner and customer. As might be expected, the cleaners and customers have
evolved certain signals to ensure that the interaction is carried out in a mutu-
ally beneficial manner (Feder, 1966; Côté, 2000). Both cleaner and customer
benefit from this mutualistic interaction because the cleaner gets food (debris,
parasites) and the customer is rid of noxious pathogens and other irritants.

CONCLUSIONS
In this chapter I have attempted an introduction to a very vast field of study
in animal behavior. In addition to the many areas that I have covered there are
many others that I simply could not include because of space limitations. Sim-
ilarly, my examples are not exhaustive, and many other, equally fascinating
examples exist for virtually every topic that I discuss. My goal was to give
you a general idea of what we mean by animal communication, introduce
you to classical concepts and perspectives, and familiarize you with some of
the more modern developments and debates. If nothing else, I hope that after
reading this chapter you have become aware of the complexity of this field and
of the need for much more research in this area.
ANIMAL COMMUNICATION 87

ACKNOWLEDGMENTS
I am grateful to Ken Yasukawa for inviting me to write this chapter and for
his careful reading of the original manuscript and many helpful suggestions for
improvement and on references. I also thank the members of the University of
Missouri–St. Louis’s Animal Behavior Discussion Group, including most spe-
cially Aimee Dunlap and Stan Braude, for illuminating discussions on the
issue of signals, information, and intentionality. Gabriel Francescoli gener-
ously e-mailed me a package of relevant references that were invaluable. I also
thank Steve Glickman, Roy Caldwell, and the late George Barlow for encour-
aging and nurturing my interest in communication ever since my graduate
school days. As always, I am grateful to Arlene Zarembka for her support
and patience during the time that I was completely engrossed in writing this
chapter, and most especially for her assistance in cross-checking references
and proofing parts of the manuscript.

REFERENCES AND SUGGESTED READING

Alexander, G. & E. Shillito-Walser. (1977). Importance of visual cues from various
body regions in maternal recognition of the young in Merino sheep (Ovis
aries). Applied Animal Ethology, 3, 137–143.
Anciães, M. & R. O. Prum. (2008). Manakin display and visiting behaviour: A com-
parative test of sensory drive. Animal Behaviour, 75, 783–790.
Arnold, K. & K. Zuberbühler. (2006). Language evolution: Semantic combinations
in primate calls. Nature, 441, 303.
Badyaev, A. V. & G. E. Hill. (2003). Avian sexual dichromatism in relation to phy-
logeny and ecology. Annual Review of Ecology and Systematics, 34, 27–49.
Baerends, G. P. & J. M. Baerends-Von Roon. (1950). An introduction to the study of
the ethology of cichlid fishes. Behaviour Supplement, 1, 1–142.
Bakke, A. (1978). Aggregation pheromone components in the bark beetle Ips acumi-
natus. Oikos, 31, 184–188.
Balshine-Earn, S. & D. J. D. Earn. (1998). On the evolutionary pathway of parental
care in mouth-brooding cichlid fish. Proceedings of the Royal Society of London,
B, 265, 2217–2222.
Barbour, M. A. & R. W. Clark. (2012). Ground squirrel tail-flag displays alter both
predatory strike and ambush site selection behaviours of rattlesnakes. Proceed-
ings of the Royal Society of London, B, 279, 3827–3833.
Barfield, C. H., Z. Tang-Martínez, & J. M. Trainer. (1994). Domestic calves (Bos
taurus) recognize their own mothers by auditory cues. Behaviour, 97, 257–264.
Bastock, M. & A. Manning. (1955). The courtship of Drosophila melanogaster. Behav-
iour, 8, 85–110.
Baylis, J. R. (1974). The behavior and ecology of Herotilapia multispinosa. Zeitschrift
für Tierpsychologie, 34, 115–146.
88 ANIMAL BEHAVIOR

Beach, F. A. (1976). Sexual attractivity, proceptivity and receptivity in female mam-

mals. Hormones and Behavior, 7, 105–138.
Beach, F. A. & R. A. Gilmore. (1949). Responses of male dogs to urine from females
in heat. Journal of Mammalogy, 30, 391–392.
Bekoff, M. (1995). Play signals as punctuation: The structure of social play in canids.
Behaviour, 132, 419–429.
Bildstein, K. L. (1983). Why white-tailed deer flag their tails. American Naturalist,
121, 709–715.
Blair, W. F. (1964). Isolating mechanisms and interspecies interactions in anuran
amphibians. Quarterly Review of Biology, 39, 334–344.
Blest, A. D. (1961). The concept of ritualization. In W. H. Thorpe & O.L. Zangwill
(eds.), Current Problems in Animal Behaviour (pp. 102–124). Cambridge, UK:
Cambridge University Press.
Boch, R. & D. A. Shearer. (1971). Chemical releasers of alarm behavior in the honey-
bee, Apis mellifera. Journal of Insect Physiology, 17, 2277–2285.
Borgia, G. (1986). Sexual selection in bowerbirds. Scientific American, 254, 92–101.
Bradbury, J. W. & S. L. Vehrencamp. (1998). Principles of Animal Communication.
Sunderland, MA: Sinauer Associates.
Breed, M. D. (1998). Recognition pheromones of the honeybee. BioScience, 48,
463–470.
Brockmann, A. & M. Sen Sarma. (2009). Honeybee dance language: Is it overrated?
Trends in Ecology and Evolution, 24, 583.
Brumm, H. & H. Slabbekoorn. (2005). Acoustic communication in noise. Advances
in the Study of Behavior, 35, 151–209.
Byers, J. A. (1983). Sex-specific response to aggregation pheromone: Regulation of
colonization density in the bark beetle Ips paraconfusus. Journal of Chemical
Ecology, 9, 129–142.
Caine, N. G., R. L. Addington, & T. L. Windfelder. (1995). Factors affecting the
rates of food calls given by red-bellied tamarins. Animal Behaviour, 50, 53–60.
Caldwell, R. L. (1985). A test of individual recognition in the stomatopod Gonadac-
tylus festae. Animal Behaviour, 33, 101–106.
Caldwell, R. L. & H. Dingle. (1975). Ecology and evolution of agonistic behavior in
stomatopods. Naturwissenschaften, 62, 214–222.
Carazo, P. & E. Font. (2010). Putting information back into biological communica-
tion. Journal of Evolutionary Biology, 23, 661–669.
Caro, T. M. (1986). The functions of stotting in Thompson’s gazelles: Some tests of
the predictions. Animal Behaviour, 54, 1147–1154.
Caro, T. M. (1995). Pursuit deterrence revisited. Trends in Ecology and Evolution, 10,
500–503.
Chivers, D. P. & R. J. F. Smith. (1994). Intra- and interspecific avoidance of areas
marked with skin extract from brook sticklebacks (Culea inconstans) in a natu-
ral habitat. Journal of Chemical Ecology, 20, 1517–1523.
Christy, J. H. (1995). Mimicry, mate choice, and the sensory trap hypothesis. Ameri-
can Naturalist, 146, 171–181.
ANIMAL COMMUNICATION 89

Cocroft, R. B. (2005). Vibrational communication facilitates cooperative foraging in a

phloem-feeding insect. Proceedings of the Royal Society of London, B, 272,
1023–1029.
Cocroft, R. B. & R. L. Rodriguez. (2005). The behavioral ecology of insect vibra-
tional communication. BioScience, 55, 323–334.
Côté, I. M. (2000). Evolution and ecology of cleaning symbioses in the sea. Oceano-
graphic Marine Biology Annual Review, 38, 311–355.
Daanje, A. (1950). On locomotory movements in birds and the intention movements
derived from them. Behaviour, 3, 48–98.
Dantzker, M. S., G. B. Deane, & J. W. Bradbury. (1999). Directional acoustic radi-
ation in the strut display of male sage grouse Centrocercus urophasianus. Journal
of Experimental Biology, 202, 2893–2909.
Darwin, C. (1872). The Expression of the Emotions in Man and the Animals. London:
John Murray.
Dawkins, R. & J. R. Krebs. (1978). Animal signals: Information or manipulation. In
J. R. Krebs & N. B. Davies (eds.), Behavioural Ecology (pp. 282–309). Oxford,
UK: Blackwell Scientific Publications.
Dean, W. R. J. & I. A. W. MacDonald. (1981). A review of African birds feeding in
association with mammals. Ostrich, 52, 135–155.
Deneubourg, J. L., S. Goss, N. Franks, & J. M. Pasteels. (1989). The blind leading
the blind: Modeling chemically mediated army ant raid patterns. Journal of
Insect Behavior, 2, 719–725.
Dunbar, I. F. (1977). Olfactory preferences in dogs: The response of male and female
to conspecific odours. Behavioral Biology, 20, 471–481.
Dunbar, R. M. (2010). The social role of touch in humans and primates: Behavioural
functions and neurobiological mechanisms. Neuroscience and Biobehavioral
Reviews, 34, 260–268.
Ehret, G. (2005). Infant rodent ultrasounds—a gate to the understanding of sound
communication. Behavioral Genetics, 35, 19–29.
Elgar, M. A. (1986). House sparrows establish foraging flocks by giving chirrup calls if
the resources are divisible. Animal Behaviour, 34, 169–174.
Elgar, M. A. (1991). Sexual cannibalism, size dimorphism, and courtship behavior of
orb-weaving spiders (Araneidae). Evolution, 45, 444–448.
Elowson, A. M., P. L. Tannenbaum, & C. T. Snowdon. (1991). Food-associated calls
correlate with food preferences in cotton-top tamarins. Animal Behaviour, 42,
931–937.
Emlen, D. J., J. Hunt, & L. W. Simmons. (2005). Evolution of sexual dimorphism
and male dimorphism in the expression of beetle horns: phylogenetic evidence
for modularity, evolutionary lability, and constraint. American Naturalist, 166,
S42–S68.
Endler, J. A. (1992). Signals, signal conditions and the direction of evolution. Ameri-
can Naturalist, 139, S125–S153.
Enquist, M. (1985). Communication during aggressive interactions with particular
reference to variation in choice of behavior. Animal Behaviour, 33, 1152–1161.
90 ANIMAL BEHAVIOR

Feder, H. M. (1966). Cleaning symbiosis in the marine environment. Symbyosis, 1,

327–380.
Fernández-Vargas, M., Z. Tang-Martínez, & S. M. Phelps. (2008). Olfactory
responses of neotropical short-tailed singing mice (Scotinomys teguina) to odors
of the mid-ventral sebaceous gland: Discrimination of conspecifics, gender,
and female reproductive condition. Journal of Chemical Ecology, 34, 429–437.
Fernández-Vargas, M., Z. Tang-Martínez, & S. M. Phelps. (2011). Singing, allog-
rooming, and allomarking behavior during intra- and inter-sexual encounters
in the neotropical short-tailed singing mouse (Scotinomys teguina). Behaviour,
148, 945–965.
Fitzgerald, T. D. & F. X. Webster. (1993). Identification and behavioral assays of the
trail pheromone of the forest tent caterpillar, Malacosoma disstria Hübner
(Lepidoptera: Lasiocampidae). Canadian Journal of Zoology, 71, 1511–1515.
Fitzpatrick, M. J. & D. A. Gray. (2001). Divergence between the courtship songs of
the field crickets Gryllus texensis and Gryllus rubens (Orthoptera: Gryllidae).
Ethology, 107, 1075–1085.
Font, E. & P. Carazo. (2010). Animals in translation: Why there is meaning (but
probably no message) in animal communication. Animal Behaviour, 79,
e1–e6.
Fusani, L. (2008). Testosterone control of male courtship in birds. Hormones and
Behavior, 54, 227–233.
Gibson, J. S. & G. W. Uetz. (2008). Seismic communication and mate choice in wolf
spiders: Components of male seismic signals and mating success. Animal
Behaviour, 75, 1253–1262.
Girard, M. B., M. M. Kasumovic, & D. O. Elias. (2011). Multimodal courtship in
the peacock spider Maratus volans (O.P.-Cambridge 1874). PLoS ONE,
e25390.
Gould, J. L. (1975). Honey bee recruitment: The dance language controversy. Sci-
ence, 189, 685–693.
Grafen, A. (1990). Biological signals as handicaps. Journal of Theoretical Biology, 144,
517–546.
Gray, D. & W. H. Cade. (2000). Sexual selection and speciation in field crickets. Pro-
ceedings of the National Academy of Sciences, USA, 97, 14449–14454.
Grieser, M. (2009). Mobbing calls signal predator category in a kin group-living bird
species. Proceedings of the Royal Society of London, B, 276, 2887–2892.
Grüter, C. & W. M. Farina. (2008). The honeybee waggle dance: Can we follow the
steps? Trends in Ecology and Evolution, 24, 242–247.
Grüter, C. & W. M. Farina. (2009). Informational conflicts created by the waggle
dance. Proceedings of the Royal Society of London, B, 275, 1321–1327.
Guillette L. M., L. L. Bloomfield, E. R. Batty, M. W. R. Dawson, & C. B. Sturdy.
(2010). Black-capped (Poecile atricapillus) and mountain chickadee (Poecile
gambeli) contact calls contain species, sex, and individual identity features.
Journal of the Acoustical Society of America, 127, 1116–1123.
ANIMAL COMMUNICATION 91

Hagedorn, M. (1986). The ecology, courtship and mating of gymnotiform electric

fish. In T. H. Bullock & W. Heiligenberg (eds.), Electroreception (pp. 497–
525). New York: Wiley.
Halpin, Z. T. (1986). Individual odors in mammals: Origins and functions. Advances
in the Study of Behavior, 16, 40–70.
Halpin, Z. T. (1991). Kin recognition cues of vertebrates. In P. G. Hepper (ed.), Kin
Recognition (pp. 220–258). Cambridge, UK: Cambridge University Press.
Harju, H. J. (1971). Spruce grouse copulation. Condor, 73, 380–381.
Harris, M. A. & J. O. Murie. (1982). Responses to oral gland scents from different
males in Columbian ground squirrels. Animal Behaviour, 30, 140–148.
Heth, G., E. Frankenberg, A. Raz, & E. Nevo. (1987). Vibrational communication in
subterranean mole rats (Spalax ehrenbergi). Behavioral Ecology and Sociobiology,
21, 31–33.
Hill, G. E. (1991). Plumage coloration is a sexually selected indicator of male quality.
Nature, 350, 337–339.
Hinde, R. A. (1970). Animal Behavior: A Synthesis of Ethology and Comparative Psy-
chology. New York: McGraw Hill.
Hinde, R. A. (1981). Animal signals: Ethological and game theory approaches are not
incompatible. Animal Behaviour, 29, 535–542.
Hunsaker, D. (1962). Ethological isolating mechanisms in the Sceloporus torquatus
group of lizards. Evolution, 16, 62–74.
Huxley, J. (1923). Courtship activities in the red-throated diver Colymbus stellatus
pontopp; together with a discussion on the evolution of courtship in birds. Jour-
nal of the Linnean Society, 2, 491–562.
Isack, H. A. & H. U. Reyer. (1989). Honeyguides and honey gatherers: Interspecific
communication in a symbiotic relationship. Science, 243, 1343–1346.
Jacobson, M. & M. Beroza. (1963). Chemical insect attractant. Science, 140, 1367–1373.
Jenssen, T. A. (1970). Female responses to filmed displays of Anolis nebulosus (Sauria:
Iguanidae). Animal Behaviour, 18, 640–647.
Johnson, D. L. (1967). Honey bees: Do they use the direction information contained
in their dance maneuver? Science, 155, 844–847.
Kaptein, N., J. Billen, & B. Gobin. (2005) Larval begging for food enhances repro-
ductive options in the ponerine ant Gnamptogenys striatula. Animal Behaviour,
69, 293–299.
Kayser, A. J. & G. E. Hill. (2000). Structurally based plumage coloration is an honest
signal of quality in male blue grosbeaks. Behavioral Ecology, 11, 202–209.
Keil, N. M., S. Imfeld-Muller, J. Aschwanden, & B. Wechsler. (2012). Are head cues
necessary for goats (Capra hircus) in recognizing group members? Animal Cog-
nition, 15, 913–921.
Kleiman, D. (1972). Social behavior in the maned wolf and bush dog: A study in con-
trast. Journal of Mammalogy, 53, 791–806.
Laidre, M. E. (2009). How often do animals lie about their intentions? An experi-
mental test. American Naturalist, 173, 337–346.
92 ANIMAL BEHAVIOR

Leal, M. (1999). Honest signaling during prey-predator interactions in the lizard Ano-
lis cristatellus. Animal Behaviour, 58, 521–526.
Leal, M. & J. A. Rodriguez-Robles. (1995). Antipredator responses of Anolis cristatel-
lus (Sauria: Polychrotidae). Copeia, 1995, 155–161.
Leger, D. W & D. H. Owings. (1978). Responses to alarm calls by California ground
squirrels: Effects of call structure and maternal status. Behavioral Ecology and
Sociobiology, 3, 177–186.
Leyhausen, P. (1956). Das verhalten der katzen (Felidae). Handbook of Zoology (Berlin),
8, 1–34.
Ligout S., M. Keller & R. H. Porter. (2004). The role of olfactory cues in the dis-
crimination of agemates by lambs. Animal Behaviour, 68,785–792.
Ligout, S. & R. H. Porter. (2004). The role of visual cues in lambs’ discrimination
between individual agemates. Behaviour, 141, 617–632.
Lorenz, K. (1972). Comparative studies on the behavior of Anatinae. In P. H. Klopfer
& J. P. Hailman (eds.), Function and Evolution of Behavior: An Historical Sam-
ple from the Pen of Ethologists (pp. 231–258). Reading, MA: Addison-Wesley.
Lovell, S. F. & M. R. Lein. (2004). Neighbor-stranger discrimination by song in a
suboscine bird, the alder flycatcher, Empidonax alnorum. Behavioral Ecology.
15, 799–804.
Magrath, R. D., B. J. Pitcher, & J. L Gardner. (2009). An avian eavesdropping net-
work: Alarm signal reliability and heterospecific response. Behavioral Ecology,
20, 745–752.
Majolo, B., R. Ventura, & N. Koyama. (2005). Postconflict behavior among male
Japanese macaques. International Journal of Primatology, 26, 321–336.
Maklakov, A. A., T. Bilde, & Y. Lubin. (2003). Vibratory courtship in a web-
building spider: Signaling quality or stimulating the female? Animal Behaviour,
66, 623–630.
Manser, M. B. (1999). Response of foraging group members to sentinel calls in suri-
cates, Suricata suricatta. Proceedings of the Royal Society of London, B, 266,
1013–1019.
Marler, P. (1961). The logical analysis of animal communication. Journal of Theoreti-
cal Biology, 1, 295–317.
Marler, P. & H. Slabbekoorn. (2004). Nature’s Music: the Science of Birdsong. San
Diego, CA: Elsevier Academic Press.
Møller, A. P. (1988). False alarm calls as a means of resource usurpation in the great
tit Parus major. Ethology, 79, 25–30.
Morgan, E. D. (2008). Trail pheromones of ants. Physiological Entomology, 34, 1–17.
Morris, D. (1956). The feather posture of birds and the problem of the origin of social
signals. Behaviour, 9, 75–113.
Morris, D. (1957). “Typical intensity” and its relation to the problem of ritualization.
Behaviour, 11, 1–12.
Moynihan, M. (1955). Some aspects of the reproductive behavior of the black-headed
gull (Larus rudibundus rudibundus L.) and related species. Behaviour Supple-
ment, 4, 1–201.
ANIMAL COMMUNICATION 93

Muller, M. N. & J. C. Mitani. (2005). Conflict and cooperation in wild chimpanzees.

Advances in the Study of Behavior, 35, 275–331.
Munz, T. (2005). The bee battles: Karl von Frisch, Adrian Wenner and the honey bee
dance language controversy. Journal of the History of Biology, 38, 535–570.
Muroyama, Y. & B. Thierry. (1998). Species differences in male loud calls and their
perception in Sulawesi macaques. Primates, 39, 115–126.
Nakamichi, M. & Y. Shizawa. (2003). Distribution of grooming among adult females
in a large, free-ranging group of Japanese macaques. International Journal of
Primatology, 24, 607–625.
Narins, P. M., E. R. Lewis, J. J. U. M. Jarvis, & J. O’Rain. (1997). The uses of seis-
mic signals by fossorial southern African mammals: A neuroethological gold
mine. Brain Research Bulletin, 44, 641–646.
Ord, T. J., R. A. Peters, B. Clucas, & J. A. Stamps. (2007). Lizards speed up visual
displays in noisy motion habitats. Proceedings of the Royal Society of London,
B, 274, 1057–1062.
Owings, D. H. & E. S. Morton. (1997). The role of information in communication:
An assessment/ management approach. In D. H. Owings, M. D. Beecher, &
N. S. Thompson (eds.), Perspectives in Ethology, Vol. 12: Communication
(pp. 359–390). New York: Plenum Press.
Owren, M. J., D. Rendall, & M. J. Ryan. (2010). Redefining animal signaling: Influ-
ence versus information in communication. Biology and Philosophy, 25,
755–780.
Payne, K. (1998). Silent Thunder: In the Presence of Elephants. New York: Simon and
Schuster.
Payne, K., W. R. Langbauer Jr., & E. M. Thomas. (1986). Infrasonic calls of the
Asian elephant (Elephas maximus). Behavioral Ecology and Sociobiology, 18,
297–301.
Pomiankowski, A. (1987). Sexual selection: The handicap mechanism does work—
sometimes. Proceedings of the Royal Society of London, B, 231, 123–145.
Proctor, H. C. (1991). Courtship in the water mite Neumania papillator: Males capi-
talize on female adaptations for predation. Animal Behaviour, 42, 589–598.
Pryke, S. R., M. J. Lawes, & S. Andersson. (2001). Agonisitic carotenoid signaling in
male red-collared widowbirds: Aggression related to the color signal of both the
territory owner and model intruder. Animal Behaviour, 62, 695–704.
Rado, R., J. Terkel, & Z. Wollberg. (1998). Seismic communication signals in the
blind mole rat (Spalax ehrenbergi): Electrophysiological and behavioral evi-
dence for their processing by the auditory system. Journal of Comparative
Physiology, A, 183, 503–511.
Rainey, H. J., K. Zuberbühler, & P. J. B. Slater. (2004). Hornbills can distinguish
between primate alarm calls. Proceedings of the Royal Society of London, B,
271, 755–759.
Randall, J. A. & E. R. Lewis. (1997). Seismic communication between the burrows of
kangaroo rats, Dipodomys spectabilis. Journal of Comparative Physiology, A, 181,
525–531.
94 ANIMAL BEHAVIOR

Randall, J. A. & M. D. Matocq. (1997). Why do kangaroo rats (Dipodomys spectabilis)

footdrum at snakes? Behavioral Ecology, 8, 404–413.
Redondo, T. & L. Arias de Reyna. (1988). Locatability of begging calls in nestling
altricial birds. Animal Behaviour, 36, 653–661.
Rendall, D., M. J. Owren, & M. J. Ryan. (2009). What do animal signals mean? Ani-
mal Behaviour, 78, 233–240.
Ribak, G., M. L. Pitts, G. S. Wilkinson, & J. G. Swallow. (2009). Wing shape, wing
size and sexual dimorphism in eye span in stalk-eyed flies (Diopsidae). Biologi-
cal Journal of the Linnean Society, 98, 860–871.
Robinson, M. H. (1982). Courtship and mating behaviors in spiders. Annual Review
of Entomology, 27, 1–20.
Roelofs, W. L. & A. Comeau. (1969). Sex pheromone specificity: Taxonomic and
evolutionary aspects in Lepidoptera. Science, 165, 398–400.
Rohwer, S. A. (1977). Status signaling in Harris sparrows: Some experiments in
deception. Behaviour, 61, 107–129.
Rohwer, S. A. & F. C. Rohwer. (1978). Status signaling in Harris’ sparrows: Experi-
mental deceptions achieved. Animal Behaviour, 26, 1012–1022.
Ruxton, G. D. & H. M. Schaeffer. (2011). Resolving current disagreements and
ambiguities in the terminology of animal communication. Journal of Evolution-
ary Biology, 24, 2574–2585.
Ryan, M. J. (1990). Signals, species, and sexual selection. American Scientist, 78,
46–52.
Ryan, M. J. & A. S. Rand. (1990). The sensory basis of sexual selection for complex
calls in the túngara frog, Physalaemus pustulosus (sexual selection for sensory
exploitation). Evolution, 44, 305–314.
Ryan, M. J. & A. S. Rand. (1993). Sexual selection and signal evolution: The ghosts
of biases past. Philosophical Transactions of the Royal Society, B, 340, 187–195.
Ryan, M. J., M. D. Tuttle, & A. S. Rand. (1982). Sexual advertisement and bat pre-
dation in a neotropical frog. American Naturalist, 119, 136–139.
Sauer, W. H. H., M. J. Roberts, M. R. Lipinsky, M. J. Smale, R. T. Hanlon, D. M.
Webber, et al. (1997). Choreography of the squids’ “nuptial dance.” Biological
Bulletin, 192, 203–207.
Schino, G. (2001). Grooming, competition, and social rank among female primates:
A meta-analysis. Animal Behaviour, 62, 265–271.
Searcy, W. A. & S. Nowicki. (2005). The Evolution of Animal Communication: Reli-
ability and Deception in Signaling Systems. Princeton, NJ: Princeton University
Press.
Seyfarth, R. M. (1980). The distribution of grooming and related behaviours among
adult female vervet monkeys. Animal Behaviour, 28, 798–813.
Seyfarth, R. M, D. L. Cheney, T. Bergman, J. Fischer, K. Zuberbühler, & K. Ham-
merschmidt. (2010). The central importance of information in studies of ani-
mal communication. Animal Behaviour, 80, 3–8.
Sherman, P. W. (1977). Nepotism and the evolution of alarm calls. Science, 197,
1246–1253.
ANIMAL COMMUNICATION 95

Sherman, P. W. (1985). Alarm calls of Belding’s ground squirrels to aerial predators:

Nepotism or self preservation? Behavioral Ecology and Sociobiology, 17,
313–323.
Sivalinghem, S., M. M. Kasumovic, A. C. Mason. M. C. B. Andrade, & D. O. Elias.
(2010). Vibratory communication in the jumping spider Phiddipus clarus: Pol-
yandry, male courtship signals, and mating success. Behavioral Ecology, 21,
1308–1314.
Slater, P. J. B. (1983). The study of communication. In T. R. Halliday & P. J. B.
Slater (eds.), Animal Behaviour, Vol. 2: Communication (pp. 9–42). Oxford,
UK: Blackwell Publishing.
Slater, P. J. B. & N. I. Mann. (2004). Why do the females of many bird species sing
in the tropics? Journal of Avian Biology, 35, 289–294.
Smith, W. J. (1977). The Behavior of Communicating: An Ethological Approach.
Cambridge, MA: Harvard University Press.
Smith, W. J. (1986). An “informational” perspective on manipulation. In R. W.
Mitchell & N. S. Thompson (eds.), Deception: Perspectives on Human and
Nonhuman Deceit (pp. 71–86). Albany: State University of New York Press.
Smotherman, W. P., R. W. Bell, J. Starzec, J. Elias, & T. A. Zackman. (1974).
Maternal responses to infant vocalizations and olfactory cues in rats and mice.
Behavioral Biology, 12, 55–66.
Spieth, H. T. (1974). Courtship behavior in Drosophila. Annual Review of Entomology,
19, 385–405.
Sugiura, H. (1998). Matching of acoustic features during the vocal exchange of coo
calls by Japanese macaques. Animal Behaviour, 55, 673–687.
Tang-Martínez, Z. (2001). The mechanisms of kin discrimination and the evolution
of kin recognition in vertebrates: A critical re-evaluation. Behavioural Processes,
53, 21–40.
Templeton, C. N. & E. Greene. (2007). Nuthatches eavesdrop on variations in het-
erospecific chickadee mobbing alarm calls. Proceedings of the National Academy
of Sciences, USA, 104, 5479–5482.
Tinbergen, N. (1951). The Study of Instinct. London: Oxford Clarendon Press.
Tinbergen, N. (1952). “Derived” activities: Their causation, biological significance,
origin and emancipation during evolution. Quarterly Review of Biology, 27,
1–32.
Tinbergen, N. (1959). Comparative studies of the behaviour of gulls (Laridae):
A progress report. Behaviour, 15, 1–70.
Trainer, J. M., D. B. McDonald, & W. A. Learn. (2002). The development of coor-
dinated singing in cooperatively displaying long-tailed manakins. Behavioral
Ecology, 13, 65–69.
Vaché, M., J. Ferron, & P. Gouat. (2001). The ability of red squirrels (Tamiasciurus
hudsonicus) to discriminate conspecific olfactory signatures. Canadian Journal
of Zoology, 79, 1296–1300.
Vandenberg, J. G. (1969). Endocrine coordination in monkeys: Male sexual
responses to the female. Physiology and Behavior, 4, 261–264.
96 ANIMAL BEHAVIOR

Von Frisch, K. (1967). The Dance Language and Orientation of Bees. Cambridge, MA:
Harvard University Press.
Von Hippel, F. A. (2000). Vigorously courting male sticklebacks are poor fathers.
Acta Ethologica, 2, 83–89.
Wager, B. R. & M. D. Breed. (2000). Does honeybee sting alarm pheromone give
orientation information to defensive bees? Annals of the Entomological Society
of America, 93, 1329–1332.
Wallace, B. (1973). Misinformation, fitness, and selection. American Naturalist, 107,
1–7.
Wenner, A. M. (1967). Honey bees: Do they use the distance information contained
in their dance maneuver? Science, 155, 847–849.
Whittaker, D. J., D. G. Reichard, A. L. Dapper, & E. D. Ketterson. (2009). Behav-
ioral responses of nesting female dark-eyed juncos, Junco hyemalis, to hetero-
and conspecific passerine preen oils. Journal of Avian Biology, 40, 579–583.
Whittaker, D. J., K. M. Richmond, A. K. Miller, R. Kiley, C. Bergeon Burns, J.
Atwell, & E. D. Ketterson. (2011). Intraspecific preen oil odor preferences in
dark-eyed juncos (Junco hyemalis). Behavioral Ecology, 22, 1256–1263.
Whittaker, D. J., H. A. Soini, J. W. Atwell, C. Hollars, M. V. Novotny, & E. D. Ket-
terson. (2010). Songbird chemosignals: Preen oil volatile compounds vary
among individuals, sexes, and populations. Behavioral Ecology, 21, 608–614.
Wickler, W. (1967). Socio-sexual signals and their intraspecific imitation among pri-
mates. In D. Morris (ed.), Primate Ethology (pp. 69–147). London: Weiden-
feld and Nicholson.
Wilson, E. O. (1975). Sociobiology. Cambridge, MA: Harvard University Press.
Wilson, E. O. & W. H. Bossert. (1963). Chemical communication among animals.
Recent Progress in Hormone Research, 19, 673–716.
Wunderlee, J. M. (1978). Differential responses of territorial yellowthroats to the
songs of neighbors and non-neighbors. Auk, 95, 389–395.
Wyatt, T. D. (2003). Pheromones and Animal Behaviour: Communication by Smell and
Taste. Cambridge, UK: Cambridge University Press.
Wyman, M. T., M. S. Mooring, B. McCowan, M. C. T. Penedos, & L. A. Hart.
(2008). Amplitude of bison bellows reflect male quality, physical condition
and motivation. Animal Behaviour, 76, 1625–1639.
Yasukawa, K., R. L Knight, & S. K. Skagen. (1987). Is courtship intensity a signal of
male parental care in red-winged blackbirds (Agelaius phoenicius)? Auk, 104,
628–634.
Young, J. C., R. M. Silverstein, & M. C. Birch. (1973). Aggregation pheromone of
the beetle Ips confusus: Isolation and identification. Journal of Insect Physiology,
11, 2273–2277.
Zahavi, A. (1975). Mate selection—selection for a handicap. Journal of Theoretical
Biology, 53, 205–213.
Zahavi, A. (1977). The cost of honesty (further remarks on the handicap principle).
Journal of Theoretical Biology, 67, 603–605.
ANIMAL COMMUNICATION 97

Zeigler, H. P. (1964). Displacement activity and motivational theory: A case study in

the history of ethology. Psychological Bulletin, 61, 362–376.
Zuberbühler, K. (2000). Interspecies semantic communication in two forest primates.
Proceedings of the Royal Society of London, B, 267, 713–718.
Zuk, M., D. Rebar, & S. Primrose Scott. (2008). Courtship song is more variable
than calling song in the field cricket Teleogryllus oceanicus. Animal Behaviour,
76, 1065–1071.
This page intentionally left blank
 4

Mating Systems and the Measurement

of Sexual Selection
Kenyon B. Mobley

INTRODUCTION
The process of sexual selection underlies the way that individuals choose
mates and produce offspring in nature (Darwin, 1871). The opportunities
and choices that an individual may make are governed by both physiological
and phylogenetic constraints, the types of social interactions and potential-
mate encounters an individual experiences, the method of parental care
employed, and a whole host of ecological factors that may influence the breed-
ing biology of individuals within a population. The sum total of these mating
and reproductive opportunities and the methods used to estimate the strength
and direction of sexual selection are the subject of this chapter: mating systems
and the measurement of sexual selection.
Mating-system research is a rapidly growing field with the specific aim to
understand the rich variety of reproductive behaviors, sex roles, and expres-
sion of secondary sexual traits that abound in nature. Historically, the term
mating system refers to a behavioral strategy employed by a population of
animals to obtain mates (Emlen & Oring, 1977), including various aspects
of sexual selection such as mate choice and mate competition, parental care,
and ecological factors that may influence the degree of multiple mating
within a population. Today, interest in mating-system research has expanded
considerably and now encompasses various alternative mating behaviors and
100 ANIMAL BEHAVIOR

strategies, genetic mating systems, and a theoretical framework for estimat-

ing the strength and direction of sexual selection based on quantitative selec-
tion theory. Still, the field of mating systems holds true to the original
goal of understanding the patterns of mating and reproductive success of a
population and how these patterns, in turn, are regulated by the environ-
mental conditions that may ultimately govern the expression of such
behaviors.
In this chapter I will provide a brief historical overview of animal mating
systems and how our understanding of sexual selection has grown since
Charles Darwin’s (1871) treatise on the subject in The Descent of Man and
Selection in Relation to Sex. Because the literature can be daunting and laden
with different perspectives, a second goal of this chapter is to provide a user-
friendly guide to help navigate the more commonly used methods to quantify
sexual selection as well as the advantages and drawbacks to each. The objective
here is to provide students and researchers a firm foundation from which to
investigate questions concerning mating systems and the measurement of sex-
ual selection in their own particular system of choice.

A BRIEF HISTORY OF MATING SYSTEMS

Among the first to recognize the importance of mating systems was Dar-
win, who, among his other important contributions to the field of sexual selec-
tion, introduced the founding principles of mating-system evolution. He
noted,
That some relation exists between polygamy and the development of secondary
sexual characters, appears nearly certain; and this supports the view that a
numerical preponderance of males would be eminently favourable to the action
of sexual selection. (Darwin, 1871, pp. 254–255)

The first principle that can be deduced from this passage is that a causal rela-
tionship exists between the mating system and the presence of secondary sex-
ual characters. Here, Darwin recognized that the degree to which secondary
sexual characters are expressed depends strongly on the degree of multiple
mating within a population. The second principle that Darwin identified is
the relationship between the strength of sexual selection and the sex ratio.
He understood that if the sex ratio was skewed toward one sex (males in this
instance) then the other sex would become limiting, and therefore intrasexual
competition would increase on the nonlimiting sex. Thus, a greater bias
towards the nonlimiting sex should result in stronger sexual selection
acting on the nonlimiting sex (Trivers, 1972; Emlen & Oring, 1977;
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 101

Clutton-Brock, 1991). Taken together, these ideas provided the raw materials
for the theoretical framework for modern-day mating-system research.
A renewed interest in sexual selection occurred in the latter part of the
twentieth century with the introduction of two theories, anisogamy and
parental investment theory. The theory of anisogamy (Parker et al., 1972)
posited that the differential investment into the production of gametes by
males and females gave rise to the evolution of sexual selection and conflict
between the sexes. The theory of anisogamy is predicated on the fact that
males produce many small, motile sperm whereas females produce large, non-
motile, and nutritious eggs. As a result of this differential investment, males
are predicted to compete more intensely for fertilization opportunities, gener-
ating a male-biased operational sex ratio under normal conditions. Although
both theories speak more to the underlying differences in the sexes rather than
the actual mechanisms of sexual selection (Kokko & Jennions, 2008), they set
the stage for further development of theories concerning the fundamental con-
flict between the sexes driving sexual selection and mating system evolution.

The Environmental Potential for Polygamy

The first cogent attempt to synthesize mating-system theory was advanced
by Stephen Emlen and Lewis Oring (1977). In their treatise, they outlined a
theoretical framework for discussing mating systems that is still in place today.
The crux of their theoretical framework rests on the ability of certain individ-
uals in a population to monopolize or limit access of competitors to potential
mates. If a few individuals of one sex can monopolize several mates, then vari-
ance in the number of mates, or mating success, will increase. This variation is
quantifiable, as we will see later in the chapter, and divides individuals of each
sex into “winners” or “losers,” or, in keeping with Emlen and Oring’s original
terminology, the “haves” and “have nots.” Thus, a precondition for the evolu-
tion of various mating systems is the ability for some individuals to monopo-
lize mates. Because the ability to monopolize mates is theoretically
constrained by phylogenetic factors, such as reproductive physiology and
parental care strategies, as well as ecological factors such as mate availability,
they coined the concept of the environmental potential for polygamy to cap-
ture the essence of both the ecological conditions necessary for and the ability
to capitalize on multiple mating.
To illustrate their point of the environmental potential for polygamy,
Emlen and Oring (1977) devised a graphical model in which two critical com-
ponents, namely the temporal availability of mates and the spatial distribution
of resources, influence the outcome of the mating system for a particular pop-
ulation. For example, if the spatial distribution of resources such as nesting
102 ANIMAL BEHAVIOR

sites is uniform and all individuals are reproductively receptive at the same
time, then the ability of a single individual to monopolize multiple mates is
diminished. If, on the other hand, mating is asynchronous such that only a
few individuals are receptive at a particular time and resources are spatially
clumped, for instance around a particularly nice patch of resources, then the
ability of a particular individual to monopolize multiple mates is greatly
increased. The resulting mating systems predicted to evolve from these scenar-
ios are monogamy and polygamy (either polygyny or polyandry, depending
on the monopolizing sex), respectively. While these two scenarios lie at the
two extremes of the environmental potential for polygamy continuum, one
can envision a multitude of situations where both the temporal availability of
mates and the spatial distribution of resources are variable and may differ
between the sexes, thereby setting the stage for the evolution of different mat-
ing systems. An important point here is that because temporal availability of
mates and spatial distributions of resources are likely to fluctuate in a particu-
lar habitat or population, this model also has the implicit assumption that
mating systems can be flexible and are not necessarily fixed for a population.
The final major contribution by Emlen and Oring (1977) was the elabora-
tion on the concept of the operational sex ratio, which quickly became a
cornerstone of mating-system theory. Originally described in Emlen (1976,
p. 309) as “the ratio of receptive females to potential mating males at one
time,” the operational sex ratio was presented as an instantaneous measure of
both the intensity and direction of mate competition, and hence sexual selec-
tion. Emlen and Oring (1977) also point out that skewed operational sex
ratios have the ability to manipulate which mating system would be most
likely to evolve in various conditions, such as polygyny under male-biased
conditions and polyandry under female-biased conditions.
In the time since Emlen and Oring’s (1977) publication, two major devel-
opments led to increased interest in mating-system research and, ultimately, a
new appreciation of the intricacies of sexual selection. The first such advance-
ment is a theoretical framework for quantifying sexual selection that combines
elements of mating-system theory with selection theory, thereby bridging the
gap between sexual selection, mating-system organization, and quantitative
selection theory (Wade, 1979; Arnold & Wade, 1984; Arnold & Duvall,
1994; Shuster & Wade, 2003; Jones, 2009). The second such advancement
was the technological development of molecular markers that opened a flood-
gate of molecular inquiries into patterns of parentage or genetic mating sys-
tems in natural populations (Avise, 2004; Jones et al., 2005). The
combination of these two major advancements has solidified the field of mat-
ing systems, propelling research into the twenty-first century.
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 103

A Theoretical Quantitative Framework

In order to discuss the first major advance in mating-system research in
recent times, namely the synthesis of a theoretical quantitative framework to
measure population-level sexual selection, we need to go back in time to
Angus J. Bateman’s classic study on the mating behavior of Drosophila mela-
nogaster, rediscovered by Trivers (1972). Based on several simple yet elegant
experiments using visible phenotypic markers of both sexes, Bateman (1948)
demonstrated that male Drosophila exhibited higher variance in mating suc-
cess, or the number of mates per male, than did females. Similarly, he demon-
strated that males exhibited higher variance in fertility, or the number of
offspring per male, than did females. Finally, Bateman postulated that the
strength of sexual selection on males was driven by a higher correlation
between mating success and reproductive success, famously stating, “Variance
in number of mates is, therefore, the only important cause of the difference in
the variance in fertility” (Bateman, 1948, p. 364). These core principles,
namely the variance in mating success, the variance in reproductive success,
and the relationship between mating and reproductive success, were dubbed
Bateman’s principles (Wade, 1979; Arnold & Wade, 1984; Arnold, 1994;
Arnold & Duvall, 1994), and a new era of quantitative appraisals of mating-
system organization was born.
After the initial introduction of Bateman’s principles as a method to quan-
tify sexual selection, a flurry of critical reviews rapidly followed, shedding
doubt on whether one can distinguish sexual selection from random mating
patterns using quantitative approaches (Sutherland, 1985, 1987; Koening &
Albano, 1986; Hubbell & Johnson, 1987) or whether quantification of sexual
selection is even interesting (Grafen, 1987). To add to this confusion, various
other methods to measure mating and reproductive inequality were also
proposed to quantify mating systems (Kokko et al., 1999), although at the
time few of these measures were actually tested and compared empirically
(Fairbairn & Wilby, 2001).
This building controversy was met with the timely publication of “Mating
Systems and Strategies” by Stephen Shuster and Michael Wade (2003), which
attempted to synthesize the burgeoning field of quantitative mating systems.
The authors’ approach combined a theoretical framework for classifying mat-
ing systems based on the temporal availability of mates and the spatial distri-
bution of resources (sensu Emlen & Oring, 1977) with a new metric for
quantifying the direction and intensity of sexual selection, Imates, or the sex dif-
ference in the opportunity for selection. Further refinements to the theory
have been added by Adam Jones (2009), who proposed an upper limit to
the standardized selection differential in units of phenotypic standard
104 ANIMAL BEHAVIOR

deviations (s0 max) and clarified some earlier misconceptions on the actual maxi-
mum strength of selection.
Today, quantitative methods to measure sexual selection and mating sys-
tems remain a controversial topic and are subject to vigorous debate (e.g.,
Klug, Heuschele, et al., 2010; Krakauer et al., 2011). Although some progress
has been made to synthesize the various methods, attempts to reach a consen-
sus on the most appropriate methods appear to be unattainable in the near
future. Theoretical advancements to mating-system theory have thus far out-
paced rigorous empirical testing, although recent work is beginning to bridge
the knowledge gap by investigating different quantitative mating-system mea-
surements under various ecological scenarios (Jones et al., 2005; Mills et al.,
2007; Mobley & Jones, 2007, 2009; Barreto & Avise, 2010; Fitze & Le
Galliard, 2011). Further empirical testing and refinement of theory is clearly
warranted and will be necessary to settle this dispute in the future.

Molecular Markers and Genetic Mating Systems

The discovery and development of molecular DNA fingerprinting tech-
niques, such as microsatellite DNA and amplified fragment length polymor-
phism (AFLP) markers, have allowed for the investigation of parentage and
kinship within natural populations in unprecedented detail (Avise, 2004).
This discovery then fueled a molecular revolution in behavioral ecology. For
example, the discovery of extra-pair paternity in socially monogamous birds
illustrated the disparity between the social mating system (i.e., social interac-
tions and parental care strategies) and the realized genetic mating system in
many species (Dunn et al., 2001; Hasselquist & Sherman, 2001; Griffith et
al., 2002). Overturning this paradigm is especially relevant to mating systems
because the seemingly widespread occurrence of gaudy secondary sexual
traits in birds was a conundrum to Darwin and subsequent researchers. To
his credit, Darwin proposed a verbal model of how sexual selection can operate
in monogamous species based on the temporal variability in breeding timing
of individuals of variable quality (Darwin, 1871). This theory is often called
the Darwin-Fisher theory of sexual selection, based on the mathematical proof
of the theory by R. A. Fisher (Fisher, 1915; Kirkpatrick et al., 1990; Jones &
Ratterman, 2009). Although the Darwin-Fisher theory of sexual selection may
help explain the existence of sexual selection in species that are genetically
monogamous, the application of molecular markers now provides ample
support for the idea that the degree of extra-pair paternity explains the exis-
tence of elaborate traits in a variety of socially monogamous bird species. This
in turn reaffirms Darwin’s original notion that a link between polygamy
and secondary sexual characters exists (Darwin, 1871; Møller & Birkhead,
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 105

1994; Petrie & Kempenaers, 1998; Griffith et al., 2002; Westneat & Stewart,
2003).
Today, it is generally recognized that a complete understanding of sexual
selection in the wild requires some knowledge of the genetic mating system.
Their widespread availability and low cost, along with advances in parentage
analysis (Jones & Ardren, 2003; Jones et al., 2010), have made molecular
markers easy to apply to many different animal mating systems. Coupled with
new quantitative approaches for the measurement of sexual selection, we now
have a framework for testing specific hypotheses concerning the evolution of
mating systems, the direction of sexual selection, and selection on individual
traits associated with mating and reproductive success.
This brief synthesis highlights the important methodological and theoreti-
cal advances in the field of mating systems to date. However, it should be
noted that this treatment glosses over significant contributions of many others
that have increased our overall understanding of mating systems. For example,
other noteworthy milestones include Fisher’s contributions to sex-ratio theory
(Fisher, 1930), James F. Crow’s formulation of the opportunity for selection
(Crow, 1958, 1962), and the mathematical formulation of spatial distribution
analysis by mean crowding by Monte Lloyd (1967), just to name a few.

TERMINOLOGY
Despite over 150 years of research in mating systems, the field suffers from
a lack of generally accepted terminology, and attempts to disambiguate various
definitions have met with limited success (Emlen & Oring, 1977; Reynolds,
1996). For example, if you initiate a literature search for “mating systems,”
you are likely to find an equal number of references concerning animal mating
systems and plant mating systems. A quick perusal would suggest zoologists
and botanists have divergent views of what comprises a mating system, com-
plete with separate lexicons (Reynolds, 1996; Sakai & Westneat, 2001).
Although the emphases of plant and animal mating systems may appear to
be different, in actuality both are concerned with patterns of parentage within
populations. Animal mating systems are primarily concerned with patterns of
sexual selection within populations (Emlen & Oring, 1977; Reynolds, 1996;
Jones et al., 2004), while plant mating systems focus on rates and consequen-
ces of selfing (Sakai & Westneat, 2001; Shuster, 2009). Thus, animal and
plant mating systems can be unified with the concept of the parental table,
which summarizes the genetic parentage of all progeny in a breeding popula-
tion by ascribing the number of progeny produced by each potential pairing
regardless of whether the offspring are produced by selfing or outcrossing
106 ANIMAL BEHAVIOR

(Arnold & Duvall, 1994). Another method for unifying plant and animal
mating systems is via spatial and temporal mean crowding (Shuster,
2009), a mathematical interpretation of the original spatial and temporal dis-
tributions of resources as discussed in Emlen and Oring (1977).
For the purposes of this chapter, I will use the general definition of a mating
system as the behavioral strategy of a population employed in obtaining mates
and reproductive opportunities, following Emlen and Oring (1977, Box 4.1).
At the most basic level, animal mating systems can be divided into monoga-
mous (having a single mate) or polygamous (having more than one mate)
(Box 4.1). Because polygamy is not a particularly useful term with respect to
the direction of sexual selection, it can be broken down into further subcatego-
ries to reflect whether males mate multiply (polygyny, stronger sexual selection

Box 4.1
Definitions of commonly used terms to describe mating systems

Mating system—The general behavioral strategy of a population

employed in obtaining mates and reproductive opportunities (e.g.,
monogamy, polygamy). The mating system may also be referred to as the
mating pattern or breeding system.
Social mating system—The social behavioral strategy of a population,
including mate monopolization strategies, the presence and characteristics
of pair bonds, and the patterns of parental care provided by each sex (e.g.,
social monogamy, resource defense polygamy, promiscuity).
Genetic mating system—Actual patterns of parentage within a population,
including all precopulatory and postcopulatory sexual selection (e.g., genetic
monogamy, genetic polyandry).
Breeding bout—The minimum unit of time required to produce viable
offspring.
Monogamy—Each sex mates a maximum of once during a breeding bout.
Polygamy—A sex or both sexes mate more than once during a breeding
bout.
Polygyny—Males have a variable number of mates during a breeding
bout; females have a maximum of one mate.
Polyandry—Females have a variable number of mates during a breeding
bout; males have a maximum of one mate.
Polygynandry—The condition where both males and females have a var-
iable number of mates during a breeding bout. Sometimes also referred to
polyandrogyny when males are more variable than females and polygynan-
dry when females are more variable than males in the number of mates.
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 107

in males), females mate multiply (polyandry, stronger sexual selection in

females), or both sexes mate multiply (polygynandry, variable strength of sex-
ual selection in both sexes, Box 4.1). Mating systems can also be classified
based on the nature and extent of their social relationships (social mating sys-
tems) or genetic relationships (genetic mating systems; Box 4.1). The social
mating system includes various behavioral aspects of mating-system organiza-
tion such as mate monopolization and parental care strategies, while the
genetic mating system encapsulates the actual patterns of parentage within a
population.
To properly discuss and compare mating systems, it is important to con-
sider the amount of time required for a male and female to produce viable off-
spring, here defined as a breeding bout (Box 4.1). A breeding bout is roughly
equivalent to a selection episode (Lande & Arnold, 1983), or the period of
interest in which selection can be measured. For some species with parental
care, a breeding bout may include the time required to raise offspring and
additional time required until an adult is physiologically ready to mate again.
This inequality in the amount of time, as well as the physiological ability of
one sex to produce more offspring than the other, sets the stage for the evolu-
tion of different strategies for obtaining sexual partners between the sexes. It
should be noted that a breeding bout in many species is variable, and the
amount of time that is required to produce offspring is dependent upon a vari-
ety of physiological parameters such as condition or age of the individual as
well as various ecological parameters such as temperature, food availability,
or population density (Kvarnemo & Ahnesjö, 1996). Additional issues arise
when individuals care for multiple broods of offspring and when the time for
a particular brood to develop from a parental pairing is difficult to measure.
In these instances, exact parentage data would be one way to distinguish the
amount of time that a successful pairing took to produce viable offspring.
Establishing a common currency of reproductive time is important to
help identify and clarify mating-system terminology between studies. In par-
ticular, considerable confusion arises when mating systems are considered
under various time scales. For example, a species where males and females
mate once per breeding bout but may change partners between breeding bouts
would be considered monogamous per breeding bout but polygamous during
a mating season or a lifetime. As the behavioral strategy employed to obtain
mates during a single breeding bout would be monogamous, this would be
the most appropriate term to describe the species in this instance. While a
breeding bout may be a difficult concept to apply in some species, it is still
important to conceptualize a minimum unit of reproduction to clarify the rel-
ative number of breeding events within a reproductive lifetime.
108 ANIMAL BEHAVIOR

The lexicon of social mating systems has been reviewed extensively, and
precise definitions, as well as the social and ecological contexts under which
social mating systems have evolved, exist (Emlen & Oring, 1977; Thornhill
& Alcock, 1983; Reynolds, 1996; Shuster & Wade, 2003). Genetic mating
systems, on the other hand, have received considerably less attention in the lit-
erature. To illustrate how to apply genetic-mating-system terminology, let us
take the breeding biology of the family Syngnathidae (pipefish, seahorses,
and seadragons) as an example. Syngnathids have a specialized reproductive
mode where females transfer eggs to the male during copulation and the male
provides all care to the developing offspring. Within this family, several
genetic mating systems have evolved in different species, including
monogamy, polyandry, and polygynandry (Jones & Avise, 2001; Wilson et
al., 2003; Coleman & Jones, 2011; Mobley et al., 2011). In the seahorses
(Hippocampus spp.), males and females develop extensive pair bonds and male
seahorses do not accept eggs from additional females. Thus, during a single
pregnancy, a male is truly genetically monogamous as judged by several par-
entage studies (Jones et al., 1998; Wilson & Martin-Smith, 2007). In all but
a few rare occasions, females do not breed a second time during the pregnancy,
as it takes nearly the same amount of time for a female to mature a new clutch
of eggs as a male takes to brood and release offspring. Therefore females can
also be considered genetically monogamous (Foster & Vincent, 2004; Mobley
et al., 2011). In pipefishes of the genus Syngnathus, females generally produce
more eggs than a single male can brood during a single male pregnancy (Ber-
glund et al., 1986, 1989; Scobell et al., 2009), and hence the potential for
polygamous mating systems, both polyandry and polygynandry, is expected
to evolve. Indeed, several species of Syngnathus exhibit polygynandry where
both males and females mate multiply during a single pregnancy, as well as
polyandry where females mate multiply but males mate with just one female
during a pregnancy (reviewed in Jones & Avise, 2001; Mobley et al., 2011).
Today, it is known that both males and females in the majority of taxo-
nomic groups can mate multiply such that strict genetic monogamy, polyg-
yny, and polyandry are rare, and therefore knowledge of different levels of
qualitative and quantitative measurements may be more useful to compare
between sexes, populations, and species.

Measuring Sexual Selection

A quick perusal of the mating-system literature shows that there are many
methods developed and advocated over the years. Measurements of sexual
selection fall into two main categories: (1) qualitative measurements that pre-
dict the direction and intensity of mate competition and hence sexual selection
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 109

and (2) quantitative measurements that capture the direction and intensity of
sexual selection using variance-based, quantitative theoretical approaches
(Table 4.1). In general, if you are interested in understanding how demogra-
phy and social interactions influence the direction and intensity of mate com-
petition, parental investment, and the organization of social mating systems,
then qualitative measurements of the mating system such as the operational
sex ratio and the potential reproductive rate are typically employed. If, on
the other hand, you would like to investigate how a particular trait is related
to mating and reproductive success, or whether or not sexual selection is likely
to occur in your own particular population or species, then quantitative mea-
surements of sexual selection based on variance in mating and reproductive
success would be more appropriate.

Qualitative Measurements of Sexual Selection

There have been several qualitative measurements of the direction and
intensity of sexual selection developed over the years, including the opera-
tional sex ratio and the potential reproductive rate. Additional qualitative mea-
sures of sexual selection such as mean spatial crowding (m*) and mean
temporal crowding (t*) focus on how ecological factors such as mate-
encounter rates, the distribution of individuals in breeding territories, and
breeding synchrony influence mating-system organization (Shuster & Wade,
2003; Shuster, 2009). Because these measures generally predict the outcome
of sexual selection by estimating the outcome of mating competition, these
measures may be considered qualitative measurements of sexual selection to
differentiate them from measurements that are related to quantitative selection
theory.

The Operational Sex Ratio

The operational sex ratio predicts the direction and intensity of sexual selec-
tion based on mating competition as judged by the number of adults ready to
mate in a population at a given time (Emlen & Oring, 1977). A bias in the
operational sex ratio should reflect the strength of this competition and there-
fore should translate into higher interspecific competition for access to the
limiting sex (Emlen & Oring, 1977; Kvarnemo & Ahnesjö, 1996; Weir et
al., 2011). The operational sex ratio can also be used to predict which sex will
compete for access to mates, thereby determining the sex roles for each sex
(Gwynne, 1991). Thus, whether a population exhibits conventional sex roles,
that is, the condition when males compete more intensely for access to
females, or sex-role reversal, or the condition that females compete more
 Table 4.1 Qualitative and quantitative methods commonly used to measure sexual selection and mating system variation.
Method Description Notes Citation

Qualitative methods
Operational sex ratio The ratio of males to females ready Predicts the direction and Emlen, 1976; Emlen & Oring,
(OSR) to mate in a population at a given intensity of mate competition 1977
time based on a population sample of
adults.
Potential reproductive The maximum number of Predicts the direction of sexual Clutton-Brock & Vincent,
rate (PRR) independent offspring that parents selection in species with biparental 1991; Clutton-Brock & Parker,
can produce per unit of time care. 1992
Mean spatial crowding The number of available mates or Predicts the direction and Lloyd, 1967; Shuster & Wade,
110

(m*) competitors that an average intensity of mate competition 2003

individual experiences per unit of based on the spatial clustering of
space individuals.
Mean temporal crowding The number of available mates or Predicts the direction and Shuster & Wade, 2003
(t*) competitors that an average intensity of mate competition
individual experiences per unit of based on the temporal receptivity
time of individuals.
Quantitative methods
Opportunity for selection Variance in reproductive success Standardized variance in Crow, 1958; Wade, 1979;
(I) divided by the mean reproductive reproductive success. Must be > 0 Arnold & Wade, 1984
success squared for selection to operate.
 Table 4.1 (Continued)

Opportunity for sexual Variance in mating success Standardized variance in mating Wade & Arnold, 1980
selection (IS) divided by the mean mating success success. Must be > 0 for selection
squared to operate.
Bateman gradient ( SS) The slope of the weighted least- Linear relationship between Arnold & Duvall, 1994
squares regression of reproductive mating and reproductive success.
success on mating success Must be > 0 for selection to
operate.
Sex difference in sexual Difference of the opportunity for In purely polygynous mating Shuster & Wade, 2003
selection (Imates) selection of males (Imales) and the systems, the sex difference in the
opportunity for selection of females opportunity for selection I
(Ifemales) at a sex ratio of unity represents the reproductive success
111

attributable to mating success.

Maximum standardized The product of the standardized The maximum potential for trait Jones, 2009
selection differential Bateman gradient ( 0 SS) and the evolution attributable to
(s0 max) square root of the opportunity of precopulatory sexual selection per
sexual selection (IS) generation in units of phenotypic
standard deviations.
Morisita’s index (I) Observed variance in mating Measure of variance based on a Morisita, 1962; Kokko et al.,
success corrected by the expected Poisson distribution. 1999
variance in mating success when all
mate acquisition probabilities are
equal.
(continued)
Table 4.1 (Continued)
Method Description Notes Citation

Index of resource Ratio of observed variance in Measure of variance based on a Ruzzante et al., 1996; Kokko et
monopolization (Q) mating success to the maximum Poisson distribution. al., 1999
possible variance in mating success
corrected by the expected variance
when all mate acquisition
probabilities are equal.
Sexual selection on
traits
Standardized selection Covariance between trait values The strength of selection on a Lande ,1979; Lande & Arnold,
differential (s0 ) and relative reproductive success. trait relating to reproductive 1983
success in the univariate case.
Standardized mating Covariance between trait values The strength of sexual selection Jones 2009
differential (m0 ) and relative mating success. on a trait relating to mating
success in the univariate case.
Standardized selection The partial regression of relative For multiple trait analysis, the Lande, 1979; Lande & Arnold,
gradient ( 0 ) reproductive success or relative overall contribution of a particular 1983; Arnold & Duvall, 1994
mating success on a trait. trait to the sexual selection process
while holding all other traits
constant.
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 113

intensely for access to males, is largely dependent upon the operational sex
ratio (Vincent et al., 1992; Kvarnemo & Ahnesjö, 1996).
The operational sex ratio is an instantaneous ratio of ready-to-mate males to
females in a given population (Emlen & Oring, 1977; Kvarnemo & Ahnesjö,
1996). For many reasons, not all adults can mate at any particular point in
time. For example, individuals may be excluded from mating due to processes
associated with inter- or intrasexual competition, or perhaps they may choose
to delay reproduction until they acquire the proper prerequisites to mating
such as a territory or sufficient food resources (Kvarnemo, 1996; Ahnesjö et
al., 2001). Moreover, adults may need sufficient time after mating to become
prepared to reproduce again. Therefore calculations of the operational sex ratio
should include all individuals that are able to mate in a population but not
those that are physiologically (but not environmentally) unprepared to mate
at a given time.
Mathematically, the operational sex ratio can be expressed either as a simple
ratio of males to females (m/f ) (Emlen & Oring, 1977) or as a relative ratio of
males to total number of adults in a population [m/(m+f )] (Kvarnemo &
Ahnesjö, 1996). While both expressions are often used interchangeably
throughout the literature, it should be noted that they are not equivalent
and have different mathematical properties. The simple ratio resembles the rel-
ative ratio near equality (i.e., equal numbers of males and females) but has
asymptotic ends at highly skewed operational sex ratios (Figure 4.1). Thus,
the simple ratio should be scaled logarithmically to help normalize the data
(Figure 4.1). However, to aid comparisons between studies and for any linear
analysis of the operational sex ratio, the relative expression is preferable
because of its linear properties (Figure 4.1). Significant deviations from an
operational sex ratio of equality are generally tested with either a χ 2
goodness-of-fit or a similar test (e.g., Fisher’s exact test), which compares the
numbers of observed individuals with the number of expected individuals
under a sex ratio of equality (or a skewed adult sex ratio in some instances).
Because this ratio is calculated on the actual numbers of individuals, these tests
are independent of the two methods to calculate the operational sex ratio (Fig-
ure 4.1).
Several variants of the operational sex ratio theme can found in the litera-
ture. First, the operational sex ratio can be expressed as the ratio of the amount
of “time in” invested by each sex (Clutton-Brock & Parker, 1992). Here, time
in is equal to the amount of time when an adult is sexually active or capable of
mating, and time out is the amount of time resting or preparing for mating.
Thus, the direction and intensity of mating competition can arise when sexes
invest differentially in such factors as parental care or spend more time
Figure 4.1. Properties of the operational sex ratio of a population of 50 sexually mature
adults calculated using a simple ratio [log10(m/f )] or as a relative ratio [m/(m/f )]. As
the operational sex ratio becomes more skewed, the simple ratio is asymptotic, whereas
the relative ratio remains linear. The zone of equality is calculated as an operational sex
ratio that does not deviate significantly from a sex ratio of unity (χ2 two-tailed P < 0.05).
Note that the zone of inequality corresponds to the same ratio of m/f in both calculations.
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 115

available for mating than the opposite sex. The operational sex ratio can also
be expressed as a competitor-to-resource ratio (CRR), or the number of poten-
tial competitors divided by the number of resource units, when resource units
are members of the opposite sex (Grant et al., 2001). The advantage of this
formulation is that it allows for the direct comparison of behavior of males
and females on a similar scale (e.g., Clark & Grant, 2010). Additionally, the
ratio of males and females qualified to mate (Q) has been advocated in situa-
tions where individuals need to acquire a specific resource such as a nest site
prior to mating (Ahnesjö et al., 2001). The authors argue that the quantity
Q provides a more accurate measure of operational sex ratio by excluding indi-
viduals that are physically able to mate but are not qualified to mate (Ahnesjö
et al., 2001). However, one should bear in mind that this measure offers an
incomplete picture of sexual selection as individuals that do not hold a re-
source are excluded via competitive interactions and, as such, are actually part
of mate competition (Shuster & Wade, 2003).
Empirical studies support that the operational sex ratio correctly predicts
the strength and direction of sexual selection in many populations and species
(reviewed in Thornhill & Alcock, 1983; Andersson, 1994; Kvarnemo &
Ahnesjö, 1996; Mitani et al., 1996; Weir et al., 2011). A recent meta-
analysis of 27 cross-taxa empirical studies comparing the operational sex ratio
with mate competition demonstrated that the operational sex ratio correctly
predicts the direction of aggression, courtship, and mate-guarding behaviors
but not necessarily postcopulatory mate competition such as sperm competi-
tion (Weir et al., 2011).
Despite the overwhelming empirical support for this theory, the opera-
tional sex ratio has been scrutinized for several reasons. Chief among the crit-
icisms is that the mathematical formulation of the operational sex ratio may
not correctly predict the intensity of mate competition and is decoupled from
mating systems (Kokko & Jennions, 2008; Shuster, 2009). For example, two
hypothetical breeding populations of five ready-to-mate males and females will
have an identical operational sex ratio of unity whether a single male monop-
olizes all five females (polygyny) or all five males have a single exclusive mate
(monogamy). Although this example is extreme, it clearly illustrates that the
operational sex ratio does not correctly predict the strength of sexual selection
among populations with disparate mating systems (Shuster, 2009).
Additional misunderstandings may arise when considering how the opera-
tional sex ratio is quantified in various species. For instance, depending on
the species of interest it may be difficult to tell whether an individual is in fact
ready to mate without invasive techniques, adding imprecision to the meas-
urement, particularly in small populations. Moreover, because the operational
116 ANIMAL BEHAVIOR

sex ratio is an instantaneous measure, calculation at different time intervals in

mobile organisms may lead to faulty conclusions. For example, take a small
patch of suitable breeding habitat where females maintain a territory around
nests while males, having no predefined breeding territory, may enter or leave
the breeding territory at will. Thus, on any given day, the operational sex ratio
is completely dependent upon the presence or absence of males, which may or
may not reflect actual mating competition. Further, new males that enter the
territory at the same time that males leave the territory will give similar esti-
mates of operational sex ratio despite the potential for males to have different
qualities such that female competition for good mates may change the mate-
competition dynamics. Lastly, one should bear in mind that mating competi-
tion may not a have a simple relationship with operational sex ratio. For exam-
ple, at highly biased operational sex ratios competition for mates may be so
fierce that it is an uneconomical strategy for members of the competing sex.
In these cases, operational sex ratio forms a dome-shaped curve with competi-
tive aggression (Grant et al., 2001; Weir et al., 2011). In spite of these short-
comings, the operational sex ratio is a useful metric to formulate hypotheses
concerning mate competition and continues to be a highly lauded theory in
behavioral ecology.

The Potential Reproductive Rate

Tim Clutton-Brock and colleagues (1991, 1992) introduced the concept of
potential reproductive rate as a method of identifying which sex limits repro-
duction based on parental investment theory. The potential reproductive rate
is equivalent to the maximum number of independent offspring that parents
can produce per unit of time (Clutton-Brock & Vincent, 1991). This method
was originally developed to calculate the direction of sexual selection in species
with biparental care where characterization of the relative parental investment
of each sex is not immediately obvious. In principle, the sex that produces
fewer offspring per unit of time is the limiting sex, and therefore sexual selec-
tion should act more intensely on the nonlimiting sex. In practice, measuring
the potential reproductive rate requires detailed information concerning the
duration of care, the size of clutches and broods, the time between clutches,
and the maximum rate of reproduction of males and females within a popula-
tion (Clutton-Brock & Vincent, 1991; Clutton-Brock & Parker, 1992).
To date, the potential reproductive rate has garnered theoretical and empir-
ical support from various animal mating systems (Clutton-Brock & Vincent,
1991; Clutton-Brock & Parker, 1992), although there are some notable
exceptions (e.g., Sogabe & Yanagisawa, 2007; Maurer et al., 2011). However,
the potential reproductive rate has been criticized on several fronts. For
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 117

instance, the original calculation of the potential reproductive rate is depen-

dent upon the rate of reproduction of the most fecund female and most virile
male in a population, and therefore information gathered on all individuals
within a population is not utilized (Arnold & Duvall, 1994; Ahnesjö et al.,
2001). An additional consideration is that the potential reproductive rate can
depend upon many ecological variables (Kvarnemo & Ahnesjö, 1996) such
that only through careful experimental manipulations that control for various
environmental factors can one measure the potential reproductive rate accu-
rately (Ahnesjö et al., 2001; Scobell et al., 2009). This, in turn, requires a focal
individual to be mated with an excess of mates in the absence of mate compe-
tition—a situation unlikely to occur under natural conditions. Thus, if under-
standing sexual selection in nature is our goal, then the potential reproductive
rate is not a particularly useful metric.

Spatial and Temporal Mean Crowding

The intensity of sexual selection and mate competition are predicted to
fluctuate if the spatial distribution and temporal receptivity of mates varies in
nature (Emlen & Oring, 1977; Shuster & Wade, 2003). This relationship
has been interpreted mathematically by applying a measure of density depen-
dence termed mean crowding to both the spatial and temporal distribution of
potential mates or same-sex competitors (Lloyd, 1967; Wade, 1995; Shuster
& Wade, 2003). Mean spatial crowding (m*) can be calculated as
, where is the mean density of individuals per unit
of area (patch size, quadrat size) and is the variance of receptive mates
per unit of area. Similarly, mean temporal crowding (t*) is equivalent to
, where is the mean number of receptive mates per unit
of time (or time interval) and is the variance of receptive mates per unit
of time. Mean crowding can be thought of as a linear measurement of how
aggregated a resource is, either spatially or temporally, and using the relation-
ship of m* to to estimate patchiness helps to visualize this relationship.
High levels of patchiness reflect high aggregations of available mates
in time or space, and stronger sexual selection should be experienced in patchy
aggregations due to increased competition. Low levels of patchiness ( )
show a more even distribution of mates or competitors and hence weaker sex-
ual selection.
Probably the most important consideration in applying mean crowding to a
particular study system is the problem of scale (e.g., Levin, 1992). Lloyd
(1967) commented in his original proposal for mean crowding that the scale
would need to be determined empirically for each case but gave no real guide-
lines as to how this might be accomplished. For behavioral studies, the
118 ANIMAL BEHAVIOR

theoretical sampling scale for mean spatial crowding might best be approxi-
mated by territory size or perhaps mean perception or interaction distance
(i.e., the mean distance at which an individual either notices or interacts with
another individual, respectively). However, this may be difficult to investigate
empirically in many species, so one might often need to make an educated
guess as to what the scale might actually be for one’s own particular study sys-
tem point (Lloyd, 1967; Shuster & Wade, 2003). Shuster and Wade (2003)
recommend a spatial scale that gives the greatest patchiness as a reasonable
starting point. Mean temporal crowding, on the other hand, can best be
defined in terms of the temporal availability of mates within a breeding season.
If all individuals of one sex are available to breed at one particular time, then
this would represent a maximum receptivity, and competition would be lim-
ited. Conversely, minimum receptivity represents a situation where all individ-
uals of one sex become sexually active sequentially such that the ability of the
opposite sex to monopolize mates becomes realistic in time. Thus, maximum
and minimum breeding receptivity for a particular population set an upper
and lower limit to mean temporal crowding, and scaling should reflect the
mean receptivity (Shuster & Wade, 2003).
Mean spatial crowding has been applied to various mating systems, and
studies show varying support for the link between mean spatial crowding
and competitive interactions, depending on the species and social mating sys-
tem (Wade, 1995; Kwiatkowski & Sullivan, 2002; Kelly, 2008; Pomfret &
Knell, 2008; Casalini et al., 2010). For example, species with resource-
defense social mating systems do not always show a relationship with mean
spatial crowding and the intensity of mate competition (Kwiatkowski & Sulli-
van, 2002; Kelly, 2008; Casalini et al., 2010), and this discrepancy is likely
due to different behavioral strategies employed under various competitive sit-
uations. Temporal crowding, on the other hand, has not been widely applied
to empirical research, although this approach could be especially rewarding
in understanding breeding synchrony in a variety of species. Mean spatial
and temporal crowding are also part of a theoretical framework for classifying
mating systems with the sex difference in sexual selection (Imates; Shuster &
Wade, 2003), so we return to these measures in that section.

Quantitative Measurements of Sexual Selection

Darwin recognized that sexual selection arises from differences in reproduc-
tive success caused by competition for access to mates (Darwin, 1871; Ander-
sson, 1994). Quantitative measurements of sexual selection are based on the
underlying principle that because all offspring are from one mother and one
father (i.e., the Fisherian condition; Queller, 1997), if an individual mates
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 119

and produces offspring with more than one mate, than the variance in mating
success is not equal among members of that sex, given a sex ratio of equality.
This nonrandom variance in Darwinian fitness generated through differential
mating success is a mathematical formulation connected directly with
Darwin’s original definition of sexual selection.
Quantitative measurements of sexual selection, particularly if they are com-
bined with precise genetic parentage information, offer an instantaneous snap-
shot of all precopulatory sexual selection in a population without any special
reference or need to quantify behavioral mate competition or specific traits
of interest. However, an underlying assumption in all variance-based methods
is that both mating and reproductive success can be accurately measured for a
reasonable number of individuals within a population. This may pose a real
problem in some species for a variety of reasons. For example, it may be diffi-
cult to assess with accuracy parentage in populations where numbers of adults
are too large to completely sample or when the number of adults and progeny
are too few to provide enough statistical power (Jones & Ardren, 2003). Fur-
thermore, variance-based approaches require careful estimation of the number
of adults that are ready to mate but do not breed in a population. These indi-
viduals that are able but are excluded from breeding are sometimes referred to
as the zero class because they have zero mating success (Shuster & Wade,
2003). While there are no hard-and-fast rules concerning who to include in
the zero class, researchers need to make a special effort to clarify and justify
who they include as these individuals can have a disproportionate effect on
the observed variance in mating and reproductive success (Shuster & Wade,
2003; Klug, Lindström, et al., 2010; Fitze & Le Galliard, 2011). Finally,
researchers should clarify which offspring they include when measuring repro-
ductive success. Although the argument can be made for including only off-
spring that reproduce, for many species this requires too much time or
expense to be a viable option. Therefore the number of developed embryos
or independent offspring produced during a breeding bout would be a reason-
able estimate of reproductive success.

Bateman’s Principles
The renewed interest in Bateman’s work in the 1970s provided a theoreti-
cal framework for sexual selection centered on both the variation in mating
and reproductive success. Based in large part on the verbal arguments made
by Bateman (1948), three specific points were articulated and are now referred
to as Bateman’s principles (Arnold, 1994). To reiterate, the three principles
are (1) the sex experiencing the strongest sexual selection has a greater variance
in reproductive success, (2) the sex experiencing the strongest sexual selection
120 ANIMAL BEHAVIOR

has a greater variance in mating success, and (3) the relationship between
reproductive success and mating success increases with the intensity of sexual
selection (Arnold & Duvall, 1994). Each of these three principles correspond
to the mathematical formulation of the opportunity for selection (I), the
opportunity for sexual selection (Is) and the sexual selection gradient (or
Bateman’s gradient; βss), respectively. Together, Bateman’s principles pro-
vide a means to predict the potential for sexual selection to occur in a sex
within a generation without necessarily knowing which traits may be under
selection.

Opportunity Measures
The first of Bateman’s principles, the opportunity for selection, is equiva-
lent to the strength of selection arising from the standard variance in fitness
attributable to reproductive success for each sex and is sometimes referred to
as reproductive skew. The opportunity for selection can be calculated as the
variance in reproductive success divided by the mean reproductive success
squared, or , where 2 is the variance in number of offspring and
is the mean number of offspring per individual in a population (Crow,
1958, 1962; Wade, 1979). An alternative and numerically identical approach
to estimating I is to calculate the variance in relative reproductive success by
dividing each individual’s absolute reproductive success by the mean repro-
ductive success for each sex (Jones, 2009).
The opportunity for selection is equivalent to the rate that absolute fitness
increases in a population relative to the standing variance in absolute fitness,
assuming all variance is due to additive genetic effects (Jones, 2009). Thus, I
represents a theoretical maximum rate at which selection may act on a popula-
tion in terms of offspring production (Wade, 1979). However, Jones (2009)
demonstrated that the true theoretical maximum response to selection of any
trait is actually , which has the added advantage of being in units of trait
phenotypic standard deviations. The change in fitness from one generation
to the next is likely to be less than predicted by I (or ) because not all vari-
ance in fitness in a population is due to additive genetic variance (Jones,
2009). Therefore researchers often refer to I (and Is) as a mathematical upper
limit for selection operating in a single generation and hence an “opportunity”
measure.
The opportunity for sexual selection (Is) is similar in all respects to the cal-
culation of I with the exception that it focuses on the standard variance in mat-
ing success for each sex in a population (Wade, 1979). Thus, Is is calculated as
the variance in mating success divided by the mean mating success squared, or
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 121

, where 2ms is the variance in number of mates and is the mean

number of mates per individual in a population.
The simplest and best method to compare I or Is statistically between two
groups (sexes, times, populations, etc.) is via an unequal variance test on rela-
tive reproductive or mating success, respectively. Several unequal variance tests
exist, such as Bartlett’s or Levene’s, although Levene’s test is preferable because
it is more robust to the assumptions of normality and therefore offers a more
conservative criterion (Sokal & Rohlf, 1995). Opportunity measures are the
most commonly used measures to quantify sexual selection and enjoy theo-
retical support in various species and mating systems (reviewed in Searcy &
Yasukawa, 1995; Shuster & Wade, 2003; Mills et al., 2007; Jones, 2009;
Klug, Heuschele, et al., 2010). The advantages of opportunity measures
include the facts that they are unitless measures that aid comparisons between
different sexes, populations, and species and that the measures are directly
linked with selection theory (Wade & Shuster, 2005; Jones, 2009). However,
opportunity measures are harshly scrutinized, and criticisms include (1) the
nonindependence of variance on population means, (2) the absence of a direct
link to traits of interest, (3) introduction of variance through random pro-
cesses, (4) the potential for opportunity measures to correctly identify sexual
selection when the strength of sexual selection is weak, and (5) that they do
not include information concerning mate quality (Downhower et al., 1987;
Grafen, 1987; Ruzzante et al., 1996; Fairbairn & Wilby, 2001; Klug,
Heuschele, et al., 2010). While criticism 1 has largely been dismissed (Nonacs,
2003; Jones, 2009) and criticism 2 is reconciled through selection differentials
on traits of interest (Lande & Arnold, 1983; Jones, 2009; Klug, Heuschele,
et al., 2010), criticisms 3, 4, and 5 present more difficult challenges. The
grievance with criticism 3 (i.e., random processes introducing variance in
opportunity measures) can be ameliorated by incorporating appropriate null
models to determine whether variances are greater than expected by chance
and by including power analyses (Klug, Heuschele, et al., 2010). Criticism
4, aimed at correctly identifying sexual selection when selection is weak, can
be partially reconciled with a multiple-test approach, and authors now advo-
cate using both opportunity measures, Bateman gradient approaches and the
maximum standardized strength of selection, whenever possible (Jones,
2009; Krakauer et al., 2011). However, when selection is weak, it will always
be difficult to differentiate between actual selection and stochastic processes,
regardless of what method you use to estimate sexual selection. The remaining
criticism applies to all qualitative and quantitative measures proposed thus far,
and therefore more empirical and theoretical work to include issues concern-
ing mate quality and other sources of variation is required to solve this issue.
122 ANIMAL BEHAVIOR

The Bateman Gradient

Stevan Arnold and David Duvall (1994) proposed that the strength of sex-
ual selection can be measured by the relationship between mating success and
reproductive success as quantified by the sexual selection gradient, ss. The
sexual selection gradient is equivalent to the slope of the weighted regression
line comparing reproductive success to mating success (Arnold & Duvall,
1994). This specific relationship was renamed the Bateman gradient by Malte
Andersson and Yoh Iwasa (1996) to differentiate it from selection gradients on
phenotypic traits (Lande & Arnold, 1983).
Not only did the Bateman gradient represent a step forward by clarifying
the relationship between mating success and Darwinian fitness expressed as
offspring production, it provided a statistically rigorous and visually appealing
method to demonstrate the strength and direction of sexual selection (Figure
4.2). The steeper the slope of the Bateman gradient experienced by a sex, the
stronger sexual selection is likely to operate on that sex. Another way to inter-
pret ss is that a sex that is limited in reproduction by extrinsic factors such as
access to mating opportunities should have a steeper  ss slope while a shal-
lower slope suggests that a sex is limited by reproduction based on intrinsic
factors such as the potential to produce gametes or provide parental care (Bate-
man, 1948; Arnold & Duvall, 1994). Because mating success necessarily
needs to translate into increased fitness, this relationship must be positive in
order for sexual selection to operate (Arnold & Duvall, 1994).
The Bateman gradient can be calculated on a particular sex using regression
analysis and is usually compared between groups with covariance methods
such as an analysis of covariance (e.g., Mobley & Jones, 2007, 2009; Barreto
& Avise, 2010). A similar measure, the standardized Bateman gradient (0 ss),
equivalent to the slope of the weighted least regression line comparing relative
reproductive success to relative mating success, is also currently used to com-
pare different groups. This measurement is preferable to the uncorrected esti-
mates of ss as it standardizes differences in the means of the two measures,
allowing for more interpretable comparisons between groups of interest
(Jones, 2009).
Bateman gradients have been applied in a variety of species, and there is
general agreement with estimates of  ss and the direction and intensity of sex-
ual selection predicted by theory (e.g., Jones et al., 2005; Kvarnemo et al.,
2007; Webster et al., 2007). Bateman gradients have also been employed to
help explain patterns of sex-role reversal (Jones et al., 2000), anisogamy (Bjork
& Pitnick, 2006), and postcopulatory sexual selection (Lorch, 2002) and even
have been ascribed as a method to quantify sexual selection in hermaphrodites
(Arnold, 1994; Anthes et al., 2010). Bateman gradients have also been
Figure 4.2. Bateman gradients for male and female broad-nosed pipefish, Syngnathus
typhle, in two experimental breeding populations (data from Jones et al., 2005). Male
Bateman gradients are represented by solid lines and closed circles, and female Bateman
gradients are represented by dashed lines and open circles. In nature, the species is sex-
role reversed and the operational sex ratio is female-biased. In the female-biased treat-
ment, the strength of sexual selection is greater in females than in males as evidenced by
the more positive slope of the female Bateman gradient. Under male-biased experimental
conditions, the strength of sexual selection among males is greater than among females as
evidenced by a more positive slope of the male Bateman gradient. All Bateman gradients
are significantly greater than zero (P < 0.001), demonstrating that sexual selection is
likely operating in both sexes. Note that Bateman gradients are not standardized.
124 ANIMAL BEHAVIOR

criticized as too simplistic a view because they do not incorporate information

such as variation in mate quality, the difficulty to procure matings, or the costs
associated with developing traits that confer a mating advantage (Jennions &
Kokko, 2010; Klug, Heuschele, et al., 2010; Fitze & Le Galliard, 2011).
However, these criticisms have been appreciated by previous authors who pro-
pose Bateman gradients as one method to estimate whether or not sexual selec-
tion could be occurring without a priori knowledge of the targets of selection
(Arnold & Duvall, 1994; Jones et al., 2004; Jones, 2009; Krakauer et al.,
2011).

Sex Difference in Sexual Selection

Closely related to Bateman’s principles, the sex difference in sexual selec-
tion (Imates) has also been used as a measurement of sexual selection (Wade,
1979, 1995; Wade & Arnold, 1980; Shuster & Wade, 2003). The origin of
this measure can be traced to Wade (1979), who considered the source of
variation in reproductive success from males under a strictly polygynous mat-
ing system. It was determined that all variance in male mating success is equiv-
alent to the sex difference in standardized reproductive success of males.
Mathematically this relationship can be expressed as Imales = (1/Ro)Ifemales +
Imates [or Imates = Imales − (1/Ro)Ifemales] when Ro is equivalent to the sex ratio
expressed as males/females. The derivation of I mates can be found in both
Wade (1979) and Shuster and Wade (2003).
In mating systems that deviate from strict polygyny, the variance in male
mating success can no longer be described as a simple function of female
reproductive success as females may split their reproductive output and invest
differently in the offspring of several males. Thus in polyandrous and polygy-
nandrous mating systems, a more generalized equation is recommended: Imales
= QIclutch + Is, where Q is the reciprocal of mean male mating success, Iclutch is
the standardized variance in female clutch size (i.e., the number of offspring
produced with each male), and Is is the opportunity for sexual selection on
males (Jones, 2009). This slightly more complicated formula requires precise
information on female clutch size, which can be estimated from parentage
analysis. However, whenever female multiple mating is infrequent, this calcu-
lation approximates the more general equation for strict polygyny.
The sex difference in sexual selection has been applied extensively to a
robust theoretical framework combining Imates in conjunction with mean spa-
tial and temporal crowding (m*, t*; see above) as a means to classify various
existing mating systems (Shuster & Wade, 2003; Wade & Shuster, 2004;
Shuster, 2009). By categorizing both the temporal and spatial variation in con-
junction with the strength and direction of sexual selection, this framework
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 125

can help explain the myriad of social and genetic mating systems in existence
including sex-role reversal, alternative mating behaviors, and plant mating sys-
tems (Shuster & Wade, 2003; Shuster, 2009).

Maximum Standardized Selection Differential

Jones (2009) proposed a method that combines elements of Bateman’s
principles into a new metric for estimating the maximum strength of sexual
selection experienced in units of phenotypic standard deviations. The maxi-
mum standardized selection differential is calculated as the product of the
standardized Bateman gradient and the square root of the opportunity for sex-
ual selection ( ). This measure is equivalent to the maximum rate
of phenotypic evolution for a given trait in one generation and therefore is
similar to other opportunity measures in that it makes no assumptions about
the heritability of traits or fitness (Jones, 2009). This new metric has the ad-
vantage of focusing specifically on selection on fitness generated by differential
mating success and therefore is more informative than either I or Is alone.
So far, the maximum standardized sexual selection differential has only been
applied in a handful of studies but adheres well to theoretical predictions. For in-
stance, values of s0max are greater in males than in females in species that have con-
ventional sex roles (Serbezov et al., 2010) and display markedly similar values in
species that appear to lack strong sexual selection between sexes (Barreto & Avise,
2010). Further application of this technique will demonstrate whether it is a
useful metric to determine the direction and intensity of sexual selection.

The Morisita Index and the Index of Resource Monopolization

Several additional variance-based methods have been introduced as alterna-
tive measures of the potential for sexual selection, including the index of re-
source monopolization (Q; Ruzzante et al., 1996) and the Morisita index
( ; Morisita, 1962; Kokko et al., 1999). The index of resource monopoliza-
tion is defined as , and the Morisita index is
, where n is the number of individuals, xi is the
value of the resource (i.e., either mating or reproductive success) for the ith
individual, is the mean value across n individuals, and 2 is the variance.
The argument for these methods is that these measures based on expected
Poisson distributions are not as sensitive to population means and therefore
are a better metric of sexual selection (Kokko et al., 1999; Fairbairn & Wilby,
2001). However, this criticism has been dismissed as fallacious (Nonacs,
2003). Empirical studies have demonstrated a disparity between the
126 ANIMAL BEHAVIOR

theoretical expectations of sexual selection theory and estimates of Q and I

(Jones et al., 2004; Croshaw, 2010), and both of these methods are heavily
criticized for their redundancy (Jones et al., 2004; Mills et al., 2007). Perhaps
the greatest drawback to applying I and Q is a lack of connection to selection
theory (Jones et al., 2004; Croshaw, 2010). As a result of the above concerns,
neither of these methods have enjoyed widespread acceptance for quantifying
sexual selection.

Selection on Sexually Selected Traits

Measuring sexual selection on traits of interest is a relatively straightforward
task in principle. One can collect a sample of breeding adults and offspring
from a population, measure traits that might be under sexual selection, and
quantify the relationship between the trait and either mating success or repro-
ductive success. However, one must bear in mind that an underlying
assumption is that the correct target of selection has been identified. If you
do not find sexual selection operating on a particular trait, then sexual selec-
tion may still exist, just not on that particular trait at the time of sampling.
Likewise, if you do see sexual selection imparted on a trait, then the trait
may be correlated with the actual target of sexual selection and not under sex-
ual selection itself. Thus, the combination of trait analysis and other quantita-
tive approaches such as I, Is, ss, and s0 max is highly recommended (Jones et al.,
2004; Shuster, 2009; Krakauer et al., 2011; but see Klug, Heuschele, et al.,
2010, for a contrasting opinion).
The metric commonly used to measure selection on traits of interest is the
standardized selection differential (s0 ), equivalent to the covariance between a
trait and either mating or reproductive success (Lande & Arnold, 1983). For
this measure, trait values are standardized by variance to have a mean of zero
and a variance of one [(z – )/(standard deviation of z), where z is an individ-
ual’s trait value and is the trait mean] so that traits can be compared in units
of phenotypic standard deviations (Lande & Arnold, 1983). In this manner,
selection differentials are an intuitive measure of selection as they predict
how much the average phenotype evolves from one generation to the next,
assuming that the genetic variance is completely additive (Lande, 1979; Lande
& Arnold, 1983). To help differentiate between the selection differentials of
particular traits to either mating success and reproductive success, the stand-
ardized mating differential (m0 ) has been coined for the specific selection dif-
ferential of traits on relative mating success (Jones, 2009).
Under the simplified scenario of one trait acting on mating and reproduc-
tive success, analyses are straightforward and linear regression models are gen-
erally applied (e.g., Jones et al., 2004, 2005). If traits do not satisfy the
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 127

assumptions of normality or equal variances, transformations (e.g., ln, log10)

should be applied before the trait has been standardized by variance. In the
more complicated case where multiple traits may be under selection simulta-
neously, the standardized selection gradient (0 ) is calculated as the partial
regression of a standardized trait value on either relative mating success or
reproductive success (s0 , m0 ) while holding the values of other traits constant
(Lande & Arnold, 1983; Arnold & Duvall, 1994). Thus, the standardized
selection gradient can be calculated using multiple regression techniques on
traits or on principal-components scores (e.g., Jones et al., 2004; Head et al.,
2008; Kelly, 2008; Hunt et al., 2009). However, because complicated genetic
relationships can give rise to multidimensional or rugged fitness surfaces, other
methods such as canonical analyses may be required to disentangle complex
nonlinear relationships of multiple traits under selection (Phillips & Arnold,
1989; Blows & Brooks, 2003).
Surprisingly, methods to quantify sexual selection on traits enjoy consider-
able agreement despite having similar issues with random processes that may
generate spurious correlations between traits and mating or reproductive suc-
cess, as do other quantitative mating system measures (Klug, Heuschele, et
al., 2010; Krakauer et al., 2011). However, because selection differentials
and gradients are correlational analyses, it is sometimes difficult to disentangle
true causative relationships between the trait and the measure of selection
(Krakauer et al., 2011). Thus, convincing evidence for traits under selection
is generally the result of manipulative experiments rather than field observa-
tions (Jones et al., 2004, 2005; Kelly, 2008; Hunt et al., 2009).

TWO EMPIRICAL EXAMPLES

To demonstrate how various measures of sexual selection behave under dif-
ferent selection regimes, I present two case studies where mating competition
was manipulated by altering the operational sex ratio of breeding adults (Table
4.2). In both experiments, adults were placed in replicated artificial breeding
populations at different sex ratios and allowed to mate freely over a period of
time. For simplicity, all data from experimental treatments are pooled
together, and therefore mating-system estimates represent the composite score
across all individuals within the same sex-ratio treatment.
The first example set comes from a study conducted by Jones and col-
leagues (2004) on rough-skinned newts (Taricha granulosa). These salaman-
ders are native to the West Coast of North America and migrate to breeding
ponds during winter months, where mating and egg laying take place. This
species is sexually dimorphic, and males are typically larger and possess a large
Table 4.2 Mating-system estimates from two experimental datasets. The first dataset represents artificial breeding populations of the
rough-skinned newt, Taricha granulosa, at two different operational sex ratios [OSR ¼ m/(m + f )]. This species is polygynadrous and
displays conventional sex roles in natural populations. Selection and mating differentials are given for tail height. The second dataset
represents artificial breeding populations of the broad-nosed pipefish, Syngnathus typhle, at three different adult sex ratios. This species is
polygynandrous and sex-role reversed in natural populations. Selection and mating differentials are given for body length. Data are
reproduced with permission from Jones et al. (2004) and Jones et al. (2005; with permission).
Conventional sex roles: Taricha granulosa
pffiffiffiffi pffiffiffi
Treatment OSR Sex N X ms ms Is Is X X rs rs I I  ss  0 ss s0 max Imates* m0 s0

Male bias 0.75 Male 48 0.44 0.46 2.42 1.56 59.2 11899.2 3.39 1.84 141.5 1.04 1.62 3.28 0.14 0.36
Female 16 1.31 0.50 0.29 0.54 177.7 10801.7 0.34 0.58 20.0 0.15 0.08 0.05 –0.14
Even 0.52 Male 48 1.35 0.87 0.48 0.69 153.2 20282.9 0.86 0.93 114.6 1.01 0.70 0.57 0.20 0.36
Female 45 1.49 0.66 0.30 0.55 164.9 8634.0 0.32 0.57 22.4 0.20 0.11 0.03 –0.10
Reversed sex roles: Syngnathus typhle
pffiffiffiffi pffiffiffi
Treatment OSR Sex N X ms ms Is Is X X rs rs I I  ss  0 ss s0 max Imates* m0 s0

Female bias 0.24 Male 13 2.23 1.19 0.24 0.49 58.4 1286.9 0.38 0.61 15.4 0.59 0.29 −6.57 0.05 0.37
Female 42 0.69 0.71 1.48 1.22 18.1 702.3 2.15 1.47 27.9 1.07 1.30 0.04 0.23
Even 0.48 Male 20 1.70 0.85 0.30 0.54 41.0 1162.4 0.69 0.83 19.0 0.79 0.43 –0.03 0.14 0.51
Female 22 1.36 0.81 0.44 0.66 34.0 763.3 0.66 0.81 17.5 0.70 0.46 0.25 0.38
Male bias 0.73 Male 32 1.31 1.45 0.84 0.92 30.8 1181.0 1.25 1.12 29.4 1.25 1.15 1.17 0.32 0.60
Female 12 2.92 1.54 0.18 0.43 70.1 942.8 0.19 0.44 14.5 0.61 0.26 −0.03 0.22
*
Imates is calculated using the general formula ascribed in Shuster & Wade (2003).
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 129

tail crest, which appears to be under sexual selection (Janzen & Brodie, 1989;
Jones et al., 2002). Males mate multiply, and females can lay multiple clutches
of mixed paternity, and therefore the species is considered to be polygynan-
drous (Jones et al., 2002).
In this experiment, Jones and colleagues (2005) mated breeding popula-
tions of either a male-biased treatment of six males and three females or an
even-sex-ratio treatment of six males and six females (Table 4.2). Under
male-biased conditions, we would expect that sexual selection would be
strongest among males either through male-male competition or female
choice. Thus, mating-system estimates such as Is, I,  0 ss, and s0 max as well as
selection differentials (m0 , s0 ) on male tail height should be higher on males
in the male-biased treatment over the even-sex-ratio treatment.
Under male-biased sex ratios, females mated on average three times as much
as males, leading to higher estimates of for females although ms was similar
for both sexes. Male was likewise reduced, and both sexes shared similar esti-
mates of rs. Under a sex ratio of equality, male and female and were
similar, and ms was also comparable, leading to similar values of Is. However,
because some males produced more offspring than others in the even-sex-ratio
treatment, the variance in male reproductive success (rs) was twice the rs for
females, leading to higher estimates of I,  ss,  0 ss, and s0 max for males. As
expected, in male-biased cases with conventional sex ratios, the strengths of sex-
ual selection as measured by Is, I,  ss, 0 ss, and s0 max were all greater in males than
in females and greater in females than in males in the even-sex-ratio treatment. It
is interesting to note that female mating-system estimates were surprisingly sim-
ilar between the two treatments, strongly suggesting that females are limited in
their reproductive potential and that the addition of more mates does not sub-
stantially influence mating and reproductive success. Finally, the sex difference
in the opportunity for selection (Imates) shows that the strength of sexual selection
is stronger in males under male-biased conditions.
For selection on the sexually selected trait of tail height, both m0 and s0 were
similar among males in both operational sex ratios despite a nearly double
potential for sexual selection on traits as judged by s0 max (Table 4.2). Females,
on the other hand, had small estimates of both m0 and s0 , indicating sexual
selection to be weak or nonexistent in females. In summary, all mating-
system estimates, including those specifically on a trait under selection, agree
with the prediction that sexual selection in this species is stronger on males,
although a more strongly biased operational sex ratio towards males does not
necessarily increase sexual selection on male traits.
The second example comes from a study conducted by Jones and
colleagues (2005) on the broad-nosed pipefish (Syngnathus typhle). Males
130 ANIMAL BEHAVIOR

and females mate multiply within a single pregnancy, and thus the particular
population of interest is polygynandrous with respect to the genetic mating
system (Jones et al., 1999). In the wild, this species generally encounters
highly female-biased operational sex ratios and is sex-role reversed in relation
to the direction of sexual selection (Berglund et al., 1986; Vincent et al.,
1994). Therefore the a priori expectation is that sexual selection should act
most strongly on females under conditions of an operational sex ratio of
equality.
In this experiment, a change in operational sex ratio from male bias to female
bias showed an increase in male and a concomitant decrease in female
(Table 4.2). These changes in means were primarily driven by many males that
did not mate in the male-biased treatment. An opposite relationship, namely a
decrease in male and an increase in female , was apparent with an
increasing female bias. While the male ms was variable between treatments,
female ms decreased with decreasing female bias, and the rs of both males
and females remained similar between treatments with some slight variations.
As the operational sex ratio shifts from a female bias to a male bias, the resulting
mating-system estimates Is, I, βss, β0 ss, and s0 max all show a clear pattern of increas-
ing in males while decreasing in females (Table 4.2, Figure 4.2). The sex differ-
ence in the opportunity for selection (Imates) also shows a clear pattern of
increasing from female bias to male bias, demonstrating that this measurement
correctly predicts sex-role reversal as evidenced by negative values encountered
at sex ratios of equality and female bias.
Interestingly, selection differentials on body size demonstrate conflicting
patterns between the sexes. The highest m0 and s0 on male body size are found
in the male-biased treatment, and this is in line with the expectation that larger
males enjoy higher mating and reproductive success because they can carry
more offspring in their brood pouch. Selection on female body size, on the
other hand, appears to be highest in the even-sex-ratio treatment, suggesting
stabilizing selection on female body size (Jones et al., 2005).

FUTURE CHALLENGES
Thus far we have explored the history and current theory concerning mat-
ing systems and the evolution of sexual selection. Although much progress
has been made, still more work is required in order to gain a deeper apprecia-
tion for the intricacies of mating-system evolution. In particular, uncovering
the causal link between social and genetic mating systems, investigating how
mating systems vary both spatially and temporally, understanding the myriad
of social and ecological influences on mating system organization, and
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 131

differentiating between precopulatory and postcopulatory sexual selection are

logical next steps in the field. Here, I briefly introduce these topics as potential
lines of further research.

Social versus Genetic Mating Systems

The degree to which social mating systems influence genetic mating systems
and vice versa is still an unresolved question in behavioral ecology. For some spe-
cies of seahorses, the social mating system may appear promiscuous despite a
purely monogamous genetic mating system (Wilson & Martin-Smith, 2007).
On the other hand, in many species of socially monogamous birds, high levels
of extra-pair copulations demonstrate that, despite the maintenance of stable pair
bonds, a portion of the population cuckolds their primary mates and solicits cop-
ulations from potentially better mates (Hasselquist & Sherman, 2001; Griffith et
al., 2002). As illustrated by these two examples, reconciling these two disparate
types of mating systems seems to depend upon many variables such as the
strength of social bonds, the degree of extra-pair mating, the presence of alterna-
tive mating behaviors, and other ecological factors.
One approach to disentangle the relationship between social and genetic mat-
ing systems can be through careful experimentation and manipulation of the
social interactions of adults in a breeding population. For example, several
manipulative experiments demonstrate a clear relationship between the opera-
tional sex ratio and various measures of the genetic mating system (e.g., Jones
et al., 2000; Mills & Reynolds, 2003; Jones et al., 2004; Mills et al., 2007;
Reichard, Ondračková, et al., 2008). These studies are the most informative in
understanding the link between demographic variables and the actual strength
of sexual selection. Comparative phylogenetic studies also have the potential
to unearth how social and genetic mating systems may evolve in concert. For
instance, studies such as those that compare sexual selection among socially co-
operative versus solitary breeding birds suggest that sexual selection on cooperative
breeders is reduced via social interactions. The nature of these cooperative social
interactions, in turn, influences the evolution of mate competition and sexually
selected traits within these species (Rubenstein & Lovette, 2009). While the
strength of the relationship between social and genetic mating systems is quite
likely to differ on a species-to-species basis, more studies along these lines are war-
ranted and necessary to gain a greater understanding of how these systems are
linked and feedback under different evolutionary scenarios.

Mating-System Variation
Sexual selection can manifest considerable variation between populations
and at different times (Mobley & Jones, 2007; Cockburn et al., 2008; Gosden
132 ANIMAL BEHAVIOR

& Svensson, 2008; Cornwallis & Uller, 2009; Mobley & Jones, 2009).
Certainly part of this variation is due to both spatial and temporal environ-
mental heterogeneity. Variation may also arise because behavioral traits and
responses to selection may be plastic in space and time (Cornwallis & Uller,
2009). Populations that experience disparate intensities of sexual selection
may theoretically drive patterns of speciation, strongly implying that mating-
system variation is important to the speciation process (Payne & Krakauer,
1997; Kraaijeveld et al., 2007). Thus, understanding how and why mating
systems vary is critical to decoding the process of sexual selection and the evo-
lution of different mating systems (Emlen & Oring, 1977; Shuster & Wade,
2003; Cornwallis & Uller, 2009).
Thus far, only a few studies concerning variation in mating systems have
been conducted, and therefore the underlying causes of mating-system varia-
tion are not well known. Often, mating-system studies focus on one exemplar
population at one specific time, and therefore the degree to which the mating
system varies is a black box in many cases. Part of the problem with con-
ducting studies within and among populations is the large expense and work
required to completely characterize the mating systems. However, studies
along these lines would pay dividends for elucidating specific factors that influ-
ence mating-system organization by mapping mating-system parameters to
specific details of the timing, habitat, and social interactions encountered over
various spatial or temporal scales. For example, studies that investigate mating
systems in multiple populations over broad geographic scales have identified a
wide array of ecological factors that may shape local mating-system dynamics
(e.g., Mobley & Jones, 2007, 2009).
Despite the relatively more studied underlying ecological causes for sexual
selection within seasons or mating periods (see “Ecological Mating-System
Influences”), there is surprisingly little known about what ecological variables
may drive temporal patterns of sexual selection between mating seasons or
years (Cornwallis & Uller, 2009). Recent studies suggest that variation
between years can be greater than within years (Cockburn et al., 2008;
Kasumovic et al., 2008), leading to the prediction that year-specific factors
may influence differences in sexual selection. Year-specific factors may be the
result of fluctuations in climate (Twiss et al., 2007; Cockburn et al., 2008)
or due to demographic processes (Kokko & López-Sepulcre, 2007; Gosden
& Svensson, 2008; Kasumovic et al., 2008). However, variation between years
might not be attributable to any particular environmental variable measured,
suggesting that in some cases sexual selection may fluctuate between years
for unknown reasons (Lehtonen et al., 2009). Thus, comparing variation in
mating and reproductive success within and between years should illuminate
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 133

whether sexual selection varies due to specific ecological variables or results

from stochastic processes such as environmental heterogeneity (Cornwallis &
Uller, 2009).

Ecological Mating-System Influences

A wide range of ecological variables have the potential to influence mating
and reproductive behaviors of individuals within a population, and under-
standing what these variables are is the first step in better understanding how
species adapt and respond to fluctuating environmental conditions. Ecological
factors can be divided into two main categories: (1) abiotic variables—
extrinsic nonbiological properties of the environment and habitat that the
population occupies—and (2) biotic variables, which include biological prop-
erties of the habitat extrinsic to the focal population as well as demographic
properties intrinsic to the population itself.
Among abiotic factors that affect mating systems, temperature is the most
studied and best understood, particularly because of its strong effect on com-
munity productivity, reproductive timing, and limiting reproductive rates,
particularly in poikilothermic organisms (Ahnesjö, 1995, 2008; Kvarnemo,
1996; Fischer et al., 2003; Olsson et al., 2011). Other abiotic factors that
may affect sexual selection include climate (Cockburn et al., 2008) and
anthropogenic pollution (Lane et al., 2011), but such factors remain an under-
studied aspect of mating systems. In contrast, a wide range of biotic factors
affecting the strength and direction of sexual selection have been identified,
including (but not limited to) factors that affect food availability and hence
fecundity and mate quality (Kvarnemo, 1997; Turner & McCarty, 1998),
habitat structure and fragmentation (Turner & McCarty, 1998; Reichard,
Ondračková, et al., 2008), parasite load (Fitze et al., 2004), resource competi-
tion (Martin & Martin, 2001), and predation (Kelly et al., 1999; Bronikowski
et al., 2002).
Still another gap in our knowledge is how and to what degree demographic
processes influence various aspects of sexual selection and mating-system evo-
lution. By far the most studied demographic factor explored to date is sex
ratio, but other factors such as population density, the length and synchrony
of the breeding season, and effective breeding population size are likely be
important to the sexual selection process (Griffith et al., 1999; Jirotkul,
1999; Spottiswoode & Møller, 2004; Kokko & Rankin, 2006; Mobley &
Jones, 2007, 2009; Reichard, Smith, et al., 2008). Fluctuations in population
density have the ability to affect mate encounter rates that influence processes
such as mate competition, mate choice, and parental care decision making by
134 ANIMAL BEHAVIOR

providing individuals with more or better choices of potential mates (Kokko

& Rankin, 2006). Variation in population density also can expose individuals
to differing levels of competition for access to mates among the same sex
(Kokko & Rankin, 2006). Moreover, population density may govern the
expression and evolution of behaviors or traits of one sex, which in turn influ-
ence the opposite sex via social and ecogenetic feedback mechanisms (Kokko
& López-Sepulcre, 2007; Alonzo, 2009). Thus the future challenge is to test
the effect of demographic factors both within and between sexes to better
understand how these are related to the sexual selection process.

Postcopulatory Sexual Selection

Notwithstanding his deep appreciation for precopulatory sexual selection
including mate competition and mate choice, Darwin did not comment on
the potential for postcopulatory sexual selection, or selection that takes place
during or after mating, to occur (Eberhard, 2009). Postcopulatory sexual
selection includes the phenomenon of gamete competition as well as the pro-
cess of cryptic mate choice (Parker, 1970; Eberhard, 1996; Birkhead &
Møller, 1998; Simmons, 2001). While these processes are usually referred to
as sperm competition and cryptic female choice, I opt for the more general
definitions to include the potential for opposite patterns to evolve in sex-
role-reversed species (i.e., egg competition and cryptic male choice) (Partridge
et al., 2009; Paczolt & Jones, 2010).
Postcopulatory processes have the potential to affect sexual selection in
many species, but the degree to which postcopulatory sexual selection influen-
ces mating-system evolution is currently unknown. In many instances we can-
not a priori distinguish between pre- and postcopulatory sexual selection, and
therefore a particular challenge for future studies is to partition the relative
influences of these two selective forces. One method to tackle this problem
is to design an experiment (or series of experiments) that can readily distin-
guish between the two processes. These experiments could be complemented
by parentage analysis utilizing molecular markers to assign unambiguous par-
entage to all offspring. In this manner, one can partition the variation experi-
enced by the two forces, making the relative contribution of each more easily
understood.
To date, there have been only a few attempts to quantify the total variance
attributed to precopulatory and postcopulatory processes in the same species
despite the potential for postcopulatory sexual selection to be strong (Hunt
et al., 2009; Simmons & Beveridge, 2010). One example comes from a study
on broad-nosed pipefish where the potential for postcopulatory sexual selec-
tion appears to be weak compared to all precopulatory sexual selection
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 135

experienced (Partridge et al., 2009). Beetles, on the other hand, demonstrate a

negative correlation between precopulatory weapons and postcopulatory testes
size, indicating that a trade-off between the two selective forces is likely among
these species (Simmons & Emlen, 2006). For the most complete picture on
the relative effects of precopulatory versus postcopulatory sexual selection,
more studies along these lines are warranted and are necessary to resolve this
outstanding issue.

CONCLUSIONS
The classification of animals by various reproductive behaviors related to
the sexual selection process has garnered much interest in the times since Dar-
win’s formulation of the theory of sexual selection. As we have seen in this
chapter, the field of animal mating systems is dynamic and prone to conten-
tious debate, making it one of the most exciting fields in animal behavior.
Although we have made much progress, a dizzying array of questions related
to mating systems and sexual selection is still left outstanding. For example,
what accounts for the stunning diversity of social and genetic mating systems?
Why do some lineages experience strong sexual selection and others do not? How
might divergence in mating systems contribute to the speciation process? Why do
various forms of parental care evolve? Why are some species sex-role reversed? What
are the demographic and ecological factors that influence mating system organiza-
tion? While definitive answers to these questions seem out of reach at the
moment, one thing is clear from our continued study of sexual selection—it
is a powerful evolutionary process that explains a staggering variety of behav-
ioral and morphological traits in animals. In the future, the study of mating
systems will continue to grow, and further theoretical refinement in concert
with carefully conceived behavioral experiments and field surveys should lead
to the ultimate goal of understanding how sexual selection operates in nature.

ACKNOWLEDGMENTS
I thank Malte Andersson, Tonje Aronsen, Anders Berglund, Inês Braga-
Gonçalves, Charlotta Kvarnemo, Gunilla Rosenqvist, Sebastian Wacker, and
Bob Wong for insightful discussions. I am particularly indebted to Adam
Jones and Steve Shuster for providing valuable advice and encouragement
throughout the years.
136 ANIMAL BEHAVIOR

REFERENCES AND SUGGESTED READING

Ahnesjö, I. (1995). Temperature affects male and female potential reproductive rates
differently in the sex-role reversed pipefish, Syngnathus typhle. Behavioral Ecol-
ogy, 6, 229–233.
Ahnesjö, I. (2008). Behavioural temperature preference in a brooding male pipefish
Syngnathus typhle. Journal of Fish Biology, 73, 1039–1045.
Ahnesjö, I., C. Kvarnemo, & S. Merilaita. (2001). Using potential reproductive rates
to predict mating competition among individuals qualified to mate. Behavioral
Ecology, 12, 397–401.
Alonzo, S. H. (2009). Social and coevolutionary feedbacks between mating and
parental investment. Trends in Ecology and Evolution, 25, 99–108.
Andersson, M. (1994). Sexual Selection. Princeton, NJ: Princeton University Press.
Andersson, M. & Y. Iwasa. (1996). Sexual selection. Trends in Ecology and Evolution,
11, A53–A58.
Anthes, N., D. Patrice, J. R. Auld, J. N. A. Hoffer, P. Jarne, J. M. Koene, et al.
(2010). Bateman gradients in hermaphrodites: An extended approach to quan-
tify sexual selection. American Naturalist, 176, 249–263.
Arnold, S. J. (1994). Bateman principles and the measurement of sexual selection in
plants and animals. American Naturalist, 144, S126–S149.
Arnold, S. J. & D. Duvall. (1994). Animal mating systems: A synthesis based on
selection theory. American Naturalist, 143, 317–348.
Arnold, S. J. & M. J. Wade. (1984). On the measurement of natural and sexual selec-
tion: Theory. Evolution, 38, 709–719.
Avise, J. C. (2004). Molecular Markers, Natural History and Evolution. Sunderland,
MA: Sinauer Associates.
Barreto, F. S. & J. C. Avise. (2010). Quantitative measures of sexual selection reveal
no evidence for sex-role reversal in a sea spider with prolonged paternal care.
Proceedings of the Royal Society of London, B, 277, 2951–2956.
Bateman, A. J. (1948). Intra-sexual selection in Drosophila. Heredity, 2, 349–368.
Berglund, A., G. Rosenqvist, & I. Svensson. (1986). Reversed sex-roles and parental
energy investment in zygotes of two pipefish (Syngnathidae) species. Marine
Ecology—Progress Series, 29, 209–215.
Berglund, A., G. Rosenqvist, & I. Svensson. (1989). Reproductive success of females
limited by males in two pipefish species. American Naturalist, 133, 506–516.
Birkhead, T. R. & A. P. Møller. (1998). Sperm Competition and Sexual Selection. San
Diego, CA: Academic Press.
Bjork, A. & S. Pitnick. (2006). Intensity of sexual selection along the anisogamy-
isogamy continuum. Nature, 7094, 742–745.
Blows, M. W. & R. Brooks. (2003). Measuring nonlinear selection. American Natu-
ralist, 162, 815–820.
Bronikowski, A. M., M. E. Clark, F. H. Rodd, & D. N. Reznick. (2002). Population-
dynamic consequences of predator-induced life history variation in the guppy
(Poecilia reticulata). Ecology, 83, 2194–2204.
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 137

Casalini, M., M. Reichard, & C. Smith. (2010). The effect of crowding and density
on male mating behaviour in the rose bitterling (Rhodeus ocellatus). Behaviour,
147, 1035–1050.
Clark, L. & J. W. A. Grant. (2010). Intrasexual competition and courtship in female
and male Japanese medaka, Oryzias latipes: Effects of operational sex ratio and
density. Animal Behaviour, 80, 707–712.
Clutton-Brock, T. H. (1991). The Evolution of Parental Care. Princeton, NJ: Prince-
ton University Press.
Clutton-Brock, T. H. & G. A. Parker. (1992). Potential reproductive rates and the
operation of sexual selection. Quarterly Review of Biology, 67, 437–456.
Clutton-Brock, T. H. & A. J. C. Vincent. (1991). Sexual selection and the potential
reproductive rates of males and females. Nature, 351, 58–60.
Cockburn, A., H. L. Osmond, & M. C. Double. (2008). Swingin’ in the rain: Con-
dition dependence and sexual selection in a capricious world. Proceedings of the
Royal Society of London, B, 275, 605–612.
Coleman, S. W. & A. G. Jones. (2011). Patterns of multiple paternity and maternity
in fishes. Biological Journal of the Linnean Society, 103, 735–760.
Cornwallis, C. K. & T. Uller. (2009). Towards an evolutionary ecology of sexual
traits. Trends in Ecology and Evolution, 25, 145–152.
Croshaw, D. A. (2010). Quantifying sexual selection: A comparison of competing
indices with mating system data from a terrestrially breeding salamander. Bio-
logical Journal of the Linnean Society, 99, 73–83.
Crow, J. F. (1958). Some possibilities of measuring selection intensity in man.
Human Biology, 30, 1–13.
Crow, J. F. (1962). Population genetics: Selection. In W. J. Burdette (ed.), Methodol-
ogy in Human Genetics (pp. 53–75). San Francisco: Holden-Day.
Darwin, C. (1871). The Descent of Man and Selection in Relation to Sex. London: John
Murray.
Downhower, J. F., L. S. Blumer, & L. B. Brown. (1987). Opportunity for selection:
An appropriate measure for evaluating variation in the potential for selection?
Evolution, 41, 1395–1400.
Dunn, P. O., L. A. Whittingham, & T. E. Pitcher. (2001). Mating systems, sperm
competition, and the evolution of sexual dimorphism in birds. Evolution, 55,
161–175.
Eberhard, W. G. (1996). Female Control: Sexual Selection by Cryptic Female Choice.
Princeton, NJ: Princeton University Press.
Eberhard, W. G. (2009). Postcopulatory sexual selection: Darwin’s omission and its
consequences. Proceedings of the National Academy of Sciences, USA, 106,
10025–10032.
Emlen, S. T. (1976). Lek organization and mating strategies in bullfrog. Behavioral
Ecology and Sociobiology, 1, 283–313.
Emlen, S. T. & L. W. Oring. (1977). Ecology, sexual selection and the evolution of
mating systems. Science, 197, 215–223.
138 ANIMAL BEHAVIOR

Fairbairn, D. J. & A. E. Wilby. (2001). Inequality of opportunity: Measuring the

potential for sexual selection. Evolutionary Ecology Research, 3, 667–686.
Fischer, K., P. M. Brakefield, & B. J. Zwaan. (2003). Plasticity in butterfly egg size:
Why larger offspring at lower temperatures? Ecology, 84, 3138–3147.
Fisher, R. A. (1915). The evolution of sexual preference. Eugenics Review, 7, 184–192.
Fisher, R. A. (1930). The Genetical Theory of Natural Selection. Oxford, UK: Claren-
don Press.
Fitze, P. S. & J. F. Le Galliard. (2011). Inconsistency between different measures of
sexual selection. American Naturalist, 178, 256–268.
Fitze, P. S., B. Tschirren, & H. Richner. (2004). Life history and fitness consequences
of ectoparasites. Journal of Animal Ecology, 73, 216–226.
Foster, S. J. & A. Vincent. (2004). Life history and ecology of seahorses: Implications
for conservation and management. Journal of Fish Biology, 65, 1–61.
Gosden, T. P. & E. I. Svensson. (2008). Spatial and temporal dynamics in a sexual
selection mosaic. Evolution, 62, 845–856.
Grafen, A. (1987). Measuring sexual selection: Why bother? In J. W. Bradbury & M.
B. Andersson (eds.), Sexual Selection: Testing the Alternatives (pp. 221–233).
Hoboken, NJ: John Wiley and Sons.
Grant, J. W. A., C. L. Gaboury, & H. L. Levitt. (2001). Competitor-to-resource ratio, a
general formulation of operational sex ratio, as a predictor of competitive aggres-
sion in Japanese medaka (Pisces: Oryziidae). Behavioral Ecology, 11, 670–675.
Griffith, S. C., I. P. F. Owens, & K. A. Thuman. (2002). Extra pair paternity in birds:
A review of interspecific variation and adaptive function. Molecular Ecology, 11,
2195–2212.
Griffith, S. C., I. R. K. Stewart, D. A. Dawson, I. P. F. Owens, & T. Burke. (1999).
Contrasting levels of extra-pair paternity in mainland and island populations of
the house sparrow (Passer domesticus): Is there an “island effect”? Biological
Journal of the Linnean Society, 68, 303–316.
Gwynne, D. T. (1991). Sexual competition among females: What causes courtship-
role reversal? Trends in Ecology and Evolution, 6, 118–121.
Hasselquist, D. & P. W. Sherman. (2001). Social mating systems and extrapair fertil-
izations in passerine birds. Behavioral Ecology, 12, 457–466.
Head, M. L., A. K. Lindholm, & R. Brooks. (2008). Operational sex ratio and density
do not affect directional selection on male sexual ornaments and behavior. Evo-
lution, 62, 135–144.
Hubbell, S. P. & L. K. Johnson. (1987). Environmental variance in lifetime
mating success, mate choice, and sexual selection. American Naturalist, 130,
91–112.
Hunt, J., C. J. Breuker, J. A. Sadowki, & A. J. Moore. (2009). Male-male competi-
tion, female mate choice and their interaction: Determining total sexual selec-
tion. Journal of Evolutionary Biology, 22, 13–26.
Janzen, F. J. & E. D. I. Brodie. (1989). Tall tails and sexy males: Sexual behavior of
rough-skinned newts (Taricha granulosa) in a natural breeding pond. Copeia,
1989, 1068–1071.
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 139

Jennions, M. D. & H. Kokko. (2010). Sexual selection. In D. F. Westneat & C. W.

Fox (eds.), Evolutionary Behavioral Ecology (pp. 343–364). Oxford, UK:
Oxford University Press.
Jirotkul, M. (1999). Population density influences male-male competition in guppies.
Animal Behaviour, 58, 1169–1175.
Jones, A. G. (2009). On the opportunity for sexual selection, the Bateman gradient
and the maximum intensity of sexual selection. Evolution, 63, 1673–1684.
Jones, A. G. & W. R. Ardren. (2003). Methods of parentage analysis in natural pop-
ulations. Molecular Ecology, 12, 2511–2523.
Jones, A. G., J. R. Arguello, & S. J. Arnold. (2002). Validation of Bateman’s princi-
ples: A genetic study of sexual selection and mating patterns in the rough-
skinned newt. Proceedings of the Royal Society of London, B, 269, 2533–2539.
Jones, A. G., J. R. Arguello, & S. J. Arnold. (2004). Molecular parentage analysis in
experimental newt populations: The response of mating system measures to
variation in the operational sex ratio. American Naturalist, 164, 444–456.
Jones, A. G. & J. C. Avise. (2001). Mating systems and sexual selection in male-
pregnant pipefishes and seahorses: Insights from microsatellite-based studies
of maternity. Journal of Heredity, 92, 150–158.
Jones, A. G., C. Kvarnemo, G. I. Moore, L. W. Simmons, & J. C. Avise. (1998).
Microsatellite evidence for monogamy and sex-biased recombination in the
Western Australian seahorse Hippocampus angustus. Molecular Ecology, 7,
1497–1505.
Jones, A. G. & N. L. Ratterman. (2009). Mate choice and sexual selection: What
have we learned since Darwin? Proceedings of the National Academy of Sciences,
USA, 106, 10001–10008.
Jones, A. G., G. Rosenqvist, A. Berglund, S. J. Arnold, & J. C. Avise. (2000). The
Bateman gradient and the cause of sexual selection in a sex-role-reversed pipe-
fish. Proceedings of the Royal Society of London, B, 267, 677–680.
Jones, A. G., G. Rosenqvist, A. Berglund, & J. C. Avise. (1999). The genetic mating
system of a sex-role-reversed pipefish (Syngnathus typhle): A molecular inquiry.
Behavioral Ecology and Sociobiology, 46, 357–365.
Jones, A. G., G. Rosenqvist, A. Berglund, & J. C. Avise. (2005). The measurement of
sexual selection using Bateman’s principles: An experimental test in the sex-
role-reversed pipefish Syngnathus typhle. Integrative and Comparative Biology,
45, 874–884.
Jones, A. G., C. M. Small, K. A. Paczolt, & N. Ratterman. (2010). A practical guide
to methods of parentage analysis. Molecular Ecology Resources, 10, 6–30.
Kasumovic, M. M., M. J. Bruce, M. C. B. Andrade, & M. E. Herberstein. (2008).
Spatial and temporal demographic variation drives within-season fluctuations
in sexual selection. Evolution, 62, 2316–2325.
Kelly, C. D. (2008). Identifying a causal agent of sexual selection on weaponry in an
insect. Behavioral Ecology, 172, 184–192.
140 ANIMAL BEHAVIOR

Kelly, C. D., J. G. J. Godin, & J. M. Wright. (1999). Geographical variation in

multiple paternity within natural populations of the guppy (Poecilia reticulata).
Proceedings of the Royal Society of London, B, 266, 2403–2408.
Kirkpatrick, M., T. D. Price, & S. J. Arnold. (1990). The Darwin-Fisher theory of
sexual selection in monogamous birds. Evolution, 44, 180–193.
Klug, H., J. Heuschele, M. D. Jennions, & H. Kokko. (2010). The mismeasurement
of sexual selection. Journal of Evolutionary Biology, 23, 447–462.
Klug, H., K. Lindström, & H. Kokko. (2010). Who to include in measures of sexual
selection is no trivial matter. Ecology Letters, 13, 1094–1102.
Koening, W. D. & S. S. Albano. (1986). On the measurement of sexual selection.
American Naturalist, 72, 358-382.
Kokko, H. & M. D. Jennions. (2008). Parental investment, sexual selection and sex
ratios. Journal of Evolutionary Biology, 21, 919–948.
Kokko, H. & A. López-Sepulcre. (2007). The ecogenetic link between demography
and evolution: Can we bridge the gap between theory and data? Ecology Letters,
10, 773–782.
Kokko, H., A. Mackenzie, J. D. Reynolds, K. Lindström, & W. J. Sutherland.
(1999). Measures of inequality are not equal. American Naturalist, 72,
358–381.
Kokko, H. & D. J. Rankin. (2006). Lonely hearts or sex in the city? Density-
dependent effects in mating systems. Philosophical Transactions of the Royal
Society of London, B, 361, 319–334.
Kraaijeveld, K., F. J. L. Kraaijeveld-Smit, & J. Komdeur. (2007). The evolution of
mutual ornamentation. Animal Behaviour, 74, 657–677.
Krakauer, A. H., M. S. Webster, E. H. Duvall, A. G. Jones, & S. M. Shuster. (2011).
The opportunity for sexual selection: Not mismeasured, just misunderstood.
Journal of Evolutionary Biology, 24, 2064–2071.
Kvarnemo, C. (1996). Temperature affects operational sex ratio and intensity of male-
male competition: Experimental study of sand gobies, Pomatoschistus minutus.
Behavioral Ecology, 7, 208–212.
Kvarnemo, C. (1997). Food affects the potential reproductive rates of sand goby
females but not of males. Behavioral Ecology, 8, 605–611.
Kvarnemo, C. & I. Ahnesjö. (1996). The dynamics of operational sex ratios and com-
petition for mates. Trends in Ecology and Evolution, 11, 404–408.
Kvarnemo, C., G. I. Moore, & A. G. Jones. (2007). Sexually selected females in the
monogamous Western Australian seahorse. Proceedings of the Royal Society of
London, B, 274, 521–525.
Kwiatkowski, M. A. & B. K. Sullivan. (2002). Geographic variation in sexual
selection among populations of an iguanid lizard Sauromalus obesus (=ater).
Evolution, 56, 2039–2051.
Lande, R. (1979). Quantitative genetical analysis of multivariate evolution, applied to
brain:body size allometery. Evolution, 33, 402–416.
Lande, R. & S. J. Arnold. (1983). The measurement of selection on correlated
characters. Evolution, 37, 1210–1226.
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 141

Lane, J. E., M. N. K. Forrest, & C. K. R. Willis. (2011). Anthropogenic influences on

natural animal mating systems. Animal Behaviour, 81, 909–917.
Lehtonen, T., B. B. M. Wong, & K. Lindström. (2009). Fluctuating mate prefer-
ences in a marine fish. Biology Letters, 6, 21–23.
Levin, S. A. (1992). The problem of pattern and scale in ecology: The Robert H.
MacArthur Award Lecture. Ecology, 73, 1943–1967.
Lloyd, M. (1967). Mean crowding. Journal of Animal Ecology, 36, 1030.
Lorch, P. D. (2002). Understanding reversals in the relative strength of sexual selec-
tion on males and females: A role for sperm competition? American Naturalist,
159, 645–657.
Martin, P. R. & T. E. Martin. (2001). Ecological and fitness consequences of
species coexistence: A removal experiment with wood warblers. Ecology, 82,
189–206.
Maurer, G., M. C. Double, O. Milenkaya, M. Süsser, & R. D. Magrath. (2011).
Breaking the rules: Sex roles and genetic mating system of the pheasant coucal.
Oecologia, 167, 413–425.
Mills, S. C., A. Grapputo, E. Koskela, & T. Mappes. (2007). Quantitative measure of
sexual selection with respect to the operational sex ratio: a comparison of selec-
tion indices. Proceedings of the Royal Society of London, B, 274, 143–150.
Mills, S. C. & J. D. Reynolds. (2003). Operational sex ratio and alternative reproduc-
tive behaviours in the European bitterling, Rhodeus sericeus. Behavioral Ecology
and Sociobiology, 54, 98–104.
Mitani, J. C., J. GrosLouis, & A. F. Richards. (1996). Sexual dimorphism, the opera-
tional sex ratio, and the intensity of male competition in polygynous primates.
American Naturalist, 147, 966–980.
Mobley, K. B. & A. G. Jones. (2007). Geographical variation in the mating system of
the dusky pipefish (Syngnathus floridae). Molecular Ecology, 16, 2596–2606.
Mobley, K. B. & A. G. Jones. (2009). Environmental, demographic and genetic mat-
ing system variation among five geographically distinct dusky pipefish (Syngna-
thus floridae) populations. Molecular Ecology, 18, 1476–1490.
Mobley, K. B., C. M. Small, & A. G. Jones. (2011). The genetics and genomics of
Syngnathidae (pipefishes, seahorses and seadragons). Journal of Fish Biology,
78, 1624–1646.
Møller, A. P. & T. R. Birkhead. (1994). The evolution of plumage brightness in birds
is related to extrapair paternity. Evolution, 48, 1089–1100.
Morisita, M. (1962). I-index, a measure of dispersion of individuals. Research in
Population Ecology, 4, 1–7.
Nonacs, P. (2003). Measuring the reliability of skew indices: Is there one best index?
Animal Behaviour, 65, 615–627.
Olsson, M., E. Wapstra, T. Schwartz, T. Madsen, B. Ujvari, & T. Uller. (2011). In
hot pursuit: Fluctuating mating system and sexual selection in sand lizards.
Evolution, 65, 574–583.
Paczolt, K. A. & A. G. Jones. (2010). Postcopulatory sexual selection and sexual con-
flict in the evolution of male pregnancy. Nature, 464, 401–404.
142 ANIMAL BEHAVIOR

Parker, G. A. (1970). Sperm competition and its evolutionary consequences in

insects. Biological Reviews of the Cambridge Philosophical Society, 45, 525–567.
Parker, G. A., R. R. Baker, & V. G. F. Smith. (1972). The origin and evolution of
gamete dimorphism and the male-female phenomenon. Journal of Theoretical
Biology, 36, 529–553.
Partridge, C., I. Ahnesjö, C. Kvarnemo, K. B. Mobley, A. Berglund, & A. G. Jones.
(2009). The effect of perceived female parasite load on post-copulatory male
choice in a sex-role-reversed pipefish. Behavioral Ecology and Sociobiology, 63,
345–354.
Payne, R. J. H. & D. C. Krakauer. (1997). Sexual selection, space, and speciation.
Evolution, 51, 1–9.
Petrie, M. & B. Kempenaers. (1998). Extra-pair paternity in birds: Explaining varia-
tion between species and populations. Trends in Ecology and Evolution, 13,
52–58.
Phillips, P. C. & S. J. Arnold. (1989). Visualizing multivariate selection. Evolution,
43, 1209–1222.
Pomfret, J. C. & R. J. Knell. (2008). Crowding, sex ratio and horn evolution in a
South African beetle community. Proceedings of the Royal Society of London,
B, 275, 315–321.
Queller, D. C. (1997). Why do females care more than males? Proceedings of the Royal
Society of London, B, 264, 1555–1557.
Reichard, M., M. Ondračková, A. Bryjová, & J. Bryja. (2008). Breeding resource dis-
tribution affects selection gradients on male phenotypic traits: Experimental
study on lifetime reproductive success in the bitterling fish (Rhodeus amarus).
Evolution, 63, 377–390.
Reichard, M., C. Smith, & J. Bryja. (2008). Seasonal change in the opportunity for
sexual selection. Molecular Ecology, 17, 642–651.
Reynolds, J. D. (1996). Animal breeding systems. Trends in Ecology and Evolution, 11,
A68–A72.
Rubenstein, D. R. & I. J. Lovette. (2009). Reproductive skew and selection on female
ornamentation in social species. Nature, 462, 786–790.
Ruzzante, D. E., D. C. Hamilton, D. L. Kramer, & J. W. A. Grant. (1996). Scaling
of the variance and the quantification of resource monopolization. Behavioral
Ecology, 7, 199–207.
Sakai, A. K. & D. F. Westneat. (2001). Mating systems. In C. W. Fox, D. A. Roff, &
D. J. Fairbairn (eds.), Evolutionary Ecology: Concepts and Case Studies (pp. 193–
206). Oxford, UK: Oxford University Press.
Scobell, S. K., A. M. Fudickar, & R. Knapp. (2009). Potential reproductive rate of a
sex-role reversed pipefish over several bouts of mating. Animal Behaviour, 78,
747–753.
Searcy, W. A. & K. Yasukawa. (1995). Polygyny and Sexual Selection in Red-winged
Blackbirds. Princeton, NJ: Princeton University Press.
Serbezov, D., L. Bernatchez, E. M. Olsen, & L. A. Vøllestad. (2010). Mating patterns
and determinants of individual reproductive success in brown trout (Salmo
MATING SYSTEMS AND THE MEASUREMENT OF SEXUAL SELECTION 143

trutta) revealed by parentage analysis of an entire steam living population.

Molecular Ecology, 19, 3193–3205.
Shuster, S. M. (2009). Sexual selection and mating systems. Proceedings of the
National Academy of Sciences, USA, 106, 10009–10016.
Shuster, S. M. & M. J. Wade. (2003). Mating Systems and Strategies. Princton, NJ:
Princeton University Press.
Simmons, L. W. (2001). Sperm Competition and Its Evolutionary Consequences in
Insects. Princeton, NJ: Princeton University Press.
Simmons, L. W. & M. Beveridge. (2010). The strength of postcopulatory sexual
selection within natural populations of field crickets. Behavioral Ecology, 21,
1179–1185.
Simmons, L. W. & D. J. Emlen. (2006). Evolutionary trade-off between weapons and
testes. Proceedings of the National Academy of Sciences, USA, 103, 16346–16351.
Sogabe, A. & Y. Yanagisawa. (2007). Sex-role reversal of a monogamous pipefish
without higher potential reproductive rate in females. Proceedings of the Royal
Society of London, B, 274, 2959–2963.
Sokal, R. R. & F. J. Rohlf. (1995). Biometry. New York: W. H. Freeman and
Company.
Spottiswoode, C. & A. P. Møller. (2004). Extrapair paternity, migration and breeding
synchrony in birds. Behavioral Ecology and Sociobiology, 15, 41–57.
Sutherland, W. J. (1985). Chance can produce a sex difference in variance in mating
success and explain Bateman’s data. Animal Behaviour, 33, 1349–1352.
Sutherland, W. J. (1987). Random and deterministic components of variance in mat-
ing success. In J. W. Bradbury & M. Andersson (eds.), Sexual Selection: Testing
the Alternatives (pp. 209–219). Chichester: Wiley.
Thornhill, R. & J. Alcock. (1983). The Evolution of Insect Mating Systems. Cambridge,
MA: Harvard University Press.
Trivers, R. L. (1972). Parental investment and sexual selection. In B. Campbell (ed.),
Sexual Selection and the Descent of Man 1871–1971 (pp. 136–179). London:
Heinemann.
Turner, A. M. & J. P. McCarty. (1998). Resource availability, breeding site selection,
and reproductive success of red-winged blackbirds. Oecologia, 113, 140–146.
Twiss, S. D., C. Thomas, V. Poland, J. A. Graves, & P. Pomeroy. (2007). The impact of
climatic variation on the opportunity for sexual selection. Biology Letters, 3, 12–15.
Vincent, A., I. Ahnesjo, & A. Berglund. (1994). Operational sex ratios and behavioral
sex differences in a pipefish population. Behavioral Ecology and Sociobiology, 34,
435–442.
Vincent, A., I. Ahnesjo, A. Berglund, & G. Rosenqvist. (1992). Pipefishes and sea-
horses: Are they all sex-role reversed? Trends in Ecology and Evolution, 7,
237–241.
Wade, M. J. (1979). Sexual selection and variance in reproductive success. American
Naturalist, 114, 742–747.
Wade, M. J. (1995). The ecology of sexual selection: Mean crowding of females and
resource-defense polygyny. Evolutionary Ecology, 9, 118–124.
144 ANIMAL BEHAVIOR

Wade, M. J. & S. J. Arnold. (1980). The intensity of sexual selection in relation to

male sexual behavior, female choice and sperm precedence. Animal Behaviour,
28, 446–461.
Wade, M. J. & S. M. Shuster. (2004). Sexual selection: Harem size and the variance
in male reproductive success. American Naturalist, 164, E83–E89.
Wade, M. J. & S. M. Shuster. (2005). Don’t throw Bateman out with the bathwater!
Integrative and Comparative Biology, 45, 945–951.
Webster, M. S., K. A. Tarvin, E. M. Tuttle, & S. Pruett-Jones. (2007). Promiscuity
drives sexual selection in a socially monogamous bird. Evolution, 61, 2205–2211.
Weir, L. K., J. W. A. Grant, & J. A. Hutchings. (2011). The influence of operational
sex ratio on the intensity of competition for mates. American Naturalist, 177,
167–176.
Westneat, D. F. & I. R. K. Stewart. (2003). Extra-pair paternity in birds: causes, cor-
relates, and conflict. Annual Review of Ecology, Evolution and Systematics, 34,
365–396.
Wilson, A. B., I. Ahnesjö, A. C. J. Vincent, & A. Meyer. (2003). The dynamics
of male brooding, mating patterns, and sex roles in pipefishes and seahorses
(family Syngnathidae). Evolution, 57, 1374–1386.
Wilson, A. B. & K. M. Martin-Smith. (2007). Genetic monogamy despite social
promiscuity in the pot-bellied seahorse (Hippocampus abdominalis). Molecular
Ecology, 16, 2345–2352.
 5

The Evolution of Ornaments and Armaments

Geoffrey E. Hill

INTRODUCTION
When Darwin conceived his theory of evolution through natural selection
(Darwin, 1859), he immediately recognized the power of the idea. Natural selec-
tion provided a unifying explanation for a tremendous range of natural phenom-
ena. Unfortunately, and much to Darwin’s consternation, the umbrella of
evolution by natural selection could not quite accommodate the whole of bio-
logical diversity. In particular, it could not adequately explain two of the most
fantastic features of animals: ornaments and armaments (Figure 5.1).
Within the realm of behavioral ecology, an ornament is a trait that enhances
the appearance (in the broad sense that potentially involves any sensory percep-
tion) of an animal, sometimes to the detriment of survival. Elongated feathers,
dewlaps, manes, fleshy wattles, bright coloration, songs, roars, conspicuous move-
ments, and an array of additional sense-stimulating traits of animals appear to be
ornaments (Figure 5.2). An armament, in contrast, is a trait that can serve as an
aid in contests with conspecifics but that is not required for foraging or
protection from predators. Armaments include traits like horns, antlers, spurs,
and—most commonly—a big, strong body. To explain ornaments and arma-
ments, Darwin had to conceive of a mechanism for evolution beyond natural
selection. Whereas Darwin (1859) proposed that natural selection promotes the
maintenance and spread of traits that enhance survival or fecundity, he presented
sexual selection as a process that promotes the maintenance and spread of traits
that enhance access to mates (Darwin, 1871). Darwin reasoned that, for a
sexually reproducing organism, staying alive and having the capacity to produce
offspring are for naught without access to a mate.
146 ANIMAL BEHAVIOR

Figure 5.1. The large antlers of this bull elk (Cervus canadensis) certainly appear to be
armaments. Research indicates, however, that females assess the size and symmetry
of antlers during mate choice, so, like many traits, elk antlers may function as both
armaments and ornaments.

Sexual selection is now accepted among evolutionary biologists as the pri-

mary force that shapes ornaments and armaments, but the specific processes
by which these types of traits evolve remains hotly debated. In this chapter I
will summarize current theory concerning how ornaments and armaments
evolved and how they function in animal populations. Sexual selection can
occur in plants, fungi, and single-celled organisms (Andersson, 1994; Skogs-
myr & Lankinen, 2002) as well as in animals, but no one has proposed that
organisms other than animals evolve ornaments or armaments. Hence, the
focus of this chapter will be sexual selection in animals.

SEXUAL SELECTION AND ANIMAL BEHAVIOR

Ornaments and armaments tend to be morphological traits, so it may seem
curious that the study of such traits is pursued primarily by students of animal
behavior. However, it is behavior, and in particular sensory perception linked
Figure 5.2. The facial structures of this mandrill (Mandrillus sphinx) are the most
colorful traits displayed by any primate and are among the most colorful traits
among mammals. They likely function as both armaments and ornaments but have
not been thoroughly studied.
148 ANIMAL BEHAVIOR

to behavioral responsiveness, that connects the drive to find a mate with the
evolution of traits such as red feathers and pheromones. Obviously, female
choice of mates and physical contests between males play out as behavioral
interactions, but more subtly, sensory perception and behavioral responsive-
ness also play roles in contests for mates when there is assessment prior to
physical confrontation. Because humans are visual animals, we tend to think
of sensory assessment in terms of vision, but animals use a wide range of sig-
naling modes to present their ornaments and armaments, including pheromo-
nal, chemical, electrical, acoustical, and tactile. An ornament can be a plume
of volatile alkaloids released by a moth (Iyengar et al., 2001) just as much as
it can be the visually stunning train of a peacock (Petrie et al., 1991).

FEMALE CHOICE AND MALE-MALE COMPETITION

Sexual selection occurs whenever there is competition among members of
one sex for sexual access to members of the opposite sex. One important factor
in determining the potential for sexual selection is investment in offspring,
including everything from the deposition of nutrients in eggs to the nourish-
ment of a growing embryo to the instruction an adult-sized offspring in hunt-
ing techniques (Ligon, 1999). The sex that invests the most in offspring
becomes a resource that is competed for by the sex that invests less in offspring
(Trivers, 1972). By definition, females have larger gametes and males have
smaller gametes, so the initial investment in offspring by females is greater
than males. In animals with parental care, males can make up the disparity
in initial investment by investing in offspring as much as or even more than
females, but for most animal species, the pattern of greater female investment
in offspring holds through reproduction (Ligon, 1999). The end result is that
females are the choosy sex in most species of animals while males are the dis-
playing and competing sex (Trivers, 1972). For the remainder of this chapter
I will assume that ornaments are the result of female mate choice (technically
termed intersexual selection). Likewise, I will assume that armaments evolve
via male-male competition (technically termed intrasexual selection). In the
unusual cases in which males provide more parental care than females, females
often have more ornaments and armaments and males are the choosy sex
(Ligon, 1999).
Other circumstances that promote sexual selection include greater availabil-
ity of potential mates of one sex compared to the other. Such a disparity in
partner availability can arise strictly from numerically greater representation
of one sex (called a skewed sex ratio; e.g., Hill et al., 1994). Mating systems
that enable one sex to monopolize more than one member of the opposite
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 149

sex can also skew the availability of mates and drive sexual selection even when
an equal number of males and females are present (Andersson, 1994; Ligon,
1999). Such a skew in operational sex ratio occurs most commonly under a
polygynous mating system in which individual males mate with more than
one female (Ligon, 1999). Extra-pair copulation can also increase the avail-
ability of a few males to many females and hence alter the operational sex ratio
of a population (Westneat & Stewart, 2003). Other circumstances that allow
individuals of one sex to monopolize more than one individual of the opposite
sex, such as a temporally staggered period of sexual receptivity, also lead to sex-
ual selection (Stutchbury & Morton, 1995).

MATE CHOICE VERSUS COMPETITION

For the purposes of explanation, it is useful to think of female choice and
male-male competition as distinct processes that lead to the evolution of dis-
tinct sets of secondary sexual traits: ornaments and armaments. For many ani-
mals, however, the processes of choosing mates and competing for access to
mates act simultaneously on the same trait (Berglund et al., 1996). Consider,
as an example, ring-necked pheasants (Phasianus colchicus). Males have dagger-
like projections from their tarsi (lower legs) called spurs that are used to evis-
cerate rivals in fights. For millennia, humans have staged fights between
pheasants and other fowl with spurs on their legs because the birds fight bru-
tally, slashing each other with their spurs. It would seem safe to surmise that
spurs are armaments that evolved through the advantages they bestowed in
male-male competition (Davison, 1985). It turns out, however, that females
assess the size of spurs when choosing a mate (von Schantz et al., 1989), and
choice by females selects for larger spurs. So the spurs of pheasants—as well
as the secondary sexual traits of many other species of animals—are simulta-
neously armaments and ornaments. To keep discussions tractable in this chap-
ter, I will refer to particular traits as being either armaments or ornaments with
the understanding that such a dichotomous characterization is commonly an
oversimplification.
How sexual selection can give rise to armaments and ornaments is a focal
topic of research within evolutionary ecology and behavioral ecology, but
explaining ornaments has proven far more challenging than explaining arma-
ments. Long before there was a theory of sexual selection or an evolutionary
perspective of any sort, it was widely appreciated that male animals used traits
like horns, tusks, and spurs to win contests for sexual access to females. Not
surprisingly, when armaments were reconsidered within the framework of
natural and sexual selection, the hypothesis that male-male competition leads
150 ANIMAL BEHAVIOR

to the evolution of armaments was widely accepted and never seriously chal-
lenged in the behavior and evolution literature (see Cronin, 1991). In con-
trast, prior to Darwin, ornamental traits were traditionally viewed as the
product of special creation whose purpose was to make the world more inter-
esting and enjoyable for humans (see Hill, 2002). From the outset, framing
evolutionary explanations for how ornamental traits evolved proved challeng-
ing and contentious. There was a long period of debate in the late nineteenth
and early twentieth centuries regarding whether female mate choice played
any role in the evolution of ornaments. It is now universally accepted that
female mate choice is an important force in the evolution of ornaments, and
debate has shifted to the process of ornament evolution. The evolution of
ornaments via female choice remains a highly contentious topic in biology
with many hypotheses proposed and no clear consensus among researchers
(Prum, 2010).
In the following sections, I will outline each of the major hypotheses that
have been proposed to explain the evolution of ornaments and armaments
and review the evidence in support of each.

THE EVOLUTION OF ARMAMENTS

The utility of armaments in battles among rival animals is easy to observe.
Contests over mates are often decided in brutal and even bloody physical
struggles. A big, strong body and use of a weapon can be the difference
between success and defeat. From the time of Darwin (1859, 1871), it has
been easy to conceive how sexual selection could lead to the evolution of traits
like armaments that aid in success in contests for mates (Figure 5.3).
A less intuitive idea regarding the outcome of intrasexual selection is that
showy traits like bright colors and melodic songs, which might otherwise be
taken as ornaments that charm females, could also function as signals of
health, vigor, and status in aggressive contests. The basic logic of this hypoth-
esis is that it is only worth engaging in a fight if there is a reasonable chance of
winning. If an assessment can be made before a fight, permitting both the
stronger and weaker contestant to perceive a mismatch, then the weaker ani-
mal benefits by retreating and the stronger animal benefits by allowing the re-
treat of the weaker (Berglund et al., 1996; West-Eberhard, 1979). But how can
the outcome of a fight be reliably predicted? What stops a weaker fighter from dis-
honestly signaling superior fighting ability? Two means to ensure the honesty of
signals of fighting ability have been proposed: social mediation and linking
trait production to individual condition (Berglund et al., 1996). I will take
up the topic of how production constraints can ensure signal honesty in the
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 151

Figure 5.3. The curved horns of a male bighorn sheep (Ovis canadensis) serve as bat-
tering rams during dramatic and violent contests for harems of females. There are
few clearer cases of secondary sexual characteristics serving as armaments.

discussion of indicator traits and intersexual selection below. Here I will

focus on social mediation.
Both signalers and receivers benefit from honest signals of status because
such honest signaling spares them both the costs of unnecessary fighting
(Rohwer, 1975). The problem with maintaining signal honesty is that low-
status males benefit even more if they can get away with dishonestly signaling
high status. The proposed solution is that signals of high status are not always
“believed”—they are routinely tested (Rohwer, 1977). So long as the costs of
drawing the aggression of dominant males outweigh the benefits of status
gained among subordinates through deception, then cheating strategies will
not succeed. Mathematical models based on game theory support the idea that
honest signaling can persist by means of such social mediation (Lachmann
et al., 2001; Ripoll et al., 2004).
In the past few decades, many studies have provided empirical support for
the idea that showy traits can be used in assessment during contests for mates
(reviewed in Berglund et al., 1996). For instance, in many species of song-
birds, males defend territories that are critical resources for reproduction. Field
152 ANIMAL BEHAVIOR

Box 5.1
Ornamental traits in territorial animals

Songbirds are among the most ornamented animals. They flash a rainbow of
colors; make some of the most complex, loud, and lovely sounds in nature;
carry long and elaborate feathers; and engage in dramatic displays. With few
exceptions, male songbirds lack an intromittent organ, so forced copulation
is typically not possible. Songbirds would seem to be a taxon in which
female mate choice has selected for a wide array of ornaments, and indeed,
in some songbirds, long tails, colorful feathers, and song are criteria in
female mate choice (Hill, 2006). But most songbirds are also highly
territorial, and the quality of the territory that a male defends has a large
effect on female reproductive success (Emlen & Oring, 1977). Perhaps
not surprisingly, in several territorial and highly ornamented songbirds,
females show no mating preference for more highly ornamented males,
but male ornamentation plays an important role in contests for territories
(e.g., eastern bluebird [Sialia sialis] [Liu et al., 2007; Siefferman & Hill,
2005]). In these songbirds, apparent ornaments are really armaments. Few
species of songbirds have been carefully studied, so at this point there is
no way to assess whether colorful feathers and complex songs most
frequently function as ornaments or armaments.

studies have shown that song and plumage coloration—two traits that appear
to be ornaments—predict male success in acquiring the most preferred territo-
ries and procuring access to mates (e.g., Pryke & Andersson, 2003; Siefferman
& Hill, 2005) (Box 5.1). Increasing the coloration of males makes potential
rivals approach them more carefully and less frequently, and decreasing
coloration causes males to be challenged more (examples in Hill, 2010).
Hence, armaments can exist either as weapons or as signals of fighting ability.

THE EVOLUTION OF ORNAMENTS

The process by which intersexual selection leads to the evolution of orna-
mental traits in animals is one of the longest-running and most contentious
debates in evolutionary biology. In the nineteenth century Charles Darwin
and Alfred Wallace were the key players in the original debate regarding the
importance of female choice in ornament evolution. Darwin argued emphati-
cally that female choice was the selective force that gave rise to ornaments
while Wallace discounted the possibility of an aesthetic sense in animals and
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 153

looked for other explanations (Cronin, 1991). Empirical studies in the late twen-
tieth and early twenty-first centuries on a range of animals from flies to primates
showed definitively that female choice can select for ornamentation (Andersson,
1994). These studies vindicated Darwin’s insistence that female choice was the
selective force that drove the evolution of ornaments, but they left unresolved
the question of why females used ornaments to choose males. Darwin provided
no insight on this fundamental question, and a resolution to how and why
female choice for ornamental traits evolves and how such choice leads to the evo-
lution of the range of ornaments seen in nature remains a major unresolved ques-
tion in behavioral and evolutionary biology. The gist of the discussion regarding
the evolution of ornamental traits focuses on whether elaboration of ornaments
occurs under selection for arbitrary markers of attractiveness or through female
demand for honest signals of the quality of perspective mates. I will consider each
of the major hypotheses for the evolution of ornaments via mate choice as well as
more specific mechanisms for ornament evolution.

Models of Intersexual Selection Based on Arbitrary Choice by Females

When evolutionary biologists refer to ornamental traits as arbitrary, they
mean that the expression of ornamental traits is not related to a male’s genetic
quality or to the quantity or quality of resources controlled by the male. Arbi-
trary traits hold value exclusively in the realm of sexual attractiveness. The
most well-known model for how sexual selection can lead to the evolution of
arbitrary traits is the Fisherian or runaway model of sexual selection (Figure
5.4). There are, however, at least three other hypotheses for how arbitrary
traits might evolve through female mate choice, and these various models of
arbitrary mate choice are not mutually exclusive. I will begin by discussing
the less prominent models of intersexual selection before focusing on the run-
away model of sexual selection.
Species recognition—The first hypothesis that was proposed for the evolution
of arbitrary traits is the species recognition hypothesis. This hypothesis was pro-
posed in the nineteenth century by Wallace, who argued that for many animals it
is vitally important both to recognize individuals of their species and to signal
their own species identity (Wallace, 1889). Wallace (1889) discussed the impor-
tance of species recognition in social contexts during flocking and other nonsex-
ual congregations, but he focused primarily on species recognition within the
context of mate choice. He concluded that mating outside species boundaries is
potentially the worst mistake that a female can make, a theme echoed half a cen-
tury later by Ronald Fisher (Fisher, 1930, 1958). Hence, Wallace proposed that
ornaments evolved to serve the very important function of unambiguously sig-
naling species identity primarily in the context of mate selection.
Figure 5.4. The elongated tail of this male paradise flycatcher (Terpsiphone paradisi) is
the sort of highly elaborated trait that is proposed to have evolved through runaway
sexual selection. Alternatively, the ornament may have evolved as a signal of male
quality.
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 155

For approximately 80 years, from the late nineteenth to the mid-twentieth

century, the species recognition hypothesis was the consensus explanation for
why ornamental traits evolved (Cronin, 1991). Since the early 1980s, the
hypothesis of ornaments as signals of species identity has fallen out of favor
in discussions of sexual selection, and many contemporary students of behav-
ioral and evolutionary biology fail to appreciate the influence and success of
this idea. For nearly a century, the species recognition hypothesis was the only
widely stated explanation for ornamental traits (Cronin, 1991; Hill, 2002).
With a surge of interest in and support for runaway and indicator models of
sexual selection (both presented below), species recognition is now given little
attention. In current discussions of sexual selection it is often not even men-
tioned. The species recognition hypothesis, however, has not been broadly fal-
sified and remains a viable hypothesis for many ornamental traits of animals.
The demise of the species recognition hypothesis came not from specific
tests or empirical observations that disproved it. The hypothesis was eventually
deemed inadequate based on theoretical considerations of the sort of traits that
should evolve as markers of species identity. Such markers, theorists
concluded, should be simple, invariant within a species, and easy to perceive.
Simple and invariant are not accurate descriptors of many ornamental traits,
particularly the elaborate ornaments that evolutionary biologists are most anx-
ious to explain. As John Maynard Smith (Maynard Smith, 1991, p. 146)
wrote, “it would be absurd to suppose that a male nightingale must sing like
that in order for a female to tell that he is not a willow warbler.” A consensus
quickly emerged in the 1980s that the species recognition hypothesis could
not explain elaborate ornaments like a peacock’s train or the roar of a red deer
(Cervus elaphus). The fact that the species recognition hypothesis is not suffi-
cient to explain highly elaborate ornamental traits, however, does not mean
that it fails as an explanation for all ornamental traits. A few empirical studies
have supported a species recognition function for simple ornaments such as
the white forehead spot on species of Ficedula flycatchers (Saetre et al., 1997).
Finally, it should be noted that the need to invoke evolutionary forces other
than species recognition to explain the evolution of a peacock’s tail does not
mean that peahens do not use the gaudy trains of males to distinguish pea-
cocks from other pheasant species. Debate regarding the species recognition
hypothesis is not about whether ornamental traits can function in species rec-
ognition; rather, it is about whether species recognition is sufficient to explain
the evolution of the traits.
Sensory exploitation—Another model for the evolution of arbitrary signals of
attractiveness is the sensory exploitation model. This hypothesis proposes
that ornaments evolve because they exploit sensory-response systems that
156 ANIMAL BEHAVIOR

typically function in contexts other than mate choice, such as foraging (Ryan
& Keddy-Hector, 1992). By this model, ornamental traits evolve because they
stimulate already existing sensory systems and trigger already existing positive
behavioral response mechanisms (Ryan & Keddy-Hector, 1992). As an exam-
ple, consider an animal for which red fruit is a choice dietary item such that
the animal is attracted to red objects in its environment. If a red patch of color
appeared on a prospective mate in such a species, females may be drawn to and
show positive sexual receptivity toward that male because they are predisposed
to respond positively to red objects (Rodd et al., 2002). The ornament simply
exploits preexisting sensory and behavioral response systems, and the positive
response by females leads to the spread of genes for the ornamental trait.
Some of the strongest empirical support for sensory exploitation as an
explanation for ornamental traits comes from studies of swordtail fish (genus
Xiphophorus). As their name implies, swordtails have a sword-like extension
from the base of their caudal fins. In all species with swords studied to date,
females prefer to mate with males with long swords (Basolo, 1990a). Sword-
tails are closely related to platyfish, which do not have swords. Through behav-
ioral studies it was discovered that, in some species of platyfish, females also
show a sexual preference for males with swords. Reconstruction of the evolu-
tionary history of the swordtails and platyfish showed that the evolution of
preference for swords predated the evolution of swords (Basolo, 1990b). The
implication was that swords had evolved in some fish lineages because there
was a preexisting bias that caused females to favor males with the trait.
Researchers found no benefit of the caudal extension other than sexual attrac-
tiveness. Studies of the calls of frogs (Ryan et al., 1990) and the colors of fish
(Egger et al., 2011) also support the sensory bias models.
The current attitude among research biologists concerning the sensory
exploitation model is that it is likely a factor in the evolution of ornamental
traits (see the runaway and chase-away models below) but that it is not suffi-
cient as a complete explanation for the evolution of most ornamental traits
of animals.
Chase-away model—Another model for the evolution of ornaments that is
related to the sensory exploitation model is the chase-away model of sexual selec-
tion. The foundation of this model is the fundamental conflict of interest
between males and females in sexually reproducing organisms. Earlier in the
chapter I described females as the “choosy sex” and males as the “displaying
sex,” but under the chase-away model, males would be better described as the
“coercive sex.” Male ornaments are proposed to stimulate females to mate in a
manner that benefits males but that is suboptimal for females—overstimulating
females to mate too often, at the wrong time, or with a low-quality partner
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 157

Figure 5.5. Flow charts illustrating the processes proposed by the three models for the
elaboration of ornamental traits that can account for elaboration of ornamental traits:
runaway model (Fisher, 1958), chase-away model (Holland & Rice, 1998) and
ratchet model (Hill, 1994). Under the runaway model, trait elaboration is inevitable
once the process ensues, but under the chase-away and ratchet models, elaboration is
dependent on the appearance of genetic variants among females that resist attraction
to the ornament or on the appearance of shortcut means to trait expression by males,
respectively. Evolutionary events are indicated in the boxed portion of the figure, and
forces of change are shown by arrows connecting boxes.

(Holland & Rice, 1998). The stimulatory male trait spreads because it enhances
male mating success, but the chase-away model proposes that evolutionary esca-
lation ensues. Because being attracted to the ornament is bad for females, they are
selected to evolve diminished response systems, thereby resisting the coercion of
males. As female responsiveness abates, the advantage to males of having the trait
declines such that there is selection on males for any new traits that enhance or
expand the stimulatory effect of the ornament. Escalation ensues as males are
selected to display more and more elaborate ornamentation to maintain stimula-
tion of females in the face of eroding responsiveness (Figure 5.5).
158 ANIMAL BEHAVIOR

The chase-away model makes two specific predictions that are different
from other models of sexual selection (Holland & Rice, 1998). First, the
model proposes declines in female responsiveness to ornamental traits over
evolutionary time. Change in female mating preference on an evolutionary
time scale is a challenging prediction to test because it requires knowledge of
female responsiveness over the evolutionary history of the trait. Evolutionary
biologists cannot go back in time to assess female responsiveness (behavior
does not typically fossilize), but they can use phylogenetic reconstructions
for estimating the past condition of traits, including behaviors (see Chapter
2). A study involving phylogenetic reconstructions of the responsiveness of
female wolf spiders (genus Schizocosa) to tufts of bristles on males supported
predictions of the chase-away model. In species of spiders without tufts, add-
ing tufts to males made them much more sexually attractive to female spiders.
In species that had tufts, however, removing or enlarging tufts had no effect on
female sexual response. Critically, phylogenetic reconstruction showed that
tufts were a derived character and that the species without tufts never had a
tufted ancestor (McClintock & Uetz ,1996). One interpretation of these
results is that in the lineage with tufts, a chase-away process had selected for
reduced response to tufts in females (Holland & Rice, 1998; see also Box 5.2).
A second prediction of the chase-away model is that females should have
reduced rather than enhanced fitness if they mate with highly ornamented
males (Holland & Rice, 1998). The strongest empirical support for this pre-
diction of reduced female fitness comes not from studies of selection for orna-
mental traits but rather from studies of mating costs among female fruit flies in
experimentally maintained monogamous versus polygynous populations
(Rice, 1996; see Chapter 6). In lines of flies that were forced to be monoga-
mous such that the reproductive output of a male was directly tied to its single
mate, females had long lives. In lines of flies in which males competed with
other males for sexual access to multiple mates, females had shortened lives
and lower fecundity. The explanation was that when multiple mating was pos-
sible, males added toxic elements to semen to benefit themselves in competi-
tion with other males even though it was costly to females (Chapman et al.,
1995). When sexual selection was eliminated, the toxic elements in semen
declined.
Costs to females of choosing highly ornamented males have proven harder
to document. In the runaway model of sexual selection as well as the indicator
models described below, females benefit by mating with highly ornamented
males. It is universally accepted that mate choice and sexual reproduction
impose a cost on females (Andersson, 1994), but a loss of female fitness result-
ing specifically from a response to male ornamentation is a difficult prediction
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 159

Box 5.2
The paradox of rooster plumes

The red jungle fowl (Gallus gallus), the wild ancestor of the domestic
chicken, is one of the most highly ornamented birds in the world. Hens
are brown and drab while roosters have long, elaborate, and colorful feathers
covering their bodies. If asked to explain these plumes, most evolutionary
biologists would respond that female mate choice or perhaps competition
among males drove the evolution of the ornaments and now maintains
them. There would be wide consensus that the showy rooster feathers are
the result of sexual selection.
Surprisingly, experimental observations do not support either female
mate choice or male-male competition as viable explanations for the plumes
of roosters. The key experiment involved presenting hens with prospective
mates that either had normal ornamental feathers or lacked them entirely.
This experiment was possible because there is a gene, called the hen-
feathered gene, that leaves males bereft of all the ornamental plumes.
In every other respect, however, hen-feathered males are normal roosters.
They have the same fleshy head ornaments as normal males, and they retain
their male aggressiveness. J. David Ligon and Patrick Zwartjes (1995) bred
red jungle fowl so as to produce brothers that either carried or did not carry
the hen-feather gene; in other words, they created brothers that had the
feathers of hens or that had full ornamental plumes. These sets of brothers
were then used in mate-choice experiments. Amazingly, hens showed no
preference for males with ornate and colorful plumes—removing all of the
long golden and glossy ornamental plumage did not make males any less
attractive as mates. In another study, Ligon and colleagues (1990) showed
that the plumes have little or no effect on male-male contests. Among pro-
posed sexual selection hypotheses, only the chase-away model can explain
why chicken ornaments exist when there is no inter- or intrasexual selection
on the trait. According to the chase-away model, the ornamental plumes
arose because they coerced females into more mating, and females have
escaped male manipulation by evolving lack of preference for the trait.

to test. To date, no specific study as been presented that clearly demonstrates a

loss of fitness for females because they were attracted to a male ornament. On
the other hand, many studies have proposed to show a net benefit to females
for mating with highly ornamented males (Andersson, 1994; Griffith &
Pryke, 2006).
160 ANIMAL BEHAVIOR

Overall, the chase-away model may be the best current explanation for why
some ornamental traits like the gaudy plumages of red jungle fowl (Gallus gal-
lus) (Box 5.2) have no influence on the mate choices made by females. The
importance of the chase-away model as a general explanation for ornamental
traits in animals remains unknown.
Runaway sexual selection—The idea that ornamental traits could evolve
through a runaway process was first presented in a brief narrative by Fisher
(1915) on which he later elaborated (Fisher 1930, 1958). Fisher proposed that
female choice and male ornamentation could enter into a self-reinforcing pro-
cess in which both the ornamental trait and preference for the ornamental trait
could become grossly exaggerated. The concept of runaway sexual selection
was subsequently formalized in quantitative genetics models, most notably
by Russell Lande (1981) and Mark Kirkpatrick (1982), and the runaway
model (also known as the Fisherian model and the Lande-Kirkpatrick model)
remains one of the most important explanations for the evolution of ornamen-
tal traits—particularly fantastic and highly exaggerated traits (Prum, 1997).
The process of runaway sexual selection is proposed to begin with a popula-
tion of animals in which some males carry genes for a trait that bestows a fitness ad-
vantage such as a slightly longer tail that enhances flight ability. (See Arnold [1983]
for a lucid and well-illustrated presentation of the runaway model). Some females
happen to carry genes that cause them to prefer longer tails—such a preexisting
genetic bias invokes sensory exploitation (see above)—and these females benefit
because their offspring inherit the genes for longer and functionally better tails.
Both longer male tails and female preference for longer tails are beneficial, and
hence genes for both greater preference and longer tails spread in the population.
As an increasing number of females inherit genes for a preference for longer tails,
two important consequences initiate the runaway process. First, the benefits for
producing sons that are attractive as mates (“sexy sons”) begins to exceed the bene-
fits of producing sons with aerodynamic tails; in other words, the benefit of the trait
is no longer determined primarily by natural selection. Moreover, the sexy-son
benefit increases with each generation as preference for the trait increases. The
stronger the female preference for a large ornament, the greater the selection will
be on males to grow even longer tails. In turn, the longer the tails grown by males,
the greater are the benefits to females in the form of sons with long and sexy tails.
A second consequence that spurs on a runaway cycle is that the genes for
longer tails and the genes for preference for longer tails begin to be inherited
together (a process called linkage disequilibrium) (O’Donald, 1962). The
offspring of a female that has chosen to mate with a long-tailed male passes
to her offspring both the genes for longer tails and the genes for preference
for longer tails. As a consequence, when a long-tailed male has mating success,
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 161

it perpetuates both the genes for longer tails and genes for long-tail preference.
The results of ever-increasing benefits for both longer tails and preferences for
longer tails are that the trait can be elaborated to extreme endpoints and that
the direction that the elaboration takes—longer, wider, colored, curled,
forked—is constrained only by the esoteric preferences of females and by what
variations on the trait arise in the population due to mutation. Eventually, the
fitness (natural selection) costs in the form of reduced survival and fecundity
from having a burdensome ornament outweigh the benefits of sexy sons and
the runaway process is halted.
For about a 10-year period from the early 1980s to the early 1990s, theoretical
biologists attempted to test the validity of the runaway model in quantitative
genetics models that made various assumptions about the mode of inheritance
of both the ornamental trait and the preference behavior, about how rare the trait
and preference were at the start of the process, and about the costs that might be
associated with mate choice (reviewed in Andersson, 1994; Møller, 1994). These
models showed that, theoretically, the runaway sexual selection process could lead
to many evolutionary outcomes including stable equilibria (Figure 5.6).
Despite many attempts over the past four decades to test its validity, the
runaway model of sexual selection has proven difficult to test empirically
because, like the chase-away model above, true tests require assessments
during the process of trait elaboration. Assessment of the end product of a pur-
ported runaway process does not allow discrimination from other possible evo-
lutionary models. From the late 1970s into the early 1990s, the runaway model
of sexual selection was the model of sexual selection favored by theoreticians, and
it was widely presented as a definitive explanation for the evolution of ornamen-
tal traits (Andersson 1986, p. 253). Through the 1990s and into the new millen-
nium, however, interest shifted to models of adaptive mate choice (see below).
With little success at testing the runaway model of sexual selection, interest in
the process declined, and in current evolution and behavior literature, relatively
little attention is paid to the runaway model of sexual selection. Nevertheless,
the runaway model of sexual selection remains the most likely model to account
for extreme forms of ornamental traits (Prum, 2010).

Models of Sexual Selection Based on Adaptive Female Choice for Ornaments

When evolutionary and behavioral biologists discuss sexual selection mod-
els based on adaptive mate choice, they are discussing hypotheses under
which ornament expression is associated with characteristics of perspective
mates that are beneficial outside of a mate-selection context. By choosing
highly ornamented males via adaptive mate choice, females receive benefits
162 ANIMAL BEHAVIOR

Figure 5.6. Two possible scenarios for runaway sexual selection according to a model
created by Lande (1981) and illustrated by Arnold (1983). Plotted on both graphs
are the mean ornament expression by males in a population (x axis) and mean orna-
ment expression preferred by females (y axis). The vertical line near the y axis indicates
the trait expression that is optimum under natural selection. The bold line slanting
through the figure is the line of equilibrium for expression and preference, which is a
function of the intensity of natural versus sexual selection. Once a point on this line of
equilibrium is reached, male expression and female preference can remain in a stable
state indefinitely. The directional lines running toward or away from the line of
equilibrium are evolutionary trajectories predicted for expression and preference
depending on the starting, nonequilibrium conditions. The trajectory of the evolving
population is a function of the genetic covariance between female trait preference and
male trait expression (see Arnold, 1983) (a) When the genetic covariance is relatively
weak, then ornamentation reaches a stable expression with moderate elaboration.
(b) When the genetic covariance exceeds a certain magnitude, then ornamentation
evolves away from the line of equilibrium toward extreme elaboration; in other words,
a runaway process ensues. See Arnold (1983) for a detailed description.

in the form of more resources for themselves or their offspring or genes that
enhance the fitness of offspring. Females might also avoid costs associated with
mating, such as sexually transmitted disease. Models of adaptive mate choice
are collectively called indicator models of sexual selection because the orna-
mental trait is proposed to indicate important qualities of a perspective mate.

MAINTAINING SIGNAL HONESTY

A central topic related to adaptive mate choice is signal honesty. Indicator
models of sexual selection propose that larger, more elaborate, and showier
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 163

ornaments are associated with high-quality males. The obvious questions

related to such a concept is what stops cheating—what prevents average or low-
quality males from sporting big ornaments? To address the question of signal
honesty, in 1975 Amotz Zahavi proposed the handicap principle (Zahavi,
1975). Zahavi hypothesized that ornamental traits served as handicaps to sur-
vival such that only robust, strong, and healthy individuals could bear the bur-
den of a big ornament. Viewing ornaments as handicaps solved the problem of
signal honesty—there is no way to dishonestly bear the burden of a handicap-
ping ornament (Zahavi, 1977). Unfortunately, the handicap model created a
new theoretical problem even as it solved the first. Under the handicap model,
females who chose males with a big ornament received the benefit of pairing
with robust and fit males, but they suffered the cost of passing the handicap
to their offspring. Theoretical models indicated that such handicapping traits
would not evolve (Maynard Smith, 1976; Bell, 1978; see Pomiankowski &
Iwasa, 1998, for an overview) even if they were associated with male quality.
In the decades since the first pioneering papers of Zahavi, the idea of costs
associated with ornamentation has become a key concept in behavioral ecol-
ogy. From the handicap model arose the condition-dependent model of orna-
ments (Kodric-Brown & Brown, 1984). This new thinking shifted the burden
of ornamentation away from survival costs onto production costs. Ornaments
were hypothesized to be honest signals of individual condition because there
was no way for an individual to meet the challenges of trait production with-
out being in good condition (reviewed in Andersson, 1994; Møller, 1994).
Once produced, ornaments need pose no survival handicap. Condition has
typically been defined as the pool of resources available to the organism such
that resources can be allocated either to ornament production or to body
maintenance (Rowe & Houle, 1996). Only individuals with large resource
pools—and hence high condition—could produce large ornaments. Math-
ematical models indicated that this form of honest signaling could evolve
(Grafen, 1990).
In discussions of allocation trade-offs, energy is generally presented as the
limiting resource that must be parsed between ornamentation and body main-
tenance (Wedekind & Folstad, 1994), but allocation trade-offs could involve
molecules such as carotenoid pigments (Figure 5.7) (Lozano, 1994), dietary
alkaloid molecules (Iyengar et al., 2001), or any other element needed for both
body maintenance and ornamentation. Defining signal honesty in terms of
allocation trade-offs, however, discounts a role for genetic contributions to
condition. Moreover, several empirical studies have shown that even when
resources were not limiting, individuals varied in ornament expression
(Bortolotti et al., 1996). To accommodate these problems, condition was
164 ANIMAL BEHAVIOR

Figure 5.7. The bright yellow coloration of the fleshy ornaments of this wattled lap-
wing (Vanellus senegallus) result from carotenoid pigments. Such carotenoid colora-
tion may reveal the functionality of vital cellular processes in the displaying male.

defined as the efficiency of vital cellular processes (Hill, 2011), allowing for
phenotype and genotype (as well as epigenetic state) to affect individual condi-
tion. With ornamentation linked to challenges of trait production, cheating is
not possible—only individuals with well-functioning cellular systems can pro-
duce elaborate ornaments (Hill, 2011).
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 165

THE LEK PARADOX

Whether expression of an ornament is hypothesized to be associated with
resource benefits or to be a signal of good genes is extremely important to
models of sexual selection. Indicator models founded on direct resource bene-
fits are relatively straightforward: if males with larger ornaments provide more
resources, then females who choose such males directly benefit (Hoelzer,
1989). By itself, this model, sometimes called the good parent model, actually
provides no mechanism for the elaboration of ornaments (discussed below),
but it can explain how such traits function and are maintained. Because males
provide no parental care and no resources to females in most animal species,
however, resource-based models cannot be universal explanations for orna-
mental traits. In particular, in many animals with highly conspicuous and
elaborate ornaments—which are the traits that the theory of sexual selection
was originally developed to explain—males provide nothing to females other
than the contents of their ejaculate. In such species, the only explanation
for ornamentation founded on adaptive mate choice is that females receive
advantageous alleles for their offspring by choosing to mate with highly orna-
mented males. This good genes hypothesis requires that males vary in genetic
quality and that ornamentation is an honest indicator of these good genes
(Williams, 1966).
Ornamentation as a signal of good genes has been an intriguing idea since it
was first proposed in the mid-twentieth century. Unfortunately, it is theoreti-
cally impossible for a population to maintain standing genetic variation associ-
ated with fitness. According to a basic tenet of evolutionary theory, called
Fisher’s fundamental theorem, at equilibrium the heritability of fitness will
be zero (Fisher, 1930). Stated another way, any alleles that improve fitness will
increase in frequency, reaching fixation in just a few generations. For decades,
theoreticians have been left with a quandary—the only benefit for female mate
choice when males provide no resources is good genes, but good genes are
theoretically impossible. This situation is termed the lek paradox (Kirkpatrick
& Ryan, 1991), and it continues to confound good genes explanations of
ornament evolution.
The lek paradox remains unresolved but is a rich area of theoretical research
in evolutionary biology. One way around the problem is to abandon the
concept of adaptive mate choice in cases where males provide no resources—
arbitrary mate choice poses no paradox. Alternatively, theoreticians have
reconsidered the assumption that most populations are at equilibrium. It has
been proposed that mutation can introduce sufficient genetic variation to
maintain selection for ornaments that signal good genes (Rice, 1988), but this
hypothesis has not been widely supported. As a resolution to the lek paradox,
166 ANIMAL BEHAVIOR

most theoreticians propose that environments are inherently unstable so that

equilibrium is never reached (Charlesworth, 1987). In particular, some evolu-
tionary biologists noted that essentially all animals exist in a perpetual evolu-
tionary arms race with their parasites. Hosts constantly evolve novel
mechanisms to suppress or defeat parasites, and parasites evolve mechanisms
to evade host resistance. Hamilton and Zuk (1982) proposed that genes for
parasite resistance will cycle in frequency in response to changes in parasite
abundance, which provides perpetual variation in genotype quality, that is,
perpetual good genes to be associated with ornamentation (Hamilton &
Zuk, 1982). Many studies have shown that parasites affect the expression of
ornamental traits, but to date, no study has linked specific genes for parasite
resistance to expression of an ornamental trait.
The genic capture model (Rowe & Houle, 1996), which is the most widely
cited resolution to the lek paradox, builds from the observation that in most
organisms there is a large amount of genetic variation associated with individ-
ual condition and proposes that when ornaments are elaborated, for example
through a runaway sexual selection process, they inevitably become associated
with condition (i.e., the ornaments become condition dependent). Hence,
ornaments become associated with good genes because as they become chal-
lenging to produce they become linked to condition, which has substantial
genetic variation (Rowe & Houle, 1996). Ornamental traits essentially cap-
ture the genetic variation that is an inherent part of individual condition.
The genic capture model relies on other models for trait elaboration and only
seeks to explain how ornamentation can be associated with good genes.
To date, despite great interest in the idea that ornamental traits might sig-
nal the genetic quality of males, no definitive tests of this hypothesis have been
conducted. With recent breakthroughs in understanding the genetic architec-
ture of animals, true definitive tests of the good gene hypothesis should soon
be possible (Chenoweth & McGuigan, 2010).

A MECHANISM FOR EVOLUTION VIA ADAPTIVE MATE CHOICE

Sexual selection hypotheses that invoke adaptive female choice are generally
explanations for the function of ornamental traits or for the evolution of an
ornament to a rudimentary level—the degree of elaboration that can be
achieved in a single mutational step. Most indicator models provide no mech-
anisms for the elaboration of ornamental traits from rudimentary states to
large and complex structures. All models of ornament elaboration proposed
to date rely on a feedback mechanism through which the ornament and pref-
erence for the ornament coevolve (Figure 5.5). Such feedback-based
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 167

mechanisms are proposed by two models of arbitrary mate choice: the run-
away model and the chase-away model. The only model for trait elaboration
via adaptive mate choice that invokes a feedback mechanism was not given a
name when it was proposed (Hill, 1994), but I will refer to it as the ratchet
model because ornament elaboration is proposed to advance in a ratchet-like
manner (Figure 5.5).
The ratchet model begins with a simple trait that is an honest signal of indi-
vidual quality. Females benefit if they show a preference for the trait because
they receive either resources or genetic benefits. Potentially, such an ornament
can persist indefinitely as a simple indicator of quality, but there will be selec-
tion on males to evolve mechanisms to display the trait independent of their
quality. If novel genes arise for such shortcut means to cheaply produce the
ornament they will spread rapidly, and the information content of the trait
will diminish. Female preference will no longer be beneficial to females. In
response to dishonest signaling by males, there will be selection on females
to shift preferences to ornamentation that restores the association with male
quality—for instance, females will be selected to respond to larger, louder, or
more complex ornaments. The coevolutionary cycle that ensues—with males
evolving means to produce ornaments without invoking costs and with
females selecting for more elaborate and costly forms of ornaments—can
result in large, costly, and elaborate traits (Figure 5.5).
The ratchet model is similar to the chase-away model (Holland & Rice,
1998) in that both models focus on a conflict of interest between males and
females as the impetus that drives coevolutionary cycles through which orna-
ments are elaborated (Figure 5.5). In the chase-away model, females are
trapped by stimulus-response systems that cause them to make maladaptive
choices when mating. The result is ornamentation that is effective at coercing
females but that reveals nothing useful about a male. In the ratchet model, in
contrast, trait elaboration occurs in response to the demand by females for
honest signaling in males. Specific contrasting tests of these two models have
not been undertaken, but the fundamental predictions of these models—a loss
of fitness through choice for male ornamentation under the chase-away and a
gain in fitness through such choice in the ratchet model—should make the
hypotheses distinguishable.

INTRASEXUAL SELECTION AS THE FOUNDATION OF INTERSEXUAL SELECTION

Anders Berglund and colleagues (Berglund et al., 1996) proposed that
ornamental traits commonly evolve as signals of status and fighting ability
either through sexual or social selection. Once ornamental traits exist as signals
168 ANIMAL BEHAVIOR

of status linked to individual condition, they can become the target of female
mate choice and begin to function in intersexual selection. As with many other
models of sexual selection, this idea can explain the current function of orna-
mental traits but it does not provide a mechanism for trait elaboration.

SUMMARY
The armaments and ornaments have fascinated biologists since they captured
the attention of the original evolutionist, Charles Darwin. It is widely accepted
that armaments are the product of aggressive contests for mates, and this hypoth-
esis has not been modified significantly since first proposed by Darwin. The evo-
lution of ornamental traits, however, remains more controversial. After a century
and a half of work on the topic, a rich and diverse set of theories has been devel-
oped to explain how ornamental traits function and how they evolved. The origi-
nal hypothesis for the elaboration of traits via sexual selection—the runaway
model—remains the most likely explanation for many of the most gaudy and
dramatic ornaments. The idea that began as Zahavi’s handicap principle has been
modified into a more general indicator hypothesis that ornamental traits are
linked to the health and vitality of individuals. This indicator model has gained
wide support through empirical studies that find links between measures of indi-
vidual quality and ornamentation. The specific idea from Zahavi that ornaments
are handicaps to survival has been generally disproven, and ornaments are now
viewed more typically as indicators of condition whose honesty is maintained
by challenges posed by trait production. Despite the success of the indicator
model in explaining the present function of many ornamental traits, and despite
100 years of thinking about the process, we still lack a well-supported mechanism
for how grand and elaborate ornaments evolve.

ACKNOWLEDGMENTS
Over the past two decades the Department of Biological Sciences and
College of Science and Mathematics at Auburn University have provided time
and resources for me to pursue my studies of sexual selection.

REFERENCES AND SUGGESTED READING

Andersson, M. (1986). Sexual selection and the importance of viability differences:
A reply. Journal of Theoretical Biology, 120, 251–254.
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 169

Andersson, M. (1994). Sexual Selection. Princeton, NJ: Princeton University Press.

Arnold, S. J. (1983). Sexual selection: The interface of theory and empiricism. In
P. P. G. Bateson (ed.), Mate Choice (pp. 67–107). Cambridge, UK:
Cambridge University Press.
Basolo, A. L. (1990a). Female preference for male sword length in the green swordtail
(Xiphophorus helleri). Animal Behaviour, 40, 332–338.
Basolo, A. L. (1990b). Female preference predates the evolution of the sword in
swordtail fish. Science, 250, 808–810.
Bell, G. (1978). The handicap principle in sexual selection. Evolution, 32, 872–885.
Berglund, A., A. Bisazza, & A. Pilastro. (1996). Armaments and ornaments: An evo-
lutionary explanation of traits of dual utility. Biological Journal of the Linnean
Society, 58, 385–399.
Bortolotti, G., J. J. Negro, J. L. Tella, T. A. Marchant, & D. M. Bird. (1996). Sexual
dichromatism in birds independent of diet, parasites and androgens. Proceed-
ings of the Royal Society of London, B, 263, 1171–1176.
Chapman, T., L. F. Liddle, J. M. Kalb, M. F. Wolfner, & L. Partridge. (1995). Cost
of mating in Drosophila melanogaster females is mediated by male accessory-
gland products. Nature, 373, 241–244.
Charlesworth, B. (1987). The heritability of fitness. In J. W. Bradbury &
M. B. Andersson (eds.), Sexual Selection: Testing the Alternatives (pp. 21–40).
London: John Wiley and Sons.
Chenoweth, S. F. & K. McGuigan. (2010). The genetic basis of sexually selected
variation. Annual Review of Ecology, Evolution, and Systematics, 81–101.
Cronin, H. (1991). The Ant and the Peacock. Cambridge, UK: Cambridge University
Press.
Darwin, C. (1859). On the Origin of Species by Natural Selection or the Preservation of
Favoured Races in the Struggle for Life. London: John Murray.
Darwin, C. (1871). The Descent of Man and Selection in Relation to Sex. London:
Murray.
Davison, G. W. H. (1985). Avian spurs. Journal of Zoology, 206, 353–366.
Egger, B., Y. Klaefiger, A. Theis, & W. Salzburger. (2011). A sensory bias has trig-
gered the evolution of egg-spots in cichlid fishes. PLoS ONE, 6, e25601.
doi:10.1371/journal.pone.0025601.
Emlen, S. T. & L. W. Oring. (1977). Ecology, sexual selection, and the evolution of
mating systems. Science, 197, 215–223.
Fisher, R. A. (1915). The evolution of sexual preference. Eugenics Review, 7,
184–192.
Fisher, R. A. (1930). The Genetical Theory of Natural Selection. Oxford, UK:
Clarendon Press.
Fisher, R. A. (1958). The Genetical Theory of Natural Selection. New York: Dover.
Grafen, A. (1990). Sexual selection unhandicapped by the Fisher process. Journal
of Theoretical Biology, 144, 473–516.
170 ANIMAL BEHAVIOR

Griffith, S. C. & S. R. Pryke. (2006). Benefits to females of assessing color displays.

In G. E. Hill & K. J. McGraw (eds.), Bird Coloration, Vol. 2: Function and
Evolution. Cambridge, MA: Harvard University Press.
Hamilton, W. D. & M. Zuk. (1982). Heritable true fitness and bright birds: a role for
parasites? Science, 218, 384–386.
Hill, G. E. (1994). Trait elaboration via adaptive mate choice: Sexual conflict in
the evolution of signals of male quality. Ethology, Ecology and Evolution, 6,
351–370.
Hill, G. E. (2002). A Red Bird in a Brown Bag: The Function and Evolution of
Ornamental Plumage Coloration in the House Finch. New York: Oxford
University Press.
Hill, G. E. (2006). Female choice for ornamental coloration. In G. E. Hill &
K. J. McGraw (eds.), Bird Coloration, Vol. 2: Function and Evolution.
Cambridge, MA: Harvard University Press.
Hill, G. E. (2010). National Goegraphic Bird Coloration. Washington, DC: National
Geographic Society.
Hill, G. E. (2011). Condition-dependent traits as signals of the functionality of vital
cellular processes. Ecology Letters, 14, 625–634.
Hill, G. E., R. Montgomerie, C. Roeder, & P. Boag. (1994). Sexual selection and
cuckoldry in a monogamous songbird: Implications for sexual selection theory.
Behavioral Ecology and Sociobiology, 35, 193–199.
Hoelzer, G. A. (1989). The good parent process of sexual selection. Animal Behaviour,
38, 1067–1078.
Holland, B. & W. R. Rice. (1998). Perspective: Chase-away sexual selection:
Antagonistic seduction versus resistance. Evolution, 52, 1–7.
Iyengar, V. K., C. Rossini, & T. Eisner. (2001). Precopulatory assessment of male
quality in an arctiid moth (Utetheisa ornatrix): Hydroxydanaidal is the only
criterion of choice. Behavioral Ecology and Sociobiology, 49, 283–288.
Kirkpatrick, M. (1982). Sexual selection and the evolution of female choice.
Evolution, 36, 1–12.
Kirkpatrick, M. & M. J. Ryan. (1991). The evolution of mating preferences and the
paradox of the lek. Nature, 350, 33–38.
Kodric-Brown, A. & J. H. Brown. (1984). Truth in advertising: the kinds of traits
favored by sexual selection. American Naturalist, 124, 309–323.
Lachmann, M., S. Szamado, & C. T. Bergstrom. (2001). Cost and conflict in animal
signals and human language. Proceedings of the National Academy of Sciences,
USA, 98, 13189–13194.
Lande, R. (1981). Models of speciation by sexual selection on polygenic traits.
Proceedings of the National Academy of Sciences, USA, 78, 3721–3725.
Ligon, D. J. (1999). The Evolution of Avian Breeding Systems. Oxford, UK: Oxford
University Press.
Ligon, J. D., R. Thornhill, M. Zuk, & K. Johnson. (1990). Male-male competition,
ornamentation, and the role of testosterone in sexual selection in red jungle
fowl. Animal Behaviour, 40, 367–373.
THE EVOLUTION OF ORNAMENTS AND ARMAMENTS 171

Ligon, J. D. & P. W. Zwartjes. (1995). Ornate plumage of male red junglefowl does
not influence mate choice by females. Animal Behaviour, 49, 117–125.
Liu, M., L. Siefferman, & G. E. Hill. (2007). An experimental test of female choice
relative to male structural coloration in eastern bluebirds. Behavioral Ecology
and Sociobiology, 61, 623–630.
Lozano, G. A. (1994). Carotenoids, parasites, and sexual selection. Oikos, 70,
309–311.
Maynard Smith, J. (1976). Sexual selection and the handicap principle. Journal of
Theoretical Biology, 57, 239–242.
Maynard Smith, J. (1991). Theories of sexual selection. Trends in Ecology and Evolu-
tion, 6, 146–151.
McClintock, W. J. & G. W. Uetz. (1996). Female choice and pre-existing bias: Vis-
ual cues during courtship in two Schizocosa wolf spiders (Araneae: Lycosidae).
Animal Behaviour, 52, 167–181.
Møller, A. P. (1994). Sexual Selection and the Barn Swallow. Oxford, UK: Oxford
University Press.
O’Donald, P. (1962). The theory of sexual selection. Heredity, 17, 541–552.
Petrie, M., T. Halliday, & C. Sanders. (1991). Peahens prefer peacocks with elaborate
trains. Animal Behaviour, 41, 323–331.
Pomiankowski, A. & Y. Iwasa. (1998). Handicap signaling: Loud and true? Evolution,
52, 928–932.
Prum, R. O. (1997). Phylogenetic tests of alternative intersexual selection mecha-
nisms: Trait macroevolution in a polygynous clade (Aves: Pipridae). American
Naturalist, 149, 668–692.
Prum, R. O. (2010). The Lande-Kirkpatrick mechanism is the null model of evolu-
tion by intersexual selection: Implications for meaning, honesty, and design
in intersexual signals. Evolution, 64, 3085–3100.
Pryke, S. R. & S. Andersson. (2003). Carotenoid-based status signalling in red-
shouldered widowbirds (Euplectes axillaris): Epaulet size and redness affect captive
and territorial competition. Behavioral Ecology and Sociobiology, 53, 393–401.
Rice, W. R. (1988). Heritable variation in fitness as a prerequisite for adaptive female
choice—the effect of mutation-selection balance. Evolution, 42, 817–820.
Rice, W. R. (1996). Sexually antagonistic male adaptation triggered by experimental
arrest of female evolution. Nature, 381, 232–234.
Ripoll, J., J. Saldana, & J. C. Senar. (2004). Evolutionarily stable transition rates in a
stage-structured model: An application to the analysis of size distributions of
badges of social status. Mathematical Biosciences, 190, 145–181.
Rodd, F. H., K. A. Hughes, G. F. Grether, & C. T. Baril. (2002). A possible
non-sexual origin of mate preference: Are male guppies mimicking fruit?
Proceedings of the Royal Society of London, B, 269, 475–481.
Rohwer, S. (1975). The social significance of avian winter plumage variability.
Evolution, 29, 593–610.
Rohwer, S. (1977). Status signaling in Harris’ sparrows: Some experiments in
deception. Behaviour , 61, 107–128.
172 ANIMAL BEHAVIOR

Rowe, L. & D. Houle. (1996). The lek paradox and the capture of genetic variance by
condition dependent traits. Proceedings of the Royal Society of London, B, 263,
1415–1421.
Ryan, M. J., J. H. Fox, W. Wilczynski, & A. S. Rand. (1990). Sexual selection for
sensory exploitation in the frog Physalaemus pustulosus. Nature, 343, 66–67.
Ryan, M. J. & A. Keddy-Hector. (1992). Directional patterns of female mate choice
and the role of sensory biases. American Naturalist, 139, S4–S35.
Saetre, G. P., T. Moum, S. Bures, M. Kral, M. Adamjan, & J. Moreno. (1997).
A sexually selected character displacement reinforces premating isolation.
Nature, 387, 589–592.
Siefferman, L. & G. E. Hill. (2005). UV-blue structural coloration and competition
for nestboxes in male eastern bluebirds. Animal Behaviour, 69, 67–72.
Skogsmyr, I. & A. Lankinen.(2002). Sexual selection: An evolutionary force in plants.
Biological Reviews, 77, 537–562.
Stutchbury, B. J. & E. S. Morton. (1995). The effect of breeding synchrony on extra-
pair mating systems in songbirds. Behaviour 132, 675–690.
Trivers, R. L. (1972). Parental investment and sexual selection. In B. Campbell (ed.),
Sexual Selection and the Descent of Man, 1871–1971 (pp. 136–179). Chicago:
Aldine.
von Schantz, T., G. Goransson, G. Andersson, I. Froberg, M. Grahn, A. Helgee, et al.
(1989). Female choice selects for a viability-based trait in pheasants. Nature,
337, 166–169.
Wallace, A. R. (1889). Darwinism. London: Macmillian.
Wedekind, C. & I. Folstad. (1994). Adaptive or nonadaptive immunosuppression by
sex hormones? American Naturalist, 143, 936–938.
West-Eberhard, M. J. (1979). Sexual selection, social competition, and evolution.
Proceedings of the American Philosophical Society, 123, 222–234.
Westneat, D. F. & I. R. K. Stewart. (2003). Extra-pair paternity in birds: Causes, cor-
relates, and conflict. Annual Review of Ecology and Evolution, 34, 365–396.
Williams, G. C. (1966). Natural selection, the costs of reproduction, and a refinement
of Lack’s principle. American Naturalist, 100, 687–690.
Zahavi, A. (1975). Mate selection—a selection for a handicap. Journal of Theoretical
Biology. 53, 205–214.
Zahavi, A. (1977). The cost of honesty (further remarks on the handicap principle).
Journal of Theoretical Biology, 67, 603–605.
 6

Sexual Conflict: All’s Not Fair

in Love—It’s War!
Zenobia Lewis

INTRODUCTION
The male lion (Panthera leo) is feared by all. But it is the female lion that
should perhaps be most cautious. The penis of the male lion is covered in tiny
barbs, made of the hard protein keratin, usually found in claws and nails.
These barbs tear the lining of the uterus during copulation, which is painful
for the female, but which allows the male to remove any previously insemi-
nated rival males’ sperm and thereby minimize sperm competition. Stimula-
tion by the barbs also sends signals to the female’s brain, which help induce
the release of eggs. Despite the unpleasant nature of copulation in this species,
during their short fertile period each month female lions will mate up to
100 times a day. It is not only female lions that are subjected to this somewhat
harsh treatment; across diverse animal species males exhibit spiky, barbed
penises and other adaptations that result in the males harming their female
mates over or during reproduction.
Until the 1970s reproduction was viewed as a venture between males and
females, a view derived from overestimation of the occurrence of lifelong
monogamy (see Table 6.1 for definitions of the types of mating systems found
in the animal world) in the wild and maintained by outdated Victorian values
with regards to sex (Birkhead, 2010). We now know differently—males and
females often have conflicting interests when it comes to reproduction, and
174 ANIMAL BEHAVIOR

Table 6.1. Types of animal mating systems.

Type Description

Monogamy Male and female within a pair mate only once.

Monandry Female typically mates with only one male.
Polyandry Female typically mates with multiple males.
Monogyny Male typically mates with only one female.
Polynygy Male typically mates with multiple females.
Promiscuity Both males and females mate with multiple partners.

these conflicts of interest can have profound evolutionary effects on both

sexes. It was Geoff Parker (1979, p. 124) who first coined the term sexual
conflict, defining it as the “the conflict in evolutionary interests between males
and females.” He noted that sexual conflict is rooted in the differential invest-
ment in reproduction by the two sexes, as described in Chapter 5. Today, two
forms of sexual conflict are recognized: intralocus sexual conflict is a form of
conflict occurring when the same genes are expressed in both males and
females but are selected in opposite directions. A classic example is that of
the human hip; it has been suggested that in our evolutionary past, there was
strong selection for wide hips in women, due to the risk of injury during child-
birth. However, opposing selection in males favoured smaller hips due to
improved locomotor function. Thus, there may have been an evolutionary
tug-of-war between males and females over the optimal hip width (Rice &
Chippindale, 2001). Intralocus sexual conflict is thought to be a powerful
and pervasive evolutionary process, yet to date surprisingly few concrete
empirical examples exist (Bonduriansky & Chenoweth, 2009; Lewis et al.,
2011). In this chapter we will focus on the second form of sexual conflict,
interlocus sexual conflict. This is where the conflict between males and
females is mediated by different genes and traits.
In many cases, sexual conflict occurs because both males and females in
the majority of species mate with more than one individual. If there is
true monogamy between a male and female, the reproductive interests of the
pair—mate, reproduce, and care for offspring where appropriate—are
perfectly aligned, and there is no sexual conflict (Rice, 2000). However, where
one or both of the pair remate with one or more other individuals, there is
the potential for sexual conflict. For example, if a female has mated with a
poor-quality mate the first time around, it would benefit her to “trade up”
should a better-quality mate arrive on the scene (reviewed in Andersson,
1994; Jennions & Petrie, 2000); this would be detrimental to the first male
SEXUAL CONFLICT : ALL ’ S NOT FAIR IN LOVE — IT ’ S WAR ! 175

in a number of ways. If the female mates with both males within the same
reproductive cycle, this would likely lead to competition between the sperm
of the two males within the female reproductive tract and could therefore
decrease the number of offspring sired by the first male. Alternatively, in spe-
cies that exhibit joint parental care, the first male may end up providing costly
paternal care to offspring that he has not sired. Thus in many species where
females mate with multiple males, males have evolved traits such as mate
guarding (reviewed in, for example, Alcock, 1994), and plugs that block the
female reproductive tract (e.g., Orr & Rutowski, 1991; Shine, Olsson, et al.,
2000) to deny access to other males until after the female has laid her eggs
or given birth. This is an example of sexual conflict over mating rate. How-
ever, conflict between the male and female can, and does, occur over any
aspect of reproduction, from courtship, through mating and fertilization, to
parental care of offspring (see Figure 6.1).

A UNIVERSAL TRUTH?
Sexual conflict seems to be universal in nature, with examples seen across all
animal groups. In social mammals such as lions and gorillas, where the females
care for their offspring for long periods of time, the females are often rendered
nonreceptive to further reproduction until their current offspring become in-
dependent or die. In such species, new males will often kill the offspring,
thereby bringing the female into estrus sooner (reviewed in Hrdy, 1979).
Infanticide is a rather extreme example of premating sexual conflict; another
example is seen in red-sided garter snakes (Thamnophis sirtalis parietalis),
where it is the females themselves that are under threat during the premating
period. This species typically exhibits populations with highly male-biased
sex ratios, resulting in strong competition between males for mates (Shine,
O’Connor, et al., 2000). On emerging from their communal overwintering
dens, the snakes form mass, writhing, mating aggregations (see Figure 6.2),
containing hundreds of males and very few females. The consequences for
the females include forced copulation, exhaustion, physiological stress, and
in some cases death (Shine, O’Connor, et al., 2000; Shine et al., 2004). Per-
haps unsurprisingly, even humans are thought to exhibit sexual conflict over,
for example, family size, fidelity within relationships, and provisioning to
offspring (reviewed in Mulder & Rauch, 2009).
Recent studies have begun examining what appears to be a fascinating
example of sexual conflict over fertilization in waterfowl. It was first docu-
mented a century ago that “rape” (forced copulation) is extremely common
in ducks, with the males attempting to coerce the females into mating so
176 ANIMAL BEHAVIOR

Figure 6.1. Sexual conflict can occur over any aspect of, and during any episode of, the
reproductive process. (Adapted from Chapman, 2006, with permission from Elsevier)

violently they will on occasion drown them (Huxley, 1912). Ducks are
unusual for birds in that the males have penises; even more remarkable is the
structure and functioning of the duck phallus. The duck penis is a highly com-
plex, coiled structure, resembling a corkscrew (Brennan et al., 2010). At rest
within the male, the penis is essentially inside out. Prior to copulation, the
SEXUAL CONFLICT : ALL ’ S NOT FAIR IN LOVE — IT ’ S WAR ! 177

Figure 6.2. A red-sided garter snake (Thamnophis sirtalis parietalis) mating aggregation;
hundreds of male compete for access to only a few females. (Photo by Jon Webb)

male does not develop an erection; instead, as copulation commences, the

penis quickly fills with fluid and is everted explosively into the female repro-
ductive tract. Penis eversion, followed by ejaculation and reversion back into
the male, all take place in a matter of seconds (Brennan et al., 2010). The
structure and functioning of the duck penis is likely to facilitate forced copu-
lations in this group; however, it seems females have evolved to counteract
this. The female reproductive tract is also highly complex, with many blind-
ending pouches and a coiled structure; however, interestingly, it coils in the
opposite direction to this male penis, making it somewhat challenging for
the male to inseminate her (Brennan et al., 2007). Although as yet untested,
Birkhead (2010) has suggested that during forced copulations, females tighten
their reproductive tract, making intromission more difficult and directing the
phallus into one of the blind-ended pouches; however, when mating with
desired mates, the female relaxes her tract, permitting intromission to occur.
The insects exhibit some of the most interesting examples of conflict over
reproduction. Perhaps the most famous is seen in the mantids, where the
female eats the male during copulation and thereby derives nutrients, which
178 ANIMAL BEHAVIOR

increase her fecundity. Lelito and Brown (2006) have shown that the male
praying mantis Tenodera aridifolia sinensis has adapted to recognize the level
of potential danger presented by the female and adopts risk-averse behavior
as appropriate. Female water striders, or “pond skaters” as they are more com-
monly known, struggle during copulation and attempt to kick the mounted
male from their back; the males have responded in evolutionary terms by
evolving spines and claspers on their legs, which assist them in gripping the
female (Arnqvist & Rowe, 2002). Closely related to the water striders, the
Zeus bugs—tiny water bugs found in New Guinea and tropical Australia—
are relatively new to science, having only been discovered in the past 10 years
(Polhemus & Polhemus, 2000; Andersen & Weir, 2001). The group exhibits
a remarkable mating system whereby the male, diminutive in size, remains
almost permanently attached to the back of the female, feeding from a pair
of glands on her back that produce a wax-like secretion (Arnqvist et al.,
2007). It has been suggested that the male behavior is an extreme form of mate
guarding; the adult sex ratio in wild populations is strongly male-biased, and
thus single males have little opportunity to find unoccupied females. Rather
than risk losing their mate by dismounting to forage for food, males stay put
and feed from the female glands. Why does the female permit this? Males of this
group take further liberties with the females, stealing prey items when they are
feeding and apparently even wounding them, evidenced by the presence of
melanized scarring on the backs of most of the females. When blocking the
female glands experimentally, Arnqvist and colleagues (2006) observed that
males increased the extent to which they steal prey items from the female, sug-
gesting that the female glands have evolved to minimize the costly behaviors
exhibited by the males.
One of the more bizarre examples of sexual conflict in insects is exhibited
by the bed bugs (see Figure 6.3, reviewed in Siva-Jothy, 2006). Bed bugs
engage in traumatic insemination; rather than employing more traditional
routes, males inject sperm directly into the female body cavity through the
abdominal wall via their syringe-like penises. Unsurprisingly, this is costly
for the female; eventually, after multiple matings, it results in death. However,
to combat this, females have evolved a specialized organ, the ectospermalege.
This structure acts essentially as a mating guide for the male, directing his
penis to a specific region of the female abdomen containing cells involved in
the immune response, which help minimize the long-term damage to the
female (Reinhardt et al., 2003). From the ectospermalege, the male’s sperm
migrates to the ovaries to inseminate the female’s eggs. It seems the male sub-
mits to this female control over the site of insemination as even low rates of
piercing outside of the ectospermalege can reduce female offspring production
SEXUAL CONFLICT : ALL ’ S NOT FAIR IN LOVE — IT ’ S WAR ! 179

Figure 6.3. Bedbug males engage in traumatic insemination, where they ejaculate their
sperm directly into the female body cavity, one of the most bizarre mating behaviors
in the animal world. (Photo by Richard Naylor)

by 50 percent, thus proving highly costly for both the female and the male
(Morrow & Arnqvist, 2003).

THE SEXUAL ARMS RACE

How does sexual conflict operate? Imagine a mutation arises in a male, which
results in an adaptation that causes harm or generates a cost to his female
mate but is beneficial to him. This would result in selection acting on the
female to evolve a counteradaptation that reduces or prevents the male-
induced harm. In turn, selection then acts on the male to evolve another adap-
tation that overcomes or reduces the action of the female counteradaptation,
and so on (Rice, 2000). This results in continual, cyclical coevolution between
the male and female, known as sexually antagonistic coevolution (SAC). The
ensuing “arms race” is akin to the host-parasite relationship, whereby the host
and parasite are locked in a continual battle to gain one-upmanship over the
other.
Chapter 5 described a number of mechanisms by which male traits
can evolve under sexual selection in the absence of direct benefits to females
such as food and parental care. Sexual conflict is inherently linked to sexual
180 ANIMAL BEHAVIOR

selection; females may mate with some males but not others as a result of sex-
ual conflict, thereby leading to variation in reproductive success among males
and contributing to the evolution and maintenance of exaggerated male traits.
Thus sexual conflict is often invoked as a further mechanism by which sexual
selection can operate. The difference with traditional models of sexual selec-
tion is that these tend to focus on direct and indirect benefits of mating pref-
erence to the female; under sexual selection via sexual conflict, female choice
is an attempt to minimize the costs of mating rather than maximize benefits
(Rice & Holland, 1997; Holland & Rice, 1998). The relative importance of
sexual conflict as an instrument of sexual selection, compared to more tradi-
tional mechanisms, is still hotly debated (e.g., Chapman et al., 2003; Cordero
& Eberhard, 2003; Córdoba-Aguilar & Contreras-Garduño, 2003; Eberhard
& Cordero, 2003; Eberhard, 2004).
For behavioral scientists, one unfortunate consequence of the manner in
which sexual conflict operates, in addition to lack of consensus over terminol-
ogy and manner of operation, is that it is difficult to study. As a result of the
nature of the sexual arms race, past adaptations are hidden, and at any given
moment it is difficult to measure the relative costs and benefits of the conflict
to each sex. However, behavioral scientists have found ways to get around this.
Currently one of the most fruitful methods of investigating sexual conflict, as
seen in the Drosophila melanogaster case study below, is through the use of
experimental evolution, where experimental treatments are imposed on repli-
cate populations of organisms in the laboratory; the organisms are then mea-
sured for traits of interest over time to see how these traits evolve. In
addition, comparative studies across species, and population crosses within
species, have proven fruitful as, again, we shall see in some of the examples
given below.

THE LINK WITH MATING SYSTEM

The degree of sexual conflict is often invoked as being related to the level of
promiscuity within a species or population, and in the past it has been sug-
gested that sexual conflict does not occur where pairs are monogamous (e.g.,
Rice, 2000). However, it is becoming increasingly recognized that this is not
the case; in fact, sexual conflict can promote monogamy in polygamous spe-
cies (reviewed in Hosken et al., 2009).
As noted above, in some species where females mate with multiple mates, a
male will guard the female after copulation or block her reproductive tract
with a mating plug, thereby protecting his reproductive success to some
degree. In some cases, the female is permanently prevented from remating.
SEXUAL CONFLICT : ALL ’ S NOT FAIR IN LOVE — IT ’ S WAR ! 181

For example, in the housefly, Musca domestica, the male transfers compounds
to the female in his ejaculate that permanently switch off her receptivity to
other males (Andrés & Arnqvist, 2001). However, females will remate with
another male if given the opportunity, and there is evidence that multiple mat-
ing could increase the reproductive success of polyandrous females through
nutrients transferred to females via the seminal fluids (Arnqvist & Andrés,
2006). Thus, in this species, males enforce monandry in females despite the
fact it may be costly to female reproductive success.
It is not only males that exert control over the remating of their partners.
The burying beetles, Nicrophorus spp., are a fascinating group of organisms
for a number of reasons, one of which is the fact that it is one of the few insect
groups in which both males and females engage in parental care (reviewed in
Scott, 1998). Mated pairs seek out, bury, and then defend the carcass of a
small vertebrate, for example a rodent or bird. The females lay their eggs in
the carcass, and one or both of the parents care for the developing offspring.
If the pair manages to find a carcass large enough to support more than one
brood of offspring, it would increase the reproductive success of the male to
rear a brood with a second female on the same carcass. This would, however,
be costly for the resident female, as her offspring would be subject to compe-
tition for the food resource. In one species, Nicrophorus defodiens, males that
are defending a large carcass will attempt to emit pheromones to attract a sec-
ond mate. However, their efforts are often thwarted by the resident female,
who will push, undercut, and bite the male to minimize the time he spends
signalling to other prospective mates (Eggert & Sakaluk, 1995) and thereby
decrease the chance of his obtaining a second mate. Thus, in this species, sex-
ual conflict over remating can result in females enforcing monogyny on males.
Sexual conflict over remating can also result in both sexes within a pair
being forced to mate only once. In many spider species, males exhibit so-
called one-shot genitalia, inseminating the female by breaking off parts of their
genitalia containing their sperm and leaving them in the female reproductive
tract (e.g., Fromhage & Schneider, 2006). This blocks the female reproductive
tract, thereby preventing her from remating with other males, but also renders
the male unable to remate despite the fact that he may survive the ordeal.
Thus, in theory, conflict over remating in such cases could result in
monogamy. However, this is not necessarily the end of the conflict; two recent
studies in different spider species suggest that both the female and the male
can, in evolutionary terms, eventually override the stalemate, in the case of
the female by adapting to remove the plug (Kuntner et al., 2009), and in the
case of the male by refraining from damaging his genitalia and instead parti-
tioning sperm across several copulations (Schneider & Michalik, 2011).
182 ANIMAL BEHAVIOR

A CASE STUDY: SEXUAL CONFLICT IN A FRUIT FLY

One of the best-understood examples of sexual conflict is over mating rate
in the fruit fly model organism, D. melanogaster (Figure 6.4). In 1989, a team
at Edinburgh University showed that for female D. melanogaster, mating has
an unfortunate consequence—they die sooner (Fowler & Partridge, 1989).
Kevin Fowler and Linda Partridge compared the adult longevity of females
housed with, and allowed to mate freely with, wild-type males (the “high-
mating” group), and females alternately housed with wild-type males and
males that had had their genitalia cauterised (the “low-mating” group). Thus
the low-mating group females were constantly exposed to males, as were the
high-mating group, but they had fewer opportunities to actually mate. It was
found that high-mating-group females had significantly shorter lifespans than
low-mating-group females, and as they died sooner they exhibited a correlated
decrease in reproductive success. But why should males “wish” their mates to die
quickly, particularly when their decreased lifespan results in their producing fewer
offspring? And how is the conflict mediated? In 1995 the same team, now based
at University College London, took further steps towards answering these
questions. They essentially repeated the experiment, but this time they
exposed females to experimental, genetically modified males of two types.

Figure 6.4. The fruit fly model organism Drosophila melanogaster presents one of the
best-understood examples of sexual conflict. (Photo by Tracey Chapman)
SEXUAL CONFLICT : ALL ’ S NOT FAIR IN LOVE — IT ’ S WAR ! 183

One group was unable to produce sperm or the major components of seminal
fluid (“DTA males”), while the other produced normal seminal fluid but not
sperm (“tudor” males) (Chapman et al., 1995). As before, cauterized males
from each of the three stocks were also used to control for nonmating exposure
of females to males. They found that females that mated to the tudor males,
and thus received the full complement of seminal fluid, died sooner than
females mated to the DTA males, suggesting that it was something in the
seminal fluid that was reducing female lifespan.
In most species, seminal fluid is a veritable cocktail of components pro-
duced by the male accessory glands, seminal vesicles, ejaculatory duct, ejacula-
tory bulb, and testes and transferred to the female during mating. These so-
called seminal fluid proteins (SFPs) have wide-ranging effects on the female,
some good and some bad (reviewed in Avila et al., 2011). For example, in
D. melanogaster, certain SFPs are essential to the female for processes such as
sperm storage in, and release from, the sperm-storage organ (e.g., Neubaum
& Wolfner, 1999; Tram & Wolfner, 1999; Avila et al., 2010). Of benefit to
the male, SFPs also increase the oviposition or egg-laying rate of the female
(e.g., Herndon & Wolfner, 1995; Heifetz et al., 2001; Ravi Ram & Wolfner,
2007), thus ensuring that she produces many eggs fertilized with his sperm.
However, this is where the conflict occurs. Although the female may now have
sufficient sperm to fertilize her eggs, she may not wish to fertilize all of them
using the male’s sperm; as we saw in Chapter 5 on sexual selection, there are
numerous reasons why the female may wish to remate with one or more other
males. If the female were to remate, the male’s sperm would likely be subject
to sperm competition. In D. melanogaster, subsequent males sire a greater
proportion of a female’s offspring (Gromko et al., 1984), and the first
male’s reproductive success would therefore decrease. However, males have
weapons to counter this. The wide-ranging effects of male SFPs on females
include manipulation of the female to decrease her interest in other males;
after receiving SFPs, mated females will actively reject courting males
(e.g., Aigaki et al., 1991).
What of the impact on female lifespan with which we began this story? To our
knowledge, the SFP-induced reduction in female longevity is an unfortunate
side effect of the overall actions of SFP; it is unlikely that male D. melanogaster
“intentionally” set out to harm their mates. As stated previously, sexual con-
flict is predicted to result in a cyclical arms race, with males and females con-
tinuously evolving to counter the selection imposed on them by the opposite
sex. Indeed, it seems female D. melanogaster have tricks up their proverbial
sleeves to counter the effects of male SFPs. A more recent study, again by
Tracey Chapman’s group, utilized experimental evolution to examine female
184 ANIMAL BEHAVIOR

responses to male-induced harm in D. melanogaster (Wigby & Chapman,

2004). They created replicate populations subject to differing levels of sexual
conflict by manipulating the adult sex ratios of each line; it was predicted that
in populations with male-biased sex ratios, females would be subject to more
frequent mating attempts from males and would therefore mate at a higher
rate than females from populations with equal-sex, and, in turn, female-
biased sex ratios. Thus, sexual conflict was predicted to be highest in male-
biased populations, followed by equal-sex-ratio populations, followed by
female-biased populations. After 31 generations of these treatments, Stuart
Wigby and Chapman (2004) measured the frequencies at which females were
copulating in the different populations. As predicted, females were courted
and mated most frequently in male-based populations, followed by the
equal-sex-ratio populations, followed by the female-biased populations. They
also compared the adult longevity of females by housing them with wild-
type (i.e., nonselected) males. Remarkably, females from the male-biased or
high-sexual-conflict populations survived longer than females from the other
populations, suggesting that after generations of being subject to high levels
of sexual conflict, the females had evolved a counteradaptation to minimize
the costs induced by males. The nature of this counteradaptation is as yet
unknown, although it is thought to be a modification in the female receptors
that are manipulated by the male SFPs.
The story does not quite end there. Another study has shown that D. mel-
anogaster females seem to prefer the males that induce the most harm; it is
known that females of this species prefer larger males, and Urban Friberg
and Göran Arnqvist (2003) showed that larger males induce a greater longev-
ity cost to females. It is currently unknown how the cost and body-size traits
are linked, or whether females actually prefer larger males or whether the target
of their preference is something linked to body size. However, this finding
provides support to models of sexual selection by sexual conflict that suggest
female preference can result in the evolution and maintenance of exaggerated
male traits despite the fact that they may impose costs on females.
It is only more recently, with the advent of the molecular age, that we have
come closer to understanding the biochemical processes underlying the sexual
conflict in D. melanogaster. We now know that one of the chief culprits medi-
ating male-induced harm to females is the SFP Acp70a, more commonly
known as sex peptide (SP) (Wigby & Chapman, 2005). Interestingly, experi-
mental application of SP to female Helicoverpa armigera moths has been found
to induce similar effects to those in D. melanogaster (Fan et al., 2000), and a
homologue of SFP—a highly genetically similar molecule—has recently been
found in mosquitoes (Dottorini et al., 2007; reviewed in Avila et al., 2011).
SEXUAL CONFLICT : ALL ’ S NOT FAIR IN LOVE — IT ’ S WAR ! 185

These studies suggest that the insect female physiological pathway, which is
targeted by males via SFPs, may be an ancient and highly conserved one.

CONSEQUENCES OF SEXUAL CONFLICT

Sexual conflict is thought to have the potential to drive reproductive isola-
tion and hence speciation (reviewed in Ritchie, 2007). Numerous theoretical
models predict that cycles of SAC between males and females can to lead to
rapid evolutionary change within species and, in turn, promote the develop-
ment of reproductive barriers between populations that can eventually result
in speciation (e.g., Parker & Partridge, 1998; Gavrilets & Waxman, 2002;
Hayashi et al., 2007). Until recently, the best evidence we had that sexual con-
flict could lead to speciation came from comparative studies across different
species. For example, Arnqvist and colleagues (2000) compared speciation rate
across pairs of groups of closely related insects that differed in their potential
for sexual conflict. Their reasoning was that in insect species where females
mate with many males, there should be greater potential for sexual conflict
than in species in which females mated only once, and thus groups of insects
with higher levels of sexual conflict should contain more species if sexual con-
flict can result in speciation. They found this was indeed the case; groups
where females mated with many males exhibited speciation rates four times
as high as in groups where females only once. Subsequent studies have pro-
duced contrasting results; for example, similar studies across butterflies, mam-
mals, and spiders (Gage et al., 2002) and another in birds (Morrow et al.,
2003) found no evidence for a relationship between female mating rate and
speciosity. However, a more recent comparative analysis of all the data of spe-
ciosity versus sexual selection, and hence sexual conflict, collected to date sug-
gests that there is a small but significant association between the two
(Kraaijeveld et al., 2011).
There is some experimental evidence that sexual conflict could lead to
reproductive isolation. Oliver Martin and David Hosken (2003) created repli-
cate populations of the dung fly (Sepsis cynipsea) under experimental evolution
where they were subject to varying levels of sexual conflict. They did this by
manipulating the densities of the populations, as sexual conflict is predicted
to be stronger in larger populations, where there are more frequent interac-
tions between males and females (Gavrilets, 2000). After 35 generations of
evolution, Martin and Hosken (2003) found that in the populations under
potentially high levels of sexual conflict, females had evolved to be better at
resisting male attempts to mate compared to females from lines with little or
no sexual conflict. In theory this could result in reproductive barriers to
186 ANIMAL BEHAVIOR

mating, as females that evolve increased resistance to males of their population

due to sexual conflict may refuse to mate with males from other populations
altogether. Over time, this barrier to gene flow between the populations could
eventually result in divergence and finally speciation. A recent study in a grass-
hopper potentially provides the first evidence of this happening in wild popu-
lations. Yoshikazu Sugano and Shin-ichi Akimoto (2007) had previously
shown that crossing different populations of the Japanese grasshopper Podisma
sapporensis revealed similar asymmetries in their mating behaviors to the exper-
imental lines described above; males of a given population mated more fre-
quently with females from a different population than with females of their
own population, yet females tended to prefer males of their own population.
Further analysis, utilizing additional populations of the grasshopper, supports
the hypothesis that there is asymmetric premating reproductive isolation
within this species as a result of geographic differences in male and female
mating propensity that could over time result in divergence (Sugano &
Akimoto, 2011).
It is not yet clear whether divergence under sexual conflict is a general phe-
nomenon; indeed, several other experimental studies have found weak or no evi-
dence that sexual conflict promotes reproductive isolation (e.g., Wigby &
Chapman, 2006; Gay et al., 2009). Increasingly, genetics and molecular studies
are informing our knowledge with regard to how speciation operates. It has been
found that proteins involved in reproduction, which we would predict to be cru-
cial players in the operation of sexual conflict, are the most fast-evolving group of
proteins (e.g., Swanson et al., 2001; reviewed in Swanson & Vacquier, 2002),
lending some support to predications that sexual conflict can lead to reproductive
isolation.
It has also been suggested that sexual conflict could lead to the extinction of
populations. A number of theoretical models suggest that, for example, male-
induced costs to females could lower the overall fitness of a population to the
extent that the population may become inviable and go extinct (reviewed in
Kokko & Brooks, 2003). For example, Rankin and colleagues (2011) theo-
rized that the evolution of male-induced “harassment” (cost) to females could
result in a positive feedback cycle whereby fewer females survived each gener-
ation as a result of the harassment, and those that did had higher mortality as a
result of the consequent high ratio of males to females in the population and
thus increased levels of harassment. Eventually, according to their model, the
population collapsed. However, further to this, they found that the evolution
of a female countertrait could ameliorate this and prevent population extinc-
tion. A comparative analysis of sexual selection and the risk of extinction in
mammals found no support for the hypothesis (Morrow & Fricke, 2004),
SEXUAL CONFLICT : ALL ’ S NOT FAIR IN LOVE — IT ’ S WAR ! 187

and to date we have no direct empirical evidence that sexual conflict has or
could drive a population to extinction.
However, there are conditions where sexual conflict could exacerbate the
population decline to extinction in conjunction with other factors. Some pop-
ulations of the fruit fly Drosophila pseudoobscura are infected with a selfish
genetic element (SGE) called sex ratio (SR) meiotic drive; SGEs are pieces
of DNA that defy the usual rules of inheritance and, through a variety of
mechanisms, promote their own transfer to the next generation at the expense
of the other genes in an organism (reviewed in Hurst & Werren, 2001). In
D. pseudoobscura, males infected with SR cannot sire sons and thus only pro-
duce daughters, which can lead to populations with female-biased sex ratios.
This can be costly to a female that mates with an SR male for several reasons;
for example, not only does she only produce daughters, but her offspring may
have reduced fitness due to inheriting the SGE (Jaenike, 2001). Price and col-
leagues (2008) showed that the presence of SR in a population results in
females evolving increased mating rates; it was suggested that this promoted
sperm competition and thus reduced the numbers of offspring that SR males
sired, thereby minimizing to some degree the costs associated with mating
with such males. However, a subsequent study showed that males countered
this by evolving an increased ability to suppress this female mating response,
suggesting that SR can promote sexual conflict in this species (Price et al.,
2009). The final twist to this tale is that if females within an SR-infected pop-
ulation are prevented from mating multiply, and thereby decreasing the num-
ber of their offspring sired by SR males, the population can go extinct due to a
lack of males (Price et al., 2010). Although not directly shown, this suggests
that in a population infected with SR, if sexual conflict prevents females from
resisting the costs associated with SR, it could lead to extinction.

CONCLUSIONS AND FUTURE DIRECTIONS

The term sexual conflict was first defined by Parker in 1979, yet it is only in
the last decade that interest in this field of research has gained momentum.
Increasingly, as with many topics within science, new technologies and innova-
tive use of preexisting technologies are allowing us to delve deeper into the mech-
anisms underlying the operation of sexual conflict. For example, a team at
Syracuse University recently produced transgenic D. melanogaster males express-
ing proteins in their sperm that fluoresced red or green under specific wave-
lengths of light (Manier et al., 2010). As a result, they could follow the sperm
on their journey through the female reproductive tract and all the while unam-
biguously discriminate between the sperm of two males. This has allowed them
188 ANIMAL BEHAVIOR

to resolve some of the long-standing questions about mechanisms of postcopula-

tory sexual selection. One of their findings, which had previously only been
inferred (Snook & Hosken, 2004), was that following copulation and before
egg-laying commences, females eject sperm from their body, presumably
allowing them to excise some control over who, out of their multiple mates, fer-
tilizes their eggs. Michal Polak’s team at the University of Cincinnati has been
using laser surgery to study the function of the microscopic structures used in
reproduction by insects. For example, using high-precision lasers to remove indi-
vidual spines—in some cases less than 1 µm in size—from the genitals of D. mel-
anogaster males, and then measuring the reproductive success of the males, they
found that the spines aided males in obtaining copulations and in competing
with rival males but had no effect on insemination and fertilization success (Polak
& Rashed, 2010). This has led Polak and Arash Rashed to suggest that the broad
morphological diversity seen in Drosophila genitalia is the result of sexual conflict
prior to insemination, supporting the idea that if the spines are an adaptation to
overcome female resistance to mating, differences in spine morphology across
species may represent differences in the form and intensity of female resistance
(Arnqvist & Rowe, 2005).
Many suspected examples of sexual conflict, some of which have been
described here, we actually know very little about. Until we can measure the
actual relative costs and benefits to males and females of their reproductive
behaviors, something that is surprisingly difficult to do, we are in danger of being
left only with anecdotal or speculative evidence. For example, it may appear that
a male trait is costly to females, yet it might be the case that, in fact, the costs of
female resistance to the trait actually outweigh the cost imposed by the trait itself.
In addition, we still have little understanding of how conflicts between males and
females might be resolved if indeed resolution occurs. Who, if anyone, wins? Or Is
there no resolution, only continuing cycles of adaptation and counteradaptation?
Quantifying animal behavior in general, including sexual conflict, is particularly
difficult under field conditions; however, it is important not to become too
focused on the laboratory, despite the diverse avenues that new techniques open
up for us. Although such laboratory studies undoubtedly increase our under-
standing of how sexual selection operates, as with all studies of evolution and
behavior, what happens in natural conditions in the field is of paramount impor-
tance. Sexual conflict is a fascinating aspect of animal behavior, but we still have a
great deal to learn.

ACKNOWLEDGMENTS
The author would like to thank the Japan Society for the Promotion of Sci-
ence for funding the research trip to Japan where much of this chapter was
SEXUAL CONFLICT : ALL ’ S NOT FAIR IN LOVE — IT ’ S WAR ! 189

conceived. She would also like to thank Professor Takahisa Miyatake and
Doctor Tom Price for valuable discussion on the topic, and her Animal Behav-
iour and Sociobiology students for providing inspiration.

REFERENCES AND SUGGESTED READINGS

Aigaki, T., I. Fleischmann, P. Chen, & E. Kubli. (1991). Ectopic expression of sex
peptide alters reproductive behaviour of female Drosophila melanogaster.
Neuron, 7, 557–563.
Alcock, J. (1994). Postinsemination associations between males and females in insects:
The mate-guarding hypothesis. Annual Review of Entomology, 39, 1–21.
Andersen, N. M. & A. T. Weir. (2001). New genera of Veliidae (Hemiptera: Heter-
optera) from Australia, with notes on the generic classification of the subfamily
Microveliinae. Invertebrate Taxonomy, 15, 217–258.
Andersson, M. (1994). Sexual Selection. Princeton, NJ: Princeton University Press.
Andrés, J. A. & G. Arnqvist. (2001). Genetic divergence of the seminal signal-
receptor system in houseflies: The footprint of sexually antagonistic coevolu-
tion. Proceedings of the Royal Society of London, B, 268, 399–405.
Arnqvist, G. & J. A. Andrés. (2006). The effects of experimentally induced polyandry on
female reproduction in a monandrous mating system. Ethology, 112, 748–756.
Arnqvist, G., M. Edvardsson, U. Friberg, & T. Nilsson. (2000). Sexual conflict pro-
motes speciation in insects. Proceedings of the National Academy of Sciences,
USA, 97, 10460–10464.
Arnqvist, G., T. M. Jones, & M. A. Elgar. (2006). Sex-role reversed nuptial feeding
reduces kleptoparasitism of females in Zeus bugs (Heteroptera; Veliidae). Biol-
ogy Letters, 2, 491–493.
Arnqvist, G., T. M. Jones, & M. A. Elgar. (2007). The extraordinary mating system
of Zeus bugs (Heteroptera: Veliidae: Phoreticovelia sp.). Australian Journal of
Zoology, 55, 131–137.
Arnqvist, G. & L. Rowe. (2002). Antagonistic coevolution between the sexes in a
group of insects. Nature, 415, 787–789.
Arnqvist, G. & L. Rowe. (2005). Sexual Conflict. Princeton, NJ: Princeton University
Press.
Avila, F. W., K. Ravi Ram, M. C. Bloch Qazi, & M. F. Wolfner. (2010). Sex peptide
is required for the efficient release of stored sperm in mated Drosophila females.
Genetics, 186, 595–600.
Avila, F. W., L. K. Sirot, B. A. LaFlamme, C. D. Rubinstein, & M. F. Wolfner.
(2011). Insect seminal fluid proteins: Identification and function. Annual
Review of Entomology, 56, 21–40.
Birkhead, T. R. (2010). How stupid to have thought of that: Post-copulatory sexual
selection. 2010. Journal of Zoology, 281, 78–93.
Bonduriansky, R. & S. F. Chenoweth. (2009). Intralocus sexual conflict. Trends in
Ecology and Evolution, 24, 280–288.
190 ANIMAL BEHAVIOR

Brennan, P. L. R., C. J. Clark, & R. O. Prum. (2010). Explosive eversion and func-
tional morphology of the duck penis supports sexual conflict in waterfowl
genitalia. Proceedings of the Royal Society of London, B, 277, 1309–1314.
Brennan, P. L. R., R. O. Prum, K. G. McCracken, M. D. Sorensen, R. E. Wilson, &
T. R. Birkhead. (2007). Coevolution of male and female genital morphology
in waterfowl. PLoS ONE, 5, e418.
Chapman, T. (2006). Evolutionary conflicts of interest between males and females.
Current Biology, 16, R744–R754.
Chapman, T., G. Arnqvist, J. Bangham, & L. Rowe. (2003). Response to Eberhard
and Cordero, and Córdoba-Aguilar and Contreras-Garduño: Sexual conflict
and female choice. Trends in Ecology and Evolution, 18, 440–441.
Chapman, T., L. F. Liddle, J. M. Kalb, M. F. Wolfner, & L. Partridge. (1995). Cost
of mating in Drosophila melanogaster females is mediated by male accessory
gland products. Nature, 373, 241–244.
Cordero, C. & W. G. Eberhard. (2003). Female choice of sexually antagonistic male
adaptations: A critical review of some current research. Journal of Evolutionary
Biology, 16, 1–6.
Córdoba-Aguilar, A. & J. Contreras-Garduño. (2003). Sexual conflict. Trends in
Ecology and Evolution, 18, 439–440.
Dottorini, T., L. Nicolaides, H. Ranson, D. W. Rogers, A. Crisanti, & F. Catteruccia.
(2007). A genome-wide analysis in Anopheles gambiae mosquitoes reveals
46 male accessory gland genes, possible modulators of female behaviour.
Proceedings of the National Academy of Sciences, USA, 104, 16215–16220.
Eberhard, W. G. (2004). Rapid divergent evolution of sexual morphology: Compara-
tive tests of antagonistic coevolution and traditional female choice. Evolution,
58, 1947–1970.
Eberhard, W. G. & C. Cordero. (2003). Sexual conflict and female choice. Trends in
Ecology and Evolution, 18, 438–439.
Eggert, A. K. & S. K. Sakaluk. (1995). Female-coerced monogamy in burying beetles.
Behavioral Ecology and Sociobiology, 37, 147–153.
Fan, Y. L., A. Rafaeli, P. Moshitzky, E. Kubli, Y. Choffat, & S. W. Applebaum.
(2000). Common functional elements of Drosophila melanogaster seminal
peptides involved in reproduction of Drosophila melanogaster and Helicoverpa
armigera females. Insect Biochemistry and Molecular Biology, 30, 805–812.
Fowler, K. & L. Partridge. (1989). A cost of mating in female fruitflies. Nature, 338,
760–761.
Friberg, U. & G. Arnqvist. (2003). Fitness effects of female mate choice: Preferred
males are detrimental for Drosophila melanogaster females. Journal of Evolution-
ary Biology, 16, 797–811.
Fromhage, L. & J. M. Schneider. (2006). Emasculation to plug up the females: The
significance of pedipalp damage in Nephila fenestrata. Behavioral Ecology, 17,
353–357.
SEXUAL CONFLICT : ALL ’ S NOT FAIR IN LOVE — IT ’ S WAR ! 191

Gage, M. J. G., G. A. Parker, S. Nylin, & C. Wiklund. (2002). Sexual selection and
speciation in mammals, butterflies and spiders. Proceedings of the Royal Society
of London, B, 269, 2309–2316.
Gavrilets, S. (2000). Rapid evolution of reproductive barriers driven by sexual con-
flict. Nature, 403, 886–889.
Gavrilets, S. & D. Waxman. (2002). Sympatric speciation by sexual conflict. Proceed-
ings of the National Academy of Sciences, USA, 99, 10533–10538.
Gay, L., P. E. Eady, R. Vasudev, D. J. Hosken, & T. Tregenza. (2009). Does repro-
ductive isolation evolve faster in larger populations via sexually antagonistic
coevolution? Biology Letters, 5, 693–696.
Gromko, M. H., D. G. Gilbert, & R. C. Richmond. (1984). Sperm transfer and use
in multiple mating system of Drosophila. In R. L. Smith (ed.), Sperm Competi-
tion and the Evolution of Animal Mating Systems (pp. 372–427). New York:
Academic Press.
Hayashi, T. I., M. Vose, & S. Gavrilets. (2007). Genetic differentiation by sexual
conflict. Evolution, 61, 516–529.
Heifetz, Y., U. Tram, & M. F. Wolfner. (2001). Male contributions to egg produc-
tion: The role of accessory gland products and sperm in Drosophila mela-
nogaster. Proceedings of the Royal Society of London, B, 268, 175–180.
Herndon, L. A. & M. F. Wolfner. (1995). A Drosophila seminal fluid protein,
Acp26a, stimulates egg laying in females for 1 day after mating. Proceedings of
the National Academy of Sciences, USA, 92, 10114–10118.
Holland, B. & W. R. Rice. (1998). Chase-away sexual selection: Antagonistic seduc-
tion versus resistance. Evolution, 52, 1–7.
Hosken, D. J., P. Stockley, T. Tregenza, & N. Wedell. (2009). Monogamy and the
battle of the sexes. Annual Review of Entomology, 54, 361–378.
Hrdy, S. B. (1979). Infanticide among animals: A review, classification, and examina-
tion of the implications for the reproductive strategies of females. Ethology and
Sociobiology, 1, 13–40.
Hurst, G. D. D. & J. H. Werren. (2001). The role of selfish genetic elements in
eukaryotic evolution. Nature Reviews, 2, 597–606.
Huxley, J. S. (1912). A “disharmony” in the reproductive habits of the wild duck
(Anas boschas L.). Biologisches Zentralblatt, 32, 621–623.
Jaenike, J. (2001). Sex chromosome meiotic drive. Annual Review of Ecology, Evolu-
tion and Systematics, 32, 25–49.
Jennions, M. & M. Petrie. (2000). Why do females mate multiply? A review of the
genetic benefits. Biological Reviews, 75, 21–64.
Kokko, H. & R. Brooks. (2003). Sexy to die for? Sexual selection and the risk of
extinction. Annales Zoologici Fennici, 40, 207–219.
Kraaijeveld, K., F. J. L. Kraaijeveld-Smit, & M. E. Maan. (2011). Sexual selection
and speciation: The comparative evidence revisited. Biological Reviews, 86,
67–377.
192 ANIMAL BEHAVIOR

Kuntner, M., S. Kralj-Fisher, J. M. Schneider, & D. Li. (2009). Mate plugging via
genital mutilation in nephilid spiders: An evolutionary hypothesis. Journal of
Zoology, 277, 257–266.
Lelito, J. P. & W. D. Brown. (2006). Complicity over sexual cannibalism? Male risk
taking in the praying mantis Tenodera aridifolia sinensis. American Naturalist,
168, 263–269.
Lewis, Z., N. Wedell, & J. Hunt. (2011). Evidence for strong intralocus sexual con-
flict in the Indian meal moth, Plodia interpunctella. Evolution, 65, 2085–2097.
Manier, M. K., J. M. Belote, K. S. Berben, D. Novikov, W. T. Stuart, & S. Pitnick.
(2010). Resolving mechanisms of competitive fertilization success in
Drosophila melanogaster. Science, 328, 354–357.
Martin, O. Y. & D. J. Hosken. (2003). The evolution of reproductive isolation
through sexual conflict. Nature, 423, 979–982.
Morrow, E. H. & G. Arnqvist. (2003). Costly traumatic insemination and a female
counter-adaptation in bed bugs. Proceedings of the Royal Society of London, B,
270, 2377–2381.
Morrow, E. H. & C. Fricke. (2004). Sexual selection and the risk of extinction in
mammals. Proceedings of the Royal Society of London, B, 271, 2395–2401.
Morrow, E. H., T. E. Pitcher, & G. Arnqvist. (2003). No evidence that sexual
selection is an “engine of speciation” in birds. Ecology Letters, 6, 228–234.
Mulder, M. B. & K. L. Rauch. (2009). Sexual conflict in humans: Variations and
solutions. Evolutionary Anthropology, 18, 201–214.
Neubaum, D. M. & M. F. Wolfner. (1999). Mated Drosophila melanogaster females
require a seminal fluid protein, Acp 36DE, to store sperm efficiently. Genetics,
153, 845–857.
Orr, A. G. & R. L. Rutowski. (1991). The function of the sphragis in Cressida cressida
(Fab) (Lepidoptera, Papilionidae): A visual deterrent to copulation attempts.
Journal of Natural History, 25, 703–710.
Parker, G. A. (1979). Sexual selection and sexual conflict. In M. S. Blum & N. A.
Blum (eds.), Selection and Reproductive Competition in Insects (pp. 123–166).
New York: Academic Press.
Parker, G. A. & L. Partridge. (1998). Sexual conflict and speciation. Philosophical
Transactions of the Royal Society of London, B, 353, 261–274.
Polak, M. & A. Rashed. (2010). Microscale laser surgery reveals adaptive function of
male intromittant genitalia. Proceedings of the Royal Society of London, B, 277,
1371–1376.
Polhemus, D. A. & T. Polhemus. (2000). Additional new genera and species of
Microveliinae (Heteroptera: Veliidae) from New Guinea and adjacent regions.
Tijdschrift voor Entomologie, 143, 91–123.
Price, T. A. R., D. J. Hodgson, Z. Lewis, G. D. D. Hurst, & N. Wedell. (2008). Self-
ish genetic elements promote polyandry in a fly. Science, 322, 1241–1243.
Price, T. A. R., G. D. D. Hurst, & N. Wedell. (2010). Polyandry prevents extinction.
Current Biology, 20, 471–475.
SEXUAL CONFLICT : ALL ’ S NOT FAIR IN LOVE — IT ’ S WAR ! 193

Price, T. A. R., Z. Lewis, D. T. Smith, G. D. D. Hurst, & N. Wedell. (2009). Sex

ratio drive promotes sexual conflict and sexual coevolution in the fly Drosophila
pseudoobscura. Evolution, 64, 1504–1509.
Rankin, D. J., U. Dieckmann, & H. Kokko. (2011). Sexual conflict and the tragedy
of the commons. American Naturalist, 177, 780–791.
Ravi Ram, K. & M. F. Wolfner. (2007). Sustained post-mating response in Dro-
sophila melanogaster requires multiple seminal fluid proteins. PLoS Genetics,
3, 2428–2438.
Reinhardt, K., R. Naylor, & M. T. Siva-Jothy. (2003). Reducing a cost of traumatic
insemination: Female bedbugs evolve a unique organ. Proceedings of the Royal
Society of London, B, 270, 2371–2375.
Rice, W. R. (2000). Dangerous liaisons. Proceedings of the National Academy of Sci-
ences, USA, 97, 12593–12955.
Rice, W. R. & K. Chippindale. (2001). Intersexual ontogenetic conflict. Journal of
Evolutionary Biology, 14, 685–693.
Rice, W. R. & B. Holland. (1997). The enemies within: Intergenomic conflict, inter-
locus contest evolution (ICE), and the intraspecific red queen. Behavioral Ecol-
ogy and Sociobiology, 41, 1–10.
Ritchie, M. G. (2007). Sexual selection and speciation. Annual Review of Ecology,
Evolution and Systematics, 38, 79–102.
Schneider, J. M. & P. Michalik. (2011). One-shot genitalia are not an evolutionary
dead end—regained male polygamy in a sperm limited spider species. BMC
Evolutionary Biology, 11, 197.
Scott, M. P. (1998). The ecology and behaviour of burying beetles. Annual Review of
Entomology, 43, 595–618.
Shine, R., D. O’Connor, & R. T. Mason. (2000). Sexual conflict in the snake den.
Behavioral Ecology and Sociobiology, 48, 392–401.
Shine, R., M. M. Olsson, & R. T. Mason. (2000). Chastity belts in gartersnakes: The
functional significance of mating plugs. Biological Journal of the Linnaean Soci-
ety, 70, 377–390.
Shine, R., B. Philips, T. Langkilde, D. I. Lutterschmidt, H. Waye, & W. T. Mason.
(2004). Mechanisms and consequences of sexual conflict in garter snakes
(Thamnophis sirtalis, Colubridae). Behavioral Ecology, 15, 654–660.
Siva-Jothy, M. T. (2006). Trauma, disease and collateral damage: Conflict in
cimicids. Philosophical Transactions of the Royal Society of London, B, 361,
269–275.
Snook, R. R. & D. J. Hosken. (2004). Sperm death and dumping in Drosophila.
Nature, 428, 939–941.
Sugano, Y. C. & S. Akimoto. (2007). Asymmetric mating in the brachypterous grass-
hopper Podisma sapporensis. Ethology, 113, 301–311.
Sugano, Y. C. & S. Akimoto. (2011). Mating asymmetry resulting from sexual con-
flict in the brachypterous grasshopper Podisma sapporensis. Behavioral Ecology,
22, 701–709.
194 ANIMAL BEHAVIOR

Swanson, W. J., A. G. Clark, H. M. Waldrip-Dail, M. F. Wolfner, & C. F. Aquadro.

(2001). Evolutionary EST analysis identifies rapidly evolving male reproduc-
tive proteins in Drosophila. Proceedings of the National Academy of Sciences,
USA, 98, 7375–7379.
Swanson, W. J. & V. D. Vacquier. (2002). The rapid evolution of reproductive pro-
teins. Nature Reviews Genetics, 3, 137–144.
Tram, U. & M. F. Wolfner. (1999). Male seminal fluid proteins are essential for
sperm storage in Drosophila melanogaster. Genetics, 153, 837–844.
Wigby, S. & T. Chapman. (2004). Female resistance to male harm evolves in
response to manipulation of sexual conflict. Evolution, 58, 1028–1037.
Wigby, S. & T. Chapman. (2005). Sex peptide causes mating costs in Drosophila
melanogaster. Current Biology, 15, 16–21.
Wigby, S. & T. Chapman. (2006). No evidence that experimental manipulation
of sexual conflict drives premating reproductive isolation in Drosophila
melanogaster. Journal of Evolutionary Biology, 19, 1033–1039.
 7

A Nest of Vipers: Conflict and

Cooperation in Families
Michelle Pellissier Scott

How sharper than a serpent’s tooth it is

To have a thankless child!
—Shakespeare, King Lear
Happy families are all alike; every unhappy family is unhappy in its
own way.
—Tolstoy, Anna Karenina

INTRODUCTION
While it may be true that happy human families are all alike, the diversity of
happy families in the nonhuman world is substantial. Evolution has sculpted
a wide variety of solutions for launching the next generation. In many species,
parental behavior may be manifested only in the choice of a place to leave the
eggs where they might have the best chance of survival. In addition, eggs may
be camouflaged or infused with a distasteful substance to protect them from
predators. Beyond that, there is a wide range of types of parental care: incuba-
tion and nest guarding, live birth, and female and/or male care of offspring.
Each increase in care places the emphasis on rearing fewer, higher-quality
young.
196 ANIMAL BEHAVIOR

Once obligate parental care beyond leaving sheltered eggs to fend for them-
selves has evolved, the trail gets twisted. Mostly family members have each
other’s best interests at heart, but there are basic conflicts of interests between
parents, between parents and offspring, and among offspring. There may be
an evolutionary benefit for fathers who desert; offspring may want more re-
sources than parents are prepared to give; and each offspring would prefer to
be first in its parents’ heart. There are questions of the optimal number of
young that will result in the greatest number of reproducing heirs, and there
are some dark ways parents have to achieve their goal.
This chapter will put parental behaviors in the perspective of natural selection.
Traits have evolved because individuals who acted in that way in a given environ-
ment, with given resources, competitors, and predators, left more young. Topics
will include an evaluation of the circumstances under which providing extended
care should be beneficial to the parent, why females are usually the caregivers in
some taxa but biparental care is common in others, and why some offspring
forgo their own reproduction to help parents rear the next generation. This chap-
ter will also explore what determines the optimal clutch or brood size and how
that is achieved, discuss how offspring communicate their needs and parents
respond, and finally conclude with the nature of the conflict between parents
and offspring as members seek to optimize their own future reproduction. The
diversity and complexity of family life across the animal kingdom not only pro-
vides interesting stories but also offers us a way to further appreciate the workings
of natural selection and adaptation.

WHY PROVIDE PARENTAL CARE?

Parental care is any form of behavior that increases the survival and ulti-
mate reproductive success of offspring. In the most basic form of care, females
provide a nutrient-rich egg. However, males in some species may contribute to
this type of care as well in the form of a food offering during courtship,
common in birds, or in the form of a nutritious mass packaged with the
sperm, commonly seen in insects. This nuptial gift, which can be substantial
(the record may be a sperm-packet “gift” weighing 25 percent of the male
katydid’s body mass), can offset some of the energetic cost to the female and
may increase the number or mass of her eggs (Gwynne, 1986, 1988a,
1988b). Nest construction, which in birds is frequently a joint effort, is
another form of prefertilization care. After fertilization, the forms and benefits
of parental care are usually obvious: eggs and live young need to be protected
and fed. Even after young are nutritionally independent, parents may facilitate
social interactions. In spotted hyenas (Crocuta crocuta) and some Old World
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 197

primates, for example, daughters inherit their mothers’ rank (Eng et al., 2000;
Bergstrom & Fedigan, 2010), which is facilitated through maternal interven-
tion, aggression towards low-ranking individuals, and coalition support. These
high-ranking daughters generally have better access to food and can better pro-
tect their young than low-ranking females.
There is a trade-off between the amount of parental care provided and the
number of young that can be reared, and there is a continuum from leaving
many young to fend for themselves at one extreme and rearing fewer, more
competitive young at the other. Mammals, with their reproductive mode of
gestation, lactation, and long period of dependence, represent the latter
extreme. But why don’t all animals provide so well for their young? This question
can be addressed by examining the costs and benefits of providing care.
Parental investment, any behavior that increases offspring fitness but
decreases the parents’ ability to invest in other offspring, is a concept that lends
itself better to testing hypotheses, because it incorporates costs, than does
parental care. In a cost/benefit analysis, the potential outcomes of the differ-
ent options are scored in terms of fitness, which refers specifically to reproduc-
tive success. It describes the ability both to survive and to reproduce. This
approach allows for a comparison of mothers and fathers using the same
currency (current and future offspring). It is the costs and benefits of any
behavior that are the grist for the mill of natural selection.
Specific costs and benefits have very much to do with life history (life span,
reproductive schedule, etc.) of each animal. Large animals with a long life
would be predicted to have fewer offspring each time they breed and to pro-
vide more care to each. They might also be predicted to be more conservative
(risk averse) when they are younger and have greater reproductive potential in
the future; older individuals might be more risk prone and willing to pay
higher energetic costs for late-born young. Jane Goodall’s chimpanzee “Flo”
was a wonderful mother and reared many successful young, but she overpro-
tected and overindulged her last, to his detriment (Goodall, 1971). In addition
to the influence of body size and life span, life history traits evolve in response
to specific environmental characteristics, such as predictability. The effect of
environmental predictability on parental behaviors is nicely illustrated with a
comparison of house mice (Mus musculus) and the similarly sized marsupial
mouse, Antechinus stuartii. House mice can have five to 10 litters of six to
eight young each in a year. Juvenile mortality is high. Females are sexually
mature in six weeks, and males are mature in eight weeks. In the wild, they
can, but usually do not, live for a year. Antechinus (Figure 7.1), on the other
hand, have a single, highly synchronized, short breeding season each year,
and within a day of female ovulation all males in the population die of
198 ANIMAL BEHAVIOR

Figure 7.1. Antechinus stuartii mother with youngsters that are almost old enough to be
left in the nest. (Photo by Diana Fisher and Andras Kazei)

stress-related symptoms. Females have six to ten young, and juvenile mortality is
lower than that of mice. There is a period of four months of gestation and lacta-
tion, and when they are weaned, the combined mass of the litter is three and a
half times that of the mother. Mice are opportunistic breeders and live in a some-
what unpredictable environment, so females breed often and fast before condi-
tions change. Antechinus lives in a predictable environment; females make a
very large energetic investment in young, and males put all their energy into a
single reproductive event as the probability of living until the next year would
have been remote. Indeed, males that put the most energy into finding mating
partners are the most successful (Braithwaite, 1979; Scott & Tan, 1985).
Birds and mammals are striking examples of parental care; offspring would
not survive without a huge expenditure of energy by the mother and, for most
birds, by the father as well. Most other taxa contain some species for which
parental care is obligatory (Clutton-Brock, 1991). Although they might not
spring to mind, some insects also provide extensive care to young. Male and
female burying beetles (Nichrophorus spp.) cooperate to bury small vertebrate
carcasses that will become food for their young (Figure 7.2). Parents remain
in the burial chamber provisioning and guarding young. Carcasses (a bonanza
resource) are rare, and finding one is unpredictable. In this case, this environ-
mental factor has selected for life history traits and behavior on the far end of
the continuum of extensive care. Once they have the opportunity (a carcass),
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 199

Figure 7.2. Burying beetle (Nichrophorus orbicollis) female feeding four-day-old larvae.
(Photo by M. W. Moffett)

females spend more than half of their reproductive life spans rearing a single
brood. Males help because two adults can successfully defend the resource
and brood from infanticidal intruders who seek to rear their own young on
this rare resource (Scott, 1990; Trumbo, 1991).
Burying beetles may be an example of particularly elaborate biparental care
in an insect, but parental care of one sort or another is widespread and diverse
(Table 7.1). It is thought to be confined to species for which eggs or young are
clumped in time and space, which in turn is dependent on the temporal and
spatial distribution of resources. The dispersion, quality, and persistence of re-
sources influence how sedentary or nomadic a species is and what competi-
tors, predators and parasites it must deal with. These in turn determine if
parental care can be an effective strategy, for instance when there is something
that parents are able to do to increase the survival of young. Because burying
beetles use a small, discrete resource, behaviors to protect it and the young
have evolved; and with burial and extended care, a suite of adaptations
(behavioral, physiological, and anatomical) coevolved from characteristics of
nonparental ancestors (Trumbo, 1996). For example, once burying beetle lar-
vae were protected underground by parents, they no longer had to be heavily
200 ANIMAL BEHAVIOR

Table 7.1. Types of parental care by insects and numbers of orders in which parents
exhibit the behaviors listed.
Maternal Paternal Biparental
Behavior care care care

Eggs covered, then abandoned 6 0 0

Parents remain and guard eggs 12 2 4
Parents remain and care for young 12 2 5
Females extend development internally 10 – –
and give birth to live young
Males provide nutrition with sperm – 9 –

protected by body armor as other (nonparental) carrion beetle larvae are.

Parents feed from the carcass, as do nonparental carrion beetles, but they must
also feed their young with predigested regurgitant.

WHO PROVIDES CARE?

Care can be provided by females, males, or both. Because this trait most likely
evolved from a condition of no care, the benefits of increased survival and quality
are assumed to outweigh the costs of providing the care. These costs may differ
for males and females, even for the same behavior. For example, pair-bonded
male and female birds may both bring food to the brood, but the cost to future
fitness may be greater for the male if he is giving up the opportunity to search
for and mate with another female who will rear his young. The female rarely
has the opportunity to find a care-giving male to rear her young. On the other
hand, if the mother ate the food she collected and left the father to feed the
chicks, she would have more energy to produce more eggs; the energy gain for
the same behavior of the father would have less direct fitness consequences.
Each offspring has one mother and one father; thus the benefits from rear-
ing it are shared equally by the mother and father regardless of which provides
the care. However, if one parent can do the job, either parent would gain by
leaving the other to do it. We can therefore expect conflict between the sexes
over the amount or duration of care the other sex provides. This conflict drives
an unusual mating system in the Eurasian penduline tit (Remiz pendulinus).
These small birds have very short pair bonds lasting only up to a week, and
both sexes practice serial polygamy. In this species one parent alone can incu-
bate and rear the brood, but 30 percent of the clutches are deserted. Both sexes
gain if they desert first and leave the other to do it—provided they can find an
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 201

unmated and willing new partner. Desertion by the male is costly to the
female because she can either stay, which take substantial time and energy,
or she can desert the eggs and leave them to die. If the female deserts, leaving
the father to incubate, she can produce a new clutch (Szentirmai et al., 2007).
Out-and-out desertion is extreme; various factors can influence the level at
which males and females work. When one parent deserts or is experimentally
removed, the remaining parent usually increases its workload, if it is able.
Burying beetle females do not increase their level of care when their mates
are removed or experimentally handicapped with extra weights, whereas males
do work harder under these circumstances. This suggests that females already
work as hard as they can; if need be, males can step up, but they usually work
less hard than their mate (Seizi & Masahiro, 2009). Parents may also adjust
their level of care based on the perceived quality of their mate. For example,
female zebra finches (Taeniopygia guttata) paired with attractive males (in this
case, males with orange or red leg bands) feed their chicks more than those
with unattractive mates (those with green or blue leg bands) (Burley, 1986).
Because females produce a comparatively large, nutrient-rich egg relative to
the male’s small, mobile sperm, their lifetime reproductive success depends
mostly on their energy intake, whereas a male’s lifetime reproductive success
depends more on the number of matings he can get. In some species the con-
sequence of this is extreme; a few males in the population mate with many
females and most mate with none. In this case, when the variance in male
success is greater than the variance in female success (i.e., most females have
closer to the average number of young in their lifetimes), there is strong selec-
tion (called sexual selection) for competition among males for access to recep-
tive females. This is not to say that females do not also undergo sexual
selection. Females may compete for high-quality males, territory, or social sta-
tus, which will increase their fitness. In some species like the polyandrous
jaçana or the sex-role-reversed spotted hyena, sexual selection is stronger on
females than on males (Clutton-Brock, 2007). But generally the greater com-
petition among males pushes for secondary sexual traits such as weapons and
ornaments. This potential for some males to do better than others in the pop-
ulation such that only a subset breed also selects for exclusive female parental
care because the costs for the breeding male to provide care is greater than that
for females (Kokko & Jennions, 2003).
There are additional reasons why female care has evolved more often than
male care. Females can be more certain that they are the mothers of their
young than males can be that they are the fathers. If females mate with multi-
ple males before eggs are fertilized, or, as with some fish, if a nonparental male
sneaks in and releases sperm just as the nest owner is spawning, males are less
202 ANIMAL BEHAVIOR

certain of paternity and benefits of parental care are devalued, theoretically.

Extra-pair copulations are quite common with pair-bonded birds, but this
does not necessarily devalue paternal care. On average, males in a population
have equal success cuckolding other males; thus a male may be caring for step
offspring while his are being cared for by another male.
Reduced parental effort rather than desertion is a more common effect of
reduced confidence in paternity. Results from theoretical modeling and experi-
ments examining the effect of paternity on parental effort are mixed (Sheldon,
2002). When either male or female collared flycatchers (Ficedula albicollis) were
temporarily removed when the female was supposed to be fertile, males seemed
to use that as an indication of cuckoldry and reduced paternity, and they did
reduce their share of provisioning the youngsters (Sheldon, 2002). For this
response to evolve, we assume that there must be a benefit in future fitness for
the males; otherwise, they would just go ahead and feed these potential step off-
spring. These male flycatchers had the same probability of survival if they
worked harder or less hard, but those that reduced their care had larger forehead
patches the following year. This is a sexually selected trait, and those with larger
patches sired a larger proportion of the offspring in their nests.
The association hypothesis (or order-of-gamete-release hypothesis)
suggests another factor that makes maternal care more likely than paternal
care (Gross & Shine, 1981). With internal fertilization, males release their
gametes first, and this puts the female in a cruel bind. The male is free to des-
ert immediately, and if he has, the female has the choice of providing care
alone or abandoning the breeding attempt (assuming that some care is neces-
sary for the survival of the young).
Male care and biparental care are common in some groups, especially in fish
and frogs. The association hypothesis offers an explanation for the frequency
of male care by nest-building fish. Females are induced to release their eggs
in the nest that males guard and are more attracted if there are already eggs
in the nest, advertising what attentive fathers the owners are. (In some species,
males are even known to steal eggs from the nests of other males [Rohwer,
1978].) Males release sperm after females spawn, so the female is free to desert
first, but the stronger force selecting for paternal care is that the cost to the
male of remaining is relatively low compared to that of the female. He remains
on his territory after the female leaves, guarding the eggs from predators and
attracting additional females. Furthermore, care is sharable, meaning that it
does not increase as more females deposit eggs in his nest.
A second reason for the commonness of paternal care in fish is that growth
is indeterminate. As females grow their fecundity increases; thus, if caring
reduces growth, they suffer a high cost to future reproduction. In species for
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 203

which size in males does not have a strong effect on fitness, the cost of care,
and consequential reduced growth, is less for them than it is for females.
Models for cost-benefit analyses have been developed using game theory
borrowed from economics. To provide care or not is seen as a trade-off.
The major factors input into these models are offspring survival with zero,
one, or two parents; the availability of additional mates if the male or female
deserts; certainty of parentage; and the increase in fecundity if the female
deserts. The potential payoff is calculated for each option, given that the other
parent will or will not provide care. Thus if offspring survival is much better
with two parents and few females are available as additional mates, males are
expected to stay and help. Imagine the case of emperor penguins (Aptenodytes
forsteri). Their harsh breeding environment has selected for biparental care.
Without the father returning to relieve the mother, the chick would have no
chance of survival. Furthermore, most females are already paired.

EXTENDED FAMILIES
The African saying that it takes a village to raise a child may be true, at least
as a metaphor, for some nonhuman species as well. Pair-bonded parents are
assisted by older siblings or even less related group members in numerous spe-
cies of birds and mammals. Cooperative breeding is good for the extended
family because groups can be more efficient than individuals through division
of labor and specialization, and there is opportunity for social learning. But
why should grown offspring give up their own chance to reproduce to help
their parents? Benefits are varied and complicated. Especially if the new brood
is comprised of full siblings, they are helping to promote the success of their
genes. (This type of selection is called kin selection.) There is a net fitness gain
if the number of young that a first-time breeder can expect is low and its assis-
tance to parents results in much better survival of the new brood. Although
the close relatedness of parents and helpers certainly facilitates the evolution
of helping, it is perhaps not the most important factor. In many species, non-
related individuals help, and in various ways this increases their personal repro-
ductive success. Helping may be the “dues” they pay to be allowed to stay.
Many recent studies show that there can be direct benefits to delaying disper-
sal from the family territory, even if they do not help with the new young. By
staying, they may increase their survival and chance of gaining a territory and
breeding in the future. They may also gain skills that increase their future
fitness (Clutton-Brock, 2002; Komdeur & Ekman, 2010).
Nothing matches social insects for extended families! Individuals in a family
of ants, termites, and many species of bees and wasps do not reproduce
204 ANIMAL BEHAVIOR

themselves but help the queen to rear their younger siblings. Honey bee (Apis
mellifera) colonies, for example, consist of tens of thousands of individuals,
and the workers do all the housework and care unstintingly for the eggs and
larvae before they become the hive’s foragers when they are about two weeks
old. Sterile worker ants are often morphologically specialized for their specific
and clearly defined roles. Leafcutter ants (Atta spp.) have highly specialized
castes. The nurse ants are relatively small, and their job, as their name implies,
is to care for the brood. They keep them clean and feed the begging larvae the
fruits from their fungal garden and even with sterile eggs that they produce for
just for nutrition. Armored soldier termites defend the family with weaponry
of huge pincers or nozzle-like heads that spray caustic substances on enemies.
The soft-bodied workers perform the mundane household duties and tend to
the youngsters’ needs. These altruistic workers in most, but not all, cases are
helping to rear closely related kin and thus pass their genes on indirectly.
However, selection at the colony level has also been important in the evolution
of this division of labor. Larger colonies that are more efficient and have grown
fast are better competitors and more likely to ultimately give rise to daughter
colonies (Wilson & Wilson, 2007). If there is a persistent food supply and
individuals have built a defensible nest, the groundwork is laid for social living.
Division of labor with some individuals specialized for reproduction and
others for caring for the young can then follow.

FAMILY PLANNING
Just as the amount of care is a trade-off between current and future off-
spring, the number of young to have and when to have them is a similar
trade-off. Large, long-lived species take a conservative strategy and save energy
for future young. Even shorter-lived species face the same “decision” of how
many young to have and how much care to provide. Parents are typically opti-
mistic or at least make room for the possibility of the best outcome. Nonethe-
less, mothers do not produce the maximum number of young that they could.
Birds tend to have an optimal clutch size (Lack, 1947; see Williams, 1966).
Mothers that lay too many eggs are overly optimistic and bring fewer young
to independence; those that lay too few may rear all but produce fewer than
the female that has chosen the intermediate. At least, that is what common
sense and theory would predict. However, this has been difficult to demon-
strate with experimental manipulation of clutches because parents differ in
quality and much is dependent on the availability of resources at the time.
Trade-offs are ultimately mediated by a number of factors including both
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 205

chance events like predation and storms, biotic factors like parasite and prey
abundance, and competition from conspecifics.
Blue tits (Parus caeruleus) are an exceptionally well-studied bird. The effects
of reproductive effort (clutch size) on offspring quantity and quality and on
parental future reproduction have been studied in populations from Scandina-
via to Corsica. An experimental manipulation of clutches, in which eggs were
removed or added to random nests, in an English population supported the
hypothesis that individuals optimize their clutch size. The number of young
surviving was best in nests that were not manipulated (Pettifor, 1993a). This
same study also found no overall negative effect of brood size on the survival
or future reproduction of the parents (Pettifor, 1993b), although there did
seem to be a significant negative effect of brood size on survival of males or
females in some individual years. The key to the variability between years
was most likely due to differences in the caterpillar crop (the main food source
for parental tits). A population of blue tits in Corsica differs significantly from
more northern ones: there is less food, less water, and higher temperatures,
and birds suffer a higher parasite load (Blondel et al., 1998). These tits have
smaller clutches and breed later than the northern ones. In this study, clutch
manipulation did not significantly affect the number of young reared, but it
did affect their quality. Young from larger broods were smaller and lighter,
and females raised in these larger broods bred later themselves and laid fewer
eggs than those from control or reduced clutches. As with the previously men-
tioned study, there was no consistent effect of clutch manipulation on adult
survival or reproduction the following year. However, in this study there was
a big effect of year-to-year variation in environmental conditions, especially
in caterpillar abundance. Not only in some years was there better overall fledg-
ling success, but also the effect of the manipulation differed. In one year,
clutches reduced by two eggs did relatively well, but those increased by two
eggs did exceptionally poorly, compared to the same manipulations in other
years (Blondel et al., 1998).
In addition to the number of young to produce, the frequency at which to
produce them and even the sex of the young are all components of family
planning. Long-lived species often do not have young every year. Large mam-
mals, with their very high costs of reproduction, have young infrequently.
African elephant females (Loxodonta africana) give birth every four to nine
years, but humpback whales (Megaptera novaeangliae), in spite of their large
size and slow growth, reproduce on average every other year. It is sometimes
the case for mammals that the sex of one offspring affects future offspring.
Rearing a larger male to independence is metabolically more costly than rear-
ing a smaller female to independence. Female red deer (Cervus elaphus) that
206 ANIMAL BEHAVIOR

have a son are more likely not to breed the following year than those that have a
daughter (Clutton-Brock et al., 1982). The effect of raising sons is even more
subtle in humans. Data from Lutheran church records of Finnish farmers and
fishermen in the eighteenth and nineteenth centuries show that having a son
takes more out of the mother than having a daughter, and succeeding offspring
have a lower birth weight. Furthermore, surviving children born after a son
were less likely to have offspring of their own than those born after a daughter
(Rickard et al., 2007), presumably because of their poor start in life.

HOW PARENTS PLAY FAVORITES

The future is uncertain, and while parents of some species produce what
nature tells them will give them the best outcome of healthy, viable offspring,
others have more alarming practices for achieving the optimal brood or clutch
size. They rely on infanticide and siblicide! The optimal clutch size for bury-
ing beetles depends entirely on the size of the carcass they have buried. For
example, Nicrophorus orbicollis can successfully rear young on a tiny 10 g
mouse or a small 60 g chipmunk, but only a few larvae survive on the mouse,
and 50 or so can be raised on the chipmunk. Because she has plenty of food to
convert to eggs and because some may not hatch, the female overproduces.
There seems to be a minimum of about 25 eggs that she oviposits (Scott,
1997). She and the male have assessed the carcass size by walking around it
and measuring (Trumbo & Fernandez, 1995). With this information both
male and female kill excess young as eggs hatch and larvae arrive on the carcass
(Bartlett, 1987; Trumbo, 1990). Although we might imagine a cartoon image
of parents counting off arriving youngsters and reaching the cut-off point after
which they eat the rest, it has been hard to assess the basis of their mathemati-
cal skills. We do know that parents are hormonally and behaviorally sensitive
to the number of begging larvae (Panaitof et al., 2004); perhaps the frequency
of interactions provides all the information they need.
Antechinus, the marsupial mice with the annual male die-off, also produce
more offspring than can be accommodated. They give birth to such underde-
veloped young that the cost of overproduction is low. Females have a limited
number of teats (that differs by population, perhaps depending on resource
availability), and the trek from the genital opening to the pouch is treacherous.
Because each young fastens onto its own teat and stays there for six weeks or
so, once the teats are occupied, no more young can survive. Mothers may
ignore or simply eat the extras (Lee & Cockburn, 1985).
It may seem both gruesome and wasteful that parents should resort to can-
nibalism to achieve the best brood size, but it is not that uncommon,
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 207

occurring in mammals, birds, fish, and reptiles as well as insects (Klug & Bon-
sall, 2007). Paternal care is common in fish. While often this consists of males
guarding a nest in which the female has left her eggs, in some cases the male’s
mouth serves as the “nest.” Mouth-brooding male cardinalfish (Pterapogon
kauderni) pick up the egg mass into their mouths immediately after spawning.
As long as things go well, eggs stay there protected from predators and well
aerated until embryos hatch. But this comes with a cost to the fathers, as they
must fast and, as the breeding season progresses, they lose condition. Partial
brood cannibalism is common, and the entire brood disappears into the
father’s stomach not infrequently. This filial cannibalism by male cardinal-
fish becomes much more prevalent later in the season as the father’s body con-
dition decreases (Okuda & Yanagisawa, 1996) and tends to happen early in
the mouth-brood phase as young eggs have greater nutritional value and less
reproductive value than younger eggs.
An energy-based explanation for filial cannibalism in teleost fish seems to
be valid in most cases (Manica, 2007), but other factors, such as brood size
or age, availability of additional mates, and the probability of paternity can
also influence this “decision.” Male threespined sticklebacks (Gasterosteus acu-
leatus), a fish that makes and guards a nest in which females deposit their eggs,
suffer from “sneaky” males that wiggle into nest of the territorial male and
deposit sperm that fertilize some of the eggs. Thus the nest owner cannot be
100 percent sure that he is the father of all the young. Clutches containing a
high proportion of foreign eggs are much more likely to be entirely cannibal-
ized. But how can the male make an informed judgment? In an experiment in
which from 0 to 100 percent of one male’s eggs were exchanged with those
of another male’s, all the eggs were destroyed and presumably eaten signifi-
cantly more often when the male was caring for fewer than 50 percent of his
own young. Moreover, the risk of total brood cannibalism was greater early
in the breeding season, but only by males with many of their own eggs; males
with few of their own eggs cannibalized the brood at a high rate throughout
the season. Some males only cannibalized part of the brood, and in these cases,
the greater the proportion of their own eggs that were in the nest, the more
foreign eggs were consumed (Mehlis et al., 2010). What cues can they use to rec-
ognize alien eggs? This experiment ruled out all but cues from the eggs them-
selves (e.g., no rival males were ever present), and olfactory cues associated
with the eggs are the most likely.
While parents sometimes do the dirty work to eliminate excess offspring
themselves, in some cases they leave it to the siblings. Sand tiger sharks
(Carcharias taurus) kill and eat brothers and sisters in utero (Gilmore et al.,
1983). Embryo pronghorn antelope (Antilocapra americana) have a spear of
208 ANIMAL BEHAVIOR

Figure 7.3. A Nazca booby mother (Sula granti) looks on while the first-born chick evicts
the second born from the nest, leaving it to certain death. (Photo by Dave Anderson)

dead tissue that they use to kill their twins in the narrow uterus (Forbes,
2005). Workers of many social insects eat eggs oviposited by other workers
because they are less related to these future offspring that they are to those of
the queen. But siblicide is best studied, and is most poignant, in birds, espe-
cially in large sea-going species that normally have small clutches. Nazca
booby (Sula granti) mothers lay two eggs but always raise only one. The older
and larger chick forces its weaker sibling out from under its mother, where it
dies a pitiful death in the beating sun (Humphries et al., 2006).
Leaving one offspring to eliminate the “extras” is a very practical evolution-
ary solution. Parents face multiple uncertainties: critical resources may vary
such that the parent cannot predict the optimal brood size, or some offspring
may not hatch, die prematurely, or be intrinsically flawed. To evaluate the
effects of these uncertainties, members of the brood are thought of as core or
marginal offspring. The core is the number of young that are normally raised
to independence, and the marginal offspring are those that make up the over-
production—those that might make it in the best of times (Mock & Forbes,
1995). The marginal member(s) of the brood are usually handicapped, often
by being hatched later, which make them easier to be killed. In Nazca boobies
siblicide is obligate (Figure 7.3)—99.95 percent of broods fledged only one
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 209

chick, and the most common cause of death was the action of the other chick
(Humphries et al., 2006); in other species, the blue-footed booby (Sula
nebouxii) for example, siblicide is facultative.
There are several hypotheses that might predict why parents overproduce
and for which species siblicide is facultative and which obligate (Mock &
Forbes, 1995). The resource-tracking hypothesis suggests that parents pro-
duce the number of young that can be raised in a good year and are prepared
to sacrifice them if resources turn out to be insufficient. Hatching asynchrony
in red-winged blackbirds (Agelaius phoeniceus) creates core and marginal off-
spring, and in this way they keep the costs of sacrificing marginal offspring
lower by creating weaker brood members. In an experiment in which the core
or marginal brood was enlarged, the growth of the core youngsters was not
affected by additional marginal ones, but alteration of the core brood signifi-
cantly reduced the growth of the marginal ones (Forbes & Glassy, 2000).
Parents feed core offspring what they need to develop properly, and marginal
ones just suffer more with extra core. Similarly, cattle egrets (Bubulcus ibis)
produce two core offspring and one marginal one, the consequence of being
last hatched. As long as there are three young in the nest the core chicks, espe-
cially the second hatched, attack the third, but it is the nutritional condition of
this third that determines the outcome; if it is weak and malnourished already,
it will succumb, but in a good year it becomes strong enough to survive (Mock
& Lamey, 1991).
A second hypothesis that might explain the overproduction and consequential
siblicide has been called the insurance hypothesis. This predicts that the junior
chick would be killed even if food were plentiful, but parents produce a back-
up chick in case something happens to the first. Whereas Nazca boobies virtually
never raise more than one chick, their cousin the blue-footed booby often raises
two or more. Nazca booby parents forage farther from land than do the blue-
footed parents. If these longer commutes limit the amount of food that can be
delivered, the resource-tracking hypothesis would be supported. However, Nazca
and blue-footed boobies have similar foraging efficiency, and both could fledge
two chicks (Anderson & Ricklefs, 1992), but the hatching success of the Nazca
booby is much lower than that of the blue-footed booby, apparently due to
exceptionally high infertility or early embryonic death (Anderson, 1990). On
average, the benefit of having the back-up chick is greater than the cost of pro-
ducing the wasteful second youngster that will just be killed. Thus fatal sibling
rivalry is in the best interests of the parents (as well of as the core chicks), and
their lack of interference is expected.
There is a third hypothesis for the evolution of siblicide, and that is as a
back-up food supply for the dominant offspring (Mock & Forbes, 1995).
210 ANIMAL BEHAVIOR

Parents thus convert relatively low-cost nutrients (marginal young) for emer-
gency rations for the more-likely-to-succeed youngsters. This probably
accounts for the sand tiger sharks’ siblicidal acts, but birds and mammals sel-
dom eat their dead siblings. Usually siblicide for nutrients is not specifically
orchestrated by the parents as it is when it occurs for insurance or resource-
tracking reasons. By practicing siblicide, the individual is eliminating some
of its own genes that could pass to the next generation; thus cannibalism for
the food value should be practiced on nonkin. The cannibalistic morph of
spadefoot toads (Spea multiplicata) is characterized by a large head and jaw
muscles. The development of this morphology is environmentally cued by
the type of food that is available when the cannibals have an advantage over
the omnivore morph. In fact, the cannibalistic toads show some ability to rec-
ognize siblings and refrain from eating them; the omnivores have no such skill
in recognizing kin (Pfenning, 1999).

“JUST TELL ME WHAT YOU NEED”

Family dynamics include complicated trade-offs between family members
and elaborate communication signals; getting and giving food is the central
arena on which these dynamics are played out. Human babies, nestling birds,
and even insect larvae are far from the helpless little creatures they may seem.
They can run their parents ragged with demands to get more for themselves
and strive to seem more needy than their siblings. In most species where
parents provision, youngsters beg. Burying beetle larvae, like baby birds, are
relatively quiet when no parents are around. But as soon as the larvae detect
a parent or the nest shakes and the parent bird casts a shadow, the young
beg energetically. Young beetles rise up, clamor over parental feet, and push
their heads into the mouth of the parent (Rauter & Moore, 1999); nestlings
raise their necks, extend their legs, open their beaks, call loudly, and jostle
for position to be the closest mouth to receive the caterpillar. Begging behavior
has even been reported in a frog. Taiwanese tree frog (Chirixalus eiffingeri) tad-
poles immediately aggregate around the mother when she returns to the nest.
They stiffen their tails, vibrate vigorously, and nip at her skin around her geni-
tal opening and thighs. This induces her to lay unfertilized eggs for food,
which the tadpoles eagerly consume (Kam & Yang, 2002). How should parents
respond to begging? Do begging youngsters honestly communicate their needs (as
opposed to their wants), and do parents respond to differences in begging vigor by
feeding the most energetic first?
This extravagant and often loud begging is assumed to be metabolically
expensive and even dangerous if it attracts predators. Offspring pay this price
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 211

to outcompete their siblings for resources. Parents are equally related to each
of their young and might favor an equitable distribution, but each individual
values its own welfare twice as much as it values it siblings’ because they
share only 50 percent of the same genes at best. So as long as parents deliver
individual packets of food, it behooves each youngster to try to get a dispro-
portionately large share of it. This suggests that begging is a means to manipu-
late parents to deliver more resources than is optimal for the parent because it
is more than that particular individual needs. However, begging is a commu-
nication signal, and as such it is thought to have evolved because is confers a
benefit to both the sender and the receiver; that is, parental response plays a
role in shaping the begging signal. If the young differ in condition, parents
might benefit if they can tell which youngster is in the greatest need to prefer-
entially feed it because it would get a greater benefit than its better-fed siblings.
Thus parents gain if greater need is signaled by more vigorous begging. The
costly nature of begging suggests that it should be an honest signal of need
(Kilner & Johnstone, 1997); that is, if it did not come with a cost, each would
demand everything it could get, and parents would gain no information and
would ignore differences.
Many studies have shown that begging intensity is an honest indication of
need. For example, yellow-headed blackbird (Xanthocephalus xanthocephalus)
nestlings in an experimentally altered brood of four consistently begged longer
than those in a brood of three (Price, 1996) and begged longer if they were
smaller than their competitors or in poor condition. Nestlings that were
deprived of food for 30, 60, or 90 minutes begged harder the longer they
had gone without. Male yellow-headed blackbird youngsters grow faster and
must reach a greater weight before fledging than their sisters. Males begged
louder and longer and ate more than females for any hunger level (Price
et al. 1996). In this species at least, begging can communicate both long-
term (males need more and begged more) and short-term (longer deprivation
led to longer begging) need. Furthermore begging worked. When begging by
the yellow-headed blackbird nestling was enhanced by adding recorded vocal-
izations, parents doubled their feeding visits, suggesting that they do use the
information provided by begging (Price, 1998).
Not surprisingly, most studies on begging have been conducted on birds.
Studies on burying beetles have also supported the prediction that begging
can be an honest signal of need. Food-deprived larvae and smaller, younger
larvae begged longer than their better-fed and larger brood mates. Begging
increased with brood size up to a point, and the time spent feeding the larvae
by both parents together was strongly associated with the time spent begging
by each larva. When the hunger levels of individuals within the brood was
212 ANIMAL BEHAVIOR

experimentally manipulated, parents fed hungry larvae more than well-fed

ones, but this bias occurred because hungry larvae spent more time begging
(Smiseth & Moore, 2002, 2007, 2008).
Parental response to the honest signal of begging may not always be to feed
the hungriest, smallest, or neediest. Earwig (Forficula auricularia) mothers
regurgitate food individually to their nymphs, and chemical cues communi-
cate offspring need. Older offspring and those in better physiological condi-
tion have higher value to parents because the expected parental investment
needed to successfully rear them to independence is lower. In this species,
rather than preferentially feeding the neediest nymphs, mothers exposed to
the chemical cues of well-fed youngsters foraged longer and gave more food
to nymphs than mothers exposed to cues from poorly fed nymphs (Mas
et al., 2009).
In spite of the above examples, results of studies on the relation of begging
intensity and need and the effect on parental provisioning have been mixed.
Offspring begging signals are often complex, with multiple components (beg-
ging loudness and duration, stretch height, etc.). Older siblings in a brood that
has hatched asynchronously may be able to stretch higher and therefore get fed
more just because they block the access of other chicks. Older chicks can beg
harder just because they have more energy to spare. Lastly, the interpretation
of begging as an honest signal of need depends on begging being costly. Exper-
imental results have been very mixed, however. In some experiments, recorded
begging calls near artificial nest have attracted predators; in other studies, there
is no difference between attraction to begging calls and white noise. Some
studies have measured the energetic cost of begging, which ranges from only
a little more to about 25 percent more than the resting metabolic rate,
but usually there seems to be no cost in either growth or survival (Kilner &
Johnstone, 1997).

PARENT-OFFSPRING CONFLICT
Although most of the time parents “want” what is best for their offspring
and vice versa, each share only half of their genes with the other. In the eyes
of the offspring, this devalues parental sacrifices, and in the eyes of the parents,
future offspring could benefit from energy saved. Thus the level of care that
maximally benefits the young is more than the parents should give if they
maximize their lifetime reproductive success. Furthermore, offspring care
more about their own welfare that than of a sibling, future or current (Trivers,
1974; see Figure 7.4). The greater the discrepancy between offspring and
parental optima, the greater the conflict. Often conflict flares towards the
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 213

Figure 7.4. The basis of parent-offspring conflict. The benefit to the offspring increases
quickly with the first investment because a little does a lot of good, but as the young
get more, additional investment makes less of a difference. Another way to think of it
is that as the offspring grow and become more independent, benefit levels off as more
food or care does not increase survival or quality very much. The cost to the parent (in
terms of future reproduction) in providing this investment is expected to be more or
less linear, each bit of increase in effort being the same cost at a high or low level. Thus
for the parent the maximum net fitness gain (B–C) is at parental investment P.
However, from the offspring’s perspective, the cost to the parent is only half as much
because the parent and future offspring only share half of the same genes. Thus the
maximum net fitness gain (B–1/2C) is at parental investment level O. Between time
period or investment level P and O, offspring want more than parents want to give,
and there is a conflict. (Redrawn from Trivers, 1974)

end of the stage when young are dependent on parental feeding. Fledglings
harass their parents for more food when they are already fatter than their
parents, and young mammals can have temper tantrums when denied milk
during weaning.
One of the most startling examples of parent-offspring conflict has been
proposed to manifest itself during human pregnancy (Haig, 1993). Some
214 ANIMAL BEHAVIOR

pregnancies progress smoothly; baby and mother appear to be in sync, and the
baby is born at a reasonable size with minimal fuss. But more often than not,
the mother has endured months of pregnancy sickness and possibly danger-
ously high blood pressure and gestational diabetes. The latter two problems
can be attributed to the demands of the fetus to get more nourishment
through the placenta than is good for the mother. Preeclampsia, or danger-
ously high blood pressure, can benefit the fetus because it increases the blood
flow through the placenta, and more blood means more nutrients. This condi-
tion is common when the placental growth is poor and the placenta is
embedded shallowly in the uterus. It appears to be the fetus’s way of compen-
sating, but the mechanism is not yet known. In gestational diabetes, the fetus
secrets a hormone (human placental lactogen) that reduces the mother’s sensi-
tivity to insulin. The result is increased blood glucose levels that the fetus uses
to good advantage.
Offspring begging can be seen as manipulation to get from the parents
more than they want to give, or, if it is costly, it may be the mediator to resolve
the conflict; parents can know what they need, not just what they want. If
parents adjust the rate of feeding in response to begging and offspring adjust
their begging in response to the amount of food received, parent-offspring
conflict can also be understood as an evolutionary process between supply
and demand, each affecting the other. Thus the level of begging and the level
of feeding have some genetic basis and have evolved together. This relation-
ship can be explored with cross-fostering experiments in which some members
of one brood are exchanged with those of another, leaving a mix of genetic and
foster offspring in each brood. Some experiments, but not others, have shown a
positive relation between begging and feeding; that is, in families where offspring
are genetically inclined to beg more, parents are genetically inclined to feed more
(Smiseth et al., 2008). This experimental approach can often reveal if parents
ultimately control feeding or if offspring begging manipulates parents to the det-
riment of their future reproductive success. Species differ.

CONCLUSIONS—NEW HORIZONS
Through exploring parental behavior, this chapter has sought to elucidate
some of the many ways that family members cooperate or are in conflict with
other members. The discipline of behavioral ecology is well developed, and
hypotheses to predict the behavioral outcome of interactions have been
proposed; this has encouraged tests of these hypotheses in a wide diversity of
species from invertebrates to mammals. Much less is known about the mech-
anisms that regulate parental and offspring behavior. Arguably the most
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 215

exciting are hormonal and genetic mechanisms. Two examples follow for
which these mechanisms have been uncovered as a suggestion for how fruitful
further exploration should be.
Many birds lay eggs that hatch asynchronously. Later-hatched chicks have a
disadvantage, and this may or may not be in the best interest of the mother.
Mothers may vary the amount of the hormone testosterone that they put into
the eggs. The general consequences are that chicks from eggs with higher tes-
tosterone beg more vigorously. The consequences of increased begging on off-
spring fitness are complicated: It can have either a negative or positive effect on
growth and survival. Most likely this is related to available resources. In lean
years, the extra effort put into begging reduces growth, and in good years off-
spring get more food. One would predict that mothers would apply more tes-
tosterone to either level the playing field for late-born youngsters or to increase
the competitive asymmetry, depending on the desired outcome to raise
all chicks or to allow siblicide to proceed at a low cost to the older chick. How-
ever, there is not strong evidence for this prediction (Smiseth et al., 2011). We
need more information on the commonness of maternal hormonal manipula-
tion of offspring condition and on how endogenous offspring hormones
(rather than experimentally altered titers) affect begging and whether this
might be a counterweapon to help regulate their begging.
More and more of the genetic basis for behavior is being discovered each
month, and among the most interesting new research is the genetic basis for
monogamy and paternal care in mammals. Monogamy is rare in mammals,
and even for these species, individuals can be “unfaithful.” Rodent mating sys-
tems vary from monogamous to promiscuous. The monogamous prairie
vole’s (Microtus ochrogaster) social behavior is characterized by pair bonding,
biparental care, and increased defense by males, whereas in the promiscuous
meadow vole (Microtus pennsylvanicus) males seek out new females after each
mating and provide little assistance rearing the litter. The dramatic increase
in the receptors for the neurotransmitter vasopressin in the appropriate brain
centers facilitates fidelity and good parenting in the monogamous males. (This
brain center, the ventral pallidum, is generally involved with addiction and
pleasure perception.) The genetic basis for this increase in vasopressin recep-
tors has been identified as a stretch of repeats in the section of DNA that reg-
ulates the gene Avpr1a; the more repeats, the more receptors, the greater
sensitivity to vasopressin, and the stronger the partner preference. When the
extra bit of DNA was introduced into the genome of the promiscuous
meadow vole, these males provided more paternal care, retrieving and licking
the pups (Nair & Young, 2006). There is increasing evidence that this
“monogamy gene” may be linked to pair-bonding behavior in many mammal
216 ANIMAL BEHAVIOR

species, including humans (Walum et al., 2008). Although the genetic basis of
behavior will no doubt prove to be very complex, the more we know about the
underpinnings of the differences in social behavior, the more exciting the
revelations become and the more we can expand our understanding into
human behavior and the evolution of behavior in general.

ACKNOWLEDGMENTS
I would like to thank my colleagues, Jessica Bolker, David Berlinsky and
Adrienne Kovach, for years of good conversations and the National Science
Foundation and the U.S. Department of Agriculture for financial support.

REFERENCES AND SUGGESTED READING

Anderson, D. J. (1990). Evolution of obligate siblicide in boobies. 1. A test of the
insurance-egg hypothesis. American Naturalist, 135, 334–350.
Anderson D. J. & R. E. Ricklefs. (1992). Brood size and food provisioning in masked
and blue-footed boobies (Sula spp.). Ecology, 73, 1363–1374.
Bartlett, J. (1987). Filial cannibalism in burying beetles. Behavioral Ecology and Socio-
biology, 21, 179–183.
Bergstrom, M. L. & L. M. Fedigan. (2010). Dominance among female white-faced
capuchin monkeys (Cebus capucinus): Hierarchical linearity, nepotism,
strength and stability. Behaviour, 147, 899–931.
Blondel, J., M. Maistre, P. Perret, S. Hurtrez-Boussès, & M. M. Lambrechts. (1998).
Is the small clutch size of a Corsican blue tit population optimal? Oecologia,
117, 80–89.
Braithwaite, R. W. & A. K. Lee. (1979). A mammalian example of semelparity.
American Naturalist, 113, 151.
Burley, N. (1986). Sexual selection for aesthetic traits in species with biparental care.
American Naturalist, 127, 415–445.
Clutton-Brock, T. H. (1991). The Evolution of Parental Care. Princeton, NJ: Princeton
University Press.
Clutton-Brock, T. H. (2002). Breeding together: Kin selection and mutualism in
cooperative vertebrates. Science, 296, 69–72.
Clutton-Brock, T. H. (2007). Sexual defection in males and females. Science, 318,
1882–1885.
Clutton-Brock, T. H., F. E. Guinness, & S. D. Albon. (1982). Red Deer: Behavior
and Ecology of Two Sexes. Edinburgh, UK: Edinburgh University Press.
Eng, A. L., K. Esch, L. Smale, & K. Holekamp. (2000). Mechanisms of maternal
rank inheritance in the spotted hyena, Crocuta crocuta. Animal Behavior, 60,
323–332.
Forbes, S. (2005). A Natural History of Families. Princeton, NJ: Princeton University
Press.
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 217

Forbes, S. & B. Glassy. (2000). Asymmetric sibling rivalry and nestling growth in red-
winged blackbirds (Agelaius phoeniceus). Behavioral Ecology and Sociobiology,
48, 413–417.
Gilmore, R. G., J. W. Dodrill, & P. A. Linley. (1983). Reproduction and embryonic
development of the sand tiger shark, Odontaspis taurus (Rafinesque). Fishery
Bulletin, 81, 201–225.
Goodall, J. v. L. & H. van Lawick. (1971). In the Shadow of Man. New York:
Houghton Mifflin.
Gross, M. R. & R. Shine. (1981). Parental care and mode of fertilization in ectother-
mic vertebrates. Evolution, 35, 775–793.
Gwynne, D. T. (1986). Courtship feeding in katydids (Orthoptera: Tettigoniidae):
Investment in offspring or in obtaining fertilizations? American Naturalist,
128, 342–352.
Gwynne, D. T. (1988a). Courtship feeding and the fitness of female katydids
(Orthoptera: Tettigoniidae). Evolution, 42, 545–555.
Gwynne, D. T. (1988b). Courtship feeding in katydids benefits the mating male’s
offspring. Behavioral Ecology and Sociobiology, 23, 373–377.
Haig, D. (1993). Genetic conflicts in human pregnancy. Quarterly Review of Biology,
68, 495–532.
Humphries, C. A., V. D. Arevalo, K. N. Fischer, & D. J. Anderson. (2006). Contri-
butions of marginal offspring to reproductive success of Nazca booby (Sula
granti) parents: Tests of multiple hypotheses. Oecologia, 147, 379–390.
Kam, Y.-C. & H.-W. Yang. (2002). Female-offspring communication in a Taiwanese
tree frog, Chirixalus eiffingeri (Anura: Rhacophoridae). Animal Behaviour, 64,
881–886.
Kilner, R. & R. A. Johnstone. (1997). Begging the question: Are offspring solicitation
behaviours signals of need? Trends in Ecology and Evolution, 12, 11–15.
Klug, H. & M. B. Bonsall. (2007). When to care for, abandon, or eat your offspring:
The evolution of parental care and filial cannibalism. American Naturalist, 170,
886–901.
Kokko, H. & M. Jennions. (2003). It takes two to tango. Trends in Ecology and
Evolution, 18, 103–104.
Komdeur, J. & J. Ekman. (2010). Adaptation and constraints in the evolution of
delayed dispersal: Implications for cooperation. In T. Szekely, A. Moore, &
J. Komdeur (eds.), Social Behavior: Genes, Ecology and Evolution (pp. 306–327).
Cambridge, UK: Cambridge University Press.
Kvarnemo, C. (2010). Parental care. In D. F. Westneat & C. W. Fox (eds.), Evolu-
tionary Behavioral Ecology (pp. 451–467). Oxford, UK: Oxford University
Press.
Lack, D. (1947). The significance of clutch size. Ibis, 89, 302–352.
Lee, A. K. & A. Cockburn. (1985). Evolutionary Ecology of Marsupials. Cambridge,
UK: Cambridge University Press.
Manica, A. (2007). Filial cannibalism in teleost fish. Biological Reviews, 77,
261–277.
218 ANIMAL BEHAVIOR

Mas, F., K. F. Haynes, & M. Kölliker. (2009). A chemical signal of offspring quality
affects maternal care in a social insect. Proceedings of the Royal Society of London,
B, 276, 2847–2853.
Mehlis, M., T. C. M. Bakker, L. Engqvist, & J. G. Frommen. (2010). To eat or not
to eat: Egg-based assessment of paternity triggers fine-tuned decisions about
filial cannibalism. Proceedings of the Royal Society of London, B, 277, 2627–
2635.
Mock, D. W. & L. S. Forbes. (1995). The evolution of parental optimism. Trends in
Ecology and Evolution, 10, 130–133.
Mock, D. W. & T. C. Lamey. (1991). The role of brood size in regulating egret
sibling aggression. American Naturalist, 138, 1015–1026.
Nair, H. P. & L. J. Young. (2006). Vasopressin and pair-bond formation: Genes to
brain to behavior. Physiology, 21, 146–152.
Okuda, N. & Y. Yanagisawa. (1996). Filial cannibalism by mouthbrooding males of
the cardinal fish, Apogon doederleini, in relation to their physical condition.
Environmental Biology of Fishes, 45, 397–404.
Panaitof, S. C., M. P. Scott, & D. W. Borst. (2004). Plasticity in juvenile hormone in
male burying beetles during breeding: Physiological consequences of the loss of
a mate. Journal of Insect Physiology, 50, 715–724.
Pettifor, R. A. (1993a). Brood-manipulation experiments. I. The number of offspring
surviving per nest in blue tits (Parus caeruleus). Journal of Animal Ecology, 62,
131–144.
Pettifor, R. A. (1993b). Brood-manipulation experiments. II. A cost of reproduction
in blue tits (Parus caeruleus). Journal of Animal Ecology, 62, 145–159.
Pfennig, D. W. (1999). Cannibalistic tadpoles that pose the greatest threat to kin are
the most likely to discriminate kin. Proceedings of the Royal Society of London, B,
266, 57–61.
Price, K. (1996). Begging as competition for food in yellow-headed blackbirds. Auk,
113, 963–967.
Price, K. (1998). Benefits of begging for yellow-headed blackbird nestlings. Animal
Behaviour, 56, 571–577.
Price, K., H. Harvey, & R. Ydenberg. (1996). Begging tactics of nestling yellow-
headed blackbirds, Xanthocephalus, xanthocephalus, in relation to need. Animal
Behaviour, 51, 421–435.
Rauter, C. M. & A. J. Moore. (1999). Do honest signalling models of spring solicitation
apply to insects? Proceedings of the Royal Society of London, B, 266, 1691–1696.
Rickard, I. J., A. F. Russell, & V. Lummas. (2007). Producing sons reduces lifetime
reproductive success of subsequent offspring in pre-industrial Finns. Proceed-
ings of the Royal Society of London, B, 274, 2981–2988.
Rohwer, S. (1978). Parental cannibalism of offspring and egg raiding as a courtship
strategy. American Naturalist, 112, 429–440.
Scott, M. P. (1990). Brood guarding and the evolution of male parental care in bury-
ing beetles. Behavioural Ecology and Sociobiology, 26, 31–39.
A NEST OF VIPERS : CONFLICT AND COOPERATION IN FAMILIES 219

Scott, M. P. (1997). Reproductive dominance and differential ovicide in the commu-

nally breeding burying beetle Nicrophorus tomentosus. Behavioral Ecology and
Sociobiology, 40, 313–320.
Scott, M. P. & T. N. Tan. (1985). Radionuclide determination of male mating suc-
cess in natural populations. Behavioural Ecology and Sociobiology, 17, 29–33.
Seizi, S. & N. Masahiro. (2009). To compensate of not? Caring parents respond
differently to mate removal and mate handicapping in the burying beetle,
Nicrophorus quadripunctatus. Ethology, 115, 1–6.
Sheldon, B. C. (2002). Relating paternity to paternal care. Philosophical Transactions
of the Royal Society of London, B, 357, 341–350.
Smiseth, P. T. & A. J. Moore. (2002). Does resource availability affect offspring beg-
ging and parental provisioning in a partially begging species? Animal Behav-
iour, 63, 577–585.
Smiseth, P. T. & A. J. Moore. (2007). Signaling of hunger by senior and junior larvae
in asynchronous broods of a burying beetle. Animal Behaviour, 74, 699–705.
Smiseth, P. T. & A. J. Moore. (2008). Parental distribution of resources in relation to
larval hunger and size rank in the burying beetle Nicrophorus vespilloides. Ethol-
ogy, 114, 789–796.
Smiseth, P. T., M. P. Scott, & C. Andrews. (2011). Hormonal regulation of offspring
begging and mediation of parent-offspring conflict. Animal Behaviour, 81,
507–517.
Szentirmai, I., T. Székely, & J. Komdeur. (2007). Sexual conflict over care: Antago-
nistic effects of clutch desertion on reproductive success of male and female
penduline tits. Journal of Evolutionary Biology. 20, 1739–1744.
Trivers, R. L. (1974). Parent-offspring conflict. American Zoologist, 14, 249–264.
Trumbo, S. T. (1990). Reproductive benefits of infanticide in a biparental burying
beetle, Nicrophorus orbicollis. Behavioral Ecology and Sociobiology, 27, 269–273
Trumbo, S. T. (1991). Reproductive benefits and the duration of paternal care in a
biparental burying beetle, Nicrophorus orbicollis. Behaviour, 117, 82–105.
Trumbo, S. T. (1996). Parental care in invertebrates. In J. S. Rosenblatt & C. T.
Snowdon (eds.), Parental Care: Evolution, Mechanisms and Adaptive Signifi-
cance (pp. 3–51). New York: Academic Press.
Trumbo, S. T. & A. G. Fernandez. (1995). Regulation of brood size by male parents
and cues employed to assess resource size by burying beetles. Ethology, Ecology
and Evolution. 7, 313–322.
Walum, H., L. Westberg, S. Henningsson, et al. (2008). Genetic variation in the vas-
opressin receptor 1a gene (AVPR1A) associates with pair-bonding behavior in
humans. Proceedings of the National Academy of Sciences, USA, 105, 14153–
14156.
Williams, G. C. (1966). Natural selection, the cost of reproduction, and a refinement
of Lack’s principle. American Naturalist, 100, 687–690.
Wilson, D. S. & E. O. Wilson. (2007). Rethinking the theoretical foundation of soci-
obiology. Quarterly Review of Biology, 82, 327–348.
This page intentionally left blank
 8

Make Space Enough between You*:

Intraspecific Variation in Animal Spacing
Nancy G. Solomon and Brian Keane

INTRODUCTION
The ways in which animals occupy and share space have intrigued biologists
for decades, and there is an extensive body of literature describing the spacing
patterns of many animal species from diverse taxa (Allee, 1931; Brown &
Orians, 1970; Maher & Lott, 2000). Many of the initial investigations of
the mechanisms causing individuals to exhibit a particular type of spacing
behavior (proximate factors) and why a specific type of spacing system
evolved (ultimate factors) focused on interspecific differences in space use.
These patterns of space use reflect the average or most common pattern of
spacing among individuals of a specific species, and all individuals of the same
species were thought to display the same species-specific pattern of spatial
dispersion and space use (Brown, 1975).
Evidence began to appear suggesting that a phylogenetic explanation for
spacing patterns was not always correct (Brown, 1975). Different spacing pat-
terns were found within the same taxa, and similarities in spacing patterns
were noted in unrelated species living under similar environmental conditions.

*
Shakespeare, Antony and Cleopatra, Act 2, Scene 3
222 ANIMAL BEHAVIOR

If spacing patterns were not related to the evolutionary history of a particular

species, then it is likely that ecological factors in the present-day environment,
such as finding food resources or avoiding predators, which are critical for sur-
vival and reproduction, are influencing spacing patterns.
Dale Lott (1984) presented considerable evidence that variation in space
use occurs within and between numerous geographically distinct populations
of the same species. Individuals of many species exhibit flexibility in space
use depending on environmental (e.g., distribution of critical resources, preda-
tion, population density) and individual (e.g., genotype, age) conditions (for
more details, see Lott, 1991). There are now hundreds of publications docu-
menting intraspecific variation in spacing systems across a broad spectrum of
animal taxa (for reviews see Lott, 1991; Maher & Lott, 2000; Adams, 2001;
Nilsen et al., 2005). These studies demonstrate that intraspecific variation in
space use between geographically distinct populations as well as within popu-
lations is common instead of being rare instances of atypical behavior as they
were previously thought to be.
Intraspecific variation in spacing patterns is typically viewed as an adaptive
response to differences in environmental conditions (e.g., conspecifics, heter-
ospecifics, and habitat structure) or to different characteristics of individuals,
such as developmental stage or genotype. Because it appears that the same
types of proximate factors that influence interspecific spacing patterns also
affect intraspecific space use in a number of species, intraspecific variation in
spacing patterns provides an excellent tool for examining the determinants of
spacing patterns without the variation being confounded by phylogenetic
influences.
Progress in understanding the proximate factors that determine the man-
ner in which individuals distribute themselves in space will greatly enrich
studies in socioecology and population biology because the spacing behavior
of individuals can have consequences for social interactions, patterns of
reproduction, and population structure and dynamics (Figure 8.1). Further-
more, a better understanding of intraspecific variation in space use should
have practical implications for the management and conservation of animal
populations.
Rather than attempting to review the extensive body of literature on intra-
specific variation in animal spacing systems in this chapter, our aim here is to
highlight several main points regarding intraspecific variation in spacing using
examples from diverse taxa. We first summarize some of the common types of
animal spacing patterns seen in nature. Next, we review some of the empirical
support for the proximate mechanisms underlying particular patterns of spa-
tial dispersion. Finally, we discuss the consequences of intraspecific variation
MAKE SPACE ENOUGH BETWEEN YOU 223

Figure 8.1. Proximate factors (e.g., ecological or genetic) can result in intraspecific
variation in space use by differentially influencing the behavior of individuals. The
manner in which individuals are spatially distributed can have substantial conse-
quences for reproductive behavior as well as population structure and dynamics.

in spacing patterns on mating behavior, population-level processes such as

population growth, and processes at even a larger scale like conservation.

TYPES OF SPACING SYSTEMS

The spacing pattern of individuals is typically viewed as a continuum rang-
ing from discrete aggregates of individuals, to a lack of spatial pattern due to
random spacing, to very uniformly spaced individuals. Random spacing is very
224 ANIMAL BEHAVIOR

unusual among animals, and most patterns of spatial dispersion tend towards
either an aggregated or uniform distribution. Although animal space use is re-
ally a continuum, the spatial distribution of individuals relative to each other
can be categorized into a small number of commonly recognized discrete
spacing systems.

Random Distribution of Individuals

A random distribution of individuals is only seen in a few animals, such as
Poli’s stellate barnacle (Chthamalus stellatus) on rocky shores in southwestern
England and southern Europe (Yahner, 2012). The position of one individual
is independent of another individual; that is, individuals do not attract or repel
each other. This pattern of distribution is thought to occur in habitats where
resources are spatially and temporally consistent.

Uniform Distribution of Individuals

Uniform spacing is often due to direct agonistic interactions between indi-
viduals that repel each other (e.g., territorial behavior) but can also result
from a uniform distribution of resources within a habitat.
Territoriality—In 1920, British ornithologist Henry E. Howard, in his
book Territory in Bird Life, first described the concept of territoriality,
although naturalists as far back as the seventeenth century described behavior
encompassed by Howard’s definition. More recently Christine Maher and
Lott (1995) reviewed definitions of territoriality that have been applied to ver-
tebrates and suggested that a practical definition should include both behav-
ioral and ecological elements. They proposed that a useful definition of
territory would be a fixed area or space containing specific resources such as
food, nest sites, or mates from which one or more individuals (territory resi-
dents) exclude other members of the same species (conspecifics). Territorial
behavior may include overt aggressive behavior, visual displays, vocalizations,
or scent marking. These behaviors typically require an increased energetic
investment and can decrease survival through injury due to fighting or a
greater risk of predation while engaged in territorial behavior. Therefore, the
evolution and maintenance of territoriality requires that the average overall fit-
ness of territorial individuals exceed that of nonterritorial individuals.
There may be sexual dimorphism in territorial behavior. In some species
males are territorial, and in others only females defend exclusive areas. Except
when paternal care is critical for offspring growth, development, or survival,
the resource that males are expected to defend to maximize their fitness is
opposite-sex conspecifics. In contrast, females would maximize their fitness
MAKE SPACE ENOUGH BETWEEN YOU 225

by defending critical resources necessary to prevent infanticide and success-

fully rear offspring. For this reason, territorial males may often benefit from
accepting additional females into their territories. However, they should repel
other males from their territories because the latter represent a threat to the
resident males’ fitness (mate guarding hypothesis; Wittenberger & Tilson,
1980; Back et al., 2002). In contrast, females should secure an area encom-
passing sufficient critical resources, such as nest sites and food, against compet-
itors of both sexes (resource defense hypothesis; Ostfeld, 1985; Back et al.,
2002). Consistent with this latter hypothesis are observations that increased
food abundance often decreases the level of female territoriality (Ostfeld,
1985, 1986; Ylönen et al., 1988). Additionally, female territoriality can be a
mechanism to prevent infanticide through territorial behavior directed at
males or other females (Wolff, 1993).
Male-female pairs—Male-female pairs are found in a number of inverte-
brates and vertebrates (see Table 1 in Matthews, 2002; Martin et al., 2007).
In these species, a single adult male and female have extensively overlapping
home ranges from which they exclude all other conspecifics. This type of spac-
ing system can occur when a male and conspecific female spend most of their
time together and coordinate behavior (associated social monogamy). One
example of this type of social monogamy (a male and female living together)
is seen in snapping shrimp (genus Alpheus) in which a male-female pair jointly
defends a territory. Lauren Matthews (2002) suggested that advantages
accrued from territorial cooperation (sharing in defense and food) may be
one factor that may have selected for social monogamy because both partners
are likely to benefit by sharing the time and energetic costs of defending and
maintaining their territory. In some of these male-female associations, both
members of the pair also participate in parental care of offspring (Kleiman,
1977; Wittenberger & Tilson, 1980; Clutton-Brock, 1991; Woodroffe &
Vincent, 1994).
In another type of social monogamy, a male and female may share a
territory but forage and sleep alone (dispersed social monogamy). Dispersed
social monogamy is much more rare than associated social monogamy and is
found in some small mammals like the Cape porcupine (Hystrix africaeaustra-
lis; Corbet & van Aarde, 1996), tree shrews (e.g., Tupaia tana; Munshi-South
et al., 2007), nocturnal prosimians (e.g., fat-tailed dwarf lemur, Cheirogaleus
medius; Fietz, 1999), and the Japanese serow (Capricornis crispus; Kishimoto,
2003), a bovid. In these species, males and females are territorial and solitary
foragers. There is overlap between the territory of one adult male and one
female, allowing a male to defend resources necessary for himself and a mate.
Further, female dispersion in the Japanese serow typically prevents males from
226 ANIMAL BEHAVIOR

defending the home ranges of more than one female, although a small propor-
tion of males (19–36 percent) manage to do so (Kishimoto, 2003).
Many hypotheses have been proposed to explain the evolution and mainte-
nance of social monogamy. When females are widely dispersed, often thought
to be the result of resource dispersion or female-female aggression, each male
home range may only be able to overlap the home range of primarily one
female, resulting in social monogamy. Territories defended by one male and
one female may occur where males do not differ much in body mass so one
male cannot prevent other males from gaining access to females and the pop-
ulation sex ratio is 1:1. Benefits due to joint territorial defense and mate guard-
ing also have been proposed to explain social monogamy. These hypotheses
are not necessarily mutually exclusive, and both joint territorial defense and
mate guarding have been thought to explain social monogamy in pair-living
snapping shrimp of the genus Alpheus, for example (Matthews, 2002). Finally,
the need for biparental care can explain social monogamy in some species (see
Table 3.2 in Brotherton & Komers, 2003), but not others.

Aggregations
Aggregated patterns of individuals may be due to social tendencies (e.g., co-
operative care of offspring or predator defense) or to a clumped distribution of re-
sources (e.g., food or mates). Some types of groups form when multiple
individuals are simply attracted to some feature of the environment (Brown &
Orians, 1970; Wittenberger, 1981) in contrast to the social groups formed by
other species such as communal nesters or eusocial animals such as hymenoptera
species (see Table 9.1 in Bourke, 1997) or naked mole-rats (Heterocephalus
glaber; Jarvis, 1981; Lacey & Sherman, 1991). In eusocial species, there is a
much more complex social structure (e.g., division of labor). Avian flocks, fish
schools, and mammalian herds would be intermediate in this continuum of
sociality. In flocks, schools, and herds, social factors hold individuals together,
but there is no division of labor and the groups are without a dominant breeding
pair or leader as seen in eusocial insects and mammals. Many classification
schemes have been proposed to define different types of groups, but we will
not discuss classification schemes or all forms of animal spatial grouping in this
chapter. Rather, we will restrict ourselves to only a few types of space use here.
Fish schools, bird flocks, or mammalian herds—Two important criteria used in
defining a group of this type are density and behavior. Density abruptly increases
when the edge of this type of group is encountered, and there is often some
type(s) of complex, coordinated behavior (such as coordinated movement) result-
ing from social interactions between individuals (Beiswenger, 1975).
MAKE SPACE ENOUGH BETWEEN YOU 227

Aggregations such as fish schools, bird flocks, or mammalian herds appear

to result from the interplay between long-range attraction and short-range
repulsion among individuals (Herbert-Read et al., 2011; Katz et al., 2011).
Groups of this type are proposed to form when predator pressure is high
(Hamilton, 1971). Flocks or schools may also form when high-quality feeding
areas are controlled by territorial conspecifics. The occurrence of a flock or
school increases intruder pressure to the point that territorial defense becomes
more costly than the benefit received, and thus territoriality will disappear
(Robertson et al., 1976). Thus, fish schools, bird flocks, and mammal herds
can also provide foraging benefits (Lazarus, 1979; Partridge, 1982).
Coloniality—When individuals breed in territories that are densely
clustered in space and these territories do not contain resources other than
breeding sites, we refer to these animals as colonial (Danchin et al., 2008).
Coloniality is seen in about 13 percent of avian species and almost all seabirds
(Chiozzi et al., 2011). Coloniality has also been reported in some mammalian
species (e.g., elephant seals, Mirounga leonine, Baldi et al., 1996; prairie dogs,
Cynomys leucurus and C. ludovicianus, Hoogland, 1981; numerous species of
bats, Davis et al., 1962; Bradbury, 1977) and some fish (e.g., damselfish,
Eupomacentrus parties, Myrberg, 1972; bluegill sunfish, Lepomis macrochirus,
Gross & MacMillan, 1981).
Coloniality is thought to have arisen multiple times so selection pressures
favoring its evolution may differ between species (Rolland et al., 1998). Colo-
niality may have evolved and be maintained due to limited favorable breeding
sites (e.g., seabirds, Veen, 1977; and elephant seals, Baldi et al., 1996),
enhanced foraging efficiency (e.g., a colony acting as an information center,
Ward & Zahavi, 1973; Brown, 1986), or allowing individuals more time to
forage if less time is needed for vigilance or protection against predation
through a number of mechanisms such as many eyes to watch for predators,
the confusion effect, or collective defense (reviewed in Wittenberger & Hunt,
1985; Siegel-Causey & Kharitonov, 1990; Danchin et al., 2008). Results
from studies testing these hypotheses are not consistent. In some studies (e.g.,
Olsthoorn & Nelson, 1990; Chiozzi et al., 2011) no nest-site limitation was
found. Furthermore, a negative relationship between per capita nest predation
risk and colony size or nearest-neighbor distance was found in some studies,
but no relationship between these variables or the opposite pattern has been
found in other studies (Møller, 1987; Anderson & Hodum, 1993). Some
investigators also have suggested that there are problems with the hypothesis
that colonies function as information centers to enhance food finding
(Mock et al., 1988; Richner & Heeb, 1995). Heinz Richner and Philipp Heeb
(1996) proposed that, in contrast to the information center hypothesis, colonies
228 ANIMAL BEHAVIOR

might serve as centers for recruiting individuals for group foraging. Additionally,
it has been proposed that coloniality may be a side effect of conspecific attraction
in habitat selection or sexual selection (Morton et al., 1990; Danchin & Wagner,
1997). Conspecific cues can lead to the formation of invertebrate colonies
(Meadows & Campbell, 1972), and evidence suggests that this can occur in
some vertebrates. For example, Jan Veen (1977) was able to attract sandwich
terns (Sterna sandvicencis) to an area using decoys. These individuals settled in
between the decoys initially and then settled on territories around the newly
established colony. Finally, Gene Morton and colleagues (1990) proposed that
colonies might form due to benefits to older males from extra-pair matings with
females whose partners were younger males. Monogamous females may benefit
from having aggregated territories like those seen in a colony for the same reason
because females with male social partners also could obtain extra-pair matings
(Wagner, 1993; Wagner et al., 1996).
Communal nesting—Communal nesting occurs when more than one breed-
ing female occupies a single nest or chamber within a burrow. Communal nest-
ing is fairly widespread among many species of insects (Wilson, 1971) and refers
to a relatively simple form of social organization where individuals share a nest
and each individual produces and provisions her own offspring. There is no
reproductive division of labor and no overlap of generations such as is seen in
eusocial insects (Michener, 1974). Communal nesting also has been reported
in numerous species of birds and mammals such as rodents and banded mongoo-
ses (Mungos mungo; Gilchrist et al., 2004). In these species, females often share in
care of offspring (Hayes, 2000; Vehrencamp & Quinn, 2004), and offspring
may benefit from an increase in care (Lewis & Pusey, 1997; Branchi, 2009).
Despite these benefits, communal nesting is often costly. In some species, such
as greater anis (Crotophaga major), in which multiple monogamous pairs nest
together, females eject each other’s eggs (see also Vehrencamp & Quinn,
2004). In this species, though, benefits may outweigh costs because pairs in larger
groups have greater reproductive success (Riehl, 2011).

PROXIMATE MECHANISMS UNDERLYING SPACING SYSTEMS

Ecological and Social Determinants of Spacing Systems
The most common ecological factor cited for its influence on spacing
patterns is food resources (Maher & Lott, 2000). This is not surprising as a
number of theoretical models predict that variation in food resources could
influence space use (Carpenter & McMillen, 1976; Hixon, 1980; Maher &
Lott, 2000; McLoughlin et al., 2000; Adams, 2001). Food resources can
MAKE SPACE ENOUGH BETWEEN YOU 229

influence spacing behavior in numerous ways, through their abundance,

distribution, accessibility (the degree to which food can be acquired by
individuals), density, predictability, quality, rate of replenishment, and type
(e.g., fruit, insect prey). Because food is relatively easy to quantify and study
compared to other ecological factors, such as predation pressure, the impor-
tance of food as a factor determining spacing systems may have received
more emphasis than its importance would warrant (Stamps, 1994). It is also
likely that more than one variable can influence spacing behavior (Maher &
Lott, 2000).
The abundance of food has been one of the most frequently studied effects
of food resources, and numerous studies have found a relationship between
food abundance and territoriality. However, the nature of this association
has not been consistent among studies. For example, Hawaiian honeycreepers
(Vestiaria coccinea; Carpenter, 1987), California voles (Microtus californicus;
Ostfeld, 1986), and grey-sided voles (Clethrionomys rufocanus; Ims, 1987)
exhibited decreased territoriality at higher food abundance. In contrast, lower
food abundance was correlated with a lack of territoriality in other species such
as the Townsend solitaire (Myadestes townsendi; Lederer, 1981), acorn wood-
pecker (Melanerpes formicivorus; Hannon et al., 1987), and pronghorn (Antilo-
capra Americana; Maher, 1994). While these findings regarding food
abundance and territorial behavior may seem irreconcilable, they are consis-
tent with several cost-benefit models that predict that the relationship between
food abundance and territoriality is an inverted U-shaped function (Carpenter
& MacMillen, 1976; Grant, 1993; Maher & Lott, 2000). These models pre-
dict that territoriality occurs when the quantity of food is moderate, but when
food is scarce or extremely abundant then territories are more energetically
costly to defend than the benefits obtained from defense (i.e., not economi-
cally defendable sensu Brown, 1964).
Food abundance can also affect the size of an individual’s territory. A neg-
ative correlation between food abundance and home range (an area that an
animal moves through in search of food and mates) or territory size also has
been observed in numerous species (Powell, 1987; Sandell, 1989; Maher &
Lott, 2000; Nilsen et al., 2005), and models of optimal territory size
predict such a relationship should often be expected (Hixon, 1980; Schoener,
1983). The results of a natural experiment indicate a causal relationship
between food abundance and territory size. Home-range size and overlap of
adult male desert iguanas (Dipsosaurus dorsalis) were compared before and after
a sandstorm that substantially reduced food resources (e.g., leaves, flowers) of
the herbivorous lizards (Krekorian, 1976). When food resources were less
abundant after the storm, home-range size was significantly greater than
230 ANIMAL BEHAVIOR

before the storm. Furthermore, although male home ranges did not overlap
prior to the storm, there was extensive overlap of home ranges afterwards.
Controlled field manipulations of the abundance of natural food sources also
have shown decreases in territory size in response to higher food levels in mol-
lusks (e.g., owl limpet, Lottia gigantean; Stimson, 1973), insects (e.g., Leucotri-
chia pictipes; Hart, 1985), fish (e.g., rainbow trout, Oncorhynchus mykiss;
Keeley & McPhail, 1998), lizards (e.g., spiny lizard, Sceloporus jarrovi; Simon,
1975), birds (rufus hummingbird, Selasphorus rufus; Hixon et al., 1983), and
mammals (e.g., eastern chipmunk, Tamias striatus; Mares et al., 1982). The
importance of these experimental studies is that they demonstrate a direct
causal link between food abundance and territory size.
The distribution of food resources is another commonly investigated varia-
ble proposed to influence spacing patterns. Some experimental manipulations
showed that territoriality occurred when food was clumped but not when it
was evenly distributed (Zahavi, 1971; Grant & Guha, 1993). In addition,
differences in the distribution of food are correlated with the distribution of
individuals such as prairie voles (Microtus ochrogaster). Dicots, which are criti-
cal in the diet of prairie voles, were more clumped at a study site in Indiana
than at a geographically distinct study site in Kansas. Adult females, males,
and nest sites also were more clumped in Indiana than in Kansas, resulting
in considerable home-range overlap among neighboring females compared to
females in the Kansas population (Streatfeild et al., 2011).
As with food abundance, results from observational studies examining food
distribution and spacing patterns are inconsistent, which may be partially due
to the fact that there are a number of differences in methodologies used in pre-
vious studies. It is also possible that the relationship between the food distribu-
tion and territoriality may be best explained, as was the relationship between
food quantity and territoriality, by an inverted U-shaped curve where moder-
ately clumped resources are the only ones that are worth defending (Craig &
Douglas, 1986).
Investigations of both food abundance and distribution generally showed
that limited, patchy food resulted in increased territoriality. Consistent with
these findings, when food was abundant and patchy, animals did not defend
territories (e.g., black bears, Ursus americanus; Rogers, 1987).
Some studies also showed that types of prey could influence space use. In
Great Britain, badgers (Meles meles) feed primarily on earthworms, which are
very abundant, but in continental Europe earthworms are typically much less
abundant and badgers tend to specialize on foods such as rabbits (Oryctolagus
cuniculus), insects, and fruit (Johnson et al., 2000). Throughout continental
Europe, badgers are typically solitary with little home-range overlap among
MAKE SPACE ENOUGH BETWEEN YOU 231

conspecifics, whereas in Great Britain badgers commonly live in large social

groups of up to 30 individuals sharing a home range.
The ease of finding and killing prey can also result in differences in spacing
patterns. Coyotes (Canis latrans) are typically solitary predators, but several
adults may hunt together sharing a common home range when prey items that
are difficult for single individuals to kill are abundant (Bowyer, 1987). A study
of coyotes in Alberta, Canada, showed that variation in the average size of prey
eaten by coyotes was positively correlated with group size, and coyotes were
more likely to be solitary when their food base was primarily small prey items
such as rodents and fawns instead of adult deer (Bowen, 1981). Finally, nest-
ing Arctic skuas (Steercorarius parasiticus) that feed primarily on lemmings
defend territories while conspecifics in the same area that feed mainly on fish
taken from other birds nest colonially (Andersson & Gotmark, 1980).
Changes in space use also can occur in response to meeting energetic
demands associated with the seasonal change in food resources. The lesser
spotted woodpecker (Dendrocopos minor) inhabits deciduous forests in
Europe. During the summer, individuals feed on surface-living adult insects,
but from autumn to spring they feed almost exclusively on beetle larvae in
dead branches of deciduous trees (Wiktander et al., 2001). Home-range size
increases from about 100 hectares in late spring to over 700 hectares in winter
for males and females. During the winter, these birds also become solitary and
nonterritorial. The pattern of space use in scatter-hoarding Eurasian red
squirrels (Sciurus vulgaris) also changes seasonally in response to food abun-
dance. Home ranges of males and females are largest in spring and summer
when seed supplies are depleted and the squirrels feed more on less preferred
food items such as flowers and buds (Wauters & Dhondt, 1992). Home
ranges are smallest in autumn and winter when seeds are abundant and
comprise 70 percent of the diet.
Spatial organization and space use is also correlated with predation risk (see
Krause & Ruxton, 2002, for a detailed overview). In general, individuals ben-
efit in a number of ways from aggregation and group formation in the pres-
ence of predation (e.g., many-eyes effect, selfish herd effect, predator
confusion, communal defense). Evidence from a number of taxa suggests that
larger groups can detect the approach of predators more effectively (i.e., the
many-eyes hypothesis, reviewed by Elgar, 1989), allowing each group member
to spend less time being vigilant and more time foraging (fish, Godin et al.,
1988; birds, Lazarus, 1979; Cresswell, 1994; mammals, Fairbanks & Dobson,
2007; van Schaik et al., 1983, but see Smith & Cain, 2009). This hypothesis
assumes that information on predator presence or approach is transmitted
clearly among group members (Krause & Ruxton, 2002) and is supported
232 ANIMAL BEHAVIOR

by some (e.g., Magurran & Higham, 1988; Treherne & Foster, 1981) but not
all experimental manipulations that have been conducted (Lima, 1995).
The many-eyes hypothesis is not necessarily distinct from the dilution
effect, which states that as group size increases, the risk of predation for an
individual decreases because the predator can only catch or eat a certain num-
ber of prey. The benefit from the dilution effect may be offset by the cost of
larger groups becoming more conspicuous. Thus, these two factors need to
be examined together to determine if there is a net benefit from grouping
(see also Turner & Pitcher, 1986). These ideas are also related to the selfish
herd effect (Hamilton, 1971), where individuals not only benefit from being
in a group versus being solitary but also benefit by keeping other group mem-
bers between themselves and a potential predator. Finally, the predator confu-
sion effect describes a reduction in the ability of a predator to single out and
kill an individual when it is in a group versus when it is solitary. This effect
would have the biggest impact when group members look very similar and
any individual that looks quite different would enable the predator to focus
upon it more easily (Krause & Ruxton, 2002).
Evidence for particular effects of grouping is limited because it is difficult to
distinguish among them. Additionally, if one of these particular benefits is not
sufficient for an individual to become part of a group, it may be that individ-
uals gain multiple benefits from grouping. In general, studies have provided
support for the effect of predation on spacing behavior, particularly for group
formation under conditions of higher predation, but it is important to also
consider the costs of group formation, such as increased conspicuousness to
a predator, when attempting to determine net benefits to grouping.
Population density is another ecological variable that has often been cited as
correlated with intraspecific variation in spacing behavior. Although prairie
voles are considered socially monogamous, within a population they may live
as single individuals, male-female pairs, or in groups containing at least two
adults of the same sex but frequently three or more voles. The proportion of
groups increases with increasing density in natural populations (Getz et al.,
1987) as well as in seminatural populations (Cochran & Solomon, 2000;
Lucia et al., 2008). Home-range size is also smaller at higher densities
(Solomon et al., 2009). As found in studies of food abundance and distribu-
tion, the results of studies in which the relationship between population den-
sity and territoriality was examined are inconsistent. Many investigators report
the occurrence of territoriality only at low or moderate population densities
and that territorial defense disappears at high density, although this not a
universal result (reviewed by Maher & Lott, 2000). It may be that different
investigators use the terms low, moderate, and high differently or that the
MAKE SPACE ENOUGH BETWEEN YOU 233

relationship between population density and territoriality also is an inverted

U-shaped function where the benefit-cost ratio is greatest when population
density is moderate. In studies where food quantity and population density
were both examined, population density appeared to have a stronger influence
on behavior, with high density resulting in a lack of territoriality (reviewed in
Maher & Lott, 2000). When food distribution and density were examined
concurrently, territoriality occurred when food was clumped. Territoriality
was also found when food was uniformly distributed and population density
was moderate but not when density was low or high (Rubenstein, 1981).
The latter result supports the hypothesis that the relationship between density
and territoriality is nonlinear.
Some investigators have hypothesized that the relationship between popu-
lation density (or food distribution) and territoriality occurs because individ-
uals can assess intruder pressure (Ferron & Ouellet, 1989). The causal
relationship that was proposed is that increased intruder pressure resulted in
decreased territoriality due to the increased costs of defense (reviewed
by Maher & Lott, 2000). This result is seen in some species of coral reef
fishes (Norman & Jones, 1984; Tricas, 1989) and birds (Myers et al., 1979;
Eberhard & Ewald, 1994) where the primary ecological correlate of territory
size is intrusion by conspecifics.
Aside from food resources, predation, and population density, numerous
other ecological variables have been implicated as contributing to intraspecific
variation in space use. Specific features of the abiotic habitat have been corre-
lated with patterns of space use. For example, rainbow trout (Salmo gairdneri)
are more territorial in fast-flowing water relative to still water (Cole & Noakes.
1980). Additionally, soil characteristics are correlated with home-range size in
tuco-tucos (Ctenomys talarum), suggesting that the energetics of digging may
influence home-range size (Cutrera et al., 2006). Individuals in several species
of rodents that are solitary in the summer may share nest sites in the winter
(red-backed vole, Clethrionomys rutilus, West, 1977; meadow voles, Microtus
pennsylvanicus, Madison, 1984; white-footed mice, Peromyscus leucopus;
Madison et al., 1984). Nest sharing in these species may save energy by lower-
ing the cost of thermoregulation. For comprehensive reviews of the potential
ecological factors related to intraspecific variation in space use see Lott
(1991) and Maher and Lott (2000).
Although most studies focus on just one ecological variable, it is likely that
spacing behavior in nature is affected by multiple variables operating simulta-
neously. Greater insight into the ecological determinants of variation in space
use will come from studies examining multiple ecological variables concur-
rently and then determining which variables best explain spacing systems.
234 ANIMAL BEHAVIOR

For example, Ernest Keeley and J. Donald McPhail (1998) showed that
territory size in juvenile steelhead trout (Oncorhynchus mykiss) was significantly
inversely related to food abundance, but there was no significant affect of
intruder pressure on territory size.

Individual Characteristics and Spacing

Intrapopulation differences in spacing behavior may be affected by charac-
teristics of individuals (Steury & Murray, 2003). Larger male sphecid wasps
(Philanthus basilaris) were significantly more likely to be territorial than small
males (O’Neill, 1983). In the spiny lizard (Sceloporus jarrovi), territory size
was influenced by sex and body size. Male territories were twice the size of
female territories after controlling for body size. Within each sex, larger lizards
had larger territories (Simon, 1975). In juvenile Atlantic salmon (Salmo salar),
88 percent of the variation in territory size was explained by differences
in body size and age, whereas food abundance only explained an additional
2 percent of the variation (Keeley & Grant, 1995).

Genetic Basis for Spacing Behavior

There are not many studies that have been conducted on the genetic influ-
ences on spacing behavior, but existing studies suggest that intraspecific differ-
ences in spatial organization may also reflect genetic differences among
individuals within or among populations (spiders, Uetz & Cangialosi, 1986;
fish, Seghers, 1974; mammals, Hammock & Young, 2005). Dominic Wright
and colleagues (2003) conducted a well-controlled breeding experiment to
determine if there were differences in schooling behavior in zebrafish (Danio
rerio) within and between populations. There was no significant difference in
schooling between populations, but there was a within-population difference,
indicating that schooling had a genetic component (Wright et al., 2003).
Geographically isolated populations of guppies (Poecilia reticulata) also show
differences in schooling behavior, even within the same stream. Guppies from
populations in areas with predatory fish have a much greater tendency to
school than guppies from areas without predatory fish. Guppies in streams
without predators tend to be solitary (Seghers, 1974). Laboratory breeding
experiments confirmed that differences in the tendency to school had an
underlying genetic basis. In a small number of species, alternative reproduc-
tive tactics (different behavior patterns shown by one sex within the same spe-
cies to enhance their reproductive success), with their concomitant differences
in space use (territorial versus nonterritorial), are genetically based, and males
do not switch tactics (e.g., bluegill sunfish, Lepomis macrochirus, and Pacific
MAKE SPACE ENOUGH BETWEEN YOU 235

salmon, Oncorhynchus spp., Gross, 1984; marine isopods, Paracerceis sculpta,

Shuster & Wade, 1991; ruffs, Philomachus pugnax, Lank et al., 1995; tree
lizards, Urosaurus ornatus, Moore et al., 1998).
The neuropeptide arginine vasopressin has been shown to be important in
the mediation of social behavior in male prairie voles through its action on the
vasopressin 1a receptor (V1aR; Insel et al., 1994; Wang et al., 1998; Phelps,
2010). Individual variation in V1aR neural expression is correlated with
length polymorphisms of microsatellite DNA within the regulatory region of
the gene (avpr1a) encoding the V1aR (Hammock & Young, 2005). In voles
maintained in outdoor rodent enclosures, neural V1aR expression in areas of
the brain associated with spatial memory strongly covaried with space use, sug-
gesting a possible link between male space use and avpr1a genotype (Ophir
et al., 2008, but see Solomon et al., 2009).

Neurobiological Basis for Spacing Behavior

During the past few decades, evidence has accumulated showing that
gonadal hormones, particularly testosterone, regulate territoriality in males at
the beginning of the breeding season by apparently affecting levels of aggres-
sion (Nelson, 2000; Sinervo et al., 2000). Intrasexual competition during
the breeding season results in transient increases in testosterone that, in turn,
induce aggression toward same-sex conspecifics (Wingfield et al., 1987,
1990). In direct contrast, during the nonbreeding season when testosterone
levels are low and the testes are regressed (Wingfield, 1994), numerous species
of male birds and some species of mammals and reptiles still display territorial-
ity (Caldwell et al., 1984; Logan, 1992; Moore & Marler, 1987; Gwinner
et al., 1994). Some evidence suggests that territoriality during the nonbreeding
season is modulated by conversion of testosterone to estrogen and the interac-
tion of estrogen with the estrogen receptor (Soma et al., 1999; Wingfield et al.,
2001). Because testosterone is typically low during the nonbreeding season,
Kiran Soma and John Wingfield (2001) suggested that a testosterone precur-
sor, dehydroepiandrosterone (DHEA), might be the substrate that is con-
verted to estrogen (see Vanson et al., 1996). DHEA is also present in the
blood plasma of temperate songbirds, some tropical birds (Hau et al., 2004),
and mammals such as red squirrels (Boonstra et al., 2008) during the non-
breeding season at significantly higher levels than testosterone or estrogen.
The seasonal pattern of DHEA parallels that of territorial aggression. Soma
and Wingfield (2001) and Rudy Boonstra and colleagues (2008) suggest that
the possible involvement of DHEA in aggression during the nonbreeding
season may be adaptive because it allows delivery of sex steroids to speci-
fic brain regions without the cost of maintaining high levels of plasma
236 ANIMAL BEHAVIOR

testosterone. In the nonbreeding season, typically winter, many birds and small
mammals increase their immune and thermogenic (heat-producing) activity so
they can survive through the winter. Because high levels of testosterone suppress
the immune system and increase energetic costs, maintenance of high levels of
testosterone would increase winter mortality (Wingfield et al., 2001).
Previous studies also suggest that neuropeptides, such as arginine vasopres-
sin and serotonin, modulate aggression in a number of species including crus-
taceans, amphibians, birds, and mammals. For example, neural infusions of
vasopressin inhibited aggressive chases and songs in two species of territorial
birds (Goodson, 1998a, 1998b). Similarly, some species of birds and mam-
mals also show an inverse relationship between serotonin levels and aggression
(Edwards & Krawitz, 1997; Sperry et al., 2003; Ten Eyck, 2008).
There is much less known about the hormonal responses to a territorial
intrusion in females. The few studies that exist suggest that hormonal control
of territorial aggression in females is different from that in males. Female song
sparrows (Melospiza melodia) that experienced simulated territorial intrusions
had significantly lower levels of testosterone and dihydrotestosterone than
controls (Elekonich & Wingfield, 2000). There were no differences between
groups with or without simulated intrusions in levels of estradiol, progester-
one, or corticosterone. In a similar study with the monogamous California
mouse (Peromyscus californicus), a decrease in progesterone and the progester-
one/testosterone ratio was the only hormonal change following simulated
territorial intrusion (Davis & Marler, 2003), again suggesting that territorial
aggression by females is not facilitated by increased testosterone as seen in
males. Androgens may limit female aggression to a short-term response rather
than facilitating it, which may prevent aggression from interfering with repro-
duction (Elekonich & Wingfield, 2000).

Developmental Effects on Spacing Behavior

Some animals display ontogenetic shifts in spacing behavior, that is,
changes in spacing behavior during maturation or between the time when
individuals are juveniles and when they are adults. More typically during early
life stages, when individuals are inexperienced and vulnerable, they tend to
aggregate, but then aggregation decreases when they are subadults or adults
as the costs of competition for critical resources become greater (Krause &
Ruxton, 2002). Often, after sexual maturation, aggregation occurs again for
reproductive purposes. It is less common to find species in which juveniles
are solitary and then aggregate as subadults. Typically juveniles of these species
are slow and use crypsis to avoid predator detection. Thus, aggregation while
MAKE SPACE ENOUGH BETWEEN YOU 237

still a defenseless juvenile would increase detection and, therefore, the risk of
predation (Tinbergen et al., 1967; Butler et al., 1999). New Zealand spiny
lobsters ( Jasus edwardsii) are solitary when they are young and small but
become more social when they get older and larger, settling with others in
dens (Butler et al., 1999). Aggregation occurs as a response to conspecific
chemical cues, but the response to chemical cues does not develop until the
lobsters are adolescents. Aggregation of subadults and adults appears to be
adaptive. Mark Butler and colleagues (1999) found no difference in survival
in solitary or aggregated juvenile lobsters, but survival more than doubled
when subadults were grouped as compared to when they were solitary. In a
few cases such as the spiny lobster, this ontogenetic shift in spacing behavior
may be due to developmental constraints. For example, some marine fish do
not begin to school until metamorphosis. At this time, the larvae begin to
show pigmentation and are more visible to predators (Gallego & Heath,
1994), so it is adaptive to school at this time.
One study of brown trout (Salmo trutta) showed that despite rearing envi-
ronment, aggression occurred in all dyadic encounters, but the level of aggres-
sion differed depending on early rearing environment (Sundstrum et al.,
2003). Brown trout that had been reared under high-density conditions (in a
hatchery) displayed more aggression and initiated aggressive encounters sooner
than individuals from low-density populations in the wild, in contrast to what
had been predicted. The authors suggested that this may be due to lack of
experience in territorial encounters and in assessing fighting ability among
hatchery-reared fish compared to wild trout. Similar results were seen in
another study where wild and hatchery-reared brown trout actually competed
for territories (Deverill et al., 1999). Interestingly, although hatchery-reared
fish initiated more aggressive encounters, they were not more successful at
obtaining better feeding territories than wild trout (Sundstrom et al., 2003;
Deverill et al., 1999). It is not clear what specific factor in the environments
may have caused the difference in levels of territorial aggression in these trout,
but it appeared that rearing environment did affect subsequent aggression.
Somewhat similar results were found in a study of the effects of population
density on schooling tendency in guppies (Chapman et al., 2008). Trinada-
dian guppies raised in low densities showed a greater tendency to school than
guppies raised in high densities. The investigators hypothesized that, in the
absence of predator cues, higher densities resulted in more aggression and thus
increased the costs of schooling relative to benefits during rearing.
In some species such as the desert locust (Schistocerca gregaria), which can
exist in a solitary or a gregarious phase, crowding or high densities not only
increases gregariousness during an individual’s lifetime but also in its offspring
238 ANIMAL BEHAVIOR

if crowding occurs during oviposition (Islam et al., 1994). Alan McCaffery

and colleagues (1998) reported that this maternal effect is due to a chemical
in the foam from the egg pods that influences subsequent behavior of
offspring.

Consequences of Intraspecific Variation in Spacing Behavior

Intraspecific variation in spacing can have a substantial impact on repro-
duction, population structure, and demography. An improved knowledge of
the ecological and genetic determinants of intraspecific variation in space use
will influence our understanding of many aspects of population and behavioral
ecology (e.g., mating and social systems, carrying capacity) as well as contrib-
ute to conservation of threatened or endangered species.

Effects on Mating Patterns and Reproductive Success

Spacing patterns can influence the social and probably also the genetic mat-
ing systems in populations. Determining the social mating system (who is
sharing space with or living with whom) can be difficult for species like
rodents that can be difficult to observe. We used to think that a reliable index
of the mating system was the spacing patterns of individuals within a popula-
tion (Emlen & Oring, 1977; Ostfeld, 1985; Anderson & Titman, 1992), but
numerous studies have shown that the genetic mating patterns differ signifi-
cantly from the mating system inferred from trapping or tracking data. We
now realize that observations, live trapping, and radio tracking reveal only
the social mating systems in animals.
In some species of mammals (Getz et al., 1993; Wolff, 2008; Schradin
et al., 2012), birds (for a review see Penteriani et al., 2011), amphibians
(Howard, 1978), fish (Gross & Charnov, 1980; Taborsky et al., 2008), and
insects (O’Neill, 1983; Peixoto & De Marco, 2009), some individuals, usually
males, exhibit territorial behavior and occupy a fixed area of residence while
other same-sex adults in the same population do not. Relative to territorial ani-
mals, the nonterritorial individuals (also referred to as wanderers, floaters, or
satellites) typically have much larger home ranges that encompass territories of
multiple opposite-sex conspecifics. In the majority of species where popula-
tions contain territorial and nonterritorial individuals, individuals may switch
roles during their lifetime (Taborsky, 2001; Taborsky et al., 2008; McGuire
& Getz, 2010; Penteriani et al., 2011). These nonterritorial individuals have
generally been portrayed as individuals of poor quality that were excluded
from territories by superior competitors and have been forced to adopt the less
preferred nonterritorial tactic as a means of making the best of a bad situation
MAKE SPACE ENOUGH BETWEEN YOU 239

when there is intense competition for suitable territories (Newton, 1992;

Taborsky, 1998, 2001). The relative fitness of territorial and nonterritorial
individuals has been measured in only a few species, and for some species
territorial individuals do have higher reproductive success (greater sac-winged
bat, Saccopteryx bilineata; Heckel & von Helversen, 2002), while in others
such as the prairie vole results differ between studies (Ophir et al., 2008;
Solomon et al., unpublished data). However, genetic analyses of paternity
showed that wandering and territorial male African stripped mice (Rhabdomys
pumilio) had similar reproductive success during a year when population den-
sity was intermediate (Schradin & Lindholm, 2011), and, in natural popula-
tions, wandering male prairie voles sired more offspring than resident males
(Solomon et al., unpublished data). Therefore, additional studies using genetic
measures of the relative reproductive success of territorial and nonterritorial
individuals are critical to understand the factors underlying the evolution
and maintenance of these two alternative patterns of space use within popula-
tions. In these studies, it is critical to include males that do not sire any
offspring (Shuster, 2011) because if some males gain a disproportionately
large number of matings, and hence reproductive success, then other males
must lose matings or be excluded from mating. An evolutionary genetics
approach requires that the relative fitness of all males, “losers” and “winners,”
be included in the calculations. For example, if most nonterritorial males do
not sire any offspring, then excluding these males may create the appearance
that there is no difference in the relative fitness of each type of male when,
actually, there is a statistically significant difference (Shuster, 2009).
As has been suggested for males, it also has been hypothesized that there
should be a relationship between spacing behavior of females and their repro-
ductive success because changes in spacing patterns can result in changes in
aggressive behavior (Monaghan & Metcalf, 1985), infanticide (Mappes
et al., 1995; Jonsson et al., 2002), or cooperative behavior (Lambin & Krebs,
1991). In bank voles (Myodes glareolus), greater overlap of female home ranges
negatively affected reproductive success of females, particularly offspring sur-
vival (Jonsson et al., 2002). Offspring survival may have decreased due to
infanticide by nearby females.

Effects on Population-level Processes

Spacing patterns may be critical to population regulation through their
effects on aggression (Watts, 1969) and breeding status (Montgomery et al.,
1997). Territoriality can have a strong influence on population dynamics
(Davies & Houston, 1984; Sutherland, 1996). If territories are a fixed size,
only a certain number of them can fit within a defined area, and if some
240 ANIMAL BEHAVIOR

individuals are excluded from territories, we expect that the number of breed-
ing males and females in the population would be limited (i.e., the population
will be regulated by density-dependent processes; Wolff, 1985; Both & Visser,
2003; López-Sepulcre & Kokko, 2005). For example, many species of spiders
are territorial, and at least some inhabit territories of fixed sizes (Reichert,
1981). Because territory size is fixed, additional individuals will not be able
to establish territories between already existing territories.
When the number of territorial individuals is limited due to constraints of
fixed territories, there would be an increase in the number of floaters, which were
assumed to be nonbreeders (Smith & Arcese, 1989; Kempenaers et al., 2001). If
they do not breed, an increase of floaters, instead of breeders, in the population
would lead to stable or decreased population density. A decrease in population
density would be expected if floaters cause direct disturbance or conflict for space
(López-Sepulcre & Kokko, 2005) such as has been reported in a number of spe-
cies of birds (Arcese et al., 1992; Komdeur, 1996), ayu fish (Plecoglossus altivelis;
Iguchi & Hino, 1996), and red squirrels (Sciurus vulgaris; Wauters & Lens,
1995). A limit on the number of breeders depends on whether or not territory
size is compressible (i.e., whether resources are divisible, Both & Visser, 2003)
and how reproductive success is affected by territory size (López-Sepulcre &
Kokko, 2005). Now that we are finding that floaters do reproduce in at least
some species (Kempenaers et al., 2001; Solomon & Keane, unpublished data),
we will need to rethink their influence on population density. The effect of
reproductive floaters on population density will depend on their contribution
to the population versus the contribution of resident breeders.
In species where the size of the territory is flexible, individuals may establish
territories between the boundaries of previously established territories or cause
territorial residents to defend smaller territories, which would allow previous
floaters or recruits to establish territories (e.g., Eurasian oystercatchers, Hae-
matopus ostralegus; Heg et al., 2000). This pattern could result in increasing
population density through the settlement of floaters or decreasing density
through decreasing birth or survival of offspring from established breeders. If
food is the limiting resource for females and the amount of food affects repro-
duction, we expect there to be density-dependent reproduction that influences
population size or density by means of territoriality. At high density, we expect
that the optimal territory size will be decreased, and thus reproduction or
reproductive success should decrease (Both & Visser, 2003). In areas where
great tits (Parus major) have large, contiguous territories, there is strong
density-dependent reproduction (Both, 1998).
The results from a number of studies on the effects of territoriality on pop-
ulation density are inconsistent, which may be due to the lack of certainty
MAKE SPACE ENOUGH BETWEEN YOU 241

about whether or not the resource being defended was critical or whether the
size of territories was flexible (Adams, 2001; Both & Visser, 2003). No effect
on population size has been found in some of the studies of rodents. Ylönen
and Viitala (1991) housed bank voles in experimental enclosures where they
were given either clumped, predictable food or evenly distributed food.
Although spacing behavior of females differed depending on the distribution
of food resources and females in the clumped food treatment produced their
first litter earlier in the spring, there were no differences in population size in
late spring. It was assumed that the critical resource that females were compet-
ing for was food, but high-quality nest sites could also have been important. In
addition, the length of the study may have been too short for an effect to be
seen. It is also not known what would have happened to population sizes if
the study had continued later into the breeding season. A subsequent study
by Alice Rémy (2011) has shown that female bank voles in plots with clumped
food were more aggregated compared to females in the dispersed-food treat-
ment. Female aggregation was a strong predictor of population size during
the breeding as well as the nonbreeding season. Analyses suggested that voles
in all treatments had similar access to food resources. Therefore, it was most
likely that differences in social interactions, due to differences in spacing,
resulted in differences in population size.

Effects on Conservation
Anthropogenic activities are substantially altering many natural ecosystems.
A better understanding of the extent and basis of intraspecific variation in
space use is vital for accurately predicting how species will respond to environ-
mental change as well as for the effective conservation and management of a
species. Spacing patterns can influence a number of factors that are relevant
to conservation including growth rates of a population, susceptibility of a pop-
ulation to disease, responses of a population to exploitation and recovery, and
recolonization of habitats (Dobson & Poole, 1998). In a number of species,
conspecific aggregation or attraction is responsible for at least some of these
effects. In many species, individuals use the presence of cues from conspecifics
to determine whether or not a particular habitat is suitable for survival and
reproduction (see work by Stamps, 1988, on conspecific attraction).
Another effect that has received some attention in the last decade is the
Allee effect, which describes a phenomenon where the reproductive rate of a
population declines with declining population density (Allee, 1931). If popu-
lation growth depends on encounter rates between opposite-sex conspecifics
or group size, then spacing patterns will have conservation consequences.
242 ANIMAL BEHAVIOR

If the encounter rates, particularly with members of the opposite sex, or group
size fall below a certain threshold, then the population may decline until
reproduction ceases and that population becomes extinct (Courchamp et al.,
1999). In addition, when the size of a group or population becomes very
small, the time it takes for that population to go extinct will decrease (Dobson
& Poole, 1998).
Dispersion of individuals within a habitat can also affect the likelihood and
timing of parasite or pathogen transmission, which also can lead to severe pop-
ulation decreases or extinction. For example, rabies and canine distemper
affected African wild dogs (Lycaon pictus) in the early 1990s, which resulted
in local population extinction (Alexander & Appel, 1994; Kat et al., 1995).
Differences in the sizes of nearest-neighbor distances among individuals living
in groups can result in differences in pathogen establishment and transmission
rates. For example, African wild dogs and lions (Panthera leo) live in relatively
small groups, but the rate of spread of pathogens between groups was slow
compared to the spread in North Sea harbor seal (Phoca vitulina) colonies,
where many individuals died because these seals live in large groups (Dobson
& Poole, 1998).
Exploitation of elephant populations has resulted in a continent-wide
decrease in numbers of greater than 95 percent in the past 200 years
(Millner-Gulland et al., 1993). The size of the group in which females live
contributes to the problem of finding a suitable mate. Females in widely dis-
persed, small groups had approximately a 39 percent reproductive rate com-
pared to females living in a single large group. Females in the latter situation
had an 87 percent reproductive rate (Dobson & Poole, 1998).
If a population of any particular species were to go extinct in a suitable hab-
itat, the presence of conspecifics or their cues may be necessary when attempt-
ing to reestablish a population in that habitat. Without the presence of
conspecifics, newly released or translocated individuals may not realize that
the habitat is suitable. To solve this problem, biologists have used decoys
and playback of species-specific calls to reestablish populations of extirpated
birds in suitable habitat patches (Kress, 1983; Ward & Schlossberg, 2004;
Parker et al., 2007). Thus, presence or spacing of individuals of the correct
sex and age class could have a positive impact on reintroduction programs.

CONCLUSIONS
A considerable body of information describing animal spacing sys-
tems across a broad range of taxa has accumulated over the last 50 years.
Flexibility in spacing behavior among individuals within a species in response
to environmental and individual conditions is commonly reported, but our
MAKE SPACE ENOUGH BETWEEN YOU 243

understanding of factors underlying intraspecific variation in space use still

needs improvement. The next challenge for research examining animal spac-
ing is to improve our understanding of the ultimate and proximate bases of
variation in spatial systems. Intraspecific variation in spatial structure can be
a valuable tool in elucidating the environmental and genetic determinants of
spatial dispersion and space use, and we suggest that progress in this area can
be achieved using two general research approaches. First, much of the empiri-
cal support for ecological determinants of space-use patterns is based on corre-
lations between ecological variables and spacing patterns. Controlled field
experiments in which ecological or other variables (e.g., early experience, hor-
mone levels) are manipulated are needed to identify casual rather than correla-
tional relationships between space use and specific factors affecting spacing.
Such studies will also be crucial for testing predictions of theoretical models
proposed to explain intraspecific variation in spacing (e.g., Maher & Lott,
2000; McLoughlin et al., 2000; Adams, 2001; Both & Visser, 2003). Second,
numerous molecular techniques are now readily available that allow investiga-
tors to assess genetic variation among large numbers of individuals with rela-
tive ease. Genetic data can be used to determine if intra- or interpopulation
variation in space use has a genetic component. Ultimately, studies simultane-
ously examining multiple ecological and individual variables as well as genetic
factors should be most enlightening because patterns of space use are likely to
influenced to varying degrees by a combination of factors.

ACKNOWLEDGMENTS
We thank former postdoctoral associates and students for assistance with
research, which has allowed us to get a glimpse into the spacing patterns of prairie
voles and sparked our interest in this topic. We thank numerous colleagues and
former students for making us aware of some of the exciting developments in spac-
ing behavior. The National Science Foundation and National Institutes of Health
funded our work, which gave us the opportunity to investigate aspects of spacing
patterns. Any opinions, findings, and conclusions or recommendations expressed
in this material are those of the authors and do not necessarily reflect the views of
the National Science Foundation. We thank Ken Yasukawa for the opportunity
to learn more about this important topic in the writing of this chapter.

REFERENCES AND SUGGESTED READING

Adams, E. S. (2001). Approaches to the study of territory size and shape. Annual
Review of Ecology and Systematics, 32, 277–303.
244 ANIMAL BEHAVIOR

Alexander, K. A. & M. J. G. Appel. (1994). African wild dogs (Lycaon pictus) endan-
gered by a canine distemper epizootic among domestic dogs near the Masai
Mara National Reserve, Kenya. Journal of Wildlife Disease, 30, 481–485.
Allee, W. C. (1931). Animal Aggregations: A Study in General Sociology. Chicago:
University of Chicago Press.
Anderson, D. C. & P. J. Hodum. (1993). Predator behavior favors clumped nesting
in an oceanic seabird. Ecology, 74, 2462–2464.
Anderson, M. G. & R. D. Titman. (1992). Spacing patterns. In B. D. J. Batt, A. D.
Afton, M. G. Anderson, C. D. Ankney, D. H. Johnson, J. A. Kadlec, & G. L.
Krapu (eds.), Ecology and Management of Breeding Waterfowl (pp. 251–289).
Minneapolis: University of Minnesota Press.
Andersson, M. & F. Götmark. (1980). Social organization and foraging ecology in the
arctic skua, Stercorarius parasiticus: A test of the food defendability hypothesis.
Oikos, 35, 63–71.
Arcese, P., J. N. M. Smith, W. M. Hochachka, C. M. Rogers, & D. Ludwig. (1992).
Stability, regulation, and the determination of abundance in an insular song
sparrow population. Ecology, 73, 805–822.
Back, S. R., L. A. Beeler, R. L. Schaefer, & N. G. Solomon. (2002). Testing func-
tional hypotheses for the behavior of resident pine voles, Microtus pinetorum,
toward non-residents. Ethology, 108, 1023–1039.
Baldi, R., C. Campagna, S. Pedraza, J. Burney, & B. J. LeBoeuf. (1996). Social effects
of space availability on the breeding behavior of elephant seals in Patagonia.
Animal Behaviour, 51, 717–724.
Beiswenger, R. E. (1975). Structure and function in aggregations of tadpoles of the
American toad, Bufo americanus. Herpetologica, 31, 222–233.
Boonstra, R., J. E. Lane, S. Boutin, A. Bradley, L. Desantis, A. E. M. Newman, &
K. K. Soma. (2008). Plasma DHEA in wild, territorial red squirrels: Seasonal
variation and effect of ACTH. General and Comparative Endocrinology, 158,
61–67.
Both, C. (1998). Experimental evidence for density dependence of reproduction in
great tits. Journal of Animal Ecology, 67, 667–674.
Both, C. & M. E. Visser. (2003). Density dependence, territoriality, and divisibility
of resources: From optimality models to population processes. American Natu-
ralist, 161, 326–336.
Bourke, A. F. G. (1997). Sociality and kin selection in insects. In J. R. Krebs & N. B.
Davies (eds.), Behavioral Ecology: An Evolutionary Approach, Fourth Edition.
(pp. 203–227). Cambridge, MA: Blackwell Science.
Bowen, W. D. (1981). Variation in coyote social organization: The influence of prey
size. Canadian Journal of Zoology, 59, 639–652.
Bowyer, R. T. (1987). Coyote group size relative to predation on mule deer. Mamma-
lia, 51, 515–526.
Bradbury, J. W. (1977). Social organization and communication. In W. A. Wimsatt
(ed.), Biology of Bats, Vol. 3 (pp. 1–72). New York: Academic Press.
MAKE SPACE ENOUGH BETWEEN YOU 245

Branchi, I. (2009). The mouse communal nest: Investigating the epigenetic influen-
ces of the early social environment on brain and behavior development. Neuro-
science and Biobehavioral Reviews, 33, 551–559.
Brotherton, P. N. M. & P. E. Komers. (2003). Mate guarding and the evolution of
social monogamy in mammals. In U. H. Reichard & C. Boesch (eds.),
Monogamy: Mating Strategies and Partnerships in Birds, Humans and Other
Mammals (pp. 42–58). Cambridge, UK: Cambridge University Press.
Brown, C. R. (1986). Cliff swallow colonies as information centers. Science, 234, 83–85.
Brown, J. L. (1964). The evolution of diversity in avian territorial systems. Wilson
Bulletin, 76, 160–169.
Brown, J. L. (1975). The Evolution of Behavior. New York: Norton.
Brown, J. L. & G. H. Orians. (1970). Spacing patterns in mobile animals. Annual
Review of Ecology and Systematics, 1, 239–262.
Butler, M. J., A. B. MacDiarmid, & J. D. Booth. (1999). The cause and consequence
of ontogenetic changes in social aggregation in New Zealand spiny lobster.
Marine Ecology Progress Series, 188, 179–191.
Caldwell, G. S, S. E. Glickman, & E. R. Smith. (1984). Seasonal aggression indepen-
dent of seasonal testosterone in wood rats. Proceedings of the National Academy
of Sciences, USA, 81, 5255–5257.
Carpenter, F. L. (1987). Food abundance and territoriality: To defend or not to
defend? American Zoologist, 27, 387–399.
Carpenter, F. L. & R. E. MacMillen. (1976). Threshold model of feeding territorial-
ity and test with a Hawaiian honeycreeper. Science, 194, 639–642.
Chapman, B. B., A. J. W. Ward, & J. Krause. (2008). Schooling and learning: Early
social environment predicts social learning ability in the guppy, Poecilia reticu-
lata. Animal Behaviour, 76, 923–929.
Chiozzi, G., G. de Marchi, & D. Semere. (2011). Coloniality in the crab plover Dro-
mas ardeola does not depend on nest site limitation. Waterbirds, 34, 77–81.
Clutton-Brock, T. H. (1991). The Evolution of Parental Care. Princeton, NJ: Prince-
ton University Press.
Cochran, G. R. & N. G. Solomon. (2000). Effects of food supplementation on the
social organization of prairie voles (Microtus ochrogaster). Journal of Mammal-
ogy, 81, 746–757.
Cole, K. S. & D. L. G. Noakes. (1980). Development of early social behavior of
rainbow trout, Salmo gairdneri (Pisces, Salmonidae). Behavioural Processes, 5,
97–112.
Corbet, N. U. & R. J. van Aarde. (1996). Social organization and space use in the Cape
porcupine in a southern African savanna. African Journal of Ecology, 34, 1–14.
Courchamp, F. T., T. Clutton-Brock, & B. Grenfell. (1999). Inverse density depen-
dence and the Allee effect. Trends in Ecology and Evolution, 14, 405–410.
Craig, J. L. & M. E. Douglas. (1986). Resource distribution, aggressive asymmetries
and variable access to resources in a nectar feeding bellbird. Behavioral Ecology
and Sociobiology, 18, 231–240.
246 ANIMAL BEHAVIOR

Cresswell, W. (1994). Flocking is an effective anti-predation strategy in redshanks,

Tringa totanus. Animal Behaviour, 47, 433–442.
Cutrera, A. P., C. D. Antinuchi, M. S. Mora, & A. I. Vassallo. (2006). Home-range
and activity patterns of the South American subterranean rodent Ctenomys
talarum. Journal of Mammalogy, 87, 1183–1191.
Danchin, E. & R. H. Wagner. (1997). The evolution of coloniality: The emergence
of new perspectives. Trends in Ecology and Evolution, 12, 342–347.
Danchin, E., L.-A. Giraldeau, & R. H. Wagner. (2008). Animal aggregations. In
E. Danchin, L.-A. Giraldeau, & F. Cézilly (eds.), Behavioural Ecology: An
Evolutionary Perspective on Behaviour (pp. 503–545). Oxford, UK: Oxford
University Press.
Davies, N. B. & A. I. Houston. (1984). Territory economics. In J. R. Krebs & N. B.
Davies (eds.), Behavioural Ecology: An Evolutionary Approach, Second Edition
(pp. 148–169). Sunderland, MA: Sinauer Associates.
Davis, E. S. & C. A. Marler. (2003). The progesterone challenge: Steroid hormone
changes following a simulated territorial intrusion in female Peromyscus califor-
nicus. Hormones and Behavior, 44, 185–198.
Davis, R. B., C. F. Herreid II, & H. L. Short. (1962). Mexican free-tailed bats in
Texas. Ecological Monographs, 32, 311–346.
Deverill, J. I., C. E. Adams, & C. W. Bean. (1999). Prior residence, aggression and
territory acquisition in hatchery-reared and wild brown trout. Journal of Fish
Biology, 55, 868–875.
Dobson, A. & J. Poole. (1998). Conspecific aggregation and conservation biology. In
T. Caro (ed.), Behavioral Ecology and Conservation Biology (pp. 193–208).
Oxford, UK: Oxford University Press.
Eberhard, J. R. & P. W. Ewald. (1994). Food availability, intrusion pressure and
territory size: An experimental study of Anna’s hummingbirds (Calypte anna).
Behavioral Ecology and Sociobiology, 34, 11–18.
Edwards, D. H. & E. A. Kravitz. (1997). Serotonin, social status and aggression. Cur-
rent Opinion in Neurobiology, 7, 812–819.
Elekonich, M. M. & J. C. Wingfield. (2000). Seasonality and hormonal control of
territorial aggression in female song sparrows (Passeriformes: Emberizidae:
Melospiza melodia). Ethology, 106, 493–510.
Elgar, M. A. (1989). Predator vigilance and group size in mammals and birds:
A critical review of the empirical evidence. Biological Review, 64, 13–33.
Emlen, S. T. & L. W. Oring. (1977). Ecology, sexual selection and the evolution of
mating systems. Science, 197, 215–223.
Fairbanks, B. & F. S. Dobson. (2007). Mechanisms of the group-size effect on vigi-
lance in Columbian ground squirrels: Dilution versus detection. Animal
Behaviour, 73, 115–123.
Ferron, J. & J. P. Ouellet. (1989). Temporal and intersexual variation in the use of
space with regard to social organization in the woodchuck (Marmota monax).
Canadian Journal of Zoology, 67, 1642–1649.
MAKE SPACE ENOUGH BETWEEN YOU 247

Fietz, J. (1999). Monogamy as a rule rather than exception in nocturnal lemurs:

The case of the fat-tailed dwarf lemur, Cheirogaleus medius. Ethology, 105,
255–272.
Gallego, A. & M. R. Heath. (1994). The development of schooling behavior in Atlan-
tic herring Clupea harengus. Journal of Fish Biology, 45, 569–588.
Getz, L. L., J. E. Hofmann, & C. S. Carter. (1987). Mating system and population fluc-
tuations of the prairie vole, Microtus ochrogaster. American Zoologist, 27, 909–920.
Getz, L. L., B. McGuire, T. Pizzuto, J. E. Hofmann, & B. Frase. (1993). Social organi-
zation of the prairie vole, Microtus ochrogaster. Journal of Mammalogy, 74, 44–58.
Gilchrist, J. S., E. Otali, & F. Mwanguhya. (2004). Why breed communally? Factors
affecting fecundity in a communal breeding mammal: The banded mongoose
(Mungos mungo). Behavioral Ecology and Sociobiology, 57, 119–131.
Godin, J.-G. J., L. J. Classon, & M. V. Abrahams. (1988). Group vigilance and shoal
size in a small characin fish. Behaviour, 104, 29–40.
Goodson, J. L. (1998a). Territorial aggression and dawn song are modulated by septal
vasotocin and vasoactive intestinal polypeptide in male field sparrows (Spizella
pusilla). Hormones and Behavior, 34, 67–77.
Goodson, J. L. (1998b). Vasotocin and vasoactive intestinal polypeptide modulate
aggression in a territorial songbird, the violet-eared waxbill (Estrilididae: Ura-
ginthus granatina). General and Comparative Endocrinology, 111, 233–244.
Grant, J. W. A. (1993) Whether or not to defend? The influence of resource distribu-
tion. Marine & Freshwater Behavior and Physiology, 23, 137–153.
Grant, J. W. A. & R. T. Guha. (1993). Spatial clumping of food increases its
monopolization and defense by convict cichlids, Cichlasoma nigrofasciatum.
Behavioral Ecology, 4, 293–296.
Gross, M. R. (1984). Sunfish, salmon, and the evolution of alternative reproductive
strategies and tactics in fishes. In R. Wooten & G. Potts (eds.), Fish Reproduc-
tion: Strategies and Tactics (pp. 55–75). London: Academic Press.
Gross, M. R. & E. L. Charnov. (1980). Alternative male life histories in bluegill sun-
fish. Proceedings of the National Academy of Sciences, USA, 77, 6937–6940.
Gross, M. R. & A. M. MacMillan. (1981). Predation and evolution of colonial nest-
ing in bluegill sunfish (Lepomis macrochirus). Behavioral Ecology and Sociobiol-
ogy, 8, 163–174.
Gwinner, E., T. Roedl, & H. Schwabl. (1994). Pair territoriality of wintering stone-
chats: Behaviour, function and hormones. Behavioral Ecology and Sociobiology,
34, 321–327.
Hamilton, W. D. (1971). Geometry for the selfish herd. Journal of Theoretical Biology,
31, 295–311.
Hammock, E. A. D., & L. J. Young. (2005). Microsatellite instability generates diver-
sity in brain and sociobehavioral traits. Science, 308, 1630–1634.
Hannon, S. J., R. L. Mumme, W. D. Koenig, S. Spon, & F. A. Pitelka. (1987). Poor
acorn crop, dominance, and decline in numbers of acorn woodpeckers. Journal
of Animal Ecology, 56, 197–207.
248 ANIMAL BEHAVIOR

Hart, D. D. (1985). Causes and consequences of territoriality in a grazing stream

insect. Ecology, 66, 404–414.
Hau, M., S. T. Stoddard, & K. K. Soma. (2004). Territorial aggression and hormones
during the non-breeding season in a tropical bird. Hormones and Behavior, 45,
40–49.
Hayes, L. D. (2000). To nest communally or not to nest communally: A review of
rodent communal nesting and nursing. Animal Behaviour, 59, 677–688.
Heckel, G. & O. von Helversen. (2002). Male tactics and reproductive success in the
harem polygynous bat Saccopteryx bilineata. Behavioral Ecology, 13, 750–756.
Heg, D., B. J. Ens, H. P. van der Jeugd, & L. W. Bruinzeel. (2000). Local dominance
and territorial settlement of nonbreeding oystercatchers. Behaviour, 137,
473–530.
Herbert-Read, J. E., A. Perna, R. P. Mann, T. M. Schaerf, D. J. T. Sumpter, &
A. J. W. Ward. (2011). Inferring the rules of interaction of shoaling fish. Pro-
ceedings of the National Academy of Sciences, USA, 108, 18726–18731.
Hixon, M.A. (1980). Food production and competitor density as the determinants of
feeding territory size. American Naturalist, 115, 510–530.
Hixon, M. A., F. L. Carpenter, & D. C. Patton. (1983). Territory area, flower den-
sity, and time budgeting in hummingbirds: An experimental and theoretical
analysis. American Naturalist, 122, 366–391.
Hoogland, J. L. (1981). The evolution of coloniality in white-tailed and black-tailed
prairie dogs (Sciuridae: Cynomys leucurus and C. ludovicianus). Ecology, 62,
252–272.
Howard, H. E. (1920). Territory in Bird Life. New York: Dutton & Co.
Howard, R. D. (1978). The evolution of mating strategies in bullfrogs, Rana catesbei-
ana. Evolution, 32, 850–871.
Iguchi, K. & T. Hino. (1996). Effect of competitor abundance on feeding territorial-
ity in the grazing fish, the ayu Plecoglossus altivelis. Ecological Research, 11, 165–
173.
Ims, R. A. (1987). Responses in spatial organization and behaviour to manipulations
of the food resource in the vole Clethrionomys rufocanus. Journal of Animal
Ecology, 56, 585–596.
Insel, T. R., Z. X. Wang, & C. F. Ferris. (1994). Patterns of brain vasopressin recep-
tor distribution associated with social organization in microtine rodents. Jour-
nal of Neuroscience, 14, 5381–5392.
Islam, M. S., P. Roessingh, S. J. Simpson, & A. L. Mccaffery. (1994). Effects of pop-
ulation density experienced by parents during mating and oviposition on the
phase of hatchling desert locusts, Schistocerca gregaria. Proceedings of the Royal
Society of London, B, 257, 93–98.
Jarvis, J. U. M. (1981). Eusociality in a mammal: Cooperative breeding in naked
mole-rat colonies. Science, 212, 571–573.
Johnson, D. D. P., D. W. Macdonald, & A. J. Dickman. (2000). An analysis and
review of models of the sociobiology of the Mustelidae. Mammal Review, 30,
171–196.
MAKE SPACE ENOUGH BETWEEN YOU 249

Jonsson, P., T. Hartikainen, E. Koskela, & T. Mappes. (2002). Determinants of

reproductive success in voles: Space use in relation to food and litter size
manipulation. Evolutionary Ecology, 16, 455–467.
Kat, P. W., K. A. Alexander, J. S. Smith, & L. Munson. (1995). Rabies and
African wild dogs in Kenya. Proceedings of the Royal Society of London, B, 262,
229–233.
Katz, Y., K. Tunstrøm, C. C. Ioannou, C. Huepe, & I. D. Couzin. (2011). Inferring
the structure and dynamics of interactions in schooling fish. Proceedings of the
National Academy of Sciences, USA, 108, 18720–18725.
Keeley, E. R. & J. W. A. Grant. (1995). Allometric and environmental correlates of
territory size in juvenile Atlantic salmon (Salmo salar). Canadian Journal of
Fisheries and Aquatic Sciences, 52, 186–196.
Keeley, E. R. & J. D. McPhail. (1998). Food abundance, intruder pressure, and body
size as determinants of territory size in juvenile steelhead trout (Oncorhynchus
mykiss). Behaviour, 135, 65–82.
Kempenaers, B., S. Everding, C. Bishop, P. Boag, & R. J. Robertson. (2001). Extra-
pair paternity and the reproductive role of male floaters in the tree swallow
(Tachycineta bicolor). Behavioral Ecology and Sociobiology, 49, 251–259.
Kishimoto, R. (2003). Social monogamy and social polygyny in a solitary ungulates
the Japanese serow (Capricornis crispus). In U. H. Reichard & C. Boesch
(eds.), Monogamy, Mating Strategies and Partnerships in Birds, Humans, and
Other Mammals (pp. 147–158). Cambridge, UK: Cambridge University Press.
Kleiman, D. G. (1977). Monogamy in mammals. Quarterly Review of Biology, 52,
39–69.
Komdeur, J. (1996). Breeding of the Seychelles magpie robin Copsychus sechellarum
and implications for its conservation. Ibis, 138, 485–498.
Krause, J. & G. D. Ruxton. (2002). Living in Groups. Oxford, UK: Oxford
University Press.
Krekorian, C. O. (1976). Home range size and overlap and their relationship to
food abundance in the desert iguana, Dipsosaurus dorsalis. Herpetologica, 32,
405–412.
Kress, S. W. (1983). The use of decoys, sounds recordings, and gull control for
re-establishing a tern colony in Maine. Colonial Waterbirds, 6, 185–196.
Lacey, E. A. & P. W. Sherman. (1991). Social organization of naked mole-rat colo-
nies: Evidence for divisions of labor. In P. W. Sherman, J. U. M. Jarvis, &
R. D. Alexander (eds.), The Biology of the Naked Mole-rat (pp. 275–336).
Princeton, NJ: Princeton University Press.
Lambin, X. & C. J. Krebs. (1991). Spatial organization and mating system of Microtus
townsendii. Behavioral Ecology and Sociobiology, 28, 353–363.
Lank, D. B., C. M. Smith, O. Hanotte, T. Burke, & F. Cooke. (1995). Genetic poly-
morphism for alternative mating behaviour in lekking male ruff Philomachus
pugnax. Nature, 378, 59–62.
Lazarus, J. (1979). Early warning function of flocking in birds: An experimental study
with captive quelea. Animal Behaviour, 27, 855–865.
250 ANIMAL BEHAVIOR

Lederer, R. J. (1981). Facultative territoriality in Townsend’s solitaire (Myadestes

townsendi). Southwestern Naturalist, 25, 461–467.
Lewis, S. E. & A. E. Pusey. (1997). Factors influencing the occurrence of communal
care in plural breeding mammals. In N. G. Solomon & J. A. French (eds.), Co-
operative Breeding in Mammals (pp. 335–363). New York: Cambridge Univer-
sity Press.
Lima, S. L. (1995). Collective detection of predatory attack by social foragers: Fraught
with ambiguity. Animal Behaviour, 50, 1097–1108.
Logan, C. A. (1992). Testosterone and reproductive adaptations in the autumnal ter-
ritoriality of northern mockingbirds Mimus polyglottos. Ornis Scandinavica, 23,
277–283.
López-Sepulcre A. & H. Kokko. (2005). Territorial defense, territory size, and popu-
lation regulation. American Naturalist, 166, 317–329.
Lott, D. F. (1984). Intraspecific variation in the social systems of wild vertebrates.
Behaviour, 88, 266–325.
Lott, D. F. (1991). Intraspecific Variation in the Social Systems of Wild Vertebrates.
Cambridge, UK: Cambridge University Press.
Lucia, K., B. Keane, L. D. Hayes, Y. K. Lin, R. L. Schaefer, & N. G. Solomon.
(2008). Philopatry in prairie voles: An evaluation of the habitat saturation
hypothesis. Behavioral Ecology, 19, 774–783.
Madison, D. M. (1984). Group nesting and its ecological and evolutionary signifi-
cance in overwintering microtine rodents. In J. F. Merritt (ed.), Winter Ecology
of Small Mammals (pp. 267–274). Pittsburgh, PA: Carnegie Museum of Natu-
ral History.
Madison, D. M., J. P. Hill, & P. E. Gleason. (1984). Seasonality in the nesting
behavior of Peromyscus leucopus. American Midland Naturalist, 112, 201–204.
Magurran, A. E. & A. Higham. (1988). Information transfer across fish shoals under
predator threat. Ethology, 78, 153–158.
Maher, C. R. (1994). Pronghorn male spatial organization: Population differences in
degree of nonterritoriality. Canadian Journal of Zoology, 72, 455–464.
Maher, C. R. & D. F. Lott. (1995). Definitions of territoriality used in the study of
variation in vertebrate spacing systems. Animal Behaviour, 49, 1581–1597.
Maher, C. R. & D. F. Lott. (2000). A review of ecological determinants of territorial-
ity within vertebrate species. American Midland Naturalist, 143, 1–29.
Mappes, T, H. Ylönen, & J. Viitala. (1995). Higher reproductive success among kin
groups of bank voles (Clethrionomys glareolus). Ecology, 76, 1276–1282.
Mares, M. A., T. E. Lacher, M. R. Willig, N. A. Bitar, R. Adams, A. Klinger, &
D. Tazik. (1982). An experimental analysis of social spacing in Tamias striatus.
Ecology, 63, 267–273.
Martin, J. K., K. A. Handasyde, A. C. Taylor, & G. Coulson. (2007). Long-term
pair-bonds without mating fidelity in a mammal. Behaviour, 144, 1419–1445.
Matthews, L. M. (2002). Territorial cooperation and social monogamy: Factors
affecting intersexual behaviours in pair-living snapping shrimp. Animal Behav-
iour, 63, 767–777.
MAKE SPACE ENOUGH BETWEEN YOU 251

McCaffery, A. R., S. J. Simpson, M. S. Islam, & P. Roessingh. (1998). A gregarizing

factor present in the egg pod foam of the desert locust Schistocerca gregaria.
Journal of Experimental Biology, 201, 347–363.
McGuire, B. & L. L. Getz. (2010). Alternative male reproductive tactics in a natural
population of prairie voles Microtus ochrogaster. Acta Theriologica, 55,
261–270.
McLoughlin, P. D., S. H. Ferguson, & F. Messier. (2000). Intraspecific variation in
home range overlap with habitat quality: A comparison among brown bears.
Evolutionary Ecology, 14, 39–60.
Meadows, P. S. & J. I. Campbell. (1972). Habitat selection by aquatic invertebrates.
Advances in Marine Biology, 10, 271–382.
Michener, C. D. (1974). The Social Behavior of the Bees: A Comparative Study.
Cambridge, MA: Belknap Press of Harvard University Press.
Milner-Gulland, E. J. & J. R. Beddington. (1993). The exploitation of elephants for
the ivory trade: An historical perspective. Proceedings of the Royal Society
London, B, 252, 29–37.
Mock, D. W., T. C. Lamey, & D. B. A. Thompson. (1988). Falsifiability and the
information center hypothesis. Ornis Scandinavica, 19, 231–248.
Møller, A. P. (1987). Advantages and disadvantages of coloniality in the swallow,
Hirundo rustica. Animal Behaviour, 35, 819–832.
Monaghan, P. & N. B. Metcalfe. (1985). Group foraging in wild brown hares: Effects
of resource distribution and social status. Animal Behaviour, 33, 993–999.
Montgomery, W. I., W. L. Wilson, & R. W. Elwood. (1997). Spatial regulation and
population growth in the wood mouse Apodemus sylvaticus: Experimental
manipulations of males and females in natural populations. Journal of Animal
Ecology, 66, 755–768.
Moore, M. C., D. K. Hews, & R. Knapp. (1998). Hormonal control and evolution of
alternative male phenotypes: Generalizations of models for sexual differentia-
tion. American Zoologist, 38, 133–151.
Moore, M. C. & C. A. Marler. (1987). Effects of testosterone manipulations on non-
breeding season territorial aggression in free-living male lizards, Sceloporus jar-
rovi. General and Comparative Endocrinology, 65, 225–232.
Morton, E. S., L. Forman, & M. Braun. (1990). Extrapair fertilizations and the evo-
lution of colonial breeding in purple martins. Auk, 107, 275–283.
Munshi-South, J., L. H. Emmons, & H. Bernard. (2007). Behavioral monogamy and
fruit availability in the large treeshrew (Tupaia tana) in Sabah, Malaysia.
Journal of Mammalogy, 88, 1427–1438.
Myers, J. P., P. G. Connors, & F. A. Pitelka. (1979). Territory size in wintering sander-
lings: The effects of prey abundance and intruder density. Auk, 96, 551–561.
Myrberg, A. A. Jr. (1972). Social dominance and territoriality in the bicolor damsel-
fish, Eupomacentrus partitus (Poey) (Pisces: Pomacentridae). Behaviour, 41,
207–231.
Nelson, R. J. (2000). An Introduction to Behavioral Endocrinology. Second Edition.
Sunderland, MA: Sinauer Associates.
252 ANIMAL BEHAVIOR

Newton, I. (1992). Experiments on the limitation of bird numbers by territorial

behaviour. Biological Reviews, 67, 129–173.
Nilsen, E. B., I. Herfindal, & J. D. C. Linnell. (2005). Can intra-specific variation in
carnivore home-range size be explained using remote-sensing estimates of envi-
ronmental productivity? Ecoscience, 12, 68–75.
Norman, M. D. & G. P. Jones. (1984). Determinants of territory size in the poma-
centrid reef fish, Parma victoriae. Oecologia, 61, 60–69.
Olsthoorn, J. C. M. & J. B. Nelson. (1990). The availability of breeding sites for
some British seabirds. Bird Study, 37, 145–164.
O’Neill, K. M. (1983). Territoriality, body size, and spacing in males of the beewolf
Philanthus basilaris (Hymenoptera: Sphecidae). Behaviour, 86, 295–321.
Ophir, A. G., J. O. Wolff, & S. M. Phelps. (2008). Variation in neural V1aR predicts
sexual fidelity and space use among male prairie voles in semi-natural settings.
Proceedings of the National Academy of Sciences, USA, 105, 1249–1254.
Ostfeld, R. S. (1985). Limiting resources and territoriality in microtine rodents.
American Naturalist, 126, 1–15.
Ostfeld, R. S. (1986). Territoriality and mating system of California voles. Journal of
Animal Ecology, 55, 691–706.
Parker, M. W., S. W. Kress, R. T. Golightly, H. R. Carter, E. B. Parsons, S. E. Schubel,
J. A. Boyce, G. J. McChesney, & S. M. Wisely. (2007). Assessment of social
attraction techniques to restore a common murre colony in central California.
Waterbirds, 30, 17–28.
Partridge, B. L. (1982). Structure and function of fish schools. Scientific American,
246, 114–123.
Peixoto, P. E. C. & P. De Marco Jr. (2009). No size or density effect on alternative
mate-locating tactics in the tropical damselfly Hetaerina rosea males (Odonata:
Calopterygidae). Revista de Biologia Tropica, 57, 361–370.
Penteriani, V., M. Ferrer, & M. M. Delgado. (2011). Floater strategies and dynamics in
birds, and their importance in conservation biology: Towards an understanding
of nonbreeders in avian populations. Animal Conservation, 14, 233–241.
Phelps, S. M. (2010). From endophenotypes to evolution: Social attachment, sexual
fidelity and the avpr1a locus. Current Opinion in Neurobiology, 20, 795–802.
Powell, R. A. (1987). Black bear home range overlap in North Carolina and the con-
cept of home range applied to black bears. International Conference on Bear
Research and Management, 7, 235–242.
Reichart, S. E. (1981). The consequences of being territorial: Spiders, a case study.
American Naturalist, 117, 871–892.
Rémy, A. (2011). Linking behaviour with individual traits and environmental condi-
tions, and the consequences for small rodent populations. Ph.D. dissertation,
University of Oslo, Oslo, Norway.
Richner, H. & P. Heeb. (1995). Is the information center hypothesis a flop? Advances
in the Study of Behavior, 24, 1–45.
Richner, H. & P. Heeb. (1996). Communal life: Honest signaling and the recruit-
ment center hypothesis. Behavioral Ecology, 7, 115–119.
MAKE SPACE ENOUGH BETWEEN YOU 253

Riehl, C. (2011). Living with strangers: Direct benefits favour non-kin cooperation in
a communally nesting bird. Proceedings of the Royal Society, London, B, 278,
1728–1735.
Robertson, D. R., H. P. A. Sweatman, E. A. Fletcher, & M. G. Cleland. (1976).
Schooling as a mechanism for circumventing the territoriality of competitors.
Ecology, 57, 1208–1220.
Rogers, L. L. (1987). The effects of food supply and kinship on social behavior, move-
ments, and population growth of black bears in northeastern Minnesota. Wild-
life Monographs, 97, 1–72.
Rolland, C., E. Danchin, & M. de Fraipont. (1998). The evolution of coloniality in
birds in relation to food, habitat, predation, and life-history traits: A compara-
tive analysis. American Midland Naturalist, 151, 514–529.
Rubenstein, D. I. (1981). Population density, resource patterning, and territoriality in
the Everglades pygmy sunfish. Animal Behaviour, 29, 155–172.
Sandell, M. (1989). The mating tactics and spacing patterns of solitary carnivores. In
J. L. Gittleman (ed.), Carnivore Behavior, Ecology, and Evolution (pp. 164–182).
Ithaca, NY: Comstock Publishers and Associates of Cornell University Press.
Schoener, T. W. (1983). Simple models of optimal feeding-territory size: A recon-
ciliation. American Naturalist, 121, 608–629.
Schradin, C. & A. K. Lindholm. (2011). Relative fitness of alternative male reproductive
tactics in a mammal varies between years. Journal of Animal Ecology, 80, 908–917.
Schradin, C., A. K. Lindholm, J. Johannesen, I. Schoepf, C. H. Yuen, B. König, &
N. Pillay. (2012). Social flexibility and social evolution in mammals: A case
study of the African striped mouse (Rhabdomys pumilio). Molecular Ecology,
21, 541–553.
Seghers, B. H. (1974). Schooling behavior in the guppy (Poecilia reticulata): An evolu-
tionary response to predation. Evolution, 28, 486–489.
Shuster, S. M. (2009). Sexual selection and mating systems. Proceedings of the
National Academy of Sciences, USA, 106, 10009–10016.
Shuster, S. M. (2011). Differences in relative fitness among alternative mating tactics
might be more apparent than real. Journal of Animal Ecology, 80, 905–907.
Shuster, S. M. & M. J. Wade (1991). Equal mating success among male reproductive
strategies in a marine isopod. Nature, 350, 608–610.
Siegel-Causey, D. & S. P. Kharitonov. (1990). The evolution of coloniality. Current
Ornithology, 7, 285–330.
Simon, C. A. (1975). The influence of food abundance on territory size in the iguanid
lizard Sceloporus jarrovi. Ecology, 56, 993–998.
Sinervo, B., D. B. Miles, W. A. Frankino, M. Klukowski, & D. F. DeNardo. (2000).
Testosterone, endurance, and Darwinian fitness: Natural and sexual selection
on the physiological bases of alternative male behaviors in side-blotched liz-
ards. Hormones and Behavior, 38, 222–233.
Smith, J. N. M. & P. Arcese. (1989). How fit are floaters? Consequences of alterna-
tive territorial behaviors in a nonmigratory sparrow. American Naturalist,
133, 830–845.
254 ANIMAL BEHAVIOR

Smith, S. M. & J. W. Cain III. ( 2009). Foraging efficiency and vigilance behavior of
impala: The influence of herd size and neighbor density. African Journal of
Ecology, 47, 109–118.
Solomon, N. G., A. R. Richmond, P. A. Harding, A. Fries, S. Jacquemin, R. L.
Schaefer, K. E. Lucia, & B. Keane. (2009). Polymorphism at the avpr1a locus
in male prairie voles correlated with genetic but not social monogamy in field
populations. Molecular Ecology, 18, 4680–4695.
Soma, K. K., K. Sullivan, & J. Wingfield. (1999). Combined aromatase inhibitor and
antiandrogen treatment decreases territorial aggression in a wild songbird
during the nonbreeding season. General and Comparative Endocrinology, 115,
442–453.
Soma, K. K. & J. C. Wingfield. (2001). Dehydroepiandrosterone in songbird plasma:
Seasonal regulation and relationship to territorial aggression. General and Com-
parative Endocrinology, 123, 144–155.
Sperry, T. S., C. K. Thompson, & J. C. Wingfield. (2003). Effects of acute treat-
ment with 8-OH-DPAT and fluoxetine on aggressive behavior in male song
sparrows (Melospiza melodia morphna). Journal of Neuroendocrinology, 15,
150–160.
Stamps, J. (1988). Conspecific attraction and aggregation in territorial species. Ameri-
can Naturalist, 131, 329–347.
Stamps, J. (1994). Territorial behavior: Testing the assumptions. Advances in the
Study of Behavior, 23, 173–232.
Steury, T. D. & D. L. Murray. (2003). Causes and consequences of individual varia-
tion in territory size in the American red squirrel. Oikos, 101, 147–156.
Stimson, J. (1973). The role of the territory in the ecology of the intertidal limpet Lot-
tia gigantea (Gray). Evolution, 54, 1020–1030.
Streatfeild, C. A., K. E. Mabry, B. Keane, T. O. Crist, & N. G. Solomon. (2011).
Intraspecific variability in the social and genetic mating systems of prairie voles,
Microtus ochrogaster. Animal Behaviour, 82, 1387–1398.
Sundström, L. F., M. Lõhmus, & J. I. Johnsson. (2003). Investment in territorial
defense depends on rearing environment in brown trout (Salmo trutta). Behav-
ioral Ecology and Sociobiology, 54, 249–255.
Sutherland, W. J. (1996). From Individual Behavior to Population Ecology. New York:
Oxford University Press.
Taborsky, M. (1998). Sperm competition in fish: “Bourgeois” males and parasitic
spawning. Trends in Ecology and Evolution, 13, 222–227.
Taborsky, M. (2001). The evolution of bourgeois, parasitic, and cooperative repro-
ductive behaviors in fishes. Journal of Heredity, 92, 100–110.
Taborsky, M., R. F. Oliveira, & H. J. Brockmann. (2008). The evolution of alterna-
tive reproductive tactics: Concepts and questions. In R. F. Oliveira, M. Tabor-
sky, & H. J. Brockmann (eds.), Alternative Reproductive Tactics: An Integrated
Approach (pp. 1–21). Cambridge, UK: Cambridge University Press.
Ten Eyck, G. R. (2008). Serotonin modulates vocalizations and territorial behavior in
an amphibian. Behavioural Brain Research, 193, 144–147.
MAKE SPACE ENOUGH BETWEEN YOU 255

Tinbergen, N., M. Impekoven, & D. Franck. (1967). An experiment on spacing-out

as a defense against predation. Behaviour, 28, 307–321.
Treherne, J. E. & W. A. Foster. (1981). Group transmission of predator avoidance
behaviour in a marine insect: The Trafalgar effect. Animal Behaviour, 29,
911–917.
Tricas, T. C. (1989). Determinants of feeding territory size in the corallivorous
butterflyfish, Chaetodon multicinctus. Animal Behaviour, 37, 830–841.
Turner, G. F. & T. J. Pitcher. (1986). Attack abatement: A model for group protec-
tion by combined avoidance and dilution. American Naturalist, 128, 228–240.
Uetz, G. W. & K. R. Cangialosi. (1986). Genetic differences in social behavior and
spacing in populations of Metepeira spinipes, a communal-territorial orb weaver
(Araneae, Araneidae). Journal of Arachnology, 14, 159–173.
van Schaik, C. P., M. A. van Noordwijk, B. Warsono, & E. Sutriono. (1983). Party size and
early detection of predators in Sumatran forest primates. Primates, 24, 211–221.
Vanson, A., A. P. Arnold, & B. A. Schlinger. (1996). 3ß-hydroxysteroid dehydrogen-
ase/isomerase and aromatase activity in primary cultures of developing zebra
finch telencephalon: Dehydroepiandrosterone as substrate for synthesis of
androstendione and estrogens. General and Comparative Endocrinology, 102,
342–350.
Veen, J. (1977). Functional and causal aspects of nest distribution in colonies of the
Sandwich tern (Sterna s. sandvicensis Lath.). Behaviour Supplement, 20, 1–193.
Vehrencamp, S. L. & J. S. Quinn. (2004). Joint laying systems. In W. D. Koenig &
J. L. Dickenson (eds.), Ecology and Evolution of Cooperative Breeding in Birds
(pp. 177–196). Cambridge, UK: Cambridge University Press.
Wagner, R. H. (1993). The pursuit of extra-pair copulations by female birds: A new
hypothesis of colony formation. Journal of Theoretical Biology, 163, 333–346.
Wagner, R. H., M. D. Schug, & E. S. Morton. (1996). Condition-dependent control
of paternity by female purple martins: Implications for coloniality. Behavioral
Ecology and Sociobiology, 38, 379–389.
Wang, Z., L. J. Young, G. J. De Vries, & T. R. Insel. (1998). Voles and vasopressin:
A review of molecular, cellular, and behavioral studies of pair bonding and
paternal behaviors. Progress in Brain Research, 119, 483–499.
Ward, M. P. & S. Schlossberg. (2004). Conspecific attraction and the conservation of
territorial songbirds. Conservation Biology, 18, 519–525.
Ward, P. & A. Zahavi. (1973). Importance of certain assemblages of birds as “infor-
mation centres” for food-finding. Ibis, 115, 517–534.
Watts, C. H. S. (1969). The regulation of wood mouse (Apodemus sylvaticus) numbers
in Wytham Woods, Berkshire. Journal of Animal Ecology, 38, 285–304.
Wauters, L. & A. A. Dhondt. (1992). Spacing behaviour of red squirrels, Sciurus
vulgaris: Variation between habitats and the sexes. Animal Behaviour, 43, 297–311.
Wauters, L. A & L. Lens. (1995). Effects of food availability and density on red squir-
rel (Sciurus vulgaris) reproduction. Ecology, 76, 2460–2469.
West, S. D. (1977). Midwinter aggregations in the northern red-backed vole,
Clethrionomys rutilus. Canadian Journal of Zoology, 55, 1404–1409.
256 ANIMAL BEHAVIOR

Wiktander, U., O. Olsson, & S. G. Nilsson. (2001). Seasonal variation in home-

range size, and habitat area requirement of the lesser spotted woodpecker (Den-
drocopos minor) in southern Sweden. Biological Conservation, 100, 387–395.
Wilson, E. O. (1971). The Insect Societies. Cambridge, MA: Belknap Press of Harvard
University Press.
Wingfield, J. (1994). Regulation of territorial behavior in the sedentary song sparrow,
Melospiza melodia morphna. Hormones and Behavior, 28, 1–15.
Wingfield, J. C., G. F. Ball, A. M. Dufty Jr., R. E. Hegner, & M. Ramenofsky.
(1987). Testosterone and aggression in birds: Tests of the “challenge hypothe-
sis.” American Scientist, 75, 602–608.
Wingfield, J. C., R. E. Hegner, A. M. Dufty Jr., & G. F. Ball. (1990). The “challenge
hypothesis”: Theoretical implications for patterns of testosterone secretion,
mating systems, and breeding strategies. American Naturalist, 136, 829–846.
Wingfield, J. C., S. Lynn, & K. K. Soma. (2001). Avoiding the “costs” of testoster-
one: Ecological bases of hormone-behavior interactions. Brain, Behavior and
Evolution, 57, 239–251.
Wittenberger, J. F. (1981). Animal Social Behavior. Boston: Duxbury Press.
Wittenberger, J. F. & G. L. Hunt Jr. (1985). The adaptive significance of coloniality
in birds. D. S. Farner, J. R. King & K. C. Parkes (eds.), Avian Biology, Vol. 8
(pp. 1–78). Orlando, FL: Academic Press.
Wittenberger, J. F. & R. L. Tilson. (1980). The evolution of monogamy: Hypotheses
and evidence. Annual Review of Ecology and Systematics, 11, 197–232.
Wolff, J. O. (1985). The effects of density, food and interspecific interference on
home range size in Peromyscus leucopus and Peromyscus maniculatus. Canadian
Journal of Zoology, 63, 2657–2662.
Wolff, J. O. (1993). Why are female small mammals territorial? Oikos, 68, 364–370.
Wolff, J. O. (2008). Alternative reproductive tactics in nonprimate male mammals.
In R. F. Oliveira, M. Taborsky, & H. J. Brockmann (eds.), Alternative Repro-
ductive Tactics: An Integrative Approach (pp. 356–372). Cambridge, UK:
Cambridge University Press.
Woodroffe, R. & A. Vincent. (1994). Mothers little helpers: Patterns of male care in
mammals. Trends in Ecology and Evolution, 9, 294–297.
Wright, D., L. B. Rimmer, V. L. Pritchard, R. K. Butlin, & J. Krause. (2003). Inter
and intra-population variation in shoaling and boldness in the zebrafish (Danio
rerio). Journal of Fish Biology, 63, 258–259.
Yahner, R. H. (2012). Wildlife Behavior and Conservation. New York: Springer.
Ylönen, H., T. Kojola, & J. Viitala. (1988). Changing female spacing behaviour and
demography in an enclosed breeding population of Clethrionomys glareolus.
Ecography, 11, 286–292.
Ylönen, H. & J. Viitala. (1991). Social overwintering and food distribution in the
bank vole Clethrionomys glareolus. Holarctic Ecology, 14, 131–137.
Zahavi, A. (1971). The function of pre-roost gatherings and communal roosts. Ibis,
113, 106–109.
 9

Let’s Get Together:

The Evolution of Social Behavior
Walter D. Koenig and Janis L. Dickinson

INTRODUCTION
Social behavior encompasses a wide variety of interactions among animals, usu-
ally of the same species, ranging from simple attraction between individuals, to
temporary feeding aggregations and mating swarms, to multigenerational family
groups with cooperative brood care. Although some types of social behavior are
intuitively obvious, others are not, because social behavior is defined by the kind
of interaction involved, not by how organisms are distributed in space. Thus,
individuals can be aggregated but not engaged in social behavior; conversely,
clumping is not a requirement for individuals to behave socially, although it
clearly increases the opportunity for meaningful social interactions to take place.
Examples of individuals engaging in social behavior while not appearing particu-
larly social include a lone female tiger moth (family Arctiidae) emitting a phero-
mone designed to attract potential male mates (Cardé & Millar, 2009) and a
male red deer (Cervus elaphus) roaring in order to signal dominance (Clutton-
Brock & Albon, 1979), attract females (Charlton et al., 2007), and enhance
female fertility (McComb, 1987).
One of the most important features of social behavior is that it is not neces-
sarily “social” in the positive sense of involving an amicable interaction or
being directed toward some mutually beneficial end. Indeed, social organisms
are often fiercely competitive and aggressive—sometimes toward the same
258 ANIMAL BEHAVIOR

individuals with whom they are highly cooperative under different circum-
stances. As one notable example, female acorn woodpeckers (Melanerpes formi-
civorus), a group-living bird found in the western United States, Mexico, and
Central America, nest communally (Koenig et al., 1995b). Two, and rarely
three, related females, usually sisters or a mother and daughter, lay eggs in
the same nest cavity. Even though they are relatives, females compete with
one another to lay their own eggs in the communal nest by removing and sub-
sequently eating their cobreeders’ eggs prior to laying their own eggs (Koenig
et al., 1995a), a process resulting in each female contributing equally to the
final clutch (Haydock & Koenig, 2002). After all birds have commenced lay-
ing, however, overt competition ends and is replaced by cooperative sharing of
incubation, brooding, and feeding of nestlings.
Social behavior has evolved to enhance an individual’s ability to garner re-
sources and form the alliances that help it survive and increase the number of
gene copies it places in future generations, whether doing so involves cooperating
with others or fighting against them. As a result, animal societies frequently
involve a delicate balance between cooperative and competitive behaviors, poten-
tially switching from one to the other depending on the ever-changing costs and
benefits of collaborative compared to competitive behaviors.

CATEGORIZING THE DIVERSITY OF SOCIAL BEHAVIOR

The diversity of behaviors that can be considered social renders the classifi-
cation of sociality difficult. Among insects, within which there is a wide range
of social behavior often within a single taxon, there is a long history of catego-
rizing sociality going back nearly a century based primarily on the involvement
of insect parents with their young (Michener, 1969, 2007). Two sequences are
recognized, the parasocial and subsocial routes, both encompassing a range
of societies that culminate in eusociality, a system in which overlapping gen-
erations of individuals live in the colony together, brood care is prolonged,
and there is a reproductive division of labor where at least some offspring act
as workers rather than reproduce.
In the parasocial route, females of the same generation assist one another, in
some cases cooperating in nest construction but otherwise rearing their broods
separately, whereas in others the broods are attended cooperatively, although
each female may still reproduce. In derived eusocial species, brood care is still
cooperative among the older generation, but generations overlap and some
members of the younger generation are workers that do not reproduce. Some
of these have achieved a complex and highly advanced eusocial system with
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 259

overlapping generations, functionally sterile workers, and distinct morphologi-

cal and behavioral castes.
In the subsocial route, females associate to a varying degree with their off-
spring, ranging from building a nest and laying an egg on a prey item that is
sufficient to allow the larva to develop into an adult, to providing direct care
for a prolonged period of time but then departing before the young emerge
as adults, to the situation where mothers stay at the nest until offspring emerge
and those newly emerged offspring then assist their mother in the rearing of
additional broods. As with the parasocial route, this route culminates in cases
where the offspring differentiate into a permanently sterile worker caste, a
stage mirroring the eusocial outcome achieved by the parasocial sequence.
Although useful, much of this terminology has not found wide favor outside
of entomological circles and even there is not easily applied to the entire euso-
ciality continuum (Sherman et al., 1995; Michener, 2007).
Apart from the attempt to classify routes to eusociality, however, there are few
other general categorizations of social behavior. One of the more comprehensive
attempts, encompassing not only reproduction but a wide range of social behav-
ior, is by the American entomologist and science writer Edward O. Wilson, who
compiled a set of 10 essential qualities of sociality (Wilson, 1975). These
included (1) group size, (2) distributions of different age and sex classes, (3) cohe-
siveness, (4) amount and pattern of connectedness, (5) permeability, or the
degree to which societies interact with one another, (6) compartmentalization,
or the extent to which subgroups operate as discrete units, (7) differentiation of
roles among group members, (8) integration of behaviors within groups, (9)
communication and information flow, and (10) the fraction of time devoted to
social behavior as opposed to individual maintenance. The overlapping nature
of these items provides a good indication of the complexities involved with clas-
sifying the highly varied social behavior exhibited by animals.
While categorizing social behavior can be a useful exercise, such classifications
can also be confusing and misleading. The current tendency is to view sociality as
a multifaceted continuum from simple aggregations to the highly organized and
complex levels of social organization found in the most derived eusocial species.

PROXIMATE VERSUS ULTIMATE CAUSATION

Social behavior, like other adaptations for survival and reproduction, is
best understood by differentiating its proximate causes—how the behavior
arises—from its ultimate causes—its evolutionary history and functional
260 ANIMAL BEHAVIOR

utility. Proximate causes include the mechanisms directly underlying the

behavior and include the hereditary, developmental, structural, cognitive,
psychological, and physiological aspects associated with the behavior. For
example, chacma baboons (Papio ursinus) live in relatively large social groups
that are subject to high rates of predation by lions, leopards, crocodiles, and
hyenas. Adult female baboons experience considerable stress following preda-
tor attacks, particularly when attacks result in animals becoming dispersed
and separated from the larger troop (Cheney & Seyfarth, 2009). Stress is indi-
cated by significantly high glucocorticoids in their blood—hormones released
by the adrenal glands in reponse to corticotropic-releasing hormones coming
from the hypothalamus when the brain perceives a physical or psychological
stressor (Sapolsky, 1998). High levels of circulating glucocorticoids have the
effect of increasing glucose availability in the bloodstream by promoting glu-
cose production and limiting glucose uptake and storage, thus mobilizing
energy, sharpening memory, and limiting nonessential but energetically costly
functions like digestion, growth, and reproduction (Nelson, 2011). The con-
sequence of these behaviors is to increase both the wariness of the animal
and its ability to search for and reunite with its troop, changes that increase
its chances of surviving under conditions of relative vulnerability to predation.
What causes these behaviors? At one level of analysis, these fear-based reac-
tions are the result of hormonal changes, specifically stress responses in the
hypothalamic-pituitary-adrenal (HPA) axis triggered by the predator attacks.
This underlying hormonal mechanism, which is triggered by the predator
attacks and separation from the herd, is a proximate cause of these and other
fear-based behaviors. At another level of analysis, the ultimate cause of these
behaviors is a range of selective processes that have shaped their past and cur-
rent expression. In this example, the increased survival that arises as a function
of reuniting with the troop has selected for individuals that exhibit and
respond to high levels of glucocorticoids following predator attacks by fleeing
and then quickly seeking to rejoin the group.
The Dutch ethologist Nikolaas Tinbergen was the first to clarify these levels
of analysis (or levels of explanation) and named four, which he referred to
as “survival value,” “causation,” “development,” and “evolutionary history”
(Tinbergen, 1963; see Volume 1, Chapters 1 and 2). Tinbergen further
emphasized the importance of addressing questions at the appropriate level
of explanation. The critical lesson to keep in mind is that determining the
underlying mechanism of a behavior does not, nor can it, address hypotheses
regarding the behavior’s historical origin or current survival value. Similarly,
understanding a behavior’s current survival value does not provide insight into
the cause of the behavior in terms of the proximate mechanisms that produce
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 261

it. Several examples of the confusion that can arise when mixing levels of
analysis by attempting to address evolutionary questions at an inappropriate
explanatory level are described by Sherman (1988), while a discussion of
how levels can complement each other is provided by MacDougall-
Shackelton (2011).

THE PROXIMATE MECHANISMS OF SOCIAL BEHAVIOR

Genes and Behavior: The Nature-Nurture Debate
Proximate mechanisms trigger the onset of a behavior. For example, rats
(Rattus norvegicus) whose ovaries and adrenal glands have been removed re-
cover some aspects of their sexual behavior when they are injected with the
hormone estradiol (Davidson et al., 1968), and testosterone implants induce
prenuptial molt—a key signal for mate choice by females—in male superb
fairy-wrens (Malurus cyaneus) (Peters et al., 2000). As discussed above, proxi-
mate mechanisms do not explain the evolutionary basis of a behavior: the fact
that testosterone causes male fairy-wrens to molt does not tell us anything
about the function of bright breeding plumage in this species. Such informa-
tion does, however, provide insight into the ways in which organisms are
adapted to perform these and other intricate and complex social behaviors.
As an example of genetic effects on an important form of social communi-
cation, male zebra finches (Taeniopygia guttata) exposed to the song of an unfa-
miliar male initiate the expression of a transcription factor-encoding gene (egr1)
in the auditory forebrain devoted to hearing, suggesting that neural processes
linked to song pattern are important to song recognition, discrimination, and
the formation of auditory associations (Mello et al., 1992). This same gene is also
important in achieving the neural plasticity that allows subordinate male Astato-
tilapia burtoni (a fish in the family Cichlidae) to become dominant within
minutes of an opportunity to do so, during which they exhibit dramatic changes
in body coloration and behavior (Burmeister et al., 2005).
Relatively few genes have been identified thus far that directly influence
social organization, but many clearly await discovery. One notable case is
found in European honey bees (Apis mellifera), where the switch from working
in the hive to foraging that takes place when bees are about two to three weeks
old is associated with a significant increase in expression of the foraging gene,
which encodes a protein kinase inducing changes in brain structure and
behavior (Ben-Shahar et al., 2002). Another example is the Gp-9 gene in the
fire ant (Solenopsis invicta), which encodes a pheromone-binding protein cru-
cial in chemical recognition of conspecifics that determines their ability to
262 ANIMAL BEHAVIOR

recognize queens and regulate their numbers—a social polymorphism that has
been key to the explosive range expansion of this species (Krieger & Ross,
2002).
Even more dramatic genetic effects on social behavior have been identified
in the social amoeba (Dictyostelium discoideum), which grows solitarily while
feeding but when starved cooperatively forms a functionally multicellular
“slug” consisting of dead stalk cells and viable spores. A genome-wide screen
detected mutations in over 100 genes that facultatively produced cheaters—
cells that disproportionately attempt to differentiate into spores rather than
the stalk during aggregation when in competition with other genotypes—
but not when clonal (Santorelli et al., 2008). Such behavioral genetics studies
demonstrate the complexities of genetic influences on social behavior and are
part of a growing field with significant potential for uncovering new informa-
tion on the relative importance of nature and nurture to behavioral develop-
ment and social behavior (Robinson et al., 2008).
A classic debate in behavioral circles is whether genetic predisposition or the
environment is more important to the development of social behavior. As we
hope is clear from the above discussion, brief as it is, this “nature-nurture”
debate, stretching back decades, has been resolved: both are invariably impor-
tant, and the interesting questions generally focus on how nature and nurture
interact (Rutter, 2006).

Vocal Learning
Environmental influences on behavior include factors like nutrition, hor-
mones, an animal’s experience of the outside world, and the social context in
which it lives. The development of bird song provides key examples of several
of these. Songbirds (order Passeriformes), in contrast to all but a handful of
other groups that include cetaceans, bats, humans, and two other avian
orders—parrots and hummingbirds—learn their vocal signals without having
to hear model signalers (Beecher & Brenowitz, 2005). This feat of learning
is accomplished by the presence of two main neural pathways in the songbird
brain. The first is a motor pathway involved in song production, and the sec-
ond is a pathway in the anterior forebrain, often referred to as a template, that
is involved in song learning and recognition. In some species, learning is
restricted to the first year of life. In others, learning is open-ended and contin-
ues long after the first year. The experience of hearing and practicing a song
provides individuals with the necessary link between the auditory and song
systems required to sing properly as adults.
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 263

In the white-crowned sparrow (Zonotrichia leucophyrs), for example, young

males exposed to conspecific (same-species) tutor songs during a critical period
of 10 to 50 days of age develop normal songs, whereas individuals raised in iso-
lation or exposed to allospecific (other-species) songs during the critical period
do not. It was originally thought that conspecific tutoring allowed birds to
learn by selectively matching their vocal output to their internal, species-
specific template (Marler, 1970; Beecher & Brenowitz, 2005). Subsequent
studies demonstrated that birds are, in fact, capable of learning allospecific
songs, including the radically different song of the unrelated strawberry finch
(Amandava amandava), over an extended period beyond 50 days of age if a
live, rather than a tape-recorded, tutor is used (Baptista & Petrinovich,
1984). More recent work on song sparrows (Melospiza melodia) suggests that
young birds may be even more likely to learn songs from eavesdropping on
songs of tutors that birds overhear singing with another young individual than
from hearing a live tutor with which they directly interact (Beecher et al.,
2007). Meanwhile, studies on other species have led to the conclusion that
there is considerable variation in songbird song learning along a variety of
dimensions, including when the song is learned, how faithfully birds copy
tutors, the importance of early song experience, and the degree to which birds
will copy tutor material when it is not produced by or similar to the song of
conspecifics (Beecher & Brenowitz, 2005).
The song of the zebra finch provides a good illustration of hormonal influ-
ences on song development and singing behavior. There is marked sexual
dimorphism in brain development in this species resulting in strikingly greater
growth in the vocal-control areas of the brain in males (Nottebohm & Arnold,
1976). Injecting females with estrogen early in life causes them to develop
male-like brains and male-typical songs, whereas females given testosterone
in adulthood do not. Castrated males, on the other hand, still develop mascu-
linized song nuclei and a song attractive to females (Wade & Arnold, 1996).
This indicates that estogens, rather than androgens (i.e., testosterone), are
the primary determinant of sexual differentiation in this species (Adkins-
Regan & Ascenzi, 1990). The effects of estogens appear to be primarily regu-
lated by factors intrinsic to the brain, most likely by the expression of genes
that influence the levels of estrogen synthesized in the brain or the responsive-
ness of brain tissue to circulating levels of estrogen (Wade & Arnold, 2004).
The critical importance of social influences on behavioral development can
be seen throughout the period of song learning in the song sparrow, in which
each male sings 5 to11 different song types (Nordby et al., 2000). There is a
sensitive period in the first summer of life when young birds learn many
264 ANIMAL BEHAVIOR

of their songs, but field studies show that learning also continues through the
first year by storing copies of older neighbors’ songs in a region of the brain
called the forebrain song nuclei. A young song sparrow occupying a territory
learns the songs of several of its neighbors, being more likely to learn song types
that are shared among two or more neighbors. The end result is that each song
sparrow holds roughly half its song types in common with neighbors, particularly
older males that have been present longer. The adaptive function of such song-
sharing behavior may be that it facilitates the rapid detection of intruders
(Beecher et al., 1997). One recent study, for example, found that neighboring
song sparrows sharing fewer songs were more aggressive with one another than
those sharing more songs (Wilson & Vehrencamp, 2001).

Kin Recognition
Another good example of the importance of proximate mechanisms to
social behavior is the study of recognition systems. Mechanisms of recognition
are essential for individuals to discriminate members of their own social group,
choose a mate of the appropriate sex, avoid incest, locate their parents, care for
the right offspring, and engage in nepotism—the preferential treatment of kin.
Early work in this area, pioneered by the Austrian zoologist Konrad Lorenz,
demonstrated how precocial birds such as geese imprint on their mothers
shortly after hatching (Lorenz, 1935; Hess, 1964).
A particularly well-studied area of recognition systems focuses on how
parents recognize their offspring. Such offspring recognition probably involves
odor in most insects and mammals. Adult birds, however, often use location of
the nest as the cue to recognizing their offspring and fail to discriminate
against the nestlings of other pairs that are artificially fostered into their nests.
Presumably the failure to evolve recognition in this context is due to the fact
that nestlings do not move around from one nest to another in the wild.
In species where the mixing of nestlings does occur, however, recognition
can be remarkably precise. In Mexican free-tailed bats (Tadarida brasiliensis),
for example, mothers exhibit spatial memory and return to within a short dis-
tance of where they last left their pups. Once in the vicinity, they are capable
of using both scent and vocalizations to locate their offspring among thou-
sands huddled in a small area of cave ceiling (Gustin & McCracken, 1987;
Balcombe, 1990). In penguin colonies, where there can also be thousands of
young huddled in crèches, offspring use vocals cues to recognize their parents.
Such cues are based on pitch alone in species that build nests, such as Adélie
penguins (Pygoscelis adeliae) and gentoo penguins (P. papua), but involve more
sophisticated acoustic cues including the frequency-modulated shape of calls
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 265

and the temporal succession of syllables in non-nest-building species such as

king penguins (Aptenodytes patagonicus) and emperor penguins (A. forsteri)
(Aubin & Jouventin, 2002; Jouventin & Aubin, 2002).
Kin recognition systems play a particularly important role in the context of
nepotism. Three basic mechanisms are recognized that animals can potentially
use to recognize kin: environmental cues, prior experience, and phenotype
matching—matching the look or smell of an individual to an internal tem-
plate independent of environmental or other external cues (Sherman et al.,
1997). Whichever is used, the mechanism must deal with the possibility of
recognition errors, balancing the possibility of being too restrictive and
thereby rejecting some individuals that are in fact kin, with that of being too
accepting, thereby increasing the probability of treating as kin individuals that
are in fact unrelated (Reeve, 1989).
After leaving the nest, some songbirds use vocalizations to recognize individ-
uals, and in a few species, such as the long-tailed tit (Aegithalos caudatus), they
favor kin when deciding whether and where to help feed nondescendent kin,
possibly based on calls learned in the nest that show distinct signatures of kinship
(Sharp et al., 2005). In Belding’s ground squirrels (Urocitellus beldingi), a species
in which nepotism is common, kin recognition involves both learning the phe-
notypes of related individuals during early development and later discriminating
these familiar relatives from other unfamiliar individuals that are presumably
nonrelatives. This is accomplished by a combination of prior association and
the learning of their own phenotype, which they later compare to unknown indi-
viduals via phenotype matching (Holmes & Sherman, 1982). Experiments have
demonstrated that phenotype matching using odors in this species is very precise,
allowing individuals to recognize first-order (mothers related by r ¼ 0.5), second-
order (grandmothers related by r ¼ 0.25), and even third-order (half-aunts
related by r ¼ 0.125) relatives (Mateo, 2002). Kin recognition is maintained
even after long periods of hibernation (Mateo, 2010).
In several systems the genetic basis of kin recognition has been decoded. In
cockroaches (Blattella germanica), for example, individuals at all developmen-
tal stages discriminate siblings from nonsiblings independent of prior associa-
tion, preferring siblings as social partners and nonsiblings as mating partners.
Kin recognition is based on quantitative differences in cuticular hydrocarbons
present in their exoskeletons perceived through their anntenae (Lihoreau &
Rivault, 2009). In several vertebrates, sophisticated kin recognition systems
appear to be based on highly polymorphic genes known as the major histo-
compatibility complex (MHC). MHC genes encode cell-surface glycopro-
teins that bind piptides and present them to lymphocytes, thereby
harnessing the mechanism for immunological self-recognition (Wedekind &
266 ANIMAL BEHAVIOR

Penn, 2000). MHC genes have been implicated in determining odor and
mating preferences in house mice (Mus musculus) and humans. In both cases,
sexual selection appears to be involved, with individuals tending to prefer
MHC-dissimilar mates (Potts et al., 1991; Wedekind & Penn, 2000; Jacob
et al., 2002).

THE ULTIMATE CAUSES OF SOCIAL BEHAVIOR

While the fitness advantages of behaviors such as mating and caring for off-
spring are obvious in that they increase the number of an individual’s own
young, those of social behaviors such as living in groups and helping others
are not. Because such behaviors are complex and paradoxical, their ultimate
cause remains a key focus of evolutionary biologists.
Social interactions can be characterized as mutualism (when both individ-
uals benefit by the behavior), altruism (when one individual, the altruist,
makes a sacrifice while the other, the recipient, benefits by the behavior), self-
ishness (when the individual performing the behavior benefits at the expense
of the recipient), and spite (when one individual performs a costly behavior
that hurts the recipient, and thus both pay a cost). Mutualisms pose no para-
dox because both individuals benefit by the interaction. Altruism is more
problematical. When individuals behaving altruistically are genetic relatives,
as is often the case, kin selection is a likely explaination, with altruistic individ-
uals gaining indirect fitness benefits by helping relatives produce additional
offspring. As discussed in Chapter 11 of this volume, the conditions for kin
altruism to evolve are summarized by Hamilton’s rule rB > C, where B is the
benefit to the recipient, C is the cost to the altruist, and r is the coefficient of
relatedness between the two (Hamilton, 1964).
Altruism between unrelated individuals is relatively rare, but it occurs and
cannot be explained by kin selection. One solution to the paradox of an indi-
vidual making sacrifices for a second unrelated individual is reciprocal altru-
ism (Trivers, 1971). If individuals interact repeatedly, altruism can be
favored as long as the altruist receives a reciprocal benefit outweighing its
initial cost.
Although reciprocal altruism is considered a potentially important evolu-
tionary force, a critical caveat is that it can only work if there is a mechanism
to punish “cheaters” that accept help without reciprocating. Consequently,
unambiguous examples of reciprocal altruism outside of humans are rare. Ap-
parent examples include male coalition formation in baboons (Papio anubis)
(Packer, 1977), food sharing in vampire bats (Desmodus rotundus) (Wilkinson,
1984), and the interactions between cleaner fish (Labroides dimidiatus) and the
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 267

“client” fish they attend, which involves complex indirect reciprocity (reci-
procity by individuals other than those originally helped) by which clients
eavesdrop on the cleaning behavior of cleaner fish and subsequently prefer to
spend more time with cleaners that engage in cooperative rather than unco-
operative cleaning (Bshary & Grutter, 2006). The possibility remains that
such forms of complex reciprocity, which are not easy to detect, may be more
common than currently recognized.
Considerable recent work has involved investigating the evolution of altru-
ism among unrelated individuals in humans, typically by using “public goods”
games in which subjects, usually university students, voluntarily choose a fraction
of their “private” goods—money or a money proxy—to add to a public pot that
is subsequently multiplied by some factor greater than one but less than the
number of players and then divided up among all participants. Although the
group’s total payoff is maximized when all players contribute all their money to
the public pot, the optimal choice for individual players is always to contribute
nothing. Thus, contributors are considered altruistic cooperators whereas non-
contributors are considered defectors or noncooperators.
The typical result of such games is that individuals initially cooperate and
contribute some fraction of their money to the public pot. Because partici-
pants are unrelated and often neither know each other nor are likely to interact
in the future, neither kin selection nor reciprocity are involved. In the absence
of some additional mechanism, however, the extent of cooperation invariably
declines rapidly to nearly zero when interactions are repeated and the game
is played iteratively.
One mechanism that potentially maintains altruistic behavior is altruistic
punishment, in which individuals are given a means to punish noncooperators
at a cost to themselves (Fehr & Gächter, 2002). Altruism can also be main-
tained by the advantages of maintaining a good reputation, a behavior predi-
cated on indirect reciprocity such as was observed in the cleaner fish example
discusssed above. Public goods games that incorporate the ability for players
to see the history of other players and act on that knowledge—thus allowing
for indirect rewarding of cooperators—result in a high level of cooperation
compared to games not incorporating such information (Milinski et al.,
2002). The evolution of cooperation by indirect reciprocity can be a potent
evolutionary force leading to reputation building, morality judgement, and
complex social interactions (Nowak & Sigmund, 2005).
Selfish behavior, when one individual benefits at the expense of another, is,
perhaps unsurprisingly, not difficult to find. In birds, for example, it is fairly
common, although often difficult to observe, for females to lay eggs in nests
other than their own, thereby parasitizing the parental care of others. Such
268 ANIMAL BEHAVIOR

brood parasitism can be interspecific, as in European cuckoos (Cuculus cano-

rus) and the North American brown-headed cowbird (Molothrus ater) (Davies,
2000) or intraspecific (also known as conspecific brood parasitism), the latter
found in birds (Yom-Tov, 1980; Lyon & Eadie, 2008) and insects (Tallamy,
2005). Such parasitic behavior would undoubtedly be even more common
were it not for strong counterselection leading to the evolution of a diverse
array of defenses designed to guard against or at least reduce the impact of
brood parasites (Rothstein, 1990; Davies, 2000).
Spite as a social interaction has traditionally been considered to be at best
unlikely and at worst paradoxical (Gadagkar, 1993), but current thinking is
that it can evolve by what amounts to the inverse of Hamilton’s rule: specifi-
cally, when two individuals are negatively related to each other (that is, less
related than the average relatedness between two individuals in the popula-
tion) and the cost to the actor is smaller than the product of the cost to the
recipient and its (negative) relatedness to the actor (West & Gardner, 2010).
One apparent example is found in the sterile soldier caste of the polyem-
bryonic parasitoid wasp (Copidosoma floridanum), in which most larvae
develop normally but a few become a sterile soldier morph that seek out and
preferentially kill larvae from other eggs, thereby freeing up resources for their
clone-mates (Garner et al., 2007). Although such “evolutionary spite” is still
thought to be quite rare, “functional spite,” in which an individual performs
a costly behavior that harms the recipient but that gains the actor some future
benefit such as increased social dominance, parasite deterrence, or offspring or
sexual partner coercion, is probably much more widespread (Jensen, 2010).

THE RANGE OF SOCIAL BEHAVIOR

Living in Groups: Costs and Benefits
Interacting with other individuals is inherently dangerous and potentially
costly. Reasons for this include an increased probability of parasite and disease
transmission, increased competition for food and other resources, increased
competition for mates, and increased interference among conspecifics. These
costs have no automatic counteracting benefits (Alexander, 1974). Why, then,
are animals ever social, and why do they live in groups? Clearly there must be
potential benefits of aggregations that outweigh the inherent costs associated
with being in close proximity with conspecifics.
On the positive side, aggregations may provide individuals with increased
access to food through information sharing, cooperative defense against pred-
ators or non-group members, or simply energy conservation by huddling
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 269

during cold or inclement weather. In any case, once animals are aggregated,
there will inevitably be selection to evolve traits that will better exploit the
potential advantages of group living, including mechanisms allowing individ-
uals to better communicate, cooperate, and recognize each other as individuals
and as kin in ways that promote their inclusive fitness through increased survi-
vorship and reproductive success. For example, aggregated nymphs of the sub-
social treehopper Umbonia crassicornis produce vibrational signals in
synchronized bursts in response to simulated predator approach. These signals
not only initiate a wave of signaling by other individuals within the aggrega-
tion but instigate defensive behavior on the part of their mother (Cocroft,
1999). In Publilia concava, another species of treehopper, adults live in mutu-
alisms with ants by providing them with honeydew in return for protection
against predators. These treehoppers produce vibrational alarm signals follow-
ing encounters with the predatory ladybird beetle Harmonia axyridis that
increase the ants’ activity, thereby increasing their likelihood of discovering
the predator and the effectiveness of their predator protection (Morales
et al., 2008). Another classic example of a finely tuned communication system
that evolved in the context of group living is that of the honey bee, a highly
eusocial species in which workers returning to the hive perform elaborate
“dances” that increase foraging efficiency by communicating to hive-mates
the direction and distance of high-quality foraging sites (von Frisch, 1967;
Seeley, 1995).
Given these complexities, a fruitful way to address the evolution of sociality
in any particular case is to determine the costs and benefits of aggregating with
others. A good example of a species in which many of these costs and benefits
have been studied and identified is the cliff swallow (Petrochelidon pyrrhonota),
a small, migratory passerine that breeds in colonies of up to several thousand
pairs in western North America, primarily underneath overhanging rock
ledges, on the sides of cliffs, and (increasingly) on artificial structures such as
bridges (Brown & Brown, 1996).
Costs associated with colonial nesting in this species include increased sus-
ceptibility to ectoparasites, including both ticks and flies; increased loss of eggs
due to disturbance by conspecifics; increased probability of losing paternity
due to extrapair copulations; increased incidence of conspecific brood parasit-
ism; increased incidence of food stealing among conspecifics (kleptoparasi-
tism); greater attraction of predators; and increased travel distance to
foraging areas. Countering these are several benefits, including increased vigi-
lance and detection of predators resulting in greater annual survivorship;
greater opportunity for conspecific brood parasitism; increased probability of
gaining paternity due to extrapair copulations; and possibly the most
270 ANIMAL BEHAVIOR

significant benefit, that of birds unsuccessful at foraging actively seeking out and
subsequently following other birds breeding nearby in the colony whose foraging
efforts have been successful, a trait that confers considerable benefits given the
highly unpredictable and variable insect food resources on which cliff swallows
depend. Note that behaviors are in some cases simultaneously a cost for some
individuals but a benefit to others. For example, the incidence of extra-pair mat-
ings, which are more frequent in larger colonies, are a benefit for the males that
succeed in obtaining them and a cost to the cuckolded males. Similarly, conspe-
cific brood parasitism is beneficial to the females that lay the eggs but a cost to the
pair of birds that ends up raising a nestling that is not theirs.
Support for the hypothesis that a colony can act as an information center
(Ward & Zahavi, 1973) has been found not only in cliff swallows but in
osprey (Pandion haliaetus) (Greene, 1987), evening bats (Nycticeius humeralis)
(Wilkinson, 1992), and colonial seabirds (Weimerskirch et al., 2010). More gen-
erally, the information center hypothesis is an example of animals taking advan-
tage of public information—social or nonsocial information that is accessible to
others—for their own benefit. Studies suggest that the use of such public infor-
mation is potentially widespread, yielding important information used by ani-
mals to inform the choices they make in foraging, breeding habitat selection,
avoiding predators, mate choice, and the transmission of cultural traits (Danchin
et al., 2004). For example, in the collared flycatcher (Ficedula albicollis), birds
monitor the current reproductive success of conspecifics using such public infor-
mation to assess local habitat quality and to choose their own subsequent breed-
ing site when they return to an area in a subsequent year (Doligez et al., 2002).
The costs and benefits of group living vary from individual to individual
depending on a variety of factors including their age, sex, and status, factors
that are often conveniently summarized by the concept of dominance, a key
way in which individuals within groups are unequal. As dominant individuals
monopolize a larger fraction of a group’s resources, group living becomes less
beneficial for subordinate individuals in the same group, which then become
more likely to leave and try to live on their own. In order for sociality to be
maintained in a population, subordinates must gain more from remaining in
the group and being social than from leaving the group and trying to repro-
duce on their own. In the case of the cliff swallows, the delicate balance
between these alternatives results in wide variation in group sizes ranging from
solitary nesting to nesting in colonies of several thousand pairs.

Passive Aggregations, the Selfish Herd, and the Dilution Effect

A general hypothesis for why individuals might aggregate is predicated on
the importance of predation and suggests that animals come together to form
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 271

a so-called selfish herd in which individuals do not directly cooperate with

each other but nonetheless passively benefit because each individual’s chances
of being eaten are reduced (the dilution effect), especially for individuals in
the interior of the group (Hamilton, 1971). For example, groups of ocean
skaters (Halobates robustus), a small marine insect that lives on the ocean sur-
face, are depredated by juvenile pilchards (Sardinops sagax) at a rate that
decreases linearly with increasing group size (Foster & Treherne, 1981).
Experiments quantifying shark attacks on rafts of seal decoys also indicate that
danger decreases proportionately with group size, thus supporting the impor-
tance of the selfish herd effect (De Vos & O’Riain, 2010).
In the simplest example, when a group-living individual encounters a predator
that will eat just one prey item, its likelihood of being eaten in a group of size n is
reduced from p, the probability when alone, to p/n. This can yield a strong ben-
efit quickly to individuals in aggregations, even if predators are differentially
attracted to larger groups, so long as the cost of increased conspicuousness does
not overtake the benefit of dilution. Where location within the group matters,
social interactions will likely sort out by social status, with some individuals gain-
ing favored positions—typically central rather than peripheral positions—by
dominance or nepotism (Krause & Ruxton, 2002).

Social Behavior Based on Protection against Predators and Intruders

Although passive protection against predators can be achieved by any spe-
cies through the selfish herd and dilution effects, social groups forming to
engage in active cooperative defense are more unusual. Examples include
muskoxen (Ovibos moschatus), which form tight rings and face outwards in
defense against wolves (Canis lupus) (Tener, 1965; Heard, 1992); mule deer
(Odocoileus hemionus), which bunch together with conspecifics and aggres-
sively defend against coyotes (Canis latrans) (Lingle, 2001); and guanacos
(Lama guanicoe), which cooperatively defend against the culpeo fox (Lycalopex
culpaeus) (Novaro et al., 2009). Cooperative defense is also found in inverte-
brates such as the tent caterpillar (Malacosoma disstria), which exhibits collec-
tive defense again parasitoids in the form of collective head flicking and
biting by groups of instars (McClure & Despland, 2011), and group-living
larvae of the Australian sawfly (Perga affinis), which store noxious Eucalyptus
oils in their foregut that they regurgitate when attacked by ants, birds, and
mice (Morrow et al., 1976). As another example, resident males of several spe-
cies of territorial fiddler crabs (Uca spp.) will leave their territories and co-
operate with neighbors to repel intruders, apparently because the benefits of
maintaining neighborhood stability outweigh the costs of repelling intruders
(Booksmythe et al., 2012).
272 ANIMAL BEHAVIOR

Alarm calls and other complex signaling behavior within aggregations can
also reduce the likelihood of predation. Such signaling may coordinate a
group’s escape from danger, confuse a predator, and prompt other nearby
individuals to seek protected sites or shelter. Alarm calls may convey informa-
tion about the type of predator and lead to the appropriate evasive behavior.
For example, vervet monkeys (Chlorocebus pygerythrus) give short, tonal alarm
calls in response to leopards (Panthera pardus); low-pitched, staccato grunts in
response to martial eagles (Polemaetus bellicosus); and high-pitched, chatter-
like calls in response to python snakes (Python sebae), prompting other mon-
keys nearby to run into trees for leopards, look up for eagles, and look down
for snakes (Seyfarth et al., 1980). Alarm calls might even provide information
regarding an individual predator’s identity and habits, although this has yet to
be demonstrated.
Alarm calling is often considered a good example of an altruistic behavior
because it appears to benefit others at some cost to the calling individual. Such
calls provide potentially valuable information to others but may endanger the
caller by attracting predators, although the evidence for the latter is equivocal,
possibly because callers are in some cases simultaneously serving as a pursuit-
deterrence signal, communicating to the predators their ability to evade cap-
ture (Blumstein, 2007).
An example of a species whose alarm-calling behavior has been extensively
studied is Belding’s ground squirrel. Individuals call more frequently when
close relatives, including noncollateral kin, are nearby, thus suggesting that
alarm calling has evolved at least in part due to kin selection—specifically
the indirect fitness benefits of aiding relatives (Sherman, 1977). Similar results
have been reported for a variety of other rodents, including several other
ground squirrels, the chipmunk (Eutamias sonomae), and both black-tailed
prairie dogs (Cynomys ludovicianus) and Gunnison’s prairie dogs (C. gunnisoni)
(Blumstein, 2007). Alarm calls are given in other contexts besides those likely
to entail kin selection, however. For example, those given by birds in mixed-
species flocks—common in both birds and monkeys—in some cases appear
to convey information regarding the proximity of a predator (Terborgh,
1990; Sharpe et al., 2010), while in others they may be used to selfishly dis-
tract flockmates and increase the caller’s foraging efficiency (Munn, 1986).
Aggregations have the potential to augment and bolster signaling systems.
This is particularly true in aposematic species that advertise their defenses to
potential predators (Orians & Janzen, 1974). There are several potential rea-
sons for this, including proportionately lower detectability, faster learning of
the aposematic signal by predators, and increased effectiveness of the apo-
sematic signal (Gamberale & Tullberg, 1998; Riipi et al., 2001). Groups of
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 273

animals may also confuse predators by looking larger than they actually are or
by moving apart in unpredictable ways (Krakauer, 1995), actions that may
cause a predator to hesitate just long enough to permit the prey’s escape.
Grouping may also serve less direct functions than reducing predation risk.
For example, the remarkable shift of the desert locust (Schistocerca gregaria) of
sub-Saharan Africa from its cryptic solitary phase to its notorious gregarious
phase, during which it can form huge swarms of individuals moving cohesively
in search of food covering as much as 200 km2, has been suggested to be a
behavioral strategy to reduce predation risk by mobile predators that would
otherwise be able to profitably forage in adjacent patches within which locust
densities are relatively high (Reynolds et al., 2009).

Social Behavior Based on Increased Foraging Efficiency

Individuals in groups may benefit by cooperating to gain access to food and
other resources. In the fruit fly (Drosophila melanogaster), for example, males
and mated females deposit an aggregation pheromone on rotting fruit that
induces aggregated oviposition; larvae subsequently feed on yeasts that develop
on the fruit. The apparent benefit of facilitating group oviposition is that at
low larval densities, fungi and molds can outcompete the larvae, resulting in
high larval mortality (Wertheim et al., 2002). Thus, larvae fare better as initial
colonists of fruit when they live in groups. This same strategy can help insect
larvae overcome plant defenses, such as in the sawfly (Neodiprion pratti),
where larvae feed on pine needles and individuals aggregate around the few lar-
vae that succeed in cutting through the tough needle cuticle (Ghent, 1960). In
other plant-feeding insects, feeding itself may affect the quality of the food
when substances in the insect’s saliva that overcome chemical defenses or alter
the metabolism of the host plant allow more efficient (for the insect) release of
nutrients (Després et al., 2007).
When predators hunt in groups, their prey may become confused, leading
to a beater effect whereby prey flushed out by group activity become easy to
capture (Swynnerton, 1915). Where predators hunt cooperatively, such as in
the foraging of large carnivores like lions, hyenas, and wolves, they can corner
and bring down prey more easily (Packer & Ruttan, 1988). This also occurs in
the group-living Harris’s hawk (Parabuteo unicinctus), where birds form hunt-
ing parties of two to six, resulting in the ability to kill prey larger than them-
selves and flush out rabbits hiding in thick cover, behaviors that result in
improved capture success and increased energy intake (Bednarz, 1988).
Group living can select for sophisticated systems of communication and co-
operation that enhance overall foraging success. A good example is the highly
274 ANIMAL BEHAVIOR

gregarious eastern tent caterpillar (Malacosoma americanum). In this species,

hungry individuals that have not yet started to feed follow silk-and-chemical trails
laid down by successful foragers returning to their communal tents to attractive
feeding sites containing young leaves that are both nutritionally superior and
offer caterpillars the ability to better repel predatory ants by regurgitating plant-
derived defensive compounds (Peterson et al., 1987; Fitzgerald & Peterson,
1988). An even more sophisticated example is the complex dance of honey bees
already mentioned, where successful foragers returning to the hive pass along
information on both the direction and distance of food sources to other workers.
Such dances, although extraordinary by themselves, are only one component of
the complex communication system of this species, which includes chemosen-
sory as well as spatial cues (Grüter & Farina, 2009).

Social Behavior Based on Reproduction

In some species, aggregations are based exclusively on mating. These
include various insect aggregations such as are found in some bees and wasps
(order Hymenoptera), flies (order Diptera), and butterflies (superfamily
Papilionoidea), in which females congregate at conspicuous landmarks
(Thornhill & Alcock, 1983), and the aggregations of males at leks—display
sites used only for mating (Höglund & Alatalo, 1995). The latter is known
from nearly 100 species of birds but has been reported in only 13 mammals,
nine of which are ungulates, along with the hammer-headed bat (Hypsignathus
monstrosus) (Bradbury, 1977).
The selective benefits of lek aggregations are a matter of intense debate. One
hypothesis is that males congregate in sites where the home ranges of many
females overlap (the “hotspot” hypothesis) (Bradbury & Gibson, 1983). Alterna-
tively, female mate choice may be driving male aggegrations because leks provide
females with a convenient means to quickly assess male quality (Queller, 1987).
Yet another hypothesis is that leks form because subdominant and novice males
benefit by associating with highly successful, dominant males (the “hotshot”
hypothesis) (Beehler & Foster, 1988). One particularly intriguing discovery is
that leks in several species are composed of clusters of related kin, raising the pos-
sibility that kin selection and inclusive fitness benefits play a role in the evolution
of at least some leks (Höglund et al., 1999; Petrie et al., 1999; Shorey et al.,
2000). In the case of wild turkeys (Meleagris gallopavo), kin selection benefits
have been shown to be sufficient to explain the evolution of cooperative court-
ship whereby several males coordinate displays (Krakauer, 2005). Although only
the dominant male in a displaying coalition actually mates in this species, subor-
dinate males gain indirect fitness benefits greater than what they could expect to
achieve on their own.
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 275

In many cases the aggregation of one sex provides opportunities for the
other. For example, in species where females aggregate due to the clumping
of food or nest sites, males are likely to aggregate at these sites as well because
they are the most efficient places to find females with which to mate. In still
others, males and females aggregate in both space and time, such as with the
explosive breeding assemblages of many frogs and toads (Wells, 1977).
Social behavior is involved in social dominance and the maintenance of ter-
ritories, regardless of whether dominance status or territories are held by indi-
viduals or by groups. Species defend territories when they are economically
defendable, meaning that the benefits of the behaviors associated with having
exclusive access to the area outweigh the costs of maintaining and defending it
from conspecifics (Emlen & Oring, 1977). In the territorial systems of many
species, overt defense in the form of direct aggressive behavior against intrud-
ers has given way to indirect defense in the form of vocalizations and scent
marking.

Social Behavior Based on Access to Clumped or Limited Resources

In most of the cases we have discussed thus far, the benefits of grouping are
apparently derived from the social interactions that take place within aggrega-
tions; that is, aggregations have formed because of some direct benefit of
sociality such as group defense or information sharing that outweighs the
inevitable costs. In some cases, however, the benefits of sociality do not appear
to be sufficient to explain aggregations; rather, groups apparently form in
order that individuals are able to gain access to localization of some critical
and limited resource (Alexander, 1974). Classic examples include safe sleeping
sites for hamadryas baboons (Papio hamadryas) (Kummer, 1968) and suitable
breeding sites for colonial marine birds and mammals. Access to clumped and
limited resources rather than intrinsic social benefits appears to be an impor-
tant driver of group living in many other systems as well, however.
Examples include cases of female (or harem) defense polygyny, in which
females aggregate, often because of clumped food or nest sites, providing dom-
inant males the opportunity to defend them and thereby gain multiple mating
opportunities (Emlen & Oring, 1977). Female-defense polygyny is found in
many ungulates (Clutton-Brock, 1989) and more rarely in other taxa, incud-
ing fishes (Seki et al., 2009); at least one species of bird, the Montezuma oro-
pendola (Psarocolius montezuma) (Webster, 1994); various insects (Heinze &
Hölldobler, 1993); and some amphipods (Just, 1988).
Cooperative breeding occurs when more than two individuals contribute to
the care of young within a single brood. The most well studied of these are
helper-at-the-nest systems in which offspring delay dispersal and remain in their
276 ANIMAL BEHAVIOR

natal groups as nonbreeding helpers, often for several years, during which time
they assist in raising younger siblings rather than breeding on their own (Brown,
1987). Although in a few of these cooperative breeders young may gain sufficient
intrinsic or social benefits by remaining at home to compensate for the reproduc-
tion they forgo, in the majority of cases this is not the case. Rather, they forgo
breeding not because it is an inferior fitness option but because they are unable
to acquire a suitable territory or mate—circumstances referred to as the “ecologi-
cal constraints” hypothesis (Koenig, 1981; Emlen, 1991).
A good example is the Florida scrub-jay (Aphelocoma coerulescens), where
long-term demographic studies have shown that helpers gain an estimated
0.14 offspring equivalents of fitness by helping compared to 0.62 offspring
equivalents gained by first-time breeders in the population (Woolfenden &
Fitzpatrick, 1984). Similar conclusions have been reached by experimental
studies in red-cockaded woodpeckers (Picoides borealis) and Seychelles war-
blers (Acrocephalus sechellensis). In the former, helpers immediately give up
their helper status and occupy artificial nest cavities when they are provided
(Walters et al., 1992), while in the latter birds transplanted to an uninhabited
island forgo group living and breed as pairs until the population saturates
the available high-quality territories (Komdeur, 1992). In all these cases, the
evidence indicates that ecological constraints (or “resource access” benefits;
S.-F. Shen and S. Emlen, personal communication), rather than social benefits
derived from group living, prompt young to remain at home. Once home,
young serve as helpers in order to “make the best of a bad job,” garnering what
inclusive benefits they can while waiting for a territory or breeding vacancy
that will allow them to reproduce (Koenig et al., 1992).
A difficulty in assessing the importance of access to a limited resource is that
once aggregations form, there will inevitably be selection to take advantage of
group living. Thus, various advantages to sociality may emerge and eventually
come to outweigh the disadvantages of grouping even though the original impe-
tus for aggregations was primarily or exclusively extrinsic constraints rather than
social benefits (Koenig et al., 1992). Further, formation of cooperative territorial
groups can lead to supersaturation of habitat with more birds than just one pair
per territory; this can impose a new selective pressure favoring group living by
increasing the constraints on outside options (Dickinson & Hatchwell, 2004).

Cooperative Breeding and Reproductive Skew

Cooperative breeding is found in approximately 9 percent of birds (Cock-
burn, 2006), a smaller proportion of mammals, a few fishes, and some insects
and arachnids. It is particularly common among birds in Australia, possibly
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 277

due to the common phylogenetic history of a large proportion of its avifauna

(Russell, 1989). In general, however, cooperative breeding is rare because it
requires parental care, which outside birds and mammals is itself a relatively
uncommon behavior.
Cooperative breeding is generally associated with species in which dispersal
is restricted and thus there are opportunities for prolonged contact between
close relatives such as occurs in permanent residents inhabiting mild climates
(Hatchwell, 2009). In birds, cooperative breeding is generally believed to be
a result of a shortage of high-quality territories or mates, and helpers will typ-
ically become breeders if given the opportunity to do so, as discussed above.
Although such ecological constraints are clearly important in many species
that have been well studied, others inhabit highly variable and unpredictable
regions where breeding is difficult, at least in some years, and in such cases
helpers may frequently be necessary for successful reproduction (Emlen,
1982; Jetz & Rubenstein, 2011). This “hard-life” hypothesis for the evolution
of cooperative breeding has been proposed to explain cooperative breeding in
African starlings (family Sturnidae) (Rubenstein & Lovette, 2007) and
humans, where raising young is a long, difficult process wrought with danger
(Hrdy, 2009). It may also be important in some species of cooperatively
breeding birds in which helping was previously thought to be driven by eco-
logical constraints as suggested by the finding that helpers are more important
when circumstances are unfavorable for breeding because of poor environ-
mental conditions or breeder inexperience (Magrath, 2001). In contrast, in
the acorn woodpecker, helpers benefit groups significantly more following a
good acorn crop when conditions for breeding are favorable than following
poor crops when conditions are not (Koenig et al., 2011).
In vertebrates, most cases of cooperative breeding involve helping at the
nest, as already described. Less common are species exhibiting cooperative po-
lygamy or mate-sharing, in which multiple cobreeders of one or both sexes
share mates of the opposite sex. In species such as the Harris’s hawk and
Galápagos hawk (Buteo galapagoensis), multiple males mate with a female in a sys-
tem of cooperative polyandry (Faaborg & Patterson, 1981; Faaborg et al., 1995).
In others, such as the European dunnock (Prunella modularis) and the acorn
woodpecker, multiple males may share and mate with more than one female in
a system of cooperative polygynandry. These two species differ, however, in that
dunnock groups are not family based: cobreeding male dunnocks are unrelated
and females breed on separate territories (Davies, 1992). In contrast, acorn
woodpecker groups, like most cooperative breeders, are family based, and
cobreeders within a sex are typically first-order relatives, either siblings or parents
and their offspring (Koenig et al., 1995b).
278 ANIMAL BEHAVIOR

The outcome of mate-sharing—the degree to which reproduction is

monopolized by one or a few individuals—is referred to as the degree of
reproductive skew and is highly variable. In egalitarian or low-skew societies,
cobreeders share parentage relatively equally, as is true for joint-nesting female
acorn woodpeckers because of the phenomenon of egg destruction mentioned
earlier. In contrast, in high-skew societies, reproduction is highly biased
toward a single individual, as is the case in most of the eusocial species we dis-
cuss next, including ant colonies, honey bees, and naked mole-rats (Heteroce-
phalus glaber). The factors determining the degree of reproductive skew is a
subject of considerable interest (Johnstone, 2000; Hager & Jones, 2009),
but most current models start with the importance of four factors: (1) the
expected success of a subordinate that reproduces on its own (corresponding
to the degree of ecological constraints), (2) the expected success of the group
if the subordinate remains in the group, (3) the genetic relatedness of group
members, and (4) the probability that a subordinate would win a fight with
a dominant without being severely injured (Keller & Reeve, 1994). These fac-
tors attempt to synthesize the relative costs and benefits of group living to a
subordinate, taking into account the potential for inclusive fitness benefits of
not breeding and the costs associated with competition for reproductive
opportunities within the group.
Cooperative breeders include species exhibiting a remarkably diverse range
of social organization and behavior (Cockburn, 1998, 2004) and include some
of the most intensively studied species in the world (Stacey & Koenig, 1990;
Koenig & Dickinson, 2004). Before moving on to a discussion of eusociality,
we briefly discuss three, two birds and one mammal. In conjunction with the
species already mentioned, they provide an introduction to the bizarre and
unusual social behavior exhibited by cooperatively breeding species.
Australian fairy-wrens in the genus Malurus encompass a series of spectacu-
larly plumaged species, all of which are cooperative breeders, usually in the
form of young males remaining as helpers in their natal groups. Experimental
work has shown that ecological constraints are key to this system, with helpers
being limited by a shortage of potential mates (Pruett-Jones & Lewis, 1990).
More surprising was the discovery that a high proportion of dominant, breed-
ing pairs appeared to consist of close relatives and that incest was apparently
common (Rowley et al., 1986). Subsequent molecular work demonstrated
that this was not the case; rather, a majority of offspring turn out to be the
product of extra-group matings, leading to the unexpected finding that there
is virtually no correlation between group composition—the social mating
system—and who actually mates with whom—the genetic mating system
(Mulder et al., 1994). The driver of this extraordinary situation appears to
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 279

be sexual selection, with females preferring to mate with older males in the
population regardless of their social mate (Double & Cockburn, 2000).
We have already mentioned acorn woodpeckers, which are a cooperatively
polygynandrous species in which cooperation and competition—the latter
most dramatically in the form of egg destruction by joint-nesting females—
are both regular features of their social behavior. Offspring in this species
typically remain as nonbreeding helpers assisting their parents in raising sub-
sequent young. In contrast to fairy-wrens there is no extra-group mating, nor
do unrelated individuals join established groups. Groups thus consist entirely
of closely related males and equally closely related females.
Of particular interest is the question of what determines reproductive roles
within groups? The two main possibilities are within-sex reproductive competi-
tion—that young birds are reproductively suppressed by their older, dominant
parents—and incest avoidance—that young birds do not breed because they
are closely related to the breeders of the opposite sex. Long-term demographic
and genetic studies have confirmed that when reproductive vacancies arise, they
are filled by unrelated birds from elsewhere rather than by subordinates of the
same sex, after which subordinates of the opposite sex are able to inherit and
cobreed along with their older, presumably dominant parents (Koenig et al.,
1998; Haydock et al., 2001). Thus incest avoidance, rather than reproductive
competition, is the primary determinant of reproductive roles in this species.
Incest avoidance is so strong that the population loses an estimated 9 to 12 per-
cent in reproductive potential due to this factor alone (Koenig et al., 1999). Incest
avoidance is similarly an important factor influencing reproductive roles in the
vast majority of, if not all, highly social vertebrates (Koenig & Haydock, 2004).
Kalahari meerkats (Suricata suricatta) provide a good example of a co-
operatively breeding mammal. Breeders of both sexes live in cooperative
groups with dominants monopolizing most of the reproduction (Hodge
et al., 2008; Spong et al., 2008). Group augmentation, a positive group-size
effect on reproduction, arises because helpers enhance pup growth and sur-
vival by babysitting, which is only done by subordinates and sometimes
involves remaining in the burrow without food for extended periods of time
(Clutton-Brock et al., 2001). There is a measurable cost of helping in terms
of weight loss, but helpers of both sexes benefit from living in the group with
fitness gains through both direct reproduction and the raising of nondescend-
ant kin (Russell et al., 2007). Although reproductive skew is relatively great,
female subordinates sometimes succeed in becoming pregnant, after which
they compete for reproductive success within the group through infanticide
(Kutsukake & Clutton-Brock, 2008). Meanwhile, subordinate males foray
to other groups where they compete to sire extra-group young (Young et al.,
280 ANIMAL BEHAVIOR

2007). While these strategies are not equivalent to breeding as a dominant,

they provide young animals with options that are relatively profitable given
the ecological constraints imposed by limited food, high susceptibility to pre-
dation, and availability of breeding vacancies.

Eusociality
Eusociality, a social organization in which young are cared for cooperatively,
generations overlap, and the society is segregated into distinct castes that provide
different services to the colony, is often considered the pinnacle of social evolu-
tion. Eusociality is found in several orders of insects, including bees and wasps
(order Hymenoptera), termites (order Isoptera), gall-making aphids (order
Hemiptera), and thrips (order Thysanoptera); Synalpheus regalis, a tropical reef-
dwelling marine shrimp dependent on sponges for survival (Duffy, 1996); and
two species of vertebrates, both African mole-rats in the family Bathyergidae,
the naked mole-rat (Heterocephalus glaber) and the Damaraland mole-rat (Crypt-
omys damarensis) (Jarvis & Bennett, 1993; Bennett & Faulkes, 2000). Reproduc-
tive skew in eusocial species is typically very high—often only a single individual
(the “queen”) reproduces out of a colony of thousands or tens of thousands.
They also exhibit extreme task specialization, which makes colonies potentially
very efficient at gathering resources. Workers may never reproduce during their
entire lives but nonetheless gain inclusive fitness benefits by aiding the reproduc-
tion of the queen, who is typically their mother. With a high proportion of the
population forgoing reproduction, often permanently, the evolution of altruism
in eusocial species is a topic of particular interest and played a key role in the
development of kin-selection theory (Hamilton, 1964, 1972).
In diploid species, including most vertebrates, sisters share half their genes
in common, and individuals who help parents produce a benefit tantamount
to the fitness they would achieve by producing an offspring of their own.
Eusocial Hymenoptera, however, are haplodiploid, a system in which unfer-
tilized (haploid) eggs become males and fertilized (diploid) eggs become
females. As a result, a group of sisters with a single mother and father, such
as may make up the workers of an ant colony or a beehive, share one-quarter
of their genes through their mothers and one-half their genes through their
father, as all sperm from a single haploid male are genetically identical. The
result is that such sisters share 75 percent of their genes through common
ancestry, whereas mothers share only 50 percent of their genes with their
own daughters. Female workers potentially transmit more copies of their genes
by helping their mother produce sisters than by producing their own daugh-
ters and sons, thus providing a potential genetic basis for the evolution of
eusociality.
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 281

As appealing as this hypothesis may be, subsequent study has indicated that
haploidiploidy is likely to be a predisposing, rather than a causal, factor in the
evolution of eusociality. Hymenoptera queens often mate with multiple males,
and thus sperm is provided by more than one source, diluting the haplodiploidy
effect on sister relatedness (Hughes et al., 2008b). In addition, multiple queens
may found wasp colonies, further diluting the average genetic relatedness among
workers. It is also relevant that many eusocial species are not haplodiploid,
including not only the two mole-rats but all termites (order Isoptera).
As a result of these and other considerations, workers in eusocial colonies are
thought to have originally given up reproduction due to ecological constraints on
independent breeding, the latter being due to high predation rates, a shortage of
nest sites, and a short breeding season (Brockmann, 1997). As in the case of co-
operatively breeding birds, opportunities to survive and reproduce away from the
colony are limited, favoring individuals that stay home. If females remain in their
natal groups, within-colony relatedness will be high and kin selection will be a
potentially important evolutionary force that favors cooperation.
There remains, however, considerable controversy regarding the evolution
of eusociality, particularly in terms of the role of kin selection (Nowak et al.,
2010). Future phylogenetic work aimed at determining the evolutionary ori-
gins of eusociality are likely to be particularly important in resolving this issue.
A recent study of female mating frequencies in eusocial hymenoptera, for
example, indicated that monogamy is ancestral in all independent eusocial lin-
eages, supporting the hypothesis that kin selection and inclusive fitness has
been key to the evolution of eusociality in this group (Hughes et al., 2008a).

THE STUDY OF SOCIAL BEHAVIOR

As should by now be clear, social behavior is a large field encompassing a
variety of issues that are studied at multiple levels using a diverse array of high-
and low-tech methods. At all levels, however, the goal is generally to deduce
causes using strong inference based on a set of critical predictions (Platt,
1964). If tests of these predictions indicate that the predictions are not met, then
the hypothesis is falsified; if the predictions are met, the hypothesis is supported.
This does not mean that the hypothesis is true, however, because there are always
likely to be alternatives that remain to be tested (see Volume 1, Chapter 1).
Consider the variability in behavior of male step parents, which may (1) kill
the offspring of their new mate, (2) tolerate the offspring, or (3) invest in the
offspring and treat them as a normal parent (Rohwer, 1986; Rohwer et al.,
1999). The first option, infanticide, was long considered aberrant behavior
but is now generally considered a form of sexual competition by which a male
282 ANIMAL BEHAVIOR

step parent gains a reproductive advantage through earlier conception by his

new female (Hrdy, 1974). More difficult to explain is why male step parents
sometimes adopt and feed the offspring of widowed females as if they were
the normal genetic parent.
One possible explanation (the current benefits hypothesis) is that they in
fact have mated with the female and thus may have genetic offspring in the
female’s nest (Meek & Robertson, 1991). An alternative (the future benefits
hypothesis) is that the adoptive male gains future reproductive benefits by vir-
tue of his foster behavior, increasing the likelihood that his female will mate
with him during her next breeding attempt. The current benefits hypothesis
predicts that some of a remated female’s nestlings have been sired by the adop-
tive father, whereas the future benefits hypothesis predicts that the adoptive
male will mate sooner, usually with the widowed female, and ultimately pro-
duce more offspring than an unpaired male that fails to adopt.
Experimental work on western bluebirds (Sialia mexicana), where about
half of male step parents feed their adoptive young while the other half tolerate
them, but are never infanticidal, found that male provisioning had no effect on
female condition or survivorship and male step parents that fed young were
not more likely to breed with the female on her subsequent attempt than were
males that did not feed (Dickinson & Weathers, 1999). This finding rejects
the future benefits hypothesis in this species but does not prove that males
gain current benefits, which would require either observational data demon-
strating that adoptive males had previously achieved extra-pair matings with
the widowed female or that one or more of her offspring had been sired by
the adoptive male (Meek & Robertson, 1991).
Ideally, hypotheses are mutually exclusive such that only one can be true.
In many cases, however, behaviors have more than one current function and
more than one hypothesis may be supported. In the case of feeding by step
parents, for example, it is possible that males might have both achieved
extra-pair matings with the female and be more likely to breed with her in
the future, thus supporting both the current and future benefits hypotheses.
Strong inference relies on clear tests and critical predictions to distinguish
between alternative hypotheses designed to explain some phenomenon at a
particular level of analysis. Predictions can be tested either with data collected
from field observations or with experiments, although the latter are often con-
sidered preferrable because confounding factors are more easily controlled.
Unfortunately, experimental manipulations may alter other factors beyond
those intended, especially where social behavior is concerned. In order to min-
imize such problems, researchers generally try to test their hypotheses using
multiple lines of evidence.
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 283

Social behavior is a complex combination of the costs and benefits of living

in groups, dominance interactions, conflict between the sexes, nepotism, com-
petition, and cooperation. The diversity of social behavior continues to pro-
vide significant material for evolutionary biologists to understand natural
selection and the process of evolution and for mechanistic biologists to under-
stand how the underlying processes governing behavior work. Through behav-
ioral ecology, these questions are framed within an ecological context to peel
back the layers and reveal the complex underpinnings of the natural world.

ACKNOWLEDGMENTS
We wish to thank the National Science Foundation for support of our
research on avian social behavior. Portions of the chapter have been adapted
from Walter D. Koenig and Janis L. Dickinson, Encyclopædia Britannica
Online, s. v. “social behaviour, animal,” accessed February 28, 2013, http://
www.britannica.com/EBchecked/topic/550897/social-behaviour-animal.

REFERENCES AND SUGGESTED READING

Adkins-Regan, E. & M. Ascenzi. (1990). Sexual differentiation of behavior in the
zebra finch: Effect of early gonadectomy or androgen treatment. Hormones
and Behavior, 24, 114–127.
Alexander, R. D. (1974). The evolution of social behavior. Annual Review of Ecology
and Systematics, 5, 325–383.
Aubin, T. & P. Jouventin. (2002). How to vocally identify kin in a crowd: The pen-
guin model. Advances in the Study of Behavior, 31, 243–277.
Balcombe, J. P. (1990). Vocal recognition of pups by mother Mexican free-tailed
bats, Tadarida brasiliensis mexicana. Animal Behaviour, 39, 960–966.
Baptista, L. F. & L. Petrinovich. (1984). Social interaction, sensitive phases and the
song template hypothesis in the white-crowned sparrow. Animal Behaviour,
32, 172–181.
Bednarz, J. C. (1988). Cooperative hunting Harris’ hawks (Parabuteo unicinctus).
Science, 239, 1525–1527.
Beecher, M. D. & E. A. Brenowitz. (2005). Functional aspects of song learning in
songbirds. Trends in Ecology and Evolution, 20, 143–149.
Beecher, M. D., J. M. Burt, A. L. O’Loghlen, C. N. Templeton, & S. E. Campbell.
(2007). Bird song learning in an eavesdropping context. Animal Behaviour,
73, 929–935.
Beecher, M. D., J. C. Nordby, S. E. Campbell, J. M. Burt, C. E. Hill, &
A. L. O’Loghlen. (1997). What is the function of song learning in songbirds?
Perspectives in Ethology, 12, 77–97.
284 ANIMAL BEHAVIOR

Beehler, B. M. & M. S. Foster. (1988). Hotshots, hotspots, and female preference in

the organization of lek mating systems. American Naturalist, 131, 203–219.
Bennett, N. C. & C. G. Faulkes. (2000). African Mole-rats: Ecology and Eusociality.
Cambridge, UK: Cambridge University Press.
Ben-Shahar, Y., A. Robichon, M. B. Sokolowski, & G. E. Robinson. (2002).
Influence of gene action across different time scales on behavior. Science,
296, 741–744.
Blumstein, D. T. (2007). The evolution of alarm communication in rodents: Struc-
ture, function, and the puzzle of apparently altruistic calling. In J. Wolff &
P. W. Sherman (eds.), Rodent Societies: An Ecological and Evolutionary Perspec-
tive (pp. 317–327). Chicago: University of Chicago Press.
Booksmythe, I., C. Hayes, M. D. Jennions, & P. R. Y. Backwell. (2012). The effects
of neighbor familiarity and size on cooperative defense of fiddler crab territo-
ries. Behavioral Ecology, 23, 285–289.
Bradbury, J. W. (1977). Lek mating behavor in the hammer-headed bat. Zeitschrift
für Tierpsychologie, 45, 225–255.
Bradbury, J. W. & R. M. Gibson. (1983). Leks and mate choice. In P. Bateson
(ed.), Mate Choice (pp. 109–138). Cambridge, UK: Cambridge University
Press.
Brockmann, H. J. (1997). Cooperative breeding in wasps and vertebrates: The role of
ecological constraints. In J. C. Choe & B. J. Crespi (eds.), The Evolution
of Social Behavior in Insects and Arachnids (pp. 347–371). Cambridge, UK:
Cambridge University Press.
Brown, C. R. & M. B. Brown. (1996). Coloniality in the Cliff Swallow: The Effect of
Group Size on Social Behavior. Chicago: University of Chicago Press.
Brown, J. L. (1987). Helping and Communal Breeding in Birds. Princeton, NJ:
Princeton University Press.
Bshary, R. & A. S. Grutter. (2006). Image scoring and cooperation in a cleaner fish
mutualism. Nature, 441, 975–978.
Burmeister, S. S., E. D. Jarvis, & R. D. Fernald. (2005). Rapid behavioral and
genomic responses to social opportunity. PLoS Biology, 3, e363. doi:10.1371/
journal.pbio.0030363.
Cardé, R. T. & J. G. Millar. (2009). The scent of a female: Sex pheromones of female
tiger moths. In W. E. Conner (ed.), Tiger Moths and Woolly Bears: Behavior,
Ecology and Evolution of the Arctiidae (pp. 127–144). Oxford, UK: Oxford
University Press.
Charlton, B. D., D. Reby, & K. McComb. (2007). Female red deer prefer the roars of
larger males. Biology Letters, 3, 382–385.
Cheney, D. L. & R. M. Seyfarth. (2009). Stress and coping mechanisms in female
primates. Advances in the Study of Behavior, 39, 1–44.
Clutton-Brock, T. H. (1989). Mammalian mating systems. Proceedings of the Royal
Society of London, B, 236, 339–372.
Clutton-Brock, T. H. & S. D. Albon. (1979). The roaring of red deer and the evolu-
tion of honest advertisement. Behaviour, 69, 145–170.
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 285

Clutton-Brock, T. H., A. F. Russell, L. L. Sharpe, P. N. M. Brotherton, G. M. McIl-

rath, S.White, & E. Z. Cameron. (2001). Effects of helpers on juvenile devel-
opment and survival in meerkats. Science, 293, 2446–2449.
Cockburn, A. (1998). Evolution of helping behavior in cooperatively breeding birds.
Annual Review of Ecology and Systematics, 29, 141–177.
Cockburn, A. (2004). Mating systems and sexual conflict. In W. D. Koenig &
J. L. Dickinson (eds.), Ecology and Evolution of Cooperative Breeding in Birds
(pp. 81–101). Cambridge, UK: Cambridge University Press.
Cockburn, A. (2006). Prevalence of different modes of parental care in birds. Proceed-
ings of the Royal Society of London, B, 273, 1375–1383.
Cocroft, R. B. (1999). Offspring-parent communication in a subsocial treehopper
(Hemiptera: Membracidae: Umbonia crassicornis). Behaviour, 136, 1–21.
Danchin, É., L.-A. Giraldeau, T. J.Valone, & R. H. Wagner. (2004). Public informa-
tion: From nosy neighbors to cultural evolution. Science, 305, 487–491.
Davidson, J. M., C. H. Rodgers, E. R. Smith, & G. J. B. Bloch. (1968). Stimulation
of female sex behavior in adrenalectomized rats with estrogen alone. Endocri-
nology, 82, 193–195.
Davies, N. B. (1992). Dunnock Behaviour and Social Evolution. New York: Oxford
University Press.
Davies, N. B. (2000). Cuckoos, Cowbirds and Other Cheats. London: T. and A. D.
Poyser.
Després, L., J.-P. David, & C. Gallet. (2007). The evolutionary ecology of insect re-
sistance to plant chemicals. Trends in Ecology and Evolution, 22, 298–307.
De Vos, A. & M. J. O’Riain. (2010). Sharks shape the geometry of a selfish seal herd:
Experimental evidence from seal decoys. Biology Letters, 6, 48–50.
Dickinson, J. L. & B. J. Hatchwell. (2004). Fitness consequences of helping, In
W. D. Koenig & J. L. Dickinson (eds.), Ecology and Evolution of Cooperative
Breeding in Birds (pp. 48–66). Cambridge, UK: Cambridge University Press.
Dickinson, J. L. & W. W. Weathers. (1999). Replacement males in the western blue-
bird: Opportunity for paternity, chick-feeding rules, and fitness consequences
of male parental care. Behavioral Ecology and Sociobiology, 45, 201–209.
Doligez, B., É. Danchin, & J. Clobert. (2002). Public information and breeding hab-
itat selection in a wild bird population. Science, 297, 1168–1170.
Double, M. & A. Cockburn. (2000). Pre-dawn infidelity: Females control extra-pair
mating in superb fairy-wrens. Proceedings of the Royal Society of London, B,
267, 465–470.
Duffy, J. E. (1996). Eusociality in a coral-reef shrimp. Nature, 381, 512–514.
Emlen, S. T. (1982). The evolution of helping. I. An ecological constraints model.
American Naturalist, 119, 29–39.
Emlen, S. T. (1991). Evolution of cooperative breeding in birds and mammals. In J. R.
Krebs & N. B. Davies (eds.), Behavioural Ecology: An Evolutionary Approach,
Third Edition (pp. 301–337). Oxford, UK: Blackwell Scientific Publications.
Emlen, S. T. & L. W. Oring. (1977). Ecology, sexual selection, and the evolution of
mating systems. Science, 197, 215–223.
286 ANIMAL BEHAVIOR

Faaborg, J., P. G. Parker, L. Delay, Tj. de Vries, J. C. Bednarz, S. M. Paz, J. Naranjo,

& T. A. Waite. (1995). Confirmation of cooperative polyandry in the Galapa-
gos hawk (Buteo galapagoensis). Behavioral Ecology and Sociobiology, 36, 83–90.
Faaborg, J. & C. B. Patterson. (1981). The characteristics and occurrence of co-
operative polyandry. Ibis, 123, 477–484.
Fehr, E. & S. Gächter. (2002). Altruistic punishment in humans. Nature, 415,
137–140.
Fitzgerald, T. D. & S. C. Peterson. (1988). Cooperative foraging and communication
in caterpillars. BioScience, 38, 20–25.
Foster, W. A. & J. E. Treherne. (1981). Evidence for the dilution effect in the selfish
herd from fish predation on a marine insect. Nature, 293, 466–467.
Gadagkar, R. (1993). Can animals be spiteful? Trends in Ecology and Evolution, 8,
232–234.
Gamberale, G. & B. S. Tullberg. (1998). Aposematism and gregariousness: The com-
bined effect of group size and coloration on signal repellence. Proceedings of the
Royal Society of London, B, 265, 889–894.
Garner, A., I. C. W. Hardy, P. D. Taylor, & S. A. West. (2007). Spiteful soldiers and
sex ratio conflict in polyembryonic parasitoid wasps. American Naturalist, 169,
519–533.
Ghent, A. W. (1960). A study of the group-feeding behaviour of larvae of the jack
pine sawfly, Neodiprion pratti banksianae Roh. Behaviour, 16, 110–148.
Greene, E. (1987). Individuals in an osprey colony discriminate between high and
low quality information. Nature, 329, 239–241.
Grüter, C. & W. M. Farina. (2009). The honeybee waggle dance: Can we follow the
steps? Trends in Ecology and Evolution, 24, 242–247.
Gustin, M. K. & G. F. McCracken. (1987). Scent recognition between females and
pups in the bat Tadarida brasiliensis mexicana. Animal Behaviour, 35, 13–19.
Hager, R. & C. B. Jones (eds.). (2009). Reproductive skew in vertebrates: Proximate
and ultimate causes. Cambridge, UK: Cambridge University Press.
Hamilton, W. D. (1964). The genetical evolution of social behaviour. I and II. Jour-
nal of Theoretical Biology, 7, 1–52.
Hamilton, W. D. (1971). Geometry for the selfish herd. Journal of Theoretical Biology,
31, 295–311.
Hamilton, W. D. (1972). Altruism and related phenomena, mainly in social insects.
Annual Review of Ecology and Systematics, 3, 193–232.
Hatchwell, B. J. (2009). The evolution of cooperative breeding in birds: Kinship,
dispersal and life history. Philosophical Transactions of the Royal Society of
London, B, 364, 3217–3227.
Haydock, J. & W. D. Koenig. (2002). Reproductive skew in the polygynandrous
acorn woodpecker. Proceedings of the National Academy of Sciences, USA, 99,
7178–7183.
Haydock, J., W. D. Koenig, & M. T. Stanback. (2001). Shared parentage and incest
avoidance in the cooperatively breeding acorn woodpecker. Molecular Ecology,
10, 1515–1525.
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 287

Heard, D. C. (1992). The effect of wolf predation and snow cover on musk-ox group
size. American Naturalist, 139, 190–204.
Heinze, J. & B. Hölldobler. (1993). Fighting for a harem of queens: Physiology of
reproduction in Cardiocondyla male ants. Proceedings of the National Academy
of Sciences, USA, 90, 8412–8414.
Hess, E. H. (1964). Imprinting in birds. Science, 146, 1128–1139.
Hodge, S. J., A. Manica, T. P. Flower, & T. H. Clutton-Brock. (2008). Determi-
nants of reproductive success in dominant female meerkats. Journal of Animal
Ecology, 77, 92–102.
Höglund, J. & R. Alatalo. (1995). Leks. Princeton, NJ: Princeton University Press.
Höglund, J., R. V. Alatalo, A. Lundberg, P. T. Rintamäki, & J. Lindell. (1999).
Microsatellite markers reveal the potential for kin selection on black grouse
leks. Proceedings of the Royal Society of London, B, 266, 813–816.
Holmes, W. G. & P. W. Sherman. (1982). The ontogeny of kin recognition in two
species of ground squirrels. American Zoologist, 22, 491–517.
Hrdy, S. B. (1974). Male-male competition and infanticide among the langurs (Pres-
bytis entellus) of Abu, Rajasthan. Folia Primatologica, 22, 19–58.
Hrdy, S. B. (2009). Mothers and Others. Cambridge, MA: Belknap Press.
Hughes, W. O. H., B. P. Oldroyd, M. Beekman, & F. L. W. Ratnieks. (2008a).
Ancestral monogamy shows kin selection is key to the evolution of eusociality.
Science, 320, 1213–1216.
Hughes, W. O. H., F. L. W. Ratnieks, & B. P. Oldroyd. (2008b). Multiple pat-
ernity or multiple queens: Two routes to greater intracolonial genetic
diversity in the eusocial Hymenoptera. Journal of Evolutionary Biology, 21,
1090–1095.
Jacob, S., M. K. McClintock, B. Zelano, & C. Ober. (2002). Paternally inherited
HLA alleles are associated with women’s choice of male odor. Nature Genetics,
30, 175–179.
Jarvis, J. U. M. & N. C. Bennett. (1993). Eusociality has evolved independently in
two genera of bathyergid mole-rats—but occurs in no other subterranean
mammal. Behavioral Ecology and Sociobiology, 33, 253–260.
Jensen, K. (2010). Punishment and spite, the dark side of cooperation. Proceedings of
the Royal Society of London, B, 365, 2635–2650.
Jetz, W. & D. R. Rubenstein. (2011). Environmental uncertainty and the global bio-
geography of cooperative breeding in birds. Current Biology, 21, 72–78.
Johnstone, R. A. (2000). Models of reproductive skew: A review and synthesis. Ethology,
106, 5–26.
Jouventin, P. & T. Aubin. (2002). Acoustic systems are adapted to breeding
ecologies: Individual recognition in nesting penguins. Animal Behaviour, 64,
747–757.
Just, J. (1988). Siphonoecetinae (Corophiidae) 6: A survey of phylogeny, distribution,
and biology. Crustaceana, 13, 193–208.
Keller, L. & H. K. Reeve. (1994). Partitioning of reproduction in animal societies.
Trends in Ecology and Evolution, 9, 98–102.
288 ANIMAL BEHAVIOR

Koenig, W. D. (1981). Reproductive success, group size, and the evolution of co-
operative breeding in the acorn woodpecker. American Naturalist, 117, 421–
443.
Koenig, W. D. & J. L. Dickinson (eds.). (2004). Ecology and Evolution of Cooperative
Breeding in Birds. Cambridge, UK: Cambridge University Press.
Koenig, W. D. & J. Haydock. (2004). Incest avoidance. In W. D. Koenig & J. L.
Dickinson (eds.), Ecology and Evolution of Cooperative Breeding in Birds (pp.
142–156). Cambridge, UK: Cambridge University Press.
Koenig, W. D., J. Haydock, & M. T. Stanback. (1998). Reproductive roles in the co-
operatively breeding acorn woodpecker: Incest avoidance versus reproductive
competition. American Naturalist, 151, 243–255.
Koenig, W. D., R. L. Mumme, M. T. Stanback, & F. A. Pitelka. (1995a). Patterns
and consequences of egg destruction among joint-nesting acorn woodpeckers.
Animal Behaviour, 50, 607–621.
Koenig, W. D., F. A. Pitelka, W. J. Carmen, R. L. Mumme, & M. T. Stanback.
(1992). The evolution of delayed dipersal in cooperative breeders. Quarterly
Review of Biology, 67, 111–150.
Koenig, W. D., P. B. Stacey, M. T. Stanback, & R. L. Mumme. (1995b). Acorn
woodpecker (Melanerpes formicivorus). In A. Poole & F. Gill (eds.), The Birds
of North America, No. 194. Philadelphia: Academy of Natural Sciences; Wash-
ington, DC: American Ornithologists’ Union.
Koenig, W. D., M. T. Stanback, & J. Haydock. (1999). Demographic consequences
of incest avoidance in the cooperatively breeding acorn woodpecker. Animal
Behaviour, 57, 1287–1293.
Koenig, W. D., E. L. Walters, & J. Haydock. (2011). Variable helper effects, ecologi-
cal conditions, and the evolution of cooperative breeding in the acorn wood-
pecker. American Naturalist, 178, 145–158.
Komdeur, J. (1992). Importance of habitat saturation and territory quality for evolu-
tion of cooperative breeding in the Seychelles warbler. Nature, 358, 493–495.
Krakauer, A. H. (2005). Kin selection and cooperative courtship in wild turkeys.
Nature, 434, 69–72.
Krakauer, D. C. (1995). Groups confuse predators by exploiting perceptual bottle-
necks: A connectionist model of the confusion effect. Behavioral Ecology and
Sociobiology, 36, 421–429.
Krause, J. & G. D. Ruxton. (2002). Living in Groups. Oxford, UK: Oxford Univer-
sity Press.
Krieger, M. J. B. & K. G. Ross. (2002). Identification of a major gene regulating com-
plex social behavior. Science, 295, 328–332.
Kummer, H. (1968). Two variations in the social organization of baboons. In P. Jay
(ed.), Primates: Studies in Adaptation and Variability (pp. 293–312). New
York: Holt, Rinehart, & Winston.
Kutsukake, N. & T. H. Clutton-Brock. (2008). The number of subordinates moder-
ates intrasexual competition among males in cooperatively breeding meerkats.
Proceedings of the Royal Society of London, B, 275, 209–216.
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 289

Lihoreau, M. & C. Rivault. (2009). Kin recognition via cuticular hydrocarbons

shapes cockroach social life. Behavioral Ecology, 20, 46–53.
Lingle, S. (2001). Anti-predator strategies and grouping patterns in white-tailed deer
and mule deer. Ethology, 107, 295–314.
Lorenz, K. Z. (1935). Der kumpan in der umwelt des vogels. Journal für Ornithologie,
83, 137–213, 287–413.
Lyon, B. E. & J. M. Eadie. (2008). Conspecific brood parasitism in birds: A life-
history perspective. Annual Review of Ecology, Evolution, and Systematics, 39,
343–363.
MacDougall-Shackleton, S. A. (2011). The levels of analysis revisited. Philosophical
Transactions of the Royal Society of London, B, 366, 2076–2085.
Magrath, R. D. (2001). Group breeding dramatically increases reproductive success of
yearling but not older female scrubwrens: A model for cooperatively breeding
birds? Journal of Animal Ecology, 70, 370–385.
Marler, P. (1970). A comparative approach to vocal learning: Song development in
white-crowned sparrows. Journal of Comparative and Physiological Psychology,
71, 1–25.
Mateo, J. M. (2002). Kin-recognition abilities and nepotism as a function of sociality.
Proceedings of the Royal Society of London, B, 269, 721–727.
Mateo, J. M. (2010). Self-referent phenotype matching and long-term maintenance
of kin recognition. Animal Behaviour, 80, 929–935.
McClure, M. & E. Despland. (2011). Defensive responses by a social caterpillar are
tailored to different predators and change with larval instar and group size.
Naturwissenschaften, 98, 425–434.
McComb, K. (1987). Roaring by red deer stags advances the date of oestrus in hinds.
Nature, 330, 648–649.
Meek, S. B. & R. J. Robertson. (1991). Adoption of young by replacement male
birds: An experimental study of eastern bluebirds and a review. Animal
Behaviour, 42, 813–820.
Mello, C. V., D. S. Vicario, & D. F. Clayton. (1992). Song presentation induces gene
expression in the songbird forebrain. Proceedings of the National Academy of
Sciences, USA, 89, 6818–6822.
Michener, C. D. (1969). Comparative social behavior of bees. Annual Review of
Entomology, 14, 299–342.
Michener, C. D. (2007). The Bees of the World. Baltimore, MD: Johns Hopkins
University Press.
Milinski, M., D. Semmann, & H. J. Krambeck. (2002). Reputation helps solve the
“tragedy of the commons.” Nature, 415, 424–426.
Morales, M. A., J. L. Barone, & C. S. Henry. (2008). Acoustic alarm signalling facil-
itates predator pretection of treehoppers by mutualist ant bodyguards. Proceed-
ings of the Royal Society of London, B, 275, 1935–1941.
Morrow, P. A., T. E. Bellas, & T. Eisner. (1976). Eucalyptus oils in the defensive oral
discharge of Australian sawfly larvae (Hymenoptera: Pergidae). Oecologia, 24,
193–206.
290 ANIMAL BEHAVIOR

Mulder, R. A., P. O. Dunn, A. Cockburn, K. A. Lazenby-Cohen, & M. J. Howell.

(1994). Helpers liberate female fairy-wrens from constraints on extra-pair mate
choice. Proceedings of the Royal Society of London, B, 255, 223–229.
Munn, C. A. (1986). Birds that “cry wolf.” Nature, 391, 143–145.
Nelson, R. J. (2011). An Introduction to Behavioral Endocrinology, Fourth Edition.
Sunderland, MA: Sinauer Associates.
Nordby, J. C., S. E. Campbell, J. M. Burt, & M. D. Beecher. (2000). Social influen-
ces during song development in the song sparrow: A laboratory experiment
simulating field conditions. Animal Behaviour, 59, 1187–1197.
Nottebohm, F. & A. P. Arnold. (1976). Sexual dimorphism in vocal control areas of
the songbird brain. Science, 194, 211–213.
Novaro, A. J., C. A. Moraga, C. Briceño, M. C. Funes, & A. Marino. (2009). First
records of culpeo (Lycalopex culpaeus) attacks and cooperative defense by gua-
nacos (Lama guanicoe). Mammalia, 73, 148–150.
Nowak, M. A. & K. Sigmund. (2005). Evolution of indirect reciprocity. Nature, 437,
1291–1298.
Nowak, M. A., C. E. Tarnita, & E. O. Wilson. (2010). The evolution of eusociality.
Nature, 466, 1057–1062.
Orians, G. H. & D. Janzen. (1974). Why are embryos so tasty? American Naturalist,
108, 581–592.
Packer, C. (1977). Reciprocal altruism in Papio anubis. Nature, 265, 441–443.
Packer, C. & L. Ruttan. (1988). The evolution of cooperative hunting. American
Naturalist, 132, 159–198.
Peters, A., L. B. Astheimer, C. R. J. Boland, & A. Cockburn. (2000). Testosterone is
involved in acquisition and maintenance of sexually selected male plumage in
superb fairy-wrens, Malurus cyaneus. Behavioral Ecology and Sociobiology, 47,
438–445.
Peterson, S. C., N. D. Johnson, & J. L. LeGuyader. (1987). Defensive regurgitation
of allelochemicals derived from host cyanogenesis by eastern tent caterpillars.
Ecology, 68, 1268–1272.
Petrie, M., A. Krupa, & T. Burke. (1999). Peacocks lek with relatives even in the
absence of social and environmental cues. Nature, 401, 155–157.
Platt, J. R. (1964). Strong inference. Science, 146, 347–353.
Potts, W. K., C. J. Manning, & E. K. Wakeland. (1991). Mating patterns in semina-
tural populations of mice influenced by MHC genotype. Nature, 352, 619–621.
Pruett-Jones, S. G. & M. J. Lewis. (1990). Sex ratio and habitat limitation promote
delayed dispersal in superb fairy-wrens. Nature, 348, 541–542.
Queller, D. C. (1987). The evoluiton of leks through female choice. Animal Behav-
iour, 35, 1424–1432.
Reeve, H., K. (1989). The evolution of conspecific acceptance thresholds. American
Naturalist, 133, 407–435.
Reynolds, A. M., G. A. Sword, S. J. Simpson, & D. R. Reyolds. (2009). Predator per-
colation, insect outbreaks and phase polyphenism. Current Biology, 19, 20–24.
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 291

Riipi, M., R. V. Alatalo, L. Lindström, & J. Mappes. (2001). Multiple benefits of gre-
gariousness cover detectability costs in aposematic aggregations. Nature, 413,
512–514.
Robinson, G. E., R. D. Fernald, & D. F. Clayton. (2008). Genes and social behavior.
Science, 322, 896–900.
Rohwer, S. (1986). Selection for adoption versus infanticide by replacement “mates”
in birds. Current Ornithology, 3, 353–395.
Rohwer, S., J. C. Herron, & M. Daly. (1999). Stepparental behavior as mating effort
in birds and other animals. Evolution and Human Behavior, 20, 367–390.
Rothstein, S. I. (1990). A model system for coevolution: Avian brood parasitism.
Annual Review of Ecology and Systematics, 21, 481–508.
Rowley, I., E. M. Russell, & M. G. Brooker. (1986). Inbreeding: Benefits may out-
weigh costs. Animal Behaviour, 34, 939–941.
Rubenstein, D. R. & I. J. Lovette. (2007). Temporal environmental variability
drives the evolution of cooperative breeding in birds. Current Biology,17,
1414–1419.
Russell, A. F., A. J. Young, G. Spong, N. R. Jordan, & T. H. Clutton-Brock. (2007).
Helpers increase the reproductive potential of offspring in cooperative meer-
kats. Proceedings of the Royal Society of London, B, 274, 513–520.
Russell, E. M. (1989). Co-operative breeding: A Gondwanan perspectic. Emu, 89,
61–62.
Rutter, M. (2006). Genes and Behavior: Nature-Nurture Interplay Explained. Oxford,
UK: Blackwell Publishing.
Santorelli, L. A., C. R. L. Thompson, E. Villegas, J. Svetz, C. Dinh, A. Parikh, et al.
(2008). Facultative cheater mutants reveal the genetic compleity of co-
operation in social amoebae. Nature, 451, 1107–1110.
Sapolsky, R. M. (1998). Why Zebras Don’t Get Ulcers. New York: W. H. Freeman and
Co.
Seeley, T. D. (1995). The Wisdom of the Hive: The Social Physiology of Honey Bee
Colonies. Cambridge, MA: Harvard University Press,
Seki, S., M. Kohda, G. Takamoto, K. Karino, Y. Nakashima, & T. Kuwamura.
(2009). Female defense polygyny in the territorial triggerfish Sufflamen chrysop-
terum. Journal of Ethology, 27. 215–220.
Seyfarth, R. M., D. L. Cheney, & P. Marler. (1980). Monkey responses to three dif-
ferent alarm calls: Evidence of predator classification and semantic communi-
cation. Science, 210, 801–803.
Sharp, S. P., A. McGowan, M. J. Wood, & B. J. Hatchwell. (2005). Learned kin rec-
ognition cues in a social bird. Nature, 434, 1127–1130.
Sharpe, L. L., A. S. Joustra, & M. I. Cherry. (2010). The presence of an avian co-
forager reduces vigilance in a cooperative mammal. Biology Letters, 6, 475–477.
Sherman, P. W. (1977). Nepotism and the evolution of alarm calls. Science, 197,
1246–1253.
Sherman, P. W. (1988). The levels of analysis. Animal Behaviour, 36, 616–619.
292 ANIMAL BEHAVIOR

Sherman, P. W., E. A. Lacey, H. K. Reeve, & L. Keller. (1995). The eusociality con-
tinuum. Behavioral Ecology, 6, 102–108.
Sherman, P. W., H. K. Reeve, & D. W. Pfennig. (1997). Recognition systems. In
J. R. Krebs & N. B. Davies (eds.), Behavioural Ecology (pp. 69–96). Oxford,
UK: Blackwell.
Shorey, L., S. Piertney, J. Stone, & J. Höglund. (2000). Fine-scale genetic structuring
on Manacus manacus leks. Nature, 408, 352–353.
Spong, G. F., S. J. Hodge, A. J. Young, & T. H. Clutton-Brock. (2008). Factors
affecting the reproductive success of dominant male meerkats. Molecular
Ecology, 17, 2287–2299.
Stacey, P. B. & W. D. Koenig. (1990). Cooperative Breeding in Birds: Long-term
Studies of Ecology and Behavior. Cambridge, UK: Cambridge University Press.
Swynnerton, C. F. M. (1915). Mixed bird parties. Ibis, 3, 346–354.
Tallamy, D. W. (2005). Egg dumping in insects. Annual Review of Entomology, 50,
347–370.
Tener, J. S. (1965). Muskoxen. Ottawa, Canada: Queen’s Printer.
Terborgh, J. (1990). Mixed flocks and polyspecific associations: Costs and benefits
of mixed groups to birds and monkeys. American Journal of Primatology, 21,
87–100.
Thornhill, T. & J. Alcock. (1983). The Evolution of Insect Mating Systems. Cambridge,
MA: Harvard University Press.
Tinbergen, N. (1963). On aims and methods of ethology. Zietschrift für Tierpsycholo-
gie, 20, 410–433.
Trivers, R. L. (1971). The evolution of reciprocal altruism. Quarterly Review of Biol-
ogy, 46, 35–57.
von Frisch, K. (1967). The Dance Language and Orientation of Bees. Cambridge, MA:
Harvard University Press.
Wade, J. & A. P. Arnold. (1996). Functional testicular tissue does not masculinize
development of the zebra finch song system. Proceedings of the National
Academy of Sciences, USA, 93, 5264–5268.
Wade, J. & A. P. Arnold. (2004). Sexual differentiation of the zebra finch song
system. Annals of the New York Academy of Sciences, 1016, 540–559.
Walters, J. R., C. K. Copeyon, & J. H. Carter III. (1992). Test of the ecological basis
of cooperative breeding in red-cockaded woodpeckers. Auk, 109, 90–97.
Ward, P. & A. Zahavi. (1973). The importance of certain assemblages of birds as
“information-centres” for food-finding. Ibis, 115, 517–534.
Webster, M. S. (1994). Female-defence polygyny in a neotropical bird, the Monte-
zuma oropendola. Animal Behaviour, 48, 779–794.
Wedekind, C. & D. Penn. (2000). MHC genes, body odours, and odour preferences.
Nephrology Dialysis Transplantation, 15, 1269–1271.
Weimerskirch, H., S. Bertrand, J. Silva, J. C. Marques, & E. Goya. (2010). Use of
social information in seabirds: Compass rafts indicate the heading of food
patches. PLoS One, 5, e9928. doi:10.1371/journal.pone.0009928.
LET ’ S GET TOGETHER : THE EVOLUTION OF SOCIAL BEHAVIOR 293

Wells, K. D. (1977). The social behaviour of anuran amphibians. Animal Behaviour,

25, 666–693.
Wertheim, B., J. Marchais, L. E. M. Vet, & M. Dicke. (2002). Allee effect in larval
resource exploitation in Drosophila: An interaction among density of adults,
larvae, and micro-organisms. Ecological Entomology, 27, 608–617.
West, S. A. & A. Gardner. (2010). Altruism, spite, and greenbeards. Science, 327,
1341–1344.
Wilkinson, G. S. (1984). Reciprocal food sharing in vampire bats. Nature, 308,
181–184.
Wilkinson, G. S. (1992). Information transfer at evening bat colonies. Animal
Behaviour, 44, 501–518.
Wilson, E. O. (1975). Sociobiology: The New Synthesis. Cambridge, MA: Belknap
Press.
Wilson, P. L. & S. L. Vehrencamp. (2001). A test of the deceptive mimicry hypoth-
esis in song-sharing song sparrows. Animal Behaviour, 62, 1197–1205.
Woolfenden, G. E. & J. W. Fitzpatrick. (1984). The Florida Scrub Jay: Demography of
a Cooperative-breeding Bird. Princeton, NJ: Princeton University Press.
Yom-Tov, Y. (1980). Intraspecific nest parasitism in birds. Biological Reviews, 55,
93–108.
Young, A. J., G. Spong, & T. Clutton-Brock. (2007). Subordinate male meerkats
prospect for extra-group paternity: Alternative reproductive tactics in a
cooperative mammal. Proceedings of the Royal Society of London, B, 274,
1603–1609.
This page intentionally left blank
 10

Ecological and Evolutionary Feedbacks

in the Evolution of Aggression
Renée A. Duckworth

INTRODUCTION
Acts of aggression are ubiquitous among animals and play a key role in survival
and reproduction. Animals often use aggressive behaviors to defend or usurp a re-
source (Stamps & Krishnan, 1997; Garcia & Arroyo, 2002), to compete for
mates (Bartoš, 1986; Hagelin, 2002), to fend off predators (Andersson et al.,
1980; Redondo & Carranza, 1989), and during foraging (Riechert, 1993). How-
ever, even though aggression is important in a number of distinct contexts, it is
also a costly behavior. These costs range from direct costs in terms of energy
expenditure and injury during aggressive conflicts to more subtle costs of dis-
rupting social bonds in communally living species. A balance between costs and
benefits has traditionally been the main explanation for variation in aggression
observed within and among individuals as well as between populations and spe-
cies; however, recent work on animal personalities has introduced the possibility
that constraints may also be important to understanding this variation.
Aggression, by definition, is a social (or antisocial) behavior in that it only
occurs in the context of two or more interacting individuals. At the most basic
level, aggression is used to either subdue another individual (as during preda-
tion or aggressive courtship) or to repel another individual (as during
296 ANIMAL BEHAVIOR

territorial defense or when fending off predators). By influencing territorial

spacing, predator-prey dynamics, and social-group membership, aggressive
interactions can influence how individuals arrange themselves in space and
time and can have large-scale ecological consequences. The importance of evo-
lutionary feedback effects resulting from ecological consequences of aggression
is a new and exciting area of research that may prove critical to our under-
standing of why populations and species vary.
Partly because of the importance of aggression to fitness and partly because
aggressive behaviors are relatively easy to observe and characterize in natural
populations, aggression is well studied in diverse contexts and species, and
thus evolutionary studies of aggression provide unique insight into the evolu-
tion of behavior more generally. In this chapter, I will first summarize recent
work on personality differences in aggression to explore the novel insights this
work poses for understanding the evolution of aggression, and I will then dis-
cuss the role of ecological and evolutionary feedbacks in maintaining popula-
tion- and species-level differences in aggression.

WHAT IS AGGRESSIVE PERSONALITY VARIATION?

Personality variation refers to consistent differences in behavior among
individuals. The study of nonhuman animal personalities is a recent addition
to the field of behavioral ecology and is set against a background of several dec-
ades of research on the evolution of animal behavior from an optimality per-
spective (Sih et al., 2004; Bell, 2007; McNamara et al., 2009), where it is
assumed that animals strategically adjust their behavior to maximize fitness
given existing trade-offs (Roff, 1994). Behavioral ecologists have used this
framework—which assumes that constraints to behavioral flexibility are weak
or nonexistent and that behavior of individuals within a population will con-
verge on a single optimal expression in a particular context—to predict when
and how individuals should reversibly adjust their behavior in different con-
texts (Krebs & Davies, 1991). Yet recent studies demonstrating the ubiquity
of animal personalities challenge these basic assumptions because they show
that individuals are often limited in the flexibility of their behavior, that there
are often pronounced differences in behavior among individuals in the same
context, and that distinct behaviors are often closely correlated in expression
(Dingemanse & Réale, 2005; Bergmüller & Taborsky, 2010).
A particularly puzzling component of personality variation is the remark-
able consistency in behavior across contexts even when changing behavior
would confer higher fitness (Sih et al., 2004). For example, in fishing spiders
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 297

(Dolomedes triton), females that are most aggressive in a foraging context are
also most aggressive in a mating context, and aggressive females are likely to
cannibalize prospective mates before copulating with them to the extent that
some females attack every single male and thus fail to mate and produce off-
spring (Arnqvist & Henriksson, 1997; Johnson & Sih, 2005). In another
example, in many species of birds, the most aggressive males invest the least
in parental care, and this often results in lower reproductive success (e.g., Tut-
tle, 2003; Duckworth, 2006a; see Ketterson & Nolan, 1994, for review).
These costs to inflexible expression of aggression raise the question of why
individuals that behave aggressively in one context cannot modify their overall
behavioral strategy to behave less aggressively in a different context.
Variation in aggressiveness has been one of the most commonly described axes
of personality variation. Consistent differences in aggression have been docu-
mented in insects, spiders, fish, birds, and mammals. Such stability in behavior
is usually detected as significant repeatability either over time or in distinct func-
tional contexts (e.g., aggression toward a rival male versus aggression towards a
predator). A meta-analysis of studies that compared repeatability of 13 different
classes of behavior showed that aggressive behavior was one of most highly repeat-
able (Bell et al., 2009). Yet there is a rich history of work on aggression showing
that it is often context dependent and strongly influenced by environmental con-
ditions. For example, increases in group size and experimental decreases of food
resources both independently increase the number of aggressive interactions in
red deer (Cervus elaphus) stags (Appleby, 1980; Bartoš, 1986). Moreover, many
studies of territorial animals show that individuals are more aggressive toward
strangers than toward neighbors (Jaeger, 1981; Temeles, 1994), demonstrating
that individuals often modulate aggressiveness depending on social context. Even
abiotic conditions have been shown to influence aggressiveness—in coral reef fish
(Pomacentrus spp.) aggressiveness increased with increases in water temperature
(Biro et al., 2010). How can we reconcile such seeming flexibility of aggression with
the growing body of work that shows aggression to be highly repeatable within individ-
uals? The key to resolving this puzzle is to recognize that most studies investigating
the effects of environmental variation on aggression often ignore individual varia-
tion and focus instead on mean population-level changes. Moreover, high repeat-
ability does not preclude the possibility that individuals modify their behavior in
different situations but simply means that the rank order of aggressiveness of a
group of individuals is consistent over time or across different contexts such that,
even if all individuals decrease their aggressive response in a particular context, the
most aggressive individuals in one context are still the most aggressive in the other
(see Figure 10.1 for an example).
Figure 10.1. Examples of consistent individual differences in a highly plastic behavior.
Coral reef fish show plasticity in activity (top graph), aggression (middle graph), and
latency (bottom graph) in relation to temperature. Rank order of individuals is con-
stant across temperatures for aggression and latency but not for activity levels. (Figure
from Biro et al., 2010. Used by permission of the Royal Society.)
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 299

Such consistent differences in aggressive behavior among individuals raise

several important questions. Why do individuals show consistency in the expres-
sion of aggression even when placed in different situations where the costs and ben-
efits of displaying aggressive behavior vary? To what extent is aggression correlated
with other traits? Why do such correlations evolve, and what are their evolutionary
implications? Are there any general patterns across species regarding which traits
are associated with aggression? Finally, how do individual, population-, and
species-level variation in aggression impact ecological and evolutionary processes?

WHY ARE INDIVIDUALS CONSISTENT IN EXPRESSION OF AGGRESSION?

There are two main answers to these questions: natural selection favors
constancy in expression (hereafter “adaptive hypothesis”) or developmental
constraints limit flexibility of aggression (hereafter “constraint hypothesis”).
The adaptive hypothesis assumes there is unlimited potential for flexibility of
aggression within individuals but that stability is adaptive. In other words, if
extreme flexibility of aggression were adaptive, then it would easily evolve
given enough time. Stability of aggression could be adaptive if natural selec-
tion favors either its correlation with other, less flexible traits (see below for
examples) or its predictability (McElreath & Strimling, 2006; Wolf et al.,
2007; McNamara et al., 2009). In contrast, under the constraint hypothesis,
there are intrinsic limits to flexibility of aggression, and thus selection
is assumed to play a minor or no role in the evolution of consistency
(Duckworth, 2010). From this perspective, the physical structures that under-
lie variation in aggression, including variation in brain anatomy, neuronal con-
nectivity, neurotransmitter synthesis and degradation, hormone secretion
patterns, hormone receptor distribution, and endocrine gland function, may
be limited in their flexibility, and in turn this limits flexibility of aggression.
In other words, under this view, no matter the intensity of natural selection,
there is a limit to how much flexibility of aggression can evolve.
The relative importance of selection and constraint in the evolution of sta-
ble differences in aggression between individuals is not clear. Many of the
adaptive hypotheses predict strong correlations between aggression and other
traits, and while such correlations are often found (see below for examples),
it is unclear whether these correlations are a cause or consequence of limited
flexibility of aggression (Duckworth, 2010). Adaptive hypotheses also predict
variability among species in whether they express consistent differences in
aggression, whereas the constraint hypothesis predicts that limits to flexibility
should be widespread across a diversity of taxa, especially if they arise from
limitations on organismal design due to physical laws (sensu Maynard Smith
300 ANIMAL BEHAVIOR

et al., 1985; Brakefield, 2006). Evidence for high repeatability of aggression

across studies supports the constraint hypothesis; however, more rigorous
comparison of flexibility of aggression across a diversity of taxa as well as a
comparison of developmental mechanisms underlying individuals differences
in behavior are needed before any conclusions on this topic can be drawn.
Unfortunately, very little is known about the developmental basis of indi-
vidual differences in aggression as most studies of behavioral development
focus on larger-scale differences between the sexes, between normal and
mutant phenotypes, or between species. Thus, there is currently very little
data linking naturally occurring individual variation in behavior to neural
and endocrine differences among individuals. However, there is some prelimi-
nary support for the idea that constraints are important. The constraint
hypothesis predicts that similar developmental pathways would underlie indi-
vidual variation in aggression across disparate taxa (Duckworth, 2010). Recent
reviews suggest that variation in aggression is linked to variation in the sero-
tonin signaling system (Figure 10.2) across a wide variety of species, from

Figure 10.2. Serotonin (5-HT) pathways that underlie variation in aggression are
remarkably conserved across disparate taxa. (a) One pathway from genes to aggres-
sion involves key enzymes involved in synthesis (TPH2), degredation (MAO A), and
transport (SERT) of serotonin. (b) A shorter pathway involves changes in 5-HT
receptors. (Figure from Popova, 2006. Used by permission of John Wiley and Sons.)
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 301

crayfish to foxes to humans (Popova, 2006), providing some preliminary evi-

dence for this prediction. Yet, at the same time, other physiological systems
that are known to influence aggression, such as hormone variation, show
extensive flexibility over evolutionary time in their association with aggression
(Hau, 2007; Wingfield et al., 2007). For example, aggression is closely linked
to testosterone levels in males of many temperate songbird species but is very
low and not responsive to territorial challenges in tropical species (Wingfield
et al., 2007). Moreover, many bird species that are territorial in the nonbreed-
ing season when gonads are regressed (reduced in size and nonfunctional) also
show dissociation between testosterone and aggression (e.g., Schwabl &
Kriner, 1991; Soma et al., 2000). Finally, in a recent study, there was no cor-
relation between testosterone levels and aggressive personality differences in
western bluebirds (Sialia mexicana) (Duckworth & Sockman, 2012). These
studies suggest that the association between testosterone and aggression is flex-
ible over an evolutionarily timescale and that endocrine responsiveness in
adulthood is not likely to be an important constraint on the evolution of
behavioral flexibility. However, hormones may still act during early develop-
ment to influence aggressive personality as recent research on both humans
and model lab organisms has demonstrated a link between personality varia-
tion and hormone exposure early in ontogeny (Carere & Balthazart, 2007;
Hines, 2008). In sum, more work is necessary to understand how natural
selection shapes aggression and how individual differences in aggression
develop in order to understand why consistent differences in aggression are
so common across a wide variety of animal species.

EVOLUTION OF CORRELATIONS BETWEEN AGGRESSION AND OTHER TRAITS

Individuals vary in morphological, behavioral, and life history traits that in-
fluence the costs and benefits of aggressive behavior either through functional
links or through trade-offs. Thus, associations between other traits and aggres-
sion can evolve if selection favors their coexpression. Once formed, correla-
tions between traits can constrain their independent evolution and can
significantly affect each trait’s future response to selection (Riska, 1989; Roff,
1997; Lynch & Walsh, 1998). In this section, I will review evidence for asso-
ciations between aggression and other traits such as body size, dominance,
parental care, and other personality traits as these are the traits most com-
monly predicted to be linked to aggression. I will also discuss the implications
of these correlations for understanding the function and evolution of aggres-
sion and the extent to which these correlations might constrain adaptive evo-
lution of aggression.
302 ANIMAL BEHAVIOR

Body Size and Aggression

Body size and aggression are often functionally linked because the costs of
initiating an aggressive encounter are less for larger compared to smaller ani-
mals. Moreover, differences in aggression can cause differences in body size.
For example, in the desert spider (Agelenopsis aperta), more aggressive individ-
uals are better foragers and thus reach a larger body size because of their ability
to acquire more food (Riechert & Johns, 2003). These clear functional links
between aggression and body size have led to the general prediction of a posi-
tive correlation between aggression and body size; however, even though larger
individuals are more aggressive in some species (e.g., Zack, 1975; Brace &
Pavey, 1978; Dowds & Elwood, 1985; Herrel et al., 2009), in many other
species aggression and body size are either unlinked or smaller individuals are
more aggressive (Just & Morris, 2003; Morrell et al., 2005). For example, in
two species of swordtail fishes (Xiphophorus nigrensis and X. multilineatus),
when the difference in size between fish was very large, contests were settled
without fights and the smaller animal retreated; however, when individuals
were more closely matched in size, 78 percent of observed fights were initiated
by the smaller individual, and in 70 percent of the fights, the fish that attacked
first lost the contest because body size was a better predictor of winning than
aggression (Morris et al., 1995). In a recent study of zebra finches (Taeniopygia
guttata), aggression was only weakly correlated with body size, and this rela-
tionship differed among the sexes such that the two traits were positively cor-
related in males and negatively correlated in females (Bolund et al., 2007). In
more than 50 percent of staged interactions between velvet swimming crabs
(Necora puber), aggression was initiated by the smaller crab, even though it
was less likely to win (Smith et al., 1994). Finally, in western bluebirds, aggres-
sion and body size are not correlated, and moreover, western bluebirds, even
though smaller than their sister species, mountain bluebirds (S. currucoides),
are more aggressive and are competitively superior to them in territorial dis-
putes (Duckworth & Badyaev, 2007). Thus, the main theme that emerges
from a review of studies measuring the relationship between body size and
aggression is that there is no consistent pattern across species in the association
of these traits.
The lack of a reliable relationship between aggression and body size suggests
that there is no consistent strategy that large and small animals pursue in con-
flict situations. In fact, the only consistent pattern that emerges is that fights
usually do not occur when individuals are extremely different in body size as
visual cues apparently provide adequate information about competitive differ-
ences such that individuals do not need to fight to determine dominance.
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 303

Dominance and Aggression

Dominance refers to a status of power of one individual over another indi-
vidual, whereas aggression refers to a specific set of behaviors such as attacks,
chases, or displacements that are elicited with an intent to inflict harm on or
instill fear into another individual. Dominance relationships are often assessed
by observing agonistic interactions between individuals and determining
which one relents. Dominance and aggression are so intricately linked that
many studies use the terms interchangeably (Drews, 1993). While aggression
and dominance are frequently correlated (e.g., Anestis, 2005; Colléter &
Brown, 2011; Riebli et al., 2011), there are many exceptions. For example,
in pronghorn antelope (Antilocapra americana), bank voles (Myodes glareolus),
and domestic pigs (Sus scrofa), aggressive personality and dominance status are
unrelated (Fairbanks, 1994; Bolhuis et al., 2005; Korpela et al., 2011). More-
over, Richard Francis (1983) showed that in paradise fish (Macropodus opercu-
laris), males that were generally dominant across multiple encounters did not
differ in their aggressiveness from males that were generally submissive. In
fact, after subjecting paradise fish to five generations of bidirectional selection
for dominance, he showed that even though the lines diverged significantly in
their dominance by the end of the experiment, they did not differ in aggres-
sion (Figure 10.3). Wolves (Canus lupus) provide another illustrative example.
Despite the prevailing view that a wolf pack is a group of individuals aggres-
sively vying for dominance, in naturally occurring wolf packs aggression is
rarely used to establish dominance (Mech, 1999). In fact, in one free-living
pack observed over a 13-year period, no aggressive interactions were ever
observed (Mech, 1999). This example appears to characterize the majority of
wild pack behavior as the typical wolf pack is a family and the breeding pair
is able to maintain its status without aggression. Submissive behavior, rather
than being a response to aggression, is performed by the offspring toward
the breeding pair or occasionally by the breeding female to the breeding male.
In the wild, only larger packs including nonkin show aggressive behavior
(Bradshaw et al., 2009).
To explain the lack of a general relationship between dominance and
aggression, Francis (1988) put forward an intriguing possibility—that domi-
nance status is not related to individual differences in aggression despite the
fact that dominance relationships are often established through use of aggres-
sive behavior. To understand this perspective it is important to distinguish
between measurements of an individual’s overall level of aggressiveness outside
the context of a specific dominance interaction and what individuals actually
do during a dominance interaction. In other words, even though dominant
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 305

Figure 10.3. Divergent selection lines. Artificial selection experiments are a tool fre-
quently used by evolutionary biologists to gain insight into the genetics of trait
variation. These experiments start with a group of individuals that vary in the trait of
interest. “High” and “low” selection lines are created from this initial group by
selecting only individuals that display either the highest or lowest expression of the
focal trait to reproduce (e.g., only individuals in the top or bottom 25 percent of trait
values are selected for breeding). These experiments can provide insight into many
aspects of the genetics of a trait, including the symmetry of response to upward and
downward selection (Pitnick & Miller, 2000) and whether there are correlated
responses in other traits to selection on the focal trait (Schwarzkopf et al., 1999).
Correlated responses are commonly caused by pleiotropy, in which one gene affects
more than one trait. Thus, artificial selection experiments can shed light on how traits
are linked. In a study by Francis (1984), an artificial selection experiment was carried
out in paradise fish (Macropodus opercularis) for five generations. Both high and low
lines for dominance were created where only individuals ranking in the top and bot-
tom ~20 percent in dominance status were selected to breed. The results showed a
significant response to selection in the downward but not the upward selected lines
(top graph). However, when the same selection lines were tested for aggression by
exposing them to an intruder fish (either another male separated by a glass partition or
a mirror), they showed no difference in aggression from one another (bottom graph).
This experiment showed that divergence in the dominance scores of the two lines was
not accompanied by any changes in aggressiveness, showing that these traits are not
linked in this species. (Figures adapted from Francis, 1984)

individuals may occasionally use aggression to establish or maintain domi-

nance, this does not mean that they are generally more aggressive than other
individuals.
If aggression plays only a limited role, then what factors are most important
in influencing dominance? Dominance is an emergent property influenced by
multiple interacting factors including both intrinsic characteristics of the indi-
vidual as well as social context (Weiß et al., 2011). Intrinsic characteristics
include differences in sex, body size, motivation, prior experience, age, and
sexual traits (e.g., Watt, 1986; Lemel & Wallin, 1993; Elwood et al., 1998;
Nosil, 2002; Duckworth et al., 2004). In general, males are dominant to
females, larger individuals are dominant to smaller ones; older, more experi-
enced individuals are dominant to younger, less experienced ones; and motiva-
tion can overturn any of these generalities (Cristol, 1992; Lemel & Wallin,
1993), especially if there is a large difference between individuals in the bene-
fits of accessing a particular resource (Enquist & Leimar, 1987). For example,
in house crickets (Acheta domesticus), body size usually determines the out-
come of dominance interactions; however, when motivation is maximized by
food deprivation, this overrides the effects of male body size (Nosil, 2002).
306 ANIMAL BEHAVIOR

Furthermore, in natural populations of birds, the effects of motivation have

been shown to be an important determinant of dominance even in species
where, all else being equal, the largest or most ornamented males win contests
(Lemel & Wallin, 1993).
Recently, a more nuanced view of aggression’s relation to dominance is
emerging where it is less important in establishing long-term dominance relation-
ships and instead is more important for short-term or initial interactions between
individuals. Support for this idea comes from studies of aggression in primate
species. Multiple experiments that attempted to elicit aggressive interactions from
stable primate groups in captivity (e.g., by making food or space more scarce)
failed (Bernstein & Gordon, 1974). The only predictable trigger of aggression
occurred when a foreign individual or individuals were introduced into a stable
group. Typically, the intruding animal was severely attacked, but initial high lev-
els of aggression declined rapidly, especially when the intruding animal behaved
submissively, suggesting that aggression was motivated primarily by the need to
maintain social order rather than as a means for mediating competition for re-
sources (Bernstein & Gordon, 1974; de Waal, 1986). In domestic cats (Felis
catus), dominance sustained without the use of aggression was more stable than
dominance formed on the basis of aggressive display (Fonberg, 1988). Many
studies of birds have shown that once individuals get to know each other well,
they do not use aggression to maintain dominance (Temeles, 1994). Finally, in
Chasmognathus crabs, Silvia Pedetta and colleagues found that aggression is the
main determinant of dominance between size-matched individuals (Pedetta et
al., 2010). In fact, aggression even overrides prior experience in this species—in
lab trials, even when pitting the same individuals against one another multiple
times, the crabs fight anew, and there seems to be no memory of previous
encounters. They suggest this makes sense in the context of this species’s natural
history as interactions with conspecifics occur mainly in the context of burrow
disputes between resident and wandering crabs, and it is rare that fights would
occur between the same opponents multiple times. Thus, there is no benefit
for individual recognition mechanisms to evolve in this species. These studies
all suggest that aggression may be more important for short-term or one-time
interactions and less of a factor for maintaining long-term dominance hierarchies.
The implications are that it may be important to understand the expected
duration of conflict situations in order to understand when and why individual
variation in aggression evolves.

Parental Care and Aggression

Aggression is costly in terms of time, energy, and risk, and thus is expec-
ted to trade off with investment in other costly behaviors and traits (Bennett
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 307

& Houck, 1983; Robertson, 1986; Johnstone & Norris, 1993; Duckworth,
2006a; Rosvall, 2011). In fact, one of the most consistent relationships
between aggression and other traits across a wide variety of species is a negative
correlation between aggression and parental care. For example, in many pass-
erine birds, males that show high levels of aggression often invest the least in
parental care (Ketterson & Nolan, 1994). Because, in these species, males
and females form strong pair bonds and both sexes contribute to offspring
care, this means that highly aggressive males often have lower reproductive
success (e.g., Duckworth, 2006a). In some of these species, variation in aggres-
sion is related to alternative male mating strategies within a population, where
aggressive males compensate for fitness costs of low parental care by actively
pursuing mating opportunities outside the pair bond, whereas other males
are nonaggressive and good fathers (Ketterson & Nolan, 1994). Why aggres-
sion, mating behavior, and parental care are frequently linked is not clear.
It is known that all of these behaviors are influenced by circulating testoster-
one levels—testosterone is necessary for sperm production and hence mating,
frequently correlates with aggression, and is antagonistic to the expression of
parental behavior. These links have led to the idea that negative correlations
between aggression and parental care are primarily due to the joint effects of
testosterone. However, whether aggression and parental behavior are linked
directly or whether they are instead only correlated through testosterone is still
an open question. If the latter, then in species where testosterone and aggres-
sion are uncoupled, there should not be a trade-off between aggression and
parental behavior. Direct links between aggression and parental care could
occur if there is a fundamental neural trade-off where individuals “wired” to
be aggressive cannot also be parental, and vice versa. Finally, it is also possible
that there is a basic time or energy trade-off that links these two behaviors such
that aggressive individuals spend so much time fighting they do not have time
or energy left for offspring care.
Recent studies suggest that the relationship between aggression and
parental behavior may be more complex than originally thought and that the
antagonistic effects of testosterone on parental behavior may not always be
present (see Lynn, 2008, for review). For example, in western bluebirds,
aggressive personality differences are unrelated to natural variation in testoster-
one even though there is a negative relationship between aggression and male
parental care in this species (Duckworth, 2006a; Duckworth & Sockman
2012). Moreover, in California mice (Peromyscus californicus), testosterone is
actually required to maintain high levels of paternal behavior, paternal behav-
ior and aggression are positively correlated, and testosterone and aggression are
not related in a simple way (Trainor & Marler, 2001). Both castration and
308 ANIMAL BEHAVIOR

experimentally increased testosterone did not influence aggressive response in

this species; instead only the control males increased their aggression in
response to a social challenge. The authors suggest that this counterintuitive
result may be because this was the only group whose level of aggression was
allowed to fluctuate naturally, suggesting that testosterone responsiveness to
social challenge, rather than mean level of testosterone, may be more impor-
tant in modulating aggressiveness in this species. Finally, even in most bird
species, recent studies are showing that some species are “behaviorally insensi-
tive” to testosterone such that experimental increases in testosterone do not
increase aggressive behavior and fail to dampen parental behaviors (Lynn,
2008). Thus, taken together, these studies show that testosterone does not
universally mediate the trade-off between aggression and parental care across
species. Moreover, this relationship was mainly developed in songbirds, and,
as shown by the California mice example, more evidence is needed from a
broader array of taxa to determine whether the trade-off between aggression
and parental care is universal.

Temperament Traits and Aggression

In a recent review, Denis Réale and colleagues characterized aggressiveness
as one of five temperament (or personality) categories that also include
shyness-boldness, exploration-avoidance, activity, and sociability (Réale et al.
2007). Correlations among these distinct behavioral axes are widespread
and are referred to as behavioral syndromes (Sih et al., 2004). Many studies
have documented an aggression-boldness syndrome—where animals that are
more aggressive are also bolder and more explorative in novel environments
(first described by Huntingford, 1976; see Norton et al., 2011, and citations
therein for examples). Although correlations between these personality axes
are common, they are not ubiquitous and appear to be maintained by natural
selection. For example, in threespine sticklebacks (Gasterosteus aculeatus) the
presence of this behavioral syndrome is correlated with predation pressure
across populations such that in populations with high predation the correla-
tion between boldness and aggression is strong, but it breaks down in popula-
tions with less intense predation (Bell & Stamps, 2004; Dingemanse et al.,
2007). As an adaptive explanation for such a pattern, Niels Dingemanse and
colleagues suggested that in ponds with predators, spatial variation in preda-
tion risk might favor evolution of alternative solitary or shoaling strategies
where solitary individuals monopolize a habitat patch that is poor in food
but safe and shoaling individuals roam patches of habitat that are relatively
dangerous but rich in food (Dingemanse et al., 2007). Competition for safe
patches would be intense, and so solitary individuals should be aggressive
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 309

and also more exploratory because they have to find patches on their own,
whereas shoaling individuals need to be more tolerant of neighbors and can
also rely on them to acquire foraging information, so they do not need to be
as explorative.
Such adaptive explanations for correlations between aggression and explor-
atory behavior still need to be tested, but trade-offs originating from differ-
ences in social strategy might be the key to understanding correlations
between personality traits as these correlations are often found in species where
there is wide variation in social strategy between individuals (Cote & Clobert,
2007; Bergmüller & Taborsky, 2010). Examples include freshwater fish,
where solitary individuals are both more explorative and aggressive than
social individuals (Ward et al., 2004); Myrmica ants, where there are
strong correlations between aggression, boldness, activity, and sociability
at the individual, caste, and colony levels (Chapman et al., 2011); the co-
operatively breeding cichlid (Neolamprologus pulcher), where an aggression-
boldness-explorativeness syndrome was linked to female helping behavior
(Schürch & Heg, 2010); and the socially polymorphic comb-footed spider
(Anelosimus studiosus), where social individuals were less aggressive, less active,
and less responsive to prey (Pruitt et al., 2008). Social conflict can select for
stable coexistence of different behavioral types, and this has led to the idea that
personality differences may evolve primarily in response to social environment
(Bergmüller & Taborsky, 2010). While this is an intriguing idea, the evidence
so far is only correlative, and the direction of causality between different per-
sonality axes is not clear. After all, differences in aggression between individ-
uals might be as likely to lead to differences in sociability as the reverse.

Correlations between Aggression and Other Traits: Ever-present, but Not Consistent
Two main themes emerge from a survey of correlations between aggression
and other traits: (1) such correlations are widespread across a diverse array of taxa
and (2) there are no consistent patterns across taxa in the specific traits that are
correlated with aggression. Aggression is closely linked to body size and domi-
nance in some species but not others. It frequently covaries with aspects of life
history investment, such as parental care, but not consistently across species.
Many species show strong correlations between aggression and other compo-
nents of personality variation, but the strength and presence of these correlations
vary across species and even across populations within a species.
Such diversity in the strength and direction of correlations has important
implications for understanding the evolution of distinct aggressive phenotypes.
First, it reinforces the importance of aggression in a wide variety of contexts
310 ANIMAL BEHAVIOR

and life histories. Second, it suggests that the evolution of aggression is not con-
strained by correlations with other traits, as these correlations can break up and
be reformed in a relatively short time span (~10,000 years in the case of stickle-
back populations that differ in the expression of behavioral correlations; Dinge-
manse et al., 2007). Finally, it supports the notion that aggression, rather than
being an emergent property of other components of the phenotype, is a trait in
its own right. After all, if aggression was always consistently correlated with other
traits such as boldness or activity levels, this could indicate that these are not re-
ally separate traits at all but simply the distinct responses reflecting a common
underlying temperament or coping style. In the next two sections, I discuss
how individual, population, and species differences in aggression can influence
ecological and evolutionary processes and how ecological and evolutionary feed-
backs on aggression might provide the key to understanding both the diversity
and ubiquity of correlations between aggression and other traits.

ECOLOGICAL CONSEQUENCES OF AGGRESSION

There is a long history of research that investigates the role of aggressive
interactions in population and community ecology (Walls, 1990; Amarase-
kare, 2002; Peiman & Robinson, 2010). By directly affecting competitive
interactions between individuals, variation in aggression can influence individ-
ual spacing patterns, population dynamics, and population cycles. Moreover,
recent studies have found links between aggression and dispersal behavior
(e.g., Rusu & Krackow, 2005; Duckworth & Badyaev, 2007; Raihani et al.,
2008). Thus, aggression may also influence population connectivity and the
dynamics of colonization. In fact, recent studies have indicated that distinct
aggressive phenotypes may play a role in species range expansions as well as
the success of invasive species. In this section, I will review the evidence that
variation in aggressive behavior has strong impacts on ecological dynamics.

Aggression, Individual Spacing, and Population Cycles

In territorial species, aggression can influence individual spacing, which in
turn can have large effects on population density and potentially even popula-
tion cycles (Adams, 2001). One of the earliest proponents of linking aggres-
sion to population cycles was Dennis Chitty (1952), who suggested that at
high density selection favors large, aggressive animals with low reproductive
rates, and at low densities it favors smaller, less aggressive animals with high
reproductive rates (Chitty, 1967). Subsequent researchers expanded on these
ideas to take into account the potential role of kin interactions (Charnov &
Finerty, 1980). However, experimental tests of these ideas have not held up
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 311

in cycling populations of microtine voles (Boonstra & Hogg, 1988; Boonstra

et al., 1994), leading to the conclusion that intrinsic changes in behavior alone
could not account for populations cycles (Stenseth & Łomnicki, 1990;
Stenseth et al., 1996).
Recent work in birds has shown that, in conjunction with extrinsic factors,
aggression can play a key role in population cycles. In red grouse (Lagopus lago-
pus scoticus), there has long been a debate about the relative importance of
intrinsic changes in aggression versus extrinsic fluctuations in parasite loads
in driving population cycles. While experimental studies have shown that
changes in parasitism can produce cycles through their effects on breeding
productivity (Hudson, 1986), there is also strong evidence that population
cycles are caused by annual variation in the aggressiveness of males through
its effects on population density and recruitment of new individuals into the
population (Mougeot et al., 2003). Researchers were able to mimic population
cycles in a Scottish population by experimentally increasing aggressiveness
with testosterone implants in four separate populations. As aggression
increased, males expanded their territories and recruitment of new males into
the population declined, resulting in a breeding density that was reduced by
50 percent, changing the populations’ trajectories from increasing to declin-
ing. However, while the effects of aggression on population cycles were clear,
it was not clear what ultimately drove changes in aggression. Subsequent stud-
ies showed that high levels of testosterone increased parasite infection, which
in turn decreased male aggressive behavior (Fox & Hudson, 2001; Seivwright
et al., 2005). Thus, the most recent consensus suggests that a combination of
extrinsic fluctuations in parasite loads and intrinsic fluctuations in aggression
are necessary to explain the observed population cycles (New et al., 2009).
Aggression is also linked to population density and population cycles in
bluebirds (Sialia spp.). Western bluebirds depend on tree cavities to breed—
a limiting resource that historically was patchily distributed and ephemeral.
Nest cavities occur at high densities following forest fires, which create suitable
habitats for bluebirds by opening up understory vegetation and creating dead
snags where nest holes are abundant. Eventually, as the forest regrows, blue-
birds are no longer able to breed in these habitat patches because snag density
decreases and regrowth of the forest eliminates the open meadows bluebirds
depend on to forage for insect prey (Power & Lombardo, 1996; Guinan
et al., 2000). Western bluebirds’ sister species, mountain bluebirds, are fre-
quently among the earliest colonizers following forest fires (Hutto, 1995),
whereas, western bluebirds often show delayed patterns of colonization (Saab
et al., 2004; Kotliar et al., 2007). Competition for nest cavities among these
and other secondary-cavity-nesting species is intense and often involves
312 ANIMAL BEHAVIOR

aggressive displacement (Gowaty, 1984; Newton, 1994; Merilä & Wiggins,

1995). Western bluebirds, while less dispersive and slower to find new habitat,
are on average more aggressive than mountain bluebirds and rapidly displace
them when they colonize new areas (Duckworth & Badyaev, 2007). The
maintenance of these cycles of species replacement depends at least partly on
the evolution of two distinct dispersal strategies in western bluebirds in which
dispersal and aggression are closely linked. Highly aggressive males tend to
leave their natal populations and disperse to new areas to breed—these newly
colonized areas initially have a very low population density, and this enables
aggressive males to obtain large territories. On the other hand, nonaggressive
males tend to remain near where they were born, which usually is a much
older population with a higher density of western bluebirds than newly colon-
ized areas (Duckworth & Badyaev, 2007; Duckworth, 2008). These patterns
of biased dispersal with respect to aggression produce a strong correlation bet-
ween population age, density, and aggressive behavior. Newly colonized pop-
ulation are less dense but highly aggressive, whereas older, well-established
populations are less aggressive and have higher densities.
These avian examples show that aggression can be an important determi-
nant of population cycles, however, not in the ways originally envisioned by
Chitty (1967)—body size and aggression are not linked in these species, and
in both examples, aggression was associated with lower population density
because more aggressive individuals are more likely to space themselves farther
apart and social tolerance among related individuals enables them to breed at
higher density (Lambin & Krebs, 1991). It remains to be seen whether cycles
of aggression might play an important role in other classic systems that show
populations cycles, such as snowshoe hares (Lepus americanus) and the
Norway lemming (Lemmus lemmus).

Aggression, Invasion, and Range Limits

Recent studies have linked differences in aggression between populations
and species to the dynamics of invasion and range expansion. The Argentine
ant (Linepithema humile) is an invasive species whose success has been attrib-
uted at least in part to its aggressive displacement of other species. Introduced
populations have undergone genetic bottlenecks that have led to very low
genetic diversity and the growth of “super-colonies” with nests that essentially
function as a single colony spread over many kilometers (Suarez et al., 2008).
Nests within these super-colonies are not aggressive toward one another due to
their genetic similarity but are very aggressive to other species. In contrast, in
the native range, Argentine ants do not form such super-colonies, display
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 313

much higher levels of intraspecific aggression, and coexist with a diverse com-
munity of ant species. These patterns suggest that changes in colony structure
and aggression in introduced populations have facilitated invasion success by
decreasing intraspecific competition and thus enabling them to outcompete
other species (Holway, 1999; Tsutsui et al., 2003). Finally, the invasive red-
eared slider (Trachemys scripta elegans) and endangered native Spanish terrapin
(Mauremys leprosa) in the Iberian Peninsula provide another example linking
aggression and invasiveness. Red-eared sliders are more aggressive than
Spanish terrapins and outcompete them during foraging, significantly restrict-
ing their access to food resources (Polo-Cavia et al., 2011).
Aggressive differences between species are not just linked to successful
invasions but have also been shown to be important in the dynamics of natural
range expansions and in determining species’ range limits (Peiman &
Robinson, 2010). Scott Pearson and Sievert Rowher (2000) found that com-
petitive superiority of Townsend’s warblers (Dendroica townsendi) over hermit
warblers (D. occidentalis) is causing a hybrid zone to move in these species,
thereby expanding Townsend’s warblers’ range at the expense of hermit war-
blers. They showed that this competitive difference between the species was
largely attributable to differences in aggression—Townsend’s warblers are
more aggressive than hermit warblers. Similarly, the recent range expansion
of the barred owl (Strix varia) at the expense of the threatened northern spot-
ted owl is at least partly due to its higher aggression (Van Lanen et al., 2011).
In bluebirds, competitive superiority of western bluebirds over mountain
bluebirds is largely due to the highly aggressive nature of western bluebirds
that colonize new populations (Duckworth, 2008). This competitive differ-
ence was most obvious during western bluebirds’ recent range expansion, in
which nest-box programs enabled them to rapidly recolonize areas in the
northwestern United States where they had previously gone extinct due to
the loss of natural nest cavities. The expansion of their range back to their his-
torical range limits was accompanied by the rapid displacement of lower-
elevation mountain bluebird populations (Duckworth & Badyaev, 2007).
The range expansion was a natural experiment that provided insight into the
competitive dynamics of these species—it showed that mountain bluebirds
are limited at the lower edge of their range by competition with western blue-
birds rather than by abiotic or other ecological factors.
Such competitive exclusion through direct aggressive interactions is a
common theme, especially in the context of species range limits across eleva-
tional gradients. In two tropical bird genera—Catharus thrushes and Henico-
rhina wrens—asymmetries in aggressiveness explained nonoverlapping ranges
across an elevational gradient in tropical forests (Jankowski et al., 2010).
314 ANIMAL BEHAVIOR

Differences in aggression among four Eutamias chipmunks occurring across an

elevational gradient in the Sierra Nevada Mountains is thought to at least par-
tially explain their nonoverlapping distributions (Heller, 1971). In Plethodon
salamanders, evidence suggests that high levels of interspecific aggression have
evolved in populations in the Great Smoky Mountains, causing elevational
range segregation between P. glutinosus and P. jordani, whereas in the Balsam
Mountains interspecific aggression was largely absent and the two species’
ranges overlapped extensively (Hairston et al., 1987). One pattern that
emerges across these studies is that the more aggressive species usually lives
in the more mild ecological conditions. This pattern is not confined to eleva-
tional gradients—red foxes (Vulpes vulpes; Figure 10.4) are more aggressive
than and competitively dominant to arctic foxes (Alopex lagopus) (Frafjord
et al., 1989). Evidence suggests that the arctic fox is limited at its southern
range edge by the more aggressive red fox and the red fox is limited at its
northern range edge by its inability to cope with the extreme climactic condi-
tions of the arctic (Hersteinsson & MacDonald, 1992; Tannerfeldt et al.,
2002). Thus, high levels of aggression may be necessary to exclude

Figure 10.4. Red foxes (Vulpes vulpes) fighting. Red foxes are more aggressive than arc-
tic foxes (Vulpes lagopus), and their competitive superiority may be at least partly
responsible for declining arctic fox populations in Scandinavia. (Alex Badyaev,
www.tenbestphotos.com)
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 315

competitors from areas of abundant resources but may come at the cost of sur-
viving and breeding in more extreme ecological conditions.

EVOLUTIONARY CONSEQUENCES OF AGGRESSION

Strong ecological consequences of aggressive interactions can affect future
evolutionary trajectory of populations. Aggressive behavior has great potential
to affect selection pressures because it is often used to obtain a breeding
territory (Stamps & Krishnan, 1997) and therefore can affect individual fitness
by determining the quality of environment in which offspring develop.
Western bluebirds provide one of the clearest examples of territorial aggression
influencing natural selection. As secondary cavity nesters, their nest sites are
extremely limited, and as a consequence, males prefer to acquire territories
with multiple nest cavities (Meek & Robertson, 1991; Plissner & Gowaty,
1995); however, only the most aggressive males are able to compete successfully
for these territories (Duckworth, 2006b). Using this knowledge, I tested the idea
that aggressive interactions over nest cavities could cause males to sort into dis-
tinct habitats, which in turn could influence evolution of morphology (Duck-
worth, 2006b). By placing a high density of nest boxes in open habitat with
very low tree cover and a low density of nest boxes in closed habitat with high
tree cover, I experimentally caused aggressive interactions to sort males into these
different habitats. Aggressive males acquired territories with multiple nest boxes
in the open habitat, and nonaggressive males were pushed into the closed habitat
where they acquired territories with only a single nest box. Most importantly,
males experienced differential selection on morphology across these habitat types.
Specifically, males with longer tails and legs were favored in open habitats where
high agility is required to forage efficiently, whereas in forested habitats, where
agility is less important, selection on morphology was weak. These results showed
that aggression can affect selection on a local scale by determining individual set-
tlement patterns. Moreover, because such sorting caused a correlation between
aggression and selection on body size, this study has important implications for
our understanding of how correlations between aggression and other traits might
originate—through a nonrandom link between aggression and habitat type.
Aggressive interactions between species can also have evolutionary conse-
quences through agonistic character displacement (Grether et al., 2009).
Classic character displacement occurs when competition between species
causes them to diverge in traits in populations where their ranges overlap com-
pared to populations where their ranges do not overlap (Grant, 1994). Such
divergence is thought to lessen interspecific competition and enable coexis-
tence in areas of overlap. Agonistic character displacement is a specialized case
316 ANIMAL BEHAVIOR

where divergence in areas of overlap occurs specifically in traits that affect the
rate, intensity, or outcome of competitive interactions. One of the clearest exam-
ples of such agonistic character displacement is the case of brook and ninespine
sticklebacks (Culaea inconstans and Pungitius pungitius, respectively). Kathryn
Peiman and Beren Robinson (2007) showed that brook sticklebacks from popu-
lations that overlap with ninespine sticklebacks are more aggressive than those
from populations that do not overlap, suggesting that there has been selection
for enhanced aggressiveness where these species come into contact. In another
example, Dean Adams (2004) found that robustness of head shape—a trait that
is associated with enhanced fighting ability—showed increases in areas of overlap
for two Plethodon salamanders, P. jordani and P. teyahalee. Moreover, he found a
positive correlation between head shape and aggression. He suggested that these
differences in head shape stemmed from aggressive interference competition—
salamanders that were more aggressive benefited from having a morphology that
increased their fighting ability.
The outcome of aggressive interactions will depend not only on an individ-
ual’s own aggressive behavior but also on the aggressive phenotype of other
individuals in the population. This social context of aggression can produce
novel evolutionary feedback dynamics for aggression and the traits associated
with aggression, particularly because the environment that elicits aggression
(other competitors) can evolve. Such influences of genotypes of other individ-
uals in the population on a focal individual’s aggression are termed indirect
genetic effects (Wolf et al., 1998). The importance of indirect genetic effects
for evolution have only recently been recognized. In one of the few empirical
papers showing indirect genetic effects on aggression, Alastair Wilson and col-
leagues (2009) found a strong positive genetic covariance between a focal
individual’s aggression and the aggression of its opponent. Such covariance
between aggression and the social environment can lead to positive evolution-
ary feedbacks and result in rapid evolution of aggression in the presence of
strong natural selection (Wolf et al., 1998).

CONCLUSIONS
Animals display aggression in a wide range of circumstances from competi-
tion over mates, food, or other resources, to territory defense and offspring pro-
tection, to the establishment of dominance hierarchies within social groups.
The ubiquity and importance of aggression has made it the focus of an immense
amount of research, making studies of the evolution of aggression a rich resource
for understanding the evolution of behavior more generally.
While early studies of aggression focused on trying to understand the opti-
mal expression of aggression within a population (Maynard Smith & Price,
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 317

1973; Parker, 1974), recent studies showing consistent differences among

individuals have shifted the focus to trying to understand the relative impor-
tance of selection and constraint in shaping aggressive phenotypes. These
studies have shown that aggression is among the most repeatable of behavioral
traits, often varies extensively among individuals within populations, and is
frequently correlated with other behaviors. Selection for integration of aggres-
sion with other traits is often cited as a key component of adaptive hypotheses
for the evolution of consistent individual differences, and while the ubiquity of
correlations between aggression and other traits supports this idea, it is not
clear whether consistent differences in aggression are a cause or consequence
of these correlations. Certainly, the diversity of correlations between aggres-
sion and other traits suggests that, if selection for integration is the main cause,
then there is a diversity of ways for selection to produce consistent individual
differences in aggression. At the same time, the idea that developmental con-
straints play an important role in the evolution of constancy in the expression
of aggression needs empirical testing. Under the constraint hypothesis, correla-
tions between aggression and other traits might be a consequence of consistent
differences rather than a cause. More work on the patterns of trait correlations
as well as the developmental basis for differences in aggression across a diver-
sity of species is needed to test these alternative hypotheses.
Finally, individual, population-, and species-level variations in aggression
can have large-scale ecological consequences by influencing population density
and species coexistence. Recent studies show that distinct aggressive pheno-
types play a role in population cycles, range expansions, and the success of
invasive species as well as competitive exclusion at range edges. In turn, these
strong ecological consequences of aggressive interactions can influence evolu-
tionary dynamics of populations, ultimately producing feedbacks that further
influence the evolution of aggression.

ACKNOWLEDGMENTS
I thank Alex Badyaev for comments, which improved this manuscript. This
work was supported by funding from the National Science Foundation (DEB
918095).

REFERENCES AND SUGGESTED READING

Adams, D. C. (2004). Character displacement via aggressive interference in Appala-
chian salamanders. Ecology, 85, 2664–2670.
Adams, E. S. (2001). Approaches to the study of territory size and shape. Annual
Review of Ecology and Systematics, 32, 277–303.
318 ANIMAL BEHAVIOR

Amarasekare, P. (2002). Interference competition and species coexistence. Proceedings

of the Royal Society of London, B, 269, 2541–2550.
Andersson, M., C. G. Wiklund, & H. Rundgren. (1980). Parental defense of
offspring: A model and an example. Animal Behaviour, 28, 536–542.
Anestis, S. F. (2005). Behavioral style, dominance rank, and urinary cortisol in young
chimpanzees (Pan troglodytes). Behaviour, 142, 1245–1268.
Appleby, M. C. (1980). Social rank and food access in red deer stags. Behaviour, 74,
294–309.
Arnqvist, G. & S. Henriksson. (1997). Sexual cannibalism in the fishing spider and a
model for the evolution of sexual cannibalism based on genetic constraints.
Evolutionary Ecology, 11, 255–273.
Bartoš, L. (1986). Dominance and aggression in various sized groups of red deer stags.
Aggressive Behavior, 12, 175–182.
Bell, A. M. (2007). Future directions in behavioural syndromes research. Proceedings
of the Royal Society of London, B, 274, 755–761.
Bell, A. M., S. J. Hankison, & K. L. Laskowski. (2009). The repeatability of behav-
iour: A meta-analysis. Animal Behaviour, 77, 771–783.
Bell, A. M. & J. A. Stamps. (2004). Development of behavioural difference between
individuals and population of sticklebacks, Gasterosteus aculeatus. Animal
Behaviour, 68, 1339–1348.
Bennett, A. F. & L. D. Houck. (1983). The energetic cost of courtship and aggression
in a plethodontid salamader. Ecology, 64, 979–983.
Bergmüller, R. & M. Taborsky. (2010). Animal personality due to social niche
specialisation. Trends in Ecology and Evolution, 25, 504–511.
Bernstein I. & T. Gordon. (1974). The function of aggression in primate societies:
Uncontrolled aggression may threaten human survival, but aggression may be
vital to the establishment and regulation of primate societies and sociality.
American Scientist, 62, 304–311.
Biro, P. A., C. Beckmann, & J. A. Stamps. (2010). Small within-day increases in tem-
perature affects boldness and alters personality in coral reef fish. Proceedings of
the Royal Society of London, B, 277, 71–77.
Bolhuis, J. E., W. G. P. Schouten, J. W. Schrama, & V. M. Weigant. (2005).
Individual coping characteristics, aggressiveness and fighting strategies in pigs.
Animal Behaviour, 69, 1085–1091.
Bolund, E., H. Schielzeth, & W. Forstmeier. (2007). Intrasexual competition in
zebra finches: The role of beak colour and body size. Animal Behaviour, 74,
715–724.
Boonstra, R., W. M. Hochachka, & L. Pavone. (1994). Heterozygosity, aggression,
and population flucuations in meadow voles (Microtus pennsylvanicus). Evolu-
tion, 48, 1350–1363.
Boonstra, R. & I. Hogg. (1988). Friends and strangers: A test of the Charnov-Finerty
hypothesis. Oecologia, 77, 95–100.
Brace, R. & J. Pavey. (1978). Size-dependent dominance hierarchy in the anemone
Actinia equine. Nature, 273, 752–753.
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 319

Bradshaw, J. W. S., E. J. Blackwell, & R. A. Casey. (2009). Dominance in domestic

dogs: Useful construct or bad habit? Journal of Veterinary Behavior, 4,
135–144.
Brakefield, P. M. (2006). Evo-devo and constraints on selection. Trends in Ecology
and Evolution, 21, 362–368.
Carere, C. & J. Balthazart. (2007). Sexual versus individual differentiation: The con-
troversial role of avian maternal hormones. Trends in Endocrinology and
Metabolism, 18, 73–80.
Chapman, B. B., H. Thain, J. Coughlin, & W. O. H. Hughes. (2011). Behavioural
syndromes at multiple scales in Myrmica ants. Animal Behaviour, 82, 391–397.
Charnov, E. L. & J. P. Finerty. (1980). Vole population cycles: A case for kin-
selection? Oecologia, 45, 1–2.
Chitty, D. (1952). Mortality among voles (Microtus agrestis) at Lake Vyrnwy,
Montgomeryshire in 1936–9. Philosophical Transactions of the Royal Society of
London, B, 236, 505–552.
Chitty, D. (1967). The natural selection of self-regulatory behavior in animal popula-
tions. Proceedings of the Ecological Society of Australia, 2, 51–78.
Colléter, M. & C. Brown. (2011). Personality traits predict hierarchy rank in male
rainbowfish social groups. Animal Behaviour, 81, 1231–1237.
Cote, J. & J. Clobert. (2007). Social personalities influence natal dispersal in a lizard.
Proceedings of the Royal Society of London, B, 274, 383–390.
Cristol, D. A. (1992). Food deprivation influences dominance status in dark-eyed
juncos, Junco hyemalis. Animal Behaviour, 43, 117–124.
de Waal, F. (1986). The integration of dominance and social bonding in primates.
Quarterly Review of Biology, 61, 459–479.
Dingemanse, N. J. & D. Réale. (2005). Natural selection and animal personality.
Behaviour, 142, 1159–1184.
Dingemanse, N. J., J. Wright, A. J. N. Kazem, D. K. Thomas, R. Hickling, &
N. Dawnay. (2007). Behavioural syndromes differ predictably between 12
populations of three-spined stickleback. Journal of Animal Ecology, 76, 1128–
1138.
Dowds, B. & R. Elwood. (1985). Shell wars 2: The influence of relative size on deci-
sions made during crab shell fights. Animal Behaviour, 33, 649–656.
Drews, C. (1993). The concept and definition of dominance in animal behaviour.
Behaviour, 125, 283–313.
Duckworth, R. A. (2006a). Behavioral correlations across reproductive contexts pro-
vide a mechanism for a cost of aggression. Behavioral Ecology, 17, 1011–1019.
Duckworth, R. A. (2006b). Aggressive behavior affects selection on morphology by
determining the environment of breeding in a passerine bird. Proceedings of
the Royal Society of London, B, 273, 1789–1795.
Duckworth, R. A. (2008). Adaptive dispersal strategies and the dynamics of a range
expansion. American Naturalist, 172, S4–S17.
Duckworth, R. A. (2010). Evolution of personality: Developmental constraints on
behavioral flexibility. Auk, 127, 752–758.
320 ANIMAL BEHAVIOR

Duckworth, R. A. & A. V. Badyaev. (2007). Coupling of aggression and dispersal

facilitates the rapid range expansion of a passerine bird. Proceedings of the
National Academy of Sciences, USA, 104, 15017–15022.
Duckworth, R. A., M. T. Mendonça, & G. E. Hill. (2004). Condition dependent
sexual traits and social dominance in the house finch. Behavioral Ecology, 15,
779–784.
Duckworth, R. A. & K. W. Sockman. (2012). Proximate mechanisms of behavioural
inflexibility: Implications for the evolution of personality traits. Functional
Ecology, 26, 559–566.
Elwood, R. W., K. E. Elwood, W. B. Gallagher, & J. T. A. Dick. (1998). Probing
motivational state during antagonistic encounters in animals. Nature, 393,
66–68.
Enquist, M. & O. Leimar. (1987). Evolution of fighting behavior: The effect of varia-
tion in resource value. Journal of Theoretical Biology, 107, 187–205.
Fairbanks, W. (1994). Dominance, age and aggression among female pronghorn,
Antilocapra americana (Family: Antilocapridae). Ethology, 97, 278–293.
Fonberg, E. (1988). Dominance and aggression. International Journal of Neuroscience,
41, 201–213.
Fox, A. & P. J. Hudson. (2001). Parasites reduce territorial behaviour in red grouse.
Ecology Letters, 4, 139–143.
Frafjord, K., D. Becker, & A. Angerbjörn. (1989). Interactions between arctic and red
foxes in Scandinavia: predation and aggression. Arctic, 42, 354–356.
Francis, R. C. (1983). Experiential effects on agonistic behavior in the paradise fish,
Macropodus opercularis. Behaviour, 85, 292–313.
Francis, R. C. (1984). The effects of bidirectional selection for social dominance on
agonistic behavior and sex ratios in the paradise fish (Macropodus opercularis).
Behaviour, 90, 25–45.
Francis, R. C. (1988). On the relationship between aggression and social dominance.
Ethology, 78, 223–237.
Garcia, J. T. & B. E. Arroyo. (2002). Intra- and interspecific agonistic behaviour in
sympatric harriers during the breeding season. Animal Behaviour, 64, 77–84.
Gowaty, P. A. (1984). House sparrows kill eastern bluebirds. Journal of Field Orni-
thology, 55, 378–380.
Grant, P. R. (1994). Ecological character displacement. Science, 266, 746–747.
Grether, G. F., N. Losin, C. N. Anderson, K. Okamoto. (2009). The role of interspe-
cific interference competition in character displacement and the evolution of
competitor recognition. Biological Reviews, 84, 617–635.
Guinan, J. A., P. A. Gowaty, & E. K. Eltzroth. (2000) Western bluebird (Sialia mex-
icana). In A. Poole & F. Gill (eds.), The Birds of North America (No. 510).
Philadelphia: The Academy of Natural Sciences; Washington, DC: The
American Ornithologists’ Union.
Hagelin, J. C. (2002). The kinds of traits involved in male-male competition: A com-
parison of plumage, behavior, and body size in quail. Behavioral Ecology, 13,
32–41.
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 321

Hairston, N. G., K. C. Nishikawa, & S. L. Stenhouse. (1987). The evolution of com-

peting species of terrestrial salamanders: Niche partitioning or interference?
Evolutionary Ecology, 1, 247–262.
Hau, M. (2007). Regulation of male traits by testosterone: Implications for the evolu-
tion of vertebrate life histories. BioEssays, 29, 133–144.
Heller, H. C. (1971). Altitudinal zonation of chipmunks (Eutamias): –nterspecific
aggression. Ecology, 52, 312–319.
Herrel, A., D. V. Andrade, J. E. de Carvalho, A. Brito, A. Abe, & C. Navas. (2009).
Aggressive behavior and performance in the Tegu lizard Tupinambis merianae.
Physiological and Biochemical Zoology, 82, 680–685.
Hersteinsson, P. & D. W. MacDonald. (1992). Interspecific competition and the
geographical distribution of red and artic foxes Vulpes vulpes and Alopex lago-
pus. Oikos, 64, 505–515.
Hines, M. (2008). Early androgen influences on human neural and behavioral devel-
opment. Early Human Development, 84, 805–807.
Holway, D. A. (1999). Competitive mechanisms underlying the displacement of
native ants by the invasive argentine ant. Ecology, 80, 238–251.
Hudson, P. J. (1986). The effect of a parasitic nematode on the breeding production
of red grouse. Journal of Animal Ecology, 55, 85–92.
Huntingford, F. A. (1976). The relationship between anti-predator behaviour and
aggression among conspecifics in the three-spined stickleback, Gasterosteus acu-
leatus. Animal Behaviour, 24, 245–260.
Hutto, R. L. (1995). Composition of bird communities following stand-replacement
fires in northern Rocky Mountain (U.S.A.) conifer forests. Conservation
Biology, 9, 1041–1058.
Jaeger, R. (1981). Dear enemy recognition and the costs of aggression between
salamanders. American Naturalist, 117, 962–974.
Jankowski, J. E., S. K. Robinson, & D. J. Levey. (2010). Squeezed at the top: nterspe-
cific aggression may constrain elevational ranges in tropical birds. Ecology, 91,
1877–1884.
Johnson, J. & A. Sih. (2005). Precopulatory sexual cannibalism in fishing spiders
(Dolomedes triton): A role for behavioral syndromes. Behavioral Ecology and
Sociobiology, 58, 390–396.
Johnstone, R. A. & K. Norris. (1993). Badges of status and the cost of aggression.
Behavioral Ecology and Sociobiology, 32, 127–134.
Just, W. & M. R. Morris. (2003). The Napoleon complex: Why smaller males pick
fights. Evolutionary Ecology, 17, 509–522.
Ketterson, E. D. & V. Nolan Jr. (1994). Male parental behavior in birds. Annual
Review of Ecology and Systematics, 25, 601–628.
Korpela, K., J. Sundell, & H. Ylönen. (2011). Does personality in small rodents vary
depending on population density? Oecologia, 165, 67–77.
Kotliar, N. B., P. L. Kennedy, & K. Ferree. (2007). Avifaunal responses to fire in
southwestern montane forests along a burn severity gradient. Ecological Appli-
cations, 17, 491–507.
322 ANIMAL BEHAVIOR

Krebs, J. R. & N. B. Davies (eds.). (1991). Behavioral Ecology: An Evolutionary

Approach. Oxford, UK: Blackwell Scientific Publications.
Lambin, X. & C. J. Krebs. (1991). Can changes in female relatedness influence
microtine population dynamics? Oikos, 61, 126–132.
Lemel, J. & K. Wallin. (1993). Status signalling, motivational condition and domi-
nance: An experimental study in the great tit, Parus major L. Animal Behaviour,
45, 549–558.
Lynch, M. & B. Walsh. (1998). Genetics and analysis of quantitative traits. Sunder-
land, MA: Sinauer Associates.
Lynn, S. E. (2008). Behavioral insensitivity to testosterone: Why and how does testos-
terone alter paternal and aggressive behavior in some avian species but not
others? General and Comparative Endocrinology, 157, 233–240.
Maynard Smith, J., R. Burian, S. Kauffman, P. Alberch, J. Campbell, B. Goodwin, R.
Lande, D. Raup, & L. Wolpert. (1985). Developmental constraints and evolu-
tion: A perspective from the Mountain Lake Conference on development and
evolution. Quarterly Review of Biology, 60, 265–287.
Maynard Smith, J. & G. R. Price. (1973). The logic of animal conflict. Nature, 246,
15–18.
McElreath, R. & P. Strimling. (2006). How noisy information and individual asym-
metries can make “personality” an adaptation: A simple model. Animal Behav-
iour, 72, 1135–1139.
McNamara, J. M., P. Stephens, S. Dall, & A. I. Houston. (2009). Evolution of trust
and trustworthiness: Social awareness favours personality differences. Proceed-
ings of the Royal Society of London, B, 276, 605–613.
Mech, L. D. (1999). Alpha status, dominance, and division of labor in wolf packs.
Canadian Journal of Zoolgy, 77, 1196–1203.
Meek, S. B. & R. J. Robertson. (1991). How do floater male eastern bluebirds benefit
by filling vacancies on territories where females already have young?Behavioral
Ecology, 3, 95–100.
Merilä, J. & D. A. Wiggins. (1995). Interspecific competition for nest holes causes
adult mortality in the collared flycatcher. Condor, 97, 445–450.
Morrell, L. J., J. Lindström, & G. D. Ruxton. (2005). Why are small males aggres-
sive? Proceedings of the Royal Society of London, B, 272, 1235–1241.
Morris, M. R., L. Gass, & M. J. Ryan. (1995). Assessment and individual recognition
of opponents in the swordtails Xiphophorus nigrensis and X. multilineatus.
Behavioral Ecology and Sociobiology, 37, 303–310.
Mougeot, F., S. M. Redpath, F. Leckie, & P. J. Hudson. (2003). The effect of aggres-
siveness on the population dynamics of a territorial bird. Nature, 421, 737–
739.
New, L. F., J. Matthiopoulos, S. M. Redpath, & S. T. Buckland. (2009). Fitting
models of multiple hypotheses to partial population data: Investigating the
causes of cycles in red grouse. American Naturalist, 174, 399–412.
Newton, I. (1994). The role of nest sites in limiting the numbers of hole-nesting
birds: a review. Biological Conservation, 70, 265–276.
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 323

Norton, W., K. Stumpenhorst, T. Faus-Kessler, A. Folchert, N. Rohner, M. P.

Harris, et al.. (2011). Modulation of Fgfr1a signaling in zebrafish reveals a
genetic basis for the aggression-boldness syndrome. Journal of Neuroscience,
31, 13796–13807.
Nosil, P. (2002). Food fights in house crickets, Acheta domesticus, and the effects of
body size and hunger level. Canadian Journal of Zoology, 80, 409–417.
Parker, G. A. (1974). Assessment strategy and the evolution of fighting behaviour.
Journal of Theoretical Biology, 47, 223–243.
Pearson, S. F. & S. Rohwer. (2000). Asymmetries in male aggression across an avian
hybrid zone. Behavioral Ecology, 11, 93–101.
Pedetta, S., L. Kaczer, & H. Maldonado. (2010). Individual aggressiveness in the crab
Chasmagnathus: Influence in fight outcome and modulation by serotonin and
octopamine. Physiology and Behavior, 101, 438–445.
Peiman, K. & B. W. Robinson. (2007). Heterospecific aggression and adaptive diver-
gence in brook stickleback (Culaea Inconstans). Evolution, 61, 1327–1338.
Peiman, K. & B. W. Robinson. (2010). Ecology and evolution of resource-related
heterospecific aggression. Quarterly Review of Biology, 85, 133–158.
Pitnick, S. & G. T. Miller. (2000). Correlated response in reproductive and life
history traits to selection on testis length in Drosophila hydei. Heredity, 84,
416–426.
Plissner, J. H. & P. A. Gowaty. (1995). Eastern bluebirds are attracted to two-box
nest sites. Wilson Bulletin, 107, 289–295.
Polo-Cavia, N., P. López, & J. Martín. (2011). Aggressive interactions during feeding
between native and invasive freshwater turtles. Biological Invasions, 13,
1387–1396.
Popova, N. K. (2006). From genes to aggressive behavior: The role of serotonergic
system. BioEssays, 28, 495–503.
Power, H. W. & M. P. Lombardo. (1996). Mountain bluebird. In A. Poole & F. Gill
(eds.), The Birds of North America (No. 222). Philadelphia: The Academy of
Natural Sciences; Washington, DC: The American Ornithologists’ Union.
Pruitt, J. N., S. E. Riechert, & T. C. Jones. (2008). Behavioural syndromes and their
fitness consequences in a socially polymorphic spider, Anelosimus studiosus.
Animal Behaviour, 76, 871–879.
Raihani, N. J., A. R. Ridley, L. E. Browning, M. J. Nelson-Flower, & S. Knowles.
(2008). Juvenile female aggression in cooperatively breeding pied babblers:
Causes and contexts. Ethology, 114, 452–458.
Réale, D., S. M. Reader, D. Sol, P. T. McDougall, & N. J. Dingemanse. (2007).
Integrating animal temperament within ecology and evolution. Biological
Reviews of the Cambridge Philosophical Society, 82, 291–318.
Redondo, T. & J. Carranza. (1989). Offspring reproductive value and nest defense in
the magpie (Pica pica). Behavioral Ecology and Sociobiology, 25, 369–378.
Riebli, T., B. Avgan, A.-M. Bottini, C. Duc, M Taborsky., & D. Heg. (2011).
Behavioural type affects dominance and growth in staged encounters of co-
operatively breeding cichlids. Animal Behaviour, 81, 313–323.
324 ANIMAL BEHAVIOR

Riechert, S. E. (1993). Investigation of potential gene flow limitation of behavioral adap-

tation in an aridlands spider. Behavioral Ecology and Sociobiology, 32, 355–363.
Riechert, S. E. & P. M. Johns. (2003). Do female spiders select heavier males for the
genes for behavioral aggressiveness they offer their offspring?Evolution, 57,
1367–1373.
Riska, B. (1989). Composite traits, selection response and evolution. Evolution, 43,
1172–1191.
Robertson, R. J. (1986). Spitefulness, altruism, and the cost of aggression: Evidence
against superterritoriality in tree swallows. Condor, 88, 104–105.
Roff, D. A. (1994). Optimality modeling and quantitative genetics: A comparison of
the two approaches. In C. Boake (ed.), Quantitative Genetic Studies of Behav-
ioral Evolution (pp. 49–66). Chicago: University of Chicago Press.
Roff, D. A. (1997). Evolutionary Quantitative Genetics. New York: Longman.
Rosvall, K. (2011). Cost of female intrasexual aggression in terms of offspring quality:
A cross-fostering study. Ethology, 117, 332–344.
Rusu, A. & S. Krackow. (2005). Agonistic onset marks emotional changes and disper-
sal propensity in wild house mouse males (Mus domesticus). Journal of Compa-
rative Physiology, 119, 58–66.
Saab, V. A., J. Dudley, & W. L. Thompson. (2004). Factors influencing occupancy
of nest cavities in recently burned forests. Condor, 106, 20–36.
Schürch, R. & D. Heg. (2010). Life history and behavioral type in the highly social
cichlid Neolamprologus pulcher. Behavioral Ecology, 21, 588–598.
Schwabl, H. & E. Kriner. (1991). Territorial aggression and song of male Europian
robins (Erithacus rubecula) in autumn and spring: Effects of antiandrogen
treatment. Hormones and Behavior, 25, 180–194.
Schwarzkopf, L., M. W. Blows, & M. J. Caley. (1999). Life history consequences of
divergent selection on egg size in Drosophila melanogaster. American Naturalist,
54, 333–340.
Seivwright, L. J., S. M. Redpathl, F. Mougeot, F. Leckie, & P. J. Hudson. (2005).
Interactions between intrinsic and extrinsic mechanisms in a cyclic species:
Testosterone increases parasite infection in red grouse. Proceedings of the Royal
Society of London, B, 272, 2299–2304.
Sih, A., A. Bell A., & J. C. Johnson. (2004). Behavioral syndromes: An ecological and
evolutionary overview. Trends in Ecology and Evolution, 19, 372–378.
Smith, I. P., F. A. Huntingford, R. J. A.Atkinson, & A. C. Taylor. (1994). Strategic
decisions during agonistic behaviour in the velvet swimming crab, Necora
puber (L.). Animal Behaviour, 47, 885–894.
Soma, K. K., A. D. Tramontin, & J. C. Wingfield. (2000). Oestrogen regulates male
aggression in the non-breeding season. Proceedings of the Royal Society of Lon-
don, B, 267, 1089–1096.
Stamps, J. A. & V. V. Krishnan. (1997). Functions of fights in territory establish-
ment. American Naturalist, 150, 393–405.
Stenseth, N. C., O. N. Bjornstad, & W. Falck. (1996). Is spacing behaviour coupled
with predation causing the microtine density cycle? A synthesis of current
ECOLOGICAL AND EVOLUTIONARY FEEDBACKS 325

process-oriented and pattern-oriented studies. Proceedings of the Royal Society of

London, B, 263, 1423–1435.
Stenseth, N. C. & A. Łomnicki. (1990). On the Charnov-Finerty hypothesis: The
unproblematic transition from docile to aggressive and the problematic transi-
tion from aggressive to docile. Oikos, 58, 234–238.
Suarez, A. V., D. A. Holway, & N. D. Tsutsui. (2008). Genetics and behavior of a
colonizing species: The invasive Argentine ant. American Naturalist, 172,
S72–S84.
Tannerfeldt, M., B. Elmhagen, & A. Angerbjörn. (2002). Exclusion by interference
competition? The relationship between red and arctic foxes. Oecologia, 132,
213–220.
Temeles, E. (1994). The role of neighbours in territorial systems: When are they
“dear enemies”? Animal Behaviour, 47, 339–350.
Trainor, B. C. & C. A. Marler. (2001). Testosterone, paternal behavior, and aggres-
sion in the monogamous california mouse (Peromyscus californicus). Hormones
and Behavior, 40, 32–42.
Tsutsui, N. D., A. V. Suarez, & R. K. Grosberg. (2003). Genetic diversity, asymmet-
rical aggression, and recognition in a widespread invasive species. Proceedings of
the National Academy of Sciences, USA, 100, 1078–1083.
Tuttle, E. M. (2003). Alternative reproductive strategies in the white-crowned spar-
row: Behavioral and genetic evidence. Behavioral Ecoogy. 14, 425–432.
Van Lanen, N. J., A. B. Franklin, K. P. Huyvaert, R. F. Reiser, & P. C. Carlson.
(2011). Who hits and hoots at whom? Potential for interference competition
between barred and northern spotted owls. Biological Conservation, 144,
2194–2201.
Walls, S. C. (1990). Interference competition in postmetamorphic salamanders:
Interspecific differences in aggression by coexisting species. Ecology, 71,
307–314.
Ward, A. J. W., P. Thomas, & P. J. B. Hart. (2004). Correlates of boldness in the
three-spined sticklebacks (Gasterosteus aculeatus). Behavioral Ecology and Socio-
biology, 55, 561–568.
Watt, D. J. (1986). Relationship of plumage variability, size and sex to social domi-
nance in Harris’ sparrows. Animal Behaviour, 34, 16–27.
Weiß, B., K. Kortschal, & K. Foerster. (2011). A longitudinal study of dominance
and aggression in greylag geese (Anser anser). Behavioral Ecology, 22, 616–624.
Wilson, A. J., U. Gelin, M. C. Perron, & D. Réale. (2009). Indirect genetic effects
and the evolution of aggression in a vertebrate system. Proceedings of the Royal
Society of London, B, 276, 533–541.
Wingfield, J. C., S. L. Meddle, I. Moore, S. Busch, D. Wacker, S. Lynn, et al. (2007).
Endocrine responsiveness to social challenges in northern and southern hemi-
sphere populations of Zonotrichia. Journal of Ornithology, 148, S435–S441.
Wolf, J. B., E. D. Brodie, J. M. Cheverud, A. J. Moore, & M. J. Wade. (1998).
Evolutionary consequences of indirect genetic effects. Trends in Ecology and
Evolution, 13, 64–69.
326 ANIMAL BEHAVIOR

Wolf, M., G. S. van Doorn, O. Leimar, & F. J. Weissing. (2007). Life-history trade-
offs favour the evolution of animal personalities. Nature, 447, 581–585.
Zack, S. (1975). A description and analysis of agonistic behaviour patterns in an opis-
thobranch mollusc, Hermissenda crassicornis. Behaviour, 5, 238–267.
 11

Altruism and Kinship

Lee Alan Dugatkin

INTRODUCTION
Animal behaviorists have a long-standing interest in understanding the
evolution of altruism. It all started when Charles Darwin obsessed over the
“altruism question” in his classic book On the Origin of Species. In particular,
Darwin was very worried about the behavior of honey bee workers. The prob-
lem was the self-sacrificial—indeed, suicidal—behavior that worker bees dis-
play when their nest is threatened. How could natural selection ever favor
this sort of behavior? Darwin’s biographer, Janet Browne (2002), describes
him as “specially exercised over honey bees” (p. 203), and Darwin paints him-
self as “half mad” over honey bee behavior. If Darwin (1859) was correct that
“natural selection acts only by the preservation and accumulation of small
inherited modifications, each profitable to the preserved being” (p. 95), then
what was going with these suicidal honeybees? And how could he explain other
prosocial behaviors such as guard duty and food sharing?
Darwin’s solution to the altruism problem revolved around kinship. But,
before we can understand what Darwin proposed, we need to be clear about
some terminology. Webster’s dictionary defines “kin” as “one’s relatives.”
But what animal behaviorists mean when they talk about kinship is some-
thing much more specific, namely genetic relatedness. Spouses and in-laws
are kin under the everyday usage of the word but not the more stringent,
evolution-based definition adopted by animal behaviorists. There is a similar
distinction to be made about the everyday definition of altruism and the
328 ANIMAL BEHAVIOR

definition used in the animal behavior literature. While the Webster’s

definition of altruism is “unselfish regard for or devotion to the welfare of
others,” the evolutionary definition is more operational—altruism is an act
that is costly to the donor and beneficial to the recipient, where costs and
benefits are measured in terms of effects on expected lifetime reproductive
success.
With these definitions and distinctions in hand, let us return to how
Darwin dealt with problems like suicidal honeybees. Darwin understood
that, somehow or another—and he was not quite sure just how—traits that
affected reproductive success were passed from parents to offspring. Of course,
he did not know about genes—Mendel was only just about to publish
his results, and they remained unnoticed until 1900—but Darwin did some-
times discuss blood-borne particles he called “gemmules” that were passed
from parents to offspring. What is more, he understood that all blood
relatives—what we would call genetic relatives—resembled one another to
varying degrees.
In a section of The Origin entitled “Objections to the Theory of Natural
Selection as Applied to Instincts: Neuter and Sterile Insects,” Darwin (1859)
proposed that the riddle of altruism like that seen in honeybee workers “disap-
pears when it is remembered that selection may be applied to the family, as well
as the individual and may thus gain the desired end” (p. 237). Acts that benefit
blood kin can be favored by natural selection. Even though a worker bee pays
a huge cost by defending the hive, this cost is compensated by the benefits
accrued by her genetic relatives at the hive. Kinship was the key to solving
Darwin’s problems with altruism.
In one sense, Darwin both posed and solved the conundrum of the evolu-
tion of altruism. The problem was confronted, and the remedy—what we
would now call kin selection—was proposed. But in some important ways,
Darwin failed to settle the issue. Without experiments or some sort of math-
ematical framework for his theory, he was never able to answer the questions
his theory brought forth, namely, precisely how does what we now call kin selec-
tion operate? For example, Just how does the degree of kinship affect the evolution
of altruism? Some blood kin like siblings are very closely related, but others,
like second cousins, are less genetically related. Does that matter, and if so, pre-
cisely how? Does it matter how costly the altruistic act is? Does it matter how great
a benefit the recipient of altruism receives? If so, how are these costs and benefits to
be measured, and how does ecology affect such costs and benefits? For the most
part, these questions lay unanswered for the next hundred or so years after
publication of The Origin of Species, though there were certainly some interest-
ing twists and turns during this time period (Dugatkin, 2007).
ALTRUISM AND KINSHIP 329

INCLUSIVE FITNESS THEORY AND THE EVOLUTION OF ALTRUISM

In 1963, William D. Hamilton published a paper entitled “The Evolution
of Altruistic Behaviour” in The American Naturalist (Hamilton, 1963). He
opened this paper as follows: “It is generally accepted that the behaviour char-
acteristic of a species is just as much the product of evolution as the morphol-
ogy” (p. 354). But, as Hamilton noted, there are some kinds of behavior that
could not easily be explained by classic evolutionary thinking: “in particular
. . . any case where an animal behaves in such a way as to promote the advan-
tages of other members of the species not its direct descendants at the expense
of its own” (p. 354). To address this issue, he built his own model examining
the role of genetic relatedness in facilitating altruism.
Hamilton asked the reader to imagine a pair of genes—gene G, which
codes for altruism, and gene g, which does not. G codes for an act that entails
a cost to the actor but a benefit to others, while g codes for no such action.
Under standard models, G is always at a selective disadvantage compared to
g and hence should never increase in frequency. But, Hamilton argued, if
the effects of genetic kinship were added to the standard model—creating an
inclusive fitness model—altruism could evolve.
To build this more inclusive model, Hamilton used only three variables,
labeled r, b, and c. Let us walk through each of these. The benefit that a recipi-
ent of an altruistic act obtained was denoted as b, and the cost paid by an altru-
ist was labeled c. For example, imagine an altruist who brings food back to her
siblings—here the benefit might be an extra chick surviving in the nest of the
altruistic bird, while the cost might entail an increased risk of death during her
foraging incursions.
To measure the genetic relatedness between altruists and those they aid,
Hamilton used Sewall Wright’s (1922) coefficient of relationship, labeled r.
In evolutionary terms, genetic relatedness centers on the probability that indi-
viduals share gene copies that they have inherited from a common ancestor or
a set of common ancestors—parents, grandparents, and so on. Gene copies
that are shared because of common ancestry are referred to as “identical by
descent.” For example, you and your siblings are genetic kin because you share
some of the same gene copies that you inherited from your mother and father.
Similarly, you and your first cousins are genetic relatives because you share a
set of common ancestors—one set of grandparents.
If we can trace the common ancestor of two or more individuals, then we
can calculate their relatedness (r). Consider the case of full siblings who have
as their recent common ancestors their mother and father. There are only
two ways that siblings can share a copy of gene X—via their mother or father.
If sibling 1 has X, then there is a 50 percent chance she received it from her
330 ANIMAL BEHAVIOR

mother; if sibling 2 has X , there is a 50 percent chance that her mother passed
this gene to sibling 2. So there is a 1 in 4 chance that the siblings share gene X
through their mother. In a similar vein, there is a 1 in 4 probability that their
father is the reason that the siblings share gene X. To calculate the chances that
our siblings share gene X through either mother or father, we add the probabil-
ities for each and obtain 1/4 + 1/4 ¼ 1/2, or 0.5. This r value can be calculated
for any set of genetic relatives, no matter how distant. For example, the genetic
relatedness between cousins is 1/8 (that is, r ¼ 0.125), between grandparent
and grandchild is 1/4 (that is, r ¼ 0.25), and between aunts or uncles and their
blood nieces and nephews is also 1/4 (that is, r ¼ 0.25).
In Hamilton’s model, natural selection favors the gene for altruism when-
ever r × b > c. This equation has become known as Hamilton’s rule, and it lies
at the heart of inclusive fitness theory (also known as kin selection). If a gene
for altruism is to evolve, then the cost (c) of altruism must be balanced by
compensating benefits to the altruist. In Hamilton’s inclusive model, the cost
is balanced by the benefits (b) accrued by genetic relatives of the altruist
because relatives may carry the gene for altruism as well. But relatives have only
some probability—measured by r—of carrying the gene in question, and so the
benefits received by the altruist must be devalued by that probability.
When we know the costs and benefits associated with an altruistic act, and
we know the relatedness of the individuals involved, Hamilton’s rule allows us
to predict whether altruism will evolve for any value of r, b, and c. Because this
rule is more easily satisfied when r is large, the more related individuals are, the
more likely altruism is to evolve. In a similar vein, the greater the benefit/cost
ratio for the altruist, the more likely altruism is to evolve (Hamilton, 1964,
1972; Grafen, 2007).
Hamilton’s rule was first constructed for cases in which an altruist helps a
single relative, but it can easily be modified to examine a single altruistic act that
has effects on many blood relatives simultaneously. In that case, altruism
evolves when In this equation, sums benefits across all genetic rel-
atives that are helped by the act of an altruist. If an altruistic act affects two sib-
lings (n ¼ 2), it will be favored as long as the benefit received by the recipient is
simply greater than the cost to the altruist—exactly half the benefit that is nec-
essary for altruism to evolve when an altruistic act benefits a single sibling.
In 1977, Paul Sherman reported the results of his University of Michigan
dissertation work on alarm calls and blood kinship in Belding’s ground
squirrels (Spermophilus beldingi) (Sherman, 1977). Though others had tested
Hamilton’s inclusive fitness theory before this paper, Sherman’s publication
was a watershed moment. He had spent more than 3,000 hours observing
ground squirrel behavior in the Sierra Nevada mountains. Sherman, like many
ALTRUISM AND KINSHIP 331

researchers before him, had noted that when a ground squirrel spots a preda-
tor, it sometimes stands ups and emits a piercing call. In response, other squir-
rels in the vicinity head for cover. But not all squirrels were equally likely to
emit such alarm calls. Why? Why were some individuals more likely than others
to emit these calls—calls that drew attention to the caller but benefited others in the
caller’s group? Sherman’s answer would require him to integrate demography,
natural history, and inclusive fitness theory (Sherman, 1981). Let us examine
how.
When male Belding’s ground squirrels reach maturity, they emigrate to
new populations to find mates. Female squirrels, however, spend their whole
lives in their natal (birth) population. What this means is that there are
important differences in how adult males and females are related to others in
their populations. By remaining in their natal populations, females—both
young and old—are surrounded by genetic relatives. Mature males, who emi-
grate to new populations, however, are often living with individuals who are
not their genetic kin.
Over the course of his three-year study, Sherman and his colleagues saw ter-
restrial predators—weasels, badgers (Taxidea taxus), dogs (Canis lupus famili-
aris), coyotes (Canus latrans), and pine martins (Martes americana)—and
ground squirrels together on 102 occasions, and predators killed a total of six
adult Belding’s ground squirrels and three juveniles. Sherman found that
when a terrestrial predator was spotted, female squirrels gave alarm calls much
more often than expected by chance, but the converse was true for males. On
average, when a predator was spotted by ground squirrels, 30 percent of the
individuals present were adult females and 20 percent were adult males. How-
ever, when he analyzed which squirrel emitted the first altruistic alarm call,
Sherman found that females did so about 65 percent of the time (more than
twice the 30 percent that would be expected by chance alone), but males did
so 2 to 3 percent of the time (about a fifth to a tenth of what would be
expected by chance). Adult females, surrounded by genetic kin, were emitting
the altruistic alarm calls; while adult males, who tend to be unrelated to those
around them, kept quiet. Sherman, who had learned about Hamilton’s rule in
graduate school, realized that alarm callers were aiding their genetic relatives,
as inclusive fitness theory predicts (Figure 11.1).
Further support for the hypothesis that kinship helps explain the distribu-
tion of alarm calls came when Sherman found that in rare instances when a
female did leave her natal group and moved into a population of unrelated
individuals, she gave alarm calls less frequently than did native females—when
a female was not surrounded by her genetic relatives she dramatically reduced
the frequency of her alarm calling.
332 ANIMAL BEHAVIOR

Figure 11.1. Alarm calls in Belding’s ground squirrels (Spermophilus beldingi) (from
Sherman, 1977, as modified in Dugatkin, 2013). Observed versus expected frequen-
cies of alarm calls in Belding’s ground squirrels. Females call much more often than
expected by chance, while the opposite is true for males.

Work on the relation between altruism and kinship is not limited to verte-
brates. Ever since Darwin first proposed that genetic relatedness may play a
role in understanding the evolution of altruism, evolutionary and behavioral
biologists have been interested in how kinship may have facilitated altruism
in the social insects: the ants, bees, and wasps (which collectively make up
the order Hymenoptera). In particular, social insect species display eusocial-
ity—a type of sociality that involves a reproductive division of labor, in which
only a small fraction of individuals in a colony breed and others in the colony
assist in cooperative rearing of offspring (Alexander et al., 1991).
Although eusociality has been uncovered in other species, it is most often
associated with ants, bees, and wasps, where it has evolved independently
at least eight times (Hughes et al., 2008). Does inclusive fitness theory help us
explain why eusociality exists in social insects? And if so, how? To answer that
question, we need to recognize that nests of eusocial species often contain
thousands of individuals, many of whom are genetic relatives. What this
means is that altruistic acts (for example, food sharing, cooperative nest
defense) benefit not just one but many many genetic relatives. But this fact
ALTRUISM AND KINSHIP 333

only gets us so far—such altruistic benefits are dispensed in any species in

which relatives live near one another, not just in eusocial hymenoptera. Why
has eusociality evolved so many times in the hymenoptera, but not elsewhere?
The answer to that question, in part, is tied to the unusual genetic system
found in hymenopterans. Social insects are haplodiploid—that is, males are
haploid, but females are diploid. In haplodiploid species, if a queen mates with
just one male, sister workers in the colony are related to one another on aver-
age by a coefficient of relatedness (r) of 0.75. The upshot of this is that hyme-
nopteran females are more related to their sisters than to their own offspring.
So if kinship is important in driving altruism and other eusocial behaviors,
we should see this primarily in female workers, not male drones. And that is
what behavioral ecologists have found over and over: females raise up broods,
defend the nest, and bring food back to the nest to share with others.
The importance of genetic relatedness in driving the evolution of eusociality
in hymenoptera has also been examined using phylogenetic techniques. When
queens mate with a single male (i.e., when queens are monandrous), genetic
relatedness in colonies should be high. But when queens mate with many males
(polyandry), the genetic relatedness in groups plummets. This difference gener-
ates a testable hypothesis: eusociality and monandry should be casually linked.
To test this hypothesis, William Hughes and his colleagues began with pre-
viously published information that eusociality has evolved nine times in hyme-
nopterans (Hughes et al., 2008; Ratnieks & Helantera, 2009). Hughes and
colleagues hypothesized that for eusociality to have been favored when it
first appeared, the ancestral mating system in these cases should have been
queens mating with a single male (monandry). Their analysis supports this
prediction—in all the lineages they analyzed (lineages that included 267 dif-
ferent species), they found that, as predicted by inclusive fitness theory,
monandry was the ancestral state.
It is important to understand that the mathematical models Hamilton and
others have derived predict that natural selection will favor behavioral traits that
increase inclusive fitness and that this will often be facilitated by remaining in
the vicinity of kin. But that is not always the case. More specifically, individuals
who have a higher inclusive fitness when remaining with their family should stay
as part of the family unit, while those who have better opportunities for increas-
ing their inclusive fitness elsewhere are predicted to depart and avail themselves of
that opportunity. For example, consider Stephen Pruett-Jones’s work with the
superb fairy wren (Malurus cyaneus), an Australian bird.
In superb fairy wrens, a breeding pair is often helped by its young (but sex-
ually mature) male offspring, who provide their siblings with additional food
and protection (Pruett-Jones & Lewis, 1990). If there are no territories
334 ANIMAL BEHAVIOR

available for these young helpers to start their own families, and if female breed-
ing partners are scarce, then remaining at their parent’s nest and helping raise
their siblings is the only thing young males can do to increase their own inclusive
fitness. But what happens when new territories open up in the vicinity of these sexu-
ally mature male helpers? To find out, Pruett-Jones removed 29 superb fairy wren
males from their territories, creating new breeding opportunities for male helpers
in the vicinity of the removals. All but one of the 32 male helpers who could have
dispersed to the newly opened territories did so, and they did so quickly—new
territories were usually occupied by former male helpers within six hours. When
a better way to increase their own inclusive fitness appeared, male helpers seized
the opportunity for a breeding territory, even though this caused the disbanding
of the family units these helpers were originally part of.

KIN RECOGNITION
Because interactions with genetic relatives will often favor the evolution of
altruism, behavioral ecologists have studied whether, and by what means,
individuals can recognize their kin (Holmes, 2004). Consider kin recognition
seen in penguins. In some species of penguins, parents travel long distances to
the sea to obtain food to bring back to inland areas where their chicks have
hatched. When they return from their journey, parents must find their young
among thousands of screaming, hungry chicks in a colony (Aubin & Jouven-
tin, 2002; Brumm & Slabbekoorn, 2005). How do parents know which chicks
are their offspring? For species like the king penguin (Aptenodytes patagonicus)
and the emperor penguin (Aptenodytes fosteri) the answer appears to center
on complex vocal cues that allow for kin recognition via vocal signatures
emitted by the young (Aubin & Jouventin, 1998; Jouventin et al., 1999;
Aubin et al., 2000; Lengagne et al., 2000).
Some penguin species are as not proficient as the king and emperor penguins
at recognizing the vocal signatures of their offspring. Individuals who build nests,
for example, are not as adept at recognizing the vocal calls of their young as are
individuals who live in dense colonies and do not nest (Jouventin & Aubin,
2002; Searby et al., 2004). Why? Researchers hypothesize that this is due to the
fact that parents in nest-building species can find their offspring by remembering
the location of their nests. Any chick in their nest is likely their offspring, and
hence natural selection to recognize offspring by vocal cues in these species is
weak. Where the problem of kin recognition is more difficult—in dense colonies
with no nests—natural selection favors the evolution of more complex vocal
recognition systems.
A number of kin-recognition models center on an internal template against
which genetic relatedness is gauged (Reeve, 1989). The basic idea is that
ALTRUISM AND KINSHIP 335

individual A determines if individual B is kin or nonkin based on how closely

individual B matches the internal template of individual A. The internal tem-
plate may be generated genetically, via learning, or via social learning, but in
all cases the animal infers the degree of kinship as some function of the extent
to which others match its own template.
A fascinating case of template matching has been studied in the behavior of
spadefoot toad tadpoles (Scaphiopus bombifrons) (Elgar & Crespi, 1992; Pfen-
nig et al., 1993). Two feeding morphs of spadefoot toads are found in ponds:
in one morph, juveniles feed on small, often drifting vegetative clumps of food
(known as detritus) and typically develop into herbivores when they mature.
In the second morph, juveniles feed on shrimp and tend to mature into car-
nivorous cannibals as they mature.
David Pfennig (1993) studied kin recognition in both herbivore and canni-
bal spadefoot morphs by testing morphs in the presence of either unfamiliar
siblings or unfamiliar nonrelatives. Herbivores preferred associating with their
siblings over unrelated individuals, presumably because of the inclusive fitness
benefits associated with interactions with genetic kin. On the other hand, the
carnivorous cannibal morphs spent more time near unrelated individuals, pre-
sumably to avoid the costs of killing their genetic kin.
Pfennig and his colleagues also offered carnivores a choice between unfa-
miliar siblings and unfamiliar nonrelatives in a protocol that allowed carni-
vores to actually eat other tadpoles. Carnivores were not only more likely to
eat unrelated individuals, but they were able to distinguish between relatives
and nonrelatives by taste. Pfennig found that carnivores were equally likely
to suck relatives and nonrelatives into their mouths, but they released their rel-
atives much more frequently than unrelated individuals.

CLOSING THOUGHT
The study of animal behavior was revolutionized by the introduction of
inclusive fitness models of the evolution of altruism. This continues to be
one of the most actively researched areas in ethology, with modern work
employing molecular genetic and phylogenetic analyses to expand the frontiers
of research in this area.

REFERENCES AND SUGGESTED READING

Alexander, R. D., K. M.Noonan, & B. J. Crespi. (1991). The evolution of eusociality.
In P. Sherman, J. U. M. Jarvis, & R. D. Alexander (eds.), The Biology of the
Naked Mole-Rat (pp. 3–44). Princeton, NJ: Princeton University Press.
336 ANIMAL BEHAVIOR

Aubin, T. & P. Jouventin. (1998). Cocktail-party effect in king penguin colonies.

Proceedings of the Royal Society of London, B, 265, 1665–1673.
Aubin, T., & P. Jouventin. (2002). How to vocally identify kin in a crowd: The pen-
guin model. Advances in the Study of Behavior, 31, 243–277.
Aubin, T., P. Jouventin, & C. Hildebrand. (2000). Penguins use the two-voice
system to recognize each other. Proceedings of the Royal Society of London, B,
267, 1081–1087.
Beecher, M. D., I. Beecher, & S. Hahn. (1981). Parent-offspring recognition in bank
swallows: II. Development and acoustic basis. Animal Behaviour, 29, 95–101.
Beecher, M. D., I. Beecher, & S. Lumpkin. (1981). Parent-offspring recognition in
bank swallows: I. Natural history. Animal Behaviour, 29, 86–94.
Beecher, M. D., B. Medvin, P. Stoddard, & P. Loesch. (1986). Acoustic adaptations
for parent-offspring recognition in swallows. Experimental Biology, 45,
179–193.
Blaustein, A. (1983). Kin recognition mechanisms: phenotype matching or recogni-
tion alleles. American Naturalist, 121, 749–754.
Browne, J. (2002). Charles Darwin: The Power of Place. New York: Knopf.
Brumm, H., & H. Slabbekoorn. (2005). Acoustic communication in noise. Advances
in the Study of Behavior, 35, 151–209.
Darwin, C. (1859). On the Origin of Species. London: J. Murray.
Dawkins, R. & J. Krebs. (1979). Arms races between and within species. Proceedings
of the Royal Society of London, B, 205, 489–511.
Dugatkin, L. A. (2007). Inclusive fitness theory from Darwin to Hamilton. Genetics,
176, 1375–1380.
Dugatkin, L. A. (2013). Principle of Animal Behavior. Third edition. W. W. Norton:
New York.
Elgar, M. & B. Crespi (eds.). (1992). Cannibalism: Ecology and Evolution among
Diverse Taxa. Oxford, UK: Oxford University Press.
Frommen, J. G., C. Luz, & T. C. M. Bakker. (2007). Kin discrimination in stickle-
backs is mediated by social learning rather than innate recognition. Ethology,
113, 276–282.
Grafen, A. (2007). Detecting kin selection at work using inclusive fitness. Proceedings
of the Royal Society of London, B, 274, 713–719.
Hamilton, W. D. (1963). The evolution of altruistic behavior. American Naturalist,
97, 354–356.
Hamilton, W. D. (1964). The genetical evolution of social behaviour. I and II.
Journal of Theoretical Biology, 7, 1–52.
Hamilton, W. D. (1972). Altruism and related phenomena, mainly in social insects.
Annual Review of Ecology and Systematics, 3, 192–232.
Holmes, W. & P. W. Sherman. (1983). Kin recognition in animals. American
Scientist, 71, 46–55.
Holmes, W. G. (2004). The early history of Hamiltonian-based research on kin
recognition. Annales Zoologici Fennici, 41, 691–711.
ALTRUISM AND KINSHIP 337

Hoogland, J. & P. W. Sherman. (1976). Advantages and disadvantages of bank swal-

low coloniality. Ecological Monographs, 46, 33–58.
Hughes, W. O. H., B. P. Oldroyd, M. Beekman, & F. L. W. Ratnieks. (2008).
Ancestral monogamy shows kin selection is key to the evolution of eusociality.
Science, 320, 1213–1216.
Jouventin, P. & T. Aubin. (2002). Acoustic systems are adapted to breeding ecologies:
Individual recognition in nesting penguins. Animal Behaviour, 64, 747–757.
Jouventin, P., T. Aubin, & T. Lengagne. (1999). Finding a parent in a king penguin
colony: The acoustic system of individual recognition. Animal Behaviour, 57,
1175–1183.
Lengagne, T., T. Aubin, P. Jouventin, & J. Lauga. (2000). Perceptual salience of indi-
vidually distinctive features in the calls of adult king penguins. Journal of the
Acoustical Society of America, 107, 508–516.
Manning, C. J., D. A. Dewsbury, E. K. Wakeland, & W. K. Potts. (1995). Commu-
nal nesting and communal nursing in house mice, Mus musculus domesticus.
Animal Behaviour, 50, 741–751.
Manning, C. J., E. K. Wakeland, & W. K. Potts. (1992). Communal nesting patterns
in mice implicate MHC genes in kin recognition. Nature, 360, 581–583.
Ortega, C. (1998). Cowbirds and Other Brood Parasites. Tucson, AZ: University of
Arizona Press.
Packer, C., S. Lewis, & A. Pusey. (1992). A comparative analysis of non-offspring
nursing. Animal Behaviour, 43, 265–281.
Pfennig, D. W., H. K. Reeve, & P. W. Sherman. (1993). Kin recognition and canni-
balism in spadefoot toads. Animal Behaviour, 46, 87–94.
Pruett-Jones, S. G. & M. J. Lewis. (1990). Sex ratio and habitat limitation promote
delayed dispersal in superb fairy-wrens. Nature 348, 541–542.
Ratnieks, F. L. W. & H. Helantera. (2009). The evolution of extreme altruism and
inequality in insect societies. Philosophical Transactions of the Royal Society of
London, B, 364, 3169–3179.
Reeve, H. K. (1989). The evolution of conspecific acceptance thresholds. American
Naturalist, 133, 407–435.
Searby, A., P. Jouventin, & T. Aubin. (2004). Acoustic recognition in macaroni pen-
guins: An original signature system. Animal Behaviour, 67, 615–625.
Sherman, P. W. (1977). Nepotism and the evolution of alarm calls. Science, 197,
1246–1253.
Sherman, P. W. (1981). Kinship, demography, and Belding’s ground squirrel
nepotism. Behavioral Ecology and Sociobiology, 8, 251–259.
Sherman, P. W. & W. Holmes, (1985). Kin recognition: Issues and evidence.
Fortschritte der Zoologie, 31, 437–460.
Waldman, B. (1987). Mechanisms of kin recognition. Journal of Theoretical Biology,
128, 159–185.
Wright, S. (1922). Coefficients of inbreeding and relationship. American Naturalist,
56, 330–338.
This page intentionally left blank
 12

Evolutionary History of Behavior

Terry J. Ord

BEHAVIORAL ECOLOGY MEETS PALEONTOLOGY AND PHYLOGENETICS

Uncovering the past history of animal behavior seems impossible; unless we
can observe behavior directly, it would appear there is no way to know what
the behavior of an extinct animal might have been like. Direct observation
seems an obvious prerequisite for the study of animal behavior, but is it really
necessary? Will we never know how the dinosaurs interacted with one another and
their environment without a time machine or a way to resurrect them? Can we say
nothing about the behavior of the evolutionary ancestors of animals that are alive
today?
Science is about deduction and inference. Direct observation of the phe-
nomena under study is an advantage, but it is by no means a necessity.
Granted, there are some things we will never know about the behavior of
extinct animals without direct observation. But much can still be learned
about the evolutionary history of animal behavior through some clever detec-
tive work and modern scientific methods.
There are two ways the history of behavior can be studied: combining fossil
evidence with information from contemporary analogues of similar species
alive today, and the reconstruction of past history using phylogenies and the
comparative method. The first approach extrapolates the behavioral ecology
of extinct species by exploiting what we know about the behavior of living spe-
cies. Behavior is often reflected in the morphology of an animal and other char-
acteristics that can leave traces in the fossil record (e.g., track marks; Mazin et al.,
340 ANIMAL BEHAVIOR

2009). These preserved characteristics can therefore be used to similarly infer the
probable behavior of prehistoric animals. The second approach is similar to the
first in the sense that it also uses information on the behavior of living species,
but it relies on phylogenetics rather than fossil evidence to infer the evolutionary
history of behavior. By comparing the behavior of species that are phylogenetic
close relatives it is possible to reconstruct the likely behavior of the common
ancestor of those species. As a basic example, if a set of closely related species
hunted in small groups, then it is reasonable to assume that the common evolu-
tionary ancestor of those species also hunted in groups. This phylogenetic com-
parative method can be taken even further to investigate the factors that might
have influenced the initial evolution of particular behaviors.
Combining fossils with information from living species and the application
of the phylogenetic comparative method has led to many important discov-
eries about the evolutionary history of behavior. These discoveries have in
turn helped behavioral scientists understand why animals in existence today
behave the way they do and how animals have adapted to their environment.
The latter is especially important for knowing how animals respond to envi-
ronmental change, which has important implications for conservation. In
the following sections, examples are presented that illustrate some of the ways
scientists have studied the history of animal behavior and what has been
learned about how animals respond to natural selection and other evolution-
ary forces.

LINKING FOSSILS TO THE BEHAVIOR OF LIVING SPECIES

Questions that often arise when contemplating the behavior of extinct
species are how social they might have been and whether their behavioral
ecology contributed to their ultimate demise. It is understandable why ani-
mal sociality is an especially evocative topic to us because we are a highly
social species ourselves. Sociality implies richness in the lives of animals that
in itself is fascinating. But the social behavior of animals is also important
in shaping the evolution of species, from how they reproduce to how they
find their food. Animal sociality is therefore a major focus of research in
behavioral ecology. Knowing the constraints on an animal’s behavior can also
help us understand why some species and not others might have gone
extinct. This is important to know so we can predict how animals today will
cope with environmental change resulting from human impacts and global
shifts in climate.
What evidence do we have for social behavior in extinct species, and how do we
know what aspects of behavior might have contributed to species extinction?
EVOLUTIONARY HISTORY OF BEHAVIOR 341

Social Sabertooths and Migratory Mastodons

Sabertooth cats (Smilodon fatalis) were about the size of the largest tigers in
existence today and roamed North America during the late Pleistocene
(1.6 million to 10,000 years ago). Fossils reveal powerful forelimbs and mas-
sive canines, indicating a formable predator. Classically, sabertooths were
believed to have been solitary, like today’s cheetahs or American mountain
lions. But the large number of fossils clumped together in the Californian tar
seeps of Rancho La Brea is at odds with this “lone hunter” lifestyle. When
fossils of the same species are found clustered together, it is often taken as evi-
dence of sociality (Ladevéze et al., 2011). For the sabertooth, however, things
were ambiguous. The consensus among paleontologists was that sabertooths
were attracted to the tar seeps by the sounds of struggling prey caught in the
tar and themselves ended up becoming trapped. Debate, however, centers on
whether sabertooths were attracted separately or in groups. There were two
competing hypotheses. The first hypothesis was that sabertooths were solitary
and attracted individually to the tar seeps. The second hypothesis was that
sabertooths were social—that is, hunted in packs—and were attracted to the
tar seeps as groups.
Testing these hypotheses presented a challenge. The number of sabertooth
cats recovered from the tar seeps was the only information paleontologists had
of what occurred in the Pleistocene. By itself, this information gives only the
final outcome, not how that outcome came about. The breakthrough came
when paleontologists teamed up with several experts on the behavior of
present-day large carnivores (Carbone et al., 2009). These behavioral ecolo-
gists had conducted many field surveys of African carnivores by “calling in”
predators through playing audio recordings of distressed prey. The researchers
argued that these playbacks, which were originally designed to estimate carni-
vore abundance, were analogous to the sounds of prey caught in the La Brea
tar seeps (Van Valkenburgh et al., 2009). That is, the types of African carni-
vores attracted by the playbacks were, in effect, a real-world simulation of
events at the La Brea tar seeps in the Pleistocence.
The behavioral ecologists knew from direct observation and other studies
which carnivores attracted to playbacks were solitary hunters (e.g., cheetahs
[Acinonyx jubatus] and leopards [Panthera pardus]) and which hunted in
groups (e.g., wild dogs [Lycaon pictus] and lions [Panthera leo]). Their data
showed clear differences in the proportion of solitary and social predators
attracted to playbacks (Figure 12.1). With this information in hand, the pale-
ontologists reevaluated the number of sabertooth cats attracted to the tar seeps
relative to other types of animals. If sabertooth cats were solitary, as tradition-
ally assumed, their numbers were vastly overrepresented in the La Brea tar
Figure 12.1. Shown in gray are the percentages of African carnivores, grouped by body
size and sociality, attracted to audio playbacks of distressed prey animals, which were
heavily skewed towards large, pack-hunting predators. In the upper panel (a) shown in
black are the percentage of fossilized remains found in the La Brea tar seeps based on
the assumption that sabertooth cats hunted in groups, while the lower panel (b) shows
the percentage of fossils based on the assumption that sabertooth cats were solitary.
Error bars are 95 percent confidence intervals. (This plot is reprinted from Carbone et
al., 2009, by permission of Highwire Press.)
EVOLUTIONARY HISTORY OF BEHAVIOR 343

seeps according to the number of living solitary predators attracted to audio

playbacks (Figure 12.1b). By contrast, if sabertooth cats were social, then their
numbers were almost exactly the number predicted by the proportion of
group-hunting predators attracted to playbacks (Figure 12.1a). While this data
cannot provide conclusive proof, it does offer compelling support for the social
hypothesis.
The study of sabertooth cats was possible because there were numerous sur-
viving analogues—for instance, cheetahs, leopards, and lions—that differed in
sociality and could be used to deduce the possible behavior of the sabertooth
based purely on their abundances in the La Brea tar seeps. But what if few
(or no) contemporary analogues exist for comparison? This was a problem faced
by paleontologists interested in the behavior of mastodons. These massive
creatures were related to present-day elephants in the sense that they both
belong to the same order of mammals (Proboscidea) and share some similar-
ities in morphology. But the similarities are generally few, and mastodons
and elephants were certainly not as genetically related as the sabetrooth was
to today’s big cats (Janczewski et al., 1992; Rohland et al., 2010). Further-
more, the elephants represented the only living analogue of mastodons, so
even if they were phylogenetic close relatives, the comparison is limited to
one. So when paleontologists wished to investigate the probable migration
patterns of mastodons, an alternative approach had to be found.
The question of mastodon migration was a topic of interest to scientists
because it might reveal why these large mammals went extinct 11,500 years
ago (Hoppe et al., 1999). Mastodons lived in North America alongside another
iconic species, the mammoth, which were close relatives to elephants (Rohland
et al., 2010). Several hypotheses had been proposed about why mastodons and
mammoths suddenly disappeared after having existed for millions of years. One
hypothesis related to the impacts of climate change on the environment and
whether the behavioral ecology of mastodons and mammoths hindered their
ability to cope with the accompanying ecological changes.
We know from many species today that migration allows animals to escape
freezing winters or scorching summers and exploit seasonally fluctuating re-
sources at different locations. A spectacular example is the annual migration
of thousands of Serengeti wildebeest (Connochaetes taurinus) that walk hun-
dreds of kilometers south at the onset of the dry season to follow rainfall and
shifting food resources (Boone et al., 2006). To discover the annual move-
ment of mastodons, paleontologists took a novel approach and examined the
fossilized remains of their teeth.
In living animals, the isotope ratio in the chemical element strontium
found in the enamel of teeth reflects the isotope signature of ingested food.
344 ANIMAL BEHAVIOR

In the case of plant foods, strontium isotope ratios depend on soil type and vary
geographically. Put simply, by measuring the strontium isotopes of mastodon
teeth, it was possible to retrace where the animals had been foraging during their
lifetime. If mastodons uncovered in southern Florida had migrated to escape the
winter from as far north as the Appalachian Mountains in Georgia, then the iso-
tope signatures of their teeth should reflect the differences in isotopes present in
the vegetation from the two regions. This is precisely what the researchers found
(Hoppe et al., 1999). Isotope signatures revealed frequent mastodon migration
over hundreds of kilometers between the Appalachians and southern Florida.
In contrast, the isotope signature of mammoth teeth also uncovered in Florida
indicated that these animals did not range nearly as far and were not, as previ-
ously hypothesized, migratory (Hoppe et al., 1999).
While the migration of mastodon implies that these animals should have
been able to buffer themselves against the ecological changes resulting from
climate shifts better than mammoths, recent studies on the migration of living
species suggested this might not have been the case. Even subtle regional
changes in temperature have lead to the mismatch of migration events with
seasonal fluctuations, placing the survival of some species in jeopardy (e.g.,
Saino et al., 2011). It is still unclear the extent to which migration in mast-
odons and its absence in mammoths contributed to their extinction, but
new data on other species—both extinct and living—should help resolve this
question. It appears, though, that climate change and whether or not species
migrated was not the primary contributing factor for the extinction of
the North American megafauna at the end of the Pleistocene (Ripple & Van
Valkenburgh, 2010).

Fatherly, Musical Dinosaurs

The repeated discovery of the remains of adult dinosaurs alongside fossil-
ized egg clutches gives a strong indication that parental care was an important
component of the behavior of some dinosaurs (Norell et al., 1995). Whether
animals provide parental care, and especially who provides parental care,
reflects the mating system of species. In birds, males often provide all or part
of the parental care (e.g., incubating eggs, feeding hatchings), whereas females
are generally the sole caregivers in mammals and in the few reptiles like the
crocodile that exhibit parental care. Were dinosaurs like crocodiles, in which
females provided all care, or more like birds, in which males provided care? If dino-
saurs were like birds, was vocal communication an integral component of their
social behavior as it is in birds? What did those calls sound like?
In living species, the size of egg clutches reflects the type of parental
care exhibited by species. When males provide all care, larger clutches are
EVOLUTIONARY HISTORY OF BEHAVIOR 345

maintained than if only females provide care (Varricchio et al., 2008). This is
presumably because males do not incur the considerable cost associated with
egg production and can invest more in incubating eggs at the expense of feed-
ing. Paleontologists used the relationship of parental care type and clutch size
to sex the caregivers in the egg-nesting Cretaceous troodontid and oviraptorid
dinosaurs. Based on the size of egg clutches, these dinosaurs were most consis-
tent with a system of exclusive male parental care and not biparental or mater-
nal care (Varricchio et al., 2008).
Further evidence for male care came from a closer examination of the bones
associated with the fossilized egg nests. Female birds and crocodiles leach large
amounts of calcium and phosphorus from their tissues during egg production,
and this leaves telltale signs in the histology of their bones. Cross sections of
fossilized bones confirmed the adult dinosaurs associated with nests were male;
there was no evidence of calcium and phosphorus leaching (Varricchio et al.,
2008). Taken together, not only were some dinosaurs building nests and incu-
bating eggs (Norell et al., 1995), but paleontologists have been able to deter-
mine that parental care was most likely provided by males and not females,
pushing back the origin of paternal care to before the evolution of birds.
Comparisons of behavior among living species have also been able to recon-
struct whether dinosaurs communicated vocally and what those calls might
have sounded like. The length of the cochlear—or inner ear—correlates
closely with the hearing sensitivity of species and can be used to predict the
frequency range, the mean frequency, and even the complexity of vocal calls
produced by animals (Walsh et al., 2009). The cochlear has been preserved
in several fossils, for example the Archaeopteryx, a bird-like precursor from
the Late Jurassic. Using the equation derived from the statistical relationship
between the length of the cochlear and the vocal characteristics of living ani-
mals, researchers have argued that Archaeopteryx had a vocal repertoire very
similar to the present-day emu (Dromaius novaehollandiae) (Walsh et al.,
2009). This also implies a reasonable complex social life for Archaeopteryx
because the complexity of communication tends to reflect the level of social
complexity in species (Freeberg et al., 2012).

Prehistoric Polygyny
It is common in nature for males to compete aggressively among themselves
for access to females. And in aggressive competition, size matters. Large males
win more fights, mate with more females, and subsequently produce more off-
spring. The selective advantage of large males in contests typically leads to the
evolution of increasing sexual size dimorphism in species. That is, males
become increasingly larger than females over evolutionary time. If body size
346 ANIMAL BEHAVIOR

is strongly skewed towards males in fossils, then this is a strong indication that
males probably competed aggressive with one another in a polygamous mating
system (monogamous species are typically sexually monomorphic in size). For
example, the sexual size dimorphism of fossils has been used to infer that
strong male-male competition and polygyny existed in the extinct relatives of
present-day marsupials from the early Palaeocene (Ladevéze et al., 2011). This
implies that the solitary nature of many marsupials today has resulted from a
loss of sociality rather than being ancestral as initially assumed.
Which male a female chooses to mate with contributes not only to the evo-
lution of male size—females preferentially mate with large males over small
males—but also the evolution of male ornamentation. Sexual ornaments in
males are common in nature and provide valuable cues to females on the qual-
ity of a male as a potential mate. This is because only males in top condition
can incur the energetic and developmental costs associated with the possession
of a large ornament. Classic examples of female-driven male ornamentation
are the elaborate plumages of many male birds (e.g., peacock [Pavo cristatus]
trains). Others include conspicuous rostral appendages, such as fleshy horns,
and large throat fans in lizards.
The presence of ornaments in an extinct animal, especially if that animal
were confirmed as male, would imply a polygamous mating system. It would
also provide strong evidence that females were highly selective about which
males they chose to mate with. In present-day animals, ornaments are also
often associated with elaborate courtship displays. Determining whether
extinct animals had ornaments is helped if the features preserved in fossils
are similar in appearance to confirmed ornaments in species today. Unfortu-
nately, common ornaments like feather plumages or fleshy structures are rarely
left in the fossil record (Archaeopteryx provides an unusually clear example of
feather ornamentation). But there are also other, often more bizarre structures
preserved in fossils that paleontologists speculate might have also functioned as
ornaments. Yet there are also frequently several plausible explanations these
structures as well.
In some instances, it has been possible to test alternative hypothesis for the
function of elaborate morphological structures (Tomkins et al., 2010). Pteran-
odon were large flying pterosaurs with wingspans of many meters found in
North America in the Late Cretaceous, some 85 million years ago. Fossils
show these pterosaurs had large, prominent crests on their head (Figure
12.2), which were especially exaggerated in males (i.e., the crests were sexually
size dimorphic). Several hypotheses for what these crests might have been used
for included a rudder to facilitate flying, a heat-dissipating or -absorbing device
to aid thermoregulation, or an ornament used to attract mates. Stranger still
EVOLUTIONARY HISTORY OF BEHAVIOR 347

Figure 12.2. Illustrations of the head crests of the flying Pteranodon and dorsal sail of
Dimetrodon alongside predicted allometric slopes if crests or sails functioned as orna-
ments, as devices for thermoregulation, or as rudders (in Pteranodon). The estimated
allometry of head crests was ×6.94, while for dorsal sails it was ×1.73; these estimates
were consistent with the allometry of ornaments. (Sketches of Pteranodon and
Dimetrodon are reprinted from Tomkins et al., 2010, by permission of University
of Chicago Press.)

were the massive sails found on the backs of Dimetrodon and Edaphosaurus
dinosaurs (Figure 12.2; these animals provided the inspiration for many a
1950s Hollywood monster movie). These large predators lived during the late
Carboniferous and early Permian around 300–260 million years ago in North
America and Europe. The classic hypothesis was these large sails were struc-
tures used in thermoregulation. But it was also possible that they might have
been ornaments.
Behavioral ecologists who have studied male ornamentation in living spe-
cies have discovered the size of ornaments commonly exhibit positive allom-
etry (Kodric-Brown et al., 2006). Positive allometry occurs when larger
individuals have disproportionately larger structures—like ornaments—com-
pared to smaller individuals. This allometric pattern is believed to happen
whenever ornaments are costly to produce. Larger animals are better able to
bear the costs of having a large ornament than smaller animals, leading to a
disproportionate increase in ornament size with overall body size. Behavioral
ecologists decided to use this phenomenon to test whether Pteranodon head
crests and Dimetrodon and Edaphosaurus sails were ornaments (Tomkins et
al., 2010). The researchers also tested the alternative hypotheses relating to
rudders and thermoregulation, which biophysics predicted would have their
own unique relationships with body size (Figure 12.2).
348 ANIMAL BEHAVIOR

By comparing the size of head crests and sails in fossils for a range of indi-
viduals varying in body size, the rudder and thermoregulation hypotheses were
rejected by the magnitude of the allometric slopes (it was too large; Figure
12.2). But these allometric slopes were well within the range for what would
be expected for ornaments (Tomkins et al., 2010). By extension, these appar-
ent ornaments—head crests and dorsal sails—implied that male Pteranodon,
Dimetrodon, and Edaphosaurus dinosaurs competed among themselves for
mating opportunities with females and that those females probably exerted a
strong preference for males with the largest ornament.

COMPARE AND CONTRAST TODAY TO RECOVER BEHAVIORS PAST

Carl Linnaeus created a classification scheme based on a nested hierarchy of
shared characteristics among organisms, a scheme that remains at the heart of
modern taxonomy. This scheme exploited the common observation of natu-
ralists that groups of species often shared physical characteristics. Darwin
argued a hundred years after Linnaeus that such shared features reflected the
shared ancestry among species (Darwin, 1859). He expanded on this idea to
formulate the theory of evolution, encapsulated by the notion of “descent with
modification”: descendent species inherit features from evolutionary ancestors,
with those features becoming modified over evolutionary time through natural
selection. Today we understand evolution as a process that is more complex
than Darwin could have appreciated in his day. But this idea of common
ancestry and its influence on the features expressed by animals—and this
includes an animal’s behavior—also provides a powerful way to uncover the
probable behavior of evolutionary ancestors.
By comparing the similarities and differences among phylogenetic close rel-
atives, it becomes possible to map behaviors onto a phylogeny and retrace their
likely origin. In this section, we review examples of some remarkable animal
behavior and how comparative biologists have uncovered the evolutionary his-
tory of those behaviors with the aid of modern phylogenetic methods.

Ancient Squirrels Exploited Smelly Predators

For a number of North American ground squirrel species, rattlesnakes pose
an acute threat to the survival of offspring. At dusk, rattlesnakes use smell to
locate burrows in which squirrel pups are sheltering. Adult ground squirrels
have evolved a number of strategies to reduce the likelihood of their pups
being eaten, such as mobbing a snake before it enters a burrow. Mobbing is
meant to harass the snake so much that it is forced to give up hunting and
leave the area. It is not always successful.
EVOLUTIONARY HISTORY OF BEHAVIOR 349

Some ground squirrel species, and in particular the females and juveniles of
those species, add another defense: they smear themselves with the chewed-up
remains of shed rattlesnake skins (rattlesnakes molt frequently). This behavior
was puzzling to researchers at first. It seemed that the squirrels were willingly
covering themselves with the scent of their main predator. Experiments later
revealed that this anointing behavior helped to minimize predation. Rattle-
snakes avoided the burrows of squirrels that had recently anointed themselves
with rattlesnake sheds (Clucas et al., 2008). The scent-application behavior
was a novel antipredator strategy. The scent of the rattlesnake not only hid
the squirrel’s own smell but also gave the false impression that a squirrel’s bur-
row had already been visited by a rattlesnake.
But what was the evolutionary history of this extraordinary anointing behavior?
Did anointing evolve multiple times independently in each squirrel species or just
once early in the history of the squirrel family and was retained in those species per-
forming the behavior today?
An extensive study of different squirrel species throughout North America
and Mexico was undertaken to determine which species performed the anoint-
ing behavior and which species did not (Clucas et al., 2010). This survey in
itself led to some fascinating findings. It revealed that squirrels not only
anointed with rattlesnake scent but weasel scent as well; weasels were another
key predator of ground squirrel pups in some areas of the Americas. That is,
the behavior was not specific to rattlesnakes but general to predators using
smell to hunt for squirrel pups.
The researchers then used the phylogenetic relationships among squirrel spe-
cies to reconstruct the evolutionary history of predator scent application. The
behavior was ancient, evolving once some 28 million years ago (Figure 12.3)
and possibly even as far back as 75 million years ago (Clucas et al., 2010). Yet,
in another twist to the story, the earliest fossils of rattlesnakes and weasels were
roughly 15 million years ago. This meant that scent application evolved well
before rattlesnakes and weasels even existed. It must therefore have evolved as
an antipredator strategy to some other, now extinct predator that also relied on
smell to hunt for prey. A good candidate was the ancient North American boa
snake that existed during the Oligocene (Clucas et al., 2010).
Further analysis of the relationship between the overlap of rattlesnakes and
weasels with living ground squirrels gave other insights as well. Squirrel species
that no longer faced predation from rattlesnakes had lost the scent-application
behavior. In a couple of cases, squirrels have since come back into contact with
rattlesnakes but have not reevolved the behavior. It seems then that applying
predator scent was easily lost in the absence of predation, was a highly unusual
event in the prehistory of the squirrel family, and has not been repeated since.
Figure 12.3. Ancestor reconstructions of the predator-scent-application behavior of
North American ground squirrels. Filled lines on the phylogeny depict reconstruc-
tions using parsimony; pie charts depict reconstructions using maximum likelihood.
The phylogeny of scent application indicates that the behavior evolved at least 28
million years ago. Dots at the tips of the phylogeny indicate which species did and did
not anoint themselves with predator scent when researchers placed scent out in the
environment for squirrels to inspect. (This figure is reprinted from Clucas et al., 2010,
by permission of John Wiley & Sons.)
EVOLUTIONARY HISTORY OF BEHAVIOR 351

The ground squirrels are a wonderful example of how field studies of

behavior on a select number of species and experiments in the laboratory doc-
umenting the adaptive significance of initially puzzling behavior can be inte-
grated with evidence from phylogeny and paleontology to show the history
of animal behavior. In doing so, researchers were able to not only date the ori-
gin of behavior but also show how unusual the evolution of such an odd
behavior like predator-scent application might have been.

Reconstructing Past Mating Calls

What the vocal communication of extinct species might have sounded like,
such as the calls of the Archaeopteryx, has been inferred from the shape of the
cochlear preserved in fossils (see “Fatherly, Musical Dinosaurs” above). But
there is another way scientists can reconstruct the calls of ancestral species.
By exploiting detailed information on the call structure of living species and
the phylogenetic relationships of those species, it has been possible to recreate
the calls of evolutionary ancestors.
In a remarkable study on Central American túngara frogs (Physalaemus
spp.) (Ryan & Rand, 1995), phylogenetic methods were used to reconstruct
the evolutionary history of mating calls (Figure 12.4). Male frogs in this spe-
cies group produce a whine to attract females. Each species has a variant on
this call, and it has been assumed from this species variation that the whine
was also important in species recognition. A female needs to recognize males
of her own species as potential mates; otherwise she is in danger of wasting
her efforts with a male from the wrong species.
Once researchers had reconstructed the ancestor calls, they synthesized the
calls on a computer and played them back to females of living species in mate-
choice experiments. The researchers wanted to test how female responses to
male calls had changed over evolutionary time.
Females responded to both the calls of conspecific males and those of males
from their immediate evolutionary ancestors. This showed that female prefer-
ences were not especially tuned to the specific call of males from their own spe-
cies. Female responses did drop off, however, as evolutionary ancestors became
older; the longer females had been separated from ancestors, the longer their
preferences have had time to change.
The study indicated in a novel way that changes in male calls were generally
accompanied by shifts in female preferences. The match was not perfect, and
this was interesting. It revealed that female responses were not the driving
force behind changes in the structure of male calls, as would be expected if
the differences in male whines among species today were the product of the
need for accurate species recognition by females. Rather, shifts in the structure
352 ANIMAL BEHAVIOR

Figure 12.4. Ancestor reconstructions of the mating call of male túngara frogs of the
genus Physalaemus. Shown are the sonograms of calls; those reconstructed at phylo-
genetic nodes were synthesized by researchers and played back to females of living
species. (This figure is reprinted from Ryan and Rand, 1995, by permission of the
American Association for the Advancement of Science.)

of the male whine were either generated from random mutation, or genetic
drift, or in response to some selection pressure other than female preference
(e.g., properties of the acoustic environment, male competition, or predation).
The fact that female preferences did generally track the evolution of male calls
showed that the coevolution of male calls and female preferences has still been
an important aspect of how communication has evolved in the group but just
that female preference has not been the engine of change in male calls.
Follow-up studies have since shown this coevolution between signal and
receiver extends beyond the whine component of the mating call. It includes
the elaboration of this call with a series of chucks added to the end of the call.
Not all species add these chucks, but the more complex call that it creates is
strongly preferred by females over just the whine by itself. Early reports using
phylogenetic reconstructions suggested that the preference for chucks pre-
dated the evolution of the chuck (Ryan & Rand, 1993), perhaps because of
EVOLUTIONARY HISTORY OF BEHAVIOR 353

a latent sensory bias for complex acoustic stimuli rather than a specific prefer-
ence for the chucks themselves. This was an exciting prospect because it sug-
gested that sensory biases in receivers could have had an important influence
on the evolution of the chuck component in the mating call. There has since
been a reexamination of this sensory bias hypothesis using a more detailed
phylogeny and broader species survey of female responses to calls that included
chucks (Ron, 2008). This later study found no evidence for females preferring
calls with chucks before the chuck evolved. Instead, the chuck produced by
males and the preference for it in females had coevolved in a similar manner
as the whine component of the call (Ryan & Rand, 1995).
This provides an important cautionary note: methods used to extrapolate
the evolutionary history of behavior are dependent on the data available at
the time the analysis is done. If these data are updated, for example if new
information comes to light, conclusions can and should be revised. This
should always be kept in mind when considering evidence from fossils or
phylogenetic comparative studies.

The Origin of Caterpillar Communication

While Charles Darwin is best known for his ideas on evolution outlined in
what are now classic works such as The Origin of Species (1859), he was also
interested in the origin of social communication. In his book The Expression
of the Emotions in Man and the Animals (1872) he pondered how some forms
of communication might have originally evolved from the grunts of physical
exertions or posturings that occurred during aggressive contests among ani-
mals. Early ethologists like Niko Tinbergen (1952) and Konrad Lorenz
(1966) explored this idea further and postulated that activities not initially
associated with communication become social signals through a process of
ritualization. For example, the aggressive head-bob displays of territorial liz-
ards might have initially evolved from moving the head up and down to aid
depth perception and help gauge the distance of territory intruders before
launching an attack. Behaviors like head-bobs that provide cues on an animal’s
intentions were hypothesized to evolve into social signals through a process in
which the behavior becomes simplified and exaggerated in structure and then
repeated in a stereotyped sequence. Testing whether this process of ritualiza-
tion explains the evolution of communication has proven difficult. It requires
that the initial precursors of a communicative behavior still be in existence to-
day alongside the very signals they are believed to have evolved into. Opportu-
nities to study ritualization are therefore quite limited.
One of the best examples is the territorial vibration signal of caterpillars
(Scott et al., 2010). Many caterpillars build shelters out of silk and leaves.
354 ANIMAL BEHAVIOR

Individuals defend these shelters against interlopers that have either lost or
failed to build shelters of their own. In some species, individuals advertise
ownership of a leaf shelter using complex vibration signals. The signals consist
of an elaborate sequence of leaf scraping using a specialized hardened “oar” on
the abdomen and rapid drumming of the mandibles against the leaf surface
(Scott et al., 2010). On closer examination, researchers found the abdomen-
scraping part of the signal had the same sequence of movements as the crawl-
ing cycle of the caterpillar along the leaf. This suggested this part of the signal
was derived from crawling behavior.
Researchers then looked at caterpillar species that did not perform vibration
signals (Scott et al., 2010). In these nonsignaling caterpillar species, shelter
owners are on alert for the vibrations of a potential intruder crawling along
the branch. When detected, shelter owners start whipping violently backwards
and forwards across the leaf in an attempt to knock the intruder off the leaf.
The intruder can feel the vibrations of this flaying behavior and sometimes re-
treat before reaching the shelter. In many cases, this does not happen, and
shelter owners must fight it out with the intruder. (In signaling caterpillars,
individuals almost never came to blows, and disputes were resolved through
the exchange of vibration signals.) Through careful analysis of the vibrations
of the flaying behavior and the vibrations produced by territorial signaling,
the researchers found the stylized mandible drumming of the leaf surface in
the signal had striking similarities with the flaying defensive behavior.
Through phylogenetic reconstructions and detail species comparisons, it
became apparent there had been a progressive ritualization of the crawling
and flaying behavior over evolutionary time to produce the synchronized
vibrations used by signaling species today. Signal movements were repeated
in long bouts, highly stereotyped, and far simpler and more exaggerated in
structure than the sporadic movements associated with crawling and flaying
defensive behavior (Figure 12.5). This met all the requirements of a ritualized
behavior: simplification, exaggeration, repetition, and stereotypy.
There are other possible examples of the ritualization of noncommunicative
behaviors into social signals, such as the foot-drumming displays of some
mammals, but formal tests are few and difficult to perform. The study of
caterpillar communication represents a particularly elegant and rare confirma-
tion of the ritualized origins of a complex communication behavior.

THE DETECTIVE WORK OF EVOLUTIONARY HISTORY

A common theme in evolutionary research is the integration of multiple
methods from a range of different disciplines. This integrative study is a gen-
eral attribute of animal behavior research, but it is especially conspicuous in
EVOLUTIONARY HISTORY OF BEHAVIOR 355

Figure 12.5. Shown in the upper panel (a) are video frames showing an encounter
between a shelter owner and an intruder. On the left is a species that does not use
territorial vibration signals, Tethea or, while on the right is a species that uses vibration
signals, Drepana arcuata. In the lower panel (b), traces from a laser vibrometer show
the vibrations produced by the flaying defensive behavior of the nonsignalling species
on the left and the vibrations of the territorial signal from the signaling species on the
right. The boxes show areas of the trace that are enlarged to show roman numerals
corresponding to components that researchers found to be modified into the
ritualized signal. (This figure is reprinted from Scott et al., 2010.)

evolutionary studies of behavior. This is because we are unable to observe

the behavior of prehistoric animals directly, so we must infer it by combining
various sources of evidence. This evidence comes in a variety of forms. The
examples discussed in this chapter used information on morphological charac-
teristics, the histology of bones, allometry, stable isotope signatures, the study
of living animals, and statistical analyses that reconstructed changes in behav-
ior into a phylogeny.
With careful consideration of the data that are available, it is possible to
study the prehistory of behavior indirectly. Evolution on a smaller scale can
also be studied using breeding experiments to identify the genes that regulate
behavior and how they are inherited from one generation to the next. Other
studies adopt the comparative approach among closely related species to inves-
tigate correlated evolutionary changes in behavior with the environment. For
example, in the scent-applying squirrels, the hypothesis was that predation
by rattlesnakes led to the evolution of squirrels anointing themselves with
356 ANIMAL BEHAVIOR

scent from shed rattlesnake skins. Another way the researchers examined this
hypothesis was by testing whether there was a correlation between squirrel spe-
cies that applied scent and geographic overlap with rattlesnakes. That is, squir-
rels that overlapped with rattlesnake ranges should apply scent, while squirrels
that did not overlap with rattlesnakes, and therefore did not suffer rattlesnake
predation, should not apply scent (given the opportunity when researchers
artificially placed shed skins out in the environment for squirrels to inspect).
A strong correlation was found and provided yet another piece of evidence that
anointing behavior in squirrels was an adaptation to predation (Clucas et al.,
2010).
To conclude, the imprint of evolutionary history is apparent in all types of
behavior to a lesser or greater degree. By placing animal behavior in a historical
context—whether it is through the study of fossils or comparisons among
living species—scientists can understand the origin of behavior and in turn
better interpret its present-day function (e.g., predator-scent application by
squirrels or the role of frog calls in species recognition). This chapter has only
briefly touched on the ways in which scientists have studied the evolution of
animal behavior. More detailed reviews can be found in Ord (2010) and
Ord and Martins (2010) for those readers who would like to delve deeper into
this exciting area of research.

REFERENCES
Boone, R. B., S. J. Thirgood, & J. G. C. Hopcraft. (2006). Serengeti wildebeest
migratory patterns modeled from rainfall and new vegetation growth. Ecology,
87, 1987–1994.
Carbone, C., T. Maddox, P. J. Funston, M. G. L. Mills, G. F. Grether, & B. Van
Valkenburgh. (2009). Parallels between playbacks and Pleistocene tar seeps
suggest sociality in an extinct sabretooth cat, Smilodon. Biology Letters 5,
81–85.
Clucas, B., T. J. Ord, & D. H. Owings. (2010). Fossils and phylogeny uncover the
evolutionary history of a unique antipredator behaviour. Journal of Evolution-
ary Biology, 23, 2197–2211.
Clucas, B., D. H. Owings, & M. P. Rowe. (2008). Donning your enemy’s cloak:
Ground squirrels exploit rattlesnake scent to reduce predation risk. Proceedings
of the Royal Society of London, B, 275, 847–852.
Darwin, C. (1859). The Origin of Species. London: Penguin Books.
Darwin, C. (1872). The Expression of the Emotions in Man and the Animals. London:
John Murray.
Freeberg, T. M., R. I. M. Dunbar, & T. J. Ord. (2012). Social complexity as a proxi-
mate and ultimate factor in communicative complexity. Philosophical Transac-
tions of the Royal Society of London, B, 367, 1785–1801.
EVOLUTIONARY HISTORY OF BEHAVIOR 357

Hoppe, K. A., P. L. Koch, R. W. Carlson, & S. D. Webb. (1999). Tracking mam-

moths and mastodons: Reconstruction of migratory behavior using strontium
isotope ratios. Geology, 27, 439–442.
Janczewski, D. N., N. Yuhki, D. A. Gilbert, G. T. Jefferson, & S. J. O’Brien. (1992).
Molecular phylogenetic inference from saber-toothed cat fossils of Rancho La
Brea. Proceedings of the National Academy of Sciences, USA, 89, 9769–9773.
Kodric-Brown, A., R. M. Sibly, & J. H. Brown. (2006). The allometry of ornaments
and weapons. Proceedings of the National Academy of Sciences, USA, 103, 8733–
8738.
Ladevèze, S., C. de Muizon, R. M. D. Beck, D. Germain, & R. Cespedes-Paz.
(2011). Earliest evidence of mammalian social behaviour in the basal Tertiary
of Bolivia. Nature, 474, 83–86.
Lorenz, K. Z. (1966). Evolution of ritualization of behaviour in animals and man:
Introduction. Philosophical Transactions of the Royal Society of London, B, 251,
273–284.
Mazin, J.-M., J.-P. Billon-Bruyat, & K. Padian. (2009). First record of a pterosaur
landing trackway. Proceedings of the Royal Society London, B, 276, 3881–3886.
Norell, M. A., J. M. Clark, L. M. Chiappe, & D. Dashzeveg. (1995). A nesting dino-
saur. Nature, 378, 774–776.
Ord, T. J. (2010). Phylogeny and the evolution of communication. In M. D. Breed
& J. Moore (eds.), Encyclopedia of Animal Behavior (pp. 652–660). Oxford,
UK: Academic Press.
Ord, T. J. & E. P. Martins. (2010). The evolution of behavior: Phylogeny and
the origin of present-day diversity. In D. F. Westneat & C. W. Fox (eds.), Evo-
lutionary Behavioral Ecology (pp. 108–128). New York: Oxford University
Press.
Ripple, W. J., & B. Van Valkenburgh. (2010). Linking top-down forces to the Pleis-
tocene megafaunal extinction. BioScience, 60, 516–526.
Rohland, N., D. Reich, S. Mallick, M. Meyer, R. E. Green, N. J. Georgiadis, A. Roca,
& M. Hofreiter. (2010). Genomic DNA sequences from mastodon and woolly
mammoth reveal deep speciation of forest and savanna elephants. PLoS Biology,
8, e1000564.
Ron, S. R. (2008). The evolution of female mate choice for complex calls in tungara
frogs. Animal Behaviour, 76, 1783–1794.
Ryan, M. J. & A. S. Rand. (1993). Sexual selection and signal evolution: The ghost of
biases past. Philosophical Transactions of the Royal Society of London, B, 340,
187–195.
Ryan, M. J. & A. S. Rand. (1995). Female responses to ancestral advertisement calls
in tungara frogs. Science, 269, 390–392.
Saino, N., R. Ambrosini, D. Rubolini, J. von Hardenberg, A. Provenzale,
K. Hüppop, O. Hüppop, A. Lehikoinen, E. Lehikoinen, K. Rainio, M.
Romano, & L. Sokolov. (2011). Climate warming, ecological mismatch at
arrival and population decline in migratory birds. Proceedings of the Royal Soci-
ety London, B, 278, 835–842.
358 ANIMAL BEHAVIOR

Scott, J. L., A. Y. Kawahara, J. H. Skevington, S.-H. Yen, A. Sami, M. L. Smith, &

J. E. Yack. (2010). The evolutionary origins of ritualized acoustic signals in
caterpillars. Nature Communications, 1, 4.
Tinbergen, N. (1952). “Derived” activities: Their causation, biological significance,
origin, and emancipation during evolution. Quarterly Review of Biology, 27,
1–23.
Tomkins, J. L., N. R. LeBas, M. P. Witton, D. M. Martill, & S. Humphries. (2010).
Positive allometry and the prehistory of sexual selection. American Naturalist,
176, 141–148.
Van Valkenburgh, B., T. Maddox, P. J. Funston, M. G. L. Mills, G. F. Grether, &
C. Carbone. (2009). Sociality in Rancho La Brea Smilodon: Arguments favour
“evidence” over “coincidence.” Biology Letters, 5, 563–564.
Varricchio, D. J., J. R. Moore, G. M. Erickson, M. A. Norell, F. D. Jackson, &
J. J. Borkowski. (2008). Avian paternal care had dinosaur origin. Science,
322, 1826–1828.
Walsh, S. A., P. M. Barrett, A. C. Milner, G. Manley, & L. M. Witmer. (2009). Inner
ear anatomy is a proxy for deducing auditory capability and behaviour in reptiles
and birds. Proceedings of the Royal Society of London, B, 276, 1355–1360.
 Glossary

Abiotic—Nonliving chemical and physical factors in the environment that affect

ecosystems.
Adaptation—The process of adjustment of an organism to its environment. The
process of evolutionary modification that improves an organism’s survival and repro-
ductive success. A trait that evolved by natural selection for its current functional
role.
Adaptationist—An unsubstantiated assumption that all or most traits are optimal
adaptations.
Adaptive mate choice—Explanations in which ornament expression is associated
with characteristics of prospective mates that are beneficial outside of a mate-
selection context.
Aggregation—A group or collection of individuals.
Aggregation pheromone—A secreted or excreted chemical factor that attracts large
numbers of conspecifics.
Aggression—Behavior used to subdue or repel another individual.
Aggressive mimicry—A predator mimicking some signal that is benign or attractive
to the prey.
Agonistic—Relating to the range of activities associated with aggressive encounters
between members of the same species, including threat, attack, appeasement, or
retreat.
Agonistic character displacement—A specialized case of character displacement in
which divergence occurs in traits that affect the rate, intensity, or outcome of com-
petitive interactions.
360 GLOSSARY

Alarm call—A signal produced by an animal in response to danger.

Alarm pheromone—A chemical secreted by an animal that alerts others of the same
species to the presence of danger.
Alarm substance—A chemical released by an injured individual that causes others
to withdraw.
Allee effect—The reproductive rate of a population declining with declining popula-
tion density.
Allogroom—To groom another individual.
Allospecific—A member of another species.
Alternative reproductive tactics—Different behavior patterns of one sex of a spe-
cies to enhance reproductive success.
Altruism—An act that is costly to the donor and beneficial to the recipient, where
costs and benefits are measured in terms of effects on expected lifetime reproductive
success.
Altruistic punishment—Individuals punishing noncooperators at a cost to themselves.
Ambivalent behavior—Behavior that appears to result from conflicting motivations.
Amplified fragment length polymorphism (AFLP)—A highly sensitive molecular
technique for detecting polymorphisms in DNA.
Anisogamy—Gametes differ is size or form.
Antithetical—Referring to the opposite.
Aposematism—Warning coloration.
Arbitrary—Traits that do not directly promote or are not tightly associated with
survival and fecundity.
Armaments—A trait that can serve as an aid in contests with members of the same
species (often of the same sex) but that is not required for foraging or protection
from predators.
Assessment/communication phase—A stage of an encounter between a predator
and prey when the prey judges how much of a risk this predator poses to its safety,
when the predator weighs the costs and benefits of attack and the likelihood of suc-
cess, and when there is overt exchange of signals or cues between the parties that
allow them to assess each other more rapidly.
Associated social monogamy—A male and female of the same species spend most
of their time together and coordinate their behavior.
Association hypothesis—An explanation for maternal care proposing that because
internally fertilizing males release their gametes first, they are free to desert immedi-
ately, forcing females to choose between providing care alone or abandoning the
breeding attempt.
Attack/flight phase—A stage of an encounter between a predator and prey when
the predator decides whether to attack and the prey decides whether to flee.
GLOSSARY 361

Autonomic behavior—Automatic physiological responses that are not under the

direct control of the animal.
Autotomy—The spontaneous casting off of a body part when the organism is
injured or under attack.
Background matching—A camouflage strategy in which coloration blends in with
the background habitat.
Back-up food supply hypothesis—An explanation proposing that some offspring are
produced as food for the dominant offspring.
Bateman’s principle—The view that females almost always invest more energy in
offspring production than males, making females a limiting resource over which
males will compete.
Beater effect—Prey flushed out by group activity becoming easy to capture.
Begging—Behavior by which depending young solicit care from their parents and
other caregivers.
Behavioral—Referring to movements, postures, or displays produced by an animal.
Behavioral ecology—An approach to investigations of how different ecological
circumstances affect animal behavior and the evolution of traits with functions that
fit these different environments.
Behavioral (ethological) isolating mechanism—Behavior that prevents members of
different species from reproducing with each other.
Behavioral syndrome—Correlation of rank-order differences among individuals
through time or across situations.
Biotic—Any living component that affects other organisms in an ecosystem.
Biparental care—Parental care provided by both the father and the mother.
Bonanza resource—A resource, especially food, that is present in vast amounts but
often for a short time.
Bower—A structure built by a male bowerbird to attract mates.
Breeding bout—The amount of time required for a male and female to produce
viable offspring.
Broadcast signal—A signal that is propagated throughout the environment without
a specific receiver.
Brood parasitism—Use of a host of the same species (intraspecific brood parasit-
ism) or a different species (interspecific brood parasitism) to care for the parasite’s
offspring.
Capture/escape phase—A stage of an encounter between a predator and prey when
the predator decides whether to attempt capture and the prey decides whether to
attempt to escape.
Caste—A distinct category of individuals that provide particular services to a colony.
362 GLOSSARY

Central-place forager—An animal that forages from a fixed home location such as
a nest.
Character displacement—Divergence in traits of species with overlapping distribu-
tions caused by competition between them.
Chase-away model—An explanation for ornaments that is based upon the funda-
mental conflict of interest between males and females in sexually reproducing organ-
isms. Male ornaments are proposed to stimulate females to mate in a manner that
benefits males but is not beneficial for females. The male ornament is favored
because it enhances male mating success, but because being attracted to the orna-
ment is bad for females, they evolve diminished response systems, which produces
selection on males for any new traits that enhance or expand the stimulatory effect
of the ornament.
Cheating—Use of a “dishonest” (unreliable) signal to manipulate a receiver.
Circadian rhythm—An endogenous activity rhythm with a period of about 24 hours.
Coefficient of relationship—A measure for the level of consanguinity (degree to
which they are descended from a common ancestor) between two given individuals.
Coevolution—The evolution of two or more species, each adapting to changes in
the other.
Colony—A dense cluster of breeding territories that do not contain only breeding
sites.
Communal nesting—More than one breeding female occupying a single nest or
nesting chamber within a burrow.
Communication—The behavior of one individual affecting the behavior of another
individual.
Communication signals—Traits (structures and behavior) that have evolved specifi-
cally because they change the behavior of receivers in ways that benefit the signaler.
Compartmentalization—The extent to which subgroups in a society operate as dis-
crete units.
Competitor—An individual of the same (intraspecific) or different (interspecific)
species that vies for the same resource (e.g., food or living space) in an ecosystem.
Condition-dependent model—An explanation that proposes that ornaments are
honest signals of condition because only individuals in good condition could pro-
duce a big ornament.
Confusion effect—A large number of prey escaping from predation at the same time
making it difficult for the predator to pursue a specific individual prey animal.
Conspecific—A member of the same species.
Constraints—Factors that act to limit response or performance.
Contact call—A signal that functions to keep an animal in touch with a group and
allow other group members to determine the location of the caller.
GLOSSARY 363

Contrast effects—Phenomena that occur when animals experience a contrast between

what they have experienced in the past and what they are experiencing now.
Conventional sex roles—Males competing more intensely than females for access to
mates.
Cooperative—Interaction of two or more individuals that is mutually beneficial, or
beneficial to one and not beneficial to the other.
Cooperative breeding—A social system in which individuals contribute care to off-
spring that are not their own at the expense of their own reproduction.
Cooperative polygamy (mate-sharing)—Mating system in which multiple
cobreeders of one or both sexes share mates of the opposite sex.
Cooperative polygynandry—Mating system in which multiple males share and
mate with more than one female.
Copulation solicitation display—A female courtship display given before and
during copulation in many species of birds.
Core—The average number of offspring reared to independence.
Cost/benefit analysis—Scoring the potential outcomes of the different options in
terms of positive and negative effects on fitness.
Counteradaptation—Evolutionary responses of one species (or sex) in response to
adverse adapations of another species (or the other sex).
Countershading—A camouflage strategy in which the back is darker in color than
the belly.
Courtship display—A stereotyped pattern of behavior in animals that functions to
attract and arouse a prospective mate.
Cross-fostering experiment—An experimental technique in which members of one
brood are exchanged with those of another, producing a mix of genetic and foster
offspring in each brood.
Crypsis—The ability of an organism to avoid observation or detection by other
organisms.
Cryptic female choice—The ability of females to manipulate paternity by choosing
which sperm fertilize their eggs.
Cuckold—A male whose mate engages in extra-pair copulation.
Cue—Circumstances that predict the likelihood of a subsequent event.
Culture—Sharing of behavior by a group of animals through social transmission
among peers and between generations.
Currency—The means by which costs and benefits are measured.
Decision rules—The cognitive mechanisms that animals use to make decisions.
Decreasing marginal returns—Situation in which it becomes harder and harder
for a forager to extract food in a particular circumstance or location.
364 GLOSSARY

Deduction—Reasoning from the general to the particular.

Defense—A stage of an encounter between a predator and prey in which the prey
can employ behavioral or morphological defenses with which to deter or repel
predators.
Detection/recognition phase—The stage during an encounter between predator
and prey when both parties can discover then determine whether the other is a
potential threat or a potential meal.
Detritus—The organic debris formed from the decay of organisms.
Developmental constraint—A bias on the development of behavior or a limitation
on phenotypic variability caused by the structure, character, composition, or dynam-
ics of the developmental system.
Dewlap—A longitudinal flap of skin that hangs beneath the lower jaw or neck of
many vertebrates.
Diel vertical migration (DVM)—A pattern of up-and-down movement within the
water column that some aquatic organisms undertake each day.
Diet selection—How animals decide whether to use a particular food resource.
Dilution effect—An individual’s risk of predation decreasing as group size increases
because the predator can only catch or eat a certain number of prey.
Diploid—A cell or an organism with a full set of paired chromosomes.
Dispersed social monogamy—A male and female of the same species sharing a
territory but foraging and sleeping alone.
Displacement activity—A seemingly inappropriate behavior thought to result from
two conflicting motivations in a particular situation.
Display—A stereotyped behavior that has evolved to function as a signal in
communication.
Disruptive coloration—A camouflage strategy in which bold blocks of color or
projections of the body break up the body outline against the background environ-
ment.
Diurnality—The behavior of animals that are active in the daytime.
Divergence—Acquisition of dissimilar traits by previously similar, related orga-
nisms.
DNA fingerprinting—A molecular technique that uses highly variable regions of
DNA to identify individuals.
Dominance—The effect of one allele at a locus being partially or entirely hidden by
another allele at the same locus. A social relationship in which one individual has
power or priority of access over another.
Dynamic—Changing.
Eavesdropping—Observing the contests, behavior, or signaling of others.
GLOSSARY 365

Economically defendable—The benefits of the behaviors associated with having

exclusive access to the area outweighing the costs of maintaining and defending it
from conspecifics.
Ectospermalege—An adaption in female bed bugs that acts as a mating guide for
the male to minimize the long-term damage of traumatic insemination of the
female.
Emancipated—Freed from its original function and motivation.
Encounter—Predator and prey coming near enough to each other that detection is
possible.
Environmental potential for polygamy—The degree to which the social and eco-
logical environment allows one sex to monopolize the other sex as mates.
Eusocial—An advanced level of social organization, in which a single female or caste
produces the offspring and nonreproductive individuals cooperate in caring for the
young.
Evolution—Descent with modification.
Evolutionary ecology—An approach to the study of ecology that explicitly considers
the evolutionary histories of species and the interactions between them.
Evolutionary tug-of-war—The evolutionary response to conflicting selection on a
single trait.
Experimental evolution—The study of evolution using experimental treatments in
the laboratory and measurement of responses of organisms over time.
Extrafloral nectaries—Nectar-producing structures not part of the flower.
Extra-pair copulation—Mating with partners other than the social mate.
Extra-pair paternity—Cases in which the social mate is not the father of particular
offspring.
Facultative—An organism that does something (e.g., act as a parasite) as an option.
Feedback mechanism—A mechanism that tends to accelerate (positive) or to inhibit
(negative) a process.
Female (harem) defense polygyny—Aggregation of females, often because of
clumped food or nest sites, providing dominant males the opportunity to defend
them and thereby gain multiple mating opportunities.
Filial cannibalism—A form of infanticide in which adults eat their own offspring.
Fisher’s fundamental theorem—Any alleles that improve fitness will increase in fre-
quency, reaching fixation in just a few generations.
Fitness—The ability to survive and reproduce; the contribution to the gene pool of
the next generation made by a specified genotype or phenotype.
Fitness proxies—The currencies that animal behavior is adapted to maximize.
Foraging—Finding and collecting resources.
366 GLOSSARY

Forced copulation—A tactic in which male ducks attempting to coerce the females
into mating so violently they will on occasion drown them.
Game theory—A mathematical method for calculating success based on the choices
of others.
Generalist forager—An animal that uses many kinds of food.
Genetic bottleneck—A sharp reduction in population size that drastically reduces
the genetic variation in a population or species.
Genetic covariance—The extent to which the genetic effects for two traits vary together.
Genetic drift—Random fluctuations in the frequency of a gene, usually in a small,
isolated population, as a result of chance.
Genetic mating system—The characterization of mating relationships based on
genetic analysis of maternity and paternity.
Genetic relatedness—A quantitative estimate of the proportion of genes shared
between the genomes of any two individuals.
Genic capture model—An explanation that proposes that elaborated ornaments
inevitably become associated with condition (become condition dependent).
Gestational diabetes—High blood sugar during pregnancy.
Good genes hypothesis—A proposed explanation for mate choice based on traits
that reflect an individual’s genetic quality.
Good parent model—Females benefit from a preference for males with large orna-
ments because they provide superior resources.
Handicap principle—An explanation first proposed by Zahavi in which ornamental
traits serve as handicaps to survival such that only robust, strong, and healthy indi-
viduals could bear the burden of a big ornament.
Haplodiploid—A system in which unfertilized (haploid) eggs become males and
fertilized (diploid) eggs become females.
Haploid—A cell or organism with a set of unpaired chromosomes.
Hatching asynchrony—A pattern in which eggs in a brood do not hatch together.
Heritability—The proportion of total variance in a trait that can be attributed to
additive genetic variance in that trait.
Heterospecific—A member of a different species.
Home range—An area in which an animal or group of animals regularly travels in
search of food or mates.
Honesty—When signals reliably predict something of use to the receiver.
Incest avoidance—Individuals not attempting to breed with close relatives.
Inclusive fitness—The ability of an individual to contribute its genes to the next
generation, taking into account the shared genes passed on by the individual’s close
relatives.
GLOSSARY 367

Index of resource monopolization (Q)—A variance-based estimate of the potential

for sexual selection.
Indicator models—Proposed explanations of mate choice based on traits that are
condition-dependent and have associated costs, and thereby reliably indicate an indi-
vidual’s genetic quality.
Indicator trait—An ornamental trait that is reliably correlated with important qual-
ities of a prospective mate.
Indirect genetic effects—Influences of genotypes of other individuals in the popula-
tion on a focal individual.
Indirect reciprocity—Reciprocity by individuals other than those originally helped.
Individual recognition—The ability to discriminate among individuals using their
identity cues.
Infanticide—A tactic by which new males increase their mating success by killing
the existing offspring.
Inference—Deriving logical conclusions from premises known or assumed to be
true. Reasoning from evidence.
Information—A reduction in uncertainty.
Information center—A gathering of individuals at which information about the
location of food can be exchanged.
Information content (message)—How a receiver behaves in response to a signal or
other stimulus.
Infrasonic—Very-low-frequency sounds that are inaudible to humans.
Insurance hypothesis—An explanation proposing that parents produce back-up off-
spring in case something happens to the first ones.
Intention movement—Behavior that occurs in preparation for another behavior.
Interlocus sexual conflict—The conflict between males and females mediated by
different genes and traits.
Intersexual selection—Differential reproductive success that results from individ-
uals of one sex choosing mates from members of the other sex.
Interspecific communication—Communication that occurs between or among
members of different species.
Intralocus sexual conflict—A form of conflict in which the same genes are
expressed in both males and females but are selected in opposite directions.
Intrasexual competition—Competition among members of one sex for mating
access to members of the other sex.
Intrasexual selection—Differential reproductive success that results from intrasex-
ual competition.
Intraspecific communication—Communication that occurs between or among
members of the same species.
368 GLOSSARY

Kin recognition—The ability to distinguish between close genetic kin and nonkin.
Kin selection—A mechanism of evolution in which characteristics favor the survival
of close relatives of an individual.
Kinship—The degree of genetic relatedness between individuals of a species.
Kleptoparasitism—Food stealing by members of the same species.
Lek—An assembly area where animals display and court potential mates.
Lek paradox—The situation that the only benefit for female mate choice when
males provide no resources is good genes, but good genes are theoretically
impossible.
Life history—The timing of key events in an organism's lifetime, as shaped by natu-
ral or sexual selection.
Life history traits—Characteristics that affect an organism’s investments in growth,
reproduction, and survivorship.
Lifelong monogamy—A form of social behavior in which individuals have only one
mate for life.
Linkage disequilibrium—The nonrandom association of alleles at two or more
loci, not necessarily on the same chromosome.
Lordosis—A female courtship display given before and during copulation in many
species of mammals.
Major histocompatibility complex (MHC)—A cell surface protein molecule that
helps the vertebrate immune system recognize foreign substances and that is
encoded by a large gene family.
Many-eyes effect—Increased antipredator vigilance ability resulting from large group
size.
Marginal—Offspring that exceed the average number that can be reared to
independence.
Masquerade—A camouflage strategy in which an organism is detectable but resem-
bles an inedible object.
Mate guarding—A tactic by which a male attempts to prevent the female from
mating with other males.
Maternal care—Parental care provided by the mother.
Mating system—A behavioral strategy employed by a population of animals to
obtain mates.
Maximum standardized selection differential—A method of estimating the maxi-
mum strength of sexual selection in units of phenotypic standard deviations.
Mean crowding—A quantitative measure of density dependence.
Mean spatial crowding (m)—A quantitative measure of the number of potential
encounters that is based on the distribution of individuals in space.
GLOSSARY 369

Mean temporal crowding (t)—A quantitative measure of the number of potential

encounters that is based on the distribution of individuals in time.
Mesopredator—Midlevel predator.
Meta-analysis—A statistical method that combines results of several studies that
address related research hypotheses.
Metacommunication—Communication about communication.
Microsatellite DNA—Repeating sequences of two to six base pairs of DNA; also
called simple sequence repeats (SSRs) or short tandem repeats (STRs).
Mobbing—An antipredator behavior in which prey individuals collectively attack or
harass a predator.
Monandry—A form of social behavior in which a female mates with only one male.
Monogamy—A form of social behavior in which an individual has only one mate at
any one time.
Monogyny—A form of social behavior in which a male mates with only one female.
Morisita index (I)—A variance-based estimate of the potential for sexual selection
that is based on the value of mating or reproductive success to each individual.
Morph—Each of several variant forms of a species of animal.
Morphological—Referring to physical characteristics of the animal.
Mutualism—Both individuals benefiting from interactions between them.
Natal—Relating to the place of an individual’s birth or hatching.
Natural selection—The nonrandom process by which traits become either more or
less common in a population as a function of differential survival or reproduction of
their bearers.
Neuropeptide—A polypeptide that acts as a neurotransmitter or hormone.
Nocturnality—The behavior of animals being active in the nighttime.
Nomadic—An organism that moves or changes location frequently.
Nuptial gift—A food offering males provide to females during courtship.
Nurse ants—Individuals that clean and feed the begging larvae of the brood.
Obligate—An organism that does something (e.g., act as a parasite) by necessity.
Olfactory cue—A source of information based on odors.
Ontogenetic—Relating to the origin and development of individual animals.
Operational sex ratio—The relative number of sexually competing males that are
ready to mate versus sexually competing females that are ready to mate.
Opportunity for selection (I)—An index of the amount of variation that could be
subject to selection in a population.
Opportunity for sexual selection (Is)—An index of the amount of variation in mat-
ing success that could be subject to sexual selection in a population.
370 GLOSSARY

Optimal foraging—Gathering resources to maximize the net energy intake (or

other foraging currency) per unit time.
Ornaments—A trait that enhances the appearance of an animal, sometimes to the
detriment of its survival.
Oviposition—The process of laying eggs by egg-laying animals.
Pair bond—An enduring preferential association between two sexually mature
adults that is characterized by selective affiliation, contact, and copulation of the
partners.
Paleontologist—A scientist who studies the fossilized remains of life.
Parasite—An organism that lives on or in an individual of another species (the
host) from which it obtains nutrients.
Parasocial—Females of the same generation assisting one another, in some cases
cooperating in nest construction but otherwise rearing their broods separately, and
in other cases attending the broods cooperatively, although each female may still
reproduce.
Parental care—Any behavior by a father that enhances the reproductive success
(fitness) of offspring.
Parental investment—Any behavior that increases offspring’s fitness but decreases
the parents’ ability to invest in other offspring.
Parental investment theory—An explanation proposed by Robert Trivers sug-
gesting that the sex that invests more in parental care would ultimately control
reproduction.
Parental table—A method of characterizing mating relationships that is based on a
summary of the genetic parentage of all progeny in a breeding population regardless
of whether the offspring are produced by selfing or outcrossing.
Parent-offspring conflict—The evolutionary conflict between parents and their off-
spring in optimal parental investment.
Paternal care—Care provided by a parent that increases the fitness of offspring.
Peak shift—Foragers rejecting items that appear too similar to unwanted items,
even to the point of rejecting many suitable items.
Permeability—The degree to which societies interact with one another.
Personality variation (personalities)—Consistent differences in behavior among
individuals.
Phenotype matching—Matching the look or smell of an individual to an internal
template independent of environmental or other external cues.
Pheromone—A secreted or excreted substance that elicits a social response in mem-
bers of the same species.
Phylogenetic—Relating to the evolutionary history of animals or behavior.
GLOSSARY 371

Phylogenetic comparative method—Use of information on the evolutionary rela-

tionships of organisms (phylogenies) to compare species.
Piscivorous—Fish-eating.
Plugs—A tactic by which a male blocks the female’s reproductive tract.
Polyandry—A form of social behavior in which a female has more than one mate;
females mate multiply, resulting in stronger sexual selection on females.
Polygamy—A form of social behavior in which individuals have more than one
mate.
Polygynandry—Both sexes mate multiply, resulting in variable strength of sexual
selection in both sexes.
Polygyny—A form of social behavior in which males have more than one mate;
males mate multiply, resulting in stronger sexual selection on males.
Population biology—The study of populations of animals, especially the regulation
of population size, life history traits, and extinction.
Positive allometry—Larger individuals have disproportionately larger structures
compared to smaller individuals.
Postcopulatory sexual selection—Selection that results from competition within a
female’s reproductive tract between sperm of different males. Involves both the traits
of the sperm and of the female.
Potential reproductive rate—The maximum number of independent offspring that
parents can produce per unit of time.
Predator—An organism that eats another (prey) species.
Preeclampsia—A condition in which a pregnant woman develops high blood
pressure.
Prenuptial molt—Replacement of all or some of a bird’s feathers prior to the breed-
ing season.
Private information—Information acquired from an individual’s own experience.
Promiscuity—A social behavior in which both males and females mate with multi-
ple partners.
Protean behavior—Fast, erratic turns to escape a predator.
Provisioning—Storing food or providing food to others.
Proximate—Referring to the immediate trigger or mechanism that controls the per-
formance of the behavior.
Public information—Social or nonsocial information that is accessible to others.
Pursuit deterrence—Communicating to a predator the ability to evade capture and
thereby causing the predator to forgo pursuit.
Random distribution—The position of one individual being independent of the
position of another individual (i.e., individuals do not attract or repel each other).
372 GLOSSARY

Random walk—A path that consists of a succession of random steps.

Ratchet model—A mechanism by which selection on an indicator trait favors males
that display the trait independent of their quality, producing selection on females to
shift preferences to ornamentation that restores the association with male quality.
The ensuing coevolutionary cycle. in which males produce ornaments without costs
and females select for more and elaborate and costly forms of ornaments, results in
large, costly, and elaborate traits
Receiver—A recipient of a signal.
Reciprocal altruism—An act by one animal that temporarily reduces its fitness
while increasing another animal’s fitness, with the likelihood that the other animal
will subsequently respond in kind.
Recognition—A stage of an encounter between a predator and prey in which the
other species is determined to be potential prey, irrelevant, a conspecific, or a poten-
tial threat.
Redirected behavior—The shifting of some behavior (e.g., aggression) away from
the primary target and toward another, less appropriate object.
Reliable—A signal that accurately predicts something useful to the receiver.
Reproductive competition—Individuals are reproductively suppressed by more
dominant individuals.
Reproductive effort—The proportion of its resources that an organism expends on
reproduction.
Reproductive skew—Variance in reproductive success of each sex that contributes
to variance in fitness, thus affecting the strength of selection. The degree to which
reproduction is monopolized by one or a few individuals.
Reproductive success—The production of fertile offspring, often measured as the
number of offspring produced by an individual.
Resource defense—Behavior by an animal that prevents another animal from gain-
ing access to an important resource.
Resource-defense social mating system—A form of social behavior in which mating
relationships are dependent on possession of critical resources by members of one
sex.
Resource-tracking hypothesis—An explanation suggesting that parents produce the
number of young that can be raised in a good year and sacrifice some if resources
turn out to be insufficient.
Risk averse—Reluctant to accept a high but uncertain payoff.
Risk prone—A behavioral strategy that favors high risk in return for potentially high
reward.
Ritualization—The evolution of signals through a process in which the behavior
becomes simplified and exaggerated in structure and then repeated in a stereotyped
sequence.
GLOSSARY 373

Ritualized displays—Stereotyped signals that evolved by ritualization.

Runaway model—A proposed explanation for sexual selection first proposed by
R. A. Fisher in which sexual selection of traits is based upon a “runaway” positive
feedback mechanism.
Satisficing—Meeting a minimum food requirement.
Scatter-hoarding—A behavior in which an animal stores (caches) food in times of
surplus for times when food is less plentiful.
Schreckstoff—Olfactory cues in the form of alarm odors of depredated conspecifics.
Search image—A mechanism by which predators learn to focus attention on the
cryptic prey type most frequently encountered during recent searching.
Secondary sexual traits—Characteristics that distinguish males and females of a
species but that are not directly part of the reproductive system.
Sedentary—An organism that is attached to the substrate or does not normally
move long distances.
Seismic communication—Signals that are propagated through the ground.
Selfish genetic element (SGE)—Segments of DNA that promote their own transfer
to the next generation at the expense of other genes.
Selfish herd—The reduction of an individual’s risk of predation by placing others
in the herd at risk.
Selfishness—The individual performing a behavior benefits at the expense of the
recipient.
Seminal fluid proteins (SFPs)—Proteins in the seminal fluid that are transferred to
the female during mating.
Sender—The originator of a signal.
Sensory bias—An explanation for the evolution of novel traits as a result of inherent
biases in the sensory system and brain.
Sensory drive—A model for signal evolution in which the environment affects sig-
nal intensity and fidelity, the sensory capabilities of the receiver evolve to achieve
the greatest sensitivity at the region of highest environmental transmission, and sig-
nal properties then evolve to match this sensory system.
Sensory exploitation model—An explanation for ornament evolution that is based
upon an arbitrary trait’s ability to exploit sensory-response systems that function in
contexts other than mate choice.
Sensory trap—Males attracting females for mating by producing signals that mimic
other stimuli occurring in different contexts and to which females are adapted to
respond.
Sequential choice theory—The view that choice depends on what other items a for-
ager is likely to encounter later.
374 GLOSSARY

Serial polygamy—A form of social behavior in which an individual has a series of

mates.
Sex difference in sexual selection (Imates)—A measure of the strength of sexual
selection that is based on the source of variation in reproductive success of males
under a strictly polygynous mating system.
Sex pheromone—A secreted or excreted substance that attracts a mate and elicits
sexual behavior.
Sex-role reversal—Females competing more intensely than males for access to
mates.
Sexual conflict—The conflict in evolutionary interests between males and females.
Sexual dimorphism—A distinct difference between males and females of the same
species.
Sexual excitant (aphrodisiac)—A secreted or excreted substance that enhances sex-
ual behavior.
Sexually antagonistic coevolution (SAC)—Cyclical coevolution between the male
and female of a species resulting from a sequence of adapations and counteradapta-
tions between them.
Sexual monomorphism—Males and females of a species being indistinguishable.
Sexual ornament—A conspicuous structure of an animal that is used in displays to
attract mates or intimidate rivals.
Sexual selection—Differential reproductive success caused by competition for
mates.
Sexual selection gradient (Bateman’s gradient)—The relationship between mating
success and reproductive success that is used to quantify the strength of sexual
selection.
Sexual size dimorphism—A difference in size between males and females of a
species.
Siblicide—The killing of a young individual by its sibling.
Signal—A characteristic that has evolved by natural selection to convey information
to other organisms.
Signal detection theory—A means of quantifying the ability to discern between
information-bearing energy patterns (signals) and random energy patterns that dis-
tract from the information (noise).
Skewed sex ratio—Numerically greater representation of one sex over the other.
Snake-harassment behavior—Prey sequentially approach and retreat from snakes
while kicking substrate at them and even biting in order to deter attack.
Social behavior—The interactions among animals, usually of the same species, rang-
ing from simple attraction between individuals, to temporary feeding aggregations and
mating swarms, to multigenerational family groups with cooperative brood care.
GLOSSARY 375

Sociality—The tendency to form social groups.

Social learning—Facilitation of the acquisition of adaptive patterns of behavior by
observing the behavior of others.
Social mating system—The characterization of mating relationships based on social
interactions and parental care strategies.
Social mediation—A mechanism that can maintain signal honesty by testing of sig-
nals and punishment of unreliable (“dishonest”) signalers.
Socioecology—The study of interactions among the members of a species and
between them and the environment.
Soldiers—Individuals that possess weapons used to defend the family.
Spatial and temporal mean crowding—A method of characterizing mating relation-
ships that is based on a mathematical interpretation of distributions in space and time.
Specialist forager (specialization)—An animal that uses one or a few kinds of
food.
Speciation—The evolutionary process by which new biological species arise.
Species recognition hypothesis—An explanation, first proposed A. R. Wallace, argu-
ing that arbitrary traits evolved to signal species identity.
Sperm competition—The process by which sperm cells of two or more different
males compete to fertilize an egg of a female.
Spite—One individual performing a costly behavior that hurts the recipient, so both
pay a cost.
Stable equilibrium—An equilibrium in which a disturbance results in a return to
the equilibrium condition.
Static—Not changing.
Stereotyped—Performed with little variation.
Stillness—Remaining motionless to avoid detection.
Stotting (spronking)—Bounding up and down while fleeing a predator.
Subsocial—Females associating to a varying degree with their offspring, ranging
from building a nest and laying an egg on a prey item that is sufficient to allow the
larva to develop into an adult, to providing direct care for prolonged period of time
but then departing before the young emerge as adults, to the situation where moth-
ers stay at the nest until offspring emerge and those newly emerged offspring then
assist their mother in rearing of additional broods.
Tail-flagging—Prey wagging their tails back and forth as a signal to a predator.
Teaching—Behavior that facilitates learning of a naïve individual at some cost to
the teaching individual.
Template—A pattern, mold, or structure that guides the formation of a duplicate
copy. An auditory template is a memory used to guide the development of bird
song.
376 GLOSSARY

Territorial behavior—The acts by which an animal (or group of animals) protects

an area from trespassing by others of the same species.
Territory—A fixed area containing specific resources such as food, nest sites, or
mates from which one or more individuals exclude other members of the same
species.
Testosterone—A steroid hormone in the androgen group, found in vertebrates.
Time in—The amount of time that an adult is sexually active or capable of mating.
Time out—The amount of time that an adult spends resting or preparing for
mating.
Trade-off—A situation that involves losing one aspect of something in return for
gaining another aspect.
Transgenic—Pertaining to a gene or genes transferred from another species, or a
species that possesses a gene or genes transferred from another species.
Transparency—A camouflage strategy in which the body is clear.
Trapline—A circuit of specific foraging sites, much like a line of traps used by
trappers.
Ultimate—Referring to how a behavior affects the survival and reproductive success
of animals performing the behavior.
Ultrasonic—Above the range of human hearing.
Uniform spacing—A distribution of individuals or territories with relatively con-
stant distances between neighbors.
Unreliable—A signal that does not accurately predict something useful to the
receiver.
Variance—A measure of the spread of values on either side of the center.
Várzea—Seasonally flooded forest or woodland along rivers in the Amazon.
Vasopressin—A peptide hormone that plays a key role in regulating water, glucose,
and salts in the blood as well as social behavior and bonding.
Vocal signature—Distinctive cues that allow individual recognition by voice.
Waggle dance—A behavior performed by a scout honey bee to indicate the direc-
tion and distance to food.
Workers—Individuals that perform household duties and care for young.
 About the Editor and Contributors

EDITOR
KEN YASUKAWA is Mead Family Professor of Biology at Beloit College and
Honorary Fellow in Zoology at the University of Wisconsin at Madison.
Using Niko Tinbergen as his model, he has studied the behavior and ecology
of birds and has focused primarily on the red-winged blackbird since 1973.

CONTRIBUTORS
JANIS L. DICKINSON is a Professor in Natural Resources and Arthur A.
Allen Director of Citizen Science at the Lab of Ornithology at Cornell Univer-
sity. Her interests are in cooperative breeding and other forms of social behav-
ior at both proximate and ultimate levels and in research on socially networked
citizen science as a collective action.

ANNA DORNHAUS is an Associate Professor in the Department of Ecology

and Evolutionary Biology at the University of Arizona. She is interested in the
emergence of complexity and increased efficiency through collective behavior,
the effects of scaling in complex systems, and the role of learning and individ-
ual variability for collective success in social insects. She has also studied many
aspects of foraging, such as search strategies, learning, optimality, communica-
tion, and recruitment.
378 ABOUT THE EDITOR AND CONTRIBUTORS

RENÉE A. DUCKWORTH is an Assistant Professor in the Department of

Ecology and Evolutionary Biology at the University of Arizona. She is inter-
ested in ecological cycles, hybridization, dispersal, colonization of novel envi-
ronments, maternal effects, and the genetics of personality.

LEE ALAN DUGATKIN is Professor and Distinguished University Scholar

at the University of Louisville. He is currently studying the evolution of co-
operation, the evolution of aggression, the interaction between genetic and
cultural evolution, the evolution of antibiotic resistance, and the evolution of
risk-taking behavior and has various projects dealing with the history of
science.

GEOFFREY E. HILL is Professor and Curator of Birds in the Department of

Biological Sciences at Auburn University. His research focuses on the function
and evolution of ornamental traits in birds and on the coevolution of hosts
and pathogens using approaches ranging from behavioral ecology and phylo-
genetic reconstruction to immunology, parasitology, proteomics, and func-
tional genomics.

BRIAN KEANE is Associate Professor in the Department of Zoology at

Miami University. His research combines the use of molecular genetic tech-
niques with field studies on natural populations to address questions in evolu-
tionary biology and behavioral ecology.

WALTER D. KOENIG is a Senior Scientist at the Lab of Ornithology and

Department of Neurobiology and Behavior at Cornell University. He is inter-
ested in animal social behavior and behavioral ecology.

ZENOBIA LEWIS is a Lecturer at the Institute of Integrative Biology at the

University of Liverpool. She is an evolutionary biologist and behavioral ecolo-
gist who uses arthropods to investigate questions of sexual selection, sexual
conflict, and reproductive biology.

KENYON B. MOBLEY is a Postdoctoral Fellow at the Max Planck Institute

for Evolutionary Biology who is interested in sexual selection, mating systems,
and molecular evolution. His research seeks to understand how behavior, the
environment, and evolutionary history interplay to shape contemporary popu-
lations in a variety of fish species including pipefishes, gobies, and sticklebacks.
ABOUT THE EDITOR AND CONTRIBUTORS 379

TERRY J. ORD is a Lecturer in Evolution at the University of New South

Wales in Sydney, Australia. He is interested in how phylogeny, natural selec-
tion, and sexual selection contribute to the evolution of behavioral diversity.

MICHELLE PELLISSIER SCOTT is Emerita Professor of Biological Sci-

ences at the University of New Hampshire. Her research centers on the use
of insects as model systems to understand the costs and benefits of reproduc-
tive strategies. She focuses on silphid beetles, a diverse group of insects that
use carrion as a resource for reproduction, because they can be studied in the
field and in the laboratory while retaining an important element of naturalism.

NANCY G. SOLOMON is Professor of Zoology at Miami University. She is

a behavioral ecologist whose research interests focus on the ecology, reproduc-
tion, and behavior of small mammals and on sociality and mating patterns.

THEODORE STANKOWICH is an Assistant Professor in the Department

of Biological Sciences at California State University–Long Beach. He studies
the evolutionary ecology of antipredator defenses in a wide variety of taxa, par-
ticularly mammals. He has worked on risk assessment during predatory
encounters, the effects of relaxed selection on predator recognition, and the
evolution of defensive weaponry and aposematism.

ZULEYMA TANG-MARTÍNEZ is Founders Professor of Biology at the

University of Missouri–St. Louis. She is interested in animal communication
and focuses on chemical communication.
This page intentionally left blank
 About the Editorial Board

Daniel T. Blumstein, Department of Ecology and Evolutionary Biology,

University of California–Los Angeles, Los Angeles, California
Dorothy L. Cheney, Department of Biology, University of Pennsylvania,
Philadelphia, Pennsylvania
David L. Clark, Department of Biology, Alma College, Alma, Michigan
Ellen S. Davis, Department of Biological Sciences, University of Wisconsin–
Whitewater, Whitewater, Wisconsin
Bennett G. Galef Jr., Department of Psychology, Neuroscience, and Behav-
iour, McMaster University, Hamilton, Ontario, Canada
James C. Ha, Department of Psychology, University of Washington, Seattle,
Washington
Kim L. Hoke, Department of Biology, Colorado State University, Ft. Collins,
Colorado
Robert L. Jeanne, Department of Entomology, University of Wisconsin–
Madison, Madison, Wisconsin
Patricia B. McConnell, Department of Zoology, University of Wisconsin–
Madison, Madison, Wisconsin
Douglas W. Mock, Department of Biology, University of Oklahoma,
Norman, Oklahoma
382 ABOUT THE EDITORIAL BOARD

Michael Noonan, Department of Biology, Canisius College, Buffalo, New

York
Sarah R. Partan, School of Cognitive Science, Hampshire College, Amherst,
Massachusetts
Robert M. Seyfarth, Department of Psychology, University of Pennsylvania,
Philadelphia, Pennsylvania
Zuleyma Tang-Martínez, Department of Biology, University of Missouri–St.
Louis, St. Louis, Missouri
 Index

Abiotic habitat, 233 consequences of, 315–16; inconsistency

Abiotic variables, 133 of correlations with other traits, 309–10;
Acorn woodpeckers, 258, 277–78 individual spacing and population cycles,
Acoustic communication, 56–57 310–12; introduction to, 295–96;
Adams, Dean, 316 invasion and range limits, 312–15;
Adaptations: antipredator, 25, 27; behavioral, parental care, 306–8; population cycles,
30, 42, 43; evolution of, 199; sexual 310–12; range limits, 312–15; spacing
conflict and, 179 behavior and, 239–40; summary, 316–17;
Adaptionists, 3–4 temperament traits, 308–9
Adaptive alert responses, 33 Aggression-boldness-explorativeness syn-
Adaptive hypothesis, 299–300 drome, 309
Adaptive mate choice, 161–62, 166–67 Aggressive mimicry, 32
AFLP (Amplified fragment length Aggressive personality variation, 296–99
polymorphism), 104 Agonistic behavior, 62–63, 69–70, 71, 75
African carnivores, 341–42 Agonistic character displacement, 315–16
African elephants, 205 Akimoto, Shin-ichi, 186
African starlings, 277 Alarm calls, 53–54, 64, 84, 272, 331–32
African wild dogs, 242 Alarm odor, 53
Aggregation, 226–28, 236–37, 268–69 Alarm pheromone, 53, 58, 67
Aggregation pheromones, 58 Alarm substance, 67
Aggression: aggressive personality variation, Allee effect, 241
296–99; agonistic behavior, 62–63; body Allocation trade-offs, 163
size, 302; consistency in expression of, Allogrooming, 58, 59
299–301; correlations with other traits, Allospecific songs, 263
301–10; definition of, 295–96; Alternative reproductive tactics, 234
dominance, 303–6; ecological Altruism: alarm calls as, 272; causes of,
consequences of, 310–15; evolutionary 266–67; definition of, 327–28; inclusive
384 INDEX

fitness theory, 329–34; introduction to, Back-up food supply, 209–10

327–28; kin recognition, 334–35; Badgers, 230–31
reciprocal, 266–67 Banded mongooses, 228
Altruistic punishment, 267 Bank voles, 241, 303
Altruistic workers, 204 Banner-tailed kangaroo rats, 85
Amazonian manatees, 28 Barbour, Matthew, 85
Ambivalent behaviors, 71, 73 Bark beetles, 58, 64
The American Naturalist, 329 Barred owls, 313
Amoeba, 262 Bateman, Angus J., 103, 119
Amplified fragment length polymorphism Bateman gradient, 111, 120, 121, 122–24
(AFLP), 104 Bateman’s principles, 103, 119–20
Andersson, Malte, 122 Bats, 15, 54, 264, 266
Androgens, 236 Beater effect, 273
Anisogamy, theory of, 101, 122 Bed bugs, 178–79
Antelope, 70 Beetles, 58, 64, 181, 199–200, 210, 269
Anthropogenic pollution, 133 Begging behavior, 210–12
Antipredator adaptations, 25, 27, 44, 46 Behavior, in aggregations, 226–27
Antithetical displays, 63, 75 Behavioral adaptations, 30, 42, 43
Ants, 6, 64, 65, 204, 261–62, 309, 312–13 Behavioral ecologists, 334
Aposematism, 33, 36–39, 272–73 Behavioral ecology, 1, 15, 145, 214, 339
Arbitrary, ornamental traits as, 153 Behavioral isolating mechanisms, 68
Arctic foxes, 314 Behavioral messages, 53
Arctic skuas, 231 Behavioral signals, 55
Argentine ant, 312–13 Behavioral syndrome, 14, 308
Arginine vasopressin, 235, 236 Behavior ecology, 339–40
Armaments. See Ornaments and armaments Belding’s ground squirrels, 265, 272, 330–32
Armor, 41–42 Berglund, Anders, 167–68
Arnold, Stevan, 122, 162 Bighorn sheep, 151
Arnqvist, Göran, 184, 185 Biotic variables, 133
Arraunt, Eduardo, 28 Bird flocks, 226–27
Artificial selection experiments, 304–5 Birds: chemical signals, 59; diversity in
Assessment, 34–35 behavior, 15; parental care, 198–99;
Assessment/communication phase, parent-offspring relations, 215; sexual
26, 34–39 conflict in, 185; siblicide, 208; vision
Associated social monogamy, 225 systems, 30–31
Association hypothesis, 202 Bird song, 53, 262–64
Atlantic salmon, 234 Black-tailed deer, 34, 43
Attack, 24, 40–41 Bluebirds, 311–12. See also Mountain
Attack/flight phase, 26, 39–41 bluebirds; Western bluebirds
Australian fairy-wrens. See Superb fairy wren Blue-footed boobies, 209
Australian sawfly, 271 Blue tits, 205
Autonomic behaviors, 71, 72 Blumstein, Dan, 34
Body size, and aggression, 302
Baboons, 260, 266, 275 Bonanza resources, 198
Baby birds, 53 Bonobos, 2
Background matching, 31 Boonstra, Rudy, 235–36
Background noise, 82 Bowerbirds, 56, 60
INDEX 385

Bowers, 56 Chemical weapons, 41, 42

Breeding bout, 106, 107 Cheney, Dorothy, 80
Broadcast signal, 53–54 Chimpanzees, 2, 15, 56
Broad-nosed pipefish, 123, 128, 129–30, Chipmunks, 314
134–35 Chitty, Dennis, 310, 312
Brockmann, Axel, 66, 67 Chivers, Doug, 44
Brood parasitism, 268 Chucks, 352–53
Brook sticklebacks, 316 Cichlid fishes, 63–64, 309
Brook trout, 44, 45 Circadian rhythms, foraging and, 6
Brown, Grant, 44 Clark, Rulon, 85
Browne, Janet, 327 Cleaner fish, 86, 266–67
Brown-headed cowbirds, 268 Cliff swallows, 12, 269–70
Brown recluse spiders, 59–60 Climate, 133
Brown trout, 237 Clumped resources, access to, 275–76
Brumm, Henrik, 82 Clutton-Brock, Tim, 116
Bumble bees, 7 Cochlear, 345, 351
Burying beetles, 181, 199–200, 210 Cockroaches, 11, 42–43, 265
Butler, Mark, 237 Coefficient of relationship, 329
Coevolution, 352
California ground squirrels, 36, 37, 43, 85 Collared flycatchers, 202
California mice, 236, 307–8 Collective defense, 227
California towhees, 31 Coloniality, 227–28, 231, 269–70
California voles, 229 Coloration, 33, 36–39
Camouflage, 31–33 Combat weapons, 41, 42
Campbell’s monkeys, 84 Comb-footed spiders, 309
Cannibalism, 206–7 Comeault, Aaron, 37–39
Cape porcupine, 225 Communal defense, 231
Capture/escape, 24 Communal nesting, 228
Carazo, Pau, 81 Communication: acoustic, 56–57; agonistic
Carcasses, 198–99 behavior, 62–63, 69–70, 71, 75;
Cardinalfish, 207 chemical, 57–59, 64; contexts of, 61–67;
Care givers, 200–203 definition of, 52; in families, 210–12;
Carnivores, 1–2, 12 food and predators, 64–67; group
Caro, T., 39 coordination, 64; honesty of, 76–79,
Casas, Jérôme, 41 150–51, 162–64, 211–12; information
Castes, 259, 280 content, 53, 67–71; interspecific, 51,
Caterpillar communication, 353–54 83–86; introduction to, 51–54;
Cattle egrets, 209 manipulation, 76; modalities of, 54–60;
Cavity nesters. See Mountain bluebirds; multiple modalities, 60; parental care,
Western bluebirds 63–64; predation and antipredator
Central-place foragers, 3, 11 behavior, 35–36, 64–67; with predators,
Chacma baboons, 260 35–36, 64–67; ritualization and origins of
Chapman, Tracey, 183–84 displays, 71–74; sexual behavior, 61–62;
Character displacement, 315–16 signals conveying information, 79–83,
Chase-away model, 156–60, 167 210–12; summary, 86; tactile and
Cheating, 151 vibrational, 59–60; typical- and variable-
Chemical communication, 57–59, 64 intensity displays, 74–76; vibrational,
386 INDEX

59–60; visual, 55–56; vocal, 345, movements, 72; sexual behavior, 61–62;
351–53 sexual conflict in, 176; tactile and
Communication signals, 12–13, 52, 79–83, vibrational communication, 59–60
210–12 Courtship displays, 346–48
Compartmentalization, 259 Cowbirds, 268
Competition, vs. mate choice, 149–50 Cox, Matt, 39
Competitors, 199 Coyotes, 23–24, 231
Competitor-to-resource ration (CRR), 115 Crabs, 306
Condition-dependent model, 163 Cresswell, Will, 40
Conflict. See Families, conflict and Crickets, 61
cooperation in Cross-fostering experiments, 214
Confusion effect, 227 Crow, James F., 105
Conservation, effects on intraspecific Crowned eagles, 84
variation in spacing behavior, 241–42 CRR (Competitor-to-resource ration), 115
Conspecific cues, 228 Crustaceans, 59–60
Conspecifics, 222 Crypsis, 236
Conspecific tutor songs, 263 Cryptic female choice, 134
Conspicuousness, 39 Cuckolding, 202
Constraint hypothesis, 299–300 Cues, 12–13, 30, 33, 44
Constraints, 4 Culture, learning and, 11
Contact calls, 64 Curio, Eberhard, 33
Contrast effects, 9 Currency, 4, 197
Conventional sex roles, 109 Current benefits hypothesis, 282
Cooper, William, 39–40 “Customer” fish, 86
Cooperation. See Families, conflict and Cuttlefish, 32
cooperation in
Cooperative breeding: based on limited Dark-eyed juncos, 59
resources, 275–76; eusociality, 281; Darwin, Charles: altruism, 327; on dogs, 63;
extended families, 203; reproductive skew, on female choice, 152–53; on honey bees,
276–80; social behavior, 276–80 327–28; mating systems, 100; On the
Cooperative defense, 271–73 Origin of Species, 327, 328; secondary
Cooperative polygamy, 277–78 sexual traits, 104; sexual selection,
Cooperative polygynandry, 277, 279 118–19, 145; on shared ancestry of species,
Cooper’s hawks, 31 348; on social communication, 353
Copulation solicitation display, 69 Darwin-Fisher theory of sexual
Coral reef fish, 298 selection, 104
Core offspring, 208 Dawkins, Richard, 76–77
Coss, Richard, 33 Deception, and communication, 76
Cost/benefit analysis, 197, 203, 213 Decision rules, 9–10
Costs of providing care, 200 Decision to flee, 39–40
Costs of remaining, 39 Decreasing marginal returns, 11
Cougars, 23–24 Deduction, 339
Counteradaptation, 179 Deep-sea anglerfish, 32
Countershading, 32 Deer, 23–24, 70, 231
Courtship behavior: in dinosaurs, 346–48; Defense against predators, 26, 41–42
honesty in communication and, 76–77; Dehydroepiandrosterone (DHEA), 235
information content in, 69; intention Density in aggregations, 226–27
INDEX 387

The Descent of Man and Selection in Relation Earwigs, 212

to Sex (Darwin), 100 Eastern tent caterpillars, 274
Desert ants, 6 Eavesdropping, 53–54, 263
Desert iguanas, 229 Ecological consequences of aggression,
Desertion, 201 310–15
Desert locust, 237–38, 273 Ecological constraints hypothesis, 276
Desert spiders, 302 Ecological mating-system influences, 133–34
Detection, 24, 30–33 Economically defendable, 275
Detection phase, 25–26 Ectospermalege, 178–79
Detritus, 335 Edaphosaurus, 347–48
Developmental constraints, 299 Edinburgh University, 182
Dewlaps, 55 Egg clutches, 344–45
DHEA (Dehydroepiandrosterone), 235 Elephants, 56, 205, 242
Diana monkeys, 54, 84 Elk, 146
Dicots, 230 Emancipation, 74
Diel vertical migration (DVM), 28, 29 Emberizine finches, 43
Diet selection, 9 Emerald cockroach wasp, 11
Dill, Lawrence, 24–25, 39 Emlen, Stephen, 101–2
Dilution effect, 232, 270–71 Emus, 345
Dimetrodon, 347–48 Encounter, 24, 28–30
Dingemancse, Niels, 308 Endler, John, 82–83
Dinosaurs, 344–48 Environment, importance of in
Diploid eggs, 280 communication, 82–83
Direct observation, 339 Environment, information about, 70
Dispersed social monogamy, 225–26 Environmental potential for
Displacement activities, 71, 72–73, 74 polygamy, 101–2
Displays, 54; ritualization and origins of, 71–74; Escape, from predators, 42–43
typical- and variable-intensity, 74–76 Estradiol, 261
Disruptive coloration, 32 Estrogen, 235, 263
Diurnality, 29 Ethological isolating mechanisms, 68
Divergence, 186 Eurasian penduline tit, 200–201
Diversity in animal behavior, 14–15 Eurasian red squirrels, 231
DNA fingerprinting, 104 European cuckoos, 268
Dogs, 61, 62–63, 71, 75, 82 Eusociality, 226, 258–59, 280–81, 332–33
Dolphins, 11 Evolution, 4, 145
Domenici, Paolo, 42–43 Evolutionary consequences of aggression,
Domestic pigs, 303 315–16
Dominance, 270–71, 303–6 Evolutionary ecology, 149
Dorsal sails, 347–48 Evolutionary history of behavior:
Dragonflies, 41 introduction to, 339–40; linking fossils to
Drosophila melanogaster. See Fruit flies behavior of living species, 340–48;
Ducks, 176–77 recovering past behaviors, 348–54; work
Dung flies, 185 of studying, 354–56
Duration of displays, 74 Evolutionary spite, 268
Duvall, David, 122 Evolutionary tug-of-war, 174
DVM (Diel vertical migration), 28, 29 “The Evolution of Altruistic Behaviour”
Dynamic signals, 55 (Hamilton), 329
388 INDEX

The Evolution of Animal Communication Fitness, 4, 27, 158–59, 197

(Searcy and Nowicki), 82 Fitness proxies, 4
Experience, in predator/prey Fleeing behavior, 36, 40
encounters, 44, 45 Flight-initiation distance, 35
Experimental evolution, 180 Flight-intention movements, 72
The Expression of the Emotions in Man and the Florida scrub-jay, 276
Animals (Darwin), 353 Flower constancy, 8–9
Extended families, 203–4. See also Families, Flycatchers, 154, 155
conflict and cooperation in Font, Enrique, 81
Extinction, 186–87, 242 Food sources, and spacing systems, 228–31
Extrafloral nectaries, 10 Foraging for food: communication and,
Extra-pair copulation, 149, 202 64–67; decision to, 5–7; efficiency,
Extra-pair paternity, 104 273–74; explaining diversity in animal
behavior, 14–15; extracting reward,
Facultative, 209 11–12; finding and selecting food, 7–8;
Families, conflict and cooperation in: care food and lifestyle, 14; foods to avoid,
givers, 200–203; communication signals 8–11; foraging trip, 5; individual tastes,
in, 210–12; extended, 203–4; family 13–14; introduction to, 1–5; social
planning, 204–6; introduction to, behavior for increased efficiency, 273–74;
195–96; parental care, 196–200; parental social foraging, 12–13
favoritism, 206–10; parent-offspring Foraging gene, 261
conflict, 212–14, 215; summary, 214–16 Forced copulation, 175–76
Family planning, 204–6 Form of displays, 74
Fathead minnows, 44, 45 Fossils, linking to behavior of living species:
Feedback mechanism, 166–67 dinosaurs with egg clutches, 344–45;
Female choice: chase-away model, 156–60, introduction to, 339–40; polygyny,
167; male-male competition and, 148–49; 345–48; sabertooth cats and mastodons,
models based on adaptive female choices, 341–44
161–62; models based on arbitrary female Fowler, Kevin, 182
choices, 153–61; in ornament evolution, Francis, Richard, 303
152–53; rooster plumes and, 159–60; Frederick, William, 39–40
runaway model, 153, 155, 157–58, Free-tailed bats, 264
160–62, 167; sensory exploitation Frequency of displays, 74
model, 155–56; species recognition Freshwater fish, 309
hypothesis, 153–55 Friberg, Urban, 184
Female defense polygyny, 275 Frog-eating bats, 54
Fernández-Juricic, Esteban, 30–31 Frogs, 37–39, 202–3
Ferrari, Maud, 44 Fruit flies: mating behavior, 103, 158; sexual
Fertilization, sexual conflict in, 176 conflict in, 180, 182–85, 187, 188; social
Fiddler crabs, 271 behavior, 273; tactile communication, 59
Filial cannibalism, 207 Functional spite, 268
Fire ants, 261–62 Future benefits hypothesis, 282
Fish, 202, 207
Fisher, Ronald, 104–5, 153, 160 Galápagos hawks, 277
Fisher’s fundamental theorem, 165 Game theory, 151, 203
Fishing spiders, 296–97 Generalist foragers, 6
Fish schools, 226–27 Genes, social behavior and, 261–62
INDEX 389

Genetic bottlenecks, 312 Heritability, 165

Genetic covariance, 316 Hermaphrodites, 122
Genetic drift, 352 Hermit crabs, 79
Genetic mating systems: definition of, 106; Hermit warblers, 313
vs. social mating systems, 104, 131; Heterospecifics, 222
terminology, 108; understanding sexual Hinde, Robert, 76, 78
selection, 105 Home ranges, 225, 232
Genetic relatedness, 327 Honesty in animal communication, 76–79,
Genic capture model, 166 150–51, 162–64, 211–12
Gestational diabetes, 214 Honey badgers, 86
Glucocorticoids, 260 Honey bees: alarm pheromone, 53, 58, 67; as
Good genes hypothesis, 165 central-place foragers, 3; costs and benefits
Good parent model, 165 of group living, 269; Darwin and,
Gopher snakes, 36, 37, 85 327–28; diversity in behavior, 14–15;
Gould, James, 66, 67 extended families, 204; foraging trips, 5;
Grey-sided voles, 229 as generalist foragers, 6; social behavior,
Ground squirrels, 68; Belding’s ground 274; social foraging, 12; waggle dance,
squirrels, 272, 330–32; California ground 65–67, 70
squirrels, 36, 37, 43, 85; smelly predators Honeyeaters, 84
and, 348–51; squirrels exploited smelly Honeyguides, 86
predators, 265 Hormones, 235–36, 260
Group augmentation, 279 Hornbill birds, 54
Group coordination, and communication, 64 Hosken, David, 185–86
Group size, and predator/prey Hotshot hypothesis, 274
behavior, 43–44 Hotspot hypothesis, 274
Grouse, 60 House crickets, 305
Guanacos, 271 Houseflies, 181
Guppies, 234, 237 House mice, 197–98
Gustatory signals, 57 Howard, Henry E., 224
HPA (Hypothalamic-pituitary-adrenal)
Habitat structure, 222 axis, 260
Hamadryas baboons, 275 Hughes, William, 333
Hamilton, William D., 329–30 Hummingbirds, 15
Hamilton’s rule, 266 Humpback whales, 205
Handicap principle, 77, 163 Hymenoptera, 280–81, 332–33
Haplodiploidy, 280–81, 333 Hypothalamic-pituitary-adrenal
Haploid eggs, 280 (HPA) axis, 260
Harassment, 186
Hard-life hypothesis, 277 Impulsivity, 9–10
Harris’s hawks, 273, 277 Incest avoidance, 279
Hatching asynchrony, 209 Inclusive fitness theory, 329–34
Hawaiian honeycreepers, 229 Index of resource monopolization,
Head bobs, 51, 52, 353 112, 125–26
Head crests, 347–48 Indiana University, 59
Hen-feathered gene, 159 Indicator models, 162–63
Herbivores, 1–2 Indicator traits, 151
Herds, 226–27 Indirect genetic effects, 316
390 INDEX

Indirect reciprocity, 267 235–36; proximate mechanisms

Individual identity, 68 underlying, 228–42; random distribution,
Individual recognition, 306 224; summary, 242–43; type of, 223–28;
Individual spacing, and aggression, 310–12 uniform distribution, 224–26. See also
Individual tastes, 13–14 Spacing behavior and systems
Infanticide, 175, 206–7, 225, 281–82 Invasion, and aggression, 312–15
Inference, 339 Irrationality, 9
Information, 52 Isotope ratios, 343–44
Information center, 12, 270 Iwasa, Yoh, 122
Information content: in bird song, 53;
environment, 70; individual identity, 68; Japanese grasshoppers, 186
metacommunication, 71; physiological Japanese serow, 225–26
condition/motivation, 69–70; sexual Jones, Adam, 103–4, 120, 125, 127–30
identity, 69; species identity, 67–68;
status, 70–71 Kalahari meerkats, 279
Infrasonic sounds, 56 Keeley, Ernest, 234
Insects: extended families, 203–4; parental Ketterson, Ellen, 59
care, 200; sexual conflict in, 177–79, Kin recognition, 68, 264–66, 334–35
185; siblicide, 208; tactile Kin selection, 203, 266, 281, 328, 330
communication, 59–60 Kinship, 327–28. See also Altruism
Insurance hypothesis, 209 Kirkpatrick, Mark, 160
Intention movements, 71–72 Kleptoparasitism, 269
Interlocus sexual conflict, 174 Krebs, John, 76–77
Intersexual selection: chase-away model,
156–60, 167; definition of, 148; La Brea tar seeps, 341–43
intrasexual selection as foundation of, Ladybird beetle, 269
167–68; models based on arbitrary female Laidre, Mark, 79
choices, 153–61; runaway model, 153, Lake trout, 44, 45
155, 157–58, 160–62, 167; sensory Lande, Russell, 160, 162
exploitation model, 155–56; species Langerhans, R. Brian, 24–25, 27
recognition hypothesis, 153–55 Leafcutter ants, 204
Interspecific communication, 51, 83–86 Leal, Manuel, 84–85
Intralocus sexual conflict, 174 Lek paradox, 165–66
Intrasexual competition, 100, 235 Leks, 274
Intrasexual selection, 148, 150, 167–68 Leopards, 84
Intraspecific communication, 51 Lesser spotted woodpecker, 231
Intraspecific variation in animal spacing: Life history traits, 197
aggregations, 226–28; aggression and Lifelong monogamy, 173
individual spacing, 310–12; consequences Ligon, J. David, 159
of, 238; developmental effects on spacing Lima, Steve, 24–25, 43
behavior, 236–38; effects on conservation, Limited resources, social behavior
241–42; effects on mating patterns and and access to, 275–76
reproductive success, 238–39; effects on Linkage disequilibrium, 160
population-level processes, 239–41; Linnaeus, Carl, 348
genetic basis for spacing behavior, Lions, 34, 173, 242
234–35; introduction to, 221–23; Living in groups, costs and benefits
neurobiological basis for spacing behavior, of, 268–70
INDEX 391

Lizards, 40, 52, 55 differential, 125; measurement of sexual

Lloyd, Monte, 105 selection, 108–9; molecular markers and
Lone hunter lifestyle, 341 genetic mating systems, 104–5; Morisita
Long-tailed tits, 265 index, 111, 125–26; operational sex ratio,
Long tails, 160–61 102, 109, 113–16, 127–30, 149; oppor-
Lordosis, 69 tunity measures, 120–21; polygamy,
Lorenz, Konrad, 264, 353 101–2, 106–7, 346; postcopulatory sexual
Lott, Dale, 222, 224, 233 selection, 122, 134–35, 188; potential
reproductive rate, 116–17; qualitative and
Magrath, Rob, 54 quantitative methods, 110–12; qualitative
Maher, Christine, 224, 233 measurements of sexual selection, 109;
Major histocompatibility complext (MHC), qualitative vs. quantitative methods for
265–66 measurement, 110–12; quantitative
Male-female pairs, spacing and, 225–26 measurements of sexual selection,
Male-male competition, 148–50, 159 118–19; selection on sexually selected
Mammalian herds, 226–27 traits, 126–27; sex difference in sexual
Mammals, and parental care, 198–99 selection, 124–25; and sexual conflict,
Mandrills, 147 180–81; social vs. genetic mating systems,
Manipulation, and communication, 76 131; spatial and temporal mean crowding,
Mantids, 177–78 109, 110, 117–18; summary, 135; termi-
Mantis shrimp, 63, 68 nology, 105–8; theoretical quantitative
Many-eyes effect, 44, 231–32 framework, 103–4; variation, 131–33
Marginal offspring, 208 “Mating Systems and Strategies” (Shuster and
Marler, Peter, 67, 70 Wade), 103
Martin, Oliver, 185–86 Mating-system variation, 131–33
Masquerade, 31–33 Matocq, Marjorie, 85
Mastodons, 341–44 Matthews, Lauren, 225
Mate choice vs. competition, 149–50 Maximum standardized selection differential,
Mate guarding, 175, 178, 180, 225 111, 125
Mate-sharing, 277–78 Maynard Smith, John, 155
Mathematical models of central-place McCaffery, Alan, 238
foragers, 3–4 McPhail, J. Donald, 234
Mating, sexual conflict in, 176 Mean crowding, 106, 117–18
Mating calls, reconstruction of, 351–53 Mean spacial crowding, 109, 110, 117–18
Mating competition, 127 Mean temporal crowding, 109, 110, 117–18
Mating patterns, 238–39, 345–48 Mendel, Gregor, 328
Mating systems and measurement of sexual Merlins, 43
selection: Bateman gradient, 111, 120, Mesopredatory fish, 28
121, 122–24; Bateman’s principles, 103, Message content, 52
119–20; commonly used terms, 106; Meta-analysis, 297
definition of, 99–100, 106; ecological Metacommunication, 71
mating-systems influences, 133–34; MHC (Major histocompatibility complext),
empirical examples, 127–30; future 265–66
challenges, 130–35; history of, 100–105; Mice, 58, 197–98, 206, 236, 307–8
index of resource monopolization, Microsatellite DNA, 104
125–26; introduction to, 99–100; Microtine voles, 311
maximum standardized selection Mobbing, 42, 67, 348
392 INDEX

Modeling studies, 15 Norway lemmings, 312

Molecular markers, and genetic mating Nowicki, Steve, 82
systems, 104–5 Nuptial gift, 196
Møller, Anders, 78 Nurse ants, 204
Monandrous, 333 Nuthatches, 54
Monandry, 174, 181
Monogamy, 102, 106, 108, Obligate, 208–9
174, 215–16, 346 Obligatory, parental care as, 198
Monogyny, 174, 181 Ocean skaters, 271
Moose, 43 Olfactory cues, 30, 33, 44, 207
Morisita index, 111, 125–26 Olfactory signals, 57, 59
Morphological adaptations, 30, 41–42, 43 On the Origin of Species (Darwin), 327, 328
Morphological signals, 55 Ontogenetic shifts, 236
Morphs, 335 Operational sex ratio (OSR), 102, 109, 110,
Morton, Gene, 228 113–16, 127–30, 149
Moths, 61 Opportunity for selection, 110, 120
Motivation, 69–70, 75–76, 78 Opportunity for sexual selection, 111, 120–21
Mountain bluebirds, 311–12, 313 Opportunity measures, 120–21
Mouthbrooders, 63–64, 207 Optimal foraging, 2
Mule deer, 271 Optimality models for decision to flee, 40
Munz, Tania, 66 Order-of-gamete-release hypothesis, 202
Muskoxen, 271 Oring, Lewis, 101–2
Mutualisms, 13, 86, 266 Ornaments and armaments: armaments,
evolution of, 150–52; in dinosaurs,
Naked mole-rats, 226 346–48; female choice and male-male
Natal populations, 331 competition, 148–49, 159; intrasexual
Natural selection: altruism, 327; antipredator selection as foundation of intersexual
adaptations, 25, 27, 44, 46; constancy in selection, 167–68; introduction to,
expression, 299; evolution by, 145; predatory 145–46; lek paradox, 165–66; mate
behavior and, 36; predatory defenses, 42 choice vs. competition, 149–50;
Nazca boobies, 208–9 mechanism for evolution via adaptive
Neotropical singing mice, 58 mate choice, 161–62, 166–67; models of
Nest construction, 196 intersexual selection based on arbitrary
Neuropeptides, 235–36 female choices, 153–61; ornamental traits
New Zealand spiny lobsters, 237 in territorial animals, 152; ornaments,
Nicrophorus orbicollis, 206 evolution of, 152–62; sexual ornaments,
Ninespine sticklebacks, 316 346–48; sexual selection, 145–49; sexual
Nobel Prize, 67 selection based on adaptive female choice
Nocturnality, 29 for ornaments, 161–62; signal honesty,
Nocturnal prosimians, 225 162–64; summary, 168
Nomadic species, 199 OSR (Operational sex ratio), 102, 109, 110,
Nonbehavioral messages, 53 113–16, 127–30, 149
Nonterritorial species, 238–39 Oviposition, 238
Noonan, Brice, 37–39 Owings, Don, 36
Northern pike, 44, 45
Northern spotted owls, 313 Pair bonds, 62
North Sea harbor seal, 242 Paleontologists, 341, 343, 345
INDEX 393

Paleontology, 339–40 Plastic learning capabilities, 33

Paradise fish, 303 Platyfish, 156
Paradise flycatcher, 154, 155 Play bow of dogs, 71
Parasites, 199 Pleiotrophy, 305
Parasociality, 258–59 Plugs, 175, 180
Parental care: aggression, 306–8; beetles, 181; Poison dart frogs, 37–39
birds, 198–99; care givers, 200–203; Polak, Michal, 188
communication and, 63–64; in dinosaurs, Poli’s stellate barnacle, 224
344–45; favoritism, 206–10; insects, 200; Polyandry, 102, 106, 108, 124,
males, 148; provision of, 196–200 174, 181, 333
Parental favoritism, 206–10 Polygamy, 101–2, 106–7, 346
Parental investment, 176, 197, 213 Polygynandry, 106, 107, 108, 124
Parental investment theory, 101 Polygyny, 102, 106, 124, 149, 174, 345–48
Parental table, 105 Pond skaters, 178
Parent-offspring conflict, 213–14, 215 Population biology, 222
Parent-offspring identification, 63 Population cycles, and aggression, 310–12
Parker, Geoff, 174, 187 Population density, 133–34, 232–33,
Partial loads, in foraging, 3 240–41, 310–11. See also Spacing
Partridge, Linda, 182 behavior and systems
Passerine birds, 307 Population-level processes, 239–41
Passive aggregations, 270–71 Positive allometry, 347–48
Past behaviors, recovery of: caterpillar Postcopulatory sexual selection,
communication, 353–54; mating calls, 122, 134–35, 188
351–53; squirrels exploited smelly Potential reproductive rate (PRR),
predators, 348–51 110, 116–17
Paternity, 202 Prairie voles, 230, 235, 239
Payne, Katy, 56 Praying mantis, 178
Peak shift phenomenon, 8 Predation and antipredator behavior:
Pearson, Scott, 313 assessment and communication, 34–39;
Pedetta, Silvia, 306 attack and flight, 39–41; communication
Peiman, Kathryn, 316 and, 35–36, 64–67; detection and
Penguins, 264–65, 334 recognition, 30–34; encounter, 24, 28–
Peregrine falcons, 43 30; ground squirrels, 349; introduction
Permeability, 259 to, 23–28; pursuit, escape, and defense,
Personalities, 13–14 41–43; social behavior as protection
Pfennig, David, 335 against, 271–73; spacing behavior and,
Phenotype matching, 265 236–37; stress and, 260; summary, 43–46
Pheromones, 57, 58 Predation risk, and spacing systems, 231–32
Phylogenetic comparative method, 340 Predator confusion, 231
Phylogenetic relationships, 349 Predator/prey encounters: four main steps,
Phylogenetics, 221, 339–40 24; frameworks for understanding,
Physiological condition, 69–70 24–25; interspecific communication, 51,
Pied flycatcher, 270 84–85; recognition, 25–26, 33–34, 43,
Piloerection, 72 63, 264–66
Pipefishes, 108, 123, 128, 129–30, 134–35 Predators, 10, 199
Piscivorous fish, 28 Preeclampsia, 214
Plasticity, 298 Prenuptial molt, 261
394 INDEX

Prey, and spacing patterns, 231 Red jungle fowl, 159–60

Primates, 59, 196–97, 306 Redshanks, 40, 43
Private information, 12 Red-sided garter snakes, 175, 177
Production costs, 163 Red-tail hawks, 31
Promiscuity, 174, 180, 215 Red-winged blackbirds, 209
Pronghorn antelope, 207–8, 229, 303 Reliable information, 77–78
Protean behavior, 42, 43 Rémy, Alice, 241
Provisioning, 6 Reproduction, and social behavior, 274–75
Proximate mechanisms: ecological and social Reproductive competition, 279
determinants of spacing systems, 228–34; Reproductive effort, 205
of social behavior, 261–66; in spacing Reproductive isolation, 185–86
systems, 221, 223; underlying spacing Reproductive skew, 120, 276–80
systems, 228–42 Reproductive success, 14, 119, 197, 238–39
PRR (Potential reproductive rate), Resource defense hypothesis, 225
110, 116–17 Resource-defense social mating systems, 118
Pruett-Jones, Stephen, 333–34 Resource monopolization, index of, 112,
Pteranodon, 346–48 125–26
Public goods games, 267 Resource-tracking hypothesis, 209
Public information, 12, 270 Rewards, extracting, 11–12
Puerto Rican lizards, 84–85 Ring-necked pheasants, 149
Pursuit and capture, 41 Ring-tailed lemurs, 60
Pursuit-deterrence signals, 84–85, 272 Risk assessment, 34–35, 39
Risk aversion, 197
Quantitative models/measurements, 3–4, Risk prone, 197
103–4, 110–12, 118–19 Risk sensitivity, 9
Quinn, John, 40 Ritualization and origins of displays:
ambivalent behaviors, 71, 73; autonomic
Rainbow trout, 44, 45, 233 behaviors, 71, 72; displacement activities,
Randall, Jan, 85 71, 72–73, 74; evolution of
Random distribution, 223–24 communication, 353–54; intention
Random walk, 7 movements, 71–72; redirected behaviors,
Range limits, and aggression, 312–15 71, 73–74
Rape. See Forced copulation Ritualized displays, 62
Rashed, Arash, 188 Robinson, Beren, 316
Ratchet model, 157, 167 Rodent mating systems, 215–16
Rats, 12, 261 Rodents, 228
Rattlesnakes, 36, 37, 85, 348–50 Rooster plumes, 159–60
Reále, Denis, 308 Rough-skinned newts, 127–29
Receiver, 52–53 Rowe, Matt, 36
Reciprocal altruism, 266–67 Rowher, Sievert, 313
Recognition, 25–26, 33–34, 43, 63, 264–66 Runaway model, 153, 155, 157–58,
Red-cockaded woodpeckers, 276 160–62, 167
Red deer, 205–6, 258 Rundus, Aaron, 36
Red-eared slider, 313 Ryan, Michael, 80
Red foxes, 314
Red grouse, 311 Sabertooth cats, 341–44
Redirected behaviors, 71, 73–74 SAC (Sexually antagonistic coevolution), 179
INDEX 395

Safe sleeping sites, 275 179–80; parental investment, 176;

Salamanders, 127–29, 314, 316 summary, 187–88; as a universal truth,
Sand tiger sharks, 207, 210 175–79
Sandwich terns, 228 Sexual dimorphism, 224–25, 263
Satisficing, 7 Sexual excitants, 61
Sawflies, 273 Sexual identity, 69
Scatter-hoarding, 231 Sexually antagonistic coevolution (SAC), 179
Scent marking behaviors, 72 Sexually monomorphic, 346
Scheel, David, 34 Sexually selected traits, 126–27
Scheuerell, Mark, 28 Sexual ornaments, 346–48
Schindler, Daniel, 28 Sexual selection: among males, 201;
Schreckstoff, 33 antipredator adaptations, 25, 27, 44, 46;
Scrounging, 12 Darwin-Fisher theory of, 104; definition
Seahorses, 108, 131 of, 99; measurement of, 108–9; models
Search image, 8, 30 based on adaptive female choice for orna-
Searcy, William, 82 ments, 161–62; opportunity for, 111,
Secondary sexual traits, 104, 149, 201 120–21; ornaments and armaments,
Sedentary species, 199 145–49; postcopulatory, 122, 134–35,
Seismic communication, 55 188; qualitative measurements of, 109;
Selection on sexually selected traits, 126–27 quantitative measurements of, 118–19;
Selfish genetic element (SGE), 187 sex difference in, 124–25; sexual conflict
Selfish herd, 231, 270–71 and, 179–80. See also Mating systems and
Selfishness, 266, 267–68 measurement of sexual selection
Seminal fluid proteins (SFPs), 183–84, 185 Sexual selection gradient, 120, 122
Sender, 52 Sexual size dimorphism, 345–46
Sen Sarma, Moushumi, 66, 67 Sexy sons, 160–61
Sensory bias, 80, 353 Seychelles warblers, 276
Sensory drive, 82–83 Seyfarth, Robert, 80–81
Sensory exploitation model, 155–56 SFPs (Seminal fluid proteins), 183–84, 185
Sensory traps, 80, 81 SGE (Selfish genetic element), 187
Sequential choice theory, 10 Sherman, Paul, 330–32
Serengeti wildebeest, 343 Shuster, Stephen, 103
Serial polygamy, 200 Siblicide, 206, 207–9
Serotonin, 236, 300 Signal detection theory, 8
Sex attractants, 61 Signal honesty, 162–64
Sex difference in sexual selection, 111, 118 Signals, 12–13, 52, 79–83, 210–12
Sex peptide (SP), 184 Skewed sex ratio, 148–49
Sex pheromones, 57, 69 Slabbekoorn, Hans, 82
Sex ratio (SR), 187 Smith, W. John, 53, 76, 78
Sex ratio meiotic drive, 187 Snake-harassment behavior, 36
Sex-role reversal, 109, 113, 122, 123, 201 Snakes: diversity in behavior, 15; gopher
Sexual behavior, and communication, 61–62 snakes, 36, 37, 85; ground squirrels and,
Sexual conflict: consequences of, 185–87; 43; Puerto Rican lizards and, 84–85;
definition of, 174; in fruit flies, 180, rattlesnakes, 36, 37, 85, 348–50;
182–85, 187, 188; in insects, 177–79, red-sided garter snakes, 175, 177; sexual
185; introduction to, 173–75; link with conflict in, 177
mating system, 180–81; operation of, Snowshoe hares, 312
396 INDEX

Social behavior: access to clumped or limited behavior, 235–36; population-level

resources, 275–76; categorization of, processes effects, 239–41; predation and
258–59; cooperative breeding and antipredator behavior, 236–37; predation
reproductive skew, 276–80; eusociality, risk, 231–32; proximate mechanisms
226, 258–59, 280–81, 332–33; foraging underlying, 228–42; random distribution,
efficiency, 273–74; genes and behavior, 224; summary, 242–43; type of, 223–28;
261–62; introduction to, 257–58; kin uniform distribution, 224–26
recognition, 68, 264–66, 334–35; living Spadefoot toads, 210, 335
in groups, 268–70; passive aggregations, Spanish terrapins, 313
selfish herd, and the dilution effect, Sparrowhawks, 40, 43
270–71; protection against predators and Spatial mean crowding, 106, 117–18
intruders, 271–73; proximate Specialist foragers, 6
mechanisms of, 261–66; proximate vs. Speciation, 186
ultimate causation, 259–61; range of, Species identity, 67–68
268–81; reproduction, 274–75; study of, Species recognition hypothesis, 153–55
281–83; ultimate causes of, 266–68; vocal Sperm competition, 134, 183
learning, 262–64 Sphecid wasps, 234
Social communication, 353–54 Spiders, 15, 41, 59–60, 181, 240. See also
Social foraging, 12–13 Brown recluse spiders; Comb-footed
Sociality, 258, 340 spiders; Desert spiders; Fishing spiders;
Social learning, 203 Wolf spiders
Social mating systems: definition of, 106; vs. Spider webs, 41
genetic mating systems, 104, 131; Spiny lizard, 234
resource-defense, 118; spacing behavior Spite, 266, 268
and, 238–39. See also Mating systems and Sponking, 84
measurement of sexual selection Spotted hyenas, 196–97
Social mediation, 150 Spronking, 70
Socioecology, 222 Squirrels, 348–51. See also Belding’s ground
Sockeye salmon, 29, 30 squirrels; California ground squirrels;
Soldier termites, 204 Eurasian red squirrels; Ground squirrels
Soma, Kiran, 235–36 SR (Sex ratio), 187
Songbirds, 151–52, 262–64 Stable equilibria, 161
Song sparrows, 236, 263–64 Standardized mating differential, 112
Sooty shearwater, 6 Standardized selection differential, 112
South American bush dog, 60 Standardized selection gradient, 112
SP (Sex peptide), 184 Stankowich, T., 39
Spacing behavior and systems: aggregations, Static signals, 55
226–28, 236–37, 268–69; aggression, Status, 70–71, 167–68
239–40, 310–12; consequences of, 238; Steelhead trout, 234
conservation effects, 241–42; Stereotyped behavior, 54
developmental effects on spacing Sticklebacks, 73, 207, 308, 316
behavior, 236–38; ecological and social Stillness, 32
determinants of, 228–34; genetic basis for Stotting, 43, 70
spacing behavior, 234–35; hormones and, Stress, 260
235–36; introduction to, 221–23; mating Submission, 62–63, 303
patterns and reproductive success effects, Subsociality, 258–59
238–39; neurobiological basis for spacing Sugano, Yoshikazu, 186
INDEX 397

Superb fairy wren, 84, 261, 278–79, 333–34 Traplines, 7–8

Super-colonies, 312 Treehoppers, 64–65, 269
Survival costs, 163 Tree shrews, 225
Survivorship, 27 Tuco-tucos, 233
Swordtail fish, 156, 302 Túngara frogs, 54, 79–80, 83, 351–52
Syngnathids, 108 Typical-intensity displays, 74–76
Syracuse University, 187
Ultimate mechanisms: of social behavior,
Tactile communication, 59–60 259–61; in spacing systems, 221
Tail autotomy, 40 Ultrasonic sounds, 56
Tail flagging, 36, 70, 85 Uniform distribution, 224–26
Taiwanese tree frogs, 210 Uniform spacing, 224–26
Tammar wallabies, 43 University College London, 182
Task specialization, 280 University of Cincinnati, 188
Teaching, and social foraging, 12 University of Michigan, 330
Teleost fish, 207 Unreliable information, 77–78
Temperament traits, 308–9 Urination in dogs, 57
Temperature, 133
Templates, 262, 334–35 Vampire bats, 266
Temporal mean crowding, 106, 117–18 Variable-intensity displays, 74–76
Tent caterpillars, 65 Variance, 201
Termites, 204 Várzea lakes, 28
Territorial animals, 151–52 Vasopressin, 215, 235, 236
Territorial behavior, 224 Veen, Jan, 228
Territoriality: effect on population dynamics, Velvet swimming crabs, 302
239–41; effects on mating patterns and Venner, Samuel, 41
reproductive success, 238–39; food Vervet monkeys, 272
abundance, 229, 233–34; hormones and, Vibrational communication, 59–60
235; uniform spacing, 224–25 Visual communication, 55–56
Territory, 224 Visual systems, 30–31
Territory in Bird Life (Howard), 224 Vocal communication, in dinosaurs, 345,
Testosterone, 215, 235, 261, 351–53
301, 307–8, 311 Vocalization, 265
Thomson’s gazelles, 43 Vocal learning, 262–64
Threespined sticklebacks, 207, 308 Vocal signatures, 334
Thrushes, 313 Von Frisch, Karl, 66–67
Tiger moth, 258
Time in, 113 Wade, Michael, 103, 124
Time out, 113 Waggle dance, 65–67, 70
Tinbergen, Niko, 73, 260–61, 353 Wallace, Alfred, 152–53
Townsend solitaires, 229 Warning coloration. See Aposematism
Townsend’s warblers, 313 Wasps, 268
Toxicity, 39 Waterfowl, 175–77
Trade-offs, 163, 197, 204–5, 296, 307, 309 Water mites, 80
Transgenic fruit flies, 187–88 Water striders, 178
Transmission of parasites and pathogens, 242 Wattled lapwing, 164
Transparency, 32 Weasels, 349
398 INDEX

Western bluebirds, 282, 302, 307, 311–12, Wrens, 313

313, 315 Wright, Dominic, 234
White-browed scrubwren, 84 Wright, Sewall, 329
White-crowned sparrows, 31, 263
White suckers, 44, 45 Ydenberg, Ron, 39
White-tailed deer, 32 Yello casqued hornbills, 84
Whittaker, Danielle, 59 Yellow-headed blackbird, 211
Wigby, Stuart, 184
Wild turkeys, 274 Zahavi, Amotz, 77, 163
Wilson, Edward O., 259 Zebra finches, 261, 263, 302
Wingfield, John, 235–36 Zebrafish, 234
“Win-stay, lose-shift,” 8 Zeus bugs, 178
Wolf spiders, 158 Zig-zag dance, 73
Wolves, 43, 303 Zooplankton, 28
Woodpeckers, 229, 258, 277–78 Zuberbühler, Klaus, 84
Worker termites, 204 Zwartjes, Patrick, 159
This page intentionally left blank
 Animal Behavior
How and Why Animals Do the Things They Do

Volume 3: Integration and Application with Case Studies

Ken Yasukawa, Editor

Copyright 2014 by Ken Yasukawa
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted, in any form or by any means, electronic, mechanical,
photocopying, recording, or otherwise, except for the inclusion of brief quotations in a
review, without prior permission in writing from the publisher.
Library of Congress Cataloging-in-Publication Data
Animal behavior : how and why animals do the things they do / Ken Yasukawa, editor.
pages cm
Includes bibliographical references and index.
ISBN 978–0–313–39870–4 (hard copy : alk. paper) — ISBN 978–0–313–39871–1
(ebook) 1. Animal behavior. I. Yasukawa, Ken, 1949–
QL751.A6498 2013
591.5—dc23 2013012228

ISBN: 978–0–313–39870–4
EISBN: 978–0–313–39871–1
18 17 16 15 14 1 2 3 4 5
This book is also available on the World Wide Web as an eBook.
Visit www.abc-clio.com for details.
Praeger
An Imprint of ABC-CLIO, LLC
ABC-CLIO, LLC
130 Cremona Drive, P.O. Box 1911
Santa Barbara, California 93116-1911
This book is printed on acid-free paper
Manufactured in the United States of America
 Contents

1. Integrated Studies of Stickleback Behavior 1

Alison M. Bell

2. Friends and Enemies: How Social Dynamics Shape

Communication and Song Learning in Song Sparrows 33
Michael D. Beecher and Çağlar Akçay

3. A Tale of Two Spiders: Investigating Communication in Two

Unique Model Species Using Video Digitization and Playback 63
George Uetz and David Clark

4. Studying Lifelong Male-Male Partnerships in a Tropical Bird 99

David B. McDonald

5. Be Loved, Be Prey, Be Eaten 121

Rachel A. Page, Michael J. Ryan, and Ximena E. Bernal

6. Siblicide in Birds 153

Douglas W. Mock

7. Referential Signals: A Window into Animal Minds 175

Carolynn L. Smith and Christopher S. Evans
iv CONTENTS

8. Cotton-top Tamarins: Research for Conservation and

Understanding Human Behavior 209
Charles T. Snowdon

9. Avian Brood Parasitism: How to Spot a Foreign Egg in the Nest? 249
Mark E. Hauber and Jason Low

10. How Animals Remember Places and Find Their Way Around 269
David F. Sherry

11. Cognitive Repertoire, Cognitive Legacies, and Collective

Reasoning: The Next Frontiers in Conservation Behavior 289
Guillermo Paz-y-Miño-C

12. Beauty in the Eye of the Beholder: Potentially Adaptive

Individual Differences in Human Mate Preferences 311
Lisa M. DeBruine, Benedict C. Jones, S. Craig Roberts,
Anthony C. Little, and Christopher D. Watkins
Glossary 349
About the Editor and Contributors 361
About the Editorial Board 365
Index 367
 1

Integrated Studies of Stickleback Behavior

Alison M. Bell

INTRODUCTION
Threespine sticklebacks (Gasterosteus aculeatus) have taught us a great deal
about animal behavior. This chapter starts with a tribute to sticklebacks,
explaining why this small fish has proven to be such a fascinating subject for
studies of animal behavior. Although there are many stories to tell about
sticklebacks, this chapter focuses on individual differences in behavior as a case
study to illustrate an integrative approach to studying animal behavior. The
chapter concludes with a forecast of what new genomic technologies are likely
to add to our understanding of individual differences in behavior and to the
study of animal behavior in general.

OVERVIEW OF STICKLEBACK BIOLOGY AND BEHAVIOR

The threespine stickleback is a small (adult size is 3–8 cm standard length) tel-
eost fish that is widely distributed throughout the northern hemisphere. The spe-
cies includes three fundamental life-history types: marine, anadromous, and
freshwater (Wootton, 1976; Baker, 1994). For many years, sticklebacks have
been a model organism in animal behavior, due in large part to the important role
they played in the development of ethological theory (Tinbergen, 1940). Niko
Tinbergen was awarded a Nobel Prize, which he shared with two other promi-
nent ethologists, Konrad Lorenz and Karl von Frisch, due in no small part to
the important discoveries that he and his collaborators made using this small fish.
2 ANIMAL BEHAVIOR

The reproductive behavior of male sticklebacks has been particularly well

studied both in the laboratory and in the field. During the breeding season,
male sticklebacks develop red nuptial coloration on the belly and blue around
the eye. They are highly aggressive toward their neighbors and vigorously
defend nesting territories. In the field, territorial male sticklebacks are fre-
quently confronted by intruders, and they are especially aggressive toward
other male sticklebacks—they chase and attack male sticklebacks that intrude
into their territory. Early ethological studies of sticklebacks emphasized the
role of sign stimuli, such as the male’s red belly, as elicitors of aggressive
behavior (Tinbergen, 1951). Besides these highly energetically demanding
defensive activities, territorial male sticklebacks actively court females and pro-
vide all of the parental care to the developing offspring, from fanning the eggs
to defending the nest and retrieving fry that stray from the nest. Therefore,
breeding male sticklebacks engage in a variety of activities, all of which are
important to reproductive fitness, but some of these activities are potentially
contradictory, even though they are not always temporarily or spatially sepa-
rated from each other. Sticklebacks have been a model system in ethology
for understanding how animals resolve conflicting demands, and early etho-
logical theory of “drives” and “motivation” were inspired by studies of stickle-
backs (Tinbergen & van Iersel, 1947). For example, high aggression
(aggression drive) might compromise a male’s courting ability (sex drive) and
might also reduce the quality of parental care (parental drive). Despite the
solid role that sticklebacks have played in studies of animal behavior, we do
not yet understand the molecular mechanisms underlying their rich behavioral
repertoire. However, this situation is likely to change in the near future aided
by emerging genomic tools, including a full genome sequence (Kingsley &
Peichel, 2007), described below.
In addition to their distinguished history as ethological subjects, stickle-
backs have also become favorites for the study of ecological speciation
(Rundle et al., 2000) and phenotypic evolution (Peichel et al., 2001; Cresko
et al., 2004; Colosimo et al., 2005). Freshwater populations of sticklebacks
exhibit dramatic population-level phenotypic variation, and all freshwater
populations are the descendants of marine or anadromous (fishes that spend
most of their lives in the sea and migrate to fresh water to breed) stickleback,
which are widely distributed in the Holarctic. Despite their wide geographic
distribution, marine sticklebacks are relatively uniform with respect to mor-
phology (Walker & Bell, 2000). Whereas the large marine populations are
likely to have changed relatively little over the last 7 to 12 million years (Bell
& Foster, 1994), the derived freshwater populations have undergone a dra-
matic diversification (Foster, 1995; McKinnon et al., 2004).
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 3

This radiation offers several unique opportunities for studying evolution-

ary processes. For example, much of the phenotypic variation that we observe
among freshwater populations is naturally replicated across independent sets
of populations. Freshwater populations rapidly evolved convergent pheno-
types in response to similar selective pressures, and they did so independently
by multiple colonizations of different freshwater drainages. Therefore different
populations of sticklebacks with convergent phenotypes are independently
derived from the marine ancestral type and are therefore biological replicates.
In other words, the radiation is a naturally replicated experiment. Because
much of the phenotypic variation among populations is adaptive and a
response to selection pressures in local environments, comparing across popu-
lations can provide clues about the important ecological factors affecting the
evolution of traits.
One of the most important selective factors shaping the biology of this
small fish is predation pressure. Sticklebacks often provide a primary food
source for a diverse assemblage of predators including predatory birds, fishes,
and some insects. Sticklebacks have lateral plates, which are rigid, bony plates
along the side of the body that offer some protection from toothed predators
such as predatory fishes, and there is tremendous variation among populations
in lateral plate morphology, often associated with predation regime (reviewed
in Reimchen, 1994). Similarly, variation in predation pressure has also shaped
the evolution of behavioral variation among stickleback populations. Stickle-
backs from areas where there are high levels of predation tend to show height-
ened sensitivity to predators when compared to their counterparts in relatively
low-predation environments (Huntingford & Coulter, 1989; Huntingford et
al., 1994; Walling et al., 2003, 2004; Messler et al., 2007).

CASE STUDY: THE BOLDNESS-AGGRESSIVENESS BEHAVIORAL SYNDROME

IN STICKLEBACKS
In a classic study, Felicity Huntingford showed that individual sticklebacks
varied along a “shy-bold” continuum. Some individuals were relatively bold
toward predators when they were young, while others were more timid. The
individuals that were relatively bold as juveniles grew up to become the indi-
viduals that were especially aggressive toward conspecifics later in life, as
adults. In other words, there was a positive correlation between boldness and
aggressiveness, or what we might call a boldness-aggressiveness behavioral
syndrome (Huntingford, 1976).
Part of the reason why this initial study is significant is that we tend to
think that behaviors in different contexts are independent of one another.
4 ANIMAL BEHAVIOR

My favorite example to illustrate this point is that standard animal behavior

textbooks often have separate chapters devoted to, for example, foraging
behavior, aggressive behavior, and territorial behavior (Bell, 2007). But there
is growing evidence for some unexpected correlations between behaviors, or
behavioral syndromes. These correlations are especially unexpected when they
are between behaviors that occur in very different contexts—it is not too
surprising that the time spent attacking a conspecific is correlated with the
number of bites delivered to a conspecific during a fight, for example. It is
more surprising to think that time spent attacking a conspecific correlates with
how an individual (imagine a male red-winged blackbird [Agelaius phoeniceus],
for example) behaves in a flock or during competition for resources, or with
how it parents. These kinds of correlations are unexpected because we tend to
assume that how an individual behaves in a particular context is independent
of how that same individual behaves in other contexts, and we tend to assume
that individuals do not have much trouble switching to the appropriate behavior
when situations change. In other words, a common assumption in behavioral
ecology is that there is no carryover between behaviors in different contexts
and that behavior is instantaneously plastic and changes according to the imme-
diate circumstances (Sih et al., 2004). But one of the interesting things about
behavioral syndromes is that if individuals have a tendency to behave a certain
way, then individuals might not be able to behave optimally in all contexts; they
have a behavioral type that limits their behavioral flexibility, and a behavioral
syndrome can be a constraint on optimal behavior (Sih et al., 2004). In what
follows, I describe two sets of experiments with sticklebacks to test whether
the previously identified boldness-aggressiveness syndrome might act as a con-
straint over either developmental or evolutionary time.

ARE BEHAVIORAL SYNDROMES CONSTRAINTS OVER DEVELOPMENTAL TIME?

Huntingford (1976) found that male sticklebacks that were relatively bold
toward predators were also relatively more aggressive toward conspecifics.
Does this apply to sticklebacks throughout their entire development? In other
words, is a bold and aggressive fish always constrained to be bold and aggres-
sive from a young fry all the way through adulthood?
We can think about the stability of behavior through development in a
variety of ways (Stamps & Groothuis, 2010), but for our purposes here, we
are interested in whether rank-order individual differences in a single behavior
are stable over time (sensu differential stability [Stamps & Groothuis, 2010]).
Imagine we measure a behavior, say behavior Y, at three points in time on five
different individuals, each represented by a different line in Figure 1.1.
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 5

Figure 1.1. The stability of a single behavior. Each line represents the behavior of a dif-
ferent individual, measured at three different points in time. Rank-order differences
between individuals are stable in (a) but not in (b).

A stable behavior is one in which the rank-order differences among individ-

uals are preserved through time—the individual that showed the highest levels
of behavior Y at time 1 also showed the highest level of behavior Y at time 2.
So according to this type of stability, we are not asking whether behavior does
not change—it can go up or down in magnitude. Instead, we are asking about
whether the rank order of individuals changes. An unstable behavior is one in
which individuals might be relatively low on a behavior at one age and then
very high at another, so that we are not able to predict how an individual will
behave at time 2 if we know how the individual behaved at time 1.
The scenario described above just considered the stability of one trait at a
time rather than an entire syndrome or correlation between behaviors. Again,
we can use hypothetical data to distinguish between a stable versus an unstable
syndrome. Imagine we observe a group of animals, and we measure boldness
and aggressiveness on each individual at two different ages. If at age 1 there
is a positive correlation between boldness and aggressiveness and at age 2 we
also see a positive correlation, we could infer that the syndrome, or the corre-
lation, is stable (Figure 1.2).
If, on the other hand, there is a positive correlation at one age but not at
another, then the syndrome is unstable. It is important to note that a stable
syndrome does not require stable single behaviors because we might observe
the same relationship between two traits at two ages, even if individual behav-
ior changes over time.
To study the development of a behavioral syndrome in sticklebacks, Judy
Stamps and I (Bell & Stamps, 2004) reared the offspring of 11 full-sibling
families from two populations (Putah Creek and the Navarro River) of
6 ANIMAL BEHAVIOR

Figure 1.2. The stability of a behavioral syndrome. Individuals are indicated with dif-
ferent data points. (a) A syndrome is stable when we observe the same correlation
between behaviors at different points in time, that is, at different ages. Note that a
stable syndrome does not necessarily imply stable single behaviors (the individual
represented by the filled circle showed the highest levels of behaviors x and y at age 1
but the lowest levels at age 2). (b) A syndrome is unstable when behaviors are corre-
lated at some ages but not others.

sticklebacks in the lab and measured each individually marked individual’s

aggressiveness and boldness on three occasions during development. There-
fore, for each fish, we had a record of its aggressive behavior and boldness at
three points during development. To measure aggressiveness, we introduced
a conspecific of the same size and sex and from the same population into a
transparent cylinder and recorded the response of the focal individual to the
conspecific (Figure 1.3).
To measure boldness, we measured each individual’s reaction to a predator.
We used different techniques to measure boldness toward a predator for each
of the different ages in order to simulate the different types of predators expe-
rienced by different age classes (Figure 1.4).
Figure 1.3. Measuring aggressiveness in sticklebacks. The focal fish is presented with
an intruder confined to a flask.

Figure 1.4. Measuring boldness, see text for details. (From Bell, 2005)
8 ANIMAL BEHAVIOR

This approach was adopted from developmental psychology, in which

there is a tradition of using different tasks to measure the same behavior at dif-
ferent ages. For example, one might measure inhibition in children in different
ways at different ages—in an infant, you might measure inhibition as attach-
ment to the mother, while in a toddler you might measure inhibition by reac-
tion to peers. The basic idea in all of these observations, though, was the same
—to record the response of the fish to predation risk—and we interpreted
behaviors such as predator inspection behavior, foraging under predation risk,
and high levels of activity after a simulated predator attack as “boldness.” On
juveniles, we measured each individual’s reaction to a live coast range sculpin
(Cottus aleuticus). On subadults, we measured each individual’s willingness
to forage under predation risk after a simulated attack by a great egret (Ardea
alba), and on adults we measured their willingness to forage under predation
risk by a live largemouth bass (Micropterus salmoides).
We found that neither boldness nor aggressiveness were stable over time
(Bell & Stamps, 2004); individuals that were bold as a juvenile did not neces-
sarily grow up to become a very bold adult, or vice versa. Similarly, a very
aggressive juvenile was not necessarily a very aggressive adult. But that was just
one trait at a time. The results for the boldness-aggressiveness syndrome are in
Table 1.1, which show the correlation coefficients between boldness and
aggressiveness for each population at each age.
In the Navarro population, boldness and aggressiveness were significantly
positively correlated with each other at all three ages. That is, as juveniles, sub-
adults, and adults, individuals that were more bold were also more aggressive.
For the Putah population, though, boldness and aggressiveness were related
to each other in juveniles but not in subadults or adults. In other words, in
the Putah population, the syndrome changed over ontogeny. However, in
the Navarro population, the single behaviors were unstable, but the syndrome
was stable. That is, boldness and aggression were positively correlated at all

Table 1.1. Correlation coefficients between boldness and aggressiveness at three dif-
ferent ages in two populations. Note that in the Navarro population, boldness and
aggressiveness were positively correlated with each other at all three ages (Bell &
Stamps, 2004).
Population Juvenile Subadult Adult

Navarro (N ¼ 35) 0.56** 0.45** 0.60**

Putah (N ¼ 28) 0.43* 0.12 0.18
*P < 0.05 **P < 0.01
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 9

three ages, so that fish that were more bold were also more aggressive, but a
fish that was bold as a juvenile did not necessarily grow up to become a bold
adult. However, if a fish was bold as a juvenile, it was also aggressive as a
juvenile.
Altogether, these results suggest that correlations between behaviors do not
necessarily mean that an individual who started out as one behavioral type, for
instance very bold and aggressive, is destined to grow up to be bold and aggres-
sive. On the one hand, this can be interpreted as a lack of developmental con-
straint because there was some flexibility of behavior. However, the pattern
that we observed in the Navarro population was that behaviors tended to
change together in a coordinated fashion, which means that an individual
could not be both shy and aggressive at the same time. Therefore some pheno-
typic possibilities were limited. The reversible coordinated shift of an entire
suite of behaviors in concert with each other is similar to other familiar behav-
ioral transitions, such as the reversible shift between different alternative
mating types. For example, when a male cichlid obtains a territory, he
becomes brightly colored, very aggressive, and courts females. But this shift
is reversible and if that male loses his territory, the entire suites of traits
changes back in a coordinated way (Hofmann et al., 1999).

ARE BEHAVIORAL SYNDROMES CONSTRAINTS OVER EVOLUTIONARY TIME?

As described above, sticklebacks exhibit tremendous population-level varia-
tion in behavior and morphology. The radiation is a powerful natural experi-
ment because there were independent colonization events and tremendous
divergence among populations. If we return to the previously identified
boldness-aggressiveness syndrome in sticklebacks (Figure 1.5), we might envi-
sion that the traits are correlated in a hypothetical marine ancestor of
sticklebacks.
If the correlation between boldness and aggressiveness were to act as a con-
straint, we can hypothesize about how we would expect the two traits to be
related to each other in descendant populations in different selective environ-
ments, such as streams that were independently colonized by a marine ances-
tor and then underwent divergence.
If the two behaviors are genetically correlated and if selection were to favor,
for example, high levels of aggressiveness in one population (open circles) but
not in another population (gray circles), then over a few generations we might
expect the individuals from the population depicted with open circles not only
to be more aggressive but also to be more bold so that the entire population
shifts up along the correlation. In other words, if the correlation between
10 ANIMAL BEHAVIOR

Figure 1.5. The evolutionary fate of correlated traits. Each data point represents a dif-
ferent individual, coded by population; population means are designated with stars.
See text for details. (From Bell, 2005)

boldness and aggression acts as a constraint, we would expect to see the behav-
ioral syndrome in different populations, and we would expect the means of the
two populations, designated by stars in Figure 1.5, to also lie along the
correlation.
Alternatively, if the correlation between boldness and aggression is not con-
strained by a genetic correlation, or if the correlation can be uncoupled, we
can envision a variety of possible outcomes. Perhaps the correlation is broken
apart in different populations. Or perhaps the mean level of one of the traits,
such as aggression, differs between the populations, but the syndrome is
present in both—the intercept is simply shifted up, but the slope is the same.
Or maybe the correlation is broken apart in one population but not the other.
My point in going through these scenarios is to point out that comparing the
relationship between behavioral traits in different populations can tell us
whether a behavioral syndrome is acting as an evolutionary constraint.
To test this hypothesis, I collected wild-caught adults from the Navarro
River and Putah Creek. I brought the adults into the lab, and I measured bold-
ness and aggressiveness on each individual, using the methods described earlier
(the subjects were actually the parents of the fish described in the previous sec-
tion). The results for the wild-caught adults are in Figure 1.6.
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 11

Figure 1.6. Aggressiveness is on the y axis, and boldness is on the x axis. These results
show principal component scores. The means are indicated with stars. Boldness and
aggressiveness were positively correlated in one population but not the other (Navarro
r ¼ 0.52, N ¼ 29, P < 0.01; Putah r ¼ 0.13, N ¼ 42, NS). (From Bell, 2005)

The first thing that was most obvious—to me, at least—when watching the
fish is that Navarro fish were very timid around the predator—while Putah
fish boldly foraged under predation risk, the Navarro fish spent most of their
time hiding, and this was a statistically significant difference between the
means of the two populations. The other thing that was immediately obvious
was the Putah fish were much less aggressive than Navarro fish. Again, the dif-
ference in population means is statistically significant. So means of the popu-
lations, although they are just two data points, do not support the idea that
the positive correlation between boldness and aggressiveness can act as a con-
straint because the population that was more bold (Putah) was actually less
aggressive.
When we look closer at the actual individual values—each of the dots rep-
resents a different individual, coded by population—we found that across
individuals from the Navarro River, there was a significant positive correlation
between boldness and aggressiveness, so that the most aggressive fish were also
12 ANIMAL BEHAVIOR

the most bold. So the boldness-aggressiveness syndrome appears to be a

general or robust feature of the Navarro population—it characterized wild-
caught adult Navarro fish; it was stable through development; it was main-
tained in the lab; and, moreover, based on quantitative genetic analyses using
both parent-offspring regression and full-sibling analysis, there was a strong
genetic correlation between boldness and aggression in the Navarro popula-
tion (Bell, 2005). In the Putah population, though, the two traits appeared
to be independent of one another. So the fact that boldness and aggression
are not always related to each other in different populations of stickleback, as
well as the fact as the means are actually negatively related to each other
between the two populations, suggests that this syndrome can be readily bro-
ken apart.
All in all, these results show that boldness and aggressiveness do not always
go together: boldness and aggressiveness are not always related to each other at
different ages or in different populations. These results beg the question as to
why behaviors might be packaged together in some groups but not others and
encourage us to develop proximate and ultimate explanations for why behav-
ioral correlations should occur sometimes but not always (for progress on that
front see (Stamps, 2007; Wolf et al., 2007; Dingemanse & Wolf, 2010; Wolf
& Weissing, 2010). In what follows, I describe a series of experiments
approaching this question empirically from both an ultimate and a proximate
perspective. From an ultimate perspective, we can approach the issue from the
top down by asking if the syndrome can be adaptive and how it is shaped by
selection. From a proximate perspective, we can ask whether boldness towards
predators and aggressiveness toward conspecifics share a common neuroendo-
crine link that might couple them together (Figure 1.7).

IS THE BOLDNESS-AGGRESSIVENESS SYNDROME ADAPTIVE UNDER PREDATION RISK?

Are there ecological differences between the two populations that might
give some insights into why boldness and aggressiveness were positively corre-
lated with each other in the Navarro River population but not in the Putah
Creek population (Bell, 2005)? Indeed, the populations differ in several
respects, but one of the ways that is most striking is with respect to predation
pressure. When I present this work in a talk I always like to ask the audience to
guess: Which population do you think is the high-predation population? Is it
the one where the two behaviors were correlated, or the one in which the two
behaviors were not correlated? Almost always there are votes for both
sides. Part of the point in asking the audience to consider this question is to
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 13

Figure 1.7. Approaching behavioral syndromes from the bottom up (by studying proxi-
mate mechanisms) and from the top down (by studying how they are shaped by
selection).

emphasize that while we are familiar with thinking about average differences
in behavior between populations, that is, we expect populations under high
predation pressure to have better-developed antipredator behavior, for exam-
ple; we have few expectations about how we expect correlations among behav-
iors to vary according to ecology or selection pressures. In other words, we
have yet to develop a conceptual framework—either mechanistic or adaptive
—that can explain why we observe consistent individual differences in behav-
ior across contexts.
Sticklebacks in the Navarro River have historically been subject to stronger
selection by piscivorous predators than fish in Putah Creek, suggesting that
the syndrome is associated with predation risk. Moreover, another study on
independent populations confirmed that sticklebacks behave more predictably
under predation pressure in multiple, replicated freshwater populations
(Dingemanse et al., 2007). Therefore the pattern is that we observe the
boldness-aggressiveness behavioral syndrome in sticklebacks when predation
pressure is high. To test whether predators are actually causally related to the
syndrome, we performed an experiment to assess whether we could generate
the syndrome by applying predation risk.
We exposed fish from the low-risk population of sticklebacks in which
boldness and aggressiveness were not correlated (Putah Creek) to real preda-
tion by trout. We measured boldness and aggressiveness on individual fish
and individually marked them. We then put groups of the fish into large semi-
natural tanks with live rainbow trout, with three trout per tank. Once half
14 ANIMAL BEHAVIOR

Figure 1.8. Boldness and aggressiveness were not correlated prior to exposure to preda-
tion (R = 0.131, N = 136, P = 0.127), but the syndrome appeared after selection (R =
0.462, N = 60, P < 0.0001. (From Bell & Sih, 2007)

of the sticklebacks had been eaten, we removed the trout from the tank and
recorded the identity of the survivors. As a control, we had a group of fish that
were treated identically except that they were not exposed to live predation.
We then remeasured the behavior of the survivors (Bell & Sih, 2007).
The results are in Figure 1.8 with each individual represented by a different
data point.
Before being subjected to predation by the trout, boldness and aggressive-
ness were not related to each other. After the predators consumed half the
sticklebacks, though, the syndrome appeared among the survivors. Impor-
tantly, the behaviors were not associated with each other in the control group
either before or after. This provides strong evidence that predation actually
generated the boldness-aggressiveness behavioral syndrome.
However, there are two ways in which this result could have come about
(Figure 1.9). The data from “before” are in Figure 1.9a.
The first way is via differential survivorship of particular behavioral types
(Figure 1.9a). That is, individuals along the diagonal were the ones that sur-
vived while individuals that were off the diagonal were eaten. Alternatively,
perhaps individuals shifted their behavior, so that the ones that were mis-
matched for the two behaviors, of the very bold and nonaggressive individuals,
and the very shy and aggressive individuals shifted their behavior, or plastically
moved onto the correlation (Figure 1.9b).
What we actually found was an intriguing combination of both of these
processes. We did a phenotypic selection analysis, which basically involves
regressing relative fitness on the standardized values for each of the traits, their
Figure 1.9. (a) Selection via disproportionate survivorship might have generated the
syndrome. (b) Alternatively, behavioral plasticity could have generated the syndrome
under predation risk.
16 ANIMAL BEHAVIOR

squared terms, and their products, to tell us about the strength of directional,
stabilizing, and correlational selection. Individuals that were more aggressive
had higher survivorship, and individuals that were less bold had lower survi-
vorship (Bell & Sih, 2007). But there was not correlational selection—or the
correlation itself was not selected for. An important caveat to this result,
though, is that we probably did not have sufficient statistical power to detect
correlational selection, which is very hard to measure (Kingsolver et al., 2001).
These results can be more easily visualized when we look at the data prior to
exposure to predators. Figure 1.10 shows individuals coded by whether or not
they survived—survivors are in open circles, but the ones that died are in
closed circles (Figure 1.10).
The rates of survivorship are marked in each corner—as you can see, the
bold and unaggressive individuals did really badly. This is consistent with
the directional selection gradients, which were going in that direction. How-
ever, the selection analysis did not reveal the whole picture because behavior
also changed: sticklebacks became less aggressive in response to the predator
(Bell & Sih, 2007), showing that there was some plasticity as well. Altogether,
we infer that predators preferentially consumed bold and unaggressive

Figure 1.10. Both selection and plasticity generated the syndrome. Filled circles re-
present individuals that died; open circles represent survivors. The rate of survival of
the different quartiles (bold and aggressive, bold and nonaggressive, etc.) are indicated
in each corner. (From Bell & Sih, 2007)
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 17

behavioral types (individuals in the top left corner) and that plasticity shifted
the entire population to the left (Figure 1.10).
In summary, we found that boldness and aggressiveness were not packaged
together as a syndrome before exposure to real predation, but the syndrome
appeared among the survivors, and both selection and plasticity induced the
correlation. Altogether this work shows that predation is, indeed, causally
related to the syndrome and shows that this behavioral syndrome can be adap-
tive, and selected for.
While just a few years ago it was surprising to consider that animals might
have “personality,” that is, that there are consistent individual differences in
the ways that animals behave, we now know that personality is widespread
and not just restricted to higher vertebrates (Gosling, 2001). We also know
that individual variation is often evolutionarily meaningful, in that it is herit-
able and related to fitness (Dingemanse & Reale, 2005). But one of the things
we have learned from sticklebacks is that the important question is not just
whether or not animals have personalities but that one of our main goals
should be to explain why personalities look the way that they do (Sih & Bell,
2008). Why do individuals behave consistently in some contexts but not
others? Why do some species, or some populations, or some individuals,
behave consistently while others do not? Why are some behaviors more consis-
tent than others? We are just beginning to develop a conceptual framework—
either from an adaptive perspective or by understanding the mechanism—that
can explain (or even better, predict) consistent individual differences in
behavior.

WHAT ARE THE PROXIMATE MECHANISMS UNDERLYING THE BOLDNESS-AGGRESSIVENESS

BEHAVIORAL SYNDROME?
While the previous section illustrated how we can understand a behavioral
syndrome from the top down, in this section, I describe insights into the
boldness-aggressiveness syndrome that come from studies of its underlying
neuroendocrine mechanisms, that is, from the bottom up (Figure 1.7). Here,
we are interested in whether boldness and aggressiveness share a common
neuroendocrine substrate that might link them together. To test this hypoth-
esis, we focused on variation in two aspects of the stress response: (1) the
release of cortisol, a glucocorticoid, and (2) the turnover of monoamines in
the brain.
There were several reasons why these were good places to start searching for
the proximate source of covariance between boldness and aggressiveness. In
several species, the shy-bold continuum and the proactive-reactive axis have
18 ANIMAL BEHAVIOR

been associated with individual differences in stress responsiveness (Koolhaas

et al., 1999). Considering the important role of predators in shaping behav-
ioral variation within and among stickleback populations and the fact that
predators are natural stressors, stress responsiveness is a promising neuroendo-
crine mechanism underlying behavioral variation. For example, how an indi-
vidual reacts to a variety of different dangerous situations, for instance a fight
with risk of injury or an encounter with a predator with risk of death, might
be influenced by the individual’s stress responsiveness. If the stress response
is generalized across behavioral contexts, and individual sticklebacks differ in
stress responsiveness, as has been shown for other animals (Koolhaas et al.,
1999), then variation in stress responsiveness could be the underlying root of
the covariance of behavioral responses in sticklebacks. Importantly, the experi-
ments described below were carried out on fish from a high-predation
(Scottish) population.
We found that both conspecifics and predators are stressors in sticklebacks,
as assessed by an increase in cortisol following either a “fight” or a “fright”
(Bell et al., 2007). Figure 1.11 shows cortisol concentrations of individuals
sacrificed at different points in time following a 15-minute exposure to either
a pike predator or a conspecific. Note the low cortisol levels of individuals
sacrificed directly from a holding tank in baseline, nonstressed conditions.
Exposure to the predator produced a dramatic stress response in terms of
cortisol released, with levels reaching their maximum within this interval at
60 minutes post exposure (Bell et al., 2007). In contrast, an aggressive interac-
tion with a conspecific also generated a stress response, but the time course
and magnitude of this response was different for these different stressors—
whereas cortisol levels were highest 15 minutes after a fight and then started
to subside, they were still elevated in the predator treatment up to 60 minutes
after exposure to the predator. Moreover, the overall magnitude of the cortisol
response was greater in response to the predator compared to the conspecific.
These results suggest that both of these kinds of interactions produce a stress
response in sticklebacks, as indicated by cortisol, but they suggest that expo-
sure to a predator might be perceived as a more serious threat than an aggres-
sive interaction, which is probably consistent with our intuition. An important
caveat to these results is that cortisol was measured in whole-body homoge-
nates so the cortisol that was detected probably includes both active, circulat-
ing cortisol in plasma as well as conjugated forms in other tissues.
For these same fish, I also measured concentrations of monoamines in dif-
ferent regions of the brain and observed several interesting differences between
these treatments. For example, concentrations of norepinephrine in the telen-
cephalon were very different following exposure to a predator versus an
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 19

Figure 1.11. Concentrations of cortisol increase following exposure to either a conspe-

cific (gray) or a predator (black). “Con” refers to individuals sampled from an unma-
nipulated tank; “settling” refers to individuals sampled from the same tank as the
experimental subjects but which were not confronted by either a conspecific or a
predator. Each of the different points represents the means of 10 individuals sacrificed
either 15, 30, or 60 minutes after visual and olfactory exposure to either a pike or a
conspecific. Note the difference in the time course between fish exposed to a predator
or a conspecific. (From Bell et al., 2007)

aggressive interactions (Figure 1.12): concentrations of norepinephrine were

very high in the predator treatment and stayed up during all the sampled
periods.
Following a fight, on the other hand, concentrations of norepinephrine in
the telencephalon went down relative to the controls and stayed down. We
interpreted these results as indicating that the fish perceived the predatory
encounter as more of a threat than the encounter with a conspecific and
20 ANIMAL BEHAVIOR

Figure 1.12. Concentrations of norepinephrine in the telencephalon following exposure

to a conspecific (gray) or a predator (black). (From Bell et al., 2007)

increased arousal or readiness to flee in that treatment (Bell et al., 2007). We

also found that serotonin concentrations decreased in both treatments relative
to the control but did so more for the conspecific treatment. This is somewhat
consistent with the idea that serotonin has an inhibitory effect on aggression.
In fact, we found that serotonin concentrations were negatively related to rates
of aggression (Figure 1.13a), whereas it was positively associated with bold
behaviors under predation risk (Figure 1.13b).
From these data, we can draw a few general conclusions. First, these results
suggest that although both fighting and a fright induce a stress response, they
are associated with markedly different monoaminergic responses, suggesting
that they really are different. Second, bold and aggressive behaviors were
actually associated with serotonin in opposite directions. We infer that
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 21

Figure 1.13a. Hypothalamic serotonin was negatively correlated with levels of aggres-
siveness at an individual level. (From Bell et al., 2007)

interactions with predators and conspecifics produce very different physiologi-

cal responses. My naïve expectation going into this project was that if the link
between boldness and aggressiveness is very strong in high-predation popula-
tions, then they just might be different manifestations of the same underlying
trait so that increasing concentrations of serotonin, for example, would be pos-
itively associated with both behaviors. These data contradict this simple
expectation and suggest instead that these two behaviors might be indepen-
dently regulated, which is consistent with our previous studies showing that
the relationship between boldness and aggressiveness is labile, both over evolu-
tionary and developmental time.

LOOKING FORWARD: GENOMICS AND ANIMAL BEHAVIOR

Genomics is transforming biology—everywhere you look, from biomedi-
cine to evolutionary biology (Rokas & Abbot, 2009) to physiology (Burggren
& Warburton, 2005) to ecology (Jackson et al., 2002), biologists are using
22 ANIMAL BEHAVIOR

Figure 1.13b. Hypothalamic serotonin was positively correlated with levels of boldness
toward a predator (predator inspection) at the individual level. (From Bell et al.,
2007)

new genomics tools to address long-standing questions. However, arguably, as

a discipline, animal behavior has been slower to embrace these technologies
(but see Robinson, 2002; Robinson et al., 2008), even though as a discipline
we are well schooled in the advantages of integrative work (Tinbergen,
1972). While up to this point this chapter has focused on the boldness-
aggressiveness behavioral syndrome in sticklebacks, the remainder of this
chapter will discuss general ideas about how genomics can add insights to
our understanding of animal behavior, especially personality in animals. Much
of this is speculative but is intended to convey some of the excitement that
new technologies can offer.

THE PHENOTYPIC GAMBIT

In a classic paper, Alan Grafen (1984) proposed that with a few simplifying
assumptions, behavioral ecology can ignore genetics. He argued that the aim
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 23

of behavioral ecology is to uncover the selective forces that shape characters

and that our method will work almost regardless of which genetic system
underlies the character. He proposed the phenotypic gambit, which was “to
examine the evolutionary basis of a character as if the very simplest genetic sys-
tem controlled it” (Grafen, 1984, p. 63). In other words, the phenotypic gam-
bit was that we do not really need to study the genetic or molecular
mechanisms underlying behavior if we are purely interested in functional
questions. For many behavioral ecologists, the gambit was appealing because
there were not tools available for tackling the mechanisms underlying complex
behaviors for the types of nonmodel organisms that are typically studied in
behavioral ecology. However, as sequencing costs plummet, it is just a ques-
tion of time until powerful genomics tools are available for virtually any
organism.
In what follows, I argue that studying the genetic and molecular mecha-
nisms underlying behavior is worth doing, using behavioral syndromes (per-
sonality) as an example. Many of the big unanswered evolutionary and
ecological questions about animal personality are either implicitly or explicitly
concerned with genetics (Bell & Aubin-Horth, 2010). For example, pleiot-
ropy (when a single gene affects multiple behaviors) is a textbook cause of lim-
ited plasticity and behavioral syndromes (Sih et al., 2004). Imagine a gene
product such as testosterone, for example, that influences aggression, parental
behavior, courtship behavior, and other traits. Individual males that have high
levels of testosterone after a fight, for example, might have trouble switching to
the appropriate behavior when the context changes, such as during parental
care (Wingfield et al., 1990). Another aim of studies of animal personality is
to explain heritable individual differences in behavior; natural selection erodes
heritable behavioral variation that is linked to fitness; why do we therefore
observe so much variation within natural populations? Finally, another central
question concerning the evolution of personality has to do with the evolution
of correlated traits—if two traits, such as boldness and aggressiveness, are
genetically correlated with each other, then selection for one behavior could
produce a correlated response to selection on the other behavior and might
constrain the independent evolution of both traits (Lynch & Walsh, 1998).
Obviously, this issue is also a question about genetics because it is concerned
with genetic correlations. Finally, there is growing interest in trying to come
up with a conceptual framework for understanding the evolution of personal-
ity traits across species. But this brings up important issues about how we
actually do this—how do we compare aggression, for example, in pythons, spi-
ders, and chimpanzees? One possibility is to use a genetic criterion for homol-
ogy, that is, aggression in pythons can be compared to aggression in
24 ANIMAL BEHAVIOR

chimpanzees if levels of aggression in both species are influenced by the same

set of genes. Therefore this question is also interested in genetics.
However, until recently, most studies interested in natural variation in eco-
logically relevant, fitness-related traits (including personality) have had to treat
the genetic mechanisms underlying their phenotypes of interest as a “black
box” (van Oers et al., 2005). For example, we might be interested in under-
standing whether the behaviors that constitute a behavioral syndrome are
genetically correlated with each other. While traditional quantitative genetic
approaches can be used to estimate a genetic correlation, an estimate of the
genetic correlation does not tell us anything about the actual genes that are
shared between the two traits.
Another approach to understanding the genetic basis of complex traits is to
try to identify regions of the genome related to the behavioral variation using
mapping approaches such as quantitative trait loci (QTL) mapping—an
obvious question might be whether the different behaviors that comprise a syn-
drome map to the same regions of the genome. However, the power of such
studies is determined by the magnitude of the genetic effects, and QTL studies
are most powerful when a few genes of large effect influence the phenotype
(Lynch & Walsh, 1998). Moreover, QTL mapping might not be a suitable
option for personality traits such boldness and aggressiveness, which are poly-
genic (with probably many genes of small effect contributing to the behavior)
and which have relatively low heritabilities (van Oers & Sinn, 2011).
Another approach is to look for associations between a candidate gene that
has been identified based on some other organism and a personality trait of
interest (Fitzpatrick et al., 2005). However, single candidate genes rarely explain
more than about 3 to 5 percent of the variation in a trait, and such studies are
renowned for failing to replicate, at least in the human literature, for example,
where associations between certain polymorphisms (e.g., serotonin transporter
or DRD4) and personality traits are found in some populations but not others
(Munafo et al., 2003). Moreover, epistatic interactions between loci make it
difficult to study the effects of single genes one at a time (Mackay, 2009).

WHOLE-GENOME EXPRESSION: ADVANTAGES AND CHALLENGES

New genomic technologies such as genome expression analysis (either
using microarrays or sequencing) for a whole genome are a good tactic for
studying the molecular mechanisms underlying behavior for several reasons
(Bell & Aubin-Horth, 2010). One of the first advantages has to do with the
fact that gene expression, like a personality trait, is a phenotype influenced
both by inherited genetic variation as well as the environment. That is,
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 25

differences between two individuals in levels of expression of a particular gene

could reflect either inherited, genetically based differences between the two
individuals or could reflect environmental effects or their interaction. Gener-
ally less than 35 percent of the variation in personality traits is due to genetic
variation (van Oers et al., 2005), and there is an important environmental
component to personality traits. Therefore one advantage of measuring gene
expression, as opposed to strictly concentrating on fixed genetic (sequence)
differences between genotypes, is that we can simultaneously study genes that
are responsive to the environment and that might be genetically variable
among groups.
While in the past it was only feasible to measure the expression of a small
number of genes at a time, the genomic revolution has introduced high-
throughput technologies such as microarrays and transcriptome sequencing
that allow researchers to measure the expression of the entire genome simulta-
neously. This makes whole-genome expression particularly appealing com-
pared to other tactics for studying the mechanisms underlying behavior
because it is likely that most personality traits in most organisms are highly
polygenic, so focusing on one gene at a time is unlikely to tell the whole story.
Instead, we can examine the coordinated action of the entire genome simulta-
neously. In theory, this unbiased approach is great for discovering new mech-
anisms in nontraditional model organisms. Collecting expression data for a
large number of genes simultaneously is advantageous because gene products
rarely act alone. Instead, they perform their functions by interacting together
in pathways and networks. As a result, the molecular changes that characterize
a phenotype do not just occur in a single marker or gene but rather in an entire
pathway. By measuring the entire transcriptome simultaneously, we have the
opportunity to look at relationships between expression of different transcripts
and to construct gene interaction networks (Robinson et al., 2008). Arguably,
by paying attention to pathways rather than specific genes in isolation, we
have an opportunity to ask about the regulation of the trait (Alaux et al.,
2009).
However, I do not want to give the impression that measuring gene expres-
sion is easy and the results are straightforward to interpret. One important
thing to keep in mind is that gene expression does not tell us anything causal
about the relationship between a gene and behavior. Gene expression can
change very rapidly, so that if we detect a difference in transcript level, for
example, between very bold versus very timid individuals following exposure
to a predator (Sanogo et al., 2011), whatever differences are observed could
be a response to the performance of the behavior, that is, attacking a predator,
or they could reflect an underlying causal tendency to behave in a particular
26 ANIMAL BEHAVIOR

way. Moreover, it can be difficult to interpret a list of differentially expressed

genes because mRNA abundance does not always predict protein abundance,
and many of the differentially expressed genes are likely to be only weakly
linked to the actual behavior.
There are also practical challenges. For example, the timing of sampling
could critically affect the results. Given the complexity of the pathways
involved, small differences in the time at sampling could produce dramatic
differences in the results that might be biologically meaningful; that is, the
genes might be involved in the initiation of the plastic response or in the main-
tenance of it (Aubin-Horth & Renn, 2009). If we are interested in identifying
genes related to behavior, do we sample during the execution of the behavior,
5 minutes after, 30 minutes after, or prior to it? Moreover, most studies inter-
ested in gene expression related to behavior measure expression in brain,
where information is centrally processed and integrated, but the brain is a
heterogeneous organ, and it is entirely likely that a particular gene might be
upregulated in one region and downregulated in another. Therefore if we
measured expression in the whole brain, the pattern would be “washed out.”
That being said, if we do elect to measure expression in a particular area, it is
nontrivial to determine which brain area is the best target for sampling.
Assuming that the problems described above are surmountable, there are
several types of simple experiments that we are only beginning to do. One
approach is to compare whole-genome expression among individuals that have
been exposed to different behavioral stimuli. For example, one could compare
individuals that were or were not exposed to an unfamiliar environment, or
were or were not confronted by an opponent, or a predator, or the like. As
results from different organisms and different treatments begin to be pub-
lished, it will be fascinating to start asking question such as whether any of
the genes that are differentially expressed in response to a predator (Sanogo
et al., 2011) are also differentially expressed in response to a conspecific
(Carney, 2007) or to a potential mate (Cummings et al., 2008), and so forth.
By looking for overlap among the lists of differentially expressed genes, we can
ask whether the same genes are expressed in different behavioral contexts to
see if the behaviors are regulated by the same mechanisms. If the same genes
are expressed in different behavioral contexts, then that suggests that the same
molecular mechanisms are involved with the two behaviors; that is, the two
behaviors are not mechanistically independent. Another approach is to com-
pare behaviorally divergent types of individuals. For example, we might com-
pare extreme behavioral types of individuals—very bold versus very timid,
for example, or individuals from different selected lines (Edwards et al.,
2006), populations, or species. Arguably, genes that are differentially expressed
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 27

between behaviorally divergent groups under baseline, unstimulated condi-

tions are more likely to be causally related to behavior than differentially
expressed genes in response to a stimulus or environment. Finally, we could
do both—compare behaviorally divergent groups’ response to the same treat-
ment—which could give insights into the evolution of plasticity (Alaux et
al., 2009; Bell & Aubin-Horth, 2010; Bell & Robinson, 2011).

CONCLUSION
Sticklebacks are likely to continue to play an important role in the study of
animal behavior and in behavioral genomics owing to their recent emergence as
a model organism for studying the genetics of adaptation (Colosimo et al.,
2005). Behavioral syndromes are also likely to be central players in integrative
studies of animal behavior because they encourage us to adopt a more holistic
view of animal behavior—rather than narrowly studying just one type of behav-
ior at a time, the behavioral syndromes view of the world emphasizes the value of
simultaneously considering relationships between diverse types of behavior (Sih
et al., 2004). Indeed, a truly integrative approach views behavior as just one
part of the organism’s entire integrated phenotype (Pigliucci & Preston, 2004).
Moreover, syndromes encourage integrative approaches because they pose both
proximate and ultimate questions (Sih et al., 2004).
Although this chapter focused on the behavior of sticklebacks in just two con-
texts—around predators and around conspecifics—there are many more fasci-
nating stories to tell about sticklebacks—from paternal care (Tulley &
Huntingford, 1987), to their complex social behavior (Ward et al., 2002), to
their strange divergence into benthic and limnetic types in certain lakes (Schluter,
1995), to their unusual “glue” used to construct nests (Jakobsson et al., 1999), to
the strange parasite-mediated changes in their behavior (Barber & Huntingford,
1995). Therefore it is likely that sticklebacks will continue to be excellent subjects
for integrative studies of animal behavior into the future.

ACKNOWLEDGMENTS
I thank Judy Stamps, Andy Sih, Felicity Huntingford, and Gene Robinson
for their mentorship, and the NSF and NIH for financial support.

REFERENCES AND SUGGESTED READING

Alaux, C., S., L. Sinha, G. J. Hasadsri, E. Hunt, G. Guzmán-Novoa, J. L. DeGrandi-
Hoffman, S. B. Uribe-Rubio, R, S. Rodriguez-Zas, & G. E. Robinson. (2009).
28 ANIMAL BEHAVIOR

Honey bee aggression supports a link between gene regulation and behavioral
evolution. Proceedings of the National Academy of Sciences, USA, 106, 15400–
15405.
Aubin-Horth, N. & S. C. Renn. (2009). Genomic reaction norms: Using integrative
biology to understand molecular mechanisms of phenotypic plasticity. Molecu-
lar Ecology, 18, 3763–3780.
Baker, J. A. (1994). Life history variation in female threespine stickleback. In M. A.
Bell & S. A. Foster (eds.), The Evolutionary Biology of the Threespine Stickleback
(pp. 144–187). Oxford, UK: Oxford University Press.
Barber, I. & F. A. Huntingford. (1995). The effect of Schistocephalus solidus (Cestoda:
Pseudophyllidea) on the foraging and shoaling behaviour of three-spined
sticklebacks, Gasterosteus aculeatus. Behaviour, 132, 1223–1240.
Bell, A. M. (2005). Differences between individuals and populations of threespined
stickleback. Journal of Evolutionary Biology, 18, 464–473.
Bell, A. M. (2007). Future directions in behavioural syndromes research. Proceedings
of the Royal Society of London, B, 274, 755–761.
Bell, A. M. & N. Aubin-Horth. (2010). What whole genome expression data can tell
us about the ecology and evolution of personality in animals. Philosophical
Transactions of the Royal Society of London, B, 365, 4001–4012.
Bell, A. M., T. Backstrom, F. A. Huntingford, T. G. Pottinger, & S. Winberg.
(2007). Variable neuroendocrine responses to ecologically-relevant challenges
in sticklebacks. Physiology and Behavior, 91, 15–25.
Bell, A. M. & G. E. Robinson. (2011). Behavior and the dynamic genome. Science,
332, 1161–1162.
Bell, A. M. & A. Sih. (2007). Exposure to predation generates personality in threes-
pined sticklebacks. Ecology Letters, 10, 828–834.
Bell, A. M. & J. A. Stamps. (2004). Development of behavioural differences between
individuals and populations of sticklebacks, Gasterosteus aculeatus. Animal
Behaviour, 68, 1339–1348.
Bell, M. A. & S. A. Foster. (1994). Introduction to the evolutionary biology of the
threespine stickleback. In M. A. Bell & S. A. Foster (eds.), The Evolutionary
Biology of the Threespine Stickleback (pp. 1–27). Oxford, UK: Oxford Univer-
sity Press.
Burggren, W. & S. Warburton. (2005). Comparative developmental physiology: An
interdisciplinary convergence. Annual Review of Physiology, 67, 203–223.
Carney, G. E. (2007). A rapid genome-wide response to Drosophila melanogaster social
interactions. BMC Genomics, 8, 288. doi:10.1186/1471-2164-8-288.
Colosimo, P. F., K. E. Hosemann, S. Balabhadra, J. Guadalupe Villarreal, M. Dick-
son, J. Grimwood, J. Schmutz, R. M. Myers, D. Schluter, & D. M. Kingsley.
(2005). Widespread parallel evolution in sticklebacks by repeated fixation of
ectodysplasin alleles. Science, 307, 1928–1933.
Cresko, W. A., A. Amores, C. Wilson, J. Murphy, M. Currey, P. Phillips, M. A. Bell,
C. B. Kimmel, & J. H. Postlethwait. (2004). Parallel genetic basis for repeated
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 29

evolution of armor loss in Alaskan threespine stickleback populations. Proceed-

ings of the National Academy of Sciences, USA, 101, 6050–6055.
Cummings, M. E., J. Larkins-Ford, C. R. L. Reilly, R. Y. Wong, M. Ramsey, & H.
A. Hofmann (2008). Sexual and social stimuli elicit rapid and contrasting
genomic responses. Proceedings of the Royal Society of London, B, 275,
393–402.
Dingemanse, N. J. & D. Reale. (2005). Natural selection and animal personality.
Behaviour, 142, 1159–1184.
Dingemanse, N. J. & M. Wolf. (2010). Recent models for adaptive personality differ-
ences: A review. Philosophical Transactions of the Royal Society of London, B,
365, 3947–3958.
Dingemanse, N., J. Wright, A. Kazem, D. Thomas, R. Hickling, & N. Dawnay.
(2007). Behavioural syndromes differ predictably between 12 populations of
three-spined stickleback. Journal of Animal Ecology, 76, 1128–1138.
Edwards, A. C., S. M. Rollmann, T. J. Morgan, & T. F. C. Mackay. (2006). Quan-
titative genomics of aggressive behavior in Drosophila melanogaster. PLOS
Genetics, 2, 1386–1395.
Fitzpatrick, M. J., Y. Ben-Shahar, H. M. Smid, L. E. M. Vet, G. E. Robinson, & M.
B. Sokolowski. (2005). Candidate genes for behavioural ecology. Trends in
Ecology and Evolution, 20, 96–104.
Foster, S. A. (1995). Understanding the evolution of behavior in threespine stickle-
back: The value of geographic variation. Behaviour, 132, 1107–1129.
Gosling, S. D. (2001). From mice to men: What can we learn about personality from
animal research? Psychological Bulletin, 127, 45–86.
Grafen, A. (1984). Natural selection, kin selection and group selection. In J. R. Krebs
& N. B. Davies (eds.), Behavioral Ecology: An Evolutionary Approach, Second
Edition (pp. 62–84). Oxford, UK: Blackwell Science.
Hofmann, H. A., M. E. Benson, & R. D. Fernald. (1999). Social status regulates
growth rate: Consequences for life-history strategies. Proceedings of the National
Academy of Sciences, USA, 96, 14171–14176.
Huntingford, F. A. (1976). The relationship between anti-predator behaviour and
aggression among conspecifics in the three-spined stickleback. Animal Behav-
iour, 24, 245–260.
Huntingford, F. A. & R. M. Coulter. (1989). Habituation of predator inspection in
the three-spined stickleback Gasterosteus aculeatus. Journal of Fish Biology, 35,
153–154.
Huntingford, F. A., P. J. Wright, & J. F. Tierney. (1994). Adaptive variation and
antipredator behaviour in threespine stickleback. In M. A. Bell & S. A. Foster
(eds.), The Evolutionary Biology of the Threespine Stickleback (pp. 277–295).
Oxford, UK: Oxford University Press.
Jackson, R. B., C. R. Linder, M. Lynch, M. Purugganan, S. Somerville, & S. S.
Thayer. (2002). Linking molecular insight and ecological research. Trends in
Ecology and Evolution, 17, 409–414.
30 ANIMAL BEHAVIOR

Jakobsson, S., B. Borg, C. Haux, & S. J. Hyllner. (1999). An 11-ketotestosterone

induced kidney-secreted protein: The nest building glue from male three-
spined stickleback, Gasterosteus aculeatus. Fish Physiology and Biochemistry, 20,
79–85.
Kingsley, D. M. & C. L. Peichel. (2007). The molecular genetics of evolutionary
change in sticklebacks. In S. Östlund-Nilsson, I. Mayer, & F. A. Huntingford
(eds.), Biology of the Three-spined Stickleback (pp. 41–81). Boca Raton, FL:
CRC Press.
Kingsolver, J. G., H. E. Hoekstra, J. M. Hoekstra, D. Berrigan, S. N. Vignieri, C. E.
Hill, A. Hoang, P. Gibert, & P. Beerli. (2001). The strength of phenotypic
selection in natural populations. American Naturalist, 157, 245–261.
Koolhaas, J. M., S. M. Korte, S. F. De Boer, B. J. Van Der Vegt, C. G. Van Reenen,
H. Hopster, I. C. De Jong, M. A. W. Ruis, & H. J. Blokhuis. (1999). Coping
styles in animals: Current status in behavior and stress-physiology. Neurosci-
ence and Biobehavioral Reviews, 23, 925–935.
Lynch, M. & B. Walsh. (1998). Genetics and Analysis of Quantitative Traits. Sunder-
land, MA: Sinauer Associates.
Mackay, T. F. C. (2009). The genetic architecture of behavior: Lessons from
Drosophila. Genetica, 136, 295–302.
McKinnon, J. S., S. Mori, B. K. Blackman, L. David, D. M. Kingsley, L. Jamieson, J.
Chou, & D. Schluter. (2004). Evidence for ecology’s role in speciation.
Nature, 429, 294–298.
Messler, A., M. A. Wund, J. A. Baker, & S. A. Foster. (2007). The effects of relaxed
and reversed selection by predators on the antipredator behavior of the threes-
pine stickleback, Gasterosteus aculeatus. Ethology, 113, 953–963.
Munafo, M. R., T. G. Clark, L. R. Moore, E. Payne, R. Walton, & J. Flint. (2003).
Genetic polymorphisms and personality in healthy adults: A systematic review
and meta-analysis. Molecular Psychiatry, 8, 471–484.
Peichel, C. L., K. S. Nereng, K. A. Ohgi, B. L. E. Cole, P. F. Colosimo, C. A.
Buerkle, D. Schluter, & D. M. Kingsley. (2001). The genetic architecture of
divergence between threespine stickleback species. Nature, 414, 901–905.
Pigliucci, M. & K. Preston (eds.). (2004). Phenotypic Integration: Studying the Ecology
and Evolution of Complex Phenotypes. Oxford, UK: Oxford University Press.
Reimchen, T. E. (1994). Predators and morphological evolution in threespine stickle-
back. In M. A. Bell & S. A. Foster (eds.), The Evolutionary Biology of the Threes-
pine Stickleback (pp. 240–273). Oxford, UK: Oxford University Press.
Robinson, G. E. (2002). Development: Sociogenomics takes flight. Science, 297,
204–205.
Robinson, G. E., R. F. Fernald, & D. F. Clayton. (2008). Genes and social behavior.
Science, 322, 896–900.
Rokas, A. & P. Abbot. (2009). Harnessing genomics for evolutionary insights. Trends
in Ecology and Evolution, 24, 192–200.
Rundle, H. D., L. Nagel, J. W. Boughman, & D. Schluter. (2000). Natural selection
and parallel speciation in sympatric sticklebacks. Science, 287, 306–308.
INTEGRATED STUDIES OF STICKLEBACK BEHAVIOR 31

Sanogo, Y. O., S. Hankison, M. Band, A. Obregon, & A. M. Bell. (2011). Brain tran-
scriptomic response of threespine sticklebacks to cues of a predator. Brain
Behavior and Evolution, 77, 270–285.
Schluter, D. (1995). Adaptive radiation in sticklebacks: Trade-offs in feeding perfor-
mance and growth. Ecology, 76, 82–90.
Sih, A. & A. M. Bell. (2008). Insights for behavioral ecology from behavioral syn-
dromes. Advances in the Study of Behavior, 38, 227–281.
Sih, A., A. M. Bell, J. C. Johnson, & R. Ziemba. (2004). Behavioral syndromes: An
integrative overview. Quarterly Review of Biology, 79, 241–277.
Stamps, J. A. (2007). Growth-mortality tradeoffs and “personality traits” in animals.
Ecology Letters, 10, 355–363.
Stamps, J. A. & T. G. G. Groothuis. (2010). The development of animal personality:
Relevance, concepts and perspectives. Biological Reviews, 85, 301–325.
Tinbergen, N. (1940). Die Ubersprungbewegung. Zeitschrift für Tierpsychologie, 4,
1–40.
Tinbergen, N. (1951). The Study of Instinct. Oxford, UK: Oxford University Press.
Tinbergen, N. (1972). The Animal in Its World: Explorations of an Ethologist.
Cambridge, MA: Harvard University Press.
Tinbergen, N. & J. J. A. van Iersel. (1947). “Displacement reactions” in the three-
spined stickleback. Behaviour, 1, 56–63.
Tulley, J. J. & F. A. Huntingford. (1987). Paternal care and the development of adap-
tive variation in anti-predator responses in sticklebacks. Animal Behaviour, 35,
1570–1572.
van Oers, K., G. de Jong, A. J. van Noordwijk, B. Kempenaers, & P. J. Drent.
(2005). Contribution of genetics to the study of animal personalities: A review
of case studies. Behaviour, 142, 1185–1206.
van Oers, K. & D. L. Sinn. (2011). Toward a basis for the phenotypic gambit: Advances
in the evolutionary genetics of animal personality. In M. Inoue-Murayama,
S. Kawamura, & A. Weiss (eds.), From Genes to Animal Behavior: Social Struc-
tures, Personalities, Communication by Color (pp. 165–183). New York: Springer.
Walker, J. A. & M. A. Bell. (2000). Net evolutionary trajectories of body shape evo-
lution within a microgeographic radiation of threespine stickleback (Gasteros-
teus aculeatus). Journal of the Zoological Society of London, 252, 293–302.
Walling, C. A., N. Dawnay, A. J. N. Kazem, R. Hickling, & J. Wright. (2003). Do
competing males cooperate? Familiarity and its effect on cooperation during
predator inspection in male three-spined sticklebacks (Gasterosteus aculeatus).
Journal of Fish Biology, 63, 243–244.
Walling, C. A., N. Dawnay, A. J. N. Kazem, & J. Wright. (2004). Predator inspec-
tion behaviour in three-spined sticklebacks (Gasterosteus aculeatus): Body size,
local predation pressure and cooperation. Behavioral Ecology and Sociobiology,
56, 164–170.
Ward, A. J. W., M. S. Botham, D. J. Hoare, R. James, M. Broom, J. G. J. Godin, &
J. Krause (2002). Association patterns and shoal fidelity in the three-spined
stickleback. Proceedings of the Royal Society of London, B, 269, 2451–2455.
32 ANIMAL BEHAVIOR

Wingfield, J. C., R. E. Hegner, A. M. Dufty, & G. F. Ball. (1990). The challenge

hypothesis: Theoretical implications for patterns of testosterone secretion,
mating systems, and breeding strategies. American Naturalist, 136, 829–846.
Wolf, M., G. S. van Doorn, O. Leimar, & F. J. Weissing. (2007). Life history trade-
offs favour the evolution of personality. Nature, 447, 581–585.
Wolf, M. & F. J. Weissing. (2010). An explanatory framework for adaptive personal-
ity differences. Philosophical Transactions of the Royal Society of London, B, 365,
3959–3968.
Wootton, R. J. (1976). The Biology of the Sticklebacks. London: Academic Press.
 2

Friends and Enemies: How Social Dynamics

Shape Communication and Song Learning
in Song Sparrows
Michael D. Beecher and Çağlar Akçay

INTRODUCTION
Songbirds have been a central focus for animal behaviorists studying the social,
ecological, and evolutionary processes that influence behavior. Their song has
long fascinated naturalists, and the relative ease of capturing, marking, and
observing songbirds has enabled detailed studies of known individuals. Several
model songbird species have emerged through decades of research. The song
sparrow (Melospiza melodia) is one of them. Starting with the pioneering work
of Margaret Nice (Nice, 1937, 1943) and continuing through the studies of
population dynamics on Mandarte Island in British Columbia (e.g., Smith,
1984; Smith et al., 2005; Smith & Zach, 1979), song sparrows have been a
major model system in North America for answering questions about behavior
and evolution. Here we describe the results of our own studies on song spar-
rows, ongoing since 1986 in Discovery Park, Seattle, Washington. We focus
on one major facet of our research, the question of how social dynamics in this
population have shaped the song system the birds use to communicate and
how it may have shaped the process of song learning.
In many ways, song sparrows are the archetypical temperate songbird: they
get their name from their beautiful, complex song, and, like most songbirds,
34 ANIMAL BEHAVIOR

they learn these songs. A male song sparrow—only males sing in this species—
learns his songs during his first year of life, the process ending when he is 9 to
10 months old and beginning his first breeding season. A major function of song
sparrow song is to mediate a male’s interactions with his territorial neighbors.
Song sparrows pair up monogamously and maintain territories that are
large enough for most of their foraging (though they do forage off the territory
some of the time). Both parents contribute roughly equally to parental care
(although the female does all the incubation) and in a good year may fledge
two clutches in a breeding season. Although pairs are socially monogamous,
they are not genetically monogamous, as about 25 percent of the young in a
nest are fathered not by the social mate but by one or more of the neighboring
males (O’Connor et al., 2006; Sardell et al., 2010; Hill et al., 2011). Extra-
pair mating appears to be the rule and not the exception in songbirds
(Griffith et al., 2002; Westneat & Stewart, 2003), and the 25 percent extra-
pair parentage rate seen in song sparrows is about average for a songbird spe-
cies. The bulk of territory defense is handled by the male, with song being
one important tool in the male’s defense arsenal. First and foremost, by sing-
ing from various posts within his territory, the male indicates the extent and
boundaries of his territory while signaling that he is home, ready to defend it
against any intruder. In addition to singing to “post” his territory, the male
can communicate, at a distance, with his neighbors. In particular, two coun-
tersinging males can use song to negotiate and remind each other of the loca-
tion of the boundary that separates their territories. These negotiations are
generally most intense in early spring before the onset of the breeding season,
especially between new neighbors (about 30 to 40 percent of song sparrows in
our population do not survive over winter).

DEAR ENEMIES
As is the case in many territorial species, long-term neighbors have an inter-
esting relationship, one that is partially competitive but also partially co-
operative. For the territorial male, his neighbor is a competitor for matings
(fertilizations), for food on the territory, and for the territory itself. At the same
time, however, he prefers the neighbor he knows to a new neighbor he does
not know because a new neighbor will not know and therefore not respect
the present boundaries and will extract costs in time and effort to reestablish
them. Thus the relationship of the territorial neighbors has been dubbed a
“dear enemy” relationship (Fisher, 1954; Temeles, 1994). We focus on this
relationship in this chapter and attempt to show how it affects both the song
communication system and the song-learning program of song sparrows.
FRIENDS AND ENEMIES 35

The dear enemy relationship is a reciprocal relationship. Just as a bird will

prefer the neighbor he knows to one he does not, so will his neighbor have
the same preference. Thus both of them should have a mutual interest in
keeping out interlopers and should cooperate to keep them out. However,
such reciprocity is difficult to demonstrate—because the interests of each indi-
vidual are essentially identical, cooperation is not easily distinguished from
pure self-interest. Nevertheless, in this chapter we will show that song spar-
rows do meet several of the conditions required for a dear enemy relationship.
We will then show how their song communication system helps them main-
tain this relationship, and, finally, how it may affect the nature of song learn-
ing in this species.
Students of behavior naturally hope to generalize at least some of their find-
ings from their study species to other species. In our case we want to generalize
our findings at least to other songbirds that have similar social systems. How-
ever, it turns out that it is a challenge to generalize even to other song sparrow
populations, as this widely distributed species (covering virtually all of North
America) is quite diverse geographically. Presently 24 subspecies of song spar-
rows have been identified (Patten, 2001; Arcese et al., 2002). Moreover, there
is already considerable evidence of differences between western (M. melodia
melospiza) and eastern (M. melodia melodia) populations. It is not clear
whether these differences are proximate effects or ultimate effects of environ-
mental and life history differences. In this chapter, we will comment from
time to time on these east-west differences but will talk mostly about our
western study population.

Background: Song in Song Sparrows

Song sparrows are like most songbirds in having multiple song types (“song
repertoires”). Rarely will a song sparrow have fewer than 6 or more than 12
song types, and most birds have repertoires ranging from 7 to 11 types. When
a song sparrow is “free singing” (i.e., simply posting his territory, or when
unpaired, singing to attract a mate), the bird uses his song types interchange-
ably, singing one type for a while before switching to another for a while; that
is, singing with eventual variety (in contrast to immediate variety singing,
where the bird changes song types after every song). When a song sparrow
switches song types, he generally turns to the song type he has not sung for
the longest while. Nevertheless, although he avoids recently sung types, he
does not sing his types in a particular order, so the order you observe on one
day will be randomly different from the order you observe on another day.
Like most songbirds, song sparrows learn their songs. This learning occurs
in their first year; a bird does not modify his song repertoire after the first year.
36 ANIMAL BEHAVIOR

Figure 2.1. Partial song repertoires of two song sparrows. Birds A and B were neighbors
and shared the first three songs in their nine-song repertoires (33% sharing). The
shared songs of birds A and B are shown in the top three rows, while six of their
remaining unshared types are shown in the bottom three rows. Frequency (vertical)
scale: 0-10 kHz, markers at 2-kHz intervals. Songs are 2 to 3 seconds long.

Because, in our western population, a song sparrow begins song learning after
he disperses from his natal area and typically learns his songs in the neighbor-
hood in which he will ultimately settle, his final repertoire consists of “local”
songs, and he will usually share song types with his neighbors. Thus a bird
might settle next to an older bird from whom he learned some of his songs,
or next to a bird with whom he has some “song tutors” in common; other
more complicated “lines of descent” in song learning are possible of course.
A typical result is shown in Figure 2.1, in which two neighboring song
sparrows share four of their nine songs.

Background: The Playback Experiment

The major method used in the studies discussed in this chapter—and
indeed in most studies of song communication in birds—is the playback
experiment: recordings of an animal vocalizations are played back to the sub-
ject, and its response is observed. Playback can consist of the animal’s own
FRIENDS AND ENEMIES 37

vocal signals, but most often they are another animal’s signals. Playback
experiments are essentially simulations: the experimenters attempt to simulate
a natural occurrence, but one in which they can control the location, timing,
context, delivery, and acoustical characteristics of the playback. The first play-
back experiments were carried out with songbirds, with the goal of determin-
ing if birds could distinguish between local and nonlocal dialects, or between
neighbor and stranger songs (Weeden & Falls, 1959). In the experiments we
describe in this chapter, we most often have tried to simulate a particular
neighbor, usually singing from his own territory but sometimes singing from
within the subject’s territory (a simulated intrusion). We have also simulated
strangers, either on or off the territory, and juvenile birds singing song that is
not fully developed. For playbacks from outside the territory, the simulation
is realistic because song is a long-distance signal most often sung by a bird that
is out of sight or obscured by foliage. However, for intrusions onto the territory
or on occasions when the subject gets close to the playback, the simulation is less
realistic because real birds do not both hide and sing full-throated song at the
same time. In some cases we have enhanced the simulation by using a taxider-
mic mount (a stuffed bird) of a song sparrow. Subjects in this case will direct
their attention to the mount, often giving aggressive displays—wing waves
(see Figure 2.2) and soft song (discussed further below) and often attacking it.
Although the mount does not move—clearly a robot would be ideal—it does
provide the subject with a plausible source of the song he is hearing as well as
an actual “intruder” to attack. We have discovered that we can make this

Figure 2.2. A male song sparrow wing-waving, an aggressive signal. Note color bands
on legs by which the bird is individually identified. (Photo by Çağlar Akçay)
38 ANIMAL BEHAVIOR

simulation most realistic by placing the mount in the foliage of a bush or a tree
in a natural singing position and by using a small (and thus relatively incon-
spicuous) loudspeaker placed close to the mount. When the mount is so placed,
over half of subjects will attack it within 15 minutes (Akçay et al., 2013).

Song Sparrow Neighbors Are Dear Enemies

In our western population, song sparrows establish their territories some-
time between the autumn of their natal year and the following spring. They
will usually stay on this territory for the rest of their lives, with only minor
shifts of territory boundaries or perhaps a lateral move into the vacated
territory of a deceased neighbor. The average territory tenure for a bird is
two to three years, although individuals can last for as long as seven or eight
years. Consequently birds typically have long-term neighbors. Moreover, they
typically have dear enemy relationships with these long-term neighbors. We
describe here three experiments that illustrate this situation.
One prediction of the dear enemy hypothesis is that a neighbor singing near
the territory boundary but still within his own territory should be tolerated.
However, a stranger singing from that very same spot, or a neighbor singing from
a different (“wrong”) boundary, should not be tolerated. This is because a neigh-
bor singing from a wrong territory would be a prospecting bird and a potential
threat, just as a new, arriving stranger singing from anywhere would be. Philip
Stoddard and colleagues (Stoddard et al., 1991) carried out a playback experi-
ment to test how a song sparrow would respond to neighbor versus stranger song
played from three locations: from within the neighbor’s territory near the boun-
dary, from the opposite territory boundary, and from within the subject’s
territory (simulating an intrusion). As predicted, these researchers found that
song sparrows responded weakly to a (simulated) neighbor singing near their
mutual boundary but strongly to a stranger singing from this same location. In
contrast, subjects responded strongly to both neighbor and stranger singing from
the opposite boundary (where that neighbor does not normally sing). And they
responded yet more aggressively to song from the center of their territory, equally
aggressively whether it was stranger song or neighbor song (Figure 2.3).
A second prediction of the dear enemy hypothesis is that the tolerance
shown to neighbors normally should dissipate if the neighbor violates their
agreed mutual boundary. We simulated such a boundary violation with a
two-stage playback experiment (Akçay et al., 2009). The design is illustrated
in Figure 2.4(a). In the first stage, we simulated a neighbor intruding into
the subject’s territory. The simulated intruding neighbor sang from the center
of the subject’s territory for 2 minutes. The second stage began after a wait of
45 minutes and consisted of two successive playbacks, each one simulating a
FRIENDS AND ENEMIES 39

Figure 2.3. Response during playback of 14 male song sparrows to songs of neighbors
and strangers played in three locations: the regular boundary of the neighbor that
sings the N song (where the neighbor song is normally heard), at the center of the
subject’s territory, and at the boundary opposite the boundary of the subject and his
neighbor. Response scores refer to the first component (PC1) of a principal compo-
nent analysis on number of flights and closest approach to the playback speaker. Error
bars are 2 SE. (From Stoddard et al., 1991. Reprinted by permission of Springer
Publishing.)

neighbor singing from his boundary with the subject. In one playback we
played songs of the earlier (simulated) intruder, this time from the intruder’s
own territory, just on his side of the boundary. In the other playback, we
played the songs of a neutral neighbor from just on his side of their mutual
boundary. These boundary playbacks were separated by 15 minutes, and their
order was counterbalanced across subjects. We predicted that if males keep
track of the behavior of their neighbors and retaliate specifically against those
who have violated the boundary agreement, that they should respond more
strongly to the boundary playback of the neighbor who had earlier intruded
40 ANIMAL BEHAVIOR

Figure 2.4. (a) Experimental design of Akçay et al. (2009). Each trial started with an
intrusion on the subject’s territory (1), followed by two boundary playbacks from
the intruder’s boundary and the neutral neighbor’s boundary (2 and 3; the order of
these were counterbalanced across subjects). (b) Results from Akçay et al. (2009).
Aggression scores refer to the first component (PC1) of a principal component
analysis on three variables: rate of flights, closest approach, and latency to respond.
Higher scores indicate higher aggression. Error bars are ≥1 SE. (Based on data from
Akçay et al., 2009)

than to the boundary playback of the neutral neighbor. Indeed, our results
showed that subjects responded significantly more aggressively towards neigh-
bors who recently intruded upon their territory (Figure 2.4[b]).
A further prediction of the dear enemy hypothesis is that a bird should base
his assessment of his neighbor not only on his direct experience with that
neighbor but on his observation of that neighbor’s interactions with other
neighbors. Birds with multiple neighbors have ample opportunity to eaves-
drop on the interactions of these neighbors with one another and identify
the breaking of a truce between two neighbors through eavesdropping. If a
neighbor breaks the truce with another neighbor, then he is more likely to
be an aggressor to oneself as well. We therefore carried out another playback
experiment and predicted that a song sparrow would (1) eavesdrop on a simu-
lated intrusion of neighbor A on neighbor B and (2) use this information in
calibrating his aggressiveness toward these neighbors in subsequent interac-
tions. Specifically, we predicted that subsequently he would behave more
aggressively toward the intruding neighbor singing from his boundary than
toward the victimized neighbor singing from his (Akçay et al., 2010).
Some recent theoretical work has challenged this prediction. In a model of
eavesdropping in agonistic interactions, Johnstone and Bshary (2004) found
that the presence of eavesdroppers increased rather than decreased their overall
aggression. The logic is that it pays to be more aggressive in the presence of
FRIENDS AND ENEMIES 41

eavesdroppers to discourage the eavesdroppers from challenging the aggressive

individual later on. Thus this model actually predicts that eavesdropping
neighbors should decrease their aggression towards aggressive neighbors.
To test whether neighbors increase or decrease aggression (retaliate or do
not) against aggressive neighbors, we carried out a playback experiment that
used a design similar to that of the previous experiment (Figure 2.5a). The
main difference was that the simulated intrusion was not on the subject’s
territory but on another neighbor’s territory (we term this neighbor the vic-
tim). We fitted the subjects with radio transmitters before the experiment to
be able to monitor their behavior during the intrusion on their neighbor.
We predicted subjects would show some interest in the intrusion happening
next door and approach the boundary of the victim.
After the intrusion, we tested the subject’s responses to the intruder and the
victim from their respective boundaries with the subject in the same way as in
Experiment 1 (the victim and the intruder were again randomly chosen from
among the subject’s neighbors, and the order of boundary trials were counter-
balanced). If subjects have eavesdropped on the simulated intrusion by their
neighbor, the intruder, onto the territory of the other neighbor, the victim,
and further were able to assign correctly the roles of “defecting” versus simply

Figure 2.5. (a) Experimental design of Akçay et al. (2010). Each trial started with an
intrusion (1) on the victim’s territory, which was adjacent to the subject’s territory.
The boundary playbacks from the victim and the intruder (2 and 3, order counter-
balanced across subjects) were directed to the subject, who did not receive the
intrusion himself but had eavesdropped on it. (b) Results from Akçay et al. (2010).
Aggression scores refer to the first component (PC1) of a principal component
analysis on three variables: rate of flights, closest approach, and latency to respond.
Higher scores indicate higher aggression. Error bars are ≥1 SE. (Based on data from
Akçay et al., 2010)
42 ANIMAL BEHAVIOR

defending one’s own territory, they should retaliate against (respond more
strongly to) the intruding neighbor.
Our results showed that subjects did in fact increase aggression towards
neighbors who had recently intruded on another neighbor’s territory (Figure
2.5b). Furthermore, 8 out of 10 subjects approached the victim’s boundary
during the intrusion, indicating that they were indeed interested in the
intrusion. This is the first evidence for eavesdropping in a territorial system
with repeated interactions between familiar individuals. In other words, an
aggressive individual will acquire a bad reputation in the eye of eavesdropping
neighbors and is likely to face retaliation from these neighbors. We identify
this threat of retaliation by eavesdropping neighbors as an additional factor
promoting dear enemy cooperation between neighbors, contrary to the sug-
gestions of models of eavesdropping in agonistic contexts (Johnstone, 2001;
Johnstone & Bshary, 2004).

HOW DO SONG SPARROWS COMMUNICATE?

The above studies demonstrate that song sparrows recognize their neigh-
bors and keep tabs on them through their direct experience with them as well
as by eavesdropping on interactions between neighbors. Now we turn to the
question of how song sparrows communicate with their neighbors through
their songs.
For song sparrows, song is a long-distance signal intended for adjacent or
once-removed territorial neighbors. Typically the birds that have a singing
interaction are out of sight of one another. This is the prototypical case for
songbirds and explains why such small birds produce such loud sounds. Most
of the rest of their vocal repertoire is much softer. At the heart of the song rep-
ertoire is the posting function—the bird is both notifying his neighbor that he
is still there occupying his territory and marking the extent of his territory.
Songbirds have no fences, and territorial boundaries need to be constantly
reaffirmed. Indeed, they may need to be renegotiated. For example, a male’s
mate may decide that the perfect spot for her nest is just over what has been
the boundary, or a neighbor may have been picked off by a hawk, leaving
some prime territory up for grabs between several of his neighbors. Song is
critical to boundary negotiations, for it provides a nonviolent way of “negoti-
ating” these disputes.
In our population of song sparrows, birds appear to follow a set of conven-
tions predicated on how they use the songs they happen to share (Figure 2.6).
As an example, consider the two neighbors whose repertoires are shown in
Figure 2.1. Suppose Bird 1 decides that their mutual boundary should be a
FRIENDS AND ENEMIES 43

Figure 2.6. Cartoon illustrating how song sparrow neighbors communicate using
shared songs. A bird typically addresses a neighbor by singing one of the songs they
share. The neighbor can reply by singing the same song type (type matching), singing
a different song type that he shares with that neighbor (repertoire matching), singing a
song type he does not share with that neighbor, or not singing at all. Type matching
appears to be an escalation and singing an unshared song a deescalation, while reper-
toire matching is an intermediate but directed signal.

few meters into Bird 2’s territory, and he aims to establish this new boundary.
They share four of these songs—we will call them types A, B, C, and D—but
no others. Bird 1 can “address” bird 2 by singing one of their shared types A,
B, C, or D (in the direction of bird 2 because other neighbors may also share
some of these types). Let us say bird 1 sings A. Bird 2 then can “acknowledge”
the signal by replying with B, C, or D (repertoire matches), reply with A (type
match), sing one of the unshared types, or ignore by not singing at all. If bird
2 type matches bird 1 (sings A), bird 2 can then continue to sing that song
type (“stay on type”), can switch to another shared song (repertoire matching;
Beecher et al., 1996), switch to an unshared type, or disengage totally by stop-
ping singing. Note that to type match, the bird need have no prior experience
with his opponent—the bird simply replies with his most similar song, and
generally the match’ will be perceptually obvious. But to reply with a shared
song or with an unshared song, the bird needs to have had some experience
with his neighbor—the bird needs to know which songs the two share and
which they do not. This pattern of singing does not normally develop until
birds have been neighbors for some period of time.
These “singing conventions” are summarized in Figure 2.7. It turns out
that each convention—type matching, repertoire matching, and switching
Figure 2.7. Schematic diagram of an interaction between two neighboring song spar-
rows. Birds A and B interact during six time periods (T1-T6), each of which con-
tains a change of state. Aggressive escalations are indicated with upward (solid line)
arrows, and deescalations with downward (dotted line) arrows. For simplicity the
diagram focuses on escalations. The interaction begins when the two birds are at a
distance, on their own territories, and probably out of visual contact, but as the
interaction escalates, the birds approach one another and are presumed to be within a
few meters of one another to the right of the dotted line. The signal and response
alternatives of each bird are indicated vertically, with more aggressive alternatives
above less aggressive ones. The song sparrow signaling system depends on the fact that
two territorial neighbors will share some of their 7 to 10 song types and not others. At
T1, bird A engages his neighbor by singing a shared song. At T2, the neighbor (bird
B) can escalate by replying to that shared song with a type match (the same song type),
deescalate by singing an unshared song, or reply at an intermediate level with a rep-
ertoire match (a different song type they share). If bird B repertoire matches (the most
common reply for established neighbors), then at T3, bird A can escalate by type
matching or deescalate by singing a different song type (either an unshared song type
or a different shared type—i.e., another repertoire match). If bird B instead type
matches at T2, then at T3, bird A can either maintain escalation by staying with the
type match or deescalate by switching song types. At T4, the birds should be close,
and bird B can escalate by switching to soft song. Bird A can meet the escalation by
switching to soft song, or he can retreat (T5). Soft song is the strongest aggressive
signal in the hierarchy, and at this point (T6), soft song is most likely followed by
attack. In this system, the type matching is a poor predictor of attack because several
layers of contingency intervene between B’s initial decision on whether to type match
and his final decision on whether to attack. Type matching can still be viewed as an
aggressive signal, however, because it does predict escalation in the interaction. (This
model is based primarily on Beecher et al., 1996; Beecher, Campbell, Burt, et al.,
2000; Burt et al., 2001; Beecher & Campbell, 2005; and Akçay et al., 2013.)
FRIENDS AND ENEMIES 45

off—has a distinct signaling function in a graded signaling system, with type

matching signaling a willingness to escalate, repertoire matching a willingness
to continue the interaction but not necessarily escalating, and switching off
signaling deescalation. We now turn to the evidence decoding this code.

The Song Sparrow Code

In the first study contrasting the different functions of singing conventions,
we used a prediction that follows from the dear enemy relationship between
neighbors. As mentioned above, early in the spring most neighbors are in the
process of negotiating their boundaries, inevitably involving more aggressive
interactions. Later in the spring, however, most boundaries have been estab-
lished and are stable, and neighbors are in the dear enemy relationship of
watchful tolerance of each other. If type matching is a signal of willingness
to escalate and repertoire matching a willingness to acknowledge the opponent
but not necessary escalate, we predicted that type matching in response to a
neighbor should be much more common early in the spring and repertoire
matching much more common later in the spring. In line with the hypothesis,
early in the spring (mid-April) most subjects (8 out of 11, 73%) replied to a
neighbor’s shared song played from that neighbor’s boundary with a type
match, the rest replying with a repertoire match. In contrast, late in the spring
(late May to early June), only 18 percent of subjects responded to their neigh-
bor’s shared song with a type match, the rest replying with a repertoire match
(Beecher, Campbell, Burt, et al., 2000), again in line with the hypothesis that
type matching is an escalation signal whereas repertoire matching is not.
Further evidence for the signaling system in Figure 2.7 came from an inter-
active playback experiment that our group carried out in 1997 and 1998 (Burt
et al., 2001). In this experiment, we wanted to more directly test the hypoth-
esis that type matching is a signal of willingness to escalate (i.e., a threat signal)
by assessing how birds responded when they were type matched versus reper-
toire matched. To that end, we set up the playback equipment at the boun-
dary of the subject whose singing we were going to simulate and waited until
the subject sang a song that was shared with this particular neighbor (not a
task for the impatient!). Then, when the subject sang a shared song (we were
able to view the sonograms of the songs that the subject sang in real time on
a laptop), we “replied” with the neighbor’s version of the song (type match),
another shared song (repertoire match), or a stranger song (song recorded from
a bird at least a kilometer away) and noted several measures of approach and
aggression. The prediction was that when subjects were type matched by their
simulated neighbor, they would respond as if they had just been threatened,
specifically, more aggressively than if the simulated neighbor had repertoire
46 ANIMAL BEHAVIOR

matched them. The stranger-playback condition was meant to provide a refer-

ence for comparison of the level of aggression to the neighbor (subjects should
respond most strongly to the stranger).
The results were clear—as expected, subjects were most aggressive when
they were countered with a stranger playback. Crucially, subjects were signifi-
cantly more aggressive in response to being type matched compared to being
repertoire matched by the simulated neighbor (Burt et al., 2001). Further-
more, among the trials where the subjects were type matched, the most aggres-
sive birds were those who stayed on the same type upon being type matched.
This suggests that not only is type matching a signal of escalation, so is staying
on the same type when you are being type matched by the opponent. Thus,
these two experiments established that type matching and repertoire matching
signal different levels of threat from the signaler. In a more recent study
(Beecher & Campbell, 2005), we tested the function of singing an unshared
song using a design similar to the above study and found that following a
simulated intrusion and retreat by a neighbor, subjects reduced their aggressive
response and departed the scene sooner if the retreating neighbor sang an
unshared song than if he sang a shared song. We also found that birds
responded more quickly when a neighbor singing from their territory boun-
dary sang a song the two shared than when he sang a song they did not share.
All in all, these studies make a strong case that song sparrow males perceive
type-matching songs, repertoire-matching songs, and unshared songs as part
of a graded signaling system with type matches being the highest threat,
unshared songs the least threatening, and repertoire matches intermediate in
threat.

The Code Reconsidered

We thought we had cracked the song sparrow code when a new study cast
doubt on the idea that type matching was a reliable threat signal. In this study
of an eastern U.S. song sparrow population (Pennsylvania), Searcy and col-
leagues (Searcy et al., 2006) challenged song sparrows with a simulated
intrusion by playing the bird his own song (“self-song”) from the middle of
his territory. They measured a host of responses, including several putative
aggressive signals such as wing waves and soft songs, as well as whether or
not the bird type matched the playback (using self-song removes any doubt
about whether the subject’s reply song is the same type or not). Then, after
five minutes of playback, they revealed a taxidermic mount of a song sparrow
that was attached to the speaker, giving the angry subject a target to attack.
The aim was to assess how reliable each signaling behavior was in predicting
a subsequent attack that seemed likely to await the hapless mount. To their
FRIENDS AND ENEMIES 47

surprise, only about 20 percent of subjects actually attacked the mount, and
only one signaling behavior, soft songs (low-amplitude songs that song spar-
rows tend to sing from close range), was significantly associated with a sub-
sequent attack. None of the other signaling behaviors, including type
matching, was associated with whether or not the subjects attacked the mount
subsequently (Searcy et al., 2006). If type matching does not predict success-
fully that an attack is impending, how can it be a threat signal?
There are at least two possible answers. First, this case may simply be
another one of the differences between eastern and western song sparrow pop-
ulations that we alluded to earlier. Some populations of eastern song sparrows,
including the Pennsylvania population that Searcy and colleagues studied,
show very limited song sharing (Hughes et al., 1998, 2007), and it is possible
that the signaling code of Washington song sparrows does not apply to Penn-
sylvania song sparrows where song sharing between neighbors is uncommon
—hence type matching and repertoire matching would not be possible most
of the time. However, a subsequent playback experiment we carried out in
our Washington population also failed to confirm that type matching was a
reliable predictor of an impending attack (Akçay et al., 2011). Thus, although
population differences may be part of the answer, they do not seem to be the
whole answer.
The second possible answer is that type matching, while a threat signal and
thus predicting escalation, is a long-distance, low-level signal that is given early
in a sequence of aggressive interaction, well before an actual physical fight. If it
were true that type matching is a low-level threat signal, it generally would not
be an appropriate signal when the intruder is already in the middle of the
territory, having boldly gone where he really should not have. In other words,
in the face of such a high-intensity threat, the subject should not be expected
to use a conventional signal that generally functions as a long-distance, early-
stage signal across territory boundaries.
We therefore carried out a two-stage experiment to assess the reliability of
type matching early in the interaction in predicting higher-level threats and
eventual attack. In such an experiment, as proposed first by Beecher, Camp-
bell, Burt, and colleagues (2000), the playback of a matchable song would
start at the boundary in the first stage, eliciting a type match or a nonmatch.
At the second stage the playback would be moved immediately to inside the
territory of the subject, a clear escalation. Adding to the mix a taxidermic
mount that is coupled with the playback at the center would also allow the
subjects to attack the intruder. With this two-stage design, we can then ask
whether type matching early in the trial will result in higher levels of aggressive
signaling and attack.
48 ANIMAL BEHAVIOR

Our prediction was that type matchers at the boundary would be more
aggressive at the center, giving more soft songs and wing waves (fluttering
one wing at a time such that the bird does not actually take off from the
branch) and attacking at a higher rate. This prediction, however, had to be
more refined given what we already knew about type matching. First, we knew
from previous experiments that birds sometimes switch off type in response to
being type matched (Burt et al., 2001). This could happen for a variety of rea-
sons, including the unwillingness of the bird to engage in a fight just then
because of conflicting activities such as feeding and mate guarding or because
of assessing the opponent as a higher-quality male than himself. Whatever the
cause, if a bird initially type matches but switches off type immediately after
the playback moves to center, he is predicted to not show high levels of aggres-
sion. On the flip side, a bird might fail to type match the playback at the
boundary for reasons other than unwillingness to escalate—the bird might
perceive the boundary playback as simply not a serious enough threat, or he
might be too busy with other activities (such as feeding young). However,
we predicted that these birds would switch to a type match once the threat
became unambiguous, that is, when the playback moved to the center. Taking
into account these possibilities we predicted that birds who (1) type matched
at the boundary and stayed on type when the playback moved to center or
(2) did not type match at the boundary but type matched as soon as the play-
back moved to the center would escalate to high-intensity threat signals and
eventually attack. We called these birds type matchers. In contrast, we pre-
dicted that birds who type matched at the boundary but switched off type
once the playback moved to the center (“switchers”), that is, who deescalated
by singing conventions, would behave less aggressively. Finally, birds that
did not type match the playback (“nonmatchers”) either at the boundary or
at the center were also predicted to show lower levels of aggression. These pre-
dictions are summarized in Figure 2.8.
The results were largely in line with our predictions. As predicted, we found
that type matchers were more aggressive than either the switchers or the non-
matchers, and all but two of them (15 out of 17, 88.2%) ended up attacking
the mount. In contrast, attack frequencies for the switchers and nonmatchers
were 50 percent and 52.2 percent, respectively, a significant difference. This
was the first clear evidence that type matching as an early threat signal is pre-
dictive of an attack later on.
There were, however, some notable exceptions. Mentioned above was the
fact that exactly half of switchers (four out of eight) also ended up attacking
the mount. We predicted that switchers would only include birds that de-
escalated after an initial threat; that is, they would be “bluffers” in the sense
FRIENDS AND ENEMIES 49

Figure 2.8. The hierarchical signaling hypothesis in the context of the experiment of
Akçay et al. (2013). The design has two phases. In the first phase, song is played
from the subject’s territory boundary. In the second (escalation) phase, the playback
switches to the center of the territory and the mount is revealed. When the bird hears
a song from his territory boundary, he may or may not consider it seriously threat-
ening. If he does consider it threatening, according to the hierarchical signaling
hypothesis he should type match and should continue to type match once the esca-
lation occurs (breaking off the type match is considered a deescalation). If the bird
does not consider song from the boundary threatening enough to warrant a type
match, by hypothesis he will certainly consider the escalation into his territory
threatening, and he should type match at this point, that is, soon after the escalation.
The common denominator of these two patterns is that the bird should type match in
the short period following escalation. The results are consistent with this prediction
with the interesting exception of “undersignalers”—birds who attack with little or no
signaling (discussed in text).

that they would threaten but not back up their threat with action once the
opponent escalates. We were wrong on this count—half the birds we consid-
ered switchers were actually fast-escalating birds who, after type matching at
the boundary, switched off type but directly into soft song and wing waves
and ended up attacking the mount. All in all, only 4 birds out of a total of
48 subjects (8.3%) could be considered bluffers, that is, birds that type
matched initially but failed to back up the threat signal with action
subsequently.
Another notable exception was that about half of the nonmatchers (12 out
of 23 birds) also attacked the mount. This result suggests that sometimes song
sparrows just forgo the early threat signals and go for an attack. Indeed, there
were even some birds who attacked without a single soft song or wing wave,
both signals of aggression we found to be independently predictive of attack.
In other words, some birds seemed to forgo aggressive signaling, opting for a
strategy of undersignaling. The presence of these “strong, silent types” is
50 ANIMAL BEHAVIOR

puzzling all the more because the signals in question—type matching, soft
song, and wing waves—are all cheap to produce. Nonetheless, it seems clear
that a good portion of birds that are otherwise very aggressive and attack the
mount in this experiment were not engaged in extensive signaling. Under-
signaling in these cases would decrease the reliability of the signal in question.
But here, the problem is not so much deceptive signaling (i.e., signaling
threat without the intention of backing it up with action), but the reverse
problem of not giving threat signals when the aggressive intent in fact is there.
One possibility is that these birds rely mostly on other behaviors to warn the
opponent of an impending attack, such as a direct and quick flight towards
it, flying around the opponent intently, and so on. These behaviors would
not be considered signals per se because they did not evolve for a signal func-
tion (Otte, 1974) but nevertheless could provide information to the receiver
regarding the likelihood of an impending attack.
In summary, these results indicate that type matching is an early threat sig-
nal that would be followed with higher-level threat signals such as soft songs
and wing waves. It would therefore seem to be beneficial for a bird to be able
to type match his opponents. To do that, however, the birds need to share at
least one song type, and, presumably, the more songs a bird shares with likely
neighbors the better. This brings us to the other and equally important aspect
of song that we studied: song learning.

THE RELATIONSHIP OF THE SONG COMMUNICATION SYSTEM TO THE PROCESS

OF SONG LEARNING
As reviewed above, song is the mechanism by which male song sparrows
communicate with their male neighbors. Song is important as well in commu-
nication with females, but our studies have focused mostly on the male-male
side of the equation (for perspective on the female side, see O’Loghlen &
Beecher, 1997, 1999). In this final section, we consider the mutual fit of
two processes: (1) the process we have been describing—how song is used
between territorial song sparrows in our population, and (2) the process of
song learning in these birds. We focus on the mutual fit because at present
we cannot really determine the direction of causality in this relationship. That
is we cannot say whether shared songs are the key to this communication sys-
tem because song sharing is an inevitable consequence of learning the songs of
your neighbors, or whether the process of song learning has been shaped by
natural selection so that song sparrows learn songs they will share with their
ultimate neighbors. More generally, it is remarkable that the function of song
learning, a fundamental and defining characteristic of the more than 4,000
FRIENDS AND ENEMIES 51

species of oscine passerines (songbirds), still remains largely a mystery after

years of intensive study (Beecher & Brenowitz, 2005). Nevertheless, in this
section we try to identify the possible advantages male song sparrows in our
population may derive from their ability to learn their songs.

What We Know about Song Learning in Western Song Sparrows

We have carried a variety of field studies and lab studies on song learning in
our population of song sparrows (Beecher, Campbell, & Stoddard, 1994;
Beecher et al., 2007; Nordby et al., 1999, 2000, 2001, 2002, 2007; Burt et
al., 2007; Nulty et al., 2010; Templeton et al., 2009; Templeton, Campbell,
et al. 2012;Templeton, Reed, et al. 2012; Akçay, Campbell, et al., in prepara-
tion). We summarize these results here as “rules of song learning” for this
population.
Rule 1—Copy songs of conspecific singers. In nature, song sparrows copy only
song sparrow song in the field, although the occasional copy of a song or song
element of a Bewick’s wren (Thryomanes bewickii; personal observation) or
white-crowned sparrow (Zonotrichia leucophrys; Baptista, 1988) are reported.
In the lab, song sparrows readily copy swamp sparrow (Melospiza georgiana)
song (Marler & Peters, 1988), so it would appear their failure to copy hetero-
specific song in the field (except for on the odd occasion) implies a mechanism
for selecting conspecific models.
Rule 2—Complete song learning by the first spring. Because adult males
(potential song tutors) in our population typically remain on their territories
from one year to the next unless they die in the interim, it is difficult for us
to determine from field data exactly when the young bird memorized his
songs. But combining our lab studies with our field studies, we conclude that
young song sparrows usually memorize their songs in the several months fol-
lowing dispersal from the nest (May, June, and July being the prime months
for most birds), that is, in the traditional lab-determined sensitive period,
roughly the second and third months of life (Marler & Peters, 1987). How-
ever, our lab studies indicate that some birds will memorize new songs they
hear for the first time the following spring, possibly because they did not
memorize enough songs the previous summer (Nordby et al., 2001; Nulty et
al., 2010; Templeton, Burt, et al., 2012), a pattern that has been demon-
strated as well in marsh wrens (Cistothorus palustris; Kroodsma & Pickert,
1980). Although most memorization occurs in the natal summer, extensive
modification of the song repertoire continues into the following spring—com-
parison of a bird’s song repertoire in early spring (e.g., January) with his final
repertoire in March or April typically reveals that he has modified his song rep-
ertoire by dropping songs, combining songs, and rearranging elements in
52 ANIMAL BEHAVIOR

songs. However, birds do not change their song repertoires after their first
breeding season (Nordby et al., 2002).
Rule 3—Copy song types completely and precisely. Young song sparrows often
develop near-perfect copies of the songs of their older neighbors. It is this fact
that first made us realize that we could trace song learning in the field (when
we began our field studies, almost all studies of song learning had been done
in the laboratory). The song similarities can be striking, with the differences
between tutor and student often being no greater than one normally sees in
repetitions of the same song sung by one bird. In contrast, laboratory studies
by Peter Marler and Susan Peters (1987, 1988) using recorded songs as the
tutor stimuli found that while song sparrows copy song elements quite pre-
cisely, they frequently combine elements from different songs to form
“hybrid” song types—songs made up of parts of different song types. That
is, they often copy song elements but use them to improvise new song types.
We are still not sure to what extent this difference reflects differences in
laboratory and field song-learning conditions (which of course are substantial),
as opposed to differences in the eastern and western populations of song spar-
rows. The substantial differences in patterns of song learning shown by differ-
ent songbird species, and in some cases by different populations of the same
species (Brenowitz & Beecher, 2005), indicate that population differences
cannot be dismissed as a possible explanation here. Indeed we have argued
elsewhere that only careful common garden experiments can support or elimi-
nate this hypothesis (Beecher, 2008).
Rule 4—Learn the songs of multiple birds. Usually it takes three to five song
tutors to account for the young bird’s entire repertoire of eight or nine song types.
For example, in Cully Nordby and colleagues’ study (Nordby et al., 1999), only 1
of the 41 subjects appeared to be a song “clone” of a single older bird. In the study
by Çağlar Akçay and colleagues (Akçay, Campbell, et al., in preparation), only 1
of 43 subjects was a song clone. Nevertheless, about half the time the young bird
learns the majority of his songs from one “primary” tutor (see Rule 7).
Rule 5—Learn from your neighbors. A bird’s song tutors almost always turn
out to have been neighbors in the young bird’s hatching summer, and, if they
survived the winter, the following spring (the young bird’s first breeding sea-
son) as well. The young bird usually establishes his territory within the
territorial range of his song tutors, often replacing a tutor that died. In the
cases where the young bird does not establish his territory among his tutor-
neighbors, the evidence suggests that he did not because he could not—
because none of his tutors had died or because other young birds moved into
this area. An example is shown in Figure 2.9 (note that it contains a larger-
than-normal level of overwinter attrition of adult tutors).
FRIENDS AND ENEMIES 53

Figure 2.9. Diagram indicating song learning and territory establishment. (a) Territo-
ries of AIRM’s tutors in his natal summer (1992). Adult birds (potential tutors) are
shown by their color bands (four-letter codes) and their territories by dotted lines. The
identified tutors of AIRM and their territories are shown by the dark hatching. (b)
Territory of AIRM the following spring (1993), overlaid on the territories of summer
1992. Adult males who did not survive the winter are crossed out. Of the 13 adult
birds shown, 8 out of 13 did not survive the winter; four out of AIRM’s five tutors did
not survive the winter. (This is an unusually high mortality rate—overwinter survival
is typically 60 to 70%.) Note that AIRM established his territory in an area overlap-
ping the former territories of three out of the four deceased tutors and next to his one
surviving tutor (OGGM). The young bird shared songs with the surviving tutor and
with other young birds that moved into that area as they had similarly learned songs of
the area. (From Nordby et al., 1999. Reprinted by permission of Oxford University
Press.)

Rule 6—Preferentially learn or retain song types of tutors surviving to your first
breeding season. Birds often have song types that can be traced to tutors that
were alive in the young bird’s natal summer but died before the next breeding
season. Nevertheless, they generally retain more songs of tutors who survive
into the next breeding season than of tutors who do not. We refer to this late
learning as late influence because it may not be de novo learning—these songs
could have been memorized in the natal summer and retained because the bird
continues to hear them the following autumn or spring. This would be the
pattern hypothesized as typical by Douglas Nelson and Marler (Nelson,
1992; Nelson & Marler, 1994)—the young bird memorizes songs during a
sensitive period in the natal summer and the following spring, retains some
of these songs, and drops others on the basis of his social interactions with
his territorial neighbors (“selective attrition”). We have recently compared
the song repertoires of young song sparrows in the plastic song phase (late
winter, early spring) and crystallized song phases and found that they do
54 ANIMAL BEHAVIOR

indeed retain songs that are more similar to those of their springtime territorial
neighbors, while dropping some songs that are less similar (Nordby et al.,
2007).
Rule 7—Pick a best tutor. Although the birds learn from multiple tutors
(Rule 4 above), not all tutors contribute to the final repertoire of the bird
equally. In rare cases, as mentioned above, the tutee’s repertoire is a complete
clone of one particular older bird. In other cases, the best tutor contributes
only slightly more than the rest of the tutors and the repertoire is more or less
made up evenly of songs from different tutors. But in most cases, the reper-
toire of the bird is skewed towards one particular tutor, who often is a neigh-
bor. Nordby and colleagues (1999) found that on average half of the songs
in a tutee’s repertoire comes from the best, or primary, tutor. The best tutor
is usually either a direct or once-removed neighbor that has survived into the
first spring of the tutee.
Rule 8—Preferentially learn tutor-shared songs. As noted earlier, in our field
population neighbors typically share a portion of their song repertoires, on
average about two to four of their eight or nine song types. We have found
that the young bird preferentially learns (or retains) song types shared by two
or more of his tutors (Beecher, Campbell, & Stoddard, 1994; Nordby et al.,
2000, 2001). There are several possible reasons that shared song types might
be particularly salient, including (1) these types are heard more than unshared
song types, (2) the “same song” is being sung by several birds, and (3) they are
heard more often in countersinging interactions than are unshared songs. We
consider the last possibility further below.
One possible function of the learning preference for tutor-shared songs is
that it represents a bet-hedging strategy to guarantee that the young bird has
song types he will share with his neighbors in his first breeding season. If
instead the bird learned tutor-unique songs, he would have songs specialized
for these particular tutor-neighbors (i.e., would share these songs with one
neighbor only). But these specialized songs would be good only until the tutor
died or moved, whereas a shared song would be good until all the birds having
it in the neighborhood died or moved, and probably even longer because other
young birds moving into the area would also preferentially learn shared types.
Rule 9—Individualize your song repertoire. The rules so far can be inter-
preted to fit the following overall rule: Learn songs that you will share with your
neighbors in your first breeding season. There is, however, an important excep-
tion to that rule (Nordby et al., 2007). In the transition from plastic song to
final crystallized song, the young bird often modifies a song so that it actually
becomes a poorer match to the model song of the putative tutor and to similar
songs of his present neighbors (who may or may not include the tutor).
FRIENDS AND ENEMIES 55

We interpret this as the bird’s “individualizing” his song, much like a pop
singer putting his or her own personal touch on an old standard. The song
may still be perceived by the birds as a shared song (even if it perhaps no
longer meets our criteria for a shared song), while at the same time being per-
ceived as his particular version of that song type. Thus the bird gets to have
songs that are both shared with his neighbors and yet unique to him. Our
many playback experiments and lab perception experiments (e.g., Stoddard
et al., 1992; Horning et al., 1993; Beecher, Campbell, & Burt, 1994) have
indicated that birds can discriminate among rather different versions of one
particular song type while recognizing their similarity. Thus individualizing a
song, as song sparrows often do, may simply stamp the song as the bird’s
own (strengthening individual distinctiveness) while maintaining his ability
to use it as a shared song in the communication system.
Rule 10—Social interaction is crucial for song learning. There is ample evi-
dence from our lab studies of song learning and from many others as well that
song learning is enhanced if the young bird experiences social stimulation
from live singing birds; indeed these “live tutors” are generally much more
effective than simple playback of recorded song (see review in Beecher,
2008). In several of our lab studies with live tutors, we found that social inter-
action was critical but that the tutee did not necessarily need to directly inter-
act with the tutor or even see him (Nordby et al., 2000, 2001; Beecher et al.,
2007). In these experiments birds often learned as well or even better from
tutors they could hear but not see and from tutors who interacted not with
them but with some other bird. That song learning might require a social
interaction context without necessarily involving direct interaction between
tutee and tutor was confirmed in a field experiment—we found that young song
sparrows in the song-learning phase found playback simulating an interaction
between two adult singers much more attractive than playback simulating a solo
singer (Templeton et al., 2009). In further field experiments, we found that
adult song sparrows are surprisingly tolerant of young, recently fledged birds,
who would often sit close to the adult bird while it interacted with a simulated
intruder (mount plus playback; Templeton, Campbell, et al., 2012).

QUESTIONS THAT REMAIN

As the review above makes it clear, we have very detailed information on
the pattern of song learning in song sparrow and the use of song in interactions
between adult males. Very interesting questions still remain, however. The
most important of these is the role of social interactions in tutor choice.
Although we have very good evidence from lab studies that social interactions
56 ANIMAL BEHAVIOR

between the tutee and tutors as well as between potential tutors matter, the
nature of the interactions that can be studied in the lab is very narrow and
inherently unnatural. For instance, in the field, a young bird can simply
move off to get away from an escalating situation, whereas in the lab he is
stuck in a cage. Therefore, the nature of social interactions is best examined
in the field.
One question we are currently pursuing is the question of why song spar-
rows choose one particular male as their primary tutor and why they choose
this particular individual rather than another. In other words, why is the rep-
ertoire of the young song sparrows so highly skewed towards a single male,
and what makes that male special?
One hypothesis is that birds skew their repertoire towards the primary tutor
because it helps them negotiate their boundaries with this particular male.
Establishing a territory is a costly affair, which usually takes persistence on
the part of the challenger (Arcese, 1987, 1989), and once a bird establishes a
territory the boundaries still need to be negotiated. In principle, any negotia-
tions with a neighbor may be costly, but some neighbors might pose a particu-
larly strong challenge because they are more aggressive. Indeed, there is
evidence that in song sparrows (as in other animals) there are individual differ-
ences in aggression that are stable over time (Hyman et al., 2004; Akçay et al.,
2013), and neighbors respond more strongly towards neighbors with more
aggressive personalities (Hyman & Hughes, 2006).
Aggression may have several effects on song learning. Most simply, a young
bird establishing his territory may have to engage in more interactions with an
aggressive neighbor and as a result may learn his songs simply as a by-product.
Alternatively, learning the songs from an aggressive neighbor may be beneficial
for the young bird. Recall that song sparrows possess a complex long-range sig-
naling system that relies on using shared songs in aggressive interactions before
these escalate into close-range interactions that may involve physical fights. If
it is true that some neighbors are more aggressive than others, it may be ben-
eficial for a young bird to optimize his repertoire for use against these neigh-
bors by learning his songs primarily from these males. Under this hypothesis,
aggressive tutors are predicted to be better tutors.
A second, alternative hypothesis is that birds choose their tutors not based
on their aggressiveness but rather on the basis of their dear enemy potential
(Akçay et al., 2009, 2010). Under this hypothesis, tutees and primary tutors
are setting up an (at least partially) cooperative relationship that may benefit
each of them, and sharing songs facilitates their coordination. In other words,
tutees learn their songs not from their “nasty neighbors” who are particularly
aggressive towards them but from neighbors with whom they are setting up
FRIENDS AND ENEMIES 57

a dear enemy relationship. This hypothesis might be termed the Mr. Chips
hypothesis after the beloved schoolteacher in the movie Goodbye, Mr. Chips.
A cooperative relationship between a tutor and a tutee should entail the
tutor and tutee showing reduced aggression towards each other and perhaps
even engaging in cooperative territory defense. We have found that song shar-
ing in our population is correlated positively with territory tenure in song spar-
rows (Beecher, Campbell, & Nordby, 2000), suggesting that there is indeed
potential for a cooperative relationship between a tutee and his primary tutor.
To the best of our knowledge, however, no one has tested this hypothesis in
song sparrows or any other songbird.
We are currently in the process of testing these two hypotheses regarding
the role of social factors in tutor choice. Our preliminary results suggest that
the aggression hypothesis is not supported in song sparrows. In a field study
where we quantified stable individual differences in aggression (i.e., “aggressive
personality”) of almost half of the potential tutors in our study site, we found
no effect of tutors’ aggression on whether or not they were selected as tutees by
the young males establishing their territories that year (Akçay, Campbell, et
al., in preparation). This result suggests that aggression does not play a signifi-
cant role in tutor choice of song sparrows. We are also planning experiments
to test the Mr. Chips hypothesis.

CONCLUSION
Song sparrows have proved themselves a fruitful model system for investi-
gating central questions in animal communication: those of reliability, social
dynamics, and development. In this chapter we have tried to integrate the
various lines of research from our laboratory on this fascinating system. We
have gained many insights into the function and development of song
repertoires in song sparrows and see many more fruitful avenues of research
opening up.

ACKNOWLEDGMENTS
This research has been supported by the National Science Foundation. We
give special thanks to Discovery Park for hosting our field research.

REFERENCES AND SUGGESTED READING

Akçay, Ç., S. E. Campbell, V. A. Reed, M. E. Tom, & M. D. Beecher. (In prepara-
tion). Aggression and tutor choice in song learning in the song sparrow.
58 ANIMAL BEHAVIOR

Akçay, Ç., V. A. Reed, S. E. Campbell, C. N. Templeton, & M. D. Beecher. (2010).

Indirect reciprocity: Song sparrows distrust aggressive neighbors based on
eavesdropping. Animal Behaviour, 80, 1041–1047.
Akçay, Ç., M. E. Tom, D. Holmes, S. E. Campbell, & M. D. Beecher. (2011). Sing
softly and carry a big stick: Signals of aggressive intent in song sparrows.
Animal Behaviour, 82, 377–382.
Akçay, Ç., M. E. Tom, S. E. Campbell, & M. D. Beecher. (2013). Song type match-
ing is an honest early threat signal in a hierarchical animal communication
system. Proceedings of the Royal Society of London, B, 280, 20122517.
Akçay, Ç., W. E. Wood, W. A. Searcy, C. N. Templeton, S. E. Campbell, & M. D.
Beecher . (2009). Good neighbour, bad neighbour: Song sparrows retaliate
against aggressive rivals. Animal Behaviour, 78, 97–102.
Arcese, P. (1987). Age, intrusion pressure and defence against floaters by territorial
male song sparrows. Animal Behaviour, 35, 773–784.
Arcese, P. (1989). Territory acquisition and loss in male song sparrows. Animal
Behaviour, 37, 45–55.
Arcese, P., M. K. Sogge, A. B. Marr, & M. A. Patten. (2002). Song sparrow (Melo-
spiza melodia). In A. Poole & F. Gill (eds.), The Birds of North America, Issue
704. Ithaca, NY: Cornell Lab of Ornithology.
Baptista, L. F. (1988). Imitations of white-crowned sparrow songs by a song sparrow.
Condor, 90, 486–489.
Beecher, M. D. (2008). Function and mechanisms of song learning in song sparrows.
Advances in the Study of Behavior, 38, 167–225.
Beecher, M. D. & E. A. Brenowitz. (2005). Functional aspects of song learning in
birds. Trends in Ecology and Evolution, 20, 143–149.
Beecher, M. D., J. M. Burt, A. L. O’Loghlen, C. N. Templeton, & S. E. Campbell.
(2007). Bird song learning in an eavesdropping context. Animal Behaviour,
73,929–935.
Beecher, M. D. & S. E. Campbell. (2005). The role of unshared songs in singing
interactions between neighbouring song sparrows. Animal Behaviour, 70,
1297–1304.
Beecher, M. D., S. E. Campbell, & J. M. Burt. (1994). Song perception in the song
sparrow: birds classify by song type but not by singer. Animal Behaviour, 47,
1343-1351.
Beecher, M. D., S. E. Campbell, J. M. Burt, C. E. Hill, & J. C. Nordby. (2000).
Song type matching between neighboring song sparrows. Animal Behaviour,
59, 21–27.
Beecher, M. D., S. E. Campbell, & J. C. Nordby. (2000). Territory tenure in song
sparrows is related to song sharing with neighbors, but not to repertoire size.
Animal Behaviour, 59, 29–37.
Beecher, M. D., S. E. Campbell, & P. K. Stoddard. (1994). Correlation of song learn-
ing and territory establishment strategies in the song sparrow. Proceedings of the
National Academy of Sciences, USA, 91, 1450–1454.
FRIENDS AND ENEMIES 59

Beecher, M. D., P. K. Stoddard, S. E. Campbell, & C. L. Horning. (1996). Repertoire

matching between neighbouring song sparrows. Animal Behaviour, 51, 917–923.
Brenowitz, E. A. & M. D. Beecher. (2005). Song learning in birds: Diversity and
plasticity, opportunities and challenges. Trends in Neuroscience, 28, 127–132.
Burt, J. M., S. E. Campbell, & M. D. Beecher. (2001). Song type matching as threat:
A test using interactive playback. Animal Behaviour, 62, 1163–1170.
Burt, J. M., A. L. O’Loghlen, C. N. Templeton, S. E. Campbell, & M. D. Beecher.
(2007). Assessing the importance of social factors in bird song learning: A test
using computer-simulated tutors. Ethology, 113, 917–925.
Fisher, J. B. (1954). Evolution and bird sociality. In J. Huxley, A. C. Hardy, & E. B.
Ford (eds.), Evolution as Process (pp. 71–83). London: Allen and Unwin.
Griffith, S. C., I. P. F. Owens, & K. A. Thuman. (2002). Extra pair paternity in birds:
A review of interspecific variation and adaptive function. Molecular Ecology, 11,
2195–2212.
Hill, C. E., Ç. Akçay, S. E. Campbell, & M. D. Beecher. (2011). Extrapair paternity,
song and genetic quality in song sparrows. Behavioral Ecology, 22, 73–81.
Horning, C. L., M. D. Beecher, P. K. Stoddard, & S. E. Campbell. (1993). Song per-
ception in the song sparrow: Importance of different parts of the song in song
type classification. Ethology, 94, 46–58.
Hughes, M., R. C. Anderson, W. A. Searcy, L. M. Bottensek, & S. Nowicki. (2007).
Song type sharing and territory tenure in eastern song sparrows: Implications
for the evolution of song repertoires. Animal Behaviour, 73, 701–710.
Hughes, M., S. Nowicki, W. A. Searcy, & S. Peters. (1998). Song-type sharing in
song sparrows: Implications for repertoire function and song learning. Behav-
ioral Ecology and Sociobiology, 42, 437–446.
Hyman, J. & M. Hughes. (2006). Territory owners discriminate between aggressive
and nonaggressive neighbours. Animal Behaviour, 72, 209–215.
Hyman, J., M. Hughes, W. A. Searcy, & S. Nowicki. (2004). Individual variation in
the strength of territory defense in male song sparrows: Correlates of age,
territory tenure, and neighbor aggressiveness. Behaviour, 141, 15–27.
Johnstone, R. A. (2001). Eavesdropping and animal conflict. Proceedings of the
National Academy of Sciences, USA, 98, 9177–9180.
Johnstone, R. A. & R. Bshary. (2004). Evolution of spite through indirect reciprocity.
Proceedings of the Royal Society of London, B, 271, 1917–1922.
Kroodsma, D. E. & R. Pickert. (1980). Environmentally dependent sensitive periods
for avian vocal learning. Nature, 288, 477–479.
Marler, P. & S. Peters. (1987). A sensitive period for song acquisition in the song spar-
row, Melospiza melodia, a case of age-limited learning. Ethology, 76, 89–100.
Marler, P. & S. Peters. (1988). The role of song phonology and syntax in vocal learn-
ing preferences in the song sparrow, Melospiza melodia. Ethology, 77, 125–149.
Nelson, D.A. (1992). Song overproduction and selective attrition lead to song sharing
in the field sparrow (Spizella pusilla). Behavioral Ecology and Sociobiology, 30,
415–424.
60 ANIMAL BEHAVIOR

Nelson, D. A. & P. Marler. (1994). Selection-based learning in bird song develop-

ment. Proceedings of the National Academy of Sciences, USA, 91, 10498–10501.
Nice, M. M. (1937). Studies in the life history of the song sparrow. I. A population
study of the song sparrow. Transactions of the Linnean Society of New York, 4,
1–247.
Nice, M. M. (1943). Studies in the life history of the song sparrow II. The behavior of
the song sparrow and other passerines. Transactions of the Linnean Society of
New York, 6, 1–328.
Nordby, J. C., S. E. Campbell, & M. D. Beecher. (1999). Ecological correlates of
song learning in song sparrows. Behavioral Ecology, 10, 287–297.
Nordby, J. C., S. E. Campbell, & M. D. Beecher. (2001). Late song learning in song
sparrows. Animal Behaviour, 61, 835–846.
Nordby, J. C., S. E. Campbell, & M. D. Beecher. (2002). Adult song sparrows do not
alter their song repertoires. Ethology, 108, 39–50.
Nordby, J. C., S. E. Campbell, & M. D. Beecher. (2007). Selective attrition and indi-
vidual song repertoire development in song sparrows. Animal Behaviour, 74,
1413–1418.
Nordby, J. C., S. E. Campbell, J. M. Burt, & M. D. Beecher. (2000). Social influen-
ces during song development in the song sparrow: A laboratory experiment
simulating field conditions. Animal Behaviour, 59, 1187–1197.
Nowicki, S. & W. A. Searcy. (2004). Song function and the evolution of female pref-
erences: Why birds sing, why brains matter. Behavioral Neurobiology of Bird-
song, 1016, 704–723.
Nowicki, S., W. A. Searcy, & S. Peters. (2002). Quality of song learning affects
female response to male bird song. Proceedings of Royal Society of London, B,
269, 1949–1954.
Nulty, B., J. M. Burt, Ç. Akçay, C. N. Templeton, E. S. Campbell, & M. D.
Beecher. (2010). Song learning in song sparrows: Relative importance of
autumn vs. spring tutoring. Ethology, 116, 653–661.
O’Connor, K. D., A. B. Marr, P. Arcese, L. F. Keller, K. J. Jeffery, & M. W. Bruford.
(2006). Extra-pair fertilization and effective population size in the song spar-
row Melospiza melodia. Journal of Avian Biology, 37, 572–578.
O’Loghlen, A. L. & M. D. Beecher. (1997). Sexual preferences for mate song types in
female song sparrows. Animal Behaviour, 53, 835–841.
O’Loghlen, A. L. & M. D. Beecher. (1999). Mate, neighbour and stranger songs: A
female song sparrow perspective. Animal Behaviour, 58, 13–20.
Otte, D. (1974). Effects and functions in the evolution of signaling systems. Annual
Review of Ecology and Systematics, 5, 385–417.
Patten, M. (2001). The roles of habitat and signalling in speciation: evidence from a
contact zone of two song sparrow (Melospiza melodia) subspecies. Ph.D. disser-
tation, University of California, Riverside.
Sardell, R. J., L. F. Keller, P. Arcese, T. Bucher, & J. M. Reid. (2010). Comprehen-
sive paternity assignment: Genotype, spatial location and social status in song
sparrows. Molecular Ecology, 19, 4352–4364.
FRIENDS AND ENEMIES 61

Searcy, W. A., R. C. Anderson, & S. Nowicki. (2006). Bird song as a signal of aggres-
sive intent. Behavioral Ecology and Sociobiology, 60, 234–241.
Smith, J. N. M. (1984). Territory size, inheritance, and fitness in the song sparrow.
American Zoologist, 24, A103–A103.
Smith, J. N. M., L. F. Keller, A. B. Marr, & P. Arcese. (2005). Biology of Small
Populations: The Song Sparrows of Mandarte Island. New York: Oxford Univer-
sity Press.
Smith, J. N. M. & R. Zach. (1979). Heritability of some morphological characters in
a song sparrow population. Evolution, 33, 460–467.
Stoddard, P. K., M. D. Beecher, C. L. Horning, & S. E. Campbell. (1991). Recogni-
tion of individual neighbors by song in the song sparrow, a species with song
repertoires. Behavioral Ecology and Sociobiology, 29, 211–215.
Stoddard, P. K., M. D. Beecher, C. L. Horning, & S. E. Campbell. (1992). Song type
matching in the song sparrow. Canadian Journal of Zoology, 70, 1440–1444.
Temeles, E. J. (1994). The role of neighbors in territorial systems: When are they
“dear enemies”? Animal Behaviour, 47, 339–350.
Templeton, C. N., Ç. Akçay, S. E. Campbell, & M. D. Beecher. (2009.) Juvenile
sparrows preferentially eavesdrop on adult song interactions. Proceedings of
the Royal Society of London, B, 277, 447–453.
Templeton, C. N., J. M. Burt, S. E. Campbell, K. Lent, E. A. Brenowitz, & M. D.
Beecher. (2012). Immediate and long-term effects of testosterone on song plas-
ticity and learning in juvenile song sparrows. Behavioural Processes, 90,
254–260.
Templeton, C. N., S. E. Campbell, & M. D. Beecher. (2012.) Territorial song spar-
rows tolerate juveniles during the early song-learning phase. Behavioral Ecology,
23, 916–923.
Templeton, C. N., V. A. Reed, S. E. Campbell, & M. D. Beecher. (2012). Spatial
movements and social networks in juvenile male song sparrows. Behavioral
Ecology, 23, 141–152.
Weeden, J. S. & J. B. Falls. (1959). Differential responses of male ovenbirds to
recorded songs of neighboring and more distant individuals. Auk, 76,
343–351.
Westneat, D. F. & I. R. K. Stewart. (2003). Extra-pair paternity in birds: Causes, cor-
relates and conflict. Annual Review of Ecology, Evolution, and Systematics, 34,
365–396.
This page intentionally left blank
 3

A Tale of Two Spiders: Investigating

Communication in Two Unique Model
Species Using Video Digitization
and Playback
George Uetz and David Clark

INTRODUCTION
In this chapter, we present two case studies based on our work together on spi-
der communication behaviors. We will review how a series of fortuitous dis-
coveries led to a better understanding of the role of communication in
species recognition and mate choice and resulted in novel technologies for
the study of animal behavior. To put things in perspective, it is important to
consider the time when some of this research began. Thirty-five years ago,
when George Uetz was finishing graduate school, computers were large enti-
ties that occupied rooms, sometimes whole buildings, and could only be
accessed remotely via coded punch cards or teletype terminals. There were
no easily available video recording devices, only audio tape recorders. Impor-
tantly, the study of behavior was different, and most invertebrate animal
behavior was assumed to be entirely ‘hard-wired’ and unchanged by experi-
ence, clearly unlike the complex behaviors exhibited by the better-studied
birds and mammals. As can be seen from this chapter, experimental animal
64 ANIMAL BEHAVIOR

behavior research has benefitted from the rapid evolution of technology, and
ideas about invertebrate behavior have changed dramatically over time.

THE BEGINNING: ETHOSPECIES IN THE GENUS SCHIZOCOSA

While conducting doctoral research on the spider fauna of a floodplain for-
est in Illinois, Uetz encountered a taxonomic conundrum. Spider species are
usually identified by the structure of their genitalia, illustrated in taxonomic
keys. However, specimens of a wolf spider from the floodplain forest, identi-
fied as the brush-legged wolf spider, Schizocosa ocreata (Hentz) on the basis
of genitalia, lacked the characteristic (and highly conspicuous) tufts of bristles
on the forelegs of males (Figure 3.1). In the newly established Uetz lab at the
University of Cincinnati, this taxonomic mystery would ultimately lead to a
research project that continues to this day.
Our colleague Jerry Rovner of nearby Ohio University, who studied wolf
spider behavior, mentioned that he had seen several male S. ocreata without

Figure 3.1a. Male wolf spider, Schizocosa ocreata. (Photo source George Uetz)
A TALE OF TWO SPIDERS 65

Figure 3.1b. Male wolf spider, Schizocosa rovneri. (Photo source George Uetz)

tufts that later emerged parasites. He suggested that this might be a case of
parasitic castration, where the internal parasite (a mermithid nematode) sup-
pressed development of the gonads and subsequent expression of adult secon-
dary sex characteristics. If so, this would represent an unusually high rate of
parasitic infection, worthy of further study. However, after dissection of many
specimens revealing no parasites, a different approach was needed, and exami-
nation of the behavior of live specimens was in order.
European arachnologists studying wolf spiders had introduced the concept
of ethospecies (Hollander et al., 1973; Hollander & Dijkstra, 1974), to iden-
tify species populations that were indistinguishable morphologically but repro-
ductively isolated by behavior during courtship and mating (commonly
known as a form of cryptic species—see Bickford et al., 2007). Spiders from
our two different populations—S. ocreata from Ohio with brush-legged males
Figure 3.2. Courtship behavior responses of male wolf spiders (left) and receptivity
responses of female wolf spiders (right) to members of the same species (conspe-
cifics) or the other species (heterospecifics). (a) S. ocreata, (b) S. rovneri. (Redrawn
from data in Uetz & Denterlein, 1979; Stratton & Uetz, 1981)
A TALE OF TWO SPIDERS 67

and the unknown species from Illinois without leg tufts—were paired in mat-
ing trials. Studies with undergrad student Gerri Denterlein found that males
from each population exhibited courtship with females from either population,
as well as their silk alone, but females were only receptive to males from their
own population. Despite having identical genitalia, members of these two popu-
lations did not interbreed (Figure 3.2), and observations of striking differences in
male courtship behavior suggested the likelihood of behavioral reproductive isola-
tion (Uetz & Denterlein, 1979). Based on these behavioral differences, a new spe-
cies, Schizocosa rovneri (Figure 3.1b), was described (Uetz & Dondale, 1979).
However, these findings raised an important question—if spiders from dif-
ferent populations do not mate, how could we test the hypothesis that the
behavioral barrier, and not some other mechanism (e.g., gametic incompatibil-
ity or developmental failure) was the cause? Studies with graduate student Gail
Stratton tested this hypothesis by anesthetizing females and allowing males to
mate with females from each population. Cross-species and within-species mat-
ings had similar results—females from each type of mating produced egg sacs
with equal frequency, and spiderlings emerged from those egg sacs, confirming
gametic and developmental compatibility (Stratton & Uetz, 1981, 1983).
Moreover, subsequent interbreeding studies demonstrated that the behaviors
characteristic of each species (the “double tap” and “jerky tap” of S. ocreata,
the “body bounce” of S. rovneri) were heritable and sorted via crosses into Men-
delian inheritance patterns (Stratton & Uetz, 1986). These studies were among
the first to demonstrate that behavior by itself could constitute a premating spe-
cies isolating mechanism, and given that S. ocreata and S. rovneri have overlap-
ping geographic ranges and nearly identical habitats, suggested that behavior
might be a driving force in sympatric speciation. Indeed, subsequent to these
early studies, Gail Stratton has gone on to discover and describe several more
cryptic ethospecies in the S. ocreata clade (Stratton, 1991, 1997, 2005).

COMMUNICATION AND SEXUAL SELECTION IN S. OCREATA, PART 1

Given the role that behavior plays in premating reproductive isolation of S.
ocreata and S. rovneri, the question was raised whether sexual selection might
play a role in evolution of the obvious species differences in male morphology.
Male leg tufts, present in S. ocreata but absent in S. rovneri, seemed a likely tar-
get of sexual selection, as they represented male secondary traits with no
obvious function other than enhancing the leg-tapping and waving displays
used in male courtship to attract female attention. Graduate student Sonja
Scheffer tested hypotheses regarding both intrasexual and intersexual selec-
tion for male tufts by shaving them off. She conducted several experiments
68 ANIMAL BEHAVIOR

with interesting results. She tested whether having tufts affected male-male
interactions by placing males in an arena for 24 hours to allow them to estab-
lish a dominance hierarchy (Aspey, 1977a, 1977b), then introducing a
female. She compared results of that experiment to mating trials with males
and females introduced simultaneously. There were no differences in mating
success associated with the presence of tufts, nor was there any evidence of
male-male conflict or dominance, suggesting the absence of intrasexual selec-
tion based on male tufts. However, something else was apparent—when males
were placed together in an arena with a female, 79 percent of females mated
with the male that initiated courtship first, regardless of whether that male
had tufts. Because shaved and unshaved males were equally likely to be the
first to court (53% versus 47%, respectively), there was a significant female
bias for the first male (Scheffer et al., 1996). These results pointed out a poten-
tial problem of research bias inherent in many experimental studies of mate
choice, which arises when a female is offered two males that initiate courtship
asynchronously (see Clark & Uetz, 1993; Schwartz et al., 2001).
Scheffer’s findings caused us to reconsider the behavior differences between
S. ocreata and S. rovneri, and we conducted additional experimental studies. In
one experiment, we compared the responses of females of S. ocreata and S. rov-
neri to isolated courtship cues from male courtship. Interestingly, female S.
rovneri showed high levels of receptivity to seismic vibration cues alone but
were far less receptive to isolated visual cues from males. In contrast, female
S. ocreata responded equally to vibration or visual cues from males. Seismic
communication was necessary and sufficient for S. rovneri, but for S. ocreata,
vibratory and visual cues were interchangeable (Scheffer et al., 1996). How-
ever, male S. ocreata produce seismic and visual signals simultaneously, as a
multimodal signal, while S. rovneri signals are clearly unimodal. These find-
ings raised questions about the evolution of divergent signaling behaviors in
these closely related species. Because seismic communication is characteristic
of all nearctic wolf spiders (Rovner, 1975), visual communication and leg tufts
in S. ocreata must be derived characters.
To test whether microhabitat might play a role in these differences, we
measured seismic signal transmission over distance in the microhabitats where
these species occur in the field, using a crude measure—amplitude of vibration
produced by a standardized pencil drop. Results suggested that the loose leaf
litter in which S. ocreata was found tended to attenuate vibration more quickly
and over shorter distances than did the compacted floodplain litter microhabi-
tat of S. rovneri. We hypothesized that leg tufts might enable S. ocreata males to
communicate with females under circumstances where seismic signals might not
be detected. Consequently, we compared responses of female S. ocreata to shaved
A TALE OF TWO SPIDERS 69

Figure 3.3. Receptivity responses of female S. ocreata to control (intact) males and
males whose tufts were removed by shaving in arenas allowing both visual and seis-
mic vibration signals or visual signals alone (Redrawn from data in Scheffer et al.,
1996)

and intact males under conditions where they could perceive both visual and
vibration cues together versus visual cues alone. With multimodal cues, females
showed no difference in response to shaved males or those with intact tufts,
but when only visual cues were present, female response to shaved males was sig-
nificantly lower (Figure 3.3). These findings suggested that leg tufts function to
increase signal efficacy in a complex litter environment (Scheffer et al., 1996).

MAEVIA INCLEMENS , A JUMPING SPIDER WITH TWO DISTINCT MALE MORPHS

Dave Clark joined the Uetz lab as a Ph.D. student in the late 1980s, with a
background in herpetology and behavior that brought a fresh perspective to
studies of spiders. During a field trip in an Arachnology class, he found two
male jumping spiders that were distinctly different in appearance, which, like
the wolf spiders above, were identified as the same species based on genitalia.
However, unlike Schizocosa, these were specimens of Maevia inclemens
(Walckenaer)—a species well recognized for over a century as having
70 ANIMAL BEHAVIOR

dimorphic males (Peckham & Peckham, 1889, 1890; Painter, 1913). The
dimorphic jumping spider (also known under the name vittata) is a jumping
spider commonly found in the leaf litter and forest-edge vegetation throughout
the eastern and midwestern United States. The two male morphs of this species
differ strikingly in both morphology and behavior (Peckham & Peckham 1889,
1890; Emerton, 1961; Painter, 1913, 1914; Barnes, 1955; Jackson, 1982;
Clark, 1994). The males have been described as variable in coloration: the
tufted morph (Figure 3.4a) has a black-colored body with three tufts of setae
above the eyes in the center of the anterior prosoma (the anterior portion of
the body of spiders that lacks segmentation), the legs are pale and white in col-
oration (except for a black band near the tips of the first pair of legs) and the
pedipalps (the second pair of head appendages of spiders) are generally entirely
black; in the gray morph (Figure 3.4b) the body has dark chevrons over a light
ground color, the sides of the abdomen and legs are striped (Barnes, 1955; Kas-
ton, 1972). The gray morph can also be distinguished from the tufted morph by
the absence of tufts above the eyes, instead having a light-colored color band on
the anterior cephalothorax (the fused head and thorax of spiders) above the
median and lateral eyes and bright yellow- to orange-colored pedipalps. Females
of the species are distinctly different from either of the males, being larger and
with cryptically colored body color patterns and a conspicuous white stripe
below the anterior eyes; they lack the tufts and brightly colored pedipalps.
However, in addition to differences in appearance, males of M. inclemens
also exhibit polyphenism (multiple discrete behavioral forms arising from a
single genotype) in courtship behavior (Clark, 1994). After orienting to a
female, the tufted morph will stand up (or stilt) with the first pair of legs held
high above the prosoma and wave them to and fro, cyclically (Figure 3.5a). In
contrast, the gray male will typically approach the female with its body low-
ered to the substratum, at the same time dropping legs I and II and assuming
a prone or crouched position (where the first two pair of legs are pointed for-
ward so that the tips touch in front in a V formation, bent at right angles).
After assuming this prone position, the gray morph moves back and forth in
a semicircle before the female, sometimes advancing, sometimes receding (Fig-
ure 3.5b). While males are behaviorally distinct during the initial phase of the
courtship sequence, the motor patterns of the two morphs converge and they
behave identically during the subsequent phase of the courtship sequence. At
this point, both males raise their front legs and move together and apart
repeatedly while at the same time zig-zagging laterally side to side as they
approach the female to initiate copulation. This zig-zag dance display is typical
of many jumping spider species, and it has been suggested that it allows female
mate assessment.
Figure 3.4. Maevia inclemens jumping spiders. (a) Male tufted morph, (b) male gray
morph. (Photos by Dave Clark)
Figure 3.5a. Behavior diagram for male courtship in Maevia inclemens jumping spiders,
tufted morph. (Redrawn from data in Clark, 1994)

72
Figure 3.5b. Behavior diagram for male courtship in Maevia inclemens jumping spiders,
gray morph. (Redrawn from data in Clark, 1994)

73
74 ANIMAL BEHAVIOR

VIDEO DIGITIZATION AND PLAYBACK WITH M. INCLEMENS —SERENDIPITY AND

ACCIDENTAL DISCOVERY LEADS TO NEW RESEARCH OPPORTUNITIES
For a part of Clark’s Ph.D. research concerning the male dimorphism in M.
inclemens, sequences of male courtship behavior were filmed with Super 8mm
film to be scored for a detailed analysis of display motor patterns. One after-
noon, while working in a small lab space surrounded by dozens of male and
female dimorphic jumping spiders, Clark was reviewing the film footage of
male courtship behavior on the wall. In a moment reminiscent of Tinbergen’s
fortuitous discovery that male stickleback fish were excited by the color red on
a passing truck, Clark noticed that spiders in the room were also watching the
film clips of the courting male projected on the wall! We immediately recog-
nized the significance of this discovery and first tested females by taping a
piece of opaque tracing paper to their small plastic cages, projecting life-sized
courting males to them, and observing their response. To our astonishment,
the females approached the projected males and displayed sexual receptivity,
soliciting the filmed images of the male to mate. Alas, the male on screen just
kept courting until the film loop ran out; never realizing his missed mating
opportunity.
This accidental discovery—that jumping spiders perceived and responded
to filmed images of courtship—led to more complete studies testing spider
perception of video images and video playback as an experimental technique
(Clark & Uetz, 1990). We recognized the difficulties associated with working
with film at the time of our discovery, but personal computers like the Com-
modore Amiga and Apple Macintosh were new to the market and provided
users with the ability to animate and show video images on small Sony Watch-
man TV screens. Remember, this was in an age before user-friendly computer
paint programs were readily available and long before the invention of the
video iPod that we use today. So first we had to convince ourselves that what
we were seeing was really evidence for jumping spiders’ ability to perceive dig-
ital video. We tested spider responses to a variety of biologically relevant
stimuli—insect prey (crickets), conspecific spiders (Maevia), and heterospe-
cific spider predators (Phidippus spp.). In all these cases, spiders behaved
appropriately toward videos on the Watchman screen—attacking crickets, dis-
playing species- and sex-appropriate behavior toward conspecifics, and retreat-
ing from predators—indicating that they recognized the images and perceived
them as if they were real (Figure 3.6).
Figure 3.6a. Female M. inclemens response to video images. Photo of female response
to video of male. (Photo source David Clark)

Figure 3.6b. Frequency of behaviors exhibited in response to live prey insects (dark
bars) or video images of prey (light bars). (Redrawn from data in Clark & Uetz, 1990)
76 ANIMAL BEHAVIOR

Figure 3.6c. Frequency of response behaviors to prey (dark bars) conspecifics (hatched
bars), and other predatory spiders (light bars). (Redrawn from data in Clark & Uetz,
1990)

DEVELOPMENT OF VIDEO PLAYBACK AS AN EXPERIMENTAL TECHNIQUE

The computer-animated stimuli used for our original studies were created
from raw video footage of courting males (see Clark & Uetz [1990, 1992,
1993] for details on constructing video stimuli). We used desktop computers
to digitize and edit video sequences into the digitized movie format. These
movies could easily be standardized for time and sequencing, and stimuli
could be videotaped against green-screen backgrounds and superimposed over
different backgrounds, depending on the study question. They could be color-
ized or otherwise manipulated in ways that live test subjects could not be
altered. Importantly, stimulus sequences could be sped up, slowed down, or
set to the mean rate of movement derived from a large sample of live males
as a representation of population attributes. Computer-animated stimuli could
also be archived by storing them digitally or downloading them to videotape.
The ability to control the test stimulus and modify variables one at a time
revolutionized our ability to ask detailed questions about the behavior of these
animals.
For example, one of our first studies to use this technique was conducted to
help us understand mate choice in females of the dimorphic jumping spider.
Jumping spiders show the same movement bias as we described with Schizo-
cosa earlier; that is, they respond to the male that initiates courtship first, rais-
ing questions about mate preference. We used video playback to test first-male
preference by presenting females with simultaneously or sequentially moving
males. We found that females show receptivity to whichever video male
A TALE OF TWO SPIDERS 77

morph began courtship first but showed no preference when video male
morphs courted simultaneously (Clark & Uetz, 1992). This finding led to fur-
ther studies that revealed the polyphenism of Maevia to represent two alterna-
tive male courtship tactics that function to attract female attention at different
distances (Clark, 1994; Clark & Uetz, 1993; Clark & Biesiadecki, 2002)
(Figures 3.7a and 3.7b).

Figure 3.7a. Phase I courtship distance of male M. inclemens relative to the female.
Angle subtended by the courtship stance. (Redrawn from data in Clark & Uetz, 1993)

Figure 3.7b. Graphic model of perceived male area as a function of distance from the
female. (Redrawn from data in Clark & Uetz, 1993)
78 ANIMAL BEHAVIOR

Subsequent studies with video playback and digital image manipulation

have enabled experiments in “virtual evolution.” For example, switching the
morphology and behavior of males of the dimorphic jumping spider and dig-
itally creating a hypothetical ancestral form allowed testing hypotheses about
the evolution and divergence of these two male types (Clark & Uetz, 1993).
This technique allowed us to decouple behavior and morphology to create a
male stimulus that looked like the gray form but courted like the tufted form
by stilting, and a tufted form that assumed the prone courtship position.
These “video chimeras” allowed testing of the role of behavior and appearance
in mate recognition. We discovered that behavior was more important for
tufted male recognition as females were receptive to a gray morph that
behaved like a tufted male. In contrast, morphology was more important for
gray morph recognition as females attacked tufted males that behaved like gray
males (Clark & Uetz, 1993). Clark and undergraduate student Carrie Morjan
(2001) tested the hypothesis that each morph may have a selective advantage
at different courtship distances. They presented courting male stimuli at dif-
ferent distances from the females and determined that the gray morph has an
advantage when close to the female, where it attracts female attention in less
time than the tufted male and experiences less predation attacks than tufted
males. In contrast, tufted morphs have a selective advantage at greater distan-
ces where the stilting display attracts female attention more quickly than the
prone display of the gray male. Furthermore, the stilting display is likely less
attractive to predators because it is a stationary display, whereas the prone dis-
play of the gray male consists of the visually noisy side-to-side movements.
Studies of Maevia inclemens continue in the Clark lab, and we are currently
investigating the risks of predation associated with male courtship display ver-
sus morphology.

COMMUNICATION AND SEXUAL SELECTION IN S. OCREATA , PART 2: VIDEO

PLAYBACK STUDIES
When Clark finished his Ph.D. and went on to Alma College to continue
with studies of jumping spiders, lizards, and toads, the Uetz lab—now
equipped with new tools of video digitization and playback—continued
research on mate preference and sexual selection in S. ocreata. The develop-
ment and use of video digitization and playback has allowed us to ask and
answer a number of questions about mate choice in S. ocreata. For example,
we were able to explore the role of female preference for larger leg tufts while
controlling for the covarying traits of male size and courtship vigor. We cre-
ated a virtual spider from a videotape of male S. ocreata courtship by using
A TALE OF TWO SPIDERS 79

frame-by-frame video rendering software to adjust body size and courtship

rates to match population means. We then manipulated individual traits
(e.g., presence/absence as well as size of male tufts) for presentation to females
(Figure 3.8a). In both single-presentation and paired-choice tests, females pre-
fer males with larger tufts (McClintock & Uetz, 1996; Uetz & Roberts, 2002;
Uetz & Norton, 2007).
Another aspect of the studies we conducted on the role of male leg tufts
concerned asymmetry, which arises as a consequence of leg loss during devel-
opment. A number of studies in other animals had suggested that fluctuating
asymmetry (an organism’s deviation from perfect bilateral symmetry), espe-
cially in male secondary traits that serve as mate choice criteria, was an indica-
tor of developmental instability and thereby important in sexual selection

Figure 3.8a. Responses of female S. ocreata to manipulation of male tufts in video play-
back studies. Receptivity of female S. ocreata to both live males (dark bars) and video
males (gray bars) with varied tuft sizes (illustrated above graph). (Redrawn from data in
Uetz, 2000)
80 ANIMAL BEHAVIOR

Figure 3.8b. Female receptivity response (%) to video males with varying types of asym-
metry (illustrated above graph). (Redrawn from data in Uetz & Smith, 1999)

(see Polak & Starmer, 2001; Polak, 2003). Male spiders that lose a leg as juve-
niles are capable of regeneration (Uetz et al., 1996; Wrinn & Uetz, 2007), and
many adult male S. ocreata with regenerated forelegs had smaller or missing
tufts. We tested whether this asymmetry had an impact on sexual selection
by comparing the success of asymmetric males (both naturally occurring and
A TALE OF TWO SPIDERS 81

experimentally shaved) in both male-male and male-female interactions. In

both cases, asymmetric tufts were disadvantageous; asymmetric males had sig-
nificantly lower mating success and were more likely to retreat in male-male
encounters (Uetz et al., 1996). Creation of varying types of tuft asymmetry
(with one tuft absent, one smaller or larger than the other, or one enlarged
and one reduced) by undergrad student Liz Smith (Figures 3.8b and 3.8c)
revealed female preference for symmetry (Uetz & Smith, 1999).
While courtship of male S. ocreata is multimodal, not all species in the genus
use both visual and seismic cues in courtship (Hebets & Uetz, 1999). Video play-
back with digitally modified virtual spiders has allowed us to conduct virtual evo-
lution experiments in which we could test whether removing decorations such as
leg tufts, or adding them to species that do not have them, might influence species
recognition and mate preference in multiple species of Schizocosa (Figure 3.9). In
their M.S. thesis research, Will McClintock and Eileen Hebets used digital
manipulation to modify tufts in several species in the ocreata clade, which vary

Figure 3.9. Receptivity responses of female Schizocosa spp. (proportion of females) to

video stimuli with tufts added or removed. (Redrawn from data in McClintock &
Uetz, 1996; Hebets & Uetz, 2000)
82 ANIMAL BEHAVIOR

in male use of visual courtship signals and the presence of leg pigmentation and
tufts (McClintock & Uetz, 1996; Hebets & Uetz, 2000). In species that exhibit
visual displays and leg ornamentation (S. ocreata and S. crassipes), digital removal
of tufts decreased female receptivity responses. In two species that have nonvisu-
ally displaying, nonornamented males, S. duplex and S. uetzi, female receptivity
did not increase when tufts were added to conspecific males (Hebets & Uetz,
2000). In contrast, female S. rovneri were more receptive to video images of con-
specific males with digitally added tufts (McClintock & Uetz, 1996). In another,
somewhat intermediate species—with a visual display and foreleg pigmentation
(S. stridulans)—the addition of foreleg tufts increased female receptivity. These
results become even more interesting when placed in the context of phylogenetic
relationships within the genus Schizocosa, as it appears that leg tufts have evolved
multiple times independently but may have been secondarily lost in some mem-
bers of the ocreata clade (McClintock & Uetz, 1996; Stratton, 2005; Hebets &
Vink, 2007). If digital addition of tufts to S. rovneri increases female receptivity,
as McClintock and Uetz (1996) found, males of this species may have secondarily
lost the trait, but females have retained the preference (Stratton, 2005). We now
know from a variety of studies that tufts and leg pigmentation play an important
role in female mate choice in a number of Schizocosa species as both condition
indicators and attention-altering traits (Uetz et al., 2002; Hebets, 2005).
The combination of video and audio playback of spider courtship has
enabled a unique form of virtual evolution experiment. In these experiments,
a speaker mounted below the substrate in front of the video screen was used
to play back synchronized seismic signals along with videos to create multimo-
dal stimuli. By combining video of S. ocreata with seismic playback of S. rov-
neri—and vice versa—we created multimodal digital chimeras to test against
controls in cue-conflict experiments to see which elements are essential in spe-
cies recognition (Uetz & Roberts, 2002). For S. rovneri, seismic signals are the
essential element, as females responded to visual cues from either species as
long as conspecific seismic signals were present. However, for S. ocreata only
the correct combination of conspecific visual and seismic signals elicited
female receptivity (Figure 3.10).
Many animal behavior studies have explored the trade-off between sexual
selection favoring larger male ornaments and natural selection from predators
acting against such decorations (see Volume 2, Chapter 5). Video digitization
and playback has allowed us to ask the following questions: Would predation
rates be lower is males had no tufts? or If larger tufts are attractive to females, what
keeps them from evolving to be even larger? and answer them with virtual evolu-
tion experiments. Andy Roberts’s doctoral research examined these questions,
using latency of orientation response of predatory American toads (Bufo
A TALE OF TWO SPIDERS 83

Figure 3.10. Response of female S. ocreata and S. rovneri (mean receptivity score) to
male signals in video/vibration cue-conflict experiments. (Redrawn from data in
Uetz & Roberts, 2002)

americanus) to virtual courting S. ocreata male stimuli with varying tuft size and
latency of orientation by female S. ocreata to the same stimuli as an index of con-
specific detection. When compared with control stimuli representing the popula-
tion average, elimination of tufts had no significant effect on predator detection
but increased orientation latency of females. Increasing the size of male tufts
had no effect on female detection but did significantly reduce detection latency
for predatory toads (Figure 3.11). Results suggest that for a courting male spider
of a given size and vigor level, variation in the expression of a secondary sexual
characteristic alone can incur differential costs and benefits by influencing latency
of detection of visual signals by predators and conspecifics.
Video playback experiments have provided other insights, for example, the
mechanisms underlying foraging behavior and decision making in wolf
84 ANIMAL BEHAVIOR

Figure 3.11a. Responses of American toads (Bufo americanus) to manipulated video

stimuli of courting male S. ocreata. Mean latency to orient to videos with tufts either
removed, unmanipulated, or enlarged. (Redrawn from data in Roberts & Uetz, 2008)

spiders. When he was a graduate student, Matt Persons was interested in how
sensory stimuli affect spider decision making—specifically, how long to stay in
a patch before moving on. He found that sensory cues, even without food
rewards, were enough to influence residence time. Spiders remained longer
in patches with visual cues alone or visual and vibratory cues together over
those with vibratory information alone (Persons & Uetz, 1996). Persons cre-
ated an artificial foraging-patch stimulus chamber with video playback of a
computer digitized, animated image of a cricket. By manipulating prey move-
ment, he found that spiders stayed significantly longer in treatments with
higher levels of cricket locomotor activity and longer when they attacked the
stimulus than when they did not. An analysis of the distribution of spider
patch-residence times indicated that S. ocreata uses a decision rule based on a
fixed probability of leaving (Persons & Uetz, 1997).
A TALE OF TWO SPIDERS 85

Figure 3.11b. Mean latency to approach (dark bars) and attack (light bars). (Redrawn
from data in Roberts & Uetz, 2008)

CURRENT VIDEO PLAYBACK STUDIES

Complex Visual Environments
Wolf spider species in the genus Schizocosa live in the visually complex envi-
ronment of forest-floor leaf litter, where body coloration of animals often
closely matches the background coloration (Endler, 1993, 2006). However,
for visual communication to be effective, animals should exhibit high contrast
against their background (Endler, 1991, 2006; Fleishman, 1992; Marchetti,
1993; Macedonia et al., 2002; Uy & Endler, 2004). In recent studies, we used
illuminance spectrophotometry and spectral analysis (a method of measur-
ing the components of light) to examine to what degree the visual color
pattern of S. ocreata matches or contrasts with the leaf-litter background
(Clark et al., 2011). When viewed from above, colors of the median stripe of
male and female S. ocreata closely overlap spectral values for dead leaves and
soil (mostly brown), suggesting that their coloration is cryptic and less visible
to well-sighted predators such as birds. However, when viewed from a
86 ANIMAL BEHAVIOR

different perspective, a side view (as seen by other spiders) shows higher
contrast color-spectrum values that are the polar opposite of leaf litter. In
addition, male leg tufts show the highest levels of background contrast.
These findings suggest that the “two-tone” color patterns of S. ocreata wolf
spiders may have evolved to provide both crypsis and contrast at the same
time, depending on the viewing perspective of the receiver (Clark et al.,
2011).
We tested the importance of background contrast using video exemplars of
courting males (with and without leg tufts) superimposed on digital images of
different backgrounds (sunny leaf litter, shady leaf litter, and a featureless
monochromatic gray background) displayed to female spiders on iPod screens
(Figures 3.12a through 3.12c). Females oriented more quickly to males on
lighter backgrounds (litter in sun, monochrome gray) than dark ones (shady
litter) (Figure 3.12d). Males with leg tufts were also more likely to be detected
than those with tufts digitally removed. Latency of females to approach was
also shorter against lighter backgrounds and longest for males without tufts
(Figure 3.12e). Female recognition and receptivity (based on summative dis-
play scores) were greater with males displayed against lighter backgrounds;
males with large tufts had the highest scores (Figure 3.12f). Results suggest
that light levels and visual complexity of the leaf-litter habitat can affect the
potential for detection and recognition of male visual courtship signals by
females, and that male foreleg tufts may serve to increase chances of detection
(and receptivity) in this complex environment (Clark et al., 2011). Ongoing
video playback studies are directed at exploring the role of the cryptic yet
high-contrast color patterns of S. ocreata in detection by visual predators such
as jumping spiders, toads, and birds.

Social Contexts
In the spring, male S. ocreata wolf spiders actively search for mates, and
because they occur at fairly high densities (2–6 spiders/m2), they are likely
to encounter each other or detect each other’s signals. Overlap of the range
of both visual and seismic signals creates a “communication network” and
increases the potential for eavesdropping and mate competition. While
aggressive interactions appear minimal in S. ocreata, mate competition
between males may take a different form, as males might increase mating
success by eavesdropping on others’ behavior and signaling whenever
courtship is detected. We tested whether males initiate courtship themselves
when exposed to video playback of visual signals of other males, but results were
A TALE OF TWO SPIDERS 87

Figure 3.12a–c. Video images of spiders superimposed on different visual backgrounds.

(a) Dark (shade) complex leaf-litter background. (b) Light (sun) complex leaf-litter
background. (c) Simple photo gray background. (From Uetz et al., 2011)

inconclusive (Roberts et al., 2006). However, field observations suggested oth-

erwise, as we have frequently seen males searching and courting in the vicinity
of other males and (hidden) females. In addition, studies have found that social
experience of juvenile female Schizocosa species can influence mate choice
behavior as adults (Hebets, 2003, Hebets & Vink, 2007; Rutledge et al.,
2010), raising questions about the role that experience might play.
Because our previous studies were conducted with naïve males reared in
lab isolation, we repeated the eavesdropping studies with field-collected
males (presumed to be exposed to male courtship or female cues) using video
playback of male courtship and got different results. We found that field-
experienced male S. ocreata exhibit eavesdropping and signal matching when
exposed to video playback of courting male conspecifics. Male spiders had
longer bouts of interaction when a courting male stimulus was on screen
(Figure 3.13a), and then showed more bouts of courtship signaling after-
wards (Figure 3.13b). Even more interesting was the finding that rates of
courtship leg tapping by observer males were correlated with the rates of
video exemplars with which they were paired. When we manipulated court-
ship display rates of on-screen males, observer males increased their rates
(Figure 3.13c; Clark et al., 2012). These findings suggest male wolf spiders
gain information by eavesdropping on courtship of conspecifics and adjust
their own signals to match their rivals. This is quite a novel finding, as such
behavior has been seen previously only in vertebrate animals.

CONCLUDING REMARKS
The case studies of our research narrative illustrate the value of pursuing
interesting discoveries that arouse one’s curiosity as well as tinkering with
88 ANIMAL BEHAVIOR

Figure 3.12d. Responses of female S. ocreata to videos. Latency to orient. (Redrawn

from data in Uetz et al., 2011)

new technologies. Video animations and digital video playback are now rec-
ognized as valuable tools in behavioral studies, as they allow manipulation of
aspects of morphology or behavior while holding others constant and pre-
vent potential interaction between focal and stimulus individuals. Video
playback has been used successfully in experimental studies of numerous

Figure 3.12e. Latency to approach. (Redrawn from data in Uetz et al., 2011)
A TALE OF TWO SPIDERS 89

Figure 3.12f. Mean receptivity score (sum displays). (Redrawn from data in Uetz et al., 2011)

animal taxa, including spiders, insects, crustaceans, amphibians, reptiles, fish,

and birds (Evans & Marler, 1991, 1992; Evans et al., 1993; Macedonia &
Stamps, 1994; Macedonia et al., 1994; Roster et al., 1995; Rowland,
1995; Rowland, Bolyard, & Halpern, 1995; Rowland, Bolyard, Jenkins, et
al., 1995; Rosenthal et al., 1996; Kodric-Brown & Nicoletto, 1997, 2001;

Figure 3.13a. Evidence for eavesdropping and social facilitation of male courtship in
S. ocreata. Males interact with courting male video when on screen. (Redrawn from
data in Clark et al., 2012)
90 ANIMAL BEHAVIOR

Figure 3.13b. Males exhibit courtship during and after presentation of video stimulus
male. (Redrawn from data in Clark et al., 2012)

Figure 3.13c. Males exhibit signal matching in response to manipulated courtship rates
of male video stimulus. (Redrawn from data in Clark et al., 2012)
A TALE OF TWO SPIDERS 91

Adret, 1997; Ord et al., 2002). There are still more questions to be asked
and more technologies to explore, especially in the interdisciplinary overlap
between animal behavior and neurobiology. For example, Beth Jakob at
the University of Massachusetts is using an eyetracker to follow jumping spi-
der eye movement while spiders view prey and conspecifics on screen. This
technology was developed by Duane Harland and Robert Jackson of the
University of Canterbury, New Zealand and allows examination of the
movement of salticid principal eyes as spiders view a stimulus, much as psy-
chologists use eyetrackers to study human eye movement. Work by Eileen
Hebets at the University of Nebraska on wolf spiders and by Sarah Partan
at Hampshire College on pigeons has combined video playback with audio
playback and live animal communication for multisensory stimuli (Hebets,
2005; Partan et al., 2005). Dave Clark and others have used programmable
and interactive animal robots as stimuli in both the lab and field (Narins et
al., 2005; Fernandez-Juricic et al., 2006; Taylor et al., 2008; Ord & Stamps,
2008, Partan & Larco, 2009; Partan et al., 2010, 2011; Macedonia et al.,
2013), and a number of other researchers have begun to use three-
dimensional holographic stimuli in studies of animal behavior (Malkiel
et al., 2003; Steyer et al., 2012). Given the recent rapid evolution of tech-
nology, we wonder, What future studies might be possible?

ACKNOWLEDGMENTS
Research described in this chapter was supported by the National Science
Foundation over the past more than 20 years through grants BNS-8920013
(GWU and DLC), IBN-9307056 (DLC), IBN-9414239 (GWU), IBN
9906446 (GWU), IBN 0239164/0238854 (GWU and DLC), and IOS-
1026995/1026817 ( GWU and DLC), as well as by the as the University of
Cincinnati and Alma College. We are especially grateful to the undergraduate
and graduate students and colleagues who have participated in these research
projects.

REFERENCES AND SUGGESTED READING

Adret, P. (1997). Discrimination of video images by zebra finches (Taeniopygia gut-
tata): Direct evidence from song performance. Journal of Comparative Psychol-
ogy, 111, 115–125.
Aspey, W. P. (1977a). Wolf spider sociobiology: I. Agonistic display and dominance-
subordinance relations in adult male Schizocosa crassipes. Behavior, 62, 103–141.
Aspey, W. P. (1977b). Wolf spider sociobiology: II. Density parameters influencing
agonistic behavior in Schizocosa crassipes. Behaviour, 62, 142–162.
92 ANIMAL BEHAVIOR

Barnes, R. D. (1955). North American jumping spiders of the genus Maevia. Ameri-
can Museum Novitates, 1746, 1–13.
Bickford, D., D. J. Lohman, N. S. Sodhi, P. K. L. Ng, R. Meier, K. Winker, K. K.
Ingram, & I. Das. (2007). Cryptic species as a window on diversity and con-
servation. Trends in Ecology and Evolution, 22, 148–155.
Clark, D. L. (1994). Sequence analysis of courtship behavior in the dimorphic jump-
ing spider, Maevia inclemens. Journal of Arachnology, 22, 94–107.
Clark, D. L. & B. Biesiadecki. (2002). Mating success and alternative reproductive
strategies of the dimorphic jumping spider, Maevia inclemens (Araneae, Saltici-
dae). Journal of Arachnology, 30, 511–518.
Clark, D. L. & C. Morjan. (2001). Attracting female attention: The evolution of
dimorphic courtship displays in the jumping spider, Maevia inclemens (Ara-
neae: Salticidae). Proceedings of the Royal Society of London, B, 268, 2461–
2465.
Clark, D. L., J. A. Roberts, M. Rector & G. W. Uetz. (2011). Spectral reflectance
properties and visual communication in the wolf spider, Schizocosa ocreata
(Hentz): Simultaneous crypsis and background contrast in visual signals.
Behavioral Ecology and Sociobiology, 65, 1237–1247.
Clark, D. L., J. A. Roberts, & G. W. Uetz. (2012). Eavesdropping and signal match-
ing in visual courtship displays of spiders. Biology Letters, 8, 375–378.
Clark, D. L. & G. W. Uetz. (1990). Video image recognition by the jumping spider
Maevia inclemens. Animal Behaviour, 40, 884–890.
Clark, D. L. & G. W. Uetz. (1992). Morph independent mate selection in a dimor-
phic jumping spider: Demonstration of movement-bias in female choice using
video controlled courtship behaviour. Animal Behaviour, 43, 247–254.
Clark, D. L. & G. W. Uetz. (1993). Signal efficacy and the evolution of male dimor-
phism, Proceedings of the National Academy of Sciences, USA, 90, 11954–
11957.
Emerton, J. H. (196l). The Common Spiders of the United States. New York: Dover
Publications.
Endler, J. A. (1991). Variation in the appearance of guppy color patterns to guppies
and their predators under different visual conditions. Vision Research, 31,
587-608.
Endler, J. A. (1993). The color of light in forests and its implications. Ecological
Monographs, 63, 1–27.
Endler, J. A. (2006). Disruptive and cryptic coloration. Proceedings of the Royal Society
of London, B, 273, 2425–2426.
Evans, C. S., L. Evans, & P. Marler. (1993). On the meaning of alarm calls: Func-
tional reference in an avian vocal system. Animal Behaviour, 46, 23–38.
Evans, C. & P. Marler. (1991). One of the uses of video images as social stimuli in
birds: Audience effects on alarm calling. Animal Behaviour, 41, 17–26.
A TALE OF TWO SPIDERS 93

Evans, C. & P. Marler. (1992). Female appearance as a factor in the responsiveness of

male chickens during anti-predator behaviour and courtship. Animal Behav-
iour, 43, 137–145.
Fernandez-Juricic, E., N. Gilak, C. McDonald, P. Pithia, & A. Valcarcel. (2006). A
dynamic method to study the transmission of social foraging information in
flocks using robots. Animal Behaviour, 71, 901–911.
Fleishman, L. (1992). The influence of the sensory system and the environment on
motion patterns in the visual displays of anoline lizards and other vertebrates.
American Naturalist, 139, 36–61.
Gordon, S. D. & G. W. Uetz. (2011). Multimodal communication of wolf spiders on
different substrates: Evidence for behavioral plasticity. Animal Behaviour, 81,
367–375.
Hebets, E. A. (2003). Subadult experience influences adult mate choice in an
arthropod: Exposed female wolf spiders prefer males of a familiar pheno-
type. Proceedings of the National Academy of Sciences, USA, 100, 13390–
13395.
Hebets, E. A. (2005). Attention-altering signal interactions in the multimodal
courtship display of the wolf spider Schizocosa uetzi. Behavioral Ecology, 16,
75–82.
Hebets, E. A. & G. W. Uetz. (1999). Female responses to isolated signals from multi-
modal male courtship displays in the wolf spider genus Schizocosa (Araneae:
Lycosidae). Animal Behaviour, 57, 865–872.
Hebets, E. A. & G. W. Uetz. (2000). Leg ornamentation and the efficacy of courtship
display in four species of wolf spider (Araneae: Lycosidae). Behavioral Ecology
and Sociobiology, 47, 280–286.
Hebets, E. A. & C. Vink. (2007). Experience leads to preference: Experienced females
prefer brush-legged males in a population of syntopic wolf spiders. Behavioral
Ecology, 18, 1010–1020.
Hollander, J. D. (1970). Morphological variation in Pardosa prativaga L. Koch. 1870,
P. prativaga var. fulvipes Collet, 1875, and P. pullata Clerck, 1757 (Araneae,
Lycosidae). Tijdschrift voor Entomologie, 113, 283–290.
Hollander, J. den & H. Dijkstra, (1974). Pardosa vlijmi sp. nov., a new ethospecies
sibling Pardosa proxima (C.L. Koch, 1848), from France, with description of
courtship display (Araneae, Lycosidae). Beaufortia, 22, 57–65.
Hollander, J. den, H. Dijkstra, H. Alleman, & L. Vlijm. (1973). Courtship behavior
as a species barrier in the Pardosa pullata group. Tijdschrift voor Entomologie,
116, 1–22.
Jackson, R. R. (1982). The behavior of communicating in jumping spiders (Saltici-
dae). In P. N. Witt & J. S. Rovner (eds.), Spider Communication: Mechanisms
and Ecological Significance (pp. 213–247). Princeton, NJ: Princeton University
Press.
94 ANIMAL BEHAVIOR

Kaston, B. J. (1972). How to Know the Spiders. Dubuque, IA: Wm. C. Brown Co.
Kodric-Brown, A. & P. Nicoletto. (1997). Repeatability of female choice in the
guppy: Responses to live and videotaped males. Animal Behaviour, 54, 369–
376.
Kodric-Brown, A. & P. Nicoletto. (2001). Age and experience affect female mate
choice in the guppy (Poecilia reticulata). American Naturalist, 157, 316–323.
Macedonia, J. (2001). Habitat light, colour variation, and ultraviolet reflectance in
the Grand Cayman anole, Anolis conspersus. Biological Journal of the Linnean
Society, 73, 299–320.
Macedonia, J. M., Y. Brandt, & D. L. Clark. (2002). Sexual dichromatism and differ-
ential conspicuousness in two populations of common collard lizard (Crotophy-
tus collaris) from Utah and New Mexico, USA. Biological Journal of the Linnean
Society, 77, 67–85.
Macedonia, J. M., D. L. Clark, R. G. Riley, & D. J. Kemp. (2013). Species recogni-
tion of color and motion signals of Anolis graham: Evidence from responses to
lizard robots. Behavioral Ecology, 24, 846–852.
Macedonia, J. M., C. S. Evans, & J. B. Losos. (1994). Male Anolis lizards discriminate
video recorded conspecific and heterospecific displays. Animal Behaviour, 47,
1220–1223.
Macedonia, J. M. & J. A. Stamps. (1994). Species recognition in Anolis grahami (Sau-
ria, Iguanidae): Evidence from responses to video playbacks of conspecific and
heterospecific displays. Ethology, 98, 246–264.
Malkiel, E., J. Sheng, J. Katz, & J. R. Strickler. (2003). The three-dimensional flow
field generated by a feeding calanoid copepod measured using digital hologra-
phy. Journal of Experimental Biology, 206, 3657–3666.
Marchetti, K. (1993). Dark habitats and bright birds illustrate the role of the environ-
ment in species divergence. Nature, 362, 149–152.
McClintock, W. & G. W. Uetz. (1996). Female choice and preexisting bias: visual
cues during courtship in two Schizocosa wolf spiders (Araneae: Lycosidae). Ani-
mal Behaviour, 52, 167-181.
Narins, P. M., D. S. Grabul, K. K. Soma, P. Gaucher, & W. Hödl. (2005). Cross-
modal integration in a dart-poison frog. Proceedings of the National Academy
of the Sciences, USA, 102, 2425–2429.
Ord, T. J., R. A. Peters, C. S. Evans, & A. J. Taylor. (2002). Digital video playback
and visual communication in lizards. Animal Behaviour, 63, 879–890.
Ord, T. J. & J. A. Stamps. (2008). Alert signals enhance animal communication in
“noisy” environments. Proceedings of the National Academy of Sciences, USA,
105, 18830–18835.
Painter, T. S. (1913). On the dimorphism of the males of Maevia vittata. Zoology
Jahrbucher Anatomie, 37, 625–636.
Painter, T. S. (1914). Spermatogenesis in spiders. Zoology Jahrbucher Anatomie, 38,
509–576.
A TALE OF TWO SPIDERS 95

Partan, S. R., A. G., Fulmer, M. A. Gounard, & J. E. Redmond. (2010). Multimodal

alarm behavior in urban and rural gray squirrels studied by means of observa-
tion and a mechanical robot. Current Zoology, 56, 313–326.
Partan, S. R., C. P. Larco, & M. J. Owens. (2009). Wild tree squirrels respond with
multisensory enhancement to conspecific robot alarm behaviour. Animal
Behaviour, 77, 1127–1135.
Partan, S. R., P. Otovic, V. L. Price, & S. E. Brown. (2011). Assessing display vari-
ability in wild brown anoles Anolis sagrei using a mechanical lizard model. Cur-
rent Zoology, 57, 140–152.
Partan, S. R., S. Velda, V. Price, & T. Shimizu. (2005). Female pigeons, Columba
livia, respond to multisensory audio/video playbacks of male courtship behav-
iour. Animal Behaviour, 70, 957–966.
Peckham, G. W. & E. G. Peckham. (1889). Observations on sexual selection in spi-
ders of the family Attidae. Occasional Papers of the Wisconsin Natural History
Society, l, 3–60.
Peckham, G. W. & E. G. Peckham. (1890). Additional observations on sexual selec-
tion in spiders of the family Attidae. Occasional Papers of the Wisconsin Natural
History Society, l, 117–151.
Persons, M. A. & G. W. Uetz. (1996). The influence of sensory information on patch
residence time in wolf spiders (Araneae: Lycosidae). Animal Behaviour, 51,
1285–1293.
Persons, M. A. & G. W. Uetz. (1997). The effect of prey movement on attack behav-
ior and patch residence decision rules of wolf spiders (Araneae: Lycosidae).
Insect Behavior, 10, 737–752.
Persons, M. A. & G. W. Uetz. (2005). Sexual cannibalism and mate choice decisions
in wolf spiders: Influence of male size and secondary sexual characters. Animal
Behaviour, 69, 83–94.
Polak, M. (2003). Developmental Instability: Causes and Consequences. New York:
Oxford University Press.
Polak, M. & W. T. Starmer. (2001). The quantitative genetics of fluctuating asym-
metry. Evolution, 55, 498–511.
Roberts, J. A., E. Galbraith, J. Milliser, P. W. Taylor, & G. W. Uetz. (2006). Absence
of social facilitation of courtship in the wolf spider, Schizocosa ocreata (Hentz)
(Araneae: Lycosidae). Acta Ethologica, 9, 71–77.
Roberts, J. A. & G. W. Uetz. (2008). Discrimination of variation in a male
signaling trait affects optimotor response in visual predators. Ethology, 114,
557–563.
Rosenthal, G. G., C. S. Evans, & W. L. Miller. (1996). Female preference for
dynamic traits in the green swordtail, Xiphophorus helleri. Animal Behaviour,
51, 811–820.
Roster, N. O., D. L. Clark, & J. C. Gillingham. (1995). Prey catching behavior in
frogs and toads using video-simulated prey. Copeia, 1995, 496–498.
96 ANIMAL BEHAVIOR

Rovner, J. S. (1975). Sound production by nearctic wolf spiders: A substratum-

coupled stridulatory mechanism. Science, 190, 1309–1310.

Rowland, W. J. (1995). Do female stickleback care about male courtship vigor?

Manipulation of display tempo using video playback. Behaviour, 132, 951–
961.
Rowland, W. J., K. J., Bolyard, & A. D. Halpern. (1995a). The dual effect of stickle-
back nuptial coloration on rivals: Manipulation of a graded signal using video
playback. Animal Behaviour, 50, 267–272.
Rowland, W. J., K. J. Bolyard, J. J. Jenkins, & J. Fowler. (1995b). Video playback
experiments on stickleback mate choice: Female motivation and attractiveness
to male colour cues. Animal Behaviour, 49, 1559–1567.
Rutledge, J., A. Miller, & G. W. Uetz. (2010). Exposure to multiple sensory cues as a
juvenile affects adult female mate preferences in wolf spiders. Animal Behav-
iour, 80, 419–426.
Rutowski, R. L. (2003). Visual ecology of adult butterflies. In C. Boggs, W. Watt, &
P. Ehrlich (eds.), Ecology and Evolution Taking Flight: Butterflies as Model
Study Systems (pp. 9–25). Chicago: University of Chicago Press.
Scheffer, S. J., G. W. Uetz, & G. E. Stratton. (1996). Sexual selection, male morphol-
ogy, and the efficacy of courtship signalling in two wolf spiders. (Araneae:
Lycosidae). Behavioral Ecology and Sociobiology, 38, 17–24.
Schwartz, J. J., B. W. Buchanan, & H. C. Gerhardt. (2001). Acoustic interactions
among male gray treefrogs, Hyla versicolor, in a chorus setting. Behavioral Ecol-
ogy and Sociobiology, 53, 9–19.
Steyer, M. M., C. Stephan, J. Aringer, W. Berger, & U. Aust. (2012). The multi-
stimulus box: An innovative learning device for the comparative study of
object perception and recognition with various types of stimuli. Behavior
Research Methods, 44, 725–732.
Stratton, G. E. (1991). A new species of wolf spider, Schizocosa stridulans (Araneae,
Lycosidae). Journal of Arachnology, 19, 29–39.
Stratton, G. E. (1997). Investigation of species divergence and reproductive isolation
of Schizocosa stridulans (Araneae, Lycosidae) from Illinois. Bulletin of the British
Arachnological Society, 10, 313–321.
Stratton, G. E. (2005). Evolution of ornamentation and courtship behavior in Schiz-
ocosa: Insights from a phylogeny based on morphology (Araneae: Lycosidae).
Journal of Arachnology, 33, 347–376.
Stratton, G. E. & G. W. Uetz. (1981). Acoustic communication and reproductive
isolation in two species of wolf spiders (Araneae: Lycosidae). Science, 214,
575–577.
A TALE OF TWO SPIDERS 97

Stratton, G. E. & G. W. Uetz. (1983). Communication via substratum-coupled

stridulation and reproductive isolation in wolf spiders (Araneae: Lycosidae).
Animal Behaviour, 31, 164–172.
Stratton, G. E. & G.W. Uetz. (1986). The inheritance of courtship behavior and its
role as a reproductive isolating mechanism in two species of Schizocosa wolf spi-
ders (Araneae: Lycosidae). Evolution, 40, 129–141.
Taylor, P. W., J. A. Roberts, & G. W. Uetz. (2008). Foreleg autotomy reduces mat-
ing success of male Schizocosa ocreata wolf spiders. Journal of Insect Behavior,
21, 193–202.
Uetz, G. W., D. L. Clark, J. A. Roberts, & M. Rector. (2011). Effect of visual back-
ground complexity and light level on detection of visual signals of male Schiz-
ocosa ocreata wolf spiders by female wolf spider receivers. Behavioral Ecology and
Sociobiology, 65, 753–761.
Uetz, G. W. & G. E. Denterlein. (1979). Courtship behavior, habitat, and reproduc-
tive isolation in Schizocosa rovneri Uetz and Dondale. Journal of Arachnology, 7,
121–128.
Uetz, G. W. & C. D. Dondale. (1979). A new wolf spider in the genus Schizocosa
(Araneae: Lycosidae). Journal of Arachnology, 7, 86–88.
Uetz, G. W., W. J. McClintock, D. Miller, E. I. Smith, & K. K. Cook. (1996). Limb
regeneration and subsequent asymmetry in a male secondary sexual character
influences sexual selection in wolf spiders. Behavioral Ecology and Sociobiology,
38, 253–257.
Uetz, G. W. & S. Norton. (2007). Preference for male traits in female wolf spiders
varies with the choice of available males, female age and reproductive state.
Behavioral Ecology and Sociobiology, 61, 631–641.
Uetz, G. W., R. Papke, & B. Kilinc. (2002). Influence of feeding regime on male sec-
ondary sexual characters in Schizocosa ocreata (Hentz) wolf spiders (Araneae:
Lycosidae): Evidence for condition-dependence in a visual signaling trait. Jour-
nal of Arachnology, 30, 461–469.
Uetz, G. W. & J. A. Roberts. (2002). Multi-sensory cues and multi-modal communi-
cation in spiders: Insights from video/audio playback studies. Brain Behavior
and Evolution, 59, 222–230.
Uetz, G. W. & E. Smith. (1999). Asymmetry in a visual signaling character and sex-
ual selection in a wolf spider. Behavioral Ecology and Sociobiology, 45, 87–93.
Uy, J. A. C. & J. A. Endler. (2004). Modification of the visual background increases
the conspicuousness of golden-collared manakin displays. Behavioral Ecology,
15, 1003–1010.
Wiley, R. H. (1991). Associations of song properties with habitats for territorial
oscine birds of eastern North America. American Naturalist, 138, 973–993.
98 ANIMAL BEHAVIOR

Wrinn, K. M. & G. W. Uetz. (2007). Impacts of leg loss and regeneration on poten-
tial fitness in a wolf spider (Schizocosa ocreata). Canadian Journal of Zoology, 85,
823–831.
Wrinn, K. M. & G. W. Uetz. (2008). The effects of autotomy and regeneration on
prey capture rate, capture efficiency and sensory detection in Schizocosa ocreata
(Araneae, Lycosidae). Behavioral Ecology, 19, 1282–1288.
 4

Studying Lifelong Male-Male Partnerships

in a Tropical Bird
David B. McDonald

INTRODUCTION
Why do animals sometimes cooperate when we might expect them to com-
pete fiercely for access to resources? I first saw a long-tailed manakin (Chirox-
iphia linearis, family Pipridae) when I was still an undergraduate. On my
first day in Monteverde, Costa Rica, in what would become my study site,
my friend Nat Wheelwright showed me a pair of male long-tailed manakins
at their dance perch, and we watched the two brilliant red, black, and blue
males do their amazing backwards leapfrog dance over each other for an all-
green female. “Why do they do that?” I asked myself. The question is still a
focus of my research many years later. The cooperative display seems particu-
larly puzzling because these birds have a lek mating system. Lek mating sys-
tems are characterized by four major features (Bradbury, 1981): (1) Males do
not provide resources such as food-rich territories, nesting sites, or paternal
care that are valuable to breeding females. (2) Males aggregate in traditional,
clustered courts for display to females. (3) Females have at least some opportu-
nity to choose among the displaying males, apparently on the basis of their
morphological features (think peacocks) or their courtship displays (think sage
grouse [Centrocercus urophasianus], manakins, or birds of paradise). In many
species, female choice is completely unconstrained—males can do nothing to
coerce a female to stay and mate with them. (4) The variance in mating success
100 ANIMAL BEHAVIOR

is often extremely high. That is, some males have incredible mating success
(mating with 100 or more females in a season), while other males have little
or no success in attracting females to their display courts and subsequently
mating with them. Lek mating systems are widely distributed through the ani-
mal kingdom (Höglund & Alatalo, 1995); lek-mating species include bats,
birds in 13 different families, African antelope, iguanas, fish, and insects. In
most taxa, however, lekking is decidedly the minority mating system. Most
African antelope, for example, have more typical polygynous (but not lek)
mating systems, but two or three species form leks. Some of my colleagues
might suppose that the Old World birds of paradise (Paradisaeidae) have the
highest proportion of lek-mating species of any avian family. That is not the
case. Only approximately half the species of birds of paradise have a lek mating
system. Approximately 85 percent of the species in the neotropical manakins
(Pipridae) have a lek mating system, making manakins the most “lekkish”
family of animals by far. In most lek-mating species, including most mana-
kins, males display on individual courts and compete fiercely for mates. Why
then, in the long-tailed manakin, do males form lifelong, cooperative partner-
ships, sing in unison, and dance a coordinated dual-male backwards leapfrog
dance for females (video clip viewable on my website at http://
www.uwyo.edu/dbmcd/mcd.html)? That question has been the central focus
of my research for about 30 years. In this chapter, I will try to describe the
answers my collaborators and I have developed over the years and explore
some of the surprising tangents that have arisen along the way. For me, tan-
gents are the best aspect of my career as an academic—when I began graduate
school I would never have guessed that I would study hybridization in endan-
gered fish (McDonald et al., 2008) or that I would spend two months of a sab-
batical year in Amazonian Brazil telling two Brazilian students, in Portuguese,
that we were living the plot of a novel by the Nobel Prize–winning Colombian
novelist Gabriel Garcia Marquez.
To set the stage for why male long-tailed manakins cooperate, I will begin
with some features of the natural history that mold the behavioral ecology of
manakins in general and of long-tailed manakins in particular. I will then
interject something that almost never appears in scientific publications—the
vicissitudes of how scientists actually do what they do, and how the progress
of a manuscript is rarely an orderly linear process. The next section explores
the extraordinary life history of male long-tailed manakins, which begin their
reproductive stage late in life (almost always at age eight or older) and then
“bop till they drop.” Following that, I explore the phenomenon of male-
male cooperation, balanced with a discussion of the role of competitive inter-
actions. With an exploration of the complex social networks that link males to
STUDYING LIFELONG MALE - MALE PARTNERSHIPS IN A TROPICAL BIRD 101

each other, I bring together many of the threads that have driven my research
over the years. I conclude by arguing that most of science is in the journey, not
in the destination. If we do our job properly, we lay the groundwork for ways
that we, or future scientists, can bring novel threads together to increase our
understanding of the great patchwork of the natural world.

NATURAL HISTORY
Frugivory
Frugivory means eating fruit. Almost all the 51 or so species of manakins
(Remsen et al., 2012; Stiles & Skutch, 1989) eat only or largely fruit. Their
frugivory is crucial to almost everything else about their natural history. For
example, because fruit is a readily available resource year-round in the tropics,
the females are capable of raising the chicks (clutch size is almost always two)
without the aid of a male partner. Males, unable to attract a female by the lure
of their paternal care, instead fall back on self-advertisement via conspicuous
plumage or behavior, or both, to attract females as mating partners. Because
females are free to choose from among the males, and because one female’s
choice does not make the male less valuable for the next female, some males
can become very successful. When some males are much more successful than
others, we talk of a high variance in male mating success. Natural and sexual
selection are stronger when variance is high—that is, the difference in fitness
between those that succeed versus those that do not exerts a powerful force
acting to increase the frequency of those morphological and behavioral traits
that occur in the successful. The extravagant result is the stunning diversity
of plumages and display behaviors in this family of birds (Pipridae). All the
species occur in forest habitats. For most of the species, their habitats are wet
or moist forests, but a few species occur in much drier forests, notably the
long-tailed manakin and the unusual helmeted manakin (Antilophia galeata)
of the cerrado (savannah-like) habitat of central Brazil. Those dry forest habi-
tats play a central role in shaping the major differences in social behavior
between long-tailed manakins on the one hand and helmeted manakins on
the other.
Long-tailed manakins, like many frugivorous tropical birds (Wheelwright
et al., 1984), eat an astonishing variety of different fruits, with different fruit-
ing times, such that fruit are the major component of the feeding list year
round. Nat Wheelwright and colleagues documented 37 species of fruits in
the diet of long-tailed manakins in Monteverde (Wheelwright et al., 1984),
but over the years I collected many species of seeds under their dance perches
that were not on that list and am therefore confident that any complete list
102 ANIMAL BEHAVIOR

would be almost twice that large. Occasionally, seasonal scarcity may force
manakins to eat unripe, less nutritious fruit (Foster, 1977). Although females
begin feeding fruit to nestlings at a very early age, availability of insects to pro-
vide extra protein may also be important to producing the healthiest possible
nestlings. The seasonality of fruit availability, along with a flush of insects fol-
lowing the onset of rains in May, likely determines the time of peak nesting
effort by females and therefore the courtship season of the males.

Distribution and Taxonomy

Manakins are exclusively neotropical birds. They occur from southern
Mexico to northern Argentina. A recent phylogeny (Tello et al., 2009)
excludes some of the traditional genera (e.g., Piprites, Schiffornis, Neopipo,
Sapayoa), while retaining two genera that seem less obviously to belong to
the family (the tiny, drab Tyranneutes tyrant-manakins and the odd Neopelma
tyrant-manakins). The genus Chiroxiphia is most closely related to the genus
Antilophia, which includes the red-and-black helmeted manakin plus the spec-
tacular, recently discovered, and endangered Araripe manakin (Antilophia
bokermanni), whose males are a very striking combination of red, black, and
white. (You should search the Internet for images of these species or have field
guides for Brazil, Ecuador or Peru, and Costa Rica open while reading this
account.) The Antilophia manakins do not have a lek mating system. In the
seasonally dry, gallery forest habitats where they occur, males seem to be able
to defend linear territories along rivers or streams. It is fascinating that these
two closely related genera, both using habitats much drier than is the norm
for the family, should have such dramatically different mating systems.
Long-tailed manakins may be the extreme among lek-mating species, with
arguably the highest variance in mating success (more extreme disparity
between the most and least successful) of any lek-mating species, plus the
bizarre twist of the dual-male courtship display. Helmeted manakins, in con-
trast, seem to have reverted to a more typical passerine mating system based
on territorial defense. My suspicion is that the resources in the cerrado habitats
of the helmeted manakin are just compact enough to allow males the opportu-
nity to defend resources valuable to breeding females and therefore to engage
in the more sure strategy of securing at least one mate. For long-tailed mana-
kins, in contrast, the resources may simply be too scattered to be defendable.
Perhaps someone who reads this chapter will do the definitive study that
would help assemble the story of how a suite of ecological factors drove the
differences between these two genera.
My primary manakin research has focused on the long-tailed manakin,
which occurs from the Pacific slope of southern Mexico to northwestern Costa
STUDYING LIFELONG MALE - MALE PARTNERSHIPS IN A TROPICAL BIRD 103

Rica. I studied them at Monteverde, fairly close to the southern limit of their
range, and at an elevation of 1,300 meters, close to the top of their elevational
range. In southwestern Costa Rica, and extending to Colombia and
Venezuela, is the lance-tailed manakin (Chiroxiphia lanceolata). Males in that
species have a shorter extension of the central tail feathers, the feature that
gives both species their English name. Three other species in the genus
Chiroxiphia occur in South America (not until 2012 did I see my “missing”
Chiroxiphia, the Yungas manakin [Chiroxiphia boliviana]). The most different
in appearance is the blue manakin (Chiroxiphia caudata), whose males have
blue undersides and blue rump and tail feathers. Blue manakins are also
unusual in that males routinely dance for females in threesomes or even four-
somes, though they do not do the tightly coordinated unison song (Trainer &
McDonald, 1993) performed by male long-tailed manakins.

Courtship Display
Male long-tailed manakins have an elaborate vocal repertoire. In birds, the
term vocal or song repertoire often refers to subtle variations in male song
(Krebs & Kroodsma, 1980). In the long-tailed manakin, vocal repertoire refers
to a suite of functionally and sonically very different calls. If you are like me,
you might have difficulty hearing the difference between variants of the songs
of a male song sparrow (Melospiza melodia; Beecher et al., 1996). You would
not, however, have any difficulty recognizing the difference between the “teea-
moo” partner-attraction call given by an alpha male to call his beta partner and
the unison “toledo” song. (These song names are onomatopoetic, meaning the
names sound like the songs.) Males give a “witwitwit” call to synchronize just
before starting a bout of toledos—the avian equivalent of a band member
calling out “and a one, and a two . . . ” Waanh, doodoodoo, owngh, weeoo,
nyanhnyownyanyownh—almost anyone can learn seven or eight distinctly
different vocalizations given by males in different contexts. The repertoire
consists of at least 15 distinctly different sounds, each given in particular con-
texts. This very large repertoire is no surprise given the complex social relation-
ships among males. I will return to the idea of social complexity later. The
most important call is the dual-male toledo song, used to attract females to
the males’ dance perch. It is given in unison (with an approximately one-
tenth of a second offset between the partners). Males from successful leks give
toledos that are closer to each other in pitch than those of less successful males
(Trainer & McDonald, 1995). Presumably, like well-tuned violins in an
orchestra, males in better tune sound more attractive to listening females. A
very successful partnership, calling at the height of the season, can perform
as many as 1,800 toledos per hour and as many as a million over the course
104 ANIMAL BEHAVIOR

of a season. Males that perform more toledos attract more females as visitors to
their dance perch (McDonald, 1989). Once a female decides to visit a pair of
males, the partners descend to the dance perch, a low vine or branch usually
half a meter or so from the ground and roughly horizontal, with a diameter
slightly thicker than your thumb. Active dancers can polish the lichen and
even some of the bark from their heavily used perch. The males then begin
their dual-male backwards leapfrog dance. One male sidles toward the female,
who watches them from the perch. As the lead male nears the female, he rises
in an arc, turning slightly sideways, and his partner sidles toward the female.
The first male lands behind his partner and begins sidling forward, and they
perform a set of 20 to 100 backwards leapfrogs. The dance was a memorable
feature of David Attenborough’s Trials of Life series on public television.
The males alternate sets of leapfrog hops with periods of 10 to 60 seconds of
what biologist Paul Slud (1957) termed “butterfly flight.” The males fly slowly
with open-slotted wings in a manner somewhat reminiscent of the flight of
blue Morpho butterflies, which are common in the same habitats favored by
long-tailed manakins.
One of the many delightful things about male manakins is their resem-
blance to animated fruit. While the females are a sensible, cryptic green, well
suited to making them difficult to detect when incubating their nests, the
males tend to be attired in striking combinations of black, red, blue, yellow,
orange, and white. By no coincidence, such colors are typical of the fruits
eaten by male and female manakins (Wheelwright & Janson, 1985). For
example, in my study area, a favorite fruit is that of Citharexylum donnell-
smithii (Verbenaceae). This small tree often occurs at forest edges, and the
fruit hang in large, asynchronously ripening panicles. The ripe fruit are black,
while the unripe fruit are orange. The resulting orange and black bunches are
visible at a considerable distance. Other favored fruits are contrasting,
attention-grabbing red and black, or white and black, and some are blue.
Courtship signals may often reflect sensory biases on the part of the female
receivers. Female túngara frogs (Physalaemus pustulosus), for example, appear
to have a preexisting bias toward preferring male vocalizations with features
similar to those of their prey (Ryan et al., 1990; see Chapter 5). Some guppies
(Poecilia reticulata) eat orange fruit and the females are attracted to males with
orange patches (Rodd et al., 2002). Female manakins, because of their reliance
on visual cues for finding fruit, have a visual system that can be co-opted to
respond strongly to the suite of color combinations that occurs in the males.
In an evolutionary sense, therefore, males that don bright plumages are
exploiting a preexisting bias in the sensory systems of the females. Studies have
also suggested that the color of light in the forest may influence which colors
STUDYING LIFELONG MALE - MALE PARTNERSHIPS IN A TROPICAL BIRD 105

are favored in the plumage. That is, drier and wetter forests may have different
light environments that affect how well plumage contrasts, or how cryptic
it is, against the general forest background (Endler & Théry, 1996). In these
and many other ways the ecological context influences the behavior of
animals.
Natural history—knowing a lot about an organism in its natural environ-
ment—is sometimes undervalued (Greene, 2005) as the basis for academic
approaches to understanding the natural world. Most great biologists, how-
ever, have an initial passion for a particular organism or taxon that inspires
all the rest of their work, even when their impact extends far beyond that par-
ticular taxon. One of the founders of behavioral ecology, as a discipline, is
Edward O. Wilson, whose book Sociobiology (Wilson, 1975) synthesized ideas
about kin selection, the costs and benefits of behavior, and many other
ongoing topics in behavioral ecology. His driving passion is ants. In his last
book, Charles Darwin (1881) focused on earthworms and their many effects
on the natural world. Careful observations of the actions of earthworms led
him to conclude that they have a powerful influence on the turnover of soils
and the subsidence of rocks and buildings into the soil. Earthworms literally
shape the surface of the earth. When Europeans arrived in North America in
the sixteenth century, earthworms did not occur in the Northeast. Their intro-
duction had huge impacts (Bohlen et al., 2004) on the vegetation, leaf-litter
load, and other aspects of the landscape. I could continue with a long list of
biologists whose general contributions to science stem from a fascination with
watching their favorite animals in natural settings.

WHAT DOES A BEHAVIORAL ECOLOGIST ACTUALLY DO?

On the one hand, to read many scientific publications, one might think
that the main concept sprang fully formed from the brain of the author, who
produced every piece of evidence in a planned and orderly sequence. That
may (I doubt it) occur in some laboratory studies where the question is some
very specific sequence of mechanisms. On the other hand, one might be
tempted to think that a behavioral ecologist simply wanders around watching
animals behave and gathers together all the observations into an orderly pat-
terned whole without having any preconceived notions of how the world, or
the animals, should behave. That is roughly as likely as the proverbial monkeys
with typewriters. In this section, I will attempt to give a glimpse into how it
works from the inside. A disclaimer—no two scientists work in the same
way. My experiences are unlikely to be representative of behavioral ecologists
in general or of how you would approach your own science.
106 ANIMAL BEHAVIOR

One of the exciting things about graduate school is that, unlike in many
undergraduate curricula, one can sample and choose and assemble ideas for
oneself, rather than in the order, and with the emphasis, chosen by a professor.
From that melting pot of ideas and the discussions with other graduate stu-
dents, I began to think of ideas and questions that I could apply to the dual-
male lek dance of long-tailed manakins. Some of my early ideas now make
me cringe. In an early proposal, I planned to handicap the alpha with small
lead weights to see whether I could tire him out and reverse the alpha and beta
roles, assuming that the alpha role was mutable from dance to dance. I now
know that I would likely simply have messed up the performance history of
any teams I treated thus. Being alpha has little to do with being the male with
the most staying power in any individual dance. The alpha’s dominance over
his beta partner has been years in the making and is not subject to temporary
reversal. Some of my earlier ideas, therefore, were complete dead ends. Not
too long ago, though, I discovered a primitive hand-drawn social network of
the relations among the males in my study area that I had created many years
ago. At the time, I had never heard of social network analysis, and it would
be about 20 years until I picked up that thread, to the point that it became
the main focus of my research. All that wasted time! Why did I not stick with
the fascination of the complicated interactions? I would like to think that at
least some of the many different things I did in the interim were essential to
a proper perspective on social networks when I finally returned to them.
One of those skills was to (start to) learn how to see what happens.
A major and long-overdue change in law enforcement practice in recent
years has been decreased reliance on eyewitness evidence, which study after
study has shown to be extremely unreliable. It is very, very hard to see what
happens. I sometimes run a short experimental demonstration of this fact in
my classes. I do five or six simple things in rapid-fire order and ask students
to write down what happened and in what order. The range of responses is
staggering. Somewhere in there, the group as a whole catches most of the
events and their sequence, but almost no one ever comes close to capturing
the whole. One student notices that I waved at them just before I flicked off
the lights. Another remembers that I whistled four notes (the Beethoven’s
Fifth notes of a golden-crowned sparrow [Zonotrichia atricapilla] song) while
the lights were out. One student has the broad sequence correct but few of
the details. Others capture one event that caught their attention but com-
pletely fail to capture the overall flow. I like to think that I would do better
than they do, and I suspect that I would be a better-than-average eyewitness
because of my training, but even I suffer from selective attention, failure to
remember accurately, and many other failures in observation. I do have one
STUDYING LIFELONG MALE - MALE PARTNERSHIPS IN A TROPICAL BIRD 107

bit of evidence that I am good at remembering what I have observed. My eye

doctor once told me that very few of his patients are as consistent as I in
reporting which of a series of letters is in better or worse focus. I suspect my
years of trying to remember multiple events and storing them for later transfer
to written notes serves me well in remembering that the fuzziness of a
given letter is equivalent to that of a different one five steps back. Of course,
video and other techniques can help one recover the details of behavioral
sequences. Even there, though, a trained eye (or a gifted one) can pick up
details that even repeated viewing would never yield. Skill and experience in
observing behavior is one of the most important tools in a behavioral ecolo-
gist’s repertoire. How does all this relate to what I actually do when observing
birds in nature?
To observe long-tailed manakins, my coworkers and I erected small black
plastic blinds, just large enough to contain a stool, with a window that faced
the dance perch of the males. The blinds were generally 8 to 10 meters from
the dance perch, and we would sit in the blinds armed with a clipboard and
observation form, a tallywhacker (click counter) for counting toledo calls, bin-
oculars, and sometimes a spotting scope or video camera. Sitting in the blinds
seems somewhat reminiscent of the way someone is said to have described
trench warfare, “Long periods of utter boredom, interspersed with minutes
or hours of sheer terror.” One waits patiently in somewhat cramped, uncom-
fortable quarters for something to happen. Then, when something happens,
too much happens. Two males (“What are their bands? [colored plastic bands
are placed on each bird’s legs to identify it individually] Is that red on the right
and black on the left, or vice versa?”) begin dancing (“When did they start the
bout of dancing?”) for a female or females (“What are her bands? Is she excited
and moving or sitting and preening?”). One is trying to write, turn on the
video camera, catch the band codes, and remember the events immediately
prior to the dance (“Were the males calling? Did the males or female arrive
first?”). With practice, one learns to prioritize one’s actions, to register the
information in a standard manner (e.g., right leg first, left leg second, regard-
less of whether the bird is facing away from or toward you), and to store infor-
mation for however long is necessary. Try memorizing the license plates of
three cars that pass you next time you are in a car (definitely do not try this
if you are the driver). You will improve with practice. Practice really does make
perfect. One of my startling lessons in that regard was to learn to count beans
thrown down on a table. With practice, most people can estimate the number
of beans to within a couple of percent (so that if the actual total is 89, one’s
estimate is almost never smaller than 87 or larger than 91). One can then
apply that skill to counting flocks of birds on the wing.
108 ANIMAL BEHAVIOR

Is sitting in a blind when nothing happens an exercise in utter boredom? It

is rarely so for me. A few times, over the years, I saw a scaled antpitta (Gral-
laria guatimalensis; imagine a bird with almost no tail that hops on the ground
and resembles a baked potato on stilts), or a margay (Leopardus wiedii; a small
spotted cat), or a tayra (Eira barbara; large marten-like member of the weasel
family) while sitting in the blinds. I often saw coveys of black-breasted wood
quail (Odontophorus leucolaemus) from the blinds—a species that is the
nemesis of many tropical bird-watchers. Once, a covey bedded down for a
midday nap in the leaf litter all around my blind. Mouse opossums (Marmosa
murina) built day nests in the roof of the blind. The wind and light would
change the scene framed by the window. I could hear emerald toucanets
(Aulacorhynchus prasinus), white-breasted wood wrens (Henicorhina leucos-
ticta), orange-billed nightingale thrushes (Catharus aurantiirostris), and lesser
greenlets (Hylophilus decurtatus) calling somewhere in the distance. I watched
a spider build its web. The animals and plants came to me rather than I to
them. And yes, sometimes I was bored or restless or impatient (“I really need
to see the band code of that male that someone thought they spotted here
yesterday!”).
For every hour spent in the field, one spends several hours making sense of
what one saw. Penciled observation forms need to be transferred to electronic
form. Data need to be compiled, crosschecked, and validated, sifted for errors
and patterns. Being a behavioral ecologist is only rarely like being Crocodile
Dundee. If you do not like the compiling, the sifting, the thinking, and the
writing and rewriting and revising, then you probably do not want to become
a behavioral ecologist (or scientist). One really must enjoy the (apparently)
tedious process as well as the moments of intense activity in the field or the
rare moments when one has finally seen a project through to completion in
the form of a published piece of work. Those obviously rewarding moments
are important. Sometimes, though, the most exciting times are late at night
when a new thought takes hold and a series of previously unconnected pat-
terns all fall into place. Few events can rival the excitement of that sort of
“aha!” moment. When a lecture clicks and seems to catch the students and
me just right, or when I feel that a presentation at a scientific conference went
particularly well, it is often because I see what can or might happen in the
future, not because I am reveling in something accomplished or behind me.
I alluded earlier to going off on tangents. One of the greatest privileges of
being an academic is the freedom to pursue tangents. When I began working
with my postdoctoral fellow Dai Shizuka, I would never have guessed that we
would end up publishing papers (Shizuka & McDonald, 2012; McDonald &
Shizuka, 2013) on orderliness in dominance hierarchies, as viewed from a
STUDYING LIFELONG MALE - MALE PARTNERSHIPS IN A TROPICAL BIRD 109

social network perspective. We had intended to work on further aspects of the

social networks in manakins. We became fascinated by various aspects of the
story told by patterns in network data. We read about something called
the triad census—a way of quantifying the relations among animals, consid-
ered three at a time, and realized with growing excitement that it might pro-
vide new ways to look at dominance hierarchies in nature. The entire
process was a tangent from what we thought was our main goal. But, luckily,
we had the freedom to follow what we found exciting and to explore the chain
of ideas, data, and logic that wound up changing much of the focus of my
ongoing research. Recently, I wrote a proposal to do experimental studies on
dominance hierarchies in chickadees. The proposal may never be funded,
but I have never had more fun writing one. Because academics sometimes
grumble about having to write grant proposals as an unpleasant and distracting
necessity, one of my colleagues told me I was a “sick puppy” for having fun
writing a proposal—the accusation may be true (I like to excuse myself by say-
ing that anyone who was born in two places at once—Trinidad and Tobago—
is inevitably going to be psychologically challenged), but not because I found
it absorbing to write that grant proposal. But perhaps we should veer back
from this tangent about tangents. Let us go back to the main theme of this
chapter—the life rhythm of long-tailed manakins.

BOP TILL YOU DROP

Male long-tailed manakins rarely succeed in mating until they are old by
the standards of other birds of the same size (16 to 18 g), or indeed even com-
pared with much larger birds such as sage grouse (3,190 g; Dunning, 1993). I
once calculated that the mean age for a copulating male long-tailed manakin
was 10.1 years (McDonald, 1993b). Why do they wait so long? They have
no choice. Females are highly unanimous in their choices and, for their mating
partners, they pick males with long and proven records of performance. The
mates they select are therefore likely to be a small subset of the older males
in the population. Given the extremely low prospects for present reproductive
success, males should engage in strategies that enhance their prospects for
future success. In the case of long-tailed manakins, a host of ecological factors
combine to make cooperation the behavioral strategy with the highest fitness
payoff. Just because it is the best (or only viable) strategy does not mean that it
is good. It does not have to be a good option; it just has to be better than any
of the alternatives. To make this last point, more dramatically, I will say that
beta males seem to have a higher mortality rate than do alpha males. Why
would males work toward beta roles, if their prospects were not great even
110 ANIMAL BEHAVIOR

then? Again, they do so because no better alternative exists. Nevertheless, why

should beta males have mortality rates higher than those of the alphas? I think
the answer lies in the stringent hurdle raised by the change from status as a
gamma male (gamma is the third letter in the Greek alphabet) to status as a
beta male. Gamma and lower-ranking males do relatively little of the toledo
calling and dual-male dancing. When a male ascends to beta status, he sud-
denly moves from “hanging with the guys” to doing as many as a million
toledos, thousands of leapfrog hops, and the energy-intensive butterfly display.
Some males may simply not be able to take the energetic strain of that height-
ened level of performance. When they move from the beta role to the alpha
role, however, the change is much less drastic. In the happy event that they
are highly successful, they may have some extra solo butterfly display to per-
form for females but nothing like the extra effort required of a beta compared
to his relatively easy life as a gamma male.
Male long-tailed manakins have a life history that is almost semelparous.
Semelparity means waiting until near the end of the lifespan to reproduce
and then having little or no postreproductive lifespan. It is sometimes called
“big bang” reproduction. Well-known examples in the natural world include
Pacific salmon (of the genus Oncorhynchus), which deteriorate physiologically
in the transition from salt to fresh water, and some agaves (century plants).
Male long-tailed manakins verge, in some ways, on having a semelparous life
history. They must wait for years to reproduce and then do so for just a few
years until they die. I have no evidence that alphas live on beyond their time
as the top male. When I first thought about the life history of males and real-
ized that females were likely to begin reproducing much earlier (as early as the
year after hatch), I wondered whether females would have a much shorter life-
span. In line with my expectation, when I analyzed my survival data to com-
pare male and female life histories, I found that selection was indeed much
stronger on adult survival for males and on fertility and juvenile survival for
females (McDonald, 1993b). In matrix-based demographic analyses, assess-
ment of the relative importance of different vital rates (age-specific fertility
and survival) is called sensitivity or elasticity analysis (McDonald & Caswell,
1993). The importance of a vital rate is assessed by its impact on lambda ()
—the population growth rate. That is, sensitivity or elasticity assesses the
effect that a change in an annual survival or fertility rate would have on the
growth rate of the population (or of that individual’s potential lineage). For
females, their elasticity profile resembled that of small “dickey birds” such as
chickadees. Males, in contrast, had demographic profiles that resembled those
of very large, long-lived birds such as yellow-eyed penguins (Megadyptes
antipodes). Surprisingly, though, female annual survival rates differed relatively
STUDYING LIFELONG MALE - MALE PARTNERSHIPS IN A TROPICAL BIRD 111

little from those of males (McDonald, 1993b). That lack of survival difference
puzzled me until I though back to the natural history. Predation rates on long-
tailed manakin nests are disturbingly high. While watching nests I have twice
seen a predator take eggs (emerald toucanet) or nestlings (squirrel). Although
females can start reproducing early, therefore, it may take them years just to
replace themselves and their mate. Furthermore, females may well improve
at nesting over time. That improvement loops back to influence the payoff
to a beta male’s patient wait for alpha status. Betas will help attract young
females to their lek. Only when those females are older, and perhaps not until
the beta has ascended to the alpha role, will the females be likely to be success-
ful nesters. Thus a beta male’s current effort may, in fact, have a delayed payoff
that accrues mostly to him. Cooperation pays—eventually.

COOPERATION
The need to build a lek’s reputation for excellent performance is critically
important to explaining a beta male's behavior. Without the leg up of a repu-
tation, inherited from his alpha partner and maintained or strengthened by the
beta, a beta male would have little hope of becoming highly successful. I do
not think the success stems from learning the craft. The analogy I would use
is Michelangelo. Michelangelo was likely a better artist, almost from the begin-
ning, than was his master, Domenico Ghirlandaio, whose work is far less well
known than that of his famous pupil. Likewise, some male long-tailed mana-
kins may be far superior in endurance, singing ability, and dancing ability to
the males who precede them in their long climb to alpha status at a perch zone.
What Michelangelo needed from his apprenticeship was as much about bene-
fitting from the reputation and future potential patrons of the work as anything
to do with learning techniques. Likewise, males who wait patiently in line in
the leks of long-tailed manakins need to be part of an unbroken chain of perfor-
mance unmarred by conflicts that would drive away potential female visitors.
When I deliver talks on my manakin studies, people sometimes ask me how
alpha males choose their beta partners. They do not. Males secure their rank
by establishing their dominance over others, not because they are chosen as
partners by higher-ranking males. An anecdote about a particular beta male
is revealing. Early in my study, the beta male at the most successful lek disap-
peared between one season and the next. The gamma that ascended to the beta
male role had a defective syrinx. His contribution to the unison toledos
sounded awful. Nevertheless, he was the next most dominant male, and his
place in line was beyond dispute. If the problem of defective syrinx were at
all common, selection would doubtless have ensured that defective singers
112 ANIMAL BEHAVIOR

did not move up to beta, where their poor performance could hurt the reputa-
tion of the lek. And the reputation of that lek did, indeed, suffer. In the first
year of the new partnership, visitation suffered, but females that had copulated
there previously came back—the beta male’s dancing was as good as that of
any other beta. In subsequent years, however, the success of the lek declined
because fewer new female visitors were attracted to the lek. When he ascended
to the alpha role, the defective beta never really established a steady partner-
ship. Although anecdotes are not usually considered scientifically acceptable
evidence, I nevertheless find it illuminating. The cooperation is not based on
mutual choice of partner. It results from each male’s struggle to secure a place
in line at a lek where he has established relations and a rank relative to the
other contenders for a place in the queue.
Cooperative behavior is often explained as resulting from the genes shared
between the recipient of an altruistic (costly) act and the donor (Clutton-
Brock, 2002). Cooperation driven by the relatedness of the interactants is
called kin selection and has been used to explain cooperation in everything
from slime molds (Queller et al., 2002) to humans. More puzzling is when co-
operation occurs between unrelated individuals (Trivers, 1971; Axelrod &
Hamilton, 1981). In 1987, I began taking small blood samples from the birds
I captured in order to assess relatedness by analyzing DNA. My colleague
Wayne Potts and I found that the male partners in long-tailed manakins are
no more closely related to each other than are pairs of males selected at ran-
dom from the population (McDonald & Potts, 1994). For kin selection (indi-
rect inclusive fitness benefits) to operate, individuals must preferentially direct
their cooperative acts toward individuals to whom they are related. In long-
tailed manakins, rather than deriving a payoff from the genes shared with their
partners, beta males receive a direct, if long-delayed, benefit from the inherit-
ance of a reputation and a suite of site-faithful females. Females tend to return
for visits and copulations to perches they have chosen in the past. Their site
fidelity transcends the turnover of alpha males. Although I have no strong evi-
dence to support the claim—here is one of many examples of questions that
you or some future researcher might answer—I strongly suspect that the
females do not individually recognize the alpha males with whom they mate.
Instead, they choose to visit leks where the toledo output has been high over
the long term. They often visit several leks in any given season, but they seem
never to copulate at more than one (McDonald, 2010). They seem to choose
the perch at which they will copulate based on the performance of the dual-
male leapfrog dance (McDonald, 1989). They do not individually choose
the male mating partner—they let the males sort out who will remain at the
end of the dance. If they return for a subsequent copulation in the same or
STUDYING LIFELONG MALE - MALE PARTNERSHIPS IN A TROPICAL BIRD 113

subsequent seasons, and the male who remains to copulate is different, so be

it. All that really matters to them is the long chain of high-level performance.

NOT ALL IS SWEETNESS AND LIGHT

Although male long-tailed manakins cooperate in courtship display, they
also engage in a long, protracted competition with other males in order to gain
a place in the queue. One of the most interesting ways that this struggle for a
place in line manifests itself is in the progression of plumages through which
young males move (Doucet et al., 2007). Immediately after they hatch, males
are all green like the females. In the subsequent breeding season, males develop
red in the cap—either two strips at the outer edge of the crown or a full red
cap. Two years after hatch, they add black feathers in the face. Three years
after hatch they add more dark plumage to the body and develop the
powder-blue back feathers. In this blue-back stage, however, one can always
find some green or grayish feathers in the plumage. Not until four years after
hatch do they attain the fully red, black, and blue plumage, called the defini-
tive plumage. The fairly simple rule for the progression is, therefore, add red,
add black, add blue, take away green. No other species of manakin has this
long a delay in plumage maturation. The age specificity of the predefinitive
plumages means that human observers, or other male manakins, can tell at a
glance exactly how old a young male is simply from his plumage. These prede-
finitive plumages appear to act as signals of social status (McDonald, 1993a).
A predefinitive male is little or no threat to an alpha or beta male. He is, how-
ever, a potential rival for males close to him in age. In what I called a “manakin
mannequin” experiment, I discovered that males at the leks reacted much
more strongly and aggressively to males in definitive plumage than to males
in the predefinitive plumages. My experiment consisted of placing stuffed
males (taxidermic mounts) near the dance perches in the various leks in my
study area. I created the taxidermic mounts from the (luckily very rare) cases
in which a male died in the course of being caught in our mist nets. Mist nets
somewhat resemble a fine fishing net, with a sagging shelf that catches and tan-
gles birds as they fly through the forest. We use the mist nets to capture the
birds for blood samples and to fit them with individual combinations of col-
ored plastic bands and a numbered metal band that allows unequivocal identi-
fication of individuals. Males reacted in various ways to the mounts but, on
average, showed much stronger aggression toward definitively plumaged
mounts than against predefinitively plumaged mounts.
One of the conclusions from the mount experiments was that the predefini-
tive plumages seem to act as a signal to older males that the younger,
114 ANIMAL BEHAVIOR

predefinitive males are not a threat to their position in the age-graded queue
for social status (McDonald, 1989b). That reduced aggression means that they
are more readily tolerated as hangers-on at the leks, giving young males a
chance to become familiar to the other members of the lek. By the time they
are four and older, I presume, they are sufficiently familiar to be individually
recognized by the other members of the lek, both by their appearance and
by their individually distinct vocalizations. Although I have never amassed
the evidence necessary to publish a paper on individual recognition among
male long-tailed manakins, I have no doubt that it occurs, as it does in many
other species of birds. Another interesting feature of the aggressive interactions
among males is that the strongest response to the definitive male taxidermic
mounts was not by the alpha or beta males at the leks but by lower-ranking
definitive males. Any definitively plumaged intruder that managed to establish
himself at the lek would be a threat to the lower-ranking definitive males.
Thus the orderliness of the queues and the surprising lack of overt aggression
that one sees as the males interact is maintained by the self-interest of all the
members of the queue. A general problem for the maintenance of any orderly
queue is that those at the end of most queues have nothing to lose from a shake-
up (Maynard Smith, 1983). Following a shakeup they cannot be any further
back in the queue and might even be further ahead. In the case of long-tailed
manakins, however, even the males at the end of the queue have something to
lose from a shakeup—the good reputation of the lek. Leks where the males fight
with each other are unattractive to females. Female choice for good behavior
therefore enforces the patient, orderly waiting that characterizes the queues of
male long-tailed manakins. Males are constrained to be orderly because the value
of the queue lies not in some physically defendable resource but in an intangible
reputation based on high levels of performance and orderly behavior. Disorderly
behavior would diminish the very resource for which males are queuing.
Although top-ranking males engage in far more behavior that is cooperative
than that is overtly aggressive, one intriguing feature of their displays hints at
the underlying tension caused by the different self-interests of the partners.
At the end of some dance bouts, whether in the presence of a female or not,
the males accelerate the pace of the leapfrogs and end with a sharp “buzz
weent” vocalization given by the dominant male and directed at his subordi-
nate partner. Alpha males fairly often direct these at their beta partners, betas
at gammas, and so forth. During dances for females the beta male often leaves
after a buzz weent, leaving the alpha male to continue, perhaps culminating in
his succeeding in copulating with the watching female.
The buzz weents do not serve to establish dominance; instead they
reinforce or express it. That is, males seem not to direct a buzz weent at
STUDYING LIFELONG MALE - MALE PARTNERSHIPS IN A TROPICAL BIRD 115

another male until they have already established a clearly dominant role. I
therefore long thought that males would never give mutual buzz weents. The
one exception provides a very interesting insight into the normal dynamics
of queuing in long-tailed manakins. In 1991, the alpha, beta, and gamma
males at the most successful lek in my study area all disappeared in quick suc-
cession. The result was a power vacuum. The lek continued to attract females
that had visited or copulated in prior years, but it was not clear which males
would assume the alpha and beta roles. Males from adjacent leks began spend-
ing time at the dance perch where the power vacuum provided a powerful
incentive to striving for the vacant alpha role. For the first and only time in
nearly 20 years of observation, I saw grappling fights between males. In look-
ing back at my data on buzz weents, I realized that mutual buzz weents had
occurred. I think that what had happened was that the power vacuum drew
in males from adjacent leks, which had current or prior ties to one or other
of the males at the lek. Some of these males were older, high-status males, per-
haps dominant to some or all of the lower-ranking males at the contended lek.
In the absence of the prior alpha, beta, and gamma, these neighboring males
had an opportunity to vie for the alpha or beta role. Critically, however, some
of these males may never have interacted directly with other contenders from
leks in a different direction from the contested lek. As a result, they had never
established their relative dominance. Each of them therefore “felt” that he was
dominant enough to direct a buzz weent at the other. In the normal, orderly
queues, all the males have clearly established relative rankings. Only in the
highly unusual case of a rapid loss of the top of the hierarchy did the estab-
lished system of orderliness break down. The contenders, some of them unfa-
miliar with each other, now had a clear and present benefit for being aggressive
because females were continuing to visit and copulate at fairly high rates. The
benefit to current strife outweighed the potential long-term cost of loss of
female patronage. Eventually, a male from one of the adjacent leks became
established as the new alpha. He was reasonably successful in his first season
as alpha. By the next season, however, the mayhem had taken its toll on the
reputation of the lek. Female visitation dropped off, as did copulatory success;
the lek never reestablished itself as the most successful in my study area.

COMPLEX SOCIAL NETWORKS

Social network analysis is a set of mathematical and visual methods for
exploring complicated connections among entities. The entities (called nodes)
can be anything from individual humans or animals to companies or airports.
The connections (edges) that link the entities can be relations such as
116 ANIMAL BEHAVIOR

cooperative behavior, dominant/subordinate relations, movement of goods in

a transportation network, or information flow across the worldwide web. I
have recently begun to use social network analysis to understand the complex
relationships among male long-tailed manakins. Why use social networks?
First, networks allow one to analyze interactions (say between A and B) in
the larger context of other interactions (A also interacts with C, D, and E,
whereas B interacts with E, F, and G). Sometimes indirect interactions may
be very important in explaining the behavior of an individual, even though
the indirectly connected individuals never interact directly. Second, the wealth
of different metrics available allows one to quantify precisely what might oth-
erwise be vague qualitative impressions of the social system. Thus, a network
perspective takes into account the whole system of interactions, all within a
very well developed framework of quantitative techniques with roots in graph
theory and linear algebra. Network analyses have a long history in the social
sciences (Granovetter, 1973; Wasserman & Faust, 1994). More recently, they
have been of interest to physicists (Watts & Strogatz, 1998), behavioral ecolo-
gists (Croft et al., 2008), and ecologists studying phenomena such as food
webs (Bascompte, 2009).
One of the first ways in which I used social network analyses was to look at
how a male’s early interactions helped predict his later status (McDonald,
2007). What I found confirmed a hunch I had long held—that a male’s early
career has long-lasting implications for his social trajectory. The long delay in
plumage maturation, and the signaling function of the predefinitive plumages,
strongly suggested that a male’s early years were critical to his later social tra-
jectory. One of the metrics available in the social network repertoire is called
information centrality. Information centrality assesses how often a male
(node) lies along paths through the network that connect other pairs of indi-
viduals. An individual that lies along many paths linking others will have high
information centrality—he is well connected. I found that males who were
well connected when young (say between the ages of two and seven) were five
times more likely to rise in social status, an average of five years later, than were
males who were less well connected as youngsters. Interestingly, a male’s cur-
rent centrality was not correlated with his rise. That is, knowing how well con-
nected a male was five years ago is a good predictor of his being a “riser,” but
his current connectivity is not a good predictor of whether or not he is a riser.
Indeed, successful alpha males tended to be less well connected than were
other males. I sometimes (loosely) use an academic analogy. For graduate stu-
dents, postdoctoral fellows, and new faculty members, being well connected
and attending conferences, making connections with academics at other insti-
tutions, and interacting with peers and leaders at conferences may be critical to
STUDYING LIFELONG MALE - MALE PARTNERSHIPS IN A TROPICAL BIRD 117

their professional prospects. Established professors with tenure may not need
to engage in such activities (even if they did so years ago).
Within a social network one can use the number of links that separate two
individuals as a measure of social distance. This social distance is perhaps most
familiar in the concept of “six degrees of separation” (Watts, 2003), the idea
that any two people can be connected by a chain of six links (A knows B,
who knows C, etc.). Six degrees of separation is the basis for the Kevin Bacon
game, where one attempts to link a given actor to Bacon by naming a chain of
movies in which they co-occurred with another actor that then acted with
another, such that the chain eventually links the original actor to Bacon. That
concept of degrees of separation, or what I call social distance, leads naturally
to the possibility that social distance is somehow related to the physical (geo-
graphic) distance between potential interactants or the genetic distance (relat-
edness or lack thereof) between potential interactants. I looked at precisely
those relations in comparing the social distance between male long-tailed man-
akins to the distance between the leks with which they were affiliated
(McDonald, 2009). I excluded males who had interacted directly (edge dis-
tance of one, geographic distance of zero) because, by definition, males that
interact have to have co-occurred at the same location at the same time. As
one might expect, for males separated by a social distance of two or greater,
the physical proximity of their respective major lek affiliations was a good pre-
dictor of their social distance. No such relation existed between social distance
and relatedness. That is, males that were socially close were no more likely to
be closely related than males that were socially distant. As with almost all social
networks, the manakin social network is a small world (Watts & Strogatz,
1998)—no two males are very many links away from each other.

CONCLUSION
Even beginning to assemble the pieces of the puzzle of why male long-tailed
manakins should display cooperatively, why they have such long lifespans, and
how they can bop till they drop has required a suite of approaches from
genetic analyses to demographic modeling and, more recently, social network
approaches. A huge number of questions remain. What hormonal mecha-
nisms underlie the tension between cooperation and conflict in the social
interactions of males? How do females benefit from choosing males that sing
and dance well? Furthermore, if most females choose the same alpha male as
a mating partner, how can they benefit at all? I think part of the answer to
the benefit to females lies in a similarity to fads. If everyone in school is wear-
ing the same hat or shoe style, one does not become cooler than others by
118 ANIMAL BEHAVIOR

wearing that very same style. One does, however, avoid the terrible fate of
being one of the very few who is demonstrably less cool. Likewise, a female
who chooses the best male does not do better than other females that make
the same choice, but she avoids doing worse. Compiling the evidence to show
that any other choice would be worth less would be a daunting task. Even
more exciting than the unanswered questions are the questions we have not
even thought to ask yet and the unquestioned assumptions that may some
day be shown to be unwarranted. Truly creative science consists of putting
together ideas that no one has previously thought to put together and then
linking them with elegant logic and evidence. I hope that some of you will
have the privilege of feeling and stimulating that sort of excitement.

REFERENCES AND SUGGESTED READING

Axelrod, R. & W. D. Hamilton. (1981). The evolution of cooperation. Science, 211,
1390–1396.
Bascompte, J. (2009). Disentangling the web of life. Science, 325, 416–419.
Beecher, M. D., P. K. Stoddard, S. E. Campbell, & C. L. Horning. (1996). Reper-
toire matching between neighbouring song sparrows. Animal Behaviour, 51,
917–923.
Bohlen, P. J., S. Scheu, C. M. Hale, M. A. McLean, S. Migge, P. M. Groffman, & D.
Parkinson. (2004). Non-native invasive earthworms as agents of change in
northern temperate forests. Frontiers in Ecology and the Environment, 2, 427–435.
Bradbury, J. W. (1981). The evolution of leks. In R. D. Alexander & D. W. Tinkle
(eds.), Natural Selection and Social Behavior: Recent Research and New Theory
(pp. 138–169). New York: Chiron Press.
Clutton-Brock, T. H. (2002). Breeding together, kin selection and mutualism in co-
operative vertebrates. Science, 296, 69–72.
Croft, D. P., R. James, & J. Krause. (2008). Exploring Animal Social Networks.
Princeton, NJ: Princeton University Press.
Darwin, C. R. (1881). The Formation of Vegetable Mould through the Action of Worms,
with Observations on Their Habits. London: John Murray.
Doucet, S. M., D. B. McDonald, M. S. Foster, & R. P. Clay. (2007). Plumage devel-
opment and molt in long-tailed manakins (Chiroxiphia linearis): Variation
according to sex and age. Auk, 124, 29–43.
Dunning, J. B., Jr. (ed.). (1993). CRC Handbook of Avian Body Masses. Boca Raton,
FL: CRC Press.
Endler, J. A. & M. Théry. (1996). Interacting effects of lek placement, display behav-
ior, ambient light, and color patterns in three neotropical forest-dwelling birds.
American Naturalist, 148, 421–452.
Foster, M. S. (1977). Ecological and nutritional consequences of food scarcity on a
tropical frugivorous bird and its fruit source. Ecology, 58, 73–85.
STUDYING LIFELONG MALE - MALE PARTNERSHIPS IN A TROPICAL BIRD 119

Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology, 6,

1360–1380.
Greene, H. W. (2005). Organisms in nature as a central focus for biology. Trends in
Ecology and Evolution, 20, 23-27.
Höglund, J. & R. V. Alatalo. (1995). Leks. Princeton, NJ: Princeton University Press.
Krebs, J. R. & Kroodsma, D. E. (1980). Repertoires and geographical variation in
bird song. In J. S. Rosenblatt, R. A. Hinde, C. Beer, & M. C. Busnel (eds.),
Advances in the Study of Behavior, Vol. 11 (pp. 143–177). New York:
Academic Press.
Maynard Smith, J. (1983). Game theory and the evolution of cooperation. In D. S.
Bendall (ed.), Evolution from Molecules to Men (pp. 445–456). Cambridge,
UK: Cambridge University Press.
McDonald, D. B. (1989). Correlates of male mating success in a lekking bird with
male-male cooperation. Animal Behaviour, 37, 1007–1022.
McDonald, D. B. (1993a). Delayed plumage maturation and orderly queues for sta-
tus: A manakin mannequin experiment. Ethology, 94, 31–45.
McDonald, D. B. (1993b). Demographic consequences of sexual selection in the
long-tailed manakin. Behavioral Ecology, 4, 297–309.
McDonald, D. B. (2007). Predicting fate from early connectivity in a social network.
Proceedings of the National Academy of Sciences, USA, 104, 10910–10914.
McDonald, D. B. (2009). Young-boy networks without kin clusters in a lek-mating
manakin. Behavioral Ecology and Sociobiology, 63, 1029–1034.
McDonald, D. B. (2010). A spatial dance to the music of time in the leks of long-
tailed manakins. In R. Macedo (ed.), Advances in the Study of Behavior, Vol
42. Behavioral Ecology of Tropical Animals (pp. 55–81). San Diego, CA:
Academic Press.
McDonald, D. B. & H. Caswell. (1993). Matrix methods for avian demography. In
D. Power (ed.), Current Ornithology, Vol. 10 (pp. 139–185). New York:
Plenum Press.
McDonald, D. B., T. L. Parchman, M. R. Bower, W. A. Hubert, & F. J. Rahel.
(2008). An introduced and a native vertebrate hybridize to form a genetic
bridge to a second native species. Proceedings of the National Academy of
Sciences, USA, 105, 10842–10847.
McDonald, D. B. & W. K. Potts. (1994). Cooperative display and relatedness among
males in a lek-mating bird. Science, 266, 1030–1032.
McDonald, D. B. & D. Shizuka. (2013). Comparative transitive and temporal order-
liness in dominance networks. Behavioral Ecology, 24, 511–520.
Queller, D. C., E. Ponte, S. Bozzaro, & J. E. Strassmann. (2002). Single-gene green-
beard effects in the social amoeba Dictyostelium discoideum. Science, 299,
105–106.
Remsen, J. V., Jr., C. D. Cadena, A. Jaramillo, M. Nores, J. F. Pacheco, J. Pérez-
Emán, M. B. Robbins, F. G. Stiles, D. F. Stotz, & K. J. Zimmer. (2012). A
classification of the bird species of South America. Version 2, April 2012.
120 ANIMAL BEHAVIOR

American Ornithologists’ Union. http://www.museum.lsu.edu/~Remsen/

SACCBaseline.html.
Rodd, F. H, K. A. Hughes, G. F. Grether, & C. T. Baril. (2002). A possible non-
sexual origin of mate preference: Are male guppies mimicking fruit? Proceedings
of the Royal Society of London, B, 269, 475–481.
Ryan, M. J., J. H. Fox, W. Wilczynski, & A.S. Rand. (1990). Sexual selection for sen-
sory exploitation in the frog Physalaemus pustulosus. Nature, 343, 66–67.
Shizuka, D. & D. B. McDonald. (2012). A social network perspective on measure-
ments of dominance hierarchies. Animal Behaviour, 83, 925–934.
Slud, P. (1957). The song and dance of the long-tailed manakin. Auk, 74, 333–339.
Stiles, F. G. & A. F. Skutch. (1989). A Guide to the Birds of Costa Rica. Ithaca, NY:
Cornell University Press.
Tello, J. G., R. G. Moyle, D. J. Marchese, & J. Cracraft. (2009). Phylogeny and
phylogenetic classification of the tyrant flycatchers, cotingas, manakins, and
their allies (Aves, Tyrannides). Cladistics, 25, 429–467.
Trainer, J. M. & D. B. McDonald. (1993). Vocal repertoire of the long-tailed mana-
kin and its relation to male-male cooperation. Condor, 95, 769–781.
Trainer, J. M. & D.B. McDonald. (1995). Singing performance, frequency matching
and courtship success of long-tailed manakins (Chiroxiphia linearis). Behavioral
Ecology and Sociobiology, 37, 249–254.
Trivers, R. L. (1971). The evolution of reciprocal altruism. Quarterly Review of Biol-
ogy, 46, 35–57.
Wasserman, S. & K. Faust. (1994). Social Network Analysis. Cambridge, UK: Cam-
bridge University Press.
Watts, D. J. (2003). Six Degrees: The Science of a Connected Age. New York: W.W.
Norton.
Watts, D. J. & S. H. Strogatz. (1998). Collective dynamics of “small-world” net-
works. Nature, 393, 440–442.
Wheelwright, N. T., W. A. Haber, K. G. Murray, & C. Guindon. (1984). Tropical
fruit-eating birds and their food plants: A survey of a Costa Rican lower mon-
tane forest. Biotropica, 16, 173–192.
Wheelwright, N. T. & C. H. Janson. (1985). Colors of fruit displays of bird-dispersed
plants in two tropical forests. American Naturalist, 126, 777–799.
Wilson, E. O. (1975). Sociobiology: The New Synthesis. Cambridge, MA: Harvard
University Press.
 5

Be Loved, Be Prey, Be Eaten

Rachel A. Page, Michael J. Ryan, and Ximena E. Bernal

INTRODUCTION
Animal communication is one of the most fundamental of all social behaviors.
It modulates interactions among neighbors and strangers, siblings and parents,
and individuals and their prospective mates. One of the most fundamental
functions of communication is to enhance the sender’s conspicuousness, to
cause the sender to stand out against the background so it can be detected
by the receiver. This is a challenge for signals that function over long distances,
such as those that function in territorial advertisement and mate attraction,
because signal intensity and fidelity decrease with distance from the sender.
At longer distances a signal is less likely to be above the receiver’s threshold
for detection and recognition, and it is also more likely to be masked by noise
as the intensity of noise at the receiver is independent of its distance from the
sender.
Another type of noise is generated by conspecifics signaling in the same
channel. We may perceive a chorus of frogs or insects as a melodious, even co-
operative unit. But to members of the chorus the calls of their neighbors are
every bit as deleterious to their own call’s ability to be detected as is environ-
mental noise. There is also “noise” within the perceptual systems of the
receiver that ameliorates signal detection. One example is habituation. Upon
continual exposure to a signal, an animal will tend to ignore it, and some of
the animal’s neurons will cease to fire. Another type of perceptual noise is
incremental forgetting of a signal once it is perceived. Some signals are more
122 ANIMAL BEHAVIOR

memorable than others, surviving longer in the receiver’s memory. The

importance of conspecific noise and perceptual noise is usually less appreciated
in animal communication studies than is environmental noise.
Selection often will favor senders to produce communication signals that
stand out against these three types of background noise. Conspicuousness is
accomplished in a variety of ways. In the acoustic domain animals can use fre-
quency bands that contain less noise; they can call longer, at a faster rate, or at
a higher amplitude; they can call during periods when others are silent; and
they can produce calls that are more complex. Visual signals can stand out
more against background when their spectral properties and spatial patterns
differ from those in the background, when colors are brighter and patterns
are more complex, and when motion patterns associated with the signal, such
as push-up displays in lizards, are different than the pattern of background
movement, such as vegetation blown by the wind (e.g., Fleishman, 1992).
Increasing signal complexity in both the acoustic and visual domains can to
some extent remedy the receiver habituating to and forgetting a signal. There
is less known about how chemical, tactile, and electrical signals can enhance
contrast with the background.
In a survey of preferences for sexual signals, Michael Ryan and Anne
Keddy-Hector (1992) showed that across modalities prospective mates
tended to prefer sexual signals that were greater in magnitude: longer and
louder, faster and brighter, more complex and with more background
contrast. Although there were numerous cases in which prospective mates
preferred signals that were near the population mean, there were few cases
in which there was a preference for signals of lesser magnitude. A simple
rule of thumb for sexual signals is that more is better. Selection for signal
efficacy, however, creates a dilemma for the sender: the curse of unintended
receivers. In general, each signal has evolved under selection to communi-
cate with a specific intended receiver. In the case of sexual advertisement sig-
nals, the intended receivers are prospective mates, usually female
conspecifics. Other receivers can also detect and respond to signals even if
there was no selection on the sender to communicate with them. We refer
to these receivers as eavesdroppers or unintended receivers. It is important
to note that the terms intended and unintended should not imply intention-
ality but instead refer to hypotheses about the selection forces that favored
the evolution of these signals. With few exceptions, senders and receivers
do not communicate in a private channel. The world is populated by unin-
tended receivers, eavesdroppers who are attendant to the signals of senders,
often to the demise of senders. In this chapter we explore the tension that
exists in sexual communication systems between being conspicuous to
BE LOVED , BE PREY , BE EATEN 123

potential mates and attracting eavesdroppers, between being loved, being

prey, and being eaten.

UNINTENDED RECEIVER-SIGNALER INTERACTIONS

Eavesdropping predators and parasites exploit the communication systems
of their hosts or prey, often by intercepting their mating signals, which they
use to localize and attack the signaler. Marlene Zuk and Gita Kolluru (1998)
showed that exploitation of sexual signals by eavesdroppers is rampant across
taxa and sensory modalities. They report 19 cases of predators, parasites,
and parasitoids, which exploit sexual signals produced by other species. While
eavesdropping insects are usually parasites and parasitoids exploiting other spe-
cies of insects, vertebrate eavesdroppers are most frequently predators that feed
on diverse invertebrate and vertebrate prey. Although several of these instances
of interspecific eavesdropping are opportunistic, eavesdroppers can be highly
specialized, depending on signal exploitation for mating or survival.
By killing or parasitizing signaling males, eavesdroppers impose strong
selective pressures on their hosts or prey and thus influence the evolution of
sexual signals as well as male signaling behavior. In some cases, signalers use
private communication channels inaccessible to signal-oriented predators
(Stoddard, 1999; Théry & Casas, 2002; Cummings et al., 2003), but most
frequently, the sensory sensitivity of eavesdroppers and their hosts or prey
overlap, and signalers must use other evasive strategies to reduce risks of
exploitation. In many instances, the response of signaling males to the pres-
ence of eavesdroppers is to stop signaling or to signal more cautiously (Tuttle
et al., 1982; Rand et al., 1997; Hedrick, 2000; Dapper et al., 2011). Other
strategies to reduce attacks include changes in the structure of the mating calls
(Zuk et al., 1993; Müller & Robert, 2001) as well as seasonal or diel shifts in
signal production, which temporally separate eavesdroppers and their hosts or
prey (Burk, 1982; Zuk et al., 1993; Cade et al., 1996). In some species, males
resort to not producing signals at all but instead take advantage of signaling
conspecifics by silently waiting nearby and intercepting approaching females.
In a review of such satellite behavior in anurans, Carl Gerhardt and Mark
Huber (2002) showed that age, size, physical condition, signal ability, and
density of calling males predict whether a male frog will wait silently by a sig-
naling one. In extreme cases, for instance in a population of field crickets para-
sitized by eavesdropping flies (Zuk et al., 2006), the majority of males may
lose the ability to produce their mating song.
Even though males usually pay the costs imposed by eavesdropping preda-
tors and parasites, females sometimes suffer as well. In decorated crickets,
124 ANIMAL BEHAVIOR

males call from the safety of burrows, protected from the attacks of predatory
geckos, which are attracted to their calls. As females approach the burrows, the
geckos intercept and attack them (Sakaluk & Bellwood, 1984). Overall, how-
ever, males are more heavily exploited by eavesdroppers than females. To be
loved, males risk being preyed upon and being eaten.

THE FROG CHORUS AS A COMMUNICATION MECCA

Nowhere is the importance of eavesdroppers in the evolution of communi-
cation better understood than in sexual communication, and in no system is
this better understood than in túngara frogs (Physalaemus pustulosus) and the
bats that eat them and the midges that feed on their blood (Figure 5.1).
Most male frogs produce conspicuous calls to attract females. The calls are
species specific in the sense that the variation among calls within the species is
far less than that among species. The calls usually have a dual function: they

Figure 5.1. Calls of male túngara frogs, Physalaemus pustulosus, attract both intended
receivers, female túngara frogs, and unintended receivers such as the predatory bat,
Trachops cirrhosus, and parasitic midges, Corethrella spp. (Bat photo by Alexander
Lang, frog photos by Alexander T. Baugh and Kathrin Lampert, midge photos by
Ximena Bernal, composite by Michael Teague O’Mara. Animals not to scale.)
BE LOVED , BE PREY , BE EATEN 125

maintain interindividual distances among males, and they attract females.

Reproductively active females of most species move toward calling males.
Conveniently, females show this same behavior in the laboratory in response
to calls broadcast from speakers. Female phonotaxis is a robust indicator of
call preference, and mate-choice phonotaxis experiments allow us to rigorously
document female preferences for various types of calls (Gerhardt & Huber,
2002; Wells, 2007).
Females are attracted to conspecific calls in preference to heterospecific
ones. There is a fairly good understanding of the biases in the auditory system
that result in conspecific call preferences. In addition to their preferences for
conspecific calls, females of many species are more attracted to the calls of
some conspecific males over others. This preference generates sexual selection
on calls as some males have greater reproductive success because they are more
likely to be chosen by a female as a mate. The details of the female’s call pref-
erences vary among species, but various studies have shown preferences for
calls that are of greater amplitude and duration, produced at a faster rate, are
of lower frequency, and are more complex (Ryan, 2009).

THE SENDER, THE MALE TÚNGARA FROG

Sexual selection and communication is better understood in túngara frogs
than in almost any other system (Ryan, 2010). These frogs are unusual for
anurans in that the mating call is one of varying complexity. All túngara frog
mating calls have a fundamental frequency that sweeps from about 900 to
400 Hz in about 300 milliseconds (Figure 5.2). The frequency sweep, or
“whine,” has several harmonics, but about half of the call’s energy is in the
fundamental frequency. The dominant frequency is about 700 Hz. A mating
call can consist of the whine only, or there can be an additional component
consisting of shorter bursts of sounds, “chucks,” added to the end of the
whine. Calls with chucks are referred to as complex calls. A chuck has a funda-
mental frequency of about 220 Hz with substantial energy in each of the 15
harmonics of the fundamental. The dominant frequency of a typical chuck is
about 2,500 Hz with a duration of approximately 35 milliseconds. A túngara
frog mating call can contain from zero to seven chucks.
The complex call of the túngara frog is an unusual acoustical feat, and males
accomplish this with an unusual larynx (Figure 5.3). Frogs produce calls by
passing air from the lungs through the larynx, where it vibrates a pair of vocal
folds and produces the pressure fluctuations that are perceived as sound. Many
species of frogs have local thickenings or small fibrous masses on the vocal
folds, which lower the frequency of vocal fold vibration and of the resulting
126 ANIMAL BEHAVIOR

Figure 5.2. Graphical depictions of túngara frog mating calls with zero, one, two, and
three chucks (from top, left to right, bottom, left to right). Waveforms are shown
above; spectrograms below.

sound. Túngara frogs have a pair of pendulous fibrous masses that hang from
the vocal folds and protrude into the passageway that connect the lungs to the
larynx.
All of the close relatives of the túngara frogs produce whine-like mating
calls, but only the túngara frog and its sister species P. petersi augment the
whine with chucks. All populations of túngara frogs studied have males that
produce complex calls, and there are no known cases in which a male was
not able to produce a complex call. P. petersi, however, has populations in
which males are able to produce complex calls and other populations in which
males only produce simple calls (Boul et al., 2007). All species and all popula-
tions that produce chucks have large fibrous masses while species and popula-
tions that produce only simple calls have much smaller fibrous masses (Boul &
Ryan, 2004). There is also experimental evidence suggesting a role for the
fibrous mass in chuck production. When the fibrous mass was surgically
removed, males attempted to add an additional component to the whine,
but they were not able to produce a chuck (Griddi-Papp et al., 2006).
Male call production is socially regulated. Males calling in isolation usually
produce only simple calls, while males in choruses are more likely to produce
complex calls. Although males can produce calls with up to seven chucks,
they rarely do so. In recordings of complete call bouts of 85 males calling in
BE LOVED , BE PREY , BE EATEN 127

Figure 5.3. Laryngeal morphology of the túngara frog. (a) Stylized model showing the
positional relationships of the larynx and lungs in the calling frog. (b) Simplified
illustration of the larynx showing the position of the two fibrous masses protruding
from the larynx towards the lungs (lungs not illustrated) and the arytenoids cartilages
(top). (c) View of the larynx from the lungs showing the fibrous masses protruding
from the larynx (d) Mid-sagittal section of the larynx showing the attachment of the
fibrous mass to the vocal fold. (Illustrations by Cristina and Marcos Griddi-Papp)

choruses in the wild, 53 percent of the calls were simple, consisting of only a
whine, while 37 percent had one chuck, 10 percent had two chucks, and
0.1 percent had had three or more chucks (Bernal, Page, et al., 2007).
Males tend to increase and decrease call complexity one chuck at a time and
add chucks in response to calls of other males (Bernal et al., 2009). In evoked-
vocalization studies using static stimuli, in which the same call is broadcast to
the male at a natural calling rate, males produced more chucks in response to
complex calls than to simple calls but tended not to produce more chucks in
response to stimuli with more versus fewer chucks (Figure 5.4; Bernal et al.,
2009). In response to dynamic playbacks, in which a computer program
counted the number of chucks in the male’s call and then responded with a
specified calling strategy, males produced more chucks when the playback
always produced one more chuck than the male (“escalate strategy”) than in
response to a playback that always produced one less chuck than the focal male
(“deescalate strategy”; Goutee et al., 2010). Thus the male’s call complexity is
influenced by calling strategies of other males. In experiments using both static
128 ANIMAL BEHAVIOR

Figure 5.4. Diagram illustrating the sequence of male túngara frog calling behavior.
Transitional probabilities from one call type to another are indicated by numbers
and arrows.

and dynamic playbacks the number of chucks produced is low, less than two,
and as in the field is far lower than the maximum number of chucks that males
are able to produce.
Females also influence a male’s call complexity (Akre & Ryan, 2011). In
nature males tend to remain fairly stationary while calling, and females are able
to approach a male unencumbered by any interference. Females choose a male
by slowly moving into the male, at which time he clasps her from the top in
amplexus. They remain in this state for several hours before constructing a
foam nest.
Anecdotal observations suggested that males can detect the approach of a
female, perhaps by detecting the vibrations of the female as she swims through
the water, and increase the number of chucks in response. Detailed behavioral
observations have shown that females possess a repertoire of behaviors in the
presence of males that do not function in mate choice directly, that is,
BE LOVED , BE PREY , BE EATEN 129

approaching and initiating amplexus with a male, but seem to function in dis-
play manipulation. In these cases females might rapidly swim past a male,
bump a male and rapidly swim away, leap over a male, or splash in his vicinity.
All of these behaviors and more (seven specific behaviors were identified)
result in males increasing the number of chucks he produces (Akre & Ryan,
2011).

THE INTENDED RECEIVER, FEMALE TÚNGARA FROGS

The main function of the mating call is to attract females. It is not possible
to understand the function and evolution of the communication system with-
out understanding its costs and benefits. As the main function of the call is to
attract females for mating, understanding female preference for mating call
variation is crucial to assessing the fitness benefits of calling for males.
As noted above, anuran mating calls are species specific. Matings between
species are often counterproductive as they usually do not produce viable off-
spring that survive and reproduce. Thus there should be strong selection on
males to produce species-specific calls and for females to prefer the calls of con-
specifics to heterospecifics. Numerous studies of anurans have demonstrated
such preferences (Gerhardt & Huber, 2002). These experiments show that
females discriminate in favor of conspecific calls or against heterospecific calls.
Another issue that is not as often addressed is that of recognition. That is, if a
female is exposed to only a heterospecific call, would she respond to it as if it
represented a viable mate? The difference between what we have termed dis-
crimination and recognition is not a trivial one. In numerous cases female
choice has shown to be context dependent, and there probably are numerous
cases in nature in which a female might encounter the call of a heterospecific
male in the absence of calling male conspecifics.
Túngara frogs are no exception in their preference for conspecific calls over
those of heterospecifics. In a series of experiments, female túngara frogs were
given a choice between a conspecific call and the call of one of seven closely
related species (Ryan & Rand, 1995, 1999). In most cases females showed
an overwhelming preference for the conspecific call. Females were also tested
in recognition experiments with the same heterospecific calls paired with a
white-noise stimulus. Female showed statistically significant recognition of
three of the seven heterospecific calls. Thus female responses to mating calls
are context specific, and female do not reject heterospecific calls in all cases.
There should be strong selection to reject heterospecific calls, but túngara
frogs are sympatric with only one other member of the same genus, and that
only occurs in a small area of the llanos of Venezuela. All of our phonotaxis
130 ANIMAL BEHAVIOR

experiments with túngara frogs are conducted in Central America, usually in

Panama, where túngara frogs are the only Physalaemus frogs present. Thus
the species-specific preferences we measure are probably due to selection for
self-recognition, as suggested by Paterson (1978), as opposed to selection to
avoid any particular heterospecific call. This interpretation was borne out in
studies of female generalization of mating calls. We constructed a series of
“acoustic transects” between the call of the túngara frogs and calls of each of
five other heterospecifics in the genus Physalaemus (Ryan et al., 2003). These
transects consisted of the túngara frogs and a heterospecific call and six calls
that were intermediate between the two. Female recognition for all of these
calls was then tested. In general, the results showed that in making recognition
decisions females generalized from the conspecific call.
The more similar the stimulus was to the conspecific call the more likely a
female was to show phonotaxis to the call—to recognize it as a potential mate.
Thus females do not seem to categorize calls as either conspecific or heterospe-
cific but instead evaluate them as more or less likely to be conspecific.
True categorical perception is important in human speech discrimination
but is not common in the animal kingdom. Categorical perception occurs
when continuous stimulus variation is labeled as being in more than one cat-
egory and when discrimination of two stimuli that differ in the same magni-
tude is stronger when the stimuli are in different categories than when they
are in the same category. The túngara frog is the only species that is known
to show categorical perception of conspecific versus heterospecific signals
(Baugh et al., 2008). Although this occurs in only one acoustic transect, and
only when the stimuli are finely partitioned, it is evidence that anurans are
capable of relying on perceptual mechanisms in mating preferences that were
thought to occur in only a few more cognitively endowed animals. Although
our data do not suggest that categorical perception is common in túngara
frogs, our data show it is possible.
Female preferences between conspecific and heterospecific calls can gener-
ate behavioral reproductive isolation and contribute to the process of
speciation. Calls often differ within a species, and when there are preferences
for some calls over others, this can generate sexual selection as some males
will enjoy greater reproductive success because females find them more
attractive.
Sexual selection by female choice has driven the evolution of complex calls
in túngara frogs. Even though males seem to be reluctant to produce complex
calls, needing to be persuaded by male vocal competition or female manipula-
tion, experiments show that females are five times more likely to choose a syn-
thetic whine-chuck over a simple whine. Addressing the question more subtly,
BE LOVED , BE PREY , BE EATEN 131

we have also shown that in 14 of 20 cases a male increases his attractiveness

when he adds a chuck to his call (Baugh & Ryan, 2010b). The differences
among males in the potency of their chucks is due to the amplitude of the
chuck. Males can vary chuck amplitude independent of whine amplitude,
and they must produce a chuck that is at least the same peak amplitude of
the whine for females to prefer this complex call to the simple call.
Not only do chucks make a call more attractive, they can also make it more
memorable (Akre & Ryan, 2010a). Most frogs and insects call in unison-bout
choruses; that is, individuals initiate and cease calling more or less at the same
time. In túngara frogs, choruses average about 50 seconds in duration and are
separated by about 25 seconds of silence. Is the female able to remember a call
over that silent interval, and if so, does call complexity influence how long the
call is remembered—what we refer to as the call’s active time? The experi-
ments to address this question were simple. A female was tested in an acoustic
chamber placed equidistant between two speakers. She was restrained inside a
mesh funnel so she could hear the calls but could not move toward them. First
the female was exposed to a simple call broadcast from one speaker and a com-
plex call from the other speaker. These broadcasts then ceased, and there was a
period of silence. The broadcasts were then resumed, but a simple call now
emanated from each speaker. The null hypothesis of no memory is that the
female should be equally likely to approach either speaker, regardless of which
one had previously been broadcasting the complex one. If the female remem-
bered which speaker had been broadcasting the complex call, she should
approach that one.
When the complex call had one chuck, there was no memory. If the com-
plex call had three chucks, then the female was attracted to the speaker that
had previously broadcast the three-chuck call. The female remembered the call
for up to 45 seconds; after 60 seconds there was a trend in preference for the
complex call suggesting perhaps some weak memory, and there was no hint
of a memory trace at 120 seconds. Thus adding additional chucks, or at least
three chucks, to a call is favored by sexual selection because males are more
likely to be remembered by females. These studies remind us that there are
subtle ways a male can make himself more likely to be chosen as a mate.
Female preference for more chucks versus fewer chucks is context depen-
dent. Any sound decreases in amplitude with greater distance from its source.
The active space of a signal is the area over which the signal is perceived by the
receiver. We normally consider the active space of a signal, but we can also
consider the active space of different signal components. Because signal ampli-
tude varies with distance, we varied the signal amplitude to vary the perceived
distance of the female from the calling male.
132 ANIMAL BEHAVIOR

The females’ preferences for chucks varied with distance (Akre & Ryan,
2010b). When the amplitude of the call at the site of the female was adjusted
to mimic a call produced by a male 12 m away, females exhibited phonotaxis
to the call, but they did not discriminate between a simple call and a complex
call with either one or three chucks (Figure 5.5). At 3 m females preferred
both complex calls (with one or three chucks) to simple calls, but they did
not prefer calls with three chucks to calls with one chuck, nor did they exhibit
memory for the complex calls. At 1 m females still preferred calls with one or
three chucks to simple calls; they also preferred complex calls with three
chucks to complex calls with one chuck; and they also exhibited memory for
calls with three chucks. Thus we can think of the call having a stratified active
space in which the potency of different components of the call varies with
distance.
The amount of evolution of a trait under sexual selection, such as the túng-
ara frog’s complex call, is dependent upon the strength of selection on that
trait. As the strength of sexual selection is related to the strength of female
preference, understanding how females perceive trait variation offers insights
into how traits have evolved.
How do frogs, and humans for that matter, compare the magnitude of
traits? If we were to be blindfolded and hold a 1-pound weight in each hand,

Figure 5.5. Active space of female discrimination of male túngara frog calls.
BE LOVED , BE PREY , BE EATEN 133

we could readily tell to which hand a quarter-pound weight was added. If we

were holding 25-pound weights in each hand, we might not detect the addi-
tion of a quarter pound. Humans usually do not use absolute differences
between stimuli to compare them but instead compare the proportional differ-
ence between them. This is called Weber’s law (Thurstone, 1927). If this is
true for animals, it could have a strong influence on the evolution of elaborate
traits under sexual selection (Cohen, 1984). For example, if the ancestral pea-
cocks had trains (often referred to as tails) that were only 5 cm in length, a
male with a train of 6 cm might be perceived as more attractive by the female.
Eventually, sexual selection by female choice would continue to favor males
with longer trains, and we would expect the evolution of the large trains
common in peacocks today. But what halts the evolution of the trait? As we
will argue in this chapter, the costs of the traits due to increased predation will
certainly slow down its evolution. But there might also be a cognitive con-
straint imposed by the female. The peahen might prefer a male with a train
of 6 cm over 5 cm, but will she prefer a male with a train of 101 cm over
100 cm? This seems less likely.
How do female túngara frogs perceive variation in chuck number? Is the
strength of preference, the probability that a female will prefer the call with
more chucks, predicted by the absolute difference in chuck number between
the two calls or by the proportional difference in chuck number? We tested
females with pairs of calls that varied in chuck number (Akre et al., 2011).
The absolute difference in chuck number was not a significant predictor of
the strength of female preference—it explained only 16 percent of the varia-
tion in responses. The proportional differences in chuck number, however,
explained 84 percent of the variation in preferences among various pairs of
calls. Thus túngara frogs follow Weber’s law in making mating decisions based
on variation in chuck number. This might be one of the reasons why males do
not add as many chucks as possible. As noted above, males almost always
increase chuck number in single increments. Adding a chuck to a call with
one chuck will have much more of an advantage than adding a chuck to a call
with four chucks. We do not know if any other animals follow Weber’s law in
mating decisions, but in the next section we talk about how a very different
animal perceptually weights variation in the number of chucks.
All of this discussion of complex calls suggests that the chuck is a very spe-
cial sound. Acoustically, it is precisely constructed with a low fundamental fre-
quency, many harmonics, and a dominant frequency in the upper harmonics
(Wilczynski et al., 1995). We expect signals to be “special” in the sense that
they elicit the specific responses from the receiver that are favored by selection.
Broadcasting the whine backwards or a whine that is composed of noise will
134 ANIMAL BEHAVIOR

not elicit a response from females. Female response to the structure of the
chuck, however, is more permissive. All of the low-frequency harmonics can
be removed from the chuck without the call suffering in attractiveness. In fact,
the high-frequency harmonics can be replaced with a single tone without
ameliorating its potency as long as that tone is in the vicinity of the dominant
frequency. What is perhaps most surprising is that the order of the call compo-
nents is not critical; a chuck-whine is still preferred to a whine only.
We have tested females with a variety of stimuli substituted for the chuck.
Our results show that a diversity of sounds, including conspecific and hetero-
specific calls as well as predator-produced and human-made sounds such as
bells and whistles, can all substitute for the chuck and still result in an increase
in the attractiveness of the whine. In addition, many of these substitutes are as
attractive as a chuck. Interestingly, we did not find any stimuli that when
paired with a whine were more attractive than a whine-chuck (Ryan et al.,
2010).

AN UNINTENDED RECEIVER, THE FROG-EATING BAT

We have shown that the strength of female preference for male mating calls
is both context dependent and distance dependent and that there is consider-
able nuance to the magnitude of female preference. In most contexts, how-
ever, female frogs strongly prefer complex calls to simple ones. If male frogs
have higher mating success when producing complex calls, why do male túng-
ara frogs ever produce simple calls? One answer comes in the form of an eaves-
dropping bat. The frog-eating bat (Trachops cirrhosus) uses frog mating calls to
detect, locate, and assess its prey. Like the female túngara frogs, this bat shows
a strong preference for complex calls over simple ones (Ryan et al., 1982).
T. cirrhosus ranges from southern Mexico to Brazil (Cramer et al., 2001)
and is highly opportunistic in its diet, consuming a large variety of arthropods
(Bonaccorso, 1978; Kalko et al., 1996; Bonato et al., 2004), as well as many
types of vertebrate prey, including lizards, birds, smaller species of bats, and
most famously, frogs (Bonato & Facure, 2000; Rodrigues et al., 2004). T. cir-
rhosus is not alone in feeding on frogs; several species of bats are frog predators.
What is extraordinary about T. cirrhosus is its ability to eavesdrop on the sexual
advertisement calls of male frogs and use these calls to obtain its prey.
Studies from Barro Colorado Island and the surrounding areas in Panama
show that T. cirrhosus responds to the calls of numerous frog species. Given
the frog mating call alone, T. cirrhosus can distinguish palatable from poison-
ous prey, and prey that are too large to capture from prey that are an appropri-
ate size (Tuttle & Ryan, 1981). It prefers higher-amplitude túngara frog calls
BE LOVED , BE PREY , BE EATEN 135

to lower-amplitude calls, and it prefers calls at a faster call rates to calls at

slower call rates (Tuttle & Ryan, 1981). It also generalizes from the calls of
known prey to the calls of unknown prey (Ryan & Tuttle, 1983). While T.
cirrhosus rely primarily on frog advertisement calls in its hunting approaches,
experiments show that at close range bats can also use echolocation and chemi-
cal cues to sequentially update their assessment of prey quality (Page et al.,
2012).
The presence of T. cirrhosus induces graded evasive responses in túngara
frog prey, beginning with call cessation, then vocal sac deflation, then partial
and finally full submersion under the water’s surface (Tuttle et al., 1982).
Túngara frogs visually detect hunting bats; they can discriminate between
the silhouette of a broad-winged frog-eating bat and a narrow-winged insec-
tivorous bat and remain silent longer following the overhead flight of a frog-
eating bat (Tuttle et al., 1982). Túngara frogs rely on public information to
resume calling; their latency to calling after a bat passes is reduced when the
frogs hear other túngara frogs and other sympatric frogs calling (Phelps et al.,
2007). Bat capture rates are high—in one study researchers observed nearly
100 capture events, at a rate exceeding six frogs captured per hour (Ryan et
al., 1981). T. cirrhosus are more successful at capturing frogs when males are
calling in sustained bouts; in contrast, when males are only sporadically pro-
ducing calls, bat capture success is low. Chorus size does not predict predation
rate (number of frogs captured per hour of observation), but an individual
frog’s risk of predation by bats is significantly lower in larger choruses (Ryan
et al., 1981).
Like female frogs, T. cirrhosus prefer complex calls (Ryan et al., 1982).
Above we discuss research conducted to understand female preference for
complex calls, but why should a predator seeking a meal prefer one call type
to another? A number of hypotheses have been proposed. It is possible that
frog-eating bats prefer complex calls because complex calls indicate higher-
quality prey and thus a better meal. Túngara frog call complexity, however,
is not correlated with length, mass, or body condition (Bernal, Page, et al.,
2007). Another possibility is that bats prefer complex calls because these calls
indicate higher prey densities. The number of males within 1 m of a calling
male is correlated both with the proportion of complex calls a male produces
and the average number of chucks per call (Bernal, Page, et al., 2007), so pred-
ators attracted to complex calls indeed are likely to find more, not fewer, túng-
ara frogs. Bats, however, could also prefer complex calls because they are easier
to localize than simple calls. The chucks associated with complex calls are
short, broadband, and have rapid onsets and offsets, all acoustic properties
predicted to maximize binaural comparisons and facilitate localization.
136 ANIMAL BEHAVIOR

Frog-eating bats indeed localize complex calls better than simple ones (Page &
Ryan, 2008). Their localization performance is a function of the difficulty of
the localization task (Figure 5.6). In more simple tasks (no background noise,
no obstacles, continuous frog calls), bats localize simple and complex calls
equally well. In more difficult localization tasks (high background noise, many
obstacles, frog calls that cease as soon as bats begin their approach), bats’ locali-
zation performance for simple and complex calls is equally poor. In conditions
of intermediate difficulty, however, T. cirrhosus consistently show better locali-
zation performance for complex calls over simple ones.
Other factors might also be involved in predator preference for call com-
plexity. Ongoing studies investigate the role of learning in predator preference,
and the degree to which complex-calling males are more distracted and easier
to capture than simple-calling males.
Cognition and perception influence how receivers process information and
respond to incoming signals. Here we discuss recent cognitive and perceptual
studies in the frog-eating bat, including the bats’ perception of multiple

Figure 5.6. Frog-eating bats localize complex calls better than simple ones when tasks
are of intermediate difficulty. When the localization task is simple or difficult
(shown here with the addition of an obstacle course and background noise), there is
no difference in localization performance for simple or complex calls.
BE LOVED , BE PREY , BE EATEN 137

chucks, how they assign prey quality to prey calls, and the potential for social
learning.
Earlier we discussed female preference for increasing male trait elaboration
and showed that for males producing complex calls, it is a case of diminishing
returns; as males increase call complexity, differences in relative attractiveness
decrease (Akre et al., 2011). Female preference for increasing number of
chucks closely follows the prediction of Weber’s law—females perceive chuck
variation proportionally, not by the absolute difference in chuck numbers, so
the same increase of a single chuck that makes a large difference to perceived
attractiveness of a male producing no chucks makes very little difference to
the attractiveness of a male producing five chucks. How do predators perceive
the increasing trait elaboration of their prey? We tested bats with pairs of calls
that varied in chuck number. Just as in the female frogs, the absolute differ-
ence in chuck number explained little of the variation in strength of bat pref-
erence (8.3 percent). The ratio of chuck number, however, explained a
substantial portion of the variation in strength of bat preference (73.9 percent;
Akre et al., 2011). So rather than increasing the strength of their preferences
linearly, frog-eating bats, like female túngara frogs, show less of an increase
in preference as chuck numbers increase.
Frog-eating bats respond to the calls of many of the palatable frogs calling
in their area (Tuttle & Ryan, 1981). Given an acoustic cue alone, T. cirrhosus
will attack speakers broadcasting the calls of palatable species and ignore
speakers broadcasting the calls of poisonous species (Tuttle & Ryan, 1981).
Clearly the bats have associations between the calls they hear and expected
prey quality. How flexible are these prey-cue/prey-quality associations? Can
they be altered given new information, or is a bat’s acoustic repertoire fixed
and resistant to change? We tested bats with species from extremes on the pal-
atability spectrum—on the palatable end, we chose túngara frogs, which are
small, easy to capture, and a preferred prey species of this bat. On the poison-
ous end, we chose Rhinella marina (=Bufo marinus), the cane toad, which is
much larger than the bat, poisonous, and has a distinct call that is very differ-
ent from the túngara frog call. Using wild-caught bats, we first measured base-
line preference levels by presenting the bats both stimuli; all bats preferred the
frog calls to the toad calls. Then, to test the flexibility of their preference, we
offered the bats five acoustic steps, each rewarded, gradually fading the frog
call into the toad call by systematically decreasing the amplitude of one and
increasing the amplitude of the other.
Bats responded to the fading stimuli quickly, completing the five steps in
approximately an hour. We then extinguished the bats’ response to túngara
frog calls by repeatedly offering the bats a call without a food reward on the
138 ANIMAL BEHAVIOR

speaker. This step took longer, but for all bats, response to the túngara frog
calls was extinguished within the course of one night. We then presented the
bats again with a choice between toad and frog calls and saw a complete rever-
sal in their preference—bats that had previously flown only to frog calls and
never to toad calls now did the opposite (Page & Ryan, 2005). These results
show that bats are highly flexible in the associations they form between prey
cues and prey quality and can quickly form new associations given changes
in prey quality. These traits should help frog-eating bats respond to temporal
and spatial fluctuations in prey availability and encounters with novel prey
species.
We next wanted to know how these prey-cue/prey-quality associations are
formed. Using a similar paradigm, we quantified the acquisition of a novel for-
aging behavior (learning to associate the calls of cane toads with palatable prey)
in three groups of bats: a social learning group, a social facilitation group, and
a trial-and-error group (Page & Ryan, 2006). We broadcast cane toad calls
from a speaker with a food reward and quantified the number of trials needed
for the bats to respond to the novel stimulus. The social learning group con-
sisted of two bats, one naïve and one experienced with the novel association.
The social facilitation group consisted of two inexperienced bats; the purpose
of this treatment was to test whether the mere presence of a conspecific would
increase the rate of learning a novel association. The trial-and-error group con-
sisted of a single inexperienced bat alone. The bats in the social learning group
rapidly acquired the novel association in a mean of 5.3 trials. Most bats in the
other groups did not approach the novel stimulus, even after 100 trials. In
each of these groups, however, there was a single bat that approached the toad
calls and learned the novel association (after 81 trials in the trial-and-error
group, and after 84 trials in the social facilitation group). These results suggest
that not only can information be transferred quickly from bat to bat, establish-
ing the potential for rapid cultural transmission of foraging information in
nature, but that bats explore novel prey and cues. Exploratory behavior, even
when rare, could explain the origin of novel foraging responses that are then
transferred socially from bat to bat.

YET ANOTHER UNINTENDED RECEIVER, THE FROG-BITING MIDGES

When male túngara frogs call, not only do they risk being eaten by frog-
eating bats, they also are in jeopardy of being parasitized by frog-biting midges
(Diptera: Corethrella spp). Female midges use túngara frog mating calls to find
and bite a calling male to obtain a blood meal to support egg production.
These midges belong to a monogeneric family, Corethrellidae, of more than
BE LOVED , BE PREY , BE EATEN 139

100 species of pantropical and subtropical flies (Borkent, 2008). Midges from this
family look similar to mosquitoes, and given these similarities, these two groups
were in the same family until the late 1980s when they were assigned to a separate
family (Wood & Borkent, 1989). Corethrellidae is currently recognized as the
sister group to the phantom midges (Chaoboridae) and mosquitoes (Culicidae).
Corethrella midges are eavesdroppers that specialize on frog mating calls and
were first recognized by Sturgis McKeever (1977). Art Borkent (2008)
reviewed the fossil, cladistic, and morphological evidence and suggests that
midges and frogs have shared a long evolutionary history, probably at least
since the Early Cretaceous. Midges are also vectors of disease. They transmit
blood parasites to frogs (Johnsons et al., 1993). In túngara frogs, a new species
of such blood parasites has been discovered, Trypanosoma tungarae (Pinto &
Bernal, in preparation). Initial evidence suggests that frogs and frog-biting
midges share a long evolutionary history with trypanosomes; their intricate
associations and interactions deserve further study.
Studies in Gamboa, Panama, revealed that at least seven species of frog-biting
midges attack calling túngara frogs (Figure 5.7). Midges are abundant, and a
speaker broadcasting túngara frog calls attracts an average of 142 midges in
30 minutes; on some nights, a single vigorously calling male can attract over
500 midges in this period (Bernal et al., 2006). Once a female midge homes
in on a calling male, she lands on his back and walks to the nostrils where, in this
frog species, most of the midges take a blood meal. Males attract midges only

Figure 5.7. Frog-biting midges attacking a túngara frog. (Photo by Alexander T. Baugh)
140 ANIMAL BEHAVIOR

while they are calling. Given that male túngara frogs call while floating in water,
when a male stops calling, the midges do not remain in flight over the water but
fly to nearby vegetation. Only those midges already on the frog continue
attempting to obtain a blood meal. The midges’ dependence on the male frog’s
call for localization creates bouts of midge attacks that parallel the call bouts of
the male frog.
As discussed earlier, both frogs and bats prefer the complex calls over the sim-
ple calls of the túngara frog. Do the frog-biting midges share this preference for
complex calls? Given that the perception of a signal is dependent on the sensory
system of the receiver (Endler, 1978, 1992), it is not necessarily expected that
receivers with such different sensory systems as frogs, bats, and midges would
share call preferences. Field playback experiments using sound traps broadcast-
ing either complex or simple calls, however, revealed that the midges too are
more likely to approach calls with chucks than calls without them (Bernal et
al., 2006). There could be several reasons for this convergence of call preferences.
One is that female frogs and eavesdroppers could all benefit from approaching
complex calls because, as mentioned earlier, calls with chucks are often associated
with a high density of males. For the bats and midges, increased effectiveness
when attacking frogs may have played a role favoring the preference for complex
calls (Bernal, Page, et al., 2007). Another possibility is that complex calls are eas-
ier for midges to localize as has been shown for bats. The landing error of midges
approaching complex and simple calls, however, does not differ depending on
the call type (Bernal et al., 2006). Although call complexity does not seem to in-
fluence the midge’s ability to land on the frog, it may affect the midge’s ability to
detect the call or its approach pathway when it cues in on the call.
It is intriguing that the midges mostly bite túngara frogs on the nostrils
(Figure 5.8) as this is not true for midge attacks on other species of frog. When
midges attack small treefrogs (Dendrosophus ebbraccatus, D. microcephalus) that
breed in the same areas as túngara frogs, they bite the treefrogs in many parts
of the body, including the nostrils, back and legs.
Túngara frogs, as their species epithet pustulosus suggests, have thick skin
with numerous small pustules that give them a toad-like appearance. It is pos-
sible that the midges are not able to pierce through the túngara’s skin in most
parts of the body and are only able to successfully obtain a blood meal from
the nostrils. Thin histological sections of túngara frog skin revealed that the
skin on their back has few, small capillaries underneath a thick layer of connec-
tive tissue farther from the surface of the skin than most midges could reach
(de Silva et al., in press). The skin of the nostrils, in contrast, is rich in capillaries
that are close to the surface, facilitating a midge’s attempts to obtain a blood meal.
In contrast to the skin of the dorsum of túngara frogs, the skin of the two small
BE LOVED , BE PREY , BE EATEN 141

Figure 5.8. Frog-biting midges attacking a túngara frog. Note the engorged midges
obtaining a blood meal from the nostril of the frog. (Photo by Alexander T. Baugh)

treefrog species from the same area has higher density of capillaries closer to the
surface of the back and nostrils, probably facilitating biting in those areas.
In addition to their thinner, capillary-rich skin, the nostrils may attract frog-
biting midges if the midges are following a CO2 gradient, as many blood-
sucking insects do. Although in anurans the majority of gas exchange takes place
at the nostrils, frogs excrete most CO2 through the skin (Boutilier et al., 1992).
Thus, it is unlikely that the midges follow a CO2 gradient that leads them to bite
at the nostrils. This area is also in close proximity to the vocal sac, however, and the
midges could be following a sound-intensity gradient that leads them to the nos-
trils. Although female midges may be using such gradients to reach the nostrils,
CO2 and sound cues alone do not determine where the midges bite their hosts.
Ongoing studies are investigating the cues used for host localization by the midges.

AUDITORY ADAPTATIONS: HOW TÚNGARA FROGS, FROG-EATING BATS,

AND FROG-BITING MIDGES HEAR FROG CALLS
In most animal communication systems there is some congruence between
the signals and receivers. A signal is only salient if it can be detected and per-
ceived by the receiver. Such patterns of congruence are best explored in audi-
tory and visual communication systems. For example, Robert Capranica’s
142 ANIMAL BEHAVIOR

(1976) seminal studies of anuran communication predicted that the tuning of

the peripheral auditory system matches the most energetic frequencies in the
male’s species-specific mating call. Gerhardt and Joshua Schwartz (2001) later
documented the accuracy of Capranica’s prediction. Similarly, in numerous
visual systems, especially in fishes, researchers have shown congruence
between the spectral tuning of the cone photopigments and the animals’ color
patterns (Cummings, 2007; Seehausen et al., 2008).
As we have noted, signals are often detected and perceived by unintended
receivers. In some cases animals have evolved signals that reside in a private com-
munication channel inaccessible to those for which the signals are not intended.
For example, many species of swordtail fishes have a band on the sword, a set
of extended rays of the caudal fin, which enhances the internal contrast of the
sword. Such a band is lacking, or appeared to be lacking, in Xiphophorus nigrensis.
Spectral analysis, however, shows the presence of a band in the ultraviolet (UV)
spectrum that is invisible to the human eye. X. nigrensis are sensitive to light in
the UV, and males with UV bands are more attractive to females than are males
that do not exhibit the band (because they were in an arena in which UV light
was filtered out; Cummings et al., 2003). The major predator of X. nigrensis is
the Mexican tetra (Astyanax mexicana; Rosenthal et al., 2001). These fish lack
photopigments that are sensitive to the UV, and behavioral experiments show
that males with and without UV bands are equally attractive to the predator
(Cummings et al., 2003). This is a case in which selection from eavesdroppers
has caused evolution of the signal. We have no evidence of such an effect on
the mating call in túngara frogs. But there are some suggestions that eavesdrop-
pers have evolved to become more sensitive to the frequencies that compose the
frog’s call. To understand the potential for an animal’s signal to be exploited it
is important to understand the sensory capabilities of the potential eavesdroppers.

AUDITORY ADAPTATIONS OF TÚNGARA FROGS

Túngara frogs follow the basic model of anuran acoustic processing. The
two inner ear organs, the amphibian papilla (AP) and the basilar papilla
(BP), are sensitive to the dominant frequencies of the whine and the chuck,
respectively (Ryan et al., 1990). The frequency to which the BP is most sensi-
tive, about 2,200 Hz, is slightly below the average chuck dominant frequency
of about 2,500 Hz. This should make the female more sensitive to the lower
chuck frequencies of larger males and probably explains why females are more
attracted to these calls and these males.
Neural responses in the inner ear reach the frog’s hindbrain through the
VIIIth cranial nerve. Most of the acoustic analysis appears to take place in
BE LOVED , BE PREY , BE EATEN 143

the midbrain, in the torus semicircularis. This brain nucleus is homologous to

the mammalian inferior colliculus, which is also a center of auditory process-
ing. Kim Hoke and colleagues used the expression of immediate early genes
(IEGs) to quantify the amount of neural response to various stimuli (Hoke
et al., 2004). IEGs are expressed in neurons when there are action potentials.
Thus quantifying the amount of IEG expression should give some estimate
of how stimulating a call might be. In the torus there is greater stimulation
in response to, in this order, the whine-chuck, the whine, the chuck only,
and the call of a heterospecific P. enesefae. The latter call contains the most
energy that would match the tuning of the túngara frog’s inner ear, but the
torus is obviously responding to more than just frequencies.
Once signals are analyzed, they need to generate decisions. Regions within
the hypothalamus are thought to play an important role in modulating behav-
ioral response to stimulus variation. Hoke, Ryan, and Walter Wilczynski
showed that in túngara frogs the behavioral salience of the signal influences
neural responses in the hypothalamic regions (Hoke et al., 2005). The key var-
iable, however, is not the amount of activity, as estimated by IEG expression,
but the correlation in activity among these regions. Thus perception of the
mating call, both the whine and the whine-chuck, shift the functional connec-
tivity within the hypothalamus compared to perception of other sounds.
These results are consistent with the principles that underlie the simultaneous
processing of sensory information in cognitive tasks.
When we consider these exquisite details of auditory processing by the túng-
ara frog we would expect strict congruence between the selectivity of the auditory
system and the acoustic properties of the mating call. But this is only partially
true. Many of the sounds contained in the whine are neither necessary nor suffi-
cient to elicit phonotaxis from gravid females. This is what we would expect from
the notion of the sign stimulus. An animal signal should contain the stimuli that
are salient to the receiver, but given the mechanics of sound production it is not
surprising that there are sounds that are not relevant. For example, about half of
the acoustic energy in the whine is in the upper harmonics, with the other half
present in the fundamental frequency. The upper harmonics, however, do not
influence the call’s attractiveness. The requirements for a salient whine are, how-
ever, quite strict as the call needs to stimulate first a high-frequency and then a
low-frequency range of the whine’s fundamental frequency (Wilczynski et al.,
1995). It is worth noting again that all of the close relatives of the túngara have
a whine-like call, and females are able to discriminate between the calls of most
heterospecifics and conspecifics by the attending to the whine alone.
The chuck presents a very different story. The whine does not need a chuck
to elicit female phonotaxis, and a chuck by itself is not responded to by
144 ANIMAL BEHAVIOR

females. The chuck appears to be “auditory cheesecake” that enhances the

stimulation of the whine.
Although rich in harmonics that encompass the sensitivity of the both the
AP and BP, the effect of the chuck can be mimicked by a single tone as long
as it is near the peak sensitivity of the BP. The requirements of a chuck are
far less stringent than those of the whine. As noted above, Ryan and colleagues
showed that numerous sounds, parts of calls of other species, environmental
noise, and even bells and whistles can enhance the attractiveness of the whine
and often can make it as attractive as a whine chuck (Ryan et al., 2010).

AUDITORY ADAPTATIONS OF FROG-EATING BATS

Like many bats that glean prey in the rainforest understory, T. cirrhosus
produces short, multiharmonic, frequency-sweep echolocation calls. These
calls range from 100 to 50 kHz, with most of the call energy at 75 kHz (Bar-
clay et al., 1981). Like other bat species, T. cirrhosus are sensitive to their own
echolocation calls, and their auditory sensitivity decreases with decreasing
stimulus frequency. Behavioral audiograms of sonic frequency stimuli show
that T. cirrhosus need progressively greater stimulus intensities to elicit
responses as frequencies drop from 15 kHz to 5 kHz. Unlike other bat species,
however, T. cirrhosus has an additional peak of auditory sensitivity below 5
kHz, in the frequency range of most frog calls (Ryan et al., 1983).
There are a number of ways in which the ear morphology of frog-eating
bats reflects their ability to hear low frequencies (Bruns et al., 1989). Like
other bats that listen for prey-emitted sounds rather than relying exclusively
on echolocation for prey detection, T. cirrhosus has large pinnae, aiding the
detection of low-intensity and low-frequency sounds. T. cirrhosus has a long
basilar membrane, expanding the frequency range to which it is sensitive.
The difference in stiffness in the basal and apical portions of the basilar mem-
brane predicts frequency range sensitivity: T. cirrhosus has an extremely large
baso-apical stiffness difference (128,600:1), indicating a very broad frequency
range (Bruns et al., 1989). Volkmar Bruns and colleagues further report that
T. cirrhosus has the highest number of cochlear neurons reported for any mam-
mal, and the second highest density of cochlear neuron innervation of any
mammal (Bruns et al., 1989). The location of neural density is telling: most
mammals have a single neural peak in the middle of the cochlea, as does T. cir-
rhosus; most bats have a second peak in neural density in the basal part of the
cochlea (the part for detecting high-frequency sounds), as does T. cirrhosus.
What is extraordinary is that T. cirrhosus has yet a third peak of neural density.
Its third peak is found in the apical portion of the cochlea, the portion for
BE LOVED , BE PREY , BE EATEN 145

detecting low-frequency sounds (von Békésy, 1960). The large number of

anatomical specializations points to extreme adaptation of the T. cirrhosus
auditory system for the detection of low-frequency sounds such as frog calls.

AUDITORY ADAPTATIONS OF FROG-BITING MIDGES

Unlike the majority of hematophagous insects, frog-biting midges use
acoustic signals to find their host. Their auditory system, however, remains a
mystery. A comparable eavesdropping parasite is the parasitoid fly Ormia
ochracea, which deposits its larvae on male field crickets (Gryllus spp.; Cade,
1975). The females of this species of fly use the cricket’s calls for localization
cues and have an extraordinarily sensitive ear that constitutes an evolutionary
innovation in flies (Lakes-Harlan & Heller, 1992; Robert et al., 1992,
1994). O. ochracea have two small ears on the prosternum, where the head
attaches to the body. These ears consist of thin, flexible tympanic membranes
coupled to provide interaural time differences that endow these flies with the
remarkable ability to detect and localize the cricket’s call (Müller & Robert,
2001). The evolution of this unique hearing structure highlights the strength
of selection promoting specializations for interspecific eavesdropping behavior.
Frog-biting midges also depend on eavesdropping for reproduction, and
one would predict similarly strong selection to evolve auditory adaptations
for hearing and localizing calls of their hosts. The antennal sensilla pediconica
of frog-biting midges were proposed as a sound receptor (McKeever, 1988)
but it is unlikely that these structures accomplish such a function given their
rigid structure and their absence in some of the midge species attracted to frog
calls. Given that studies of antennal hearing in mosquitoes have recently chal-
lenged common assumptions of antennal sensitivity (Cator et al., 2009;
Nadrowski et al., 2011), it is possible that the antennae and the associated
Johnston organ of frog-biting midges are more sensitive than previously
thought. Field experiments, however, suggest that the midges respond to frog
calls in the far field where particle displacement is substantially diminished
(Bernal et al., 2006). This suggests that a more elaborate pressure-sensitive
organ is involved in hearing in this group. Ongoing studies are addressing
the mystery of how frog-biting midges hear the mating calls of their hosts.

OTHER EAVESDROPPERS
Mating signals are species specific, conspicuous, and often easy to localize,
characteristics that make them particularly vulnerable to exploitation by eaves-
droppers. Not surprisingly, more than one unintended receiver may exploit a
146 ANIMAL BEHAVIOR

given communication system. We have discussed how male calling túngara frogs
are attacked by both bats and midges, but these are not the only eavesdroppers
taking advantage of this frog’s mating calls. Túngara frogs are surrounded by a
network of exploiters. Merlin Tuttle and colleagues (1982) experimentally
showed that four-eyed opossums (Philander opossum), for instance, also use the
calls of túngara frog males to localize and eat them. There is anecdotal evidence
that cane toads, which also prey on túngara frogs, use the frog’s call to localize
them (Jaeger, 1976). Similarly, South American bullfrogs (Leptodactylus savagei)
often share their breeding sites with túngara frogs. They are voracious predators
of túngara frogs (Ryan et al., 1981), and because their call shares many of the
frequencies in the túngara frog’s call, the bullfrogs should be quite sensitive to
the calls of their prey and might also rely on the túngara frog’s call for localization
cues. Although eavesdropping by L. savagei has not been demonstrated, female
túngara frogs are less attracted to a mating call which is broadcast in the presence
of the call of L. savagei (Bonachea & Ryan, 2011).
Additional predators, which could be potential eavesdroppers, have been
reported eating túngara frogs. Crabs (Potamorcacinus richmondi) and spiders, for
instance, have been observed eating calling males at túngara frog choruses (Ryan
et al., 1981; personal observation). These arthropods could use the vibrations pro-
duced by the inflation and deflation of the vocal sac in the water or other compo-
nents of the acoustic display of male frogs to increase their hunting efficiency.
Although eavesdroppers are more likely to attack males, female túngara
frogs also fall victim to eavesdroppers attracted to the breeding area. Light lev-
els, for instance, influence the responses of females to mating calls, suggesting
an effect of predation risk. Stanley Rand and colleagues investigated female
phonotaxis under lower (dark) and higher (dim) light conditions, finding that
females find darker conditions less risky (Rand et al., 1997). Using acoustic
playback experiments in which the intensity and complexity of the calls was
adjusted as the female moved to a given speaker, Alexander Baugh and Ryan
(2010a) showed that female túngara frogs’ choosiness is also influenced by
light levels. Under dim-light conditions, females are more likely to commit
to an initial call choice even if the attractiveness of the call produced by that
chosen speaker is experimentally reduced. In dark conditions, however,
females are more likely to reverse their choices and approach a suddenly more
attractive call at the expense of increasing the amount of time and movement
navigating the breeding area to select a male.
Both studies support the hypothesis that under high-light conditions
females are more vulnerable to predation, and they adjust their mate-choice
strategies to minimize such risk. High-light conditions may increase the
chances that a female will be captured by a visually orienting predator as well
BE LOVED , BE PREY , BE EATEN 147

as by approaching eavesdroppers that could opportunistically attack her if she

is close to a calling male. Consistently, females are preferentially attracted to
calls without a predator-related sound compared to the same call followed by
the sound of an approaching eavesdropper such as the one produced by the
wings of a flying frog-eating bat (Bernal, Rand, et al., 2007) or, as noted
above, the call of a predator. Although some studies have considered the influ-
ence of predation risk on female mate-choice behavior, fewer have examined
the responses and strategies of females to minimize risk to eavesdroppers.

SUMMARY
In its simplest version, communication involves a sender who produces signal
and a receiver who detects and perceives it. The relative costs and benefits of com-
municating, along with the constraints on signal production and reception, are
crucial to understanding the evolution of the communication system. As we illus-
trate, eavesdroppers can impose substantial costs on these systems and might be
one of the primary reasons for the tension between natural selection and sexual
selection. Not only can eavesdroppers influence the evolution of their victims,
but the victims’ communication systems can in turn drive sensory adaptations
in eavesdroppers. This communication mecca of intended and unintended
receivers, as we illustrated here with details of the frog-bat-midge system, offers
endless opportunities to unweave this tangled web of evolution and function.

ACKNOWLEDGMENTS
We are grateful to our late mentor and collaborator A. Stanley Rand for his
generosity with his exquisite knowledge of tropical biology and his encourage-
ment to pursue our ideas. He is sorely missed. We also thank Merlin Tuttle,
whose ability to know what it is like to be a bat opened the door to the frog-
eating bat saga. For decades of support and funding, we are grateful to the
Smithsonian Tropical Research Institute, the National Science Foundation,
and the University of Texas at Austin.

REFERENCES AND SUGGESTED READING

Akre, K. L., H. E. Farris, A. M. Lea, R. A. Page, & M. J. Ryan. (2011). Signal perception
in frogs and bats and the evolution of mating signals. Science, 333, 751–752.
Akre, K. A. & M. J. Ryan. (2010a). Complexity increases working memory for mat-
ing signals. Current Biology, 20, 502–505.
Akre, K. L. & M. J. Ryan. (2010b). Proximity-dependent response to variably complex
mating signals in túngara frogs (Physalaemus pustulosus). Ethology, 116, 1138–1145.
148 ANIMAL BEHAVIOR

Akre, K. L. & M. J. Ryan. (2011). Female túngara frogs elicit more complex mating
signals from males. Behavioral Ecology, 22, 846–853.
Barclay, R. M. R., B. Fenton, M. D. Tuttle, & M. J. Ryan. (1981). Echolocation calls
produced by Trachops cirrhosus (Chiroptera: Phyllostomatidae) while hunting
for frogs. Canadian Journal of Zoology, 59, 750–753.
Baugh, A. T., K. L. Akre, & M. J. Ryan. (2008). Categorical perception of a natural,
multivariate signal: Mating call recognition in túngara frogs. Proceedings of the
National Academy of Sciences, USA, 105, 8985–8988.
Baugh, A. T. & M. J. Ryan. (2010a). Ambient light alters temporal-updating behav-
iour during mate choice in a neotropical frog. Canadian Journal of Zoology, 88,
448–453.
Baugh, A. T. & M. J. Ryan. (2010b). The relative value of call embellishment in
túngara frogs. Behavioral Ecology and Sociobiology, 65, 359–367.
Bernal, X., K. L. Akre, A. T. Baugh, A. S. Rand, & M. J. Ryan. (2009). Female and
male behavioral response to advertisement calls of variable complexity in túng-
ara frogs, Physalaemus pustulosus. Behavioral Ecology and Sociobiology, 63,
1269–1279.
Bernal, X. E., A. S. Rand, & M. J. Ryan. (2006). Acoustic preferences and localization
performance of blood-sucking flies (Corethrella Coquillett) to túngara frog
calls. Behavioral Ecology, 17, 709–715.
Bernal, X. E., R. A. Page, A. S. Rand, & M. J. Ryan. (2007). Cues for eavesdroppers:
Do frog calls indicate prey density and quality? The American Naturalist, 169,
409–415.
Bernal, X. E., A. S. Rand, & M. J. Ryan. (2007). Sexual differences in the behavioral
response of túngara frogs, Physalaemus pustulosus, to cues associated with
increased predation risk. Ethology, 113, 755–763.
Bonaccorso, F. J. (1978). Foraging and reproductive ecology in a Panamanian bat
community. Bulletin of the Florida State Museum, Biological Sciences, 24,
359–408.
Bonachea, L. A. & M. J. Ryan. (2011). Simulated predation risk influences female
choice in túngara frogs, Physalaemus pustulosus. Ethology, 117, 400–407.
Bonato, V. & K. G. Facure. (2000). Bat predation by the fringe-lipped bat, Trachops
cirrhosus (Phyllostomidae, Chiroptera). Mammalia, 64, 241–243.
Bonato, V., K. G. Facure, & W. Uieda. (2004). Food habits of bats of subfamily
Vampyrinae in Brazil. Journal of Mammalogy, 85, 708–713.
Borkent, A. (2008). The frog-biting midges of the world (Corethrellidae: Diptera).
Zootaxa, 1804, 1–456.
Boul, K. E., W. C. Funk, C. R. Darst, D. C. Cannatella, & M. J. Ryan. (2007). Sex-
ual selection drives speciation in an Amazonian frog. Proceedings of the Royal
Society of London, B, 274, 399–406.
Boul, K. E. & M. J. Ryan. (2004). Population variation of complex advertisement
calls in Physalaemus petersi and comparative laryngeal morphology. Copeia,
2004, 624–631.
BE LOVED , BE PREY , BE EATEN 149

Boutilier, R. G., D. F. Stiffler, & D. P. Toews. (1992). Exchange of respiratory gases,

ions, and water in amphibous and aquatic amphibians. In M. E. Feder & W.
W. Burggren (eds.), Environmental Physiology of the Amphibians (pp. 81–
124). Chicago: University of Chicago Press.
Bruns, V., H. Burda, & M. J. Ryan. (1989). Ear morphology of the frog-eating bat
(Trachops cirrhosus, Family: Phyllostomidae): Apparent specializations for
low-frequency hearing. Journal of Morphology, 199, 103–118.
Burk, T. (1982). Evolutionary significance of predation on sexually signaling males.
Florida Entomologist, 65, 90–104.
Cade, W. H. (1975). Acoustically orienting parasitoids: fly phonotaxis to cricket song.
Science, 190, 1312–1313.
Cade, W. H., M. Ciceran, & A.-M. Murray. (1996). Temporal patterns of parasitoid
fly (Ormia ochracea) attraction to field cricket song (Gryllus integer). Canadian
Journal of Zoology, 74, 393–395.
Capranica, R. R. (1976). The auditory system. In B. Lofts (ed.), Physiology of the
Amphibia (pp. 443–466). New York: New York Academic Press.
Cator, L. J., B. J. Arthur, L. C. Harrington, & R. R. Hoy. (2009). Harmonic conver-
gence in the love songs of the dengue vector mosquito. Science, 323, 1077–
1079.
Cohen, J. (1984). Sexual selection and the psychophysics of female choice. Zeitschrift
für Tierpsychologie, 64, 1–8.
Cramer, M. J., M. R. Willig, & C. Jones. (2001). Trachops cirrhosus. Mammalian Spe-
cies: American Society of Mammalogists, 656, 1–6.
Cummings, M. E. (2007). Sensory trade-offs predict signal divergence in surfperch.
Evolution, 61, 530–545.
Cummings, M. E., G. G. Rosenthal, & M. J. Ryan. (2003). A private ultraviolet
channel in visual communication. Proceedings of the Royal Society of London,
B, 270, 897–904.
Dapper, A. L., A. T. Baugh, & M. J. Ryan. (2011). The sounds of silence as an alarm
cue in túngara frogs, Physalaemus pustulosus. Biotropica, 43, 380–385.
de Silva, P., C. Jaramillo, & X. E. Bernal. (In press). Selection of biting sites on anu-
ran hosts by Corethrella Coquillett species. Journal of Insect Behavior.
Endler, J. A. (1978). A predator’s view of animal color patterns. Evolutionary Biology,
11, 319–364.
Endler, J. A. (1992). Signals, signal conditions, and the direction of evolution. The
American Naturalist, 139, S125–S153.
Fleishman, L. J. (1992). The influence of the sensory system and the environment on
motion patterns in the visual displays of anoline lizards and other vertebrates.
The American Naturalist, 139, S36–S61.
Gerhardt, H. C. & F. Huber. (2002). Acoustic Communication in Insects and Anurans.
Chicago: University of Chicago Press.
Gerhardt, H. C. & J. J. Schwartz. (2001). Auditory tunings and frequency preferences
in anurans. In M. J. Ryan (ed.), Anuran Communication (pp. 73–85). Wash-
ington DC: Smithsonian Institution Press.
150 ANIMAL BEHAVIOR

Goutee, S., N. M. Kime, T. F. I. Argo, & M. J. Ryan. (2010). Calling strategies of

male túngara frogs in response to dynamic playback. Behaviour, 147, 65–83.
Griddi-Papp, M., A. S. Rand, & M. J. Ryan. (2006). Complex call production in
túngara frogs. Nature, 441, 38.
Hedrick, A. V. (2000). Crickets with extravagant mating songs compensate for preda-
tion risk with extra caution. Proceedings of the Royal Society of London, B, 267,
671–675.
Hoke, K. L., S. S. Burmeister, R. D. Fernald, A. S. Rand, M. J. Ryan, & W. Wilczyn-
ski. (2004). Functional mapping of the auditory midbrain during mate call
reception. Journal of Neuroscience, 24, 11264–11272.
Hoke, K. L., M. J. Ryan, & W. Wilczynski. (2005). Acoustic social cues shift func-
tional connectivity in the hypothalamus. Proceedings of the National Academy
of Sciences, USA, 102, 10712–10717.
Jaeger, R. G. (1976). A possible prey-call window in anuran auditory perception.
Copeia, 4, 833–834.
Johnsons, R. N., D. G. Young, & J. F. Butler. (1993). Trypanosome transmission by
Corethrella wirthi (Diptera: Chaoboridae) to the green treefrog, Hyla cinerea
(Anura: Hylidae). Journal of Medical Entomology, 30, 918–921.
Kalko, E. K. V., C. O. Handley, & D. Handley. (1996). Organization, diversity, and
long-term dynamics of a neotropical bat community. In M. L. Cody & J. A.
Smallwood (eds.), Long-term Studies of Vertebrate Communities (pp. 503–
553). San Diego: Academic Press.
Lakes-Harlan, R. & K.-G. Heller. (1992). Ultrasound-sensitive ears in a parasitoid fly.
Naturwissenschaften, 79, 224–226.
McKeever, S. (1977). Observations of Corethrella feeding on treefrogs (Hyla). Mos-
quito News, 37, 522–523.
McKeever, S. (1988). A new species of Mexican Corethrella (Diptera, Chaoboridae)
and a description of a new antennal sensillum. Annals of the Entomological Soci-
ety of America, 81, 400–402.
Müller, P. & D. Robert. (2001). A shot in the dark: The silent quest of a free-flying
phonotactic fly. Journal of Experimental Biology, 204, 1039–1052.
Nadrowski, B., T. Effertz, P. R. Senthilan, & M. C. Göpfert. (2011). Antennal hear-
ing in insects: New findings, new questions. Hearing Research, 273, 7–13.
Page, R. A. & M. J. Ryan. (2005). Flexibility in assessment of prey cues: Frog-eating
bats and frog calls. Proceedings of the Royal Society of London, B, 272, 841–847.
Page, R. A. & M. J. Ryan. (2006). Social transmission of novel foraging behavior in
bats: Frog calls and their referents. Current Biology, 16, 1201–1205.
Page, R. A. & M. J. Ryan. (2008). The effect of signal complexity on localization per-
formance in bats that localize frog calls. Animal Behaviour, 76, 761–769.
Page R. A., T. Schnelle, E. K. V. Kalko, T. Bunge, & X. E. Bernal. (2012). Reassess-
ment of prey through sequential use of multiple sensory cues by an eavesdrop-
ping bat. Naturwissenschaften, 99, 505–509.
Paterson, H. E. H. (1978). More evidence against speciation by reinforcement. South
African Journal of Science, 74, 369–371.
BE LOVED , BE PREY , BE EATEN 151

Phelps, S. M., A. S. Rand, & M. J. Ryan. (2007). The mixed-species chorus as public
information: Túngara frogs eavesdrop on a heterospecific. Behavioral Ecology,
18, 108–114.
Pinto, C. M. & X. E. Bernal. (In preparation). A new Trypanosoma species from neo-
tropical frogs, with notes on the taxonomy of anuran trypanosomes.
Rand, A. S., M. E. Bridarolli, L. Dries, & M. J. Ryan. (1997). Light levels influence
female choice in túngara frogs: Predation risk assessment? Copeia, 1997,
447–450.
Robert, D., J. Amoroso, & R. R. Hoy. (1992). The evolutionary convergence of hear-
ing in a parasitoid fly and its cricket host. Science, 258, 1135–1137.
Robert, D., M. P. Read, & R. R. Hoy. (1994). The tympanal hearing organ of the
parasitoid fly Ormia ochracea (Diptera, Tachinidae, Ormiini). Cell Tissue
Research, 275, 63–78.
Rodrigues, F. H. G., M. L. Reis, & V. S. Braz. (2004). Food habits of the frog-eating
bat, Trachops cirrhosus, in Atlantic forest of northeastern Brazil. Chiroptera
Neotropical, 10, 180–182.
Rosenthal, G. G., T. Y. F. Martinez, F. J. G. de Leon, & M. J. Ryan. (2001). Shared
preferences by predators and females for male ornaments in swordtails. The
American Naturalist, 158, 146–154.
Ryan, M. J. (2009). Communication in frogs and toads. In L. R. Squire (ed.), Ency-
clopedia of Neuroscience (pp. 1159–1166). Oxford, UK: Academic Press.
Ryan, M. J. (2010). The túngara frog: A model for sexual selection and communica-
tion. In M. D. Breed & J. Moore (eds.), Encyclopedia of Animal Behavior (pp.
453–461). Oxford, UK: Academic Press.
Ryan, M. J., X. E. Bernal, & A. S. Rand. (2010). Female mate choice and the poten-
tial for ornament evolution in the túngara frog Physalaemus pustulosus. Current
Zoology, 56, 343–357.
Ryan, M. J., J. H. Fox, W. Wilczynski, & A. S. Rand. (1990). Sexual selection for
sensory exploitation in the frog Physalaemus pustulosus. Nature, 343, 66–67.
Ryan, M. J. & A. Keddy-Hector. (1992). Directional patterns of female mate choice
and the role of sensory biases. The American Naturalist, 139, S4–S35.
Ryan, M. J. & A. S. Rand. (1995). Female responses to ancestral advertisement calls
in tungara frogs. Science, 269, 390–392.
Ryan, M. J. & A. S. Rand. (1999). Phylogenetic influences on mating call preferences in
female túngara frogs (Physalaemus pustulosus). Animal Behaviour, 57, 945–956.
Ryan, M. J., W. Rand, P. L. Hurd, S. M. Phelps, & A. S. Rand. (2003). Generaliza-
tion in response to mate recognition signals. The American Naturalist, 161,
380–394.
Ryan, M. J. & M. D. Tuttle. (1983). The ability of the frog-eating bat to discriminate
among novel and potentially poisonous frog species using acoustic cues. Ani-
mal Behaviour, 31, 827–833.
Ryan, M. J., M. D. Tuttle, & R. M. R. Barclay. (1983). Behavioral responses of the
frog-eating bat, Trachops cirrhosus, to sonic frequencies. Journal of Comparative
Physiology A, 150, 413–418.
152 ANIMAL BEHAVIOR

Ryan, M. J., M. D. Tuttle, & A. S. Rand. (1982). Bat predation and sexual advertise-
ment in a neotropical anuran. The American Naturalist, 119, 136–139.
Ryan, M. J., M. D. Tuttle, & L. K. Taft. (1981). The costs and benefits of frog chor-
using behavior. Behavioral Ecology and Sociobiology, 8, 273–278.
Sakaluk, S. K. & J. J. Bellwood. (1984). Gecko phonotaxis to cricket calling song: A
case of satellite predation. Animal Behaviour, 32, 659–662.
Seehausen, O., Y. Terai, I. S. Magalhaes, K. L. Carleton, H. D. J. Mrosso, R. Miyagi,
I. van der Sluijs, M. V. Schneider, M. E. Maan, H. Tachida, H. Imai, & N.
Okada. (2008). Speciation through sensory drive in cichlid fish. Nature, 455,
620–626.
Stoddard, P. K. (1999). Predation enhances complexity in the evolution of electric
fish signals. Nature, 400, 254–256.
Théry, M. & J. Casas. (2002). Predator and prey views of spider camouflage. Nature,
415, 133.
Thurstone, L. L. (1927). A law of comparative judgment. Psychological Review, 34,
273–286.
Tuttle, M. D. & M. J. Ryan. (1981). Bat predation and the evolution of frog vocal-
izations in the neotropics. Science, 214, 677–678.
Tuttle, M. D., L. K. Taft, & M. J. Ryan (1982). Acoustical location of calling frogs by
Philander opossums. Biotropica, 13, 233–234.
Tuttle, M. D., L. K. Taft, & M. J. Ryan. (1982). Evasive behavior of a frog in
response to bat predation. Animal Behaviour, 30, 393–397.
von Békésy, G. (1960). Experiments in Hearing. New York: McGraw-Hill.
Wells, K. D. (2007). The Ecology and Beahvior of Amphibians. Chicago: University of
Chicago Press.
Wilczynski, W., A. S. Rand, & M. J. Ryan. (1995). The processing of spectral cues by
the call analysis system of the túngara frog, Physalaemus pustulosus. Animal
Behaviour, 49, 911–929.
Wood, D. M. and A. Borkent. (1989). Phylogeny and classification of the Nemato-
cera. In J. F. Alpine & D. M. Wood (eds.), Manual of Nearctic Diptera 3
(pp. 1333–1370). Research Branch, Agriculture Canada, Monograph 32.
Zuk, M. & G. R. Kolluru. (1998). Exploitation of sexual signals by predators and par-
asitoids. Quarterly Review of Biology, 73, 415–438.
Zuk, M., J. T. Rotenberry, & R. M. Tinghitella. (2006). Silent night: Adaptive disap-
pearance of a sexual signal in a parasitized population of field crickets. Biology
Letters, 2, 521–524.
Zuk, M., L. W. Simmons, & L. Cupp. (1993). Calling characteristics of parasitized
and unparasitized populations of the field cricket Teleogryllus oceanicus. Behav-
ioral Ecology and Sociobiology, 33, 339–343.
 6

Siblicide in Birds
Douglas W. Mock

INTRODUCTION
The word siblicide entered the Oxford English Dictionary (OED) several years
ago (simultaneously with grunge!) as a biological gender-neutral alternative to
fratricide and a Bible-free alternative to Cainism. Like many English words,
it has since been used in various ways, but it is most useful when referring to
the death of a dependent offspring that was caused at least in part by a
nursery-mate’s overt aggression. Cain slew Abel; some degree of intentional
mayhem is implied. As such, it is a subset of the wider category of early mor-
tality known as brood reduction (frequency-dependent mortality of depen-
dent young, which commonly occurs without such hostility when parents
just run short of food and allocate the limited supply unevenly). Siblicide is
thus more active, seemingly nastier, and perpetrated by nestmates. The extent
of sibling aggression can be highly variable and is usually documented via
direct behavioral observations but can sometimes be inferred from the nature
of superficial wounds.
Siblicide is not particularly common; indeed, it is relatively rare so far as we
know. It is interesting because it is extreme, counterintuitive, and instructive.
Often, natural selection favors behavior and other traits that increase the wel-
fare of close genetic relatives (see Volume 2, Chapter 11), so this category rep-
resents an exception that challenges us to explain why it has evolved. That is,
we must seek to understand the special circumstances under which individuals
apparently can increase their genes’ replication rates by destroying the very
154 ANIMAL BEHAVIOR

individuals (kin) that carry more of such material than anyone else. Further-
more, the Darwinian interests of other family members are automatically
affected by internecine aggression, leading to secondary questions about what
those individuals should do in response. Although siblicide has been studied
most closely in birds, largely for practical reasons, parallels are found in such
diverse taxa as fruiting plants, sharks, foxes, and barnyard piglets (Sus scrofa
domestica)—and even human royal families (reviewed in Mock & Parker,
1997; Mock, 2004). That diversity suggests that it evolves repeatedly in
response to a constellation of widely shared ecological circumstances that we
have now made some progress toward identifying and testing. Thus it has
its theoretical roots in the concept of inclusive fitness (see Volume 2, Chapter
11), but whereas that construct is usually cited to help explain very positive
aspects of social behavior (e.g., generosity, cooperation, and altruism), sibli-
cide dramatically reveals its upper ecological limits. In short, siblicide is the
apex of selfishness, evolving only when the toleration of others, even those
with whom the individual’s interests overlap maximally, can no longer be
afforded.

THE FAMILY AS CRUCIBLE OF COOPERATION AND CONFLICT

To understand the ecological factors affecting family social dynamics, the
basics of sexual reproduction must be held in mind. This can be done for-
mally, but a quip attributed to evolutionary geneticist J. B. S. Haldane cap-
tures the essence neatly: “I would risk my life for two brothers or eight
cousins.” In sexual diploid organisms like us, sperm and eggs are formed in
the meiosis version of cell division, wherein each chromosome aligns with its
partner (the homolog), then drifts apart into the new cells without duplicating
first. Most obviously, this means that the resulting sperm and ova contain
unpaired chromosomes (are haploid), subsequently uniting with a similarly
haploid gamete from the mating partner to restore diploid condition for the
offspring. In this simple sense, then, a thinking gene that “wanted” to direct
its genial behavior to copies of itself but could not identify which of its siblings
happened to be a carrier might have no better option than being half as nice to
twice as many nestmates. The flip side of that logic is that it might do better by
investing in itself, a guaranteed carrier (Dawkins, 1976). Without going into
the argument further, it works out clearly enough that there is a considerable
range of circumstances under which its overall replication success is higher
by withholding largess to nestmates (O’Connor, 1978). From there, it is easy
to see that parental sexual habits feed into this drama. If two nestmates have
the same mother and two different fathers (because the resident female
SIBLICIDE IN BIRDS 155

accepted sperm from more than one male), the incentive for sibling generosity
is halved. Ceteris paribus, we expect full siblings (same genetic parents) to be
nicer to one another than half-siblings; in practice, the species’s mating system
is more likely to tune the location of this threshold, but there are suspected
cases where actual detection of relatedness may exist (see Volume 2, Chapters
7 and 9). An extreme case is represented by brood parasites (Davies, 2000),
wherein totally unrelated conspecifics may be introduced into someone else’s
nest (e.g., American coots [Fulica americana]); of course, this trick has been
known for centuries to operate between species (e.g., European cuckoos
[Cuculus canorus]).
But even in a simple nuclear family with nonphilandering parents, the sit-
uation is far from simple. For one thing, each of the concurrent offspring
should see the family budget through a similar lens, such that each would pre-
fer to receive somewhat more than its fair share of limited resources. This is
the essence of sibling rivalry, a mismatch between the supply (typically pro-
vided by one or both of the genetic parents) and demand (the growth and
energy expenditures of dependent offspring). In practice, the limited com-
modity is usually food, so we can explore further by focusing on that currency.
Being young, dependent offspring tend to grow very rapidly while also being
relatively vulnerable to predation, thermal stresses, and other dangers. Parents
are often under pressure to provide at a high rate. All parties have clear but dis-
parate interests in the family’s total food budget (Clutton-Brock, 1991).
In this process, the primary role of a parent is to create, protect, and support
offspring, but its fitness interests are also shaped by the environment (e.g.,
local traffic in predators, current availability of food, vagaries of weather,
etc.) and by history. In most species, parents breed more than once, so the
amount laid out for an earlier brood may abridge what they have to offer the
current offspring. Similarly, if parents are young and have a considerable
future to protect, they may do well to conserve their energy and pace them-
selves. In the process of conserving, they are lowering the disposable income
available to the current brood and altering the choices available to those off-
spring. In this light, Robert Trivers’s (1974) theory of parent-offspring con-
flict (POC) has been immensely provocative. Following the simplified
genetic logic sketched above, POC puts a clear focus on social asymmetries.
Basically, parents are equally related to all of their offspring (cuckoldry and
such set aside for the moment) and thus might be expected to value all of them
to an equal degree. More specifically, because of how meiosis works, a diploid
sexual parent has a coefficient of relatedness of 0.50 to every progeny (each
got half its DNA from that parent). But, as we have just seen, offspring regard
each other as asymmetrically valuable as bearers of gene copies. Insofar as the
156 ANIMAL BEHAVIOR

genes themselves are concerned, then, parents might be expected to favor

equal allocation of parental investment (see Volume 2, Chapter 7; essentially
this means parental effort and resources are finite), while each brood member
should generally try to skew such investment preferentially toward itself.
Finally, the two parents are very seldom close genetic relatives of each other
(lest their offspring suffer from inbreeding depression, or the concentration of
harmful recessive alleles), yet they are engaged in a cooperative venture (sexual
reproduction) for which they need each other. In taxa where offspring require
considerable postmating parental care (e.g., most birds and mammals), the mat-
ter of how those expenses will be met indicates another zone of potential conflict
that is likely to have ramifications for offspring welfare. The adult partnerships
vary tremendously in temporal duration and level of mutual cooperation. In
most sexual animals the male parent provides only sperm, leaving all subsequent
investment to the mother, but in fishes that fertilize ova externally the reverse
occurs, with the female deserting quickly while the male is still engaged in his
slightly deferred half of the sexual union, such that he is the only one remaining
to provide protection and other services. Between those extremes are a relatively
few species (the great majority of them feathered) in which biparental care is the
norm. Even within the birds, the two adults may work together during just one
breeding cycle, such that neither has a stake in whether the other emerges in
good condition (see Volume 2, Chapter 6). At the other extreme, swan and
goose partners may remain together for all future cycles, with each standing to
gain from ensuring that the other has a long and vigorous future.
Overall, then, even a “simple nuclear family” can be regarded as having
three distinct social dimensions likely to interact in key respects. The two
parents may be selfish or cooperative, young or old, healthy or frail—all of
which will shape the amount of parental investment available for raising the
current brood. The offspring may routinely (see next two sections) demand
more investment than is available, producing sibling competition and related
manifestations of nestmate selfishness (including aggression). And the parents
may dispute the partitioning of their own investment among multiple brood
members. In the context of siblicide, for example, an obvious question is
whether the parents should try to thwart the fighting.

THE LIFE HISTORY BACKDROP: WHY PARENTS OVERPRODUCE

The fundamental behavioral decision on family size is made by parents, pri-
marily mothers. In many taxa, more offspring are created than normally sur-
vive to independence (sometimes a lot more), raising the obvious question of
why natural selection does not favor smaller families in the first place. Some
SIBLICIDE IN BIRDS 157

of the most puzzling examples, dating from Aristotle, are large birds (eagles,
cranes, certain seabirds) that routinely lay two eggs but virtually never raise
more than one chick. This seems fundamentally wasteful, especially when
the secondary downward adjustment to brood size involves horrifying levels
of bloodshed and looks vaguely exhausting even for the victorious chick. In
one nest of an African black eagle (Aquila verreauxi), for example, the first-
hatched nestling (the “A chick”) was observed as it pecked relentlessly at a hap-
less younger sibling (“B chick”) from the moment of the latter’s hatching until
its death three days later, and a meticulous observer counted 1,579 blows! If
one could assess the nutrient and energetic costs of building that second egg
and incubating it for several weeks, then fueling the A chick’s sustained assault
on its victim (also paid for with parentally delivered food), the option of sim-
ply stopping with one egg has a certain attraction. It follows that parental fit-
ness probably obtains a boost from the initial overproduction that
adequately compensates, on average, for all these costs. Otherwise, one is left
with what one eagle observer called “an inexplicable example of apparent bio-
logical waste” (Brown et al., 1977). Furthermore, siblicidal eagles are trashing
only one offspring, whereas many plants shed hundreds or even thousands of
immature seeds and fruits, as if skewing investment to a smaller number that
meet some unknown affordability criteria. In short, there seems to be a wide-
spread trade-off between quantity and quality of offspring for a given repro-
ductive cycle that parents can regulate by dropping unripe fruit or, as in
eagles, by leaving a messy execution task to certain offspring.
The incentives for initial parental overproduction fall into three nonexclu-
sive categories, and we can formalize the problem in terms of labeling one sub-
set of the offspring as core members (the ones that parents might normally
expect to raise fully) and the extras as marginals (Mock & Forbes, 1995).
This discussion will employ the modest family size of two eaglets to keep
things simple, so there is one core egg and one marginal. The most famous
explanation for laying that second egg has already been mentioned and centers
on the parents’ imperfect information about what hunting conditions will be
like several weeks after laying is finished: only by aiming high (the “optimistic
clutch size”) can parents capitalize on potential good fortune. For the nontri-
vial price of metabolizing that second egg and providing it with an extra mea-
sure of body warmth, parents may be able to double their cycle’s output.
Because natural selection is essentially a reproductive race against other mem-
bers of the same population, genes promoting second-egg production may
spread for this reason. Several factors will shape this potential payoff, including
just how often those beneficent hunting conditions will materialize. This idea
was sketched out many years ago (Lack, 1947) and can be tagged as a
158 ANIMAL BEHAVIOR

“resource-tracking” strategy: parents that overproduce can realize their repro-

ductive potential more fully if and when resources turn out to be high. But
there is yet another cost to hold in mind, namely that associated with trim-
ming family size secondarily on those occasions when resources are less favor-
able (which conditions may hold most of the time). Before moving on to the
next incentive, it is worth noting that this argument does not mean that eagle
parents have to work any of these probabilities out mathematically, as an edu-
cated human might, just that those genetically predisposed to build a second
egg will turn out to outperform those with genes that incline them to stop
after laying only one. The temptation to anthropomorphize, imputing our
own cognitive skills onto such parents, may be calmed by realizing that many
flowering plants face (and solve) the same numerical problem. A fruit tree, for
example, makes the equivalent of a decision about whether to retain or abscise
a given green fruit based on how well that offspring is developing. If it has
been damaged by insects or birds, for example, it tends to be dropped early.
If the tree is not getting enough sunlight, rain, or soil nutrients, it terminates
large numbers of progeny. Accordingly, brain-equipped orchard farmers may
reduce the loss of peaches by adding fertilizer at the right moments, effectively
inflating the family budget of their trees and shifting the physiological perfor-
mance of the developing fruits. Needless to say, the trees do not need to
understand any life-history theory to participate in this process (nor, for that
matter, do the farmers).
The second fitness incentive parents can gain from adding a marginal off-
spring comes as a biological equivalent of insurance against loss of the core
members of the brood. Any number of mishaps may befall a given dependent
young, ranging from hatching failure (e.g., ontogenetic malfunctions) to
single-offspring seizures by predators and even stochastic accidents like falling
out of a tree nest. To illustrate, if we imagine a 10 percent risk of loss for a sin-
gle core offspring, the presence of a backup (also carrying a 10 percent chance
of failure) reduces the parents’ overall likelihood of total disappointment to
just 1 percent (because the probability that both will fail is an exponent of risk,
here 0.102). From the parents’ vantage, this is often highly attractive, even
though it may lead to an unpleasant and short life for a marginal that is not
promoted to core status. In many songbird species, a typical brood of four
contains one runt that hatches a day later than its three older siblings and is
usually the first to starve, but it has three slots that it potentially could fill if
one of its senior nestmates were to falter; using our 10 percent risk level in this
context, such a marginal chick stands a 30 percent chance of being promoted.
An elegant field experiment using one such system revealed that adding a sec-
ond marginal chick (backup 2) does not affect the survival rates of core brood
SIBLICIDE IN BIRDS 159

members, whose age and size advantages insulate them from competitive pres-
sures, but adding an extra core chick spells disaster for any marginal backups
(Forbes et al., 1997).
Lastly, a marginal offspring can sometimes provide tangible goods or serv-
ices to core siblings. Minimally, it adds to the nestmate huddle, reducing
everyone’s exposure if air temperatures are chilly. More dramatically, it may
be sacrificed cannibalistically, in which case it may represent a final buffer
against starvation in dire food shortages. And in some species, it may serve
later in life as a nonbreeding worker to enrich the reproductive success of sib-
lings or parents. In principle, these three classes of fitness payoffs (resource-
tracking, insuring, and aiding core siblings) may be additive if, for example,
the marginal offspring serves temporarily as an insurance back-up, but if local
food conditions suddenly bloom, it may thrive and add value as either a future
breeder (creating nieces and nephews) or a servant.

THE ECOLOGICAL CRUNCH

The upshot of parental overproduction is that the family food budget is likely
to prove inadequate under many or most circumstances, such that truly com-
petitive conditions impinge upon sibling dynamics as a routine matter of course.
After all, hatchling birds are usually tiny, naked (or nearly so), and ectothermic
(“cold-blooded”), requiring thermal assistance from their warm parents and
increasing rates of delivery for high-quality food that will support extraordinarily
rapid growth and development in the near future. As endothermy is achieved
and the need for a brooding parent is relaxed, both parents are likely to forage
nearly all the time on behalf of their increasingly ravenous brood. Even so, an
inclement spell of rain or chill can easily depress food availability or increase
physiological maintenance costs such that the broodmates’ collective demands
exceed parental delivery capacity. It is thus quite routine for a shortfall to materi-
alize, in which case something must give. Such a crunch might be dealt with in
several ways, including all nestmates taking an across-the-board pay cut and
growing at a slower rate or stronger ones taking self-promoting measures that
concentrate the misfortune on less competitive siblings. It is here that the inter-
ested researcher must analyze the various evolutionary interests of each family
member dispassionately to see where their inclusive fitness payoffs are maxi-
mized. In many cases, the option of a family-wide slowdown is not viable
because of other threats to the vulnerable brood, notably discovery by local
predators. So, with the clock running steadily on danger to the whole parental
enterprise, insufficient food usually means that someone must be sacrificed for
the good of the others, the general solution known as brood reduction.
160 ANIMAL BEHAVIOR

PARENTAL CONTROL AND MANIPULATION OF SIBLING ASYMMETRIES

During a resource shortage, all family members face these shared problems,
but the options available to each may differ dramatically by virtue of their
varying attributes. A marginal offspring, for example, presumably would prefer
not to be the designated victim in a trimming of family size, but it seldom gets
much of a choice. Similarly, contemporaneous nestlings of all status types
might like parents to work harder in promoting their immediate welfare, even
though this is likely, in principle, to diminish production of future siblings.
However, all they can do is importune. The transcendent difference between
the two generations is that the adult players are larger, stronger, more experi-
enced, and better informed about ecological conditions beyond the nest’s con-
fines. Moreover, they commonly have the first move, setting certain processes
into motion before the youngsters have any power at all. On the other hand, it
has been proposed that there may be various things offspring can do to
manipulate parents (e.g., by providing misinformation about how desperate
their physical condition is), essentially tricking adults into serving their inter-
ests. Provocative though such suggestions are, offspring psychological weap-
onry usually pales in comparison with the array of tools available to parents
for dealing with family decisions. As noted above, parents make the initial
move of overproducing that creates sibling rivalry in the first place; there are
also despotic manipulations parents can use to shape and reshape the nature
of that rivalry when it arises.
The most drastic of these parental options is simple desertion: adults can
leave at any time, allowing dependents to perish. Of course, that sacrifices
the investment already made and, at many latitudes, the chance for successful
reproduction that year, so it is chosen only under dire conditions. For exam-
ple, in long-lived aquatic birds like gulls and herons, parents may abandon
en masse when the nesting colony becomes a favorite hunting patch for a
predator they cannot deter and poses a threat to their future reproductive pros-
pects. Absent a direct threat to the adults themselves, deteriorating foraging
conditions may lead parents to forsake their broods rather than trying to invest
in a lost cause. Wood storks (Mycteria americana) in the American Southeast,
for example, abandon whole colonies of half-grown young if persistent rains
materialize during the dry season, flooding the local pools. These storks hunt
fish by touch, which works well only when prey are concentrated in shrinking
catchments. So unseasonal rains cut into the food budget and make sustained
effort a worse option than deserting and trying again next year, a viable option
considering that the adults may live 20 years and more. In various mammals,
mothers facing dire food shortages or imminent threat from predators may
desert or even cannibalize their litters, recycling the nutrients. On a less
SIBLICIDE IN BIRDS 161

extreme level, the Old World black stork (Ciconia nigra) has been known to
perform selective removal of an individual nestling, dropping it from the tree.
Mammalian mothers may sacrifice partial litters at various stages. Indeed, mar-
supial females divide litter production into two distinct stages, often produc-
ing more offspring in their vaginal parturition than they have nipples in the
pouch: the unattached extras are shed like small green applets dropping from
a tree.
Usually, parents combine overproduction with some preemptive engineer-
ing of brood members’ competitive abilities such that core young enjoy physi-
cal advantages over their less fortunate marginal siblings. Because
overproduction is typically followed by corrective brood reduction, parental
issuance of advantages to certain progeny (and/or handicaps to others) can
minimize the costliness of that secondary adjustment, especially if enforce-
ment is left to the core offspring. Quite simply, a brood of unequals can be
trimmed less expensively than one of evenly matched rivals. There are several
ways to do such “evolutionary engineering,” the best known being the wide-
spread avian habit of parents commencing effective incubation prior to the
completion of laying. Avian eggs that are not incubated seldom reach the tem-
perature threshold for efficient metabolism and growth (so-called physiologi-
cal zero, generally about 30°C). The deliberate application of parental
warmth facilitates biochemical processes supporting growth and thus acceler-
ates embryonic development within any eggs present. No birds are known to
lay more than one egg per day (larger birds often take two or more days
between eggs), so early incubation gives early embryos a head start, which
translates into hatching asynchrony. Upon emerging from the rigid confines
of its shell, a hatchling is no longer exclusively reliant on its yolk supply for
fuel and starts receiving supplements from parents. It also can exercise more
freely, gaining strength and size. The span of emergence times varies across
birds from zero (total synchrony,* which is provocatively found in precocial
birds like ducklings that tend to feed themselves and thus do not share a
parentally delivered food budget) to a week or more for certain owls and par-
rots. Size disparities among even slightly different-aged nestmates closely
match their position in the hatching order. Not surprisingly, the last to hatch
is usually the first to die in times of shortage.

*
For those wondering how eggs can hatch synchronously despite being laid at least 24 hours apart, the key
detail is that development requires maintenance at a minimum temperature for efficient cell division. In
most habitats the air temperature is below this threshold; thus eggs must be boosted by the application of
parental warmth (incubation). Thus, parents can “engineer” rough synchrony merely by not incubating
before the last egg is laid. In addition, synchronous hatching may be coordinated to a finer degree by the
young vocalizing while still inside the egg.
162 ANIMAL BEHAVIOR

Hatching asynchrony is not the only means parents have for playing favor-
ites. In some birds, mothers adjust the physical sizes of the eggs themselves,
and a burgeoning literature is now documenting the degree to which yolk ste-
roids may vary as a function of laying order. Even after all have hatched,
parents may continue to skew investment toward some offspring at the
expense of others. In mammals, there are better and worse portions of the
uterus for embryos to implant and even varying milk supplies associated with
particular nipples in litter-bearing species.
It is periodically suggested that overproduction serves parental interests as
a screening process. The idea is that creating a surplus initially could allow
parents a chance to assess and then select the most promising individuals,
thereby upgrading those who will be retained for full investment. This has
intuitive appeal for anyone who has test-driven several cars before purchase,
but it is easy to overlook the fact that test drives are free. Once a female bird
has produced an egg (typically constituting a nontrivial fraction of her body
mass) or a female mammal has carried a fetus to term, the costs are substan-
tial. By contrast, in various plants, such an argument looks promising. For
one thing, the scale of overproduction can be orders of magnitude higher
than for most animals. Moreover, a female plant’s opportunity for premat-
ing choosiness of sexual partner (pollen source) is severely constrained by
her inability to move or otherwise avoid fertilization by a male with which
she is genetically incompatible. Some wonderful experiments have shown
that fruit abscission (separation of fruit from a branch) can depend critically
on paternity (i.e., on pollen genotypes), lending clear support for this view.
Specifically, ova fertilized by pollen from donor plants adapted to similar
conditions as the receiving plant are retained, while those pollinated by
males with less compatible genes tend to be abscised. The genetic comple-
mentarity of the two mated parents plays a key role in determining offspring
success.
This logic of parental choice seems ill suited to siblicidal birds for several
reasons, not least of which being that the whole handicapping exercise
(hatching asynchrony, yolk steroids, etc.) would tend to mask genetic varia-
tion, thereby eliminating the basis for making beneficial choices. A black
eagle that might benefit from hatching two eggs and “letting them fight it
out” to see which chick is more deserving of full investment would ruin
its own plan by randomly awarding one rival a three-day growth advantage;
the genetically superior chick would be killed by its inferior older sib
roughly half the time for reasons unrelated to their relative levels of intrinsic
quality.
SIBLICIDE IN BIRDS 163

THE USE OF OVERT AGGRESSION

Watching week-old egrets repeatedly pecking nestmates in the face is an
uncomfortable experience. Most sibling fights are instigated by a member of
the core brood and directed at the youngest, smallest, and weakest nestmate,
especially if that individual has the temerity to beg as a food-bearing parent
arrives. There is no escaping the sense that this is bullying and fundamentally
unfair. Even if the junior sibling tries to retaliate, which is fairly common for
awhile, the onslaught is most likely to intensify. Fights may be drawn-out
affairs with literally hundreds of seesawing pecks exchanged before one com-
batant concedes by crouching. Gradually, over a few days, they truncate to a
blow or two and even to simple threats (the aggressor assuming a tall posture
from which a jab might be delivered). Sometimes crouching does not suffice
to placate the aggressor, and the loser tries to flee by scrambling about within
the very tight confines of a nest cup; if that does not work, it may try to dangle
its head over the rim to get the nape, a favorite target, out of reach. Bruising
and lesions are often evident. The eventual outcome of these repeated bouts
is the establishment of a linear dominance hierarchy, usually matching the sib-
lings’ hatching order, and refreshed with a few fresh pecks each time a parent
returns from hunting and before boluses of freshly caught fish are regurgitated.
It seems quite clear that the point of the exercise is to intimidate lower-ranking
siblings from getting access to the food until their stronger comrades have
eaten their fill.
Parent great egrets (Ardea alba) capture and swallow large numbers of
rather small fish, transporting them home in the crop (storage stomach) and
regurgitating them as a tubular column that breaks apart as boluses of 5 to
15 such prey. The youngsters vocalize in apparent excitement and reach out
to grip the parent’s bill. In the first days, such scissoring lacks any clear func-
tion because the wads of fish are dropped to the nest floor, from which the
chicks hastily pluck individual morsels, quickly orienting each fish for head-
first swallowing. Fights during this period appear rather pointless (nobody
seems particularly intimidated) and may actually distract bullies somewhat
from the fish-gulping scramble type of competition. By the end of the first
week, however, the first-hatched A chick begins to catch boluses before they
hit the nest floor, and two mysteries become clear. The scissoring of the
parent’s bill is the necessary first step toward intercepting whole boluses,
which helps the well-positioned chick monopolize more of the total food.
Acquiring and holding that advantageous position is enhanced by the cumula-
tive effects of the early fighting, as junior sibs increasingly hesitate slightly
before scissoring. Naturally, the second-hatched B chick gets the knack of
scissoring soon, as does C (and D if such exists), but another dynamic is
164 ANIMAL BEHAVIOR

transforming the brood. Having established itself as the dominant nestling, A

becomes less involved in the actual fighting, though it will escalate harshly if a
lesser sibling gets in the way. In a modal brood of three, the spotlight shifts to
the B-C relationship with the cumulative intimidation of C appearing as B’s
chief objective. The first two or three boluses often go straight to A, while B
is pummeling C. Soon, A’s long neck is too engorged with food to bend or
deliver punishment. With A surfeited for the moment, B pivots from disci-
pline to gluttony. And C may end up with nothing by the time B is full. Over-
all, each of the two seniors consumes about 40 percent of the total food,
leaving only a half dose for C. In broods where total intimidation of C is
achieved, its share approaches zero and it succumbs, either crumpling in the
nest or on the ground below after being driven over the rim. Interestingly,
when the family food budget was experimentally doubled by providing
chopped fish supplements, sibling persecution of C chicks continued apace,
but these marginal chicks survived at a higher rate than controls despite the
physical persecution (Mock & Parker, 1997; Mock, 2004).
Prey small enough for nestlings to monopolize by intercepting boluses
directly thus seems to provide incremental incentives for aggression, as even
slight hesitation reduces a rival’s access to the key resource, the descending
parental bill. Great blue herons (Ardea herodias) in Texas breed alongside the
great egrets but seldom fight; intriguingly, their prey fish are too large for
chicks to catch shy of the nest floor. But when cross-fostered to great egret
parents (and thus receiving discrete boluses of small fish), these heron nestlings
also fought at high rates. A field study in a Quebec great blue heron colony,
where the adults catch mostly smaller fish, showed that these young also fight
at egret levels. It thus appears that aggression may develop where food is of a
type that can be monopolized through aggression.
Other factors also help account for the patchy distribution of sibling aggres-
sion, including weaponry (egret and heron bills make fine pikes) and nest
architecture. Kookaburras (Dacelo novaeguineae) illustrate both factors nicely,
with burrow nests that make the parents’ direction of arrival perfectly predict-
able and hook-shaped bills used to punish junior siblings (Legge, 2000). Com-
bined, these two features make it relatively easy for strong nestlings to assert
control over how the finite food deliveries will be divided; to say it another
way, they make it substantially easier for parents to realize their own preferred
optimal brood size.
Similarly, barnyard piglets are born with lateral deciduous teeth that are
used in side-directed slashing motions for acquiring and defending the (more
productive) anterior teats on the sow. If these are clipped experimentally from
the senior siblings but left intact on the runts, the latter press their weaponized
SIBLICIDE IN BIRDS 165

advantage and claim better positions, thus growing more rapidly than other-
wise (Fraser & Thompson, 1990). In general, one expects to find aggression
used any time there is something at stake (limited resources) that is economi-
cally favorable (gains exceeding costs). When those conditions do not hold,
there may still be just as much brood reduction but through less flamboyant
behavioral processes. In many songbirds, for example, comparable rates of
runt mortality have been reported as for these large siblicidal birds, but there
is a striking lack of evidence for overt aggression playing a central role. Of
course, a marginal chick that is simply unable to compete effectively for paren-
tal food deliveries and ends up succumbing to starvation is every bit as dead as
one that has been bludgeoned first by nestmates, so many of the puzzles
remain.

FITNESS CONSEQUENCES (COSTS AND BENEFITS)

As has been evident to this point, the basic cost/benefit analytical approach
that characterizes other subfields in behavioral ecology applies to family social
dynamics, but this is a good place to set up that framework more formally
while holding focus on the dodgy fate of a marginal brood member in a small
family. If raised successfully to fledge, then surviving adolescence and the risky
transition to independence to reach breeding age, its subsequent lifetime
reproductive success provides a useful measure of Darwinian fitness. Each of
its own successful offspring carries copies of half its genes, from which one
can derive the “direct component” of its overall or inclusive fitness. Mean-
while, if it had siblings that also thrived, its gene copies housed in those bodies
are replicating and being passed along, so it gains reduced (halved) dividend
from the production of nieces and nephews. Because that second component
of its inclusive fitness is derived from the performance of lateral kin (its sib-
lings), it constitutes “indirect” fitness.
Accordingly, whenever a senior sibling behaves so selfishly that its marginal
sibling dies, a unit of its indirect fitness is forfeited. Fatal selfishness by a core
offspring thus can be viewed as a trade-off between two strategies: tolerance
(in which case it stands to gain indirect fitness) versus intolerance (sacrificing
same). Natural selection blindly favors the first strategy if food is so plentiful
that the core offspring need sacrifice little or nothing from its own nutrition,
but it tips the balance in the other direction if that sacrifice is too great. It is easy
to show that this tipping point exists where the next unit of parental invest-
ment would enhance the marginal chick’s future prospects by at least twice as
much as the same unit would contribute the core chick. That solution applies
to a two-offspring brood, where the senior sibling is stronger and is essentially
166 ANIMAL BEHAVIOR

making the only decision. The model gets a bit less genial if there are two or
more senior siblings because then there are multiple parties carrying slightly
dissimilar agendas (slightly different tipping points). As well, as brood size
enlarges the loss of one chick represents a smaller fraction of the total fitness
pool affecting the surviving family members’ inclusive fitness scores. This argu-
ment need not be extended here, but it casts light on some interesting facets.
First, selection operates on results at the cycle’s end, so a few peculiar phe-
nomena that seemed anomalous initially must be considered in a somewhat
extended time frame. Early observations of siblicide in black eagles showed
that the senior chick typically kills its one junior sibling long before an actual
food shortage is realized. This gave the clear impression that the fighting must
not be over food (executions were documented amid piles of prey carcasses
strewn about). Confusion stemmed from assuming a temporal immediacy that
may not apply. That is, a surfeit of food early (when offspring are too small to
consume much) may well be the wrong cue on which to base a tolerance/intol-
erance decision if it is highly likely that the two growing eaglets will need so
much more food later that parental delivery rates will soon prove inadequate.
If that future crunch occurs often enough, then selection may well reward a
preemptive strategy that eliminates the sole rival early, while it is still small
and weak, rather than waiting until it is formidable and able to retaliate dan-
gerously. Black eagles are pigeonholed as practicing obligate siblicide, mean-
ing that the stronger hatchling virtually always kills its nestmate quickly,
which is comprehensible if deferred shortages are likely.
For species with less predictable resource shortfalls, the outcome is less pre-
determined. Here, the relevant cues are likely to materialize more gradually
and to vary across species that differ ecologically. Moreover, because death is
irreversible, nonlethal solutions that nevertheless suffice to buffer core brood
members during a risky period of ecological duress have the attractive feature
of flexibility by not crossing the point of no return. Specifically, if stronger off-
spring can control investment allocation well enough to avoid jeopardizing
themselves while retaining the marginal sibling’s viability —keeping it in a
temporary limbo from which it may still recover—toleration seems preferable.
Sparing the rod may allow them to redeem the runt’s indirect contributions,
but only if that does not spoil the child. In the rosy version of this scenario,
the runt may experience a mere delay in its growth trajectory. This has been
suspected for some time (e.g., Forbes & Ydenberg, 1992), but the necessary
data for testing it are elusive. Specifically, one needs a facultative (i.e.,
context-conditional) brood-reducing species in which the young produced in
one year can be rediscovered reliably some years down the line when they have
become adult breeders. Once located, these individuals must be followed for
SIBLICIDE IN BIRDS 167

enough of their reproductive lives to provide a clear picture of whether those

that had been marginals turned out to be “just as good” as their core nestmates
that had enjoyed more cosseted early lives.
These onerous empirical conditions have been met for a few long-lived sea-
birds, notably kittiwakes and boobies, both of which are facultatively siblicidal
and highly philopatric (returning faithfully to their home nesting colonies).
Kittiwakes are cliff-nesting gulls of higher latitudes, where parents lay their
two eggs on minimal nests on narrow ledges. These precarious nest sites pre-
vent most ground predators from bothering the family but also provide the
option of using gravity to help get rid of a junior nestmate that is judged to
be of insufficient remaining value to the senior chick (Braun & Hunt,
1983). Of interest in the present context, long-term analyses have now been
performed for a population of marked individuals from the French coast,
showing that junior siblings take longer to fledge and suffer much higher mor-
tality in the short term, both before and after fledging, but those juniors that
reach breeding age do not show reduced breeding success as adults (Cam et
al., 2003).
Several recent discoveries for boobies show mixed results that still point
toward rewarding toleration. Blue-footed boobies (Sula nebouxii), which have
been individually banded and followed for more than two decades, show that
subordinate siblings recruit into the breeding population as often as their
dominant nestmates (or singletons that had no rival to begin with), even
though they were physically abused, deprived of food, and physiologically
stressed (as shown by hormone profiles) as chicks (Drummond et al., 2011).
Moreover, their early ordeal does not seem to leave them with a subordinate
behavioral profile as adults, at least as shown by how they respond to a simu-
lated (mounted photograph) male intruder as adults (Sánchez-Macouzet &
Drummond, 2011). In a different species, though, adult Nazca boobies (Sula
granti) that had been bullied by their dominant nestmates as chicks are more
likely to exhibit violent behavior, including sexual assaults, on unrelated
young than are adults who had been spared such abuse when young. This pat-
tern seems to parallel the so-called cycle of violence reported for human chil-
dren (Müller et al., 2011).
Personality effects aside, these remarkable data from the kittiwake and
booby systems are very important to our understanding of why facultative
siblicide evolves. These birds seem to have flexible developmental programs
that enable individuals to withstand severe-looking food deprivation and
physical abuse early in life and to recover subsequently—perhaps fully in the
blue-footed booby—thus becoming functional reproductives in the breeding
population. To whatever degree they manage to succeed, their breeding
168 ANIMAL BEHAVIOR

success incentivizes the decisions made by siblings to tolerate (not kill) them as
nestlings. These are rare and precious findings but are limited to two groups of
marine birds that are both siblicidal and sufficiently philopatric to render these
long-term effects accessible. There is, of course, no guarantee that all siblicidal
species will follow suit. If birds that seldom show clemency to lesser siblings
(the so-called obligate siblicide of various cranes, pelicans, eagles, etc.), such
lack of toleration may have been shaped by lingering effects of early food dep-
rivation. Clearly further studies will be needed to help resolve this puzzle.

WHY NO ADAPTIVE SUICIDE?

In principle, a handicapped (marginal) sibling might maximize its own
inclusive fitness by dying quickly and quietly, inconveniencing other family
members as little as possible (O’Connor, 1978). This paradoxical possibility
might require only that its demise redirect enough food to siblings to enhance
their viability to a degree that compensates the altruist for its own lost repro-
duction. In practice, this is highly unlikely for several reasons, not least of
which being the victim should almost always achieve higher net fitness by sac-
rificing someone else (so it would prefer to push a nestmate over the side!).
Even when realistic options are considered, suicide may yield little competitive
release because a marginal chick’s share of the food is often too meager to
boost anyone by much. The typical pattern seems to be that the stronger core
siblings help themselves to disproportionate shares of the pie, promoting their
own chances for success and incrementally lessening a designated victim’s
value through gradual food deprivation. If continued, this dynamic widens
growth disparities. Just how the senior siblings enforce that social starvation
varies from simple self-feeding efforts (jockeying for advantageous position
near the delivering parents’ preferred nest-entry points, escalating signals of
conspicuousness to parents, etc.) to intimidation of junior siblings through
overt assault. In most well-studied avian siblicide systems, the cause of death
is a mélange of physical abuse and starvation, where these contributions are
inseparable. In egrets, for example, the bloodied and emaciated runt may suc-
cumb slowly in the nest over a period of many days, or it may be driven out in
dramatic fashion to fall from the tree and die on the colony floor below. When
contemplating the fitness impacts of a victim’s death (on parental and/or sib-
ling interests) it is important to realize that its actual reproductive value has
often eroded drastically prior to its actual execution: specifically, a bullying
senior sibling is not really discarding a family member that is still worth half
as much as itself (that value, derived from gene-sharing alone, is only its maxi-
mum theoretical value). In practice, the cumulative effects of ongoing food
SIBLICIDE IN BIRDS 169

thefts and physical abuse already meted out are likely to have abridged that
value considerably.
The unique feature of death in this context is, of course, that it is irrevers-
ible. Even if voluntary suicide is less likely than execution of a nestmate to
reduce competition for food, the benefits accruing to surviving siblings are
far from simple. In addition to the withering victim scenario sketched above,
we know of two species—cattle egrets (Bubulcus ibis) and brown pelicans
(Pelecanus occidentalis)—in which parents have been shown to respond to
brood reduction by keeping more of the total food budget for themselves
(Mock & Lamey, 1991; Ploger, 1997). In such birds, the fitness incentives
available for killing a nestmate are not readily apparent but surely include
the survivors’ viability being affordable to their parents under a wider array
of ecological conditions.

JUST AS DEAD
This article is on avian siblicide per se, but a few words must be offered
about forms of fatal sibling rivalry that do not involve overt aggression. Non-
aggressive forms of brood reduction are much more common and surely share
similar costs and benefits as the more brutal alternatives, so we may learn
many things about siblicide’s fitness trade-offs from species that do not
actually practice siblicide. Many avian taxa exhibit the patterns of overproduc-
tion and early parental manipulations of competitive ability, so the limitation
of food still seems to play the central role, even though the behavioral mecha-
nisms by which these supplies are distributed involve no violent activities.
Instead, uneven allocation is effected by such mundane processes as begging
(i.e., signaling to the food providers), vying for nest locations where food is
most likely to be transferred, and parents simply doling out food to certain
mouths preferentially. In these families, the victim typically dies from starva-
tion that is not enforced or accelerated by physical abuse. The end result is
much the same.
One way of looking at this comparison of siblicidal versus nonviolent brood
reduction is to postulate that the first category’s overt aggression is an attempt
by senior siblings to usurp parental control over family size. To the degree that
this is valid, these nonsiblicidal species may provide a clearer picture of what
parents are trying to accomplish. For example, tiny nestlings may signal their
desire for more food in various vivid ways, but they cannot coerce their parents
in any direct way to comply with those requests. Some clever indirect argu-
ments have been offered (e.g., that loud begging serves as a form of blackmail,
forcing parents to deliver more before a predator detects the nest), but for the
170 ANIMAL BEHAVIOR

most part begging is seen as merely the offspring’s weaker end of the dynamic.
There is much controversy surrounding the issues of just what these signals
convey, ranging from a helpless plea for aid (most exaggerated when the signal-
er’s need is highest; Godfray, 1991, 1995) to an advertisement of offspring
promise (most exaggerated when the signaler’s need is lowest; Mock et al.,
2011). All such discussions need to consider the possibility that parents are
not necessarily committed to bringing all nestlings to maturity, which merely
recognizes that the equivocation implied by sibling asymmetries may well
indicate brood reduction as part of the overall parental strategy.

PARENTAL COMPLICITY
The initially puzzling fact that the powerful adults stand by passively during
the lengthy process of siblicide, which ranges from a couple days up to multi-
ple weeks across species, is similarly best explained in terms of parents favoring
brood reduction under various circumstances. Long before the nestlings have
any real say over optimal family size, avian parents have set two preconditions
(an optimistic clutch size and a within-brood competitive hierarchy) that seem
superficially at odds with each other, but which can now be considered as two
sides of the same coin. If the initial overproduction carries sufficient value (as
protection for raising fewer or less viable young than can be afforded), then
manipulations of the competition keep brood-reduction costs low so as to
maximize the parents’ net fitness benefits.
This Machiavellian view of parental manipulation leads naturally to the
question of why parents do not simply do their own dirty work. Why is filial
(parental) infanticide not more common? We know that it is quite easy for
adults to kill young nestlings in many species because they readily dispatch
neighbors’ offspring (e.g., in dense nesting colonies) and may even use them
as food. So why not play hit man within the family once ecological conditions
have deteriorated to the point where parents should favor brood reduction?
This puzzle has not been discussed at much length, but the problem may
reduce to whether parents have any clear means of recognizing the hypotheti-
cal tipping point at which their interests are better served by the victim’s
death. There are simple enough reasons for imagining that such a point may
be reached sooner for senior siblings than for parents (O’Connor, 1978), so
the family member(s) most likely to take overt action are the victim’s peers.
Our expectations that parents should try to thwart siblicide can now be seen
as built on the assumption that these different thresholds are recognizable to
both generations, but a general explanation for laissez-faire parenting may rest
on the imperfect information that all parties have about the ecological realities
SIBLICIDE IN BIRDS 171

(i.e., family budget) under which the drama is unfolding. Once parents have
established an effective mechanism for family size to adjust itself, they may
have no better tactic than to let it play out on its own terms. It may be worth
noting that filial infanticide, when it has been observed, is very swift and sure,
so a more extended drama might be advantageous before the point of no
return is reached.

POSTSCRIPT: THE PAS DE DEUX OF THEORY AND DATA

The recorded natural history of avian siblicide dates back to Aristotle, and it
was almost certainly known long before that. As an extreme form of behavior,
it seems mildly surprising that more than two millennia passed before it began
to be studied closely until we remember that the earliest accounts concerned
great eagles, which are inherently difficult to study. By the late 1970s, though,
siblicide was observed in several taxa of colony-nesting birds (kittiwakes,
egrets, pelicans, and boobies), in addition to those raptors, so the logistics were
suddenly quite favorable: an observer could sit quietly (usually inside an obser-
vation blind) and see many such families at once. Of course, an even more
important thing that had happened since Aristotle’s time was the rise of Dar-
winism in the nineteenth century and its refinement ever since, particularly
the melding with Mendelian genetics. In addition, the subsequent extension
of gene thinking known as inclusive fitness theory (Hamilton, 1964) brought
family cooperation and conflict into clear focus and made sense out of various
behavioral oddities, including sterile castes in social insects and siblicide. Sig-
nificantly, if the siblicidal tendencies of colony-nesting birds had been noticed
even 20 years earlier—before the implications of inclusive fitness were appre-
ciated—they would probably have been treated as something peculiar that
was either “an inexplicable example of biological waste” or pathological. For
all we know, it was noticed earlier but nobody got excited. Einstein wrote “it
is theory which determines what we can observe” (Heisenberg, 1971), which
seems likely for particle physics, but for field biology a topic like siblicide
shows us that theory may be necessary for recognizing that something we have
known about for 23 centuries is important enough to understand. The killing
of close kin provides a fine example of phenomena for which theory and
empiricism must operate in concert.

ACKNOWLEDGMENTS
I thank Ken Yasukawa for his patience and Matt Taylor and Elise Knowl-
ton for providing comments on an earlier draft of this manuscript.
172 ANIMAL BEHAVIOR

REFERENCES AND SUGGESTED READING

Braun, B. M. & G. L. Hunt Jr. (1983). Brood reduction in black-legged kittiwakes.
Auk, 100, 469–476.
Brown, L. H., V. Gargett, & P. Steyn. (1977). Breeding success in some African
eagles relative to theories about sibling aggression and its effects. Ostrich, 48,
65–71.
Cam, E., J.-Y. Monnat, & J. E. Hines. (2003). Long-term fitness consequences of
early conditions in the kittiwake. Journal of Animal Ecology, 72, 411–424.
Clutton-Brock, T. H. (1991). The Evolution of Parental Care. Princeton, NJ: Prince-
ton University Press.
Davies, N. B. (2000). Cuckoos, Cowbirds and Other Cheats. Princeton, NJ: Princeton
University Press.
Dawkins, R. (1976). The Selfish Gene. London: Oxford University Press.
Drummond, H., C. Rodríguez, & D. Oro. (2011). Natural “poor start” does not
increase mortality over the lifetime. Proceedings of the Royal Society of London,
B, 278, 3421–3427.
Forbes, L. S. (1990). Insurance offspring and the evolution of avian clutch size. Jour-
nal of Theoretical Biology, 147, 345–359.
Forbes, L. S., S. Thornton, B. Glassey, M. Forbes, & N. J. Buckley. (1997). Why
parent birds play favourites. Nature, 390, 351–352.
Forbes, L. S. & R. C. Ydenberg. (1992). Sibling rivalry in a variable environment.
Theoretical Population Biology, 41, 335–360.
Fraser, D. & B. K. Thompson. (1990). Armed sibling rivalry among piglets. Behav-
ioral Ecology and Sociobiology, 29, 9–15.
Godfray, H. C. J. (1991). The signalling of need by offspring to their parents. Nature,
353, 328–330.
Godfray, H. C. J. (1995). Signalling of need between parents and young: parent-
offspring conflict and sibling rivalry. American Naturalist, 146, 1–24.
Hamilton, W. D. (1964). The genetical evolution of social behaviour. Journal of
Theoretical Biology, 7, 17–52.
Heisenberg, W. (1971). Physics and Beyond: Encounters and Conversations. New York:
Harper & Row.
Lack, D. (1947). The significance of clutch-size. Parts 1 and 2. Ibis, 89, 302–352.
Legge, S. (2000). Siblicide in the cooperatively breeding laughing kookaburra (Dacelo
novaeguineae). Behavioral Ecology and Sociobiology, 48, 293–302.
Mock, D. W. (2004). More than Kin and Less than Kind. Cambridge, MA: Belknap/
Harvard University Press.
Mock, D. W., M. B. Dugas, & S. A. Strickler. (2011). Honest begging: Expanding
from signal of need. Behavioral Ecology, 22, 909–917.
Mock, D. W. & L. S. Forbes. (1992). Parent-offspring conflict: A case of arrested
development? Trends in Ecology and Evolution, 7, 409–413.
Mock, D. W. & L. S. Forbes. (1995). The evolution of parental optimism. Trends in
Ecology and Evolution, 10, 130–134.
SIBLICIDE IN BIRDS 173

Mock, D. W. & T. C. Lamey. (1991). The role of brood size in regulating egret sib-
ling aggression. American Naturalist, 138, 1015–1026.
Mock, D. W. & G. A. Parker. (1986). Advantages and disadvantages of ardeid brood
reduction. Evolution, 40, 459–470.
Mock, D. W. & G. A. Parker. (1997). The Evolution of Sibling Rivalry. London:
Oxford University Press.
Mock, D. W. & G. A. Parker. (1998). Siblicide, family conflict, and the evolutionary
limits of selfishness. Animal Behaviour, 56, 1–10.
Müller M. S., E. T. Porter, J. K. Grace, J. A. Awkerman, K. T. Birchler, A. R. Gun-
derson, E. G. Schneider, M. A. Westbrock, & D. J. Anderson. (2011). Mal-
treated nestlings exhibit correlated maltreatment as adults: Evidence of a
“cycle of violence” in Nazca boobies (Sula granti). Auk, 128, 615–619.
O’Connor, R. J. (1978). Brood reduction in birds: Selection for infanticide, fratricide,
and suicide? Animal Behaviour, 26, 79–96.
Ploger, B. J. (1997). Does brood reduction provide nestling survivors with a food
bonus? Animal Behaviour, 54, 1063–1076.
Sánchez-Macouzet, O. & H. Drummond. (2011). Sibling bullying during infancy
does not make wimpy adults. Biology Letters, 7, 869–871.
Trivers, R. L. (1974) Parent-offspring conflict. American Zoologist, 14, 249–264.
This page intentionally left blank
 7

Referential Signals: A Window into

Animal Minds
Carolynn L. Smith and Christopher S. Evans

REFERENTIAL SIGNALS: INSIGHT OR ILLUSION?

Many people imagine that only humans have any ability to share information
about the world. But this is quite wrong. Moreover, it’s a belief that is relatively
recent. Before Descartes, there was a widespread feeling that animals had a rich
mental life, including sophisticated communication (Cheney & Seyfarth,
1990; Seyfarth & Cheney, 2010). Part of the problem is a misunderstanding
of Charles Darwin. In his 1879 book, On the Expression of the Emotions in
Man and the Animals, an extraordinary piece of scholarship that still repays study
today, he builds a painstaking case for continuity between animal signals of affec-
tive state (e.g., fear or aggression) and our own. Treating human signals as just
another display emancipated the analysis from the domain of linguistics and
opened up the possibility of comparative study for the first time. It also made
clear and testable predictions. For example, if humans signal basic emotions like
animals, the facial expressions for them should be universal, not socially acquired
like language or birdsong. This turns out to be essentially true (Ekman, 1973).
The misunderstanding is the belief that Darwin had the Cartesian view that ani-
mals only signaled about emotion. On the contrary, he makes an argument for
continuity between language and animal signals, even citing several systems in
which the ability to denote external objects has since been demonstrated, thereby
anticipating the scientific literature by more than a century (Evans, 1997).
176 ANIMAL BEHAVIOR

Naturally, any claim that potentially reduces the uniqueness of language—

long a shibboleth for the proposition that humans are special—is going to be
controversial. The suggestion that bees communicate the location of food by
encoding direction and distance in their “dance” (von Frisch, 1967) was no
exception. Perhaps the general incredulity was increased because the first
exception to Descartes’s dictum was not one of our close primate relatives
but a humble social insect with a brain no bigger than the head of a pin. Inevi-
tably, strenuous efforts were made to disprove von Frisch’s ideas (Wenner,
2002, 2006). As more data have become available from an international
research effort over more than four decades, his original claim has been vindi-
cated, extended, and strengthened (Dyer, 2002; Srinivasan et al. 2006; Griffin
et al., 2012). At first, the proponents of Darwin’s view were in the awkward
position of having only a single compelling example, and this a rather unlikely
one. The next evidence came 10 years later, when Thomas Struhsaker
described the extraordinary alarm call system of vervet monkeys (Chlorocebus
pygerythrus) (Struhsaker, 1967; details below). Robert Seyfarth and colleagues
(1980) provided the crucial demonstration that the calls alone are sufficient
to allow companions to select an appropriate response tailored to the attack
method of the original predator.
Over the last few decades, there has been a spate of additional examples
across a large taxonomic range (see Table 7.1 for examples). Yet the idea that
animals communicate about more than their affective state remains highly
controversial. There is a thriving cottage industry, in both linguistics and

Table 7.1. Species in which functionally referential signaling has been demon-
strated.
Category of functionally
Species referential signals Source

Fowl Food, terrestrial and aerial Marler et al., 1986;

(Gallus gallus) predators
Evans & Evans, 1999
House sparrow Food Elgar, 1986
(Passer domesticus)
Cliff swallow Food Brown et al., 1991
(Hirundo pyrrhonota)
Carolina chickadee Food Mahurin & Freeberg,
2009
Table 7.1. (Continued)

(Poecile carolinensis)
Raven Food Heinrich & Marzluff,
(Corvus corvax) 1991
Bugnyar et al., 2001
Japanese great tit Terrestrial and aerial Suzuki, 2012
(Parus major minor) predators
Chimpanzee Food Hauser & Wrangham,
(Pan troglodytes) 1987;
Slocombe &
Zuberbühler, 2006
Bonobo Food Clay & Zuberbühler,
(Pan paniscus) 2009;
Clay & Zuberbühler,
2011
Mandrill Food Laidre, 2006
(Mandrillus sphinx)
Tufted capuchin monkey Food, aerial predators Di Bitetti, 2003, 2005;
(Cebus apella nigritus) Wheeler, 2010
White-faced capuchin Food, aerial predators Digweed et al., 2005;
monkey
(Cebus capucinus) Gros-Louis, 2004a,
2004b, 2006
Red-bellied tamarin Food Caine et al., 1995
(Saguinus labiatus)
Black-fronted titi Terrestrial and aerial Cäsar et al. 2012
monkeys predator
(Callicebus nigrifrons)
Marmoset Food Kitzmann & Caine,
(Callithrix geoffroyi) 2009
Diana monkey Terrestrial and aerial Zuberbühler et al.,
(Cercopithecus diana) predators 1999
Gunnison’s prairie dog Terrestrial and aerial Kiriazis &
predators Slobodchikoff, 2006
(continued)
178 ANIMAL BEHAVIOR

Table 7.1. (Continued)

Category of functionally
Species referential signals Source

(Cynomys gunnisoni)
African mongoose Terrestrial and aerial Manser et al., 2002
(Surakarta suricatta) predators
Meerkat Terrestrial and aerial Manser et al., 2001
(Suricata suricatta) predators

philosophy, concerned with identifying alternative explanations for the empir-

ical facts. This is a good thing. It ensures that proponents refine their methods
and counter new theoretical arguments, which are reviewed later in this chap-
ter. Briefly, objections include the suggestion that signals encode the future
behavior of the sender, that they reflect highly specific affective responses, that
they should be classed as instructions rather than descriptions, and that they
encode only the urgency of threat. Most recently, there has been a spirited
attack on the fundamental proposition that animal signals can meaningfully
be thought of as providing information at all!

EVER SINCE DARWIN: MOUNTING EVIDENCE

Darwin (1871) marveled at the dog’s ability to respond correctly to its
master’s verbal commands and the way in which trained parrots use human
words to identify objects or events. He recognized that while animals have a
limited ability to produce vocalizations, at least when compared to human
speech, their comprehension of these sounds and their ability to connect
them to specific ideas is quite remarkable. Since that time, it has been dem-
onstrated repeatedly that animals respond to subtle differences in the charac-
teristics of sounds and that listeners respond as if the sounds contain a
wealth of information. For example, baboons (Papio hamadryas ursinus) are
able to recognize the rank, identity, and kinship of individuals based on
the sounds they produce during social interactions (Seyfarth et al., 2005).
But of greater interest is the evidence that animals communicate about
events external to themselves, such as the presence of predators or food,
and that these displays are distinct enough for listeners to respond
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 179

appropriately in the absence of any other contextual cues. Vervet monkeys

provided the first evidence that a nonhuman primate might possess this
capability to refer to environmental events. Vervets are omnivorous primates
that inhabit a large part of sub-Saharan Africa. Their semiarboreal and sem-
iterrestrial lifestyle exposes them to multiple types of predators including
leopards (Panthera pardus), martial eagles (Polemaetus bellicosus), and snakes.
Each predator exhibits a different hunting strategy, which requires the mon-
keys to perform different and incompatible escape strategies. Leopards hunt
by stealth and pose the greatest threat to the monkey on the ground,
whereas eagles hunt using speed and easily snatch monkeys from tree
branches. Snakes are slow moving and tend to stay in an area for a long
time but are easily avoided by monkeys once discovered. Researchers work-
ing in Amboseli National Park in Kenya noted that the vervets produced
distinct vocalizations in the presence of each predator type and that nearby
individuals reacted differently to each sound. The call given to a leopard
caused the troop to run towards the trees; an eagle alarm caused them to
move to the center of the tree or down towards the ground. Lastly, the
monkeys stood tall and peered into the grass upon hearing the alarm
associated with a snake. Playbacks of the audio in the absence of the preda-
tors confirmed that the sounds alone were sufficient to induce the appropri-
ate response. Signals, such as these, that appear to refer to specific events
have been termed functionally referential (Evans et al., 1993; Macedonia
& Evans, 1993; Hauser, 1996). The term functional acknowledges that
we are interpreting the behaviors in light of their outcomes rather
than claiming to know the underlying cognitive processes (Marler et al.,
1992).
Since that first example in a nonhuman primate, it has become apparent
that this type of communication is relatively common. Most species of pri-
mates produce functionally referential communication in the presence of
predators, as do many other mammals such as meerkats (Suricata suricatta)
(Manser et al., 2001) and Gunnison prairie dogs (Cynomys gunnisoni) (Kiriazis
& Slobodchikoff, 2006). Several species of birds also possess this ability,
including ravens (Corvus corax) (Bugnyar et al., 2001), chickadees (Poecile atri-
capillus) (Templeton et al., 2005), and Japanese great tits (Parus major minor)
(Suzuki, 2012).
Predators are not the only significant things that animals encounter. Food
also plays a central role in their lives, and individuals often announce the pres-
ence of food using specific vocalizations, known as food calls. Primates, such
as chimpanzees (Pan troglodytes) (Slocombe & Zuberbühler, 2005), bonobos
(Pan paniscus) (Clay & Zuberbühler, 2009), marmosets (Callithrix geoffroyi)
180 ANIMAL BEHAVIOR

(Kitzmann & Caine, 2009), tufted capuchins (Sapajus apella) (Di Bitetti,
2003), and rhesus macaques (Macaca mulatta) (Hauser & Marler, 1993) and
birds, including ravens (Bugnyar et al., 2001), house sparrows (Passer domesti-
cus) (Elgar, 1986), and cliff swallows (Petrochelidon pyrrhonota) (Brown et al.,
1991), give functionally referential calls in the presence of food.
Referential communication is not limited to vocalizations. Ravens use refer-
ential gestures to show or offer items to other ravens (Pika & Bugnyar, 2011),
and chimpanzees use pointing to solicit grooming on inaccessible parts of their
bodies (Pika & Mitani, 2006). In captivity, chimpanzees have also been
known to use gestures to elicit assistance from humans (Leavens et al., 2004;
Liebal & Pika, 2007).

WHAT FACTORS DRIVE REFERENTIAL SIGNALS AND SIGNAL STRUCTURE?

Although functionally referential signaling is unexpectedly common and
not taxonomically limited, it is by no means ubiquitous. There are many spe-
cies that produce graded variation within a class of signal rather than distinct
signals in response to environmental stimuli as well as species in which the
same signal occurs in multiple contexts. Furthermore, even within a species
that produces referential signals, not all signals given are functionally referen-
tial. For example, species that produce functionally referential alarm calls
may produce only food-associated calls. Certainly all species produce sounds
that are purely affect-based, such as a yelp of pain or exclamation of surprise.
These tend to be involuntary, and receivers cannot infer any additional infor-
mation about the environment based on them. In truth, all animal signals can
be placed on a theoretical continuum. On one end of the scale are purely
affect-based signals, which reflect the signaler’s internal state. On the far end
are signals that are referential, denoting a specific object or event. Categorizing
a signal as referential is equivalent to postulating a threshold value on such an
underlying continuum and then demonstrating empirically that the properties
of the signal are such that it is exceeded.
Several hypotheses have been proposed to explain why some species pro-
duce functionally referential calls while other produce signals that convey less
specific information. One hypothesis suggests that referential alarm signaling
evolved in response to pressure from predators with different hunting strate-
gies that require incompatible escape behaviors (Marler, 1976; Cheney &
Seyfarth, 1990). For example, ring-tailed lemurs (Lemur catta) exhibit spe-
cific escape strategies depending on the type of predator. Information about
the type of approaching predator is crucial to eliciting an adaptive escape
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 181

response, and this information is encoded in their alarm calls (Macedonia &
Evans, 1993). In contrast, urgency of response may be more important than
the specific type of predator approaching for species with only one escape
strategy, for example fleeing to the nearest cover or burrow. An excellent
example of this type of urgency-base alarm signaling is found in sciurids, par-
ticularly marmots (Blumstein, 2007). Of 14 species of marmot studied, none
produced functionally referential alarm calls. Although these ground-dwelling
species face threats from the air and the land, their escape response depends
on the immediacy of the threat rather than the hunting tactic of the predator.
Other factors, such as sociality and habitat structure, have been suggested as
possible drivers of referential alarm signals (Evans, 1997). Referential alarm
signals may have been more likely to evolve in species with stable social
groups that may include several generations of descendant kin or where the
likelihood of repeated interactions and thus reciprocity between individuals
is higher. Evidence in support of this hypothesis is thus far lacking. Habitat
structure has been presented as an argument for more specificity in alarm sig-
naling. Consider the trade-off between alarm calling to the approach of an
innocuous species (type I error) and failing to respond to a dangerous preda-
tor (type II error). Each type of error imposes a cost on the signaler and
receiver, the first in wasted time and the second in loss of life. Species that
inhabit open areas should have longer to assess potential threats and may
therefore be more specific in their responses (Walters, 1990). This, however,
may not be the determinant of whether a species produces a referential alarm
but instead may determine the specificity of the stimuli that are categorized
as a predator by that species. Interestingly, meerkats do not fit the incompat-
ible escape hypothesis because they only have one escape strategy but pro-
duce referential signals. Nor does habitat complexity have explanatory
power because other sympatric ground-dwelling species do not produce
referential alarms. Instead, the need to coordinate group movements for for-
aging efficiency may have exerted a sufficient selective pressure in this species
(Furrer & Manser, 2009).
It is important to recognize that there is no inherent incompatibility
between response-urgency-based and referential signals. There are examples
of species that produce both, such as Gunnison’s prairie dogs and meerkats,
and others, such as the red-fronted lemurs (Eulemur fulvus rufus) and white
sifakas (Propithecus verreauxi verreauxi) of Madagascar, that produce referential
signals for one class of predator and urgency-based signals for another (Fitchel
& Kappeler, 2002).
Unlike alarm calls, food-associated communication does not appear to
have a unifying function, and the selective forces that lead to functionally
182 ANIMAL BEHAVIOR

referential food calls are still poorly understood (Clay et al., 2012). Far fewer
species produce referential food calls than produce referential alarm signals.
The majority of species produce food-associated calls under multiple con-
texts, and therefore the calls do not meet the production criteria for func-
tional reference. Instead, they appear to serve multiple functions, such as
advertisement of resource ownership or of the signaler’s status, or are a direct
indication of the affective state of the caller (reflecting excitement over the
discovery of food or other scarce resources (e.g., Toque macaques [Macaca
sinica] using the same call for rain after a drought and upon finding food
[Dittus, 1988]). In further contrast to alarm signals, species tend to produce
one type of call and vary the call rate in response to characteristics of the
feeding opportunity rather than giving distinct calls for each type of food.
Further research is needed into the forces that may have selected for referen-
tial signaling about food.

Signal Structure
Communication, referential or otherwise, imposes costs on the signaler and
receiver. For such behavior to persist, the benefits must equal or outweigh
these costs. At first glance, alarm calls present a serious conundrum to the
study of animal communication because the calls potentially expose the ani-
mal to a predator, which may increase the likelihood of capture and death (a
high cost to pay). On the other hand, the benefit to the receiver is obvious,
as long as the signal reliably indicates the presence of a predator (see the sec-
tion on deception below). However, there are ways in which alarm signaling
may benefit the signaler. Calls may lessen the individual’s likelihood of capture
if they spread the risk among other prey that is present. This may be achieved
by manipulating other prey into revealing their location, by synchronizing
escape responses, or by recruiting conspecifics to mob the predator (Bradbury
& Vehrencamp, 1998). Calls may also affect predator behavior by deterring a
current attack (“perception advertisement hypothesis,” Curio, 1978) or by
lessening the likelihood of future attacks (Shultz, 2001). Calls may indirectly
benefit the signaler if the group is composed of mates, offspring, or other close
relatives, which are more likely to survive following the warning signal (May-
nard Smith, 1965).
Some species also employ risk compensation or risk management tactics,
which may help to offset the cost of alarm signaling (Owings & Hennessy,
1984). Two such tactics are selectivity in call production and variation in the
acoustic characteristics of the call. The audience effect is a common phenome-
non, wherein a behavior is only produced in the presence of an appropriate
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 183

receiver. Paul Sherman’s (1980, 1985) studies of Belding’s ground squirrels

(Urocitellus beldingi) revealed that females only signal in the presence of
ground predators if their descendent kin are nearby. By selectively signaling,
rather than signaling any time a predator is present, the caller increases the
potential for benefiting, directly or indirectly, from the behavior. A second
means of adjusting the risk associated with alarm calling is to produce signals
that are conspicuous to companions but inconspicuous to predators. The
acoustic structure of the alarm call will dictate its audibility and localizability
(Marler, 1955). For example, whistles and squeals are often produced during
encounters with predators that rely on speed of attack rather than stealth to
capture their prey. These high-pitched sounds are more difficult to localize
for some predators. They also attenuate more quickly compared to broadband
sounds (like barks or claps) and may therefore warn nearby conspecifics with-
out alerting more distant predators (Wood et al., 2000). Thus the prey species
may remain cryptic while warning conspecifics. The acoustic structures of
these calls are in contrast to signals that function as predator deterrents or to
recruit conspecifics for group defense, which should be audible and localizable
by both the predator and distant companions.
The production of calls in the presence of food presents a similar problem
regarding the benefit to the signaler of sharing or even giving up a preferred
food. Again, the benefit to the receiver is not difficult to determine, but there
are several means by which the signaler could benefit as well. Recruiting con-
specifics or heterospecifics to a large, divisible resource may reduce the signal-
er’s risk of predation while foraging (known as the dilution effect) or may
allow the signaler to more effectively exploit an ephemeral resource, such as
an insect swarm. Food calling may also strengthen the bond between mated
pairs (Stokes, 1971). Conversely, it may reduce the likelihood of aggression
from more dominant individuals (Hauser & Marler, 1993). The structure of
these calls should be easily localizable. Similarly to alarm calls, the cost of call-
ing may be managed by selectively calling in the presence of an appropriate
audience.
Acoustic signals are only one possible mode of communication. Both
food-associated and warning signals can be composed of two different com-
ponents that are produced simultaneously (e.g., sounds and movements).
This type of communications is known as multimodal signaling because
the signals are transmitted and can be detected using at least two different
sensory systems (Partan & Marler, 1999, 2005; Hebets & Papaj, 2005).
Researchers frequently use the receiver’s response to determine the signal’s
content. Redundant (or “backup”; Johnstone, 1996) signals generate the
184 ANIMAL BEHAVIOR

same response to each channel, which suggests each modality transmits the
same information. These signals can be further divided into two subcatego-
ries (Partan & Marler, 1999, 2005). Equivalent signals evoke the same
responses singly or combined, whereas signals classified as enhancement gen-
erate a higher level of response when combined. The second major category
of signal type is nonredundant (or multiple messages; Møller & Pomiankow-
ski, 1993; Johnstone, 1996). These signals can convey different information
in each modality, and the combined signal can alter the information content.
An example of nonredundant signaling that many people are probably famil-
iar with, but do not realize it, is the “play-bow.” Dogs often simultaneously
bark and lower their heads towards the ground in a bowing posture. The ears
are often forward. Barking with the ears forward is an aggressive display, but
when combined with the bowing posture the message changes from aggres-
sion to a solicitation to play.
Redundant, equivalent signals are the most common in animal communi-
cation. These signals have several potential benefits. Producing two signals
may enhance the detectability of the display as well as increase the ease with
which the receiver can determine the message (known as discriminability).
This may be critical in a noisy environment where the signaler may be com-
peting with individuals of its own and other species as well as masking noise
produced by wind, water, or even humans. Previous research also suggests that
multiple signals increase the memorability of the signal and signaler (Rowe,
1999).
In some systems, the components of a multimodal signal can be flexibly
combined (also called “fluid” or “free” signals; Wickler, 1978). Most commu-
nication occurs within a matrix of receivers, such as competitors, predators, or
parasites, which may impose an additional cost on the signaler. Redundant sig-
nals allow the signaler to produce just one channel without altering the mes-
sage. This flexibility may reduce the cost of signaling if it limits the
attraction of these unwanted individuals (Johnstone, 1996; Peake, 2006;
McGregor, 2009).
These general principles for the structure of signals appear to be valid for
both referential and nonspecific communication. Indeed, it is often, but not
always (see Figure 7.1), the case that signals that encode information about
very different types of events have contrasting structures. However, the analy-
sis of the meaning of signals has provided little in the way of insights into the
specific form of referential signals. Comparative studies are needed to deter-
mine if there is any correlation between signal structure and meaning across
taxa.
Figure 7.1. Acoustic structures of calls with different functions. Calls that encode infor-
mation about different types of events often have contrasting structures, such as the
calls produced by male fowl (Gallus gallus) in response to aerial compared to ground
predators. However, calls produced in the context of food have very similar structure
to those elicited by ground predators. It is apparent that signal structure alone cannot
be used to predict signal meaning. Instead, selective pressures arising from the sig-
naling context, such as the need to remain cryptic around aerial predators, may have
shaped the signal’s characteristics.
186 ANIMAL BEHAVIOR

IDENTIFYING FUNCTIONALLY REFERENTIAL SIGNALS

There are two criteria that signals must meet to be considered functionally
referential. The first relates to the specificity of the signal production. The sig-
nal must be structurally distinct from other sounds or displays, and there must
be a close association between the context of the eliciting event and the specific
signal production such that it would allow listeners to extract meaning from
the signals regarding the external event (Seyfarth & Cheney, 2010). The sec-
ond requires the receiver to respond with a specific and adaptive behavior
without the need for other contextual information (for example, without the
sight, sound, or smell of a particular predator [Evans, 1997]).
Ever-advancing technology increases our ability to study referential com-
munication. Major advances in the study of sounds began in the 1950s with
the increasing availability of the tape recorder to record and play back sounds
and the subsequent development of the sound spectrogram, which allowed
researchers to visualize and analyze the recorded sound (Baker, 2001). This
new technology opened up the world of animal acoustics to researchers who
could now quantify the structural differences between two sounds. This
enabled the statistical analysis of the characteristics of different sounds, which
allowed the researchers to make predictions about which sounds would be dis-
tinct enough for a receiver to distinguish between them. These data, combined
with the observations of the eliciting event, allowed researchers to determine if
a sound met the production criterion for functional reference. The next step
required the researchers to test the animal’s responses in the absence of the
putative trigger. Playback experiments to free-living animals, such as those
conducted with the vervet monkeys, allow the researchers to examine the
function of the calls within the natural environment. One potential difficulty
with field experiments is that they must be conducted opportunistically, wait-
ing until the eliciting event is absent from the area. An alternative approach is
to bring the animal into a controlled environment, such as the laboratory.
This facilitates the manipulation of the social as well as environmental con-
texts of the call. A combination of both approaches (observation in the field
and controlled tests in the laboratory) may provide the most complete picture.
As mentioned previously, acoustic signals are not the only form of referen-
tial communication, and the study of visual signals has also benefited from
recent advances. Research in the 1990s revealed that video images would elicit
natural responses from birds and spiders (Clark & Uetz, 1990; Evans &
Marler, 1991). At the time, VHS camcorders and the standard-definition tele-
visions provided only 240 lines of resolution, but even with this low resolution
video was sufficient to replace a live companion (Evans & Marler, 1991).
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 187

Experiments revealed that birds could recognize the feeding movements of a

companion on the video and discriminate these from other types of move-
ments (McQuoid & Galef, 1993). Furthermore, preferences acquired from
video sequences, such as for a particular color of food bowl, transferred to their
real equivalents (McQuoid & Galef, 1993). The current high-definition cam-
era and television technology provide approximately five times more detail
than those pioneering video playback studies (Smith & Evans, 2008, 2009).
The greater resolution provides enough detail that birds will even attempt to
take food items directly from the screen (Smith & Evans, 2008). Advances in
computer technology have created the ability to manipulate visual signals using
3D animation (birds: Watanabe & Troje, 2006; Smith et al., 2009; lizards:
Van Dyk & Evans, 2008). Using these techniques, individual aspects of the
signaling movement can be changed and tested to determine their function.

OPENING THE WINDOW ON ANIMAL MINDS

During any interaction, there are aspects that we can directly observe and
others that are, by their nature, hidden from us. We can observe the eliciting
event, and we can quantify the individual’s physical and signaling responses.
In this same way, we can measure the responses of the receivers. What we can-
not directly assess is the underlying cognitive process in the signaler prior to call-
ing and the receiver during the response. For example, there is strong evidence
that alarm calls and food calls are not reflexive. Instead, the likelihood of calling
appears to be dependent upon the specific circumstances surrounding each
event. In many species, individuals will only signal if an appropriate audience
is present. Furthermore, stimuli with similar characteristics appear to generate
comparable signaling responses. This suggests that there is an intervening step
between the external event and the signal production. Based on observations
of natural groups and experimental manipulation of the eliciting factors, we
can infer that the stimuli are categorized by a set of characteristics. Using a sim-
ilar testing method, we may determine the breadth of each putative category. A
similar type of categorization may occur during the receiver’s perception of the
signal. The receiver may first determine which class the signal corresponds to
and then select an adaptive response. The study of referential signals may hence
provide insight into the underlying cognitive processes of communication.

Evidence of Cognitive Processing and Information

Functionally referential signals reveal that individuals can use the associa-
tion between different signals and events to produce an adaptive response.
However, the highly predictive nature of the signal may be sufficient to
188 ANIMAL BEHAVIOR

explain the response without invoking any form of more complex cognition.
That is, the animal’s response could be based on perceptual processing, using
the physical form of the signal rather than the associated meaning. On the
other hand, referential signals may form a mental representation of that event
in the mind of the receiver. If the response to the calls relies upon a mental
process, then the signal must stimulate retrieval of stored information that
then determines receiver response. This property corresponds to Charles Gal-
listel’s (1990) “nominal representation,” which is the lowest level of cognitive
complexity, wherein the signals “stand for” something in the environment.
“Prime-probe” experiments are one tool that researchers use as a means of
investigating the potential underlying cognitive processes of animal signals. In
these experiments, the animal experiences two stimuli in sequence that are either
congruent or incongruent. For example, the eagle alarm call of a Diana monkey
(Cercopithecus diana) is played to a group of monkeys. A short time later, another
sound is played, either the shriek of an eagle (congruent with the previous sound)
or a leopard alarm call (representing a different predator class). When the two
pieces of information are in agreement, the Diana monkeys do not give new
alarm calls. When the second stimulus provides different information, the mon-
keys call. This suggests that the sound of the eagle alarm call generates a mental
representation of a specific predator type in the mind of the listener and that this
expectation affects the reaction to the subsequent event (Zuberbühler et al.,
1999). This type of response appears to require the storage and retrieval of past
experiences or knowledge, which are then compared to the current conditions,
which is why these signals are known as representational.

Deception
The term deception is extremely contentious in the context of animal com-
munication. It is often used in two contexts. The first is in a functional sense,
in which the initiator of the “deceptive” act benefits by behaving in a manner
that causes another to alter its behavior in a costly way. It is agnostic about the
underlying mental state of both the signaler and the responder and focuses sim-
ply on the outcome of the behavior. The second usage has cognitive implica-
tions. Under this definition, the “deceptive” act can be evaluated using four
criteria: two relating to the signaler and two to the responder. The first is the
mental state of the signaler. In the most rigorous definition of cognitive decep-
tion, the signaler would have the ability to attribute mental states to itself and
to others. This ability is known as “theory of mind.” The second criterion deals
with the intentionality of the signaler’s action to change the behavior of the
responder (i.e., the initiator’s behavior is not simply a reflexive response to a
heightened emotional state). The remaining two criteria focus on the responder.
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 189

The first criterion asks about the responder’s mental state. Does the signaler’s
action create a mental representation of an external event in the responder’s
mind? The final criterion is whether or not the responder’s actions are based
on that mental representation.
We can examine each of these criteria using a fascinating example of deceptive
behavior by fork-tailed drongos (Dicrurus adsimilis). These birds live in the Kala-
hari Desert and share the same habitat, and eat many of the same things, as meer-
kats. Meerkats eavesdrop on the alarm calls of the drongos and respond to them
as if a predator is approaching. If an alarm call is given while a meerkat is feeding,
it will often drop the food and run for cover. However, drongos occasionally
alarm call in the absence of a predator. This generates the same response from
the meerkat as if the predator were present. Once the meerkat has dropped the
food, the drongo flies down and takes it (Flower, 2010). In the functional sense,
the drongo has deceived the meerkat by producing a false alarm call (i.e., in the
absence of a predator), thereby causing the meerkat to give up its meal. What
about in the cognitive sense? It would be tempting to imagine that the drongo
intends to deceive the meerkat by creating a false belief in the meerkat that a
predator is approaching. However, juvenile drongos often watch adults steal
food from other species; therefore the behavior may actually reflect trial-and-
error learning rather than a cognitive task requiring the drongo to have a theory
of mind. Because a conclusive experimental test for theory of mind remains elu-
sive, let us discount the first criterion and examine the later three. It appears that
the drongo’s false alarm calls are not reflexive (although they have not been
shown to be specific enough to be referential) because they are more likely to
be given when the target animal has a small food item that can be taken and
quickly consumed (Flower & Gribble, 2012). This also suggests that the calls
are not simply a reflection of the caller’s excitement over the potential for food.
If this were the case, larger food sources would be predicted to have a great excita-
tory effect and hence to elicit more false alarm calls. This further suggests that the
call production is under intentional control of the drongo, which meets the sec-
ond signaler criteria. From the responder perspective, meerkat alarm calls are
functionally referential (Manser et al., 2001), which suggests that the response
to the drongo call is not reflexive. However, it has not been experimentally dem-
onstrated that the alarms of either species are representational and hence meet
the second signaler and responder criteria. This example demonstrates the diffi-
culties and pitfalls of attempting to classify a behavior as cognitively deceptive.
But we cannot ignore deceptive behavior. There are a tremendous number of
examples of deceptive behavior (in the functional sense) in nature (e.g., primates:
Byrne & Whiten, 1988; fishes: Bshary, 2002; stomatopods: Adams & Caldwell,
1990), and deception may have shaped animal communication.
190 ANIMAL BEHAVIOR

Recently, Michael Lachmann and Carl Bergstrom (2004) proposed that

deception may be the underlying reason why so few animals use combinations
of signals in their communication. The hallmark of complex communication
in humans is ability to use combinations of different sounds to convey different
messages. Unlike human communication, even combining of a few sounds to
generate specific signals is uncommon in animals. One of the rare examples is
the putty-nosed monkey (Cercopithecus nictitans). These West African primates
combine different sounds to alert group mates to the specific types of predator
that are present. Alone the sounds are not specific enough to identify the predator
class, but in combination the calls are (Arnold & Zuberbühler, 2008). In this sys-
tem only two sounds are used, which means the receiver has the opportunity to
experience all possible combinations multiple times and to verify that these com-
binations correspond to the conditions the signaler is indicating. Lachmann and
Bergstrom (2004) point out that as more sounds are added and more combina-
tions are created that indicate more novel conditions, the receiver will be more
likely to encounter signals that are outside of its firsthand experience. In this case,
the receiver must infer the combined signal meaning based on the single compo-
nent’s meanings. This creates the opportunity for signalers to benefit by creating
novel false signals that the receiver will respond to as true.

FUNCTIONALLY REFERENTIAL SIGNALING IN FOWL

Perhaps surprisingly, some of the best-described examples of referential com-
munication come from chickens (Gallus gallus). The ancestral form of domestic
chickens, red jungle fowl, evolved in the dense forests of Indonesia, Ceylon, and
China. In the wild, social groups consist of a dominant pair and several subordi-
nates of both sexes. Groups are typically stable, but juveniles and subordinate
males may move between groups over the course of a season (McBride et al.,
1969). Both males and females have a dominance hierarchy, and intrasexual
aggressive displays are common. The birds have a completely lateralized brain,
meaning that visual stimuli from each eye are processed in different hemispheres
of the brain. The left eye (right hemisphere) appears to be superior for spatial
location tasks, while the right eye (left hemisphere) is better at the categorization
of objects, such as food (Rogers, 1980, 1990). They are hunted by both terres-
trial and aerial predators and have several potential escape behaviors. Chickens
have an extensive repertoire of visual and vocal signals (Collias, 1987). As one
might predict based on their habitat, sociality, and antipredator behavior, chick-
ens produce four functionally referential signals: two alarm signals (Evans,
Evans, & Marler, 1993; Evans, Macedonia, & Marler, 1993) and two food sig-
nals (Evans & Evans, 1999; Smith & Evans, 2008, 2009).
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 191

Referential Alarm Signals: Ground Predators

Terrestrial predators, such as dogs and foxes, are ambush predators that
hunt by stealth, stalking their prey. If detected at a distance, they are more
likely to terminate the hunt. The visual and vocal behavior of fowl when
encountering a terrestrial predator appears adaptive in light of these predator
traits. Both males and females produce distinct calls composed of a series of
loud, short pulses (Figure 7.2). The rate of calls often increases over the course
of the encounter with the predator and may continue even after the predator
has left. Adult group members respond by producing their own ground alarm
calls while standing erect and scanning in the horizontal plane. There is no
effect of audience on ground alarm vocalizations, even solitary individuals call,
suggesting that the calls may also be directed at the predator, as well as poten-
tially coordinating the group’s escape behavior. These calls are structurally dis-
tinct from sounds produced in other antipredator contexts. Playback
experiments of the vocalizations to isolated males and females reveal that the
sound alone is sufficient to elicit appropriate escape behavior (Evans, Evans,
& Marler, 1993).

Referential Alarm Signals: Aerial Predators

In contrast to the response to ground predators, the response to aerial pred-
ators is much more nuanced. Unlike terrestrial predators, which rely on stealth
to capture their prey, aerial predators rely on speed of attack from a distance,
which means that vocalizing in their presence may increase the likelihood of

Figure 7.2. Ground alarm calls of male fowl and an example of the female’s visual
response. Ground alarm calls are high-amplitude, short pulses given in long bouts
by both males and females. Production of these calls is not subject to the audience
effect. The calls are easily localizable, suggesting that they may be directed at the
predator as well as conspecifics. Both males and females respond to the calls by
standing erect and scanning in the horizontal plane.
192 ANIMAL BEHAVIOR

Figure 7.3. Aerial alarm calls of male fowl and example of females’ visual response. Aer-
ial alarm calls are low-amplitude vocalizations consisting of an initial pulse followed by
a relatively long second component. Male fowl flexibly change the composition of
their aerial alarm signals, omitting the more conspicuous broadband component, in
subsequent calls. The narrow bandwidth and short duration of the calls reduces their
localizability, allowing the caller to remain cryptic. This class of call is subject to the
audience effect. Both males and females react to the vocalizations by crouching down,
sleeking their feathers, and looking upward.

being detected and caught. The production of aerial alarm calls by both males
and females is hence governed by an audience effect: any conspecific potenti-
ates calling in males, but females will only call if they have chicks. In fact,
males are significantly more likely to call if they have recently mated (Wilson
& Evans, 2008). In stark contrast to the structure of ground alarm calls, aerial
alarm calls are more tonal and significantly quieter. These call characteristics
make the caller difficult for the predator to localize but still function to alert
his mate (Wood et al., 2000; Bayly & Evans, 2003). Males also vary the com-
position and duration of their calls. During an encounter, males often produce
a series of aerial alarm calls. The first call often begins with a brief (<200 ms)
broadband pulse followed by a longer tonal component. The broadband pulse
is omitted from subsequent calls, suggesting that this first sharp sound serves
an alerting function to attract the attention of the hen. Once the call is
detected, the broadband pulse would serve no further function (Figure 7.3)
(Bayly & Evans, 2003). In addition, the likelihood, composition, and dura-
tion of the calls are dependent upon the male’s personal risk. Males give
longer-duration alarms, which increase the likelihood of signal detection by
both the female and the predator, when they are closer to a refuge. They are
also more likely to call in response to their rival’s call when they are protected
under the cover of a tree or shrub. Lastly, dominant males produce longer calls
when their subordinate is nearby (Kokolakis et al., 2010). By selectively calling
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 193

when the subordinate is close by, the dominant may further reduce his risk of
capture by giving the predator more than one target (the dilution effect for
predators choosing between multiple targets [Bertram, 1978] and the confu-
sion effect when prey are fleeing [Schradin, 2000]).
These functionally referential aerial alarm calls are also urgency-based signals.
Fowl communicate the size, speed, and proximity of flying birds using graded
variation in their calls. Flying birds that are closer, larger, and faster generate
alarm calls that are shorter, louder, and lower in frequency (Wilson & Evans,
2012). It is also interesting to note that unlike some species, males call in
response to a broad range of stimuli moving overhead, including both predators
and innocuous objects. The birds appear to use a general rule of thumb, calling
to any overhead objects that exceed a specific threshold for size and apparent
speed (Evans et al., 1993). This may be an adaptive response to the dense forest
habitat in which fowl evolved. The birds would have only a short period of time
to assess the threat, and incorrectly categorizing a true threat may be more costly
than the time spent responding to nonthreatening stimuli.
The females’ response to an aerial predator alarm call is also very different from
their reaction to a ground alarm call. Females crouch down, sleek their feathers,
and fixate upwards using their left eye. This behavior reduces the hen’s conspic-
uousness and brings to bear the brain hemisphere best suited for locating small,
rapidly moving targets, such as aerial predators. Experimental studies using audio
and video (including live footage and animations) reveal that both the produc-
tion and perception of fowl alarm calls meet the criteria for functional reference.

Referential Food Signals: Visual and Vocal

Fowl also produce one of the few multimodal signals wherein both compo-
nents are functionally referential (Smith & Evans, 2008, 2009). Discovery of a
palatable item by a male in the presence of a hen reliably elicits “tidbitting.”
This signal is composed of a series of pulsatile vocalizations, which are audible
up to 30 m from the male, and rhythmic motions of the head and neck (Davis
& Domm, 1943). The pulsatile nature of the food calls should make them
easy to localize in the same way as other sounds in the fowl vocal repertoire
with similar spectral and temporal structure (Wood et al., 2000). The vocal-
izations are individually distinctive, which allows females out of visual contact
with the male to determine that a specific male has located food (Stokes &
Williams, 1971, 1972; Evans & Evans, 1999). The visual component of the
tidbitting display is made up of three distinct motor patterns, including a
series of side-to-side lateral movements of the head—“short bobs”—in which
the head is brought approximately halfway to the ground and back to the
upright position, and “long bobs”— long sweeps of the head in the vertical
194 ANIMAL BEHAVIOR

Figure 7.4. Tidbitting display and female’s response. Upon discovery of a food item in
the presence of a female, males perform visual and vocal signals that specifically indi-
cate the availability of food. The visual display is composed of a series of movements,
including side-to-side twitches of the head, short bobs of the head and neck down
towards the ground and back to the upright position, and long bobs that bring the
head down towards the ground before rapidly sweeping up again. Males often pick up
food during long bobs and hold it in their mandibles for a period of time before either
dropping it or ingesting it. The visual display is usually accompanied by pulsatile
vocalizations. Males selectively perform the tidbitting display when a female is
present. Both components of the display are functionally referential. Females often
approach the male and begin food searching near him by binocularly fixating on the
substrate (as shown in the right-most image). Females food search with equal intensity
to both the visual display alone and vocalizations alone.

plane. The male often uses these long sweeps to pick up and manipulate the
food item before dropping it again (Figure 7.4).
Under natural conditions, this performance often entices one or more hens
to approach the tidbitting male and food search near him, sometimes taking
the food item directly from his mandibles (Gyger & Marler, 1988; Marler
et al., 1986). Playback experiments confirmed that females discriminate
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 195

between the vocalizations and structurally similar calls, preferentially food

searching only in response to the putative food calls (Evans & Evans, 1999).
Video playback manipulating the visual display revealed similar specificity of
response to the motor patterns (Smith & Evans, 2008, 2009).
Because both components of the tidbitting display are redundant, a receiver
detecting either component would recognize that the signaler is indicating the
presence of food. The subordinate male exhibits behavioral flexibility that
takes advantage of this fact. Dominant males respond aggressively to subordi-
nates that tidbit, often chasing them away from the food and then presenting
the food to the female (Stokes & Williams, 1972). To avoid this aggression,
subordinate males change the way they signal based on the attentional state
of the alpha male. Subordinates often omit the more conspicuous vocaliza-
tions and perform only the visual display when the dominant male is attentive.
Conversely, when the dominant male is distracted, the subordinate performs
the combined display, which may benefit the subordinate in the long term
by alerting eavesdropping females to his food-providing behavior and may
improve his mating success. This flexibility allows the subordinate to balance
the social costs and benefits of tidbitting. Unlike many other species, the sub-
ordinate male’s behavior is adjusted in response to his rival rather than the
female (Smith et al., 2011).
Food calls are also remarkable because they appear to be one of the few ani-
mal signals that has been demonstrated to be not only referential but also rep-
resentational (Evans & Evans, 2007). The female’s response to food calls is
specifically mediated by selective retrieval of information about the discovery
of food. Females perform the characteristic food-searching behavior when
food calls are heard in a setting in which food had not recently been found.
In contrast, the same signal heard minutes after ingesting a small quantity of
food elicits no food-searching behavior. This difference in behavior to pre-
cisely the same signal suggests that when food calls provide no new informa-
tion about the hens’ immediate environment, they are ignored.
This suggests that food calls are nominally representational (Gallistel,
1990). Representational signals are most likely more taxonomically widespread
than the current evidence suggests. Comparative studies are needed to identify
common factors responsible for the evolution of such signaling systems.
The referential nature of tidbitting creates the opportunity for another
unexpected behavior: deception, at least in the functional sense. Males occa-
sionally signal while holding an inedible item. This behavior also occurs more
often when the hen is farther away from the male. Females often approach a
male that is calling without an edible item but are less likely to approach than
if the male has food (Gyger & Marler, 1988). Experimental data show that the
Figure 7.5. Model of fowl communication. Using the observable events during signal-
ing interactions facilitates inferences about the underlying cognitive processes. There
are three observable layers (environmental events, signals, and responses) and two that
may be inferred from the sender’s and receiver’s responses (visual category and signal
category). Based on the observed environmental events, we can experimentally deter-
mine the characteristics of each type of event that will elicit a specific behavior. We
can infer from this that the signaler is making a similar categorization. The receiver
produces an adaptive response to each signal type, and thus we can infer that the
receiver categorizes the signals by the corresponding event. For examples, the discov-
ery of corn (environmental event) in the presence of a hen (audience effect) elicits
pulsatile calls and stereotypic movements (signals) by the male. If the female has not
recently found food in that location (representation of event), she may approach, fix-
ate on the substrate, and peck the ground (receiver response). From these observable
actions, we can make several inferences about the cognitive processes of sender and
receiver. We can infer from the male’s behavior that the items found are part of a
larger category (food). Experimental results reveal that the female’s response is not
reflexive but is specifically dependent upon her prior experience with food, from
which we can infer that the calls evoke specific retrieval of prior information. Because
females respond with an appropriate response to food calls, we can infer that they
categorize these signals in a similar manner to the males and that calls are nominally
representational (i.e., the sounds “stand for” food). (Redrawn from Evans, (1997),
Figure 1, with kind permission from Springer Science+Business Media B.V.)
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 197

females stop responding if the male calls too often in the absence of food
(Evans & Evans, unpublished data). Despite meeting three of the four criteria
for cognitive deception, we are still unable to conclusively categorize this as a
positive example. Because the study was entirely observational, there are too
many unanswerable questions regarding the male’s behavior (e.g., a human
observer classified the object as inedible). Further research is still needed to
understand this apparently deceptive behavior.
Chickens provide a robust and tractable model in which we can observe
and experimentally test the function of specific signals. The distinct and easily
quantifiable nature of the receiver response allow for inferences into the under-
lying cognitive processes of animal communication (see Figure 7.5).

DESCARTES’S GHOST, OR RADICAL BEHAVIORISM REVISITED

Recently, Drew Rendall and colleagues (2009) have raised the most funda-
mental challenge of all. They argue not just that claims for external reference
are unnecessary but that no animal signal contains information of any kind.
They dismiss the very concept of information as naïve and illogical, claiming
that it is incompatible with evolutionary theory. If true, these arguments
would be absolutely fatal not just to the proposition that animals are capable
of signaling about events in the world but also to dozens of other research pro-
grams. This is because the argument is a broad-brush attack on the utility of
information as an explanatory and hypothesis-generating concept rather than
on any particular use of it.
Rendall and colleagues (2009) claim that information locks the field into a
language-based model that distracts researchers from a more appropriate focus
on the issue of the physical form of signals—the classic problem of design.
They further suggest that the effects of signals can all be explained by their
structure interacting with properties of the receiver’s nervous system. For
example, many begging calls are harsh and unpleasant, at least to us. Rendall
and colleagues (2009) assert that the responses of parental animals are invol-
untary reactions to the repeated stimulation of their auditory systems by such
sounds. With a nod to Skinner’s (1974) “behaviorism,” we refer to this per-
spective as “radical structuralism.”
No one imagines that animal signals have arbitrary structures. Indeed, the
demonstration that they do not was one of the foundation papers of bioacous-
tics (Marler, 1955). There is good reason to think that there might be “rules”
linking form to function (Morton, 1977), and Patricia McConnell and Jeffrey
Baylis (1985) have shown that the same signal structures are used by humans
to increase and decrease activity in dogs across language groups and in
198 ANIMAL BEHAVIOR

geographically separated communities. The animals are thus shaping us to

produce sounds that best address their auditory systems, and they most likely
have their effects just as Rendall and colleagues (2009) claim. There is no con-
flict between these and other similar findings and the idea that animal signals
have their effects in part by providing information. For example, chicken food
calls and ground alarms have similar structure, and both cause recipients to
become more active, but in very different ways. Ground alarms trigger scanning
in the horizontal plane in a head-up posture (Figure 7.2) (Evans, Macedonia, &
Marler, 1993). In contrast, food calls evoke substrate searching, for which the
birds bring their heads close to the ground in order to engage the myopic frontal
field on which depth perception relies (Evans & Evans, 1999, 2007; Figure 7.4).
Structuralism provides a parsimonious explanation for the shared effect of the
two signals on general activity, but it fails to account for the differences in detail,
which produce an exquisite match with the original eliciting events—an
approaching terrestrial predator requiring increased vigilance in one case, and
discovery of a small food item in the other. Information is critical to explaining
these different responses. Note that the two theoretical positions can be comple-
mentary, each tackling a different aspect of the recipient’s response.
Radical structuralism rules out all cognitive processes, or at least treats them
as inaccessible to science. In this regard, Rendall and colleagues (2009) part
company with modern learning theory (Shettleworth, 2010), which considers
animals to have evolved by natural selection to identify predictive relation-
ships. From this perspective, there is no real difference between the mecha-
nisms by which conditioned stimuli and signals have their effects because
both are mediated by reliable relationships between the signal or synthetic
cue and an event that matters (Evans, 1997). In the jargon of animal learning
theory, these are contingencies, regardless of whether they acquire their
salience through experience or selection. Contingencies are information in
the formal sense that animals that attend to them have reduced uncertainty
about their physical or social environment (Seyfarth et al., 2010).
There is a clear consensus among researchers working on animal communi-
cation that signal design is an important question. Over the last 20 years, great
advances have been made by combining sophisticated measurement of struc-
ture with perceptual experiments conducted in a comparative framework
(Ryan & Rand, 1998; Basolo, 1998; Endler & Basolo, 1998; Wilczynski
et al., 2001). This research program has transformed our understanding of
signal evolution by revealing the selective role of sensory biases, particularly
in the context of female mate choice. The controversial issue is whether signal
structure is sufficient to answer the full range of questions being asked, from
moment-to-moment changes in behavior to ecological time scales.
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 199

This is not a new argument. Over 30 years ago, Richard Dawkins and John
Krebs (1978) suggested that the concept of information was inherently mutu-
alistic and hence incompatible with selfish-gene models of evolution. Robert
Hinde (1981) rebutted this charge, pointing out that Dawkins and Krebs
had erected a straw man and that there was no such theoretical conflict. The
attack on information was taken up next by Daniel Owings and Eugene Mor-
ton (1998), who renewed the call to expunge the concept, arguing that an
“assessment/management” perspective would better serve the field. Like
Rendall and colleagues (2009), Owings and Morton (1998) rejected informa-
tion as an explanation for the effects of signals, although they focused more on
function than mechanism.

REJOINDER: WHY FEW THINGS IN ANIMAL COMMUNICATION

MAKE SENSE WITHOUT INFORMATION
Seyfarth and colleagues (2010) provide an able defense of the ethological
approach. They point out that structure alone is unable to explain several
well-established phenomena. Similar sounds can evoke very different
responses, and sounds with qualitatively distinct structures can evoke the same
response. Neither of these findings, which have been demonstrated in multi-
ple taxa, is predicted by the structuralist model. In addition, Rendall and col-
leagues (2009) are unable to account for the enormous diversity of variation in
the physical form of signals that should all have their effects by being aversive.
Nor are they able to explain the design of nonacoustic signals because their
model is specific to the sounds (Seyfarth et al., 2010).
Seyfarth and colleagues (2010) also identify several theoretical or logical
flaws in Rendall and colleagues’ (2009) argument. We believe that the most
serious of these is the lopsided fixation on the signaler. All researchers in the
field recognize that signals are an efficient way to alter the behavior of recipi-
ents. This much is common ground between those who give information a
central place and those who think the concept a liability. Where the two sides
part company is their view of signal receivers. The radical structuralist position
is that responses are involuntary, while we take a more symmetrical view of
signaling interactions. The ubiquitous phenomenon of habituation reveals
that animals quickly stop responding to signals or other events that fail to pre-
dict anything. Similarly, selection should swiftly cause animals to ignore sig-
nals when it does not increase their fitness to respond to them. Because all
signaling has costs, selection should also expunge ineffective signals from the
repertoire (Searcy & Nowicki, 2005). For both of the above reasons, the struc-
turalist model predicts an outcome that is evolutionary unstable. Moreover, it,
200 ANIMAL BEHAVIOR

not the information model, is inconsistent with current theory on signal evo-
lution (Grafen, 1990; Searcy & Nowicki, 2005; Seyfarth et al., 2010).
To these arguments, we would add one additional consideration: the
experiments involving manipulation of putative referents (reviewed earlier)
change only what animals know about the world. The playback sounds are
identical in the various treatments, so there is no structural variation to appeal
to. Essentially, these studies are purely manipulations of “information.” It fol-
lows that the effects obtained are attributable to information. There is a grow-
ing list of such studies in primates and birds (see Table 7.1). It is the nature of
absolute arguments that a single compelling counterexample is sufficient to
disprove them (Popper, 1959). The first of these experiments (Zuberbühler
et al., 1999) thus revealed that the concept of information is an indispensible
component of any general theory of animal communication, while the steady
accumulation of additional findings over the last decade should have made this
indisputable.

ACKNOWLEDGMENTS
We wish to thank L. Evans for her enduring support throughout. Chris
Evans was fortunate to be taught by Pat Bateson and Robert Hinde, who intro-
duced him to the study of animal behavior using an integrative approach encap-
sulated by Tinbergen’s four questions. He later worked with Peter Marler, who
introduced him both to the topic of referential signaling and to fowl as a model
system. He studied these extraordinary creatures for over 20 years.

REFERENCES AND SUGGESTED READING

Adams, E. S. & R. L. Caldwell. (1990). Deceptive communication in asymmetric
fights of the stomatopod crustacean Gonodactylus bredini. Animal Behaviour,
39, 706–716.
Arnold, K. & K. Zuberbühler. (2008). Meaningful call combinations in a non-human
primate. Current Biology, 18, R202–R203.
Baker, M. (2001). Bird song research: The past 100 years. Bird Behavior, 14, 3–50.
Basolo, A. L. (1998). Evolutionary change in a receiver bias: A comparison of female
preference functions. Proceedings of the Royal Society of London, B, 265, 2223–
2228.
Bayly, K. L. & C. S. Evans. (2003). Dynamic changes in alarm call structure: A strat-
egy for reducing conspicuousness to avian predators? Behaviour, 140, 352–369.
Bertram, B. C. R. (1978). Living in groups: Predators and prey. In N. B. Davies (ed.),
Behavioural Ecology: An Evolutionary Approach (pp. 64–96). Oxford, UK:
Blackwell Scientific.
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 201

Blumstein, D. T. (2007). The evolution, function, and meaning of marmot alarm

communication. Advances in the Study of Behavior, 37, 371–401.
Blumstein, D. T. & K. B. Armitage. (1997). Alarm calling in yellow-bellied marmots.
I. The meaning of situationally variable alarm calls. Animal Behaviour, 53,
143–171.
Bradbury, J. W. & S. L. Vehrencamp. (1998). Principles of Animal Communication.
Sunderland, MA: Sinauer Associates.
Brown, C. R., M. B. Brown, & M. L. Shaffer. (1991). Food-sharing signals among
socially foraging cliff swallows. Animal Behaviour, 42, 551–564.
Bshary, R. (2002). Biting cleaner fish use altruism to deceive image scoring reef fish.
Proceedings of the Royal Society of London, B, 269, 2087–2093.
Bugnyar, T., M. Kijne, & K. Kotrschal. (2001). Food calling in ravens: Are yells refer-
ential signals? Animal Behaviour, 61, 949–958.
Byrne, R. W. & A. Whiten. (1988). Machiavellian Intelligence: Social Expertise and the
Evolution of Intellect in Monkeys, Apes and Humans. Oxford, UK: Clarendon.
Caine, N. G., R. L. Addington, & T. L. Windfelder. (1995). Factors affecting the
rates of food calls given by red-bellied tamarins. Animal Behaviour, 50, 53–60.
Cäsar, C., R. W. Byrne, W. Hoppitt, R. J. Young, & K. Zuberbühler. (2012). Evi-
dence for semantic communication in titi monkey alarm calls. Animal Behav-
iour, 84, 405–411.
Cheney, D. L. & R. M. Seyfarth. (1990). The representation of social relations by
monkeys. Special Issue: Animal cognition. Cognition, 37, 167–196.
Clark, D. L. & G. W. Uetz. (1990). Video image recognition by the jumping spider,
Maevia inclemens (Araneae: Salticidae). Animal Behaviour, 40, 884–890.
Clay, Z., C. L. Smith, & D. Blumstein. (2012). Food-associated vocalizations in
mammals and birds: What do these calls really mean? Animal Behaviour, 83,
323–330.
Clay, Z. & K. Zuberbühler. (2009). Food-associated calling sequences in bonobos.
Animal Behaviour, 77, 1387–1396.
Clay, Z. & K. Zuberbühler. (2011). Bonobos extract meaning from call sequences.
PLoS ONE, 6, e18786.
Collias, N. E. (1987). The vocal repertoire of the red junglefowl: A spectrographic
classification and the code of communication. Condor, 89, 510–524.
Curio, E. (1978). The adaptive significance of avian mobbing. I. Telenomic hypoth-
esis and predictions. Zeitschrift für Tierpsychologie, 48, 175–183.
Darwin, C. (1871). The Descent of Man, and Selection in Relation to Sex. London:
John Murray.
Darwin, C. (1872). The Expression of the Emotions in Man and Animals. London: John
Murray.
Davis, D. E. & L. V. Domm. (1943). The influence of hormones on the sexual
behavior of domestic fowl. In T. Cowles & H. M. Evans (eds.), Essays in Biol-
ogy: In Honour of Herbert M. Evans (pp. 171–181). Berkeley: University of
California Press.
202 ANIMAL BEHAVIOR

Dawkins, R. & J. R. Krebs. (1978). Animal signals: Information or manipulation? In

J. R. Krebs & N. B. Davies (eds.), Behavioral Ecology: An Evolutionary
Approach (pp. 282–309). Sunderland, MA: Sinauer Associates.
Di Bitetti, M. S. (2003). Food-associated calls of tufted capuchin monkeys
(Cebus apella nigritus) are functionally referential signals. Behaviour, 140,
565–592.
Di Bitetti, M. S. (2005). Food-associated calls and audience effects in tufted capuchin
monkeys, Cebus apella nigritus. Animal Behaviour, 69, 911–919.
Digweed, S. M., L. M. Fedigan, & D. Rendall. (2005). Variable specificity in the
anti-predator vocalizations and behaviour of the white-faced capuchin, Cebus
capucinus. Behaviour, 142, 997–1021.
Dittus, W. (1988). An analysis of toque macaque cohesion calls from an ecological
perspective. In D. Todt, P. Goedeking, & D. Symmes (eds.), Primate Vocal
Communication (pp. 31–50). Berlin: Springer.
Dyer, F. C. (2002). The biology of the dance language. Annual Review of Entomology,
47, 917–949.
Ekman, P. (1973). Cross-cultural studies of facial expression. In P. Ekman (ed.), Dar-
win and Facial Expression: A Century of Research in Review (pp. 169–222). New
York: Academic Press.
Elgar, M. (1986). House sparrows establish foraging flocks by giving chirrup calls if
the resources are divisible. Animal Behaviour, 34, 169–174.
Endler, J. A. & A. L. Basolo. (1998). Sensory ecology, receiver biases and sexual selec-
tion. Trends in Ecology and Evolution, 13, 415–420.
Evans, C. S. (1997). Referential signals. In D. H. Owings, M. D. Beecher, & N. S.
Thompson (eds.), Perspectives in Ethology (pp. 99–143). New York: Plenum.
Evans, C. S. & L. Evans. (1999). Chicken food calls are functionally referential. Ani-
mal Behaviour, 58, 307–319.
Evans, C. S. & L. Evans. (2007). Representational signalling in birds. Biology Letters,
3, 8–11.
Evans, C. S., L. Evans, & P. Marler. (1993). On the meaning of alarm calls: Func-
tional reference in an avian vocal system. Animal Behaviour, 46, 23–38.
Evans, C. S., J. M. Macedonia, & P. Marler. (1993). Effects of apparent size and
speed on the response of chickens, Gallus gallus, to computer-generated simu-
lations of aerial predators. Animal Behaviour, 46, 1–11.
Evans, C. S. & P. Marler. (1991). On the use of video images as social stimuli in
birds: Audience effects on alarm calling. Animal Behaviour, 41, 17–26.
Fichtel, C. & P. Kappeler. (2002). Anti-predator behavior of group-living Malagasy
primates: Mixed evidence for a referential alarm call system. Behavioral Ecology
and Sociobiology, 51, 262–275.
Flower, T. (2010). Fork-tailed drongos use deceptive mimicked alarm calls to steal
food. Proceedings of the Royal Society of London, B, 278, 1548–1555.
Flower, T. P. & M. Gribble. (2012). Kleptoparasitism by attacks versus false alarm
calls in fork-tailed drongos. Animal Behaviour, 83, 403–410.
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 203

Furrer, R. D. & M. B. Manser. (2009). The evolution of urgency-based and function-

ally referential alarm calls in ground-dwelling species. American Naturalist,
173, 400–410.
Gallistel, C. R. (1990). The Organization of Learning. Cambridge, MA: MIT Press.
Grafen, A. (1990). Biological signals as handicaps. Journal of Theoretical Biology, 144,
517–546.
Griffin, S. R., M. L. Smith, & T. D. Seeley. (2012). Do honeybees use the directional
information in round dances to find nearby food sources? Animal Behaviour,
83, 1319–1324.
Gros-Louis, J. (2004a). Responses of white-faced capuchins (Cebus capucinus) to
naturalistic and experimentally presented food-associated calls. Journal of Com-
parative Psychology, 118, 396–402.
Gros-Louis, J. (2004b). The function of food-associated calls in white-faced capuchin
monkeys, Cebus capucinus, from the perspective of the signaller. Animal Behav-
iour, 67, 431–440.
Gros-Louis, J. (2006). Acoustic analysis and contextual description of food-associated
calls in white-faced capuchin monkeys (Cebus capucinus). International Journal
of Primatology, 27, 273–294.
Gyger, M. & P. Marler. (1988). Food calling in the domestic fowl, Gallus gallus: The
role of external referents and deception. Animal Behaviour, 36, 358–365.
Hauser, M. D. (1996). The Evolution of Communication. Cambridge, MA: MIT
Press.
Hauser, M. D. & P. Marler. (1993). Food-associated calls in rhesus macaques
(Macaca mulatta). 2. Costs and benefits of call production and suppression.
Behavioral Ecology, 4, 206–212.
Hauser, M. D., P. Teixidor, L. Field, & R. Flaherty. (1993). Food-elicited calls in
chimpanzees: Effects of food quantity and divisibility? Animal Behaviour, 45,
817–819.
Hauser, M. D. & R. W. Wrangham. (1987). Manipulations of food calls in captive
chimpanzees: A preliminary report. Folia Primatologica, 48, 207–210.
Hebets, E. I. & D. R. Papaj. (2005). Complex signal function: Developing a frame-
work of testable hypotheses. Behavioral Ecology and Sociobiology, 57, 197–214.
Heinrich, B. & J. Marzluff. (1991). Do common ravens yell because they want to
attract others? Behavioral Ecology and Sociobiology, 28, 13–21.
Hinde, R. A. (1981). Animal signals: Ethological and games-theory approaches are
not incompatible. Animal Behaviour, 29, 535–542.
Johnstone, R. A. (1996). Multiple displays in animal communication: “Backup sig-
nals” and “multiple messages.” Philosophical Transactions of the Royal Society
of London, B, 351, 329–338.
Kiriazis, J. & C. N. Slobodchikoff. (2006). Perceptual specificity in the alarm calls of
Gunnison’s prairie dogs. Behavioural Processes, 73, 29–35.
Kitzmann, C. & N. Caine. (2009). Marmoset (Callithrix geoffroyi) food-associated
calls are functionally referential. Ethology, 115, 439–448.
204 ANIMAL BEHAVIOR

Kokolakis, A., C. L. Smith, & C. S. Evans. (2010). Aerial alarm calling by male fowl
(Gallus gallus) reveals subtle new mechanisms of risk management. Animal
Behaviour, 79, 1373–1380.
Lachmann, M. & C. T. Bergstrom. (2004). The disadvantage of combinatorial com-
munication. Proceedings of the Royal Society of London, B, 271, 2337–2343.
Laidre, M. E. (2006). Manipulation without mind-reading: Information suppression
and leakage during food discovery by mandrills (Mandrillus sphinx). Behaviour,
143, 365–392.
Leavens, D. A. & W. D. Hopkins. (1998). Intentional communication by chimpan-
zees (Pan troglodytes): A cross-sectional study of the use of referential gestures.
Developmental Psychology, 34, 813–822.
Leavens, D. A., W. D. Hopkins, & R. K. Thomas. (2004). Referential communica-
tion by chimpanzees (Pan troglodytes). Journal of Comparative Psychology, 118,
48–57.
Liebal, K. & S. Pika. (2007). “Hands-on communication”: Use of gestures in apes
and humans. In Proceedings of Interacting Bodies Conference, Lyons, France,
2005. http://gesture-lyon2005.ens-lyon.fr/article.php3?id_article=240http://
gesture-lyon2005.ens-lyon.fr/article.php3?id_article=240.
Macedonia, J. M. & C. S. Evans. (1993). Variation among mammalian alarm call sys-
tems and the problem of meaning in animal signals. Ethology, 93, 177–197.
Mahurin, E. J. & T. M. Freeberg. (2009). Chick-a-dee call variation in Carolina
chickadees and recruiting flockmates to food. Behavioral Ecology, 20, 111–116.
Manser, M. B., M. B. Bell, & L. B. Fletcher. (2001). The information that receivers
extract from alarm calls in suricates. Proceedings of the Royal Society of London,
B, 268, 2485–2491.
Manser, M. B., Seyfarth, R. M. & Cheney, D. L. (2002). Suricate alarm calls signal
predator class and urgency. Trends in Cognitive Sciences, 6, 55–57.
Marler, P. (1955). Characteristics of some animal calls. Nature, 176, 6–8.
Marler, P. (1976). Social organization, communication, and graded signals: The
chimpanzee and the gorilla. In P. P. G. Bateson & R. A. Hinde (eds.), Growing
Points in Ethology (pp. 239–280). Cambridge, UK: Cambridge University
Press.
Marler, P., A. Dufty, & R. Pickert. (1986). Vocal communication in the domestic
chicken: I. Does a sender communicate information about the quality of a
food referent to a receiver? Animal Behaviour, 34, 188–193.
Marler, P., C. S. Evans, & M. Hauser. (1992). Animal signals: Motivational, referen-
tial, or both? In H. Papousek, U. Jurgens, & M. Papousek (eds.), Nonverbal
Vocal Communication: Comparative and Developmental Approaches (pp. 66–
86). Cambridge, UK: Cambridge University Press.
Maynard Smith, J. (1965). The evolution of alarm calls. American Naturalist, 99, 59–
63.
McBride, G., I. P. Parer, & F. Foenander, (1969). The social organisation and behav-
iour of the feral domestic fowl. Animal Behaviour Monographs, 2, 127–181.
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 205

McConnell, P. B. & J. R. Baylis. (1985). Interspecific communication in cooperative

herding: Acoustic and visual signals from human shepherds and herding dogs.
Zeitschrift für Tierpsychologie, 67, 302–328.
McGregor, P. K. (2009). Communication networks and eavesdropping in animals. In
L. R. Squire (ed.), Encyclopedia of Neuroscience (pp. 1179–1184). Oxford, UK:
Elsevier.
McGregor, P. K. & T. M. Peake. (2000). Communication networks: Social environ-
ments for receiving and signalling behaviour. Acta Ethologica, 2, 71–81.
McQuoid, L. M. & B. G. Galef Jr. (1993). Social stimuli influencing feeding behav-
iour of Burmese red junglefowl: A video analysis. Animal Behaviour, 46, 13–
22.
Møller, A. P. & A. Pomiankowski. (1993). Why have birds got multiple sexual orna-
ments? Behavioral Ecology and Sociobiology, 32, 167–176.
Morton, E. S. (1977). On the occurrence of motivation-structural rules in some birds
and mammal sounds. American Naturalist, 111, 855–869.
Overington, S. E., J. Morand-Ferron, N. J. Boogert, & L. Lefebvre. (2009). Techni-
cal innovations drive the relationship between innovativeness and residual
brain size in birds. Animal Behavior, 78, 1001–1010.
Owings, D. H. & D. F. Hennessy. (1984). The importance of variation in sciurid vis-
ual and vocal communication. In J. O. Murie & G. R. Michener (eds.), The
Biology of Ground-dwelling Squirrels (pp. 169–200). Lincoln: University of
Nebraska Press.
Owings, D. H. & E. S. Morton. (1998). Animal Vocal Communication: A New
Approach. Cambridge, UK: Cambridge University Press.
Partan, S. R. & P. Marler. (1999). Communication goes multimodal. Science, 283,
1272–1273.
Partan, S. R. & P. Marler. (2005). Issues in the classification of multimodal commu-
nication signals. American Naturalist, 166, 231–245.
Peake, T. (2006). Animal communication networks. In K. Brown (ed.), Encyclopedia
of Language & Linguistics, Second Edition (pp. 264–267). Oxford, UK:
Elsevier.
Pika, S. and T. Bugnyar. (2011). The use of referential gestures in ravens (Corvus
corax) in the wild. Nature Communications, 2, 560.
Pika, S. & J. Mitani. (2006). Referential-gestural communication in wild chimpan-
zees (Pan troglodytes). Current Biology, 16, R191–R192.
Popper, K. (1959). The Logic of Scientific Discovery. Translation of Logik der For-
schung. London: Hutchinson.
Rendall, D., M. J. Owren, & M. Ryan. (2009). What do animal signals mean? Ani-
mal Behaviour, 78, 233–240.
Robinson, S. R. (1980). Antipredator behavior and predator recognition in Belding’s
ground squirrels. Animal Behaviour, 28, 840–852.
Robinson, S. R. (1981). Alarm communication in Belding’s ground squirrels. Zeits-
chrift für Tierpsychologie, 56, 150–168.
Rogers, L. (1980). Lateralisation in the avian brain. Bird Behaviour, 2, 1–12.
206 ANIMAL BEHAVIOR

Rogers, L. (1990). Light input and the reversal of functional lateralization in the
chicken brain. Behavioural Brain Research, 38, 211–221.
Rowe, C. (1999). Receiver psychology and the evolution of multicomponent signals.
Animal Behaviour, 58, 921–931.
Ryan, M. J. & A. S. Rand. (1998). Evoked vocal response in male túngara frogs: Pre-
existing biases in male responses? Animal Behavior, 56, 1509–1516.
Schradin, C. (2000). Confusion effect in a reptilian and a primate predator. Ethology,
106, 691–700.
Searcy, W. A. & S. Nowicki. (2005). The Evolution of Animal Communication.
Princeton, NJ: Princeton University Press.
Seyfarth, R. M., & D. L. Cheney. (2010). Production, usage, and comprehension in
animal vocalizations. Brain and Language, 115, 92–100.
Seyfarth, R. M., D. L. Cheney, & T. Bergman. (2005). Primate social cognition and
the origins of language. Trends in Cognitive Sciences, 9, 264–266.
Seyfarth, R. M., D. L. Cheney, T. Bergman, J. Fische, K. Zuberbühler, &
K. Hammerschmidt. (2010). The central importance of information in studies
of animal communication. Animal Behaviour, 80, 3–8.
Seyfarth, R. M., D. L. Cheney, & P. Marler. (1980). Monkey responses to three dif-
ferent alarm calls: Evidence of predator classification and semantic communi-
cation. Science, 210, 801–803.
Sherman. P. (1980). The meaning of nepotism. The American Naturalist, 116, 604–606.
Sherman, P. W. (1985). Alarm calls of Belding’s ground squirrels to aerial predators:
Nepotism or self-preservation? Behavioral Ecology and Sociobiology, 17,
313–323.
Shettleworth, S. (2010). Cognition, Evolution, and Behavior. Oxford, UK: Oxford
University Press.
Shultz, S. M. (2001). Notes on interactions between monkeys and African crowned
eagles in Taï National Park, Ivory Coast. Folia Primatologica, 72, 248–250.
Skinner, B. F. (1974). About Behaviorism. New York: Vintage.
Slocombe, K. E. & K. Zuberbuhler. (2005). Functionally referential communication
in a chimpanzee. Current Biology, 15, 1779–1784.
Slocombe, K. E. & K. Zuberbühler. (2006) Food-associated calls in chimpanzees:
Responses to food types or food preferences? Animal Behaviour, 72, 989–999.
Smith, C. L. & C. S. Evans. (2008). Multimodal signaling in fowl, Gallus gallus. Jour-
nal of Experimental Biology, 211, 2052–2057.
Smith, C. L. & C. S. Evans. (2009). Silent tidbitting in male fowl, Gallus gallus: A
referential visual signal with multiple functions. Journal of Experimental Biol-
ogy, 212, 835–842.
Smith, C. L., A. Taylor, & C. S. Evans. (2011). Tactical multimodal signalling in
birds: Facultative variation in signal modality reveals sensitivity to social costs.
Animal Behavior, 82, 521–527.
Smith, C. L., D. A. Van Dyk, P. W. Taylor, & C. S. Evans. (2009). On the function
of an enigmatic ornament: Wattles increase the conspicuousness of visual dis-
plays in male fowl. Animal Behaviour, 78, 1433–1440.
REFERENTIAL SIGNALS : A WINDOW INTO ANIMAL MINDS 207

Srinivasan, M., S. Zhang, & J. Reinhard. (2006). Small brains, smart minds: Vision,
perception, navigation, and “cognition” in insects. In E. Warrant & D. E.
Nilsson (eds.), Invertebrate Vision (pp. 462–493). Cambridge, UK: Cambridge
University Press.
Stokes, A. W. (1971). Parental and courtship feeding in red jungle fowl. Auk, 88,
21– 29.
Stokes, A. W. & H. W. Williams. (1971). Courtship feeding in gallinaceous birds. Auk,
88, 543–559.
Stokes, A. W. & H. W. Williams. (1972). Feeding calls in gallinaceous birds. Auk,
89, 177–180.
Struhsaker, T. T. (1967). Auditory communication among vervet monkeys (Cerco-
pithecus aethiops). In S. Altmann (ed.), Social Communication among Primates
(pp. 281–324). Chicago: University of Chicago Press.
Suzuki, T. N. (2012). Referential mobbing calls elicit different predator-searching
behaviours in Japanese great tits. Animal Behaviour, 84, 53–57.
Templeton, C. N., E. Greene, & K. Davis. (2005). Allometry of alarm calls: Black-
capped chickadees encode information about predator size. Science, 308,
1934–1937.
Van Dyk, D. & C. S. Evans. (2008). Opponent assessment in lizards: Examining the
effect of aggressive and submissive signals. Behavioral Ecology, 19, 895–901.
von Frisch, K. (1967). The Dance Language and Orientation of Bees. Cambridge, MA:
Harvard University Press.
Walters, J. R. (1990). Anti-predatory behavior of lapwings: Field evidence of discrimi-
native abilities. Wilson Bulletin, 102, 49–70.
Watanabe, S. & N. F. Troje. (2006). Towards a “virtual pigeon”: A new technique for
investigating avian social perception. Animal Cognition, 9, 271–279.
Wenner, A. M. (2002). The elusive honey bee dance “language” hypothesis. Journal
of Insect Behavior, 15, 859–878.
Wenner, A. M. (2006). A failure to “resolve” the bee language controversy. American
Bee Journal, 146, 561–561.
Wheeler, B. (2010). Production and perception of situationally variable alarm calls in
wild tufted capuchin monkeys (Cebus apella nigritus ). Behavioral Ecology and
Sociobiology, 64, 989–1000.
Wickler, W. (1978). A special constraint on the evolution of composite signals. Zeitschrift
für Tierpsychologie, 48, 345–348.
Wilczynski, W., A. S. Rand, & M. J. Ryan. (2001). Evolution of calls and auditory
tuning in the Physalaemus pustulosus species group. Brain Behavior and Evolu-
tion, 58, 137–151.
Wilson, D. R. & C. S. Evans. (2008). Mating success increases alarm-calling effort in
male fowl, Gallus gallus. Animal Behaviour, 76, 2029–2035.
Wilson, D. R. & C. S. Evans. (2012). Fowl communicate the size, speed and proxim-
ity of avian stimuli through graded structure in referential alarm calls. Animal
Behaviour, 83, 535–544.
208 ANIMAL BEHAVIOR

Wood, S. R., K. J. Sanderson, & C. S. Evans. (2000). Perception of terrestrial and

aerial alarm calls by honeyeaters and falcons. Australian Journal of Zoology,
48, 127–134.
Zuberbühler, K., D. L. Cheney, & R. M. Seyfarth. (1999). Conceptual semantics in a
nonhuman primate. Journal of Comparative Psychology, 113, 33–42.
 8

Cotton-top Tamarins: Research for

Conservation and Understanding
Human Behavior
Charles T. Snowdon

WHAT IS A TAMARIN?
Few people have heard about tamarins, often confusing them with a tropical
fruit (tamarind). Tamarins are small monkeys (about the size of a squirrel),
weighing about 1 to 1.5 pounds (450–650 g), that live in the forests of South
America. All species are arboreal, spending almost all their lives in trees, where
they eat tree exudates (sap or gum), small insects and vertebrates, fruits, and
legumes.
Tamarins are part of the family Callitrichidae, which are the most primitive
(ancient) of true monkeys. Most taxonomists consider five genera in the fam-
ily, with tamarins (Saguinus), lion tamarins (Leontopithecus), marmosets (Cal-
lithrix), pygmy marmosets (Cebuella), and Goeldi’s monkey (Callimico).
Many species are found in the Amazon basin in Brazil, Peru, Ecuador, and
Colombia, but a few species are found on the Caribbean coast in Panama
and Colombia and in remnants of the Atlantic coastal forest in eastern Brazil.
Marmosets have specialized teeth that allow them to gouge holes in bark to
create an exudate flow, whereas tamarins and lion tamarins lack these special-
ized teeth and can only obtain exudates when other species open holes in bark.
210 ANIMAL BEHAVIOR

Unlike other monkeys all marmosets and tamarins have claws that allow them
to climb trees readily.
All species of marmosets and tamarins live in small family groups and are
known as cooperative breeders. This means that not only mothers, or moth-
ers and fathers, but all group members work together to help rear infants.
Mothers usually give birth to twins, and the twins at birth weigh about 10 per-
cent of their mother’s weight (imagine a 130-pound woman giving birth to
two 13-pound babies). Because monkeys (unlike birds) do not build nests,
the heavy infants must be carried by someone as the group moves throughout
its daily range. And because mothers can become pregnant again within a few
weeks of giving birth, the value of having many others help with infant care is
enormous. Typically only one female reproduces at a time. This type of co-
operative child-rearing system is rare among mammals, with meerkats, mon-
gooses, and wolves among the other cooperative breeders. The only other
primate species characterized by cooperative breeding is our own (Hrdy,
1999).

WHY STUDY TAMARINS?

I have spent more than 30 years studying one species of tamarin—the
cotton-top tamarin (Saguinus oedipus). Although I have worked with several
other species over the same period, the cotton-top tamarin has been my main
focus. What is special about cotton-top tamarins, and why spend so much of
my life studying them? There are two main reasons. First, cotton-top tamarins
are highly endangered, and so there is an ethical need to learn as much about
them as possible not only to help create a reason for preserving them but also
to preserve knowledge about them in case they become extinct. Second, I am
an animal behaviorist working within a psychology department, and there is
strong pressure from my colleagues and students to make my research relevant
to human behavior. As a psychologist I am also fascinated by human behavior
and see research with other species as an important tool to learn more about
ourselves and our potential.
I did not start my career intending to study cotton-top tamarins, or indeed
primates of any kind. My involvement with primate behavioral research was
due to a series of accidents. The first species I studied was the pygmy marmo-
set (Cebuella pygmaea), the world’s smallest true monkey, found in the western
Amazon. An undergraduate student had seen some in a back room at the
nearby Wisconsin National Primate Center, and I learned that no one was
studying them. I brought them to the psychology building and began to study
how they communicate with one another. I realized quickly that I needed to
COTTON - TOP TAMARINS 211

learn more about primate behavior and spent a year working with one of the
world’s foremost primatologists, Thelma Rowell at the University of Califor-
nia, Berkeley. She had a colony of cotton-top tamarins acquired as part of a
divorce settlement some years earlier. One of her graduate students insisted
that her ex-husband and Thelma write a grant so she could study male paren-
tal care. A small number of monkeys were imported just before the species was
declared to be endangered in 1970. By the time I arrived, the grant money had
run out and the student had dropped out of graduate school. Thelma offered
me the monkeys, and I jumped at the opportunity to bring them to
Wisconsin.
Why were cotton-top tamarins endangered? The cotton-top tamarin is
found only in Colombia and only in the northern part of the country, outside
of the Amazon basin adjacent to the Caribbean. This is also one of the areas of
the country with the greatest human population, in two major cities, Barran-
quilla and Cartagena. As the human population expanded, more and more
forest was cut to create grazing land for cattle and to provide building materials
and firewood. Even more critically, because of its relatively close access to the
United States, an estimated 20,000 to 30,000 tamarins had been exported for
biomedical research in the two decades before the species was determined to
be endangered. Today the plight of the cotton-top tamarin is even worse.
More habitat is being destroyed every year, and a survey, published in 2010,
estimated approximately 7,300 cotton-top tamarins remained in the wild
(Savage et al., 2010).
Although habitat preservation is crucial, Colombia has several areas of
political instability, making the creation and enforcement of preserves diffi-
cult. Furthermore, in order to make intelligent decisions about conservation,
good research is needed to know about social organization, behavioral devel-
opment, migration, and more. In the absence of the possibility to create
adequate natural reserves for an endangered species, captive breeding may be
essential. A successful captive breeding environment must allow for multiple
generations to be reared with little or no intervention from humans. Further-
more, the captive environments must allow animals to develop species-typical
behaviors, and, if some behaviors must be acquired through learning, then
adequate learning opportunities must be provided. From a conservation per-
spective, behavioral research has great importance.
But what value can a small, squirrel-sized monkey have for understanding
human behavior? Are not the species closest to us, the great apes, most relevant
for understanding our origins? It is true that tamarins are distantly related to
humans, but they share with us a similar social system that involves small,
close-knit family-like groups. There are two main models for understanding
212 ANIMAL BEHAVIOR

the processes of evolution. We are most familiar with diverging evolution

where two populations become increasingly different from one another, even-
tually forming separate species. Traits that are shared between two different
species can be attributed to the nearest common ancestor of both species.
Thus, if we share behavioral traits with our nearest relatives, chimpanzees
(Pan troglodytes) and bonobos (Pan paniscus), we can infer that these arose in
our common ancestor.
However, there is also a concept of converging evolution. In this case spe-
cies with very different evolutionary backgrounds that may have diverged long
ago converge in similar behavioral traits due to similarities in social systems or
environments that lead to similar solutions to adaptive problems. By the logic
of convergent evolution, diverse species with shared environmental or social
systems may evolve similar behavioral patterns.
If humans are cooperative breeders, then there are several interesting prob-
lems that can be studied in a nonhuman primate that is also a cooperative
breeder. Human behavior is frequently influenced by cultural variables that
may mask the biological basis of our behavior or may obscure some human
behavioral potential. By studying a nonhuman cooperative breeder we may
better understand the mechanisms underlying our own behavior, the develop-
mental processes involved, and the adaptive significance of behavior.
Male parental care and cooperative breeding more broadly create several
problems for traditional evolutionary explanations. No male mammal can ever
be certain that he is the father of any child (prior to DNA fingerprinting), but
female mammals are biologically committed to their infants through the costs
of pregnancy and nursing. Under such conditions it has been argued that it is
more adaptive for a male to invest in mating with as many females as possible
while paying little attention to infant care. Yet cooperative breeders are charac-
terized by high levels of paternal care, along with care by other group mem-
bers. How can it be adaptive for males to engage in infant care? Are there
specific developmental processes and physiological changes that lead males
and others to become involved in infant care? Cooperatively breeding primates
are also characterized by a strong emotional relationship or pair bond between
breeding adults. How does this bond develop, and what sustains the relation-
ship? Only one female in a group breeds at a time, with other females not even
ovulating. How do breeding adults keep other group members from breeding,
and how do they retain helpers to provide quality infant care? What benefits
do helpers gain by caring for someone else’s infants rather than striking out
and breeding on their own?
In addition, tamarins have a well-developed vocal repertoire with many
calls and call sequences. Forest-living primates use vocal communication
COTTON - TOP TAMARINS 213

much more often than our closest relatives, the great apes, and we humans are
also a highly vocal species. Chimpanzees are remarkably quiet in comparison
to humans. Are there any possible parallels between the vocal communication
of tamarins and aspects of our communication and language?
Human beings appear to have a capacity for cooperation and social support
that is not always seen in great apes. Is there something about a cooperative
breeding system that makes individuals more attentive to other group mem-
bers, more likely to learn from them, and more likely to share resources and
even teach others? Tamarins and their relatives appear to be more capable of
learning from one another, of sharing resources and information about the
location of resources, and even of teaching their young compared with chim-
panzees and our other close relatives.
The study of cotton-top tamarins has provided knowledge about successful
captive breeding and has provided many ideas that can be applied to conserva-
tion. At the same time these studies inform us about interesting aspects of
human behavior that appear to be the result of convergence due to similarities
of cooperative breeding. Now let me tell you about the research methods we
have developed, and then I will move on to some of the interesting results.
Several other researchers have studied aspects of behavior communication
and cognition in cotton-top tamarins, but rather than provide an exhaustive
review, I will focus here primarily on work that my students and collaborators
have done.

HOW DO I STUDY TAMARINS?

I received 11 tamarins from Thelma Rowell in 1977, not enough to do
much research, and the history of captive breeding of the species was also
rather dismal. Most biomedical research facilities with cotton-top tamarins
had great breeding success with wild-caught animals, but few captive-born
animals were successfully reproducing. Thus, if I was going to learn anything
about tamarins, I would need to develop a self-sustaining breeding colony
and find ways to increase the size of my colony to have adequate sample sizes
for research.

Successful Breeding
What was the reason for the lack of breeding success for captive tamarins?
As I looked at the reports from other colonies, I noticed that most facilities
removed infants from their parents, often before natural weaning. Many labs
had nurseries where infants could be fed by hand. The managers of these labs
made two assumptions. The first assumption was that if infants were weaned
214 ANIMAL BEHAVIOR

early, mothers would be more likely to start cycling again, because lactation
was thought to suppress ovulation, and thus overall reproductive output
would be greater. The second assumption was that parenting skills were innate
and experience played no role in successful parenting. Both assumptions were
wrong.
Our first hormonal studies showed that mothers ovulated within two to
four weeks after giving birth regardless of whether they were still nursing or
not (Ziegler, Bridson, et al., 1987). Thus there was no suppression due to
nursing. Summaries of colony breeding records from closely related species
showed that breeding success was high when both parents had previous expe-
rience helping others raise their young (Epple, 1978). When only the mother
had prior experience and the father was naïve with respect to infant care, the
number of surviving infants was much lower, and when neither parent had
prior infant care experience, survival rate of their offspring was near zero.
We did an informal experiment when the tamarins were still at Berkeley.
Two males had grown up with no chance to help care for someone else’s off-
spring, and so we took a young infant from another group and presented it
to these two males. As soon as the infant started to cry, one of the males ran
to pick it up, but as soon as the infant was on its back, the male acted as if
he was very uncomfortable and started to bite at the infant and push it off.
As soon as the infant was off, it began crying again, and the males went back
to retrieve it, repeating the process over and over until we retrieved the infant
ourselves and put it back in its family.
All of this suggested that we had to find a new method for husbandry. We
decided that young born into a family should stay long enough to have expe-
rience looking after at least two other sets of offspring. In order to allow for
this we needed large enclosures to house several family members. I also noted
when observing the animals at Berkeley that they were very hesitant to
approach the food and water dishes that we placed on the floor of their cages.
It took some time for me to realize that for monkeys that usually live in trees,
they would naturally be fearful of approaching the ground. So we built large,
high walk-in cages and built platforms for all the food and water at least 1 m
above the floor. Tamarins weigh about 1 percent of an average adult human
and, based on this ratio, the floor space of a family enclosure was the equiva-
lent of a 5,800 square foot (540 m2) house, and that of an enclosure for just
two monkeys the equivalent of a 2,000 square foot (185 m2) house.
Because tamarins have claws and they scent mark a lot, we thought it is
important to have lots of branches and ropes in the cages to simulate an arbo-
real environment. But we ran into trouble with our veterinarian. He wanted
hard surfaces that could be sterilized, but these would be dangerous for
COTTON - TOP TAMARINS 215

climbing and jumping and would not retain the scent marks important for
communication. We finally reached a compromise whereby we were allowed
to use branches and ropes, but we agreed to remove them completely every
three months and replace them with new structures. This had an added ben-
efit of providing new climbing and jumping patterns for the animals to master
and thus keep their sensorimotor skills sharp (Snowdon et al., 1985).

Noninvasive Research Methods

We developed a variety of noninvasive research methods. Because we had
constructed large cages to house family groups, we realized that it would be
difficult to catch an animal every time we wanted to do a study. Not only
would the process of catching an animal make life hard for us, it would defi-
nitely increase the stress level of the animals as well so that we could not be
confident of the quality of any data we were gathering. Therefore we found a
variety of ways to study tamarins in their home cages without having to handle
them or disturb the rest of the family. We put a scale on a moveable cart and
could readily train each animal with small treats to sit briefly on the scale while
we gathered the weight. We could monitor the growth of young infants by
weighing the carrier and the infant together, waiting until the infant was car-
ried by someone else, and then weighing the first carrier without the infant.
We developed ways to place experimental apparatus on the sides of the cage
and could distract other group members while testing one animal at a time.
We could hide speakers in the lights or behind curtains to do playback studies
to see how animals respond to vocal signals from familiar or strange individ-
uals, and we could habituate animals to the presence of an observer and take
careful behavioral observations.
Perhaps our greatest success has been with the development of noninva-
sive hormone assays. Many of the questions that we wanted to answer about
reproductive behavior required that we have measurements of hormones.
When we started, the gold standard for hormonal measurement was the level
found in the serum. But collecting serum samples has several problems. First,
an animal has to be captured and a blood sample taken. Second, because
sampling serum also requires removing red blood cells as well, sampling can-
not occur very frequently in small-bodied animals without compromising the
health of the animal. Third, many hormones are released in pulses, and
unless one takes frequent samples, one might miss the pulse. How do we
avoid these problems? Our solution was to collect urine samples. Because
we had walk-in cages, we could collect urine samples be simply following
an animal around and holding a bucket under it until it urinated (Ziegler,
Bridson, et al., 1987; Ziegler, Savage, et al., 1987). We quickly realized that
216 ANIMAL BEHAVIOR

tamarins, like humans, often urinate when they first wake up in the morning,
so it became relatively easy to wake up a cage of animals and go in to follow
the animals from which we wanted to collect a urine sample. This early-
morning collection had another important advantage. Because the monkeys
do not appear to urinate during the night, our samples reflected all of the
hormonal activity from the preceding 12 hours. Thus any pulses of hormonal
release in the blood would be collected and averaged over the 12 hours.
Because we did not have to catch or handle the animals and because animals
urinated several times a day, we did not have to worry about inducing stress
or about reducing red blood cell levels. We could sample an animal every
day, and we could even obtain hormonal samples from a mother on the
day she gave birth and throughout nursing.
But we had to convince skeptics that our measures were valid. My former
student Jeffrey French, now at the University of Nebraska–Omaha, developed
the first assays for estrogen and luteinizing hormones (French et al., 1984),
and my close colleague and collaborator Toni Ziegler of the Wisconsin
National Primate Center has developed and validated many other assays. In
the early validations, we often had to inject a small number of animals with
radioactive hormone tracers so that we could demonstrate that labeled hor-
mones were excreted in urine in proportion to what we had injected (e.g., Zie-
gler et al., 1993). Later on we were able to dispense with using radioactive
hormones, but we have still had to show for each hormone that what we are
measuring makes sense biologically. We have been able to validate assays to
measure the gonadal steroid hormones, testosterone, dihydrotestosterone,
estrogen, and estradiol; the adrenal steroid hormones, cortisol and corticoster-
one; the peptides, oxytocin and vasopessin; and several protein hormones
including prolactin and luteinizing hormone. Oddly enough, we could not
measure progesterone as our radiolabeled studies showed that all of the labeled
progesterone was excreted in the feces and none in the urine (Ziegler et al.,
1993).
The development of hormonal assays for both urine and feces has had far-
reaching effects. Toni Ziegler has collaborated with scientists around the
world to develop noninvasive assays for more than 30 different species,
including many endangered species. As a result field researchers can now
study reproductive function in many species, and efforts to conserve threat-
ened or endangered animals can be based on sound knowledge of reproduc-
tive biology. Today fecal, urinary, and saliva assays have virtually replaced
serum sampling as the method of choice, and we can study a broad array
of species, including humans, with noninvasive methods (Snowdon &
Ziegler, 2007b).
COTTON - TOP TAMARINS 217

Laboratory versus Field Studies

Are laboratory studies valid ways of understanding the behavior of any spe-
cies? Many field biologists will say no. Yet there are several advantages to be
gained in studies in captivity. It is easy to identify individuals, and one can fol-
low an individual from birth through reproductive years and beyond, carrying
out longitudinal studies that may be difficult in the field. One can also do
experimental manipulations in captivity with controls that might be impos-
sible in the wild. It is much easier to obtain hormonal samples on a regular
basis than could be done otherwise. And finally, it is easier to habituate captive
animals to human observers, and we can observe behavior and communication
at a much closer distance than could be done in the wild, allowing us to
observe subtle behaviors and quiet vocalizations that might not be detected
in a field study. However, for these manipulations and observations to have
any meaning, it is critical to have captive animals living in natural social groups
with enough space to be able to carry out most of their natural behaviors.
And yet we cannot begin to understand the ecological validity of any cap-
tive research without a good understanding of life in the wild. Good captive
research must be complemented by field research. For an endangered species
such as the cotton-top tamarin, it is critical to understand life in the wild,
but one must temper those observations with the fact that the remaining hab-
itat in which we can observe tamarins today may not be anything like the
ancestral habitat in which the tamarins evolved.
I worked with an outstanding undergraduate, Anne Savage, now Conserva-
tion Director, Disney’s Animals, Science and Environment. She published a
paper on twin orangutans based on work she began in her freshman year,
and her senior thesis won the best student paper award from the American
Society of Primatologists in competition with Ph.D. students. Anne remained
to do her Ph.D. with me and wanted to do a field study on cotton-top tama-
rins. We began meeting with government officials and were invited to visit
Colombia. There was a field station near the small town of Coloso with lodg-
ing and several outdoor captive environments to house tamarins that had been
confiscated from people trying to sell them as pets. The field station was on
the edge of a remnant forest that had a small population of wild tamarins.
On our first visit to Coloso, we noted how little forest we could see on the
drive from Barranquilla. At the field station we were shown a pastured hillside
that 10 years earlier had been forested, with five species of monkeys found
there. Even as pasture the soil did not provide much nutrients, and the cattle
seemed emaciated. However, the first morning we saw wild cotton-top tama-
rins chorusing with the captive animals. This remnant forest was along a
stream used by villagers to water their animals and get their own water.
218 ANIMAL BEHAVIOR

The noise from human activities overwhelmed our ability to record tamarin
vocalizations, and we frequently saw tamarins on the ground where the forest
canopy was not dense enough to allow animals to travel from tree to tree.
Nevertheless, Anne Savage and her collaborators studied tamarins at Coloso
for six years until guerilla activity closed off access to the village and its forest.
I will say more about specific field results below.

WHAT HAVE WE LEARNED?

Over the years we have learned much about the behavioral biology of
cotton-top tamarins through both captive and field research. I will describe
some of the most interesting findings here, and occasionally I will supplement
results on tamarins with results from closely related species that I have studied.
There are six broad areas that I will cover. (1) For a cooperative breeding sys-
tem to work there has to be a strong social relationship between the breeding
animals. How are these relationships formed, and what are the mechanisms that
maintain relationships? (2) What mechanisms are involved in letting fathers and
other group members play an important role in parenting, and what are the bene-
fits to infants from having multiple caretakers? (3) How do parents recruit helpers
and maintain enough nonreproductive helpers to sustain a cooperative breeding
system? (4) What are the benefits to nonreproductive helpers of helping with infant
care rather than breeding on their own? (5) How do group members communicate
with one another, and what do they communicate about? How do communication
skills develop in young tamarins? (6) Tamarins cooperate with one another to
provide infant care. Does this cooperation also translate into cooperation among
adults in other realms? Does family life predispose animals for increased attention
to one another and hence to rapid social learning? Is there evidence of teaching
behavior in tamarins?

How Do Tamarins Form and Maintain Relationships?

A basic principle of behavioral ecology is that mammalian females invest
more energy than do males in producing infants, both through pregnancy
and nursing. Also, no male mammal can ever be certain that he is the father
of a given infant, whereas every mammalian mother is absolutely certain that
a given infant is hers. This combination of male uncertainty and female invest-
ment in offspring has led to the view that females must be very choosy about
mates. This argument would seem to predispose males from ever being
involved in infant care, and yet there are many species of mammals (and more
among primates than among any other groups of mammals) where fathers
actively help care for infants.
COTTON - TOP TAMARINS 219

If males are to be involved in parenting then there needs to be some set

of mechanisms to allow a pair to develop a close relationship (often called
a pair bond) whereby the male can be relatively confident that any infants
he cares for are likely to be his own and whereby the female can be reason-
ably confident that the male will stay with her long enough to assist with
infant care.
One reason for a male staying with his mate and being involved in infant
care is if his care is essential for the survival and well-being of his infants. Over
the course of our field study we kept track of group size and survival of infants
and found a close relationship between group size and survival. Only with a
group size of five (mother, father, and three helpers) did we find 100 percent
survival in infants, and we never saw a group of two attempt to breed. We
gathered similar statistics from our captive colony and were surprised to find
the same results. That is, only with a group size of five or more did we find
100 percent infant survival, even though our captive animals had a much
smaller daily travel length, had all the food they could eat, and did not have
to worry about predators. We never had a single mother be able to successfully
rear her infants, although frequently a pair of captive animals could rear
infants successfully (Snowdon, 1996). Here is part of the answer, then. If a
father’s care is necessary for his infants to survive, then it will be adaptive for
him to stay with one mate and help her.
But this leads to the question: How do animals get together and form a rela-
tionship? In the management of our captive colony we had to be careful to
avoid inbreeding, and so we arranged for the pairing of animals rather than
allowing animals to choose their own mates. After several years of field
research with both cotton-top tamarins and the related common marmoset,
found in Brazil, we found little evidence of mate choice in the wild. Breeding
opportunities are rare, and animals have to make the best of whatever potential
mate might be available.
When we first put a new male and a new female together we were
impressed by the high rates of sexual and affiliative behavior we observed
within a few hours (Savage et al., 1988). Copulation could occur several times
an hour, and there was a high rate of mutual grooming and huddling. Over
time these rates decreased, but even in long-term pairs we saw mating on an
average of once a day. Why do new couples and even old, established couples
mate so often? Some have argued that if a female “wants” to keep her mate
with her she needs to conceal when she ovulates. By concealing ovulation a
male can never be sure when a female is fertile, and by the time a pregnancy
has actually been established, the male may actually want to stay with that
female (Burley, 1979).
220 ANIMAL BEHAVIOR

Toni Ziegler and I had a friendly argument about concealed ovulation. She
had noted that in 85 percent of the ovulations that we observed, the female
became pregnant. She argued that to have such a high rate of conception
females must be communicating when they ovulate to their mates. I argued
that there does not need to be any communication so long as you copulate
at least once a day. Toni developed a brilliant behavioral bioassay where she
collected scent marks each day from an ovulating female and then transferred
the marks to cages with a mated pair where the female was already pregnant.
She found that there was a significant increase in male erections and mounting
of the female on the days with odors from when a novel female was ovulating
compared to odors from other times in her cycle, and the females also solicited
sex more often when exposed to odors of novel ovulating females (Washa-
baugh & Snowdon, 1998). Thus ovulation is concealed only to human scien-
tists, and cotton-top tamarins can readily detect when their mates are
ovulating.
So why do tamarins have sex so often? Much of their sexual behavior is
occurring at times when conception is impossible. What could be the value
of nonconceptive sex? I started looking at data from studies where we had
experimentally manipulated a relationship, and I found increased rates of sex-
ual behavior, huddling, and grooming in reunions after a brief (30 minute)
separation from a mate as well as in response to novel odors from breeding
females. This suggested that nonconceptive sex and other affiliation behavior
may be important in restoring a relationship after some disturbance and that
these behaviors may also be important in the formation of a relationship
(Snowdon & Ziegler, 2007a).
We also noted a clear asymmetry in grooming behavior. In every pair of
wild common marmosets and captive tamarins we have observed, the male
always grooms its mate more than the female grooms him. Furthermore, in
studies of brief separations between tamarins in captivity, we found that males
gave twice as many separation calls as females (Snowdon & Ziegler, 2007a).
Both of these findings are consistent with male paternal uncertainty. If a male
is uncertain of paternity, then he should play a more active role in attending to
his mate, and he should be more anxious when he is separated from her.
In recent years the hormone oxytocin, which was thought to be mainly
involved in uterine contractions at birth and in controlling the milk let-
down reflex in nursing, has been shown to be involved in the formation and
maintenance of pair bonds in female voles, but only in monogamous species
(Carter, 1998). In humans administration of oxytocin through a nasal spray
has been shown to increase feelings of trust in men (e.g., Kosfeld et al.,
2005). Does oxytocin play a role in pair bonding in tamarins?
COTTON - TOP TAMARINS 221

We collected urine samples from 14 long-established pairs of tamarins over

the female’s three-week ovulation cycle. During the same period we made
regular observations of the behavior of the pairs and then correlated the behav-
ior with urinary oxytocin levels. We also collected urine samples from several
animals in the week before they were paired and in the week after pairing.
We observed a significant increase in oxytocin levels in newly paired animals.
We also found no difference in oxytocin levels between male and female tam-
arins overall and also a strong correlation between mates. If a male had high
oxytocin, his mate did too. If a female had a low level of oxytocin, her mate
did too (Snowdon et al., 2010).
We observed a great deal of variation in both oxytocin levels and in the
amount of sexual and affiliation behavior that any pair would show. We found
that variation in male oxytocin levels could be best explained by how often
they had sex, and variation in female oxytocin levels could be explained by
how much cuddling and grooming a female received. In the pairs with high
oxytocin levels, females solicited sex more frequently from their mates and
males initiated cuddling and grooming more often, as though each mate
“understood” what its partner needed most (Snowdon et al., 2010). Thus,
relationships between tamarin mates are maintained by nonconceptive sex,
grooming, and cuddling. If this strikes you as at all similar to human relation-
ships, then you can see how we might learn from observing tamarins.

How Does Parenting Develop?

Many of us speak loosely about parenting being an instinct, something that
is completely natural, particularly when we think about mothers. Surely natu-
ral selection must have led to a maternal instinct so that offspring will have a
high chance of survival. I mentioned earlier that the breeding success of cap-
tive tamarins was extremely poor and that one cause of this might be that
colony managers frequently removed infants for hand-rearing in the mistaken
belief that mothers would ovulate and become pregnant sooner, whereas field
studies showed that some tamarins can breed twice a year. With a five- to six-
month gestation and two- to three-month nursing phase, this means that
mothers can become pregnant while they are still nursing. Data from other
colonies found a very low rate of surviving infants in parents who had been
removed from their own families at an early age (Tardiff et al., 1984; Johnson
et al., 1991). Tamarins, both males and females, need some sort of experience
in looking after someone else’s infants before they can be successful breeders.
We housed our families in multigenerational groups and never removed an
animal until it had cared for at least two other sets of infants, and we have
not found the high rate of infant mortality that others have reported.
222 ANIMAL BEHAVIOR

We do not know exactly what type of experience is critical, but it may

involve becoming accustomed to tolerating the carrying of a squirming infant
on one’s back. There appear to be progressive stages in how helpers engage in
infant care, with monkeys just one litter older than the current infants not
being allowed to interact with infants until the infants are four to six weeks
old, and then only for brief periods of time (Achenbach & Snowdon, 1998).
Only the animals at least two litters older than the current infants are allowed
to help carry infants from the first days of life.
Despite the experience gained in the family we still found a relatively high
rate of infant mortality in first-time parents. Because we had found that infant
survival was related to group size, one possible source of difficulty is that infant
survival requires multiple caregivers. But we also found that first-time mothers
often had difficulty orienting infants to their nipples, with infants often held
upside down. Even with a great deal of experience in carrying someone else’s
infants, a first-time mother will never have had experience in nursing an
infant. Another variable that might account for the low survival rate of infants
born to first-time parents is that first-time mothers were much less likely to let
other group members carry their infants. In field observations, we observed
first-time mothers carrying infants an average of 90 percent of observation
time in the first two weeks, whereas experienced mothers carried infants
50 percent of observations in the first week and much less in successive weeks
(Savage, Snowdon, et al., 1996).
To humans, male parental care is perhaps the most interesting feature of
tamarin life. Why do males show so much interest in infants? What mecha-
nisms help them to become skilled and involved parents, and can we translate
that to our own families? Males play two important roles in direct care of
infants—they carry infants significantly more than mothers do (Zahed et al.,
2010) and at the time of weaning fathers and male helpers share solid food
with infants, distracting them from trying to nurse (Joyce & Snowdon, 2007).
We know of two major hormonal changes that affect mammalian mothers.
The hormone prolactin is important for preparing the mammary glands to
produce milk, and oxytocin is involved in uterine contractions at delivery
and in maintaining the milk let-down reflex during suckling. These hormones
have also been shown to have an important role in the behavioral response of
mothers to newborn infants. Could these same hormones be active in males as
well? Considerable research has been done on monogamous ring-neck doves
(Streptopelia risoria) where both parents secrete a crop milk (based on epi-
thelial cells sloughed from the crop membrane) that feeds young squab. The
production of crop milk is directly influenced by prolactin in both sexes
(Buntin, 1996). Do male mammals that are active in infant care also produce
COTTON - TOP TAMARINS 223

prolactin? If they do, is it after the infants are born, or can males anticipate a
birth by increasing prolactin before birth?
Studies on common marmosets have shown that serum prolactin levels are
higher in both fathers and male helpers while they are carrying infants than
when they are not directly carrying infants (Dixson & George, 1982; Mota
et al., 2006). Other studies have injected a prolactin inhibitor and have found
that fathers were unaffected (Almond et al., 2006), although juvenile males
were less inclined to retrieve infants when injected with a blocker (Roberts
et al., 2001). Our urinary measures were not sensitive enough to differentiate
between when an animal was carrying or not carrying an infant, but we found
that fathers had prolactin levels that were as high as those of nursing mothers
(Ziegler et al., 1996). Surprisingly, we found that levels of prolactin during
the month prior to birth were the same as the levels when fathers were active
in infant care. Does this mean that fathers can somehow anticipate the timing
of infant births?
To find out if fathers can hormonally anticipate the birth of infants, we fol-
lowed a group of first-time fathers and experienced fathers throughout their
mates’ pregnancy, taking hormonal samples and making behavioral observa-
tions several times a week throughout the sixth month of pregnancy (Ziegler
et al., 2004). To our initial surprise, we found that experienced fathers showed
many hormonal changes during the second half of pregnancy. There was a sig-
nificant rise in testosterone, estrogen, dihydrotestosterone, and prolactin,
mainly in the last month or two of pregnancy. There was also a significant rise
in cortisol just at the fourth month. These changes were not seen in first-time
fathers except that their prolactin levels were increased significantly in the last
month of pregnancy.
These results lead to further questions. Why do we see so many “female-
typical” hormones in males, and how do males “know” when their mates are
pregnant? Women typically tell their partners when they are pregnant, but
how can a female tamarin communicate to her mate that she is pregnant?
The female-typical hormones are easiest to explain. Testosterone can be
reduced by enzymes into either estrogen or dihydrotestosterone. Paradoxically
it is the effect of reduction of testosterone to estrogen that leads to masculine-
typical behavior in mammals, but estrogen also works to stimulate production
of prolactin and oxytocin. Thus increased levels of testosterone may be neces-
sary to produce estrogen to stimulate prolactin production in males.
One clue to how males might “know” when their mates are pregnant comes
from the peak in cortisol in the fourth month. We found that all pregnant
females showed an increase in excretion of glucocorticoid hormones in the
fourth month of pregnancy. Within a week of this increased excretion, all
224 ANIMAL BEHAVIOR

experienced fathers displayed an increase in glucocorticoids, and following this

increase they displayed the cascade of other hormonal changes. But why do
the pregnant females produce such high glucocorticoid levels at the fourth
month? It is at this stage that the fetal adrenal gland becomes active and starts
secreting glucocorticoids, but an excess of glucocorticoids is harmful to the
mother, so she excretes the excess in her urine. In a way we can say that the
developing fetus is communicating through its mother to prepare the father
for parenting within a couple of months (Ziegler et al., 2004).
The fact that experienced fathers but not first-time fathers respond to the
changes in female excretions suggests that some learning process must be
involved to be able to detect the changes because first-time mothers, like expe-
rienced mothers, produced the same signal. An experienced male paired with a
first-time mother still showed a hormonal response during mid-pregnancy
(Almond et al., 2008). But how is it that first-time fathers are able to have high
prolactin levels a month before their infants are born? Finally, our behavioral
observations became important. We found no changes in behavioral interac-
tions between experienced pairs during pregnancy. Indeed, most of their time
was taken up by interacting with other family members rather than with each
other. But first-time parents showed a significant increase in affiliative behav-
ior in the month before birth (Ziegler et al., 2004). We were unable to deter-
mine how this change was initiated, but we think that increased affiliative
behavior might stimulate prolactin production. In fact, we have found that
prolactin levels in both male and female tamarins are directly correlated with
how much affiliative behavior each animal receives (Snowdon & Ziegler, in
preparation).
Hormones are not the only changes that occur in males during their mates’
pregnancies. We found that males increase body weight throughout preg-
nancy, perhaps in anticipation of the weight loss that will come during infant
care (Ziegler et al., 2006). Taken together, these hormonal and weight
changes are similar to the “couvade syndrome” experienced in many human
cultures, where the father takes on the symptoms and behavior of the mother
during pregnancy and childbirth. Perhaps undergoing couvade prepares men
for acting as fathers.
There is evidence that brains of female mammals are changed when they
become mothers (Kinsley & Lambert, 2006). Do fathers also undergo changes
in their brains? We think so based on several experiments with the closely
related common marmoset. We were interested in how male marmosets
would react to the odors of females, and, using noninvasive brain imaging
technology, we found in single and paired males who were not yet fathers an
increase in neuronal activity in brain areas associated with sexual arousal when
COTTON - TOP TAMARINS 225

males were presented with odors of novel ovulating females (Ferris et al.,
2001). We evaluated the behavioral and hormonal responses of males to these
same odors and found that males showed an increase in exploration and in
erections when exposed to the odors of novel ovulating females; and when
we took a blood sample 30 minutes after the first presentation of the odor,
we found a significant increase in testosterone (Ziegler et al., 2005). However,
there was a lot of variation, and some males showed no interest in the odors
and no change in testosterone. Every one of these nonresponding males was
a father with offspring in his family. Thus, being a father seems to increase
mate fidelity (or at least inhibit any interest in a novel female).
We also asked whether fathers were more sensitive to infant cues than non-
fathers and tested this with both auditory and olfactory stimuli. We put
together two cages, with a bridge separating the two cages; in one cage we
placed a live infant, and in the other cage we placed a male, either a father or
a nonfather. Fathers responded by crossing the bridge and attempting to
retrieve infants more often and with a shorter latency than nonfathers. It did
not seem to matter to fathers whether the infants were their own or not. They
responded just as well to their own infant as to an unfamiliar infant. To isolate
the cues involved we repeated the study with an MP3 player producing infant
cries and again found that fathers responded significantly more often than
nonfathers and responded equally to cries of their own and to unfamiliar
infants (Zahed et al., 2008).
Subsequently we tested fathers with the odors of their own versus other
infants and took a blood sample 20 minutes after initial exposure to the odors.
Fathers showed a rapid hormonal response to odors of their own infants but
not to odors of unfamiliar infants. Fathers showed a significant decrease in testos-
terone and a significant increase in estrogen (perhaps converted from testoster-
one). This hormonal response to infant odors was seen only when infants were
still dependent on being carried. When fathers were presented with their infants’
odors at a later age, the fathers were unresponsive (Ziegler et al., 2011). There
appears to be clear evidence that becoming a father makes males more faithful
and less interested in other females and that fathers are much more sensitive to
infant cues than nonfathers, suggesting that fatherhood does change male brains.
Although from a mother’s perspective it may seem attractive to have other
group members do most of the infant care, there are a few downsides. One
of these is with whom the infant develops an attachment. Harry Harlow found
that rhesus macaque (Macaca mulatta) infants were more attached to a furry
surrogate mother even though all the milk was provided from a wire-frame
surrogate. When infants were frightened or in a novel environment, they
clung to the furry surrogate. We wanted to evaluate attachment in tamarins
226 ANIMAL BEHAVIOR

in a more natural way, and so we kept track throughout infancy of how much
time each family member carried an infant and how often they shared food with
the infant. Then, when the infants were old enough to move on their own and
were foraging for their own food, we presented a frightening stimulus (a person
dressed to catch an animal for a veterinary exam) and kept track of whom the
infants ran to for comfort (Kostan & Snowdon, 2002). In every single case the
young tamarins ran to the individual that had carried them the most and had
shared food with them the most, and in no case was it the mother!
Do infants benefit from cooperative care in anything more than increased
survival? In species where mothers do almost all of the infant care (macaques,
baboons, and vervet monkeys, Chlorocebus pygerythrus) scientists have
described distinct styles of mothering that can be described on two indepen-
dent dimensions (acceptance versus rejection, and laissez faire versus restric-
tive; Fairbanks, 1996). A daughter growing up experiencing a specific
maternal style is highly likely to use the same style with her own infants.
But what happens when an infant has multiple caregivers that are all active
to varying degrees? We studied the type and amount of care that infants
received in 30 litters where some of the parents were highly experienced and
others were parenting for the first time; the number of group members varied
between two and seven animals. We observed lots of variation in what each
family member did and in how it parented, but when we analyzed the data
from the perspective of what infants received, we found no differences as a
function of either experience or group size. All infants received about the same
amount and quality of care (Washabaugh et al., 2002). Thus, one important
benefit of cooperative parenting may be to reduce the variance in caretaking,
with different caretakers buffering the extreme effects of others. Maternal style
appears not to influence infant outcomes in cooperatively breeding monkeys.

How Are Helpers Recruited and Maintained?

Why are helpers necessary? I have mentioned data from both the field and
captivity that infant survival does not reach 100 percent until group size is five
or greater. What could account for the need for helpers even in captivity? I
noted earlier that tamarins typically give birth to twins that weigh 20 percent
of the mother’s weight at birth. With a postpartum ovulation that usually results
in conception, mothers are simultaneously lactating and pregnant with the next
litter. So someone must carry around the increasingly heavy infants all day long.
This is more than a single mother can do, and so helpers play a necessary role.
In the field tamarins have a travel path of about 2.4 km a day, climbing up
and down trees, so there is considerable energy expended just in foraging
throughout the day. But what explains the need for helpers in captivity?
COTTON - TOP TAMARINS 227

I was talking about my work at Indiana University, and Jeffrey Alberts, a lead-
ing researcher on thermoregulation, asked if it was typical for infants to be car-
ried on the back of adults. After all, most terrestrial primates carry infants on
their chests. I suggested that riding on the back might be best if animals are
running up and down tree trunks all day, but Alberts suggested another
hypothesis. Brown fat, which generates heat, is located between our shoulders,
and infants riding between the shoulders may be using the brown fat of the
one carrying them as a sort of radiator for keeping warm. In support of this
idea are findings that males lose significant amounts of weight during the
period of intensive infant care (up to 10 percent in eight weeks in fathers with-
out additional helpers; Sanchez et al., 1999; Achenbach & Snowdon, 2002).
Thus, infant thermoregulation may be an important function of parental
behavior that requires shared help from others.
But how do parents recruit helpers? There are several issues involved. First,
it would be important to limit reproduction to just one female so that the re-
sources for infant care are not diluted among litters from multiple females.
Second, parents must prevent helpers from joining other groups or leaving
to reproduce elsewhere and instead remain to help the parents. Are there spe-
cific reward mechanisms that parents can use to retain helpers?
One of our first hormonal studies looked at reproductive function in female
tamarins. In mothers we could determine ovulation through a rise in estrogen
levels and the appearance of a spike in luteinizing hormone (Ziegler, Bridson,
et al., 1987). As noted earlier we did not observe many ovulations because
females typically ovulated within two to four weeks of giving birth and often
became pregnant on this first ovulation. However, when we looked at hormo-
nal patterns of daughters, we saw no evidence of ovulation. Estrogen levels
were very low, and there was no sign of luteinizing hormone activity (French
et al., 1984; Ziegler, Savage, et al., 1987). Basically, daughters living at home
with their mothers were not ovulating. This lack of ovulation was not simply
due to maturation as we observed daughters living in family groups who were
two or three years older than some reproducing females, and yet the daughters
were not ovulating.
What can cause this failure to ovulate in females living with a reproductive
female? One hypothesis supported by data from other species is that there is an
increased level of stress in subordinate females that keeps them from reproduc-
ing. However, we rarely noted any signs of direct antagonism between moth-
ers and other females, and when we measured cortisol levels as an index of
stress, we found that these nonovulating females had the lowest cortisol levels
of any group member (Ziegler et al., 1995). Thus, stress cannot be invoked as
a mechanism of reproductive failure.
228 ANIMAL BEHAVIOR

We hypothesized that cues from the mother might inhibit ovulation in

others, and because reproductive females have better developed scent glands
and scent mark more than any other group members (French & Cleveland,
1984), we tested the idea that scent marks could inhibit ovulation. We paired
several nonovulating females with new males, and for half of the females we
transferred each day boards that the reproductive female had scent marked
the previous day; for the other half of the females we placed an unmarked
board in their cage each morning. The results were striking. Each of the con-
trol females ovulated (in as little as eight days), and all become pregnant within
the 60 days of the transfer experiment. None of the females who received scent
marks from a familiar reproductive female even ovulated in the 60 days of
scent transfer (Savage et al., 1988). Thus, information in the scent secretions
from a breeding female was sufficient to maintain the inhibition of ovulation
even when no other cues were present.
But when a female leaves the presence of a breeding female she is also meet-
ing a novel male. Does a novel male play any role in stimulating reproductive
function in female tamarins? We removed some additional nonovulating
females from any sort of contact with the breeding female and housed them
with a familiar male—their father or brother—and found no signs of ovula-
tion (Widowski et al., 1990). Thus the cues from the breeding female are
not sufficient to keep a female from ovulating. When these same females were
then moved to a cage adjacent to novel male where they could see, hear, and
smell but not have direct physical contact with a novel male, they ovulated
quickly (Widowski et al., 1992). Thus in order for ovulation to begin a female
must be removed from olfactory cues from a familiar reproductive female and
also be stimulated by a novel male. We do not know what specific aspects of a
novel male are important, but this does provide a nice mechanism to avoid
inbreeding.
In our field studies in Colombia, we monitored pregnancies regularly, and
in almost every group and every year we found only one female pregnant.
However, in the two cases where we found two females pregnant in the same
group at the same time, a new male had moved into the group prior to concep-
tion, supporting the mechanism that we observed in captivity (Savage,
Giraldo, et al., 1996). No more than one set of surviving infants was observed
at the time of birth, suggesting that a group could support only one set of
infants. In the related common marmoset there is often more than one female
ovulating in a captive group, and in a field study we found that two females
were often pregnant at the same time, but one female would harass or kill
the infants of the other pregnant female, leaving a single set of infants to be
reared by the group (Lazaro-Perea et al., 2000).
COTTON - TOP TAMARINS 229

What about reproductive abilities in males? We could find no evidence of

social or hormonal mechanisms of reproductive inhibition in males. Males,
whether living with an adult breeding male or not, all demonstrated hormonal
maturation at a similar age, and all displayed rates of erection and sexual
behavior as often as the breeding male (Ginther et al., 2001, 2002). The
notable exception was that these males typically mounted other males in the
groups as well as nonovulating females. On occasion they would attempt to
mount their mothers, but rather than showing aggression, mothers simply
moved to a position where a successful mount would be impossible, and after-
wards both the breeding male and female would groom the young male
(Ginther & Snowdon, in preparation). So tolerance of the sexual activities of
helper males appears to be necessary. Once helper males have mates of their
own, the homosexual activities seen when they are helpers disappear.
We know that grooming plays an important role in primate social relation-
ships, and I noted earlier how grooming plays an important role in pair rela-
tionships. Generally grooming is thought to be directed by subordinates
toward dominants, and there is considerable evidence that the one being
groomed gains several physiological benefits from grooming: increased levels
of brain opioids (Keverne et al., 1989), reduced heart rate, and reduced levels
of the stress hormone cortisol (Shutt et al., 2007). In a field study of common
marmosets, we found that parents directed significantly more grooming to
helpers than helpers directed toward parents, suggesting that parents may be
rewarding helpers by increased grooming (Lazaro-Perea et al., 2004).
We pursued the idea of grooming as a reward by looking closely at patterns
of grooming in captive cotton-top tamarins. We focused our attention on
males because we found that male helpers are much more active in infant care
than are female helpers. We had good records on how much effort each male
helper had invested in caring for the previous litter of infants when we
observed them in the last half of the reproductive female’s pregnancy. We kept
track of grooming by both the breeding female and the breeding male. We
found a significant linear relationship between grooming time by parents
and how much work a helper male had done with the previous litter. Mothers
groomed most the males that had spent the most effort carrying infants previ-
ously, but fathers groomed most the males that had provided the least effort
with the previous litter (Ginther & Snowdon, 2009). What could account
for these very different responses between sexes? We found that time spent
carrying infants by mothers was reduced significantly with only one additional
helper, whereas carrying effort by fathers was reduced further with each suc-
cessive helper (Zahed et al., 2010). Thus it might benefit a mother to reward
the helper that had done the most care of the previous litter whereas a father
230 ANIMAL BEHAVIOR

may benefit most by recruiting additional helpers to reduce his workload with
the next litter.
So far I have portrayed tamarin family life as harmonious, with every indi-
vidual working for the benefit of the breeding pair. There appears to be a sys-
tem of tolerance of the sexual activities of male helpers and a reward system,
with grooming being used strategically to reward or recruit helpers. But there
is also a dark side to tamarin (and marmoset) life. Infrequently we observed
the onset of severe fights between group members. The targets of the fights
were almost always within sexes, with males fighting with other males and
females fighting with other females. In all cases involving males, it was typi-
cally an older male helper who initiated aggression toward a younger male sib-
ling, one that the older male had helped to carry when it was an infant. Among
females both sister-sister and mother-daughter conflicts were seen. In the case
of mother-daughter fights they occurred when the daughter started engaging
in scent-marking behavior and had measurable levels of estrogen (although
there was no evidence of the cyclic pattern typical of ovulation; Snowdon &
Pickhard, 1999). I need to emphasize that these fights were rare and usually
occurred when captive groups were at the maximum group size seen in the
field (7–8 animals). We quickly learned to remove (disperse) animals when
group size approached this critical value.

Why Should Helpers Help?

We have already hinted at a couple of reasons why it might benefit helpers
to help care for infants other than their own. There are several other possible
explanations. Some have argued that the most parsimonious single explana-
tion for both male parental care and the presence of helpers is as a courtship
strategy (Dunbar, 1995; Smuts & Gubernick, 1992). While this may be one
plausible explanation, there are several others, and one should not expect that
a single adaptive explanation should apply throughout the entire life history of
an animal. Let us explore some other options. First, many of the helpers (and
all of those in our captive colony) were related to the infants for which they
provided care. Thus, kin selection may play an important role in leading help-
ers to engage in infant care. Second, from our field studies with tamarins in
Colombia and marmosets in Brazil, we know that breeding opportunities are
rare, and so helpers may benefit most from helping their kin but being ready
to take advantage of any possible chance to breed in another group. Indeed,
we found that common marmosets had almost daily encounters with other
groups that appeared to be territorial encounters. But we also found a male
from one group slipping away with a female from another group for a quick
copulation (Lazaro-Perea, 2001). These encounters never resulted in
COTTON - TOP TAMARINS 231

pregnancy until after the death of the breeding female, in which case new
groups formed over the next few months with the females from a family join-
ing some adjacent males and the males forming a new breeding pair with adja-
cent females (Lazaro-Perea et al., 2000). Third, an animal may join a new
group that may have some future potential for breeding, but in return for
being a member of the group, the immigrant may have to “pay its dues” by
helping with infant care. This could be especially important if the immigrant
was the last offspring in its natal group. In this case it would have to join a
new group and help with care of unrelated animals simply to acquire infant-
care skills. Even joining a different group may not lead to a chance to repro-
duce. In Colombia we found both males and females dispersed equally from
family groups, but despite being able to monitor several dispersing animals
over several years we never found one that had successfully bred (Savage,
Giraldo, et al., 1996). Throughout an individual’s life the adaptive value of
helping care for other infants may change—initially helping to care for others
is likely to benefit kin and, through the need to learn skills, benefits individual
fitness as well. As an animal gets older, in the absence of open territories or
breeding opportunities, continuing to serve as a helper may be the best option.
If one moves to a new group then becoming active in infant care may be
important to fit in with the group, and, only secondarily, infant care may serve
as a courtship strategy, impressing a potential mate.

How Do Tamarins Communicate with Each Other?

Because tamarins live in dense vegetation where visibility is poor at a dis-
tance, they have relatively few visual displays, but they have a very large and
complex vocal repertoire. We have described more than 35 different types of
calls and call sequences (Cleveland & Snowdon, 1982; Campbell & Snowdon,
2007). There is a set of short, frequency-modulated calls that sound like
chirps. Through careful listening, looking at sound spectrograms, and close
behavioral observations, we could distinguish eight different types of chirps
that were used each in a specific situation. We did playback studies and found
that tamarins showed a behavioral discrimination between the most similar
call types (Bauers & Snowdon, 1990), and, when we created contexts in which
we expected each chirp type to appear, we found that adults, but not infants,
gave the call type that we predicted should be produced (Castro & Snowdon,
2000).
Many calls are produced in sequences of several notes, and these sequences
appear to have a rudimentary grammar (syntax), with notes appearing in a
predictable sequence. Tamarins appear able to combine call types into sequen-
ces as a function of context. Thus, when an alarm call has been given, animals
232 ANIMAL BEHAVIOR

take cover and freeze, but after this a few individuals will combine an alarm
call with a chirp that is usually given promoting contact within groups. This
sequence appears to act as an “all-clear” signal, and animals resume activity
after this sequence.
We attempted to simulate intergroup encounters in our colony by opening
doors between adjacent rooms, which created the appearance of a strange
group approaching. We found one type of call given primarily by males and
another type by females in the early stages of these vocal encounters, but when
vocal exchanges reached their peak, both sexes produced combined calls that
had the components of the calls given by each sex early on (McConnell &
Snowdon, 1986). However, the clear sex differences in calling that we
observed in the early study were completely reversed when we replicated the
study 20 years later (Scott et al., 2006). One should be cautious in drawing
conclusions about sex differences in a species that shows few differences
between sexes in physical or behavioral traits.
There has been great interest in whether species other than humans can
communicate about objects or events in their environment rather than simply
communicate about emotional state. We found that two of the chirp types
given by tamarins were used only in the context of feeding (Elowson et al.,
1991). One chirp was given as animals approached food, and the other type
was given after the animal had moved away from the food dish with a piece
of food. However, emotional variables are still involved in this signal. We
evaluated the preferences of each individual for six different types of food
and then presented each individual with each food type and recorded the
number of chirps it gave. There was a strong correlation with each individual’s
preference and the number of chirps it gave. Thus tamarins appear to be com-
municating honestly about their preferences.
Many researchers have noted that primate vocal communication often
appears to be highly stereotyped, with little evidence of modification of the
types of calls primates make, although there are clear developmental processes
in when and how calls are used. Although adults use food calls only with
respect to edible objects, young tamarins appear to overgeneralize and produce
food calls to any small object that they can manipulate. Furthermore, there
calls were not as well formed as those of adults, and young tamarins often gave
many other calls types when they were feeding—calls we never heard from
feeding adults (Roush & Snowdon, 1994). We thought this might simply
reflect a maturation of vocal control, but to our surprise, young tamarins con-
tinued to use these imperfect food calls and other vocalizations long past
puberty. This suggested that the “immature” form of food calling might be
related to the role of being helpers rather than breeding adults. To test this,
COTTON - TOP TAMARINS 233

we recorded food calls from animals living in a subordinate role within a fam-
ily and then from the same individuals for several weeks after they had been
removed from the family and paired with a mate of their own. Within one
to two weeks of pairing the tamarins stopped using the other types of calls
and used only food chirps, and by four to five weeks all displayed adult call
structure and usage (Roush & Snowdon, 1999). It seems likely that the tama-
rins were inhibited from using adult-like calls while living as helpers in their
families. We also found that young tamarins rarely produced adult-like chirps
when we created specific contexts to elicit calls in adults. Many young tamarins
would occasionally produce an appropriate chirp, but if they produced the
appropriate chirp on one occasion, they would be unlikely to produce it consis-
tently thereafter (Castro & Snowdon, 2000). This is not a formal, experimental
test of vocal learning, but the results suggest that young tamarins are not born
with the ability to produce all of their calls and that social interactions with
others may influence when and how they call, suggesting more developmental
flexibility in communication than previously thought (Snowdon, 2009).
We have found flexibility in the contexts in which tamarins use alarm calls.
We presented tamarins with a highly preferred food (tuna) that we had adul-
terated with white pepper, making it less palatable (Snowdon & Boe, 2002).
The tamarins that first tasted the adulterated tuna gave alarm calls, using these
calls in a completely novel context. Other group members paid attention and
rarely sampled the pepper-laced tuna. The animal sampling the tuna also made
facial expressions of disgust and produced food calls at a very low rate, and all
these signals taken together may have made for an effective communication
for others to avoid the food.
We were also interested in recording mobbing calls for a planned study on
response to predators and found that captive tamarins responded with mob-
bing vocalizations to events that were relevant to the ecology of the lab—
cage-cleaning equipment or someone dressed in veterinary garb (Campbell
& Snowdon, 2007), but they showed no mobbing calls and very low levels
of arousal when we placed a live boa constrictor in their cage (Hayes & Snow-
don, 1990; Campbell & Snowdon, 2009). Furthermore, although we tried to
use mobbing calls as a way to condition fear of snakes in captive tamarins, we
had no success (Campbell & Snowdon, 2009). In addition, captive tamarins
did not distinguish between playbacks of vocalizations of natural predators
versus nonpredators (Friant et al., 2008). This lack of response to both live
predators and calls of predators suggests that considerable effort would be
needed to prepare captive-born animals for reintroduction to the wild.
Many of our results concerning how a tamarin produces calls, how it can
adapt calls to novel contexts and to the ecology of captivity, and that it fails
234 ANIMAL BEHAVIOR

to respond with fear to natural predators all suggest that a tamarin’s ability to
communicate is the result of learning through experience. This is yet another
example (along with parental care) of the important role of learning in the
development of tamarin behavior.
Tamarins can distinguish individuals through the structure of the calls they
make. We and others have found that the “long call” (named because it is a
series of notes that form the longest call in their repertoire) has specific features
that allow tamarins to distinguish familiar animals from strangers (Snowdon
et al., 1983), but do animals remember family members after they disperse?
To test this we played back long calls recorded from family members from
whom tamarins had been separated for periods from six months to five and a
half years, and we also played the calls from the animal’s current mate and
from a completely unrelated and unfamiliar animal. There was a clearly differ-
ent response, with high degrees of alerting and arousal, to the calls of an unfa-
miliar tamarin and very low response to the calls of a familiar mate. The
response to playbacks of calls of former family members did not differ from
the response to a familiar mate, regardless of the duration of separation, sug-
gesting a long-term memory (up to at least five and a half years) for calls of
familiar animals (Matthews & Snowdon, 2011). This could be an important
mechanism for avoiding inbreeding, especially for tamarins, where both sexes
can disperse from their natal groups (Savage, Giraldo, et al., 1996).
A few years ago I was approached by David Teie, a professional cellist and
composer, who had some ideas about how music influences our emotions.
Because we all grow up with lots of experience with music and have developed
clear preferences, we are not good subjects for testing his ideas, but our tama-
rins had not been exposed to human music before. David hypothesized that
the tempo of a resting heart rate would be calming whereas a much faster
tempo could be arousing. Short, quick notes (staccato) would be arousing
and long notes calming. Harmony indicates affiliation, but dissonance and
noise indicate fear or aggression. But we faced a problem. Josh McDermott
and Marc Hauser (2007) had tested both cotton-top tamarins and common
marmosets, giving them a choice between Mozart and German heavy metal.
The monkeys preferred Mozart, but when they were offered a choice between
Mozart and silence, the monkeys preferred silence. Thus, McDermott and
Hauser concluded that monkeys are indifferent to music. But tamarins have
a much faster heart rate than humans, and their vocal range is three octaves
higher than the human voice. Perhaps the test was unfair to tamarins.
David Teie composed several musical pieces in the frequency range and
tempo of tamarins, with two pieces having musical structures that should be
calming and two having arousing and fearful structures. We played these
COTTON - TOP TAMARINS 235

pieces to the tamarins along with two calming and two arousing human musi-
cal pieces. The tamarins were generally indifferent to the human music, but
they reduced activity and foraged more after hearing tamarin-calming music,
and they showed more activity, more anxious behavior, and more huddling
with one another after hearing the tamarin-arousing music (Snowdon & Teie,
2010). Thus tamarins do respond to music if it is composed in their frequency
range and with appropriate tempos. This study suggests the emotional compo-
nents of music may have a long evolutionary history and emphasizes the emo-
tional impact that music has for us.
Cotton-top tamarins also make use of odors in communication. There is a
clear sexual dimorphism, with females scent marking much more often than
males (French & Cleveland, 1984) and reproductively inhabited females
marking much less often than reproductive females (French & Cleveland,
1984). The scent glands become much larger and secrete more fluid when a
female becomes a breeding adult (French et al., 1984). The two main glands
are in the anogenital region and the suprapubic area, with anogenital marking
appearing to provide information about reproductive status (Ziegler et al.,
1993) and suprapubic marking being used in territorial conflicts (French &
Snowdon, 1981; McConnell & Snowdon, 1986).

What Cognitive Skills Do Tamarins Have?

Although tamarins and marmosets are evolutionarily primitive among pri-
mates, they have displayed remarkable cognitive abilities, especially with
respect to social aspects of cognition. I will describe two studies on social
learning. Although many other species are able to learn to avoid noxious
foods from observing others, there was little evidence in nonhuman primates.
A study in Italy of capuchin monkeys (Cebus apella) presented animals with
mozzarella cheese to which white pepper had been added. Monkeys quickly
learned to avoid the peppered cheese, but there was no evidence of social
learning (Visalberghi & Addessi, 2000). We thought that cooperatively
breeding tamarins might be more successful, and so we presented a favorite
food, tuna, with white pepper added. We presented this adulterated tuna to
44 monkeys in eight different social groups. Only a third of the animals ever
sampled the food, and, of those, very few sampled tuna on subsequent trials.
When we began to serve tuna without any pepper added, there were several
individuals who had never tasted the peppered tuna who continued to avoid
tuna for more than a year (Snowdon & Boe, 2002). As I noted earlier, the
animals tasting the tuna gave alarm calls and made facial expression of dis-
gust, something capuchin monkeys never did (E. Addessi, personal
communication).
236 ANIMAL BEHAVIOR

In another study we presented tamarins with two identical foraging frames

that each had five different-colored cylinders. All were baited with food, but
four were locked and only one could be opened. We trained one animal in
each pair alone until it had learned to respond to the correct cylinder, and then
we introduced the mate. Our goal was to see if the naïve mate could learn from
the trained partner. All naïve animals rapidly learned the novel motor skill of
opening the cylinder, and all learned the correct cylinder within a very few tri-
als (two to six), much more quickly than the original animals. However, the
naïve animals learned without any food reward because they followed closely
behind the mate, which always ate the food first. When we tested the animals
18 months later, all of them still remembered the task and found the correct
cylinder with few mistakes (Moscovice & Snowdon, 2006). This rapid social
learning without any reward is extraordinary among primates.
In the initial study spatial location and the color of the cylinder were both con-
sistent. In a follow-up study we were interested in whether tamarins would learn
spatial or color cues more quickly. We trained half of the animals to a consistent
spatial location regardless of color and the other half to a consistent color regardless
of location, and then, after animals showed good learning of each cue type, we
reversed them so that those trained with spatial cues now had to respond to color
and those trained to color had to respond to spatial location. Tamarins had a
much more difficult time learning to respond to a color cue than a spatial cue,
and animals initially trained on color learned the spatial cue significantly faster
after reward cues were reversed (Gaudio & Snowdon, 2008). In the wild, tamarins
would use spatial location as the primary cue to find food resources, and once the
food was found they might use color to evaluate ripeness, so the more rapid learn-
ing about spatial cues in captivity has relevance to behavior in the wild.
Teaching is considered to be a high-level cognitive skill, and reports of
teaching behavior in great apes have been rare. However, tamarins share food
with offspring at the time of weaning, and this food sharing may have elements
of teaching involved. When food is offered to young, adults give much more
rapid and intense versions of the food calls that they give when foraging on
their own, and infants are rarely able to get food from adults unless these calls
are given. We have found a direct correlation between the onset of food sharing
by adults and the onset of when infants first forage successfully on their own
(Roush & Snowdon, 2001; Joyce & Snowdon, 2007). By the age of five
months young tamarins are completely weaned and food sharing is absent.
We then created a novel foraging apparatus that had two opaque tubes with
a small cup with food in it suspended in the tube. We trained one parent from
each family group in one of two methods for obtaining the food. One parent
learned to go above the tubes and pull up the string hand over hand, and the
COTTON - TOP TAMARINS 237

other parent was trained to climb on a branch below the tube and reach up to
pull the food down. We then tested juveniles (seven to nine months old), one
with each parent, and found that the parents began giving food-sharing calls
again and would share food with the juveniles. However, as soon as a juvenile
successfully learned to obtain the food by itself, the parent stopped calling and
sharing food (Humle & Snowdon, 2008). When we tested naïve adults on the
same task in the company of a mate who knew how to solve the task, there was
no evidence of calling or food sharing (Dillis et al., 2010). These results taken
together suggest a rudimentary form of teaching, with parents providing assis-
tance until a young tamarin masters a task and then withholding support
thereafter. This is not an artifact of captivity because Rapaport and Ruiz
Miranda (2002) have reported similar behavior in wild lion tamarins (Leonti-
pithecus rosalia). When we studied mother-infant interactions in wild chim-
panzees foraging for biting ants in Africa, we saw absolutely no evidence of
any help by mothers. Young chimpanzees could observe their mothers, but
mothers never assisted them in making tools or correcting their foraging
behavior (Humle et al., 2009). The contrast between the teaching in tamarins
and the lack of any help by chimpanzee mothers, even with biting ants, is
quite striking.
So far I have shown that cooperatively breeding tamarins appear to outper-
form other species in terms of social learning and teaching, and one hypothesis
for this ability is that cooperative breeders must coordinate their behavior
closely and communicate frequently with one another. But do they actually
cooperate with one another in contexts other than infant care? Katherine
Cronin was an undergraduate student who became interested in this question,
and she developed a clever apparatus—a transparent box with two sliding
trays, with two holes in each tray. When the trays were pulled out from each
side of the box at the same time, food placed in the holes would drop to the
base of the apparatus where the tamarins could each retrieve a piece of food.
Animals were trained individually to learn that if they pulled the trays they
could obtain food. We then attached elastic bands to the trays so they would
have to be pulled simultaneously for anyone to obtain food and then tested a
pair of animals together. The tamarins quickly learned to coordinate their
behavior, and they had a high rate of success (>97%). Frequently one animal
would pull on its tray and wait until the other animal pulled. We then sepa-
rated the animals and found that the animal left alone with the apparatus
showed an interest in it but rarely attempted to pull on a tray, suggesting that
the animal realized the need for a partner to be successful (Cronin et al.,
2005). When Cronin returned for her Ph.D. she devised several variations
and found that if only one animal was rewarded in a session, both animals
238 ANIMAL BEHAVIOR

continued to pull at a high rate and that performance remained strong even
after several days elapsed between when one animal received rewards and the
other did (Cronin & Snowdon, 2008).
At the same time donation behavior was being studied in chimpanzees. An
animal could pull one of two trays—one with one piece of food for itself and
the other with two pieces of food, one for itself and one for a companion in
an adjacent cage. Even though there was no cost to pulling the tray with two
pieces of food in terms of what the actor would get, there was no evidence that
chimpanzees would voluntarily share food with another (Silk et al., 2005;
Vonk et al., 2008). We attempted the same experiment with tamarins, hoping
to find that tamarins would share, but we also failed to find any donation of
food either with no cost to the donor or when only the recipient could get
food (Cronin et al., 2009). Yet at the same time an independent group found
that common marmosets would donate food, even with no reward for the
donor and only the recipient benefitting (Burkart et al., 2007). We eventually
developed a test where one animal provided food continuously for itself and its
mate for five minutes and on another day the tray was blocked and no one
received a reward. We then reversed roles so that the prior recipient could pro-
vide food only for the former donor. Tamarins needed some time to adjust,
but within a couple of minutes they began pulling the tray to provide food
to their mate regardless of whether the mate had just rewarded them or not
(Cronin et al., 2010). So both tamarins and marmosets show true donation
behavior to others. Thus, cooperative breeding monkeys are truly cooperative.

CONCLUSION
Conservation
At the beginning of the chapter, I asserted that studying tamarins, even in
captivity, would have benefits for conservation. Let me now back up that
assertion. The noninvasive hormone-collection techniques we developed in
captivity allowed us to monitor hormones in wild tamarins, and we found a
strong seasonal breeding pattern with births linked to the onset of the rainy
season (Savage et al., 1997). However, our captive tamarins (and those that
had been confiscated and housed in outdoor cages in Colombia) were able to
breed year round. This suggests that availability of high-quality food, which
is linked to the rainy season, may be critical for reproduction in the wild. Sup-
porting this, we found that one year when the rains did not start on time, all
females lost their infants, but they became pregnant again quickly and gave
birth six months out of phase with the normal seasons. By the following year
they were back on a seasonal breeding pattern.
COTTON - TOP TAMARINS 239

We found that typically only one female bred at a time in the wild, with the
only exceptions occurring when a novel male joined a group prior to the date
of conception, just as predicted from our captive work (Savage, Giraldo, et al.,
1996). And we found close parallels between parental care in first-time and
experienced mothers and of group size and infant survival in both captivity
and the field. A recent survey to census the number of tamarins remaining in
Colombia used playbacks of long calls of other tamarins to elicit vocal
responses in wild tamarins, allowing them to be counted accurately (Savage
et al., 2010). This method was directly based on playback studies we did with
captive tamarins (Snowdon et al., 1983).
However, the most important influence on conservation has been to get
others excited about tamarins in Colombia and elsewhere. While she was
doing her fieldwork, Anne Savage involved the local community at Coloso
in several ways. First, they did not realize that cotton-top tamarins were
unique to their part of Colombia, and as Anne recruited journalists and TV
producers to visit and write about the tamarins for the national media, the
townspeople developed great pride in “their” monkeys.
Anne also involved school children. There were no Spanish biology text-
books that dealt with local wildlife, so she began teaching once a week in
11th-grade biology. Subsequently, she raised some money to buy radio trans-
mitters and receivers and trained the 11th-graders to track the animals that
were living in the backyards of the village after school each day. Wearing spe-
cial t-shirts and carrying radio-tracking equipment, the children provided an
enthusiastic and highly visible message about conservation every afternoon.
Anne has also worked to develop new economic resources to reduce the
pressures on cutting wood. She worked with a potter to develop a novel oven
that can cook food using corn stalks and other vegetation while preserving
trees. And several women are involved in making purses and carry bags from
leftover plastic bags. Anne has also worked closely with the Nature Conserv-
ancy, Disney’s Animal Kingdom, and the Colombian government to create a
reserve for tamarins. You can learn more about her work in Colombia at
www.proyectotiti.com.
We have also donated more than 300 of our tamarins to zoos and to under-
graduate colleges in North America so that the general public can learn more
about cotton-top tamarins and why they are interesting and important.
Many people ask about reintroduction as a conservation strategy. I think
that reintroduction has too many problems to be a viable strategy. First, I have
mentioned above that our captive-born tamarins have not shown any signs of
fear of natural predators and that we were unable to condition a fear of preda-
tors. Second, a long-term study of golden lion tamarins (Leontopithecus rosalia)
240 ANIMAL BEHAVIOR

reintroduced to Brazil found motor deficits even in animals that were in the
second generation of wild-born offspring to reintroduced parents (Stoinski
et al., 2003). Third, there are complex foraging skills that are needed: catch
insects, process fruit, and discriminate between toxic and nontoxic foods that
captive-born tamarins would have to learn before they could be released.
Finally, there must be enough suitable habitat to support additional tamarins
without affecting the local wild population. Because habitat has been steadily
decreasing over the last several decades, adding reintroduced tamarins to com-
pete with the limited resources available to the remaining wild tamarins simply
makes no sense. We will be in a better position to save cotton-top tamarins by
creating and funding suitable refuges for wild tamarins and by providing alter-
native economic resources so that humans can survive without having to make
further incursions on the remaining forest.

Understanding Human Behavior

I also asserted that studies of tamarins might also help us understand human
behavior. At the most basic level tamarins show us that behaviors that we casu-
ally think of as innate require learning and experience. Parenting skills and fear
of predators are not features that are “built in” to human and nonhuman pri-
mates, and we must acknowledge that many basic skills that are important to
survival must be acquired through experience. We have seen that males have
the ability to be outstanding parents, but they also require a strong pair bond
with their mate to provide confidence of paternity, and mothers must also allow
their partners to become active in parenting. Tamarins also illustrate an impor-
tant point that there may be adaptive value in deferring reproduction until
circumstances are good in terms of mate and resources. Indeed many of you
reading this will have already delayed reproduction to complete your education.
The vocal communication skills of tamarins and the role of developmental
experience in acquiring communication competence provide an important
antidote to the notion of a “language instinct.” Although each species has its
own unique communication system, social learning and social interactions
play an important role in development, suggesting that we cannot assume that
human language would develop automatically in a social vacuum. The paral-
lels between emotional communication and music suggest that music is not
uniquely human and that music may have emerged from the emotional
aspects of animal and human communication.
The apparent disconnect between social learning, cooperation, and teach-
ing in cooperatively breeding species and the comparative lack of these in
our closest ape ancestors emphasizes the importance of considering convergent
evolutionary processes. Some recent theories about human cognitive evolution
COTTON - TOP TAMARINS 241

have drawn on these differences to argue that humans have advanced on the
individual cognitive and problem-solving abilities of the apes while bringing
as well the social learning, cooperation, and teaching skills of cooperatively
breeding animals (Burkart et al., 2009; Hrdy, 2009). Thus studies of co-
operatively breeding primates may be leading us to a broader, more complete
understanding of our own cognitive evolution.
I hope that you can now see why cotton-top tamarins have fascinated me
for so many years and also see the wide range of studies that can be pursued
within a single species.

ACKNOWLEDGMENTS
I am grateful to the National Institute of Mental Health for long-term sup-
port of our captive research and the National Science Foundation for it sup-
port of field work in Colombia. I am also grateful to Toni Ziegler for her
long-term collaboration on captive research and Anne Savage for initiating
and continuing field research on cotton-top tamarins. I am also grateful to
the many people who have worked with me on this research—postdoctoral
fellows, graduate students, undergraduate students, and research staff. With-
out their dedication to the tamarins and the research, we would not have
accomplished as much as we have.

REFERENCES AND SUGGESTED READING

Achenbach, G. G. & C. T. Snowdon. (1998). Response to sibling birth in juvenile
cotton-top tamarins (Saguinus oedipus), Behaviour, 135, 845–862.
Achenbach, G. G. & C. T. Snowdon. (2002). Costs of caregiving: Weight loss in cap-
tive adult male cotton-top tamarins (Saguinus oedipus) following the birth of
infants. International Journal of Primatology, 23, 179–189.
Almond, R. E A., G. R. Brown, & E. B. Keverne. (2006). Suppression of prolactin
does not reduce infant care by parentally experienced common marmosets
(Callithrix jacchus). Hormones and Behavior, 49, 673–680.
Almond, R. E. A., T. E. Ziegler, & C. T. Snowdon. (2008). Changes in prolactin and
glucocorticoid levels in cotton-top tamarin fathers during their mate’s preg-
nancy: The effect of infants and paternal experience. American Journal of
Primatology, 70, 560–565.
Bauers, K. A. & C. T. Snowdon. (1990). Discrimination of chirp vocalizations in the
cotton-top tamarin. American Journal of Primatology, 21, 53–60.
Buntin, J. D. (1996). Neural and hormonal control of parental behavior in birds. In J.
S, Rosenblatt & C. T. Snowdon (eds.), Parental Care: Evolution, Mechanism
and Adaptive Significance (pp. 161–213). San Diego, CA: Academic Press.
242 ANIMAL BEHAVIOR

Burkart, J. M., E. Fehr, C. Efferson, & C. P. van Schaik. (2007). Other-regarding

preferences in a non-human primate: Common marmosets provision food
altruistically. Proceedings of the National Academy of Sciences, USA, 104,
19762–19766.
Burkart, J. M., S. B. Hrdy, & C. P. van Schaik. (2009). Cooperative breeding and
human cognitive evolution. Evolutionary Anthropology, 18, 175–186.
Burley, N. (1979). The evolution of concealed ovulation. American Naturalist, 114,
835–858.
Campbell, M. W. & C. T. Snowdon. (2007). Vocal response of captive-reared Sagui-
nus oedipus during mobbing. International Journal of Primatology, 28, 257–
270.
Campbell, M. W. & C. T. Snowdon. (2009). Can auditory playback condition
predator mobbing in captive-reared cotton-top tamarins? International Journal
of Primatology, 30, 93–102.
Carter, C. S. (1998). Neuroendocrine perspectives on social attachment and love. Psy-
choneuroendocrinology, 23, 779–818.
Castro, N. A. & C. T. Snowdon. (2000). Development of vocal responses in infant
cotton-top tamarins. Behaviour, 137, 629–646.
Cleveland, J. & C. T. Snowdon. (1982). The complex vocal repertoire of the adult
cotton-top tamarin (Saguinus oedipus). Zeitschrift für Tierpsychologie, 58, 231–
270.
Cronin, K. A., A. V. Kurian, & C. T. Snowdon. (2005). Cooperative problem solving
in a cooperatively-breeding primate, the cotton-top tamarin (Saguinus oedipus).
Animal Behaviour, 69, 133–142.
Cronin, K. A., K. K. Schroeder, E. D. Rothwell, J. A. Silk, & C. T. Snowdon. (2009).
Cooperatively breeding cottontop tamarins (Saguinus oedipus) do not preferen-
tially donate rewards to their long-term mates. Journal of Comparative Psychol-
ogy, 123, 231–241.
Cronin, K. A. K. K. Schroeder, & C. T. Snowdon. (2010). Prosocial behaviour
emerges independent of reciprocity in cottontop tamarins. Proceedings of the
Royal Society of London, B, 277, 3845–3851.
Cronin, K. A. & C. T. Snowdon. (2008). The effects of unequal reward distributions
on cooperative problem solving by cottontop tamarins (Saguinus oedipus).
Animal Behaviour, 75, 245–257.
Dillis, C., T. Humle, & C. T. Snowdon. (2010). Socially biased learning among adult
cottontop tamarins (Saguinus oedipus). American Journal of Primatology, 72,
287–295.
Dixson, A. F. & L. George. (1982). Prolactin and parental behavior in a male New
World primate. Nature, 299, 551–553.
Dunbar, R. I. M. (1995). The mating system of callitrichid primates II. Impact of
helpers. Animal Behaviour, 50, 1071–1089.
Elowson, A. M., P. L. Tannenbaum, & C. T. Snowdon. (1991). Food associated calls
correlate with food preferences in cotton-top tamarins. Animal Behaviour, 42,
931–937.
COTTON - TOP TAMARINS 243

Epple, G. (1978). Reproductive and social behaviors of marmosets with special refer-
ence to captive breeding. Primates in Medicine, 10, 50–62.
Fairbanks, L. A. (1996). Individual differences in maternal style: Causes and conse-
quences for mothers and offspring. In J. S. Rosenblatt & C. T. Snowdon
(eds.), Parental Care: Evolution, Mechanism and Adaptive Significance (pp.
579–611). San Diego, CA: Academic Press.
Ferris, C. F., C. T. Snowdon, J. A. King, T. Q. Duong, T. E. Ziegler, K. Ugurbil, R.
Ludwig, N. J. Schultz-Darken, Z. Wu, D. P. Olson, J. M. Sullivan, Jr., P. L.
Tannebaum, & J. T. Vaughn. (2001). Functional imaging of brain activity
in conscious monkeys responding to sexually arousing cues. NeuroReport, 12,
2231–2236.
French, J. A., D. H. Abbott, & C. T. Snowdon. (1984). The effect of social environ-
ment on estrogen secretion, scent marking and sociosexual behavior in tama-
rins (Saguinus oedipus). American Journal of Primatology, 6, 155–167.
French, J. A. & J. Cleveland. (1984). Scent marking in the tamarin (Saguinus oedi-
pus): Sex differences and ontogeny. Animal Behaviour, 32, 615–618.
French, J. A. & C. T. Snowdon. (1981). Sexual dimorphism in responses to unfamil-
iar intruders in the tamarin (Saguinus oedipus). Animal Behaviour, 29,
822–829.
Friant, S. C., M. A. Campbell, & C. T. Snowdon. (2008). Captive-born cotton-top
tamarins (Saguinus oedipus) respond similarly to vocalizations of predators
and non-predators. American Journal of Primatology, 70, 707–710.
Gaudio, J. L. & C. T. Snowdon. (2008). Spatial cues more salient than color cues in
cotton-top tamarin (Saguinus oedipus) reversal learning. Journal of Comparative
Psychology, 122, 441–444.
Ginther, A. J., A. A. Carlson, T. E. Ziegler, & C. T. Snowdon. (2002). Neonatal and
pubertal development in males of a cooperatively breeding primate, the cotton-
top tamarin (Saguinus oedipus oedipus). Biology of Reproduction, 66, 282–290.
Ginther, A. G. & C. T. Snowdon. (2009). Expectant parents groom adult sons
according to previous alloparenting in a biparental cooperatively breeding pri-
mate. Animal Behaviour, 78, 287–297.
Ginther, A. J., T. E. Ziegler, & C. T. Snowdon. (2001). Reproductive biology of cap-
tive male cotton-top tamarin monkeys as a function of social environment.
Animal Behaviour, 61, 65–78.
Hayes, S. L. & C. T. Snowdon. (1990). Predator recognition in cotton-top tamarins
(Saguinus oedipus). American Journal of Primatology, 20, 283–291.
Hrdy, S. B. (1999). Mother Nature. New York: Ballantine Books.
Hrdy, S. B. (2009). Mothers and Others. Cambridge, MA: Belknap Press.
Humle, T. & C. T. Snowdon. (2008). Socially biased learning in the acquisition of a
complex foraging task in juvenile cottontop tamarins (Saguinus oedipus). Ani-
mal Behaviour, 75, 267–277.
Humle, T., C. T. Snowdon, & T. Matsuzawa. (2009). Social influences on the
acquisition of ant dipping among the wild chimpanzees (Pan troglodytes verus)
of Bossou, Guinea, West Africa. Animal Cognition, 12, S37–S48.
244 ANIMAL BEHAVIOR

Johnson, L. D., A. J. Petto, & P. K. Sehgal. (1991). Factors in the rejection and sur-
vival of captive cotton-top tamarins (Saguinus oedipus). American Journal of
Primatology, 25, 91–102.
Joyce, S. M. & C. T. Snowdon. (2007). Developmental changes in food transfers in
cotton-top tamarins (Saguinus oedipus). American Journal of Primatology, 69,
955–965.
Keverne, E. B., N. D. Martens, & B. Tuite. (1989). Beta-endorphin concentrations
in cerebrospinal fluid of monkeys as influenced by grooming relationships. Psy-
choneuroendocrinology, 14, 155–161.
Kinsley, C. H. & K. G. Lambert. (2006). The maternal brain. Scientific American,
January, 72–79.
Kosfeld, M., M. Heinrichs, P. J. Zak, U. Fischbacher, & E. Fehr. (2005). Oxytocin
increases trust in humans. Nature, 435, 673–676.
Kostan, K. M. & C. T. Snowdon. (2002). Attachment and social preferences in
cooperatively-breeding cotton-top tamarins. American Journal of Primatology,
57, 131–139.
Lazaro-Perea, C. (2001). Intergroup interactions in wild common marmosets (Calli-
thrix jacchus): Territorial defence and assessment of neighbours. Animal Behav-
iour, 62, 11–21.
Lazaro-Perea, C., M. F. Arruda, & C. T. Snowdon. (2004). Grooming as reward?
Social functions of grooming in cooperatively breeding marmosets. Animal
Behaviour, 67, 627–636.
Lazaro-Perea, C., C. S. S. Castro, R. Harrison, A. Araujo, M. F. Arruda, & C. T.
Snowdon. (2000). Behavioral and demographic changes following the loss of
the breeding female in cooperatively breeding marmosets. Behavioral Ecology
and Sociobiology, 48, 137–146.
Matthews, S. A. & C. T. Snowdon. (2011). Long-term memory for calls of relatives
in cotton-top tamarins (Saguinus oedipus). Journal of Comparative Psychology,
125, 366–369.
McConnell, P. B. & C. T. Snowdon. (1986). Vocal interactions between unfamiliar
groups of captive cotton top tamarins. Behaviour, 97, 273–296.
McDermott, J. & M. D. Hauser. (2007). Nonhuman primates prefer slow tempos
but dislike music overall. Cognition, 104, 654–668.
Moscovice, L. R. & C. T. Snowdon. (2006). The role of social context and individual
experience in novel task acquisition in cotton-top tamarins (Saguinus oedipus).
Animal Behaviour, 71, 933–943.
Mota, M. T. S., C. R. Franci, & M. B. C. Sousa. (2006). Hormonal changes related
to paternal and alloparental care in common marmosets (Callithrix jacchus).
Hormones and Behavior, 49, 293–302.
Rapaport, L. G. & C. R. R. Miranda. (2002). Tutoring in wild golden lion tamarins.
International Journal of Primatology, 23, 1063–1070.
Roberts, R. L., K. T. Jenkins, T. Lawler Jr., F. H. Wegner, & J. D. Newman. (2001).
Bromocryptine administration lowers serum prolactin and disrupts parental
COTTON - TOP TAMARINS 245

responsiveness in common marmosets (Callithrix j. jacchus). Hormones and

Behavior, 39, 106–112.
Roush, R. S. & C. T. Snowdon. (1994). Ontogeny of food associated calls in cotton-
top tamarins. Animal Behaviour, 47, 263–273.
Roush, R. S. & C. T. Snowdon. (1999). The effects of social status on food-associated
calling behavior in captive cotton-top tamarins, Animal Behaviour, 58,
1299–1305.
Roush, R. S. & C. T. Snowdon. (2001). Food transfers and the development of feed-
ing behavior and food-associated vocalizations in cotton-top tamarins. Ethol-
ogy, 107, 415–429.
Sanchez, S., F. Peleaz, C. Gil-Burmann, & W. Kaumann. (1999). Costs of infant car-
rying in the cotton-top tamarin. American Journal of Primatology, 48, 99–111.
Savage, A., H. Giraldo, L. Soto, & C. T. Snowdon. (1996a). Demography, group
composition and dispersal of wild cotton-top tamarin groups. American Jour-
nal of Primatology, 38, 85–100.
Savage, A., S. E. Shideler, L. H. Soto, J. Causado, L. H. Giraldo, B. L. Lasley, & C. T.
Snowdon. (1997). Reproductive events of wild cotton-top tamarins (Saguinus
oedipus) in Colombia. American Journal of Primatology, 43, 329–337.
Savage, A., C. T. Snowdon, H. Giraldo, & H. Soto. (1996b). Parental care patterns
and vigilance in wild cotton-top tamarins (Saguinus oedipus). In M. Norconk,
A. Rosenberger, & P. A. Garber (eds.), Adaptive Radiations of Neotropical Pri-
mates (pp. 187–199). New York: Plenum.
Savage, A. C., L. Thomas, K. A. Leighty, L. H. Soto, & F. S. Medina. (2010.) Novel
survey method finds dramatic decline of wild cotton-top tamarin population.
Nature Communications, 1, 1–7. doi: 10.1038/ncomms1030.
Savage, A. C., T. E. Ziegler, & C. T. Snowdon. (1988). Sociosexual development,
pairbond formation and mechanisms of fertility suppression in female
cotton-top tamarins (Saguinus oedipus oedipus). American Journal of Primatol-
ogy, 14, 345–359.
Scott, J. J., K. L. Carlson, & C. T. Snowdon. (2006). Lability of sex differences in
long calling in cotton-top tamarins. American Journal of Primatology, 68,
153–160.
Shutt, K., A. MacLarnon, M. Heistermann, & S. Semple. (2007). Grooming in
Barbary macaques: Better to give than to receive?Biology Letters, 3, 231–233.
Silk, J. B., S. F. Brosnan, J. Vonk, J. Henrich, D. J. Povinelli, A. S. Richardson, et al.
(2005). Chimpanzees are indifferent to the welfare of unrelated group mem-
bers. Nature, 437, 1357–1359.
Smuts, B. B. & D. G. Gubernick. (1992). Male-infant relationships in nonhuman
primates: Paternal investment or mating effort? In B. S. Hewlett (ed.),
Father-Child Relations: Cultural and Biosocial Contexts (pp. 1–31). New York:
Aldine de Gruyter.
Snowdon, C. T. (1996). Parental care in cooperatively breeding species. In J. S. Rose-
nblatt & C. T. Snowdon (eds.), Parental Care: Evolution, Mechanisms and
Adaptive Significance (pp. 643–689). San Diego, CA: Academic Press.
246 ANIMAL BEHAVIOR

Snowdon, C. T. (2009). Plasticity of communication in nonhuman primates. Advan-

ces in the Study of Behavior, 40, 239–276.
Snowdon, C. T. & C. Y. Boe. (2003). Social communication about unpalatable foods
in tamarins. Journal of Comparative Psychology, 117, 142–148.
Snowdon, C. T., J. Cleveland, & J. A. French. (1983). Responses to context- and
individual-specific cues in cotton-top tamarin long calls. Animal Behaviour,
31, 92–101.
Snowdon, C. T. & J. J. Pickhard. (1999). Family feuds severe aggression among co-
operatively breeding cotton-top tamarins. International Journal of Primatology,
20, 651–663.
Snowdon, C. T., B. A. Pieper, C. Y. Boe, K. A. Cronin, A. V. Kurian, & T. E. Zie-
gler. (2010). Variation in oxytocin levels is associated with variation in affilia-
tive behavior in monogamous pairbonded tamarins. Hormones and Behavior,
58, 614–618.
Snowdon, C. T., A. Savage, & P. B. McConnell. (1985). A breeding colony of
cotton-top tamarins. Laboratory Animal Science, 35, 477–481.
Snowdon, C. T. & D. Teie. (2010). Affective responses in tamarins elicited by
species-specific music. Biology Letters, 6, 30–32.
Snowdon, C. T. & T. E. Ziegler. (2007a). Growing up cooperatively: Family pro-
cesses and infant development in marmosets and tamarins. Journal of Develop-
mental Processes, 2, 40–66.
Snowdon, C. T. & T. E. Ziegler. (2007b). Reproductive hormones. In J. Cacioppo,
L. Tassinary, & G. Berntsen (eds.), Handbook of Psychophysiology, Third Edi-
tion (pp. 319–346). Cambridge, UK: Cambridge University Press.
Stoinski, T. S., B. B. Beck, M. A. Bloomsmith, & T. L. Maple. (2003). A behavioural
comparison of captive-born, reintroduced golden lion tamarins and their wild-
born offspring. Behaviour, 140, 137–160.
Tardif, S. D., C. N. Richter, & R. L. Carson. (1984). Effects of sibling experience on
future reproductive success in two species of Callitrchidae. American Journal of
Primatology, 6, 377–380.
Visalberghi, E. & E. Addessi. (2000). Response to changes in food palatability in
tufted capuchin monkeys, Cebus apella. Animal Behaviour, 59, 231–238.
Vonk, J., S. F. Brosnan, J. B. Silk, J. Henrich, A. S. Richardson, S. P. Lambeth, et al.
(2008). Chimpanzees do not take advantage of very low cost opportunities to
deliver food to unrelated group members. Animal Behaviour, 75, 1757–1770.
Washabaugh, K. & C. T. Snowdon. (1998). Chemical communication of reproduc-
tive status in female cotton-top tamarins (Saguinus o. oedipus). American Jour-
nal of Primatology, 45, 337–349.
Washabaugh, K. F., T. E. Ziegler, & C. T. Snowdon. (2002). Variations in care for
cotton-top tamarin (Saguinus oedipus) infants as a function of parental experi-
ence and group size. Animal Behaviour, 63, 1163–1174.
Widowski, T. M., T. A. Porter, T. E. Ziegler, & C. T. Snowdon. (1992). The stim-
ulatory effect of males on the initiation, but not the maintenance, of ovarian
COTTON - TOP TAMARINS 247

cycling in cotton-top tamarins (Saguinus oedipus). American Journal of Prima-

tology, 26, 97–108.
Widowski, T. M., T. E. Ziegler, A. M. Elowson, & C. T. Snowdon. (1990). The role
of males in stimulation of reproductive function in female cotton-top tamarins,
Saguinus o. oedipus. Animal Behaviour, 40, 731–741.
Zahed, S. K., A. V. Kurian, & C. T. Snowdon. (2010). Social dynamics and individ-
ual plasticity of infant care behavior in cooperatively breeding cotton-top tam-
arins. American Journal of Primatology, 72, 296–306.
Zahed, S. R., S. L. Prudom, C. T. Snowdon, & T. E. Ziegler. (2008). Male parenting
and response to infant stimuli in the common marmoset (Callithrix jacchus).
American Journal of Primatology, 70, 84–92.
Ziegler, T. E., W. E. Bridson, C. T. Snowdon, & S. Eman. (1987). Urinary gonado-
tropin and estrogen excretion during the postpartum estrus, conception and
pregnancy in the cotton-top tamarin (Saguinus oedipus oedipus). American Jour-
nal of Primatology, 12, 127–140.
Ziegler, T. E., G. Epple, C. T. Snowdon, T. A. Porter, A. Belcher, & I. Kuederling.
(1993). Detection of the chemical signals of ovulation in the cotton-top tam-
arin, Saguinus oedipus, Animal Behaviour, 45, 313–322.
Ziegler, T. E., L. J. Peterson, M. E. Sosa, & A. M. Barnard. (2011). Differential
endocrine responses to infant odors in common marmoset (Callithrix jacchus)
fathers. Hormones and Behavior, 59, 265–270.
Ziegler, T. E., S. L. Prudom, N. J. Schultz-Darken, A. V. Kurian, & C. T. Snowdon.
(2006). Pregnancy weight gain: Marmoset and tamarin dads show it too.
Biology Letters, 2, 181–183.
Ziegler, T. E., A. Savage, G. Scheffler, & C. T. Snowdon. (1987). The endocrinology
of puberty and reproductive functioning in female cotton-top tamarins (Sagui-
nus oedipus) under varying social conditions. Biology of Reproduction, 37, 618–
627.
Ziegler, T. E., G. Scheffler, & C. T. Snowdon. (1995). The relationship of cortisol
levels to social environment and reproductive functioning in female cotton-
top tamarins, Saguinus oedipus. Hormones and Behavior, 29, 407–424.
Ziegler, T. E., N. J. Schultz-Darken, J. J., Scott, C. T. Snowdon, & C. F. Ferris.
(2005). Neuroendocrine response to female ovulatory odors depends upon
social condition in male common marmosets, Callithrix jacchus. Hormones
and Behavior, 47, 56–64.
Ziegler, T. E., K. F. Washabaugh, & C. T. Snowdon. (2004). Responsiveness of
expectant male cotton-top tamarins, Saguinus oedipus, to mate’s pregnancy.
Hormones and Behavior, 45, 84–92.
Ziegler, T. E., F. H. Wegner, & C. T. Snowdon. (1996). A hormonal role for male
parental care in a New World primate, the cotton-top tamarin (Saguinus oedi-
pus). Hormones and Behavior, 30, 287–297.
Ziegler, T. E., D. J. Wittwer, & C. T. Snowdon. (1993). Circulating and excreted
hormones during the ovarian cycle in the cotton-top tamarin, Saguinus oedipus.
American Journal of Primatology, 31, 55–65.
This page intentionally left blank
 9

Avian Brood Parasitism: How to Spot a

Foreign Egg in the Nest?
Mark E. Hauber and Jason Low

OVERVIEW
To claim and understand the uniqueness of any physical, chemical, or biologi-
cal system, it is necessary to use the same set of approaches, tools, and analyses
to probe other systems. Accordingly, to assess whether and how people are sin-
gularly different from other animals in our perceptual abilities, cognitive skills,
and behavioral repertoires, a parallel set of studies is required to examine how
human and nonhuman animals would respond. Brood parasites have pro-
vided a long-standing and popular model system for the experimental analysis
of self-recognition, which is critical for all social aspects of behavior. Specifi-
cally, when faced with a nest containing own and foreign eggs, the paramount
question is how individuals discriminate themselves from others by utilizing
each of their recognition system’s components: perception, cognition, and
response. Specifically, brood-parasitic birds lay their eggs in others’ nests,
and the costs of parental care and reduced reproductive success are borne by
the hosts. Hosts, in turn, may reject costly parasitism by ejecting foreign prog-
eny or deserting parasitized nests. The cues used by hosts to perceive, recog-
nize, discriminate, and respond to foreign eggs have been well studied in a
variety of avian host-parasite systems. How, in turn, the hosts’ recognition sys-
tems, including their sensory and cognitive processes, receive, sort through,
and determine the behavioral responses to these self-versus-foreign cues allows
250 ANIMAL BEHAVIOR

us to construct and understand general concepts and specific mechanisms of

self-recognition. An overview of recent experimental analyses of egg rejection
behaviors suggests that foreign-egg rejection is triggered through comparisons
with internal filters, or recognition templates, even when hosts’ own eggs are
not present. Such studies provide detailed new information for our under-
standing of parasitic birds’ impacts on hosts’ perceptual processes and their
behavioral outcomes. The final aim of this chapter is then to provide some
starting points towards a sample set of methodological and analytical tools to
determine whether and how other species, including humans, may use similar
perceptual, cognitive, and decision rules to detect foreign items in odd-egg-out
behavioral paradigms.

BACKGROUND
Brood-parasitic birds deposit their eggs in the nests of other individuals or
species, sparing the cost of providing care for their own young (Croston &
Hauber, 2010). This relatively simple act of sneaking an egg into a foreign
nest may result in a coevolutionary arms race of sophisticated host defences
to circumvent or reject parasitism and the escalating steps of the parasites’
breaching these host defences (Rothstein & Robinson, 1998; Lyon & Eadie,
2004). Brood parasitism may be obligate, with all eggs laid in others’ nests,
or facultative, with some own eggs incubated by the parent(s) and others laid
in foreign nests (Davies, 2000). Interspecific brood parasitism involves the
laying of eggs in other species’ nests, whereas in intraspecific parasitism eggs
are laid in nests of conspecifics (Davies, 2000). The benefits and the costs of
brood parasitism to elicit parental care for genetically unrelated young re-
present an important evolutionary selection pressure, so that, respectively, typ-
ical avian brood parasites show morphological and behavioral adaptations to
exploit hosts, and these hosts often show antiparasite adaptations to reduce
the costs of parasitism (Dawkins & Krebs, 1979; Krüger, 2007; Rothstein &
Robinson, 1998).
Irrespective of the type of parasitism, theory predicts that hosts should
reduce or eliminate the many costs of parasitism to increase the benefits of
parental care for their own, genetically related progeny (Servedio & Hauber,
2006). To prevent rejection, some parasites trick hosts by producing perceptu-
ally mimetic eggs (Avilés, 2008), laid into a clutch of several of the hosts’ own
eggs during the laying stage (Moskát & Hauber, 2007), with their thicker egg-
shells making it more difficult or even impossible to puncture them and eject
the chick (Antonov et al., 2009). The sensory, perceptual, and behavioral
responses triggered by parasitic eggs therefore set the stage for testing the limits
AVIAN BROOD PARASITISM : HOW TO SPOT A FOREIGN EGG IN THE NEST ? 251

Figure 9.1. The outline of the recognition-systems approach to understanding behav-

ioral discrimination patterns in animals. (Modified from Sherman et al., 1997)

of each component of the hosts’ recognition systems, categorized as sensory,

perceptual, and behavioral processes (Sherman et al., 1997) (Figure 9.1).

DISCRIMINATING OWN FROM FOREIGN EGGS

The overarching aim of this chapter is to advocate the use of a unified
recognition-systems-based approach to explore, study, and understand the
processes involved in the evolutionary responses of hosts of avian brood para-
sites to foreign eggs in their nests. We therefore propose to modify the classic
tripartite recognition-system definition (Sherman et al., 1997) by forgoing the
analysis of the recognition cue itself, because that is the property of the egg of
the parasite, not of the host, and instead focus on the sensory processes of the
hosts involved in detecting the foreign egg, the cognitive decision rules
involved in assessing and determining the response (if any) to these cues,
and, finally, the behavioral displays involved in responding to the foreign egg’s
cues (Figure 9.2).

Figure 9.2. An alternative construction of a recognition-systems-based approach,

applicable to understanding avian hosts’ behavioral responses to brood-parasitic eggs
in their nests.
252 ANIMAL BEHAVIOR

The most extensively studied avian brood parasite is the common cuckoo
(Cuculus canorus) (hereafter: cuckoo), whose females form host-specialized
races (“gentes”), showing egg mimicry of a particular host species as perceived
by the avian hosts themselves (Avilés, 2008); in turn, most hosts of the cuckoo
show behavioral strategies to discriminate and reject imperfect mimic foreign
eggs (e.g., Moksnes, Røskaft, Braa, et al., 1991; Davies & Brooke, 1988;
Røskaft et al., 2002; Hauber et al., 2006). Cuckoo eggs typically resemble host
eggs in color and pattern (Davies & Brooke, 1988, 1989; Moksnes & Røskaft,
1995) but not in egg shape (Bán et al., 2011), to generate a sensory image that
closely mimics the hosts’ own eggs’ appearance (Cassey, Honza, et al., 2008).
The cognitive basis of behavioral responses to brood parasitism has been
considered to include both self-referenced and socially learned signals of the
parasitic eggs’, chicks’, and adults’ phenotypes (Lotem et al., 1995, Moskát
& Hauber, 2007). For example, the discrimination of own versus foreign eggs
appears to depend on how the avian sensory system receives visual information
(Avilés, 2008; Cassey, Honza, et al., 2008; Spottiswoode & Stevens, 2010;
Stoddard & Stevens, 2010, 2011) and interprets the differences between the
physical patterns of coloration and maculation (spotting pattern) of con- or
heterospecific eggshells (Davies & Brooke, 1989; Moksnes, Røskaft , & Braa,
1991; Cherry et al., 2007; Honza, Polaciková, et al., 2007; Moskát, Székely,
et al., 2008).
The behavioral process of egg discrimination can be divided into several
main parts, as defined by recognition-systems analyses (Sherman et al.,
1997): (1) the perception of the eggshell, (2) the cognitive algorithms involved
in recognizing and generating a response to perceived cues, and (3) the behav-
ioral responses (if any) to the egg, including the cases when hosts recognize the
foreign eggs but are unable to pierce or grasp them, or decide not to reject
them (Davies et al., 1996; Hauber & Sherman, 2001; Moskát & Hauber,
2007; Antonov et al., 2009; Spottiswoode, 2010). Most studies to date have
used data on the subjects’ displays of behavioral patterns and outcomes of
egg rejection to infer perceptual abilities of egg discrimination (Avilés, 2008;
Cassey, Honza, et al., 2008; Spottiswoode & Stevens, 2010; Stoddard &
Stevens, 2010, 2011), while others used video observations on duration of
inspection, latency to reject, and patterns of egg-pecking to reveal that dis-
crimination can take place before or in the absence of egg rejection (Soler
et al., 2002; Honza, Pozgayova, et al., 2007; Antonov et al., 2009). Identify-
ing the perceptual and cognitive basis of egg-discrimination cues and thresh-
olds requires in-depth data collection using a combination of artificial
parasitism and dynamic experimental manipulation of host egg appearance
in space and time (Hauber et al., 2006).
AVIAN BROOD PARASITISM : HOW TO SPOT A FOREIGN EGG IN THE NEST ? 253

Another approach of testing alternative cognitive mechanisms underlying

egg discrimination by potential hosts is to apply a standard experimental
design to a large sample of host nests and concurrently test alternative cogni-
tive decision rules contributing to egg rejection (Hauber & Sherman, 2001;
Moskát & Hauber, 2007). For example, Csaba Moskát and colleagues dem-
onstrated that a regular host of the cuckoo, the great reed warbler (Acrocepha-
lus arundinaceus), relied on both of two different cognitive methods of egg
recognition (Moskát et al., 2010). One of these mechanisms was “discord-
ancy,” where hosts reject the egg phenotype in the minority but accept the
egg phenotype in the majority in the clutch (Rothstein, 1974). Accordingly,
in the great reed warbler, 10 percent of these hosts ejected even their own eggs
(N = 30) when clutches contained four parasitic eggs and one own egg
(Moskát et al., 2010). The alternative, but not mutually exclusive, mechanism
was template recognition, whereby birds compared each eggs against an inter-
nal recognition template of own eggs, probably learned shortly after laying
(Moskát & Hauber, 2007). The experimental separation of the two decision
mechanisms required specially designed treatments of several different manip-
ulations of dyeing a single egg, dyeing all but one egg in the nest, and dyeing
all eggs in the nest, conducted concurrently in different host clutches, to test
alternative cognitive mechanisms of foreign-egg recognition (Moskát et al.,
2010). The results revealed the simultaneous use of multiple cognitive algo-
rithms by these hosts in rejecting foreign eggs. This finding was consistent
with the hypothesis that multiple decision rules working in parallel may be
beneficial in reducing recognition errors through redundancy (Hauber et al.,
2000), including cuckoo egg rejection in those host populations where parasit-
ism rates are high, egg mimicry is good, and clutches are exposed to instances
of multiple parasitism (Moskát et al., 2009).
Critically, neither discordancy based on clutch learning (Rothstein, 1974,
1975; Hauber et al., 2004; Hoover et al., 2006) nor recognition templates
(Lotem, 1993; Lotem et al., 1995; Moskát & Hauber, 2007) need to involve
the concurrent inspection and comparison of own and foreign eggs at the same
time in the nest (i.e., through a process of in situ, simultaneous, online self-
referenced comparison; Hauber & Sherman, 2001). In turn, when faced with
high and multiple parasitism rates, where cuckoos replace host eggs with para-
sitic eggs, hosts’ own eggs may frequently be in the minority or altogether
absent in host nests during the laying or during the incubation period
(Rothstein, 1974; Trine, 2000; Hoover, 2003). As such, hosts faced with high
parasitism rates are predicted to recognize foreign eggs in the absence of own
eggs in the clutch. Where cuckoos parasitize great reed warblers at unusually
high rates (e.g., in Hungary), causing more than 50 percent parasitism rates
254 ANIMAL BEHAVIOR

in habitats where trees are available for vantage points for cuckoos (Røskaft
et al., 2002; Moskát, Hansson, et al., 2008), a potential consequence of heavy
cuckoo parasitism, timed closely during the hosts’ laying cycle (Moskát et al.,
2006), is the high frequency of multiple parasitism. However, when faced
with multiple parasite eggs, great reed warbler hosts show fewer rejection
responses towards any of the cuckoo eggs than in nests with a single parasite
egg, including in natural clutches with multiple parasitism (Moskát et al.,
2009) and in experimental clutches with multiple variably-mimetic dyed host
eggs (Moskát, Székely, et al., 2008). Similarly, rates of rejections of natural
cuckoo eggs decrease with more host eggs in the nest or when own eggs are
experimentally exchanged with foreign eggs during the hosts’ laying period
(Moskát & Hauber, 2007). Critically, in all these cases, the relative numbers
of own versus foreign eggs were variable, thus implying a potential role for
the in situ simultaneous, online comparison of own and foreign eggs in the
nests (Hauber & Sherman, 2001).
The results of these experimental studies lend strong support for the con-
ceptual benefits of a recognition-systems-based approach for the scientific
understanding of what cognitive processes underlie hosts’ rejection behaviors.
In these studies, by identifying and distinguishing the roles that sensory, cog-
nitive, and behavioral processes play in recognizing foreign eggs, it becomes
possible and feasible to separate potential confounds of abilities to perceive
differences and evolved behavioral repertoires, which in turn generate the
observable and quantifiable recorded outcomes of rejection behaviors.

RECOGNITION RESPONSES OF THE HOSTS OF BROWN-HEADED COWBIRDS VERSUS

COMMON CUCKOOS
To illustrate the benefits of this approach, let us consider the example that,
at the interspecific level, there is a long-standing conundrum in that hosts of
the North American obligate brood parasite the brown-headed cowbird
(Molothrus ater) (hereafter: cowbird) do not reject foreign eggs even though
this parasite lays a highly nonmimetic egg (Hauber, 2003). In turn, common
cuckoos lay highly mimetic eggs, many of which are ejected by hosts (Moskát
& Hauber, 2007).
In our experience, simply talking about and presenting the findings and
images often elicits statements and questions from students, colleagues, and
the general public suggestive of severe cognitive limitations (i.e., intellectual
shortcomings) regarding nonrejecter host birds of avian brood parasites. Such
questions include references to whether some birds might not be “sharp”
enough to realize that there is a blatantly dissimilar and clearly foreign egg in
AVIAN BROOD PARASITISM : HOW TO SPOT A FOREIGN EGG IN THE NEST ? 255

their nests. Indeed, evolutionary biologists have accepted just such a scenario
by proposing that some host species have simply not had the time to evolve
the cognitive architecture and decision rules associated with rejection of intact
but dissimilarly colored eggs (Rothstein & Robinson, 1998). For example,
eastern phoebes (Sayornis phoebe) historically have nested in caves and along
streams, two microhabitats typically avoided by brood-parasitic brown-
headed cowbirds (Hauber, 2001). In turn, both phoebes and cowbirds have
extensively benefitted from cattle grazing and suburban development in
eastern North America during the past 400 years, with, for instance, up to
30 percent of phoebe nests parasitized by cowbirds in the early spring near
Ithaca, New York (Hauber et al., 2004). It is therefore no surprise that at such
a short evolutionary timescale (of less than 400 generations), even though
phoebes readily remove broken eggshells from their nests of clear white eggs,
they have not evolved the complex decision rules to grab and toss out intact
partially or fully speckled foreign (cowbird) eggshells from their clutches
(Hauber, 2003).
Alternatively, other folks in the audience of public talks on brood parasites
wonder whether acceptor hosts are intellectually and physically able to recog-
nize and remove foreign objects in their nests but decide not to do so. Exper-
imental evidence again provides some support for such a counterintuitive
scenario. For example, prothonotary warblers (Protonotaria cinerea) are one
of the few frequently parasitized cavity-nesting host species of cowbirds, and
these hosts pay a sizeable fitness cost, through lower reproductive output per
nesting attempt, following the acceptance of parasitic eggs and young in their
nests (Hoover, 2003). Yet these warblers do not reject cowbird eggs, do not
abandon parasitized clutches, and return, year after year, to repeatedly parasit-
ized nest sites (Hoover et al., 2006). Critically, however, rejection behavior of
cowbirds in this host-parasite system is not favored by natural selection
because parasitized nests from which cowbird eggs are experimentally removed
are revisited by female cowbirds to peck on host eggs and toss out host young;
the outcome of such mafia-like cowbird behavior is that acceptance of parasit-
ism generates a greater reproductive output than rejection followed by para-
sitic retaliation (Hoover & Robinson, 2007).
In light of these questions and possibilities to explain why hosts may or may
not remove foreign eggs from their nests, the recognition-systems approach
advocated here allows us to dissect any such apparent sensory, cognitive, and
behavioral dissonances. Accordingly, from the perspective of the sensation of
the cues of hosts versus parasites, it is no surprise that visual models of the
avian sensory system repeatedly confirm that differences in the background
coloration and the maculation of host versus parasite eggs can be generally
256 ANIMAL BEHAVIOR

detected both by hosts of interspecific avian brood parasites (Stoddard &

Stevens, 2010, 2011) and even by phylogenetically related, nonhost bird spe-
cies and humans (Cassey, Ewen, et al., 2008; Cassey et al., 2009, 2010)—
hosts can clearly see that the eggs are different. Therefore, the lack of sensory
discrimination does not appear to be the answer to resolve the puzzle. That
said, a focus on sensory processes alone results in a novel alternative proposal:
instead of being moderately mimetic (Klippenstein & Sealy, 2008), perhaps
cowbird eggs are simply cryptic against the backgrounds of many of their
hosts’ finely woven open-cup nest bottoms (Bill Strausberger, personal com-
munication). This hypothesis predicts that cowbird eggs are accepted not
because they cannot be identified in comparison with the hosts’ own eggs
but because they cannot be delineated and detected against the background
pattern of the host nest’s illumination; this hypothesis thus far remains
untested in cowbirds and their typically open-cup nesting hosts (Hoover
et al., 2006) but received support for the nonmimetic dark eggs of Chalcites
cuckoos parasitizing enclosed-nest breeding hosts in Australia (Langmore
et al., 2009); similarly, reduced nest illumination was also associated with
reduced rejection behaviors even in the open-cup-nesting hosts of Cuculus
cuckoos in the Czech Republic (Honza et al., 2011), although lower environ-
mental light levels and rarer parasitic egg-rejection behavior from host nests
are also consistent with some psychophysical mechanisms of mimetic egg
discrimination.
Extensive experimental work on the sensory bases of host-parasite egg rejec-
tion, conducted using subjects within the same species, consistently supported
the pattern that increased mimicry between own and foreign eggs leads to
reduced rejection by the hosts (Hauber et al., 2006; Rutila et al., 2006). How-
ever, in most studies only a single foreign egg was introduced into experimen-
tal nests, thereby confounding the sensory mechanisms of concurrent contrast
of own versus foreign eggs and sensory filters assessing the sensory inputs from
the foreign egg alone. Therefore, to disentangle these confounding factors,
experimental tests of the sensory mechanisms of rejecter hosts require nest
manipulations where the entire clutch has been altered. Recently, Moskát
and colleagues (2010) conducted such a set of experiments, where one, several,
or all eggs in great reed warblers’ nests were manipulated by changing macula-
tion patterns of the hosts’ own eggs. The experiments revealed that egg-
rejection rates increased above background or control levels of the accidental
rejection rate of unmanipulated own eggs by the hosts, including those
clutches where the appearance of all host eggs was manipulated. This implies
that these hosts do not require viewing of their own eggs to recognize and
reject foreign eggs in the same clutch. Still, even in this series of experiments,
AVIAN BROOD PARASITISM : HOW TO SPOT A FOREIGN EGG IN THE NEST ? 257

whenever the full clutch was manipulated, all host eggs were dyed in the same
manner, so that any potential use of the comparison of more or less mimetic
foreign eggs to a perceptual recognition template of own eggs in eliciting
acceptance versus rejection responses could not be tested. To identify the plas-
ticity of the discrimination threshold in generating acceptance versus rejection
decision requires the use of differently mimetic eggs in the same clutch
(Hauber et al., 2006). In that experiment, we used a moderately mimetic egg:
great reed warbler host egg dyed with speckles versus a fully nonmimetic egg:
a host egg dyed fully brown. The moderately mimetic egg was accepted in most
cases, and the nonmimetic egg was rejected in most cases. However, after the
initial rejection of the nonmimetic egg, dyeing a second host egg with the mod-
erately mimetic speckles suddenly resulted in most eggs being rejected; this
experiment showed that acceptance versus rejection decisions are not based
on stationary acceptance thresholds but instead use flexible decision rules to
recognize and reject foreign egg colors and patterns (Hauber et al., 2006).
Similarly, in a separate set of experiments Moskát, Tamás Székely, and col-
leagues (2008) also manipulated all host eggs by painting variable numbers of
spots on hosts’ own eggs during the laying stage and found for the first time
clear experimental evidence for a reduced rejection rate of foreign eggs with
increasing intraclutch variability. But again, that experiment did not test the
relative acceptance threshold of foreign eggs that themselves vary in their rela-
tive appearance compared to the original host eggs. Experiments are now
needed to assess whether hosts possess a consistent absolute or relative accep-
tance threshold to detect foreign egg colors, and whether those foreign eggs
that elicit discrimination responses, and are behaviorally rejected, may be
modified by the overall appearance of the hosts’ full clutch, including variation
in the proportion of hosts’ own eggs present in the nest.
A critical interpretation of the increasing amounts of experimental data on
egg-rejection behaviors of hosts of brood-parasitic birds is that hosts recognize
more eggs than they actually reject. This conclusion also has implications for
the cognitive decision rules involved in mediating behavioral responses
through sensation and perception. Specifically, the new results reveal that the
lack of behavioral responses to foreign eggs cannot be considered as necessarily
being due to limited cognitive complexity of host birds (Moskát & Hauber,
2007; Antonov et al., 2009). In other words, hosts of brood parasites are likely
much smarter than they appear in response to natural and experimental para-
sitism. This is, of course, no surprise to any ornithologists or other natural his-
torian: for example, eastern phoebe hosts of brown-headed cowbirds are
sophisticated flycatchers that capture prey in midair using acrobatic flight
and precise vision. In addition, phoebes also engage in several thousands of
258 ANIMAL BEHAVIOR

miles of migration for the winter but return to the exact same nest site where
they had bred in the previous year, down to the nearest meter (Hauber,
2002). As such, phoebes’ lack of behavioral responses to speckled cowbird eggs
is unlikely to be due to limited sensory or cognitive traits and repertoires.

EGG-MANIPULATION EXPERIMENTS
The broader implication of the range of experimental methods used in the
study of avian brood-parasitism and egg-rejection behaviors, including the critical
uses of parallel and concurrent egg manipulations of rejecter hosts, is that these
experimental manipulations can provide the basis of several parallel experiments
using photographic and image-presentation approaches for an “odd-one-out”
paradigm to study human behavioral responses and cognitive decision rules in
the recognition-systems approach. In one experiment, a set of reference images
would be provided to the human study subjects to familiarize them with the range
of acceptable eggs (images with clutches of host eggs only), thereby providing a
range and variation of acceptable phenotypic traits present in the population.
Following the familiarization step, a novel set of images would then be pro-
vided (ideally, standardized for egg size and clutch size: number of eggs), and
the subjects would be asked (people) or trained (nonhuman animals) to detect
and point to the foreign egg. By using a suite of different clutch images, with
natural or artificially manipulated foreign egg(s) placed in the nest, including
either single foreign eggs or multiple (including all) foreign eggs in the clutch,
the concurrent experimental approach advocated here, modified from Moskát
and colleagues (2010) could be applied into this paradigm. The study, for
example, could focus on cognitive mechanisms based on discordancy (i.e.,
the detection of the odd egg relative to the majority egg type present in the
nest) versus conspecific familiarity (i.e., the detection of eggs with an unfamil-
iar phenotypic trait, which would fall outside the range presented during the
familiarization stage of the experiment using acceptable, conspecific eggs
only). Because these cognitive mechanisms make different predictions about
the eggs identified as foreign during the experiment, it would be possible to
identify which algorithm is used by subjects to recognize foreign eggs.
Critically, a second conclusion associated with the schematics of such an
experimental design is that studies with natural or artificial cuckoo eggs that
are based on the presence or introduction of a single parasitic egg in the nest
cannot discriminate between decision rules based on discordancy versus con-
specific familiarity—in these situations, the predictions are identical. Instead,
to contrast the predictions of these alternatives, it is required that additional
manipulations of several eggs in each clutch take place.
AVIAN BROOD PARASITISM : HOW TO SPOT A FOREIGN EGG IN THE NEST ? 259

Furthermore, by altering the relative proportion of host versus foreign eggs

or changing the absolute number of host and foreign eggs in the nest, this
experimental schematic will also make it feasible to identify and test specific
predictions of alternative or novel cognitive mechanisms not previously identi-
fied in rejection decisions by hosts of obligate avian brood parasites (e.g.,
numerosity of host versus parasite eggs; Lyon, 2003; Low et al., 2009; White
et al., 2009). For example, some host species are poorer at identifying foreign
eggs in the nest with increasing number of their own eggs because more eggs
represent greater phenotypic variability against which foreign eggs are com-
pared, as predicted by each or a combination of the discordancy-, familiarity-,
and internal-template-based mechanisms (Moskát & Hauber, 2007; Moskát,
Avilés, et al., 2008; Moskát et al., 2009). To discriminate among these alterna-
tives again requires a concurrent experimental design.

CONCLUSIONS
The main conclusion of this chapter is that future work should incorporate
both the conceptual and the practical details developed in the studies reviewed
and advocated here in order to better understand the evolutionary parallels
and innovations required to detect parasitic eggs. In addition, it is essential
to examine concurrently whether multiple cognitive processes are involved in
the behavioral discrimination of own and foreign eggs, for instance discord-
ancy and self-referenced familiarity. Yet, in the absence of the availability of
either of these mechanisms, an internal threshold-based discrimination deci-
sion may also be available to hosts detecting foreign eggs; the combination of
these non-mutually-exclusive alternative cognitive processes can clearly pro-
duce efficient and beneficial rejection decisions. That cognitive mechanisms
do not act mutually exclusively but instead supplement and combine with
each other is consistent with the scenario that increasingly fail-safe mecha-
nisms of accurate parasite-egg detection function in these hosts (Hauber &
Sherman, 2001; Moskát et al., 2010). Future work, therefore, should also
examine the novel prediction that the evolutionary response to arms races with
increasingly mimetic hosts is the concurrent reliance on multiple cognitive
decision rules to converge on the detection of and the behavioral decision to
reject the foreign egg(s) from the nest.

ACKNOWLEDGMENTS
We are grateful to editor Ken Yasukawa for his kind invitation to contrib-
ute and careful comments on a draft of this book chapter; to Zachary Aidala,
260 ANIMAL BEHAVIOR

Michael Anderson, Phill Cassey, Rebecca Croston, Brian Gill, Tomas Grim,
Jeffrey Hoover, Csaba Moskát, Peter Samas, and many others for our years
of collaboration on parasitic cowbirds and cuckoos; and to the Human Fron-
tier Science Program, the Marsden Fund, the National Geographic Society,
and PSC-CUNY for research funding.

REFERENCES AND SUGGESTED READING

Antonov, A., B. G. Stokke, A. Moksnes, & E. Røskaft. (2009). Evidence for egg dis-
crimination preceding failed rejection attempts in a small cuckoo host. Biology
Letters, 5, 169–171.
Antonov, A., B. G. Stokke, J. R. Vikan, F. Fossøy, P. S. Ranke, E. Røskaft, A.
Moksnes, A. P. Møller, & J. A. Shykoff. (2010). Egg phenotype differentiation
in sympatric cuckoo Cuculus canorus gentes. Journal of Evolutionary Biology,
23, 1170–1182.
Avilés, J. M. (2008). Egg colour mimicry in the common cuckoo Cuculus canorus as
revealed by modelling host retinal function. Proceedings of the Royal Society
London, B, 275, 2345–2352.
Avilés, J. M. & A. P. Møller. (2003). Meadow pipit (Anthus pratensis) egg appearance
in cuckoo (Cuculus canorus) sympatric and allopatric populations. Biological
Journal of the Linnean Society, 79, 543–549.
Avilés, J. M., J. R. Vikan, F. Fossøy, A. Antonov, A. Moksnes, E. Røskaft, et al.
(2010). Avian colour perception predicts behavioural responses to experimen-
tal brood parasitism in chaffinches. Journal of Evolutionary Biology, 23, 293–
301.
Bán, M., Z. Barta, R. A. Munoz, F. Takasu, H. Nakamura, & C. Moskát. (2011).
The analysis of common cuckoo’s egg shape in relation to its hosts’ in two geo-
graphically distant areas. Journal of Zoology, 284, 77–83.
Barabás, L., B. Gilicze, F. Takasu, & C. Moskát. (2004). Survival and antiparasite
defense in a host metapopulation under heavy brood parasitism: A source-
sink dynamic model. Journal of Ethology, 22, 143–151.
Bates, D. & M. Maechler. (2009). lme4: Linear mixed-effects models using S4
classes. R package version 0.999375-32. http://CRAN.R-project.org/
package=lme4.
Braa, A. T., A. Moksnes, & E. Røskaft. (1992). Adaptations of bramblings and chaf-
finches towards parasitism by the common cuckoo. Animal Behaviour, 43,
67–78.
Brooke, M. L., N. B. Davies, & D. G. Noble. (1998). Rapid decline of host defences
in response to reduced cuckoo parasitism: Behavioural flexibility of reed war-
blers in a changing world. Proceedings of the Royal Society London, B, 265,
1277–1282.
Cassey, P., T. M. Blackburn, & K. L. Evans. (2005). Changes in egg size of exotic
passerines introduced to New Zealand. Notornis, 52, 243–246.
AVIAN BROOD PARASITISM : HOW TO SPOT A FOREIGN EGG IN THE NEST ? 261

Cassey, P., J. G. Ewen, T. M. Blackburn, M. E. Hauber, M. Vorobyev, & N. J.

Marshall. (2008). Maternal investment of yolk carotenoids is not related to
eggshell color variation in European thrushes. Naturwissenschaften, 95,
713–721.
Cassey, P., J. G. Ewen, N. J. Marshall, M. Vorobyev, T. M. Blackburn, & M. E.
Hauber. (2009). Are avian eggshell colours effective intraspecific communica-
tion signals? A perceptual modeling approach. Ibis, 151, 689–698.
Cassey, P., M. Honza, T. Grim, & M. E. Hauber. (2008). The modelling of avian
visual perception predicts behavioural rejection responses to foreign egg col-
ours. Biology Letters, 4, 515–517.
Cassey, P., S. J. Portugal, J. G. Ewen, R. L. Boulton, M. E. Hauber, & T. M. Black-
burn. (2010). Variability in eggshell colour: A comparative study of museum
eggshells. PLoS ONE, 5, e12054.
Cherry, M. I., A. T. D. Bennett, & C. Moskát. (2007). Host intra-clutch variation,
cuckoo egg matching and egg rejection by great reed warblers. Naturwissen-
schaften, 94, 441–447.
Cramp, S. (ed.). (1994). Warblers. In The birds of the Western Palearctic, Vol. 4. Terns
to Woodpeckers. Oxford, UK: Oxford University Press,
Croston, R. & M. E. Hauber. (2010). The ecology of avian brood parasitism. Nature
Education Knowledge, 1, 3.
Davies, N. B. (2000). Cuckoos, Cowbirds and Other Cheats. London: T. & A. D.
Poyser.
Davies, N. B. & M. Brooke. (1988). Cuckoos versus reed warblers: Adaptations and
counteradaptations. Animal Behaviour, 36, 262–284.
Davies, N. B. & M. Brooke. (1989). An experimental study of co-evolution between
the cuckoo, Cuculus canorus, and its hosts. I. Host egg discrimination. Journal
of Animal Ecology, 58, 207–224.
Davies, N. B., M. Brooke, & A. Kacelnik. (1996). Recognition errors and probability
of parasitism determine whether reed warblers should accept or reject mimetic
cuckoo eggs. Proceedings of the Royal Society London, B, 263, 925–931.
Dawkins, N. B. & J. R. Krebs. (1979). Arms races between and within species. Pro-
ceedings of the Royal Society London, B, 205, 489–511.
Endler, J. A. & P. Mielke Jr. (2005). Comparing entire colour patterns as birds see
them. Biological Journal of the Linnean Society, 86, 405–431.
Erard, C. & G. Armani. (1986). Reflexions sur un cas de parasitisme et d’aide au
nourrissage mettant en cause Turdus merula, T. philomelos et Erithacus rube-
cula. Alauda, 54, 138–144.
Evans, K. L., R. P. Duncan, T. M. Blackburn, & H. Q. P. Crick. (2005). Investigat-
ing geographic variation in clutch size using a natural experiment. Functional
Ecology, 19, 616–624.
Formánek, J. (1958). Migration of blackbird (Turdus merula) and song thrush (T.
ericetorum) in Czechoslovakia. Sylvia, 15, 23–41.
Futuyma, D. J. (1998). Evolutionary Biology. Third edition. Sunderland, MA: Sinauer
Associates.
262 ANIMAL BEHAVIOR

Grendstad, L. C., A. Moksnes, & E. Røskaft. (1999). Do strategies against conspecific

brood parasitism occur in redwings Turdus iliacus? Ardea, 87, 101–111.
Grim, T. (2005a). Mimicry vs. similarity: Which resemblances between brood para-
sites and their hosts are mimetic and which are not? Biological Journal of the
Linnean Society, 84, 69–78.
Grim, T. (2005b). Host recognition of brood parasites: Implications for methodology
in studies of enemy recognition. Auk, 122, 530–543.
Grim, T. (2006). Cuckoo growth performance in parasitized and unused hosts: Not
only host size matters. Behavioral Ecology and Sociobiology, 60, 716–723.
Grim, T. (2007). Equal rights for chick brood parasites. Annales Zoologici Fennici, 44,
1–7.
Grim, T. & M. Honza. (2001). Differences in behaviour of closely related thrushes
(Turdus philomelos and T. merula) to experimental parasitism by the common
cuckoo Cuculus canorus. Biologia, 56, 549–556.
Grim, T., J. Rutila, P. Cassey, & M. E. Hauber. (2009). Experimentally constrained
virulence is costly for common cuckoo Cuculus canorus chicks. Ethology, 115,
14–22.
Grim, T., P. Samaš, C. Moskát, O. Kleven, M. Honza, A. Moksnes, E. Røskaft, & B.
G. Stokke. (2011). Constraints on host choice: Why do parasitic birds rarely
exploit some common potential hosts? Journal of Animal Ecology, 80, 508–518.
Gross, M. R. (1996). Alternative reproductive strategies and tactics: Diversity within
sexes. Trends in Ecology and Evolution, 11, 92–98.
Hale, K. & J. V. Briskie. (2007). Response of introduced European birds in New
Zealand to experimental brood parasitism. Journal of Avian Biology, 38, 198–
204.
Hargitai, R., M. Moskát, M. Bán, D. Gil, I. López-Rull, & E. Solymos. (2010). Egg-
shell characteristics and yolk composition in the common cuckoo Cuculus
canorus: Are they adapted to brood parasitism? Journal of Avian Biology, 41,
177–185.
Hauber, M. E. (2001). Site selection and repeatability in brown-headed cowbird
(Molothrus ater) parasitism of eastern phoebe (Sayornis phoebe) nests. Canadian
Journal of Zoology, 79, 1518–1523.
Hauber, M. E. (2002). Is reduced clutch size a cost of parental care in eastern phoebes
(Sayornis phoebe)? Behavioral Ecology and Sociobiology, 51, 503–509.
Hauber, M. E. (2003). Egg-capping is a cost paid by hosts of interspecific brood para-
sites. Auk, 120, 860–865.
Hauber, M. E. & C. Moskát. (2008). Shared parental care is costly for nestlings of
common cuckoos and their great reed warbler hosts. Behavioral Ecology, 19,
79–86.
Hauber, M. E., C. Moskát, & M. Bán. (2006). Experimental shift in hosts’ acceptance
threshold of inaccurate-mimic brood parasite eggs. Biology Letters, 2, 177–180.
Hauber, M. E. & P. W. Sherman. (2001). Self-referent phenotype matching: Theo-
retical considerations and empirical results. Trends in Neurosciences, 24,
609–616.
AVIAN BROOD PARASITISM : HOW TO SPOT A FOREIGN EGG IN THE NEST ? 263

Hauber, M. E., P. W. Sherman, & D. Paprika. (2000). Self-referent phenotype

matching in a brood parasite: The armpit effect in brown-headed cowbirds
(Molothrus ater). Animal Cognition, 3, 113–117.
Hauber, M. E., P. J. Yeh, & J. O. L. Roberts. (2004). Patterns and coevolutionary
consequences of repeated brood parasitism. Proceedings of the Royal Society Lon-
don, B, 271, S317–S320.
Higgins, P. J. (ed.). (2006). Handbook of Australian, New Zealand and Antarctic Birds,
Vol. 7B. Boatbill to Starlings. Melbourne, Australia: Oxford Universiy Press.
Honza. M. & C. Moskát (2008). Egg rejection behaviour in the great reed warbler
(Acrocephalus arundinaceus): The effect of egg type. Journal of Ethology, 26,
389–395.
Honza, M., L. Polaciková, L., & P. Procházka. (2007). Ultraviolet and green parts of
the colour spectrum affect egg rejection in the song thrush (Turdus philomelos).
Biological Journal of the Linnean Society, 92, 269–276.
Honza, M., M. Pozgayova, P. Procházka, & E. Kadlec. (2007). Consistency in egg
rejection behaviour: Responses to repeated brood parasitism in the blackcap
(Sylvia atricapilla). Ethology, 113, 344–351.
Honza, M., P. Procházka, K. Morongová, M. Čapek, & V. Jelínek (2011). Do nest
light conditions affect rejection of parasitic eggs? A test of the light environ-
ment hypothesis. Ethology, 117, 538–546.
Honza, M., P. Procházka, B. G. Stokke, A. Moksnes, E. Røskaft, M. Capek, & V.
Mrlik. (2004). Are blackcaps current winners in the evolutionary struggle
against the common cuckoo? Journal of Ethology, 22, 175–180.
Hoover, J. P. (2003). Decision rules for site fidelity in a migratory bird, the prothono-
tary warbler. Ecology, 84, 416–430.
Hoover, J. P. & S. K. Robinson (2007). Retaliatory mafia behavior by a parasitic cow-
bird favors host acceptance of parasitic eggs. Proceedings of the National Acad-
emy of Sciences of the USA, 104, 4479–4483.
Hoover, J. P., K. Yasukawa, K., & M. E. Hauber. (2006). Spatially and temporally
structured avian brood parasitism affects the fitness benefits of hosts’ rejection
strategies. Animal Behaviour, 72, 881–890.
Hosoi, S. A., & S. I. Rothstein. (2000). Nest desertion and cowbird parasitism: Evidence
for evolved responses and evolutionary lag. Animal Behaviour, 59, 823–840.
Jackson, W. M. (1998). Egg discrimination and egg-color variability in the northern
masked weaver: The importance of conspecific versus interspecific parasitism.
In S. I. Rothstein & S. K. Robinson (eds.), Parasitic Birds and Their Hosts
(pp. 407–416). New York: Oxford University Press.
Kilner, R. M. (2006). The evolution of egg colour and patterning in birds. Biological
Reviews, 81, 383–406.
Kilner, R. M., J. R. Madden, & M. E. Hauber. (2004). Brood parasitic cowbird nest-
lings use host young to procure resources. Science, 305, 877–879.
Klippenstine, D. R. & S. G. Sealy. (2008). Differential ejection of cowbird eggs and
non-mimetic eggs by grassland passerines. Wilson Journal of Ornithology, 120,
667–673.
264 ANIMAL BEHAVIOR

Kosciuch, K. L., T. H. Parker, & B. K. Sandercock. (2006). Nest desertion by a cow-

bird host: An anti-parasite behavior or a response to egg loss? Behavioral
Ecology, 17, 917–924.
Krüger, O. (2007). Cuckoos, cowbirds and hosts: Adaptations, trade-offs and con-
straints. Philosophical Transactions of the Royal Society of London, B, 362,
1873–1886.
Langmore, N. E., S. Hunt, & R. M. Kilner. (2003). Escalation of a coevolutionary
arms race through host rejection of brood parasitic young. Nature, 422,
157–160.
Langmore, N. E., R. M. Kilner, S. H. M. Butchart, G. Maurer, N. B. Davies, A.
Cockburn, N. A. Macgregor, A. Peters, M. J. L. Magrath, & D. K. Dowling.
(2005). The evolution of egg rejection by cuckoo hosts in Australia and
Europe. Behavioral Ecology, 16, 686–692.
Langmore, N. E., M. Stevens, G. Maurer, R. Heinsohn, M. L. Hall, A. Peters, & R.
M. Kilner. (2011). Visual mimicry of host nestlings by cuckoos. Proceedings of
the Royal Society of London, B, 278, 2455–2463
Langmore, N. E., M. Stevens, G. Maurer, & R. M. Kilner. (2009). Are dark cuckoo
eggs cryptic in host nests? Animal Behaviour, 78, 461–468.
Latif, Q. S., J. L. Grenier, S. H. Heath, G. Ballard, & M. E. Hauber. (2006). First
evidence of conspecific brood parasitism and egg ejection in song sparrows,
with comments on methods sufficient to document these behaviors. Condor,
108, 452–458.
Lotem, A. (1993). Learning to recognize nestlings is maladaptive for cuckoo Cuculus
canorus hosts. Nature, 362, 743–745.
Lotem, A., H. Nakamura, & A. Zahavi. (1995). Constrains on egg discrimination
and cuckoo-host co-evolution. Animal Behaviour, 49, 1185–1209.
Lovaszi, P. & C. Moskát. (2004). Break-down of arms race between the red-backed
shrike (Lanius collurio) and common cuckoo (Cuculus canorus). Behaviour,
141, 245–262.
Low, J., K. C. Burns, & M. E. Hauber. (2009). Wild number sense in brood parasitic
brown-headed cowbirds. Ibis, 151, 775–777.
Lyon, B. E. (2003). Egg recognition and counting reduce costs of avian conspecific
brood parasitism. Nature, 422, 495–499.
Lyon, B. E., & J. M. Eadie. (2004). An obligate brood parasite trapped in the intra-
specific arms race of its hosts. Nature, 432, 390–393.
Martin-Vivaldi, M., M. Soler, & A. P. Møller. (2002). Unrealistically high costs of
rejecting artificial model eggs in cuckoo Cuculus canorus hosts. Journal of Avian
Biology, 33, 295–301.
Moksnes, A. & E. Røskaft. (1995). Egg-morph and host preferences in the common
cuckoo (Cuculus canorus): An analysis of cuckoo and host eggs from European
museum collections. Journal of Zoology, 236, 625–648.
Moksnes, A., E. Røskaft, V. Bicik, M. Honza, & I. J. Øien. (1993). Cuckoo Cuculus
canorus parasitism on Acrocephalus warblers in Southern Moravia in the Czech
Republic. Journal of Ornithology, 134, 425–434.
AVIAN BROOD PARASITISM : HOW TO SPOT A FOREIGN EGG IN THE NEST ? 265

Moksnes, A., E. Røskaft, & A. T. Braa. (1991). Rejection behavior by common

cuckoo hosts towards artificial brood parasite eggs. Auk, 108, 348–354.
Moksnes, A., E. Røskaft, A. T. Braa, L. Korsnes, H. M. Lampe, & H. C. Pedersen.
(1991). Behavioural responses of potential hosts towards artificial cuckoo eggs
and dummies. Behaviour, 116, 64–89.
Møller, A. P. (1976). One more case of blackbird Turdus merula as a cuckoo Cuculus
canorus host. Dansk Ornitologisk Forenings Tidsskrift, 70, 71.
Moskát, C., J. M. Avilés, M. Bán, R. Hargitai, & A. Zölei. (2008). Experimental sup-
port for the use of egg uniformity in parasite egg discrimination by cuckoo
hosts. Behavioral Ecology and Sociobiology, 62, 1885–1890.
Moskát, C., M. Bán, M., T. Székely, J. Komdeur, R. W. G. Lucassen, A. L. van
Boheemen, et al. (2010). Discordancy or template-based recognition? Dissect-
ing the cognitive basis of the rejection of foreign eggs in hosts of avian brood
parasites. Journal of Experimental Biology, 213, 1976–1983.
Moskát, C., Z. Barta, M. E. Hauber, & M. Honza. (2006). High synchrony of egg
laying between common cuckoos (Cuculus canorus) and their great reed war-
bler (Acrocephalus arundinaceus) hosts. Ethology Ecology & Evolution, 18,
159–167.
Moskát, C., B. Hansson, L. Barabás, I. Bártol, & Z. Karcza. (2008). Common
cuckoo Cuculus canorus parasitism, antiparasite defence and gene flow in
closely located populations of great reed warblers Acrocephalus arundinaceus.
Journal of Avian Biology, 39, 663–671.
Moskát, C. & M. E. Hauber. (2007). Conflict between egg recognition and egg rejec-
tion decisions in common cuckoo (Cuculus canorus) hosts. Animal Cognition,
10, 377–386.
Moskát, C., M. E. Hauber, J. M. Avilés, M. Bán, R. Hargitai, & M. Honza. (2009).
Increased host tolerance of multiple cuckoo eggs leads to higher fledging suc-
cess of the brood parasite. Animal Behaviour, 77, 1281–1990.
Moskát, C., Z. Karcza, & T. Csörgö. (2003). Egg rejection in European blackbirds
(Turdus merula): The effect of mimicry. Ornis Fennica, 80, 86–91.
Moskát, C., E. C. Rosendaal, M. Boers, A. Zölei, M. Bán, & J. Komdeur. (2011).
Post-ejection nest-desertion of common cuckoo hosts: A second defense
mechanism or avoiding reduced reproductive success? Behavioral Ecology and
Sociobiology, 65, 1045–1053.
Moskát, C., T. Székely, I. C. Cuthill, & T. Kisbenedek. (2008). Hosts’ responses to
parasitic eggs: Which cues elicit hosts’ egg discrimination? Ethology,
114,186–194.
Moskát C., J. Szentpéteri, & Z. Barta. (2002). Adaptations by great reed warblers to
brood parasitism: A comparison of populations in sympatry and allopatry with
the common cuckoo. Behaviour, 139, 1313–1329.
Nakamura, H. (1990). Brood parasitism by the cuckoo Cuculus canorus in Japan and
the start of new parasitism on the azure-winged magpie Cyanopica cyana. Jap-
anese Journal of Ornithology, 39, 1–18.
266 ANIMAL BEHAVIOR

Partecke, J. & E. Gwinner. (2007). Increased sedentariness in European blackbirds

following urbanization: A consequence of local adaptation? Ecology, 88, 882–
890.
Peer, B. D., S. I. Rothstein, K. S. Delaney, & R. C. Fleischer. (2007). Defence behav-
iour against brood parasitism is deeply rooted in mainland and island scrub-
jays. Animal Behaviour, 73, 55–63.
Peer, B. D. & S. G. Sealy. (2004). Correlates of egg rejection in hosts of the brown-
headed cowbird. Condor, 106, 580–599.
Ranjard, L., M. G. Anderson, M. J. Rayner, R. B. Payne, I. G. McLean, J. V. Briskie,
H. A. Ross, D. H. Brunton, S. M. N. Woolley, & M. E. Hauber. (2010). Bio-
acoustic distances between the begging calls of brood parasites and their host
species: A comparison of metrics and techniques. Behavioral Ecology and Socio-
biology, 64, 1915–1926.
R Development Core Team. (2009). R: A Language and Environment for Statistical
Computing. Vienna, Austria: R Foundation for Statistical Computing, http://
www.R-project.org.
Reeve, H. K. (1989). The evolution of conspecific acceptance thresholds. American
Naturalist, 133, 407–435.
Rodríguez-Gironés, M. A., & A. Lotem. (1999). How to detect a cuckoo egg: A
signal-detection theory model for recognition and learning. American Natural-
ist, 153, 633–648.
Ringsby, T. H., A. Moksnes, E. Røskaft, & H. E. Lerkelund. (1993). Do conspecific
brood parasitism and antiparasite strategies occur in fieldfares Turdus pilaris?
Fauna norvegica Series C, Cinclus, 16, 45–53.
Røskaft, E., A. Moksnes, B. G. Stokke, C. Moskát, & M. Honza. (2002). The spatial
habitat structure of host populations explains the pattern of rejection behavior
in hosts and parasitic adaptations in cuckoos. Behavioral Ecology, 13, 163–168.
Rothstein, S. I. (1974). Mechanisms of avian egg recognition: Possible learned and
innate factors. Auk, 91, 796–807.
Rothstein, S. I. (1975). An experimental and teleonomic investigation of avian brood
parasitism. Condor, 77, 250–271.
Rothstein, S. I. (1982). Mechanisms of avian egg recognition: Which egg parameters
elicit responses by rejecter species? Behavioral Ecology and Sociobiology, 11,
229–239.
Rothstein, S. I. (2001). Relic behaviours, coevolution and the retention versus loss of
host defences after episodes of avian brood parasitism. Animal Behaviour, 61,
95–107.
Rothstein, S. I. & S. K. Robinson (eds.). (1998). Parasitic Birds and Their Hosts. New
York: Oxford University Press.
Rutila, J., J. Jokimaki, J. M. Avilés, & M. L. Kaisanlahti-Jokimak. (2006). Responses
of parasitized and unparasitized common redstart (Phoenicurus phoenicurus)
populations against artificial cuckoo parasitism. Auk, 123, 259–265.
SAS Institute Inc. (1995). JMP Statistics and Graphics Guide. Version 3.2. Cary, NC:
SAS Institute.
AVIAN BROOD PARASITISM : HOW TO SPOT A FOREIGN EGG IN THE NEST ? 267

SAS Institute Inc. (2000). SAS Online Doc, Version 8. Cary, NC: SAS Institute.
Schuetz, J. G. (2005). Reduced growth but not survival of chicks with altered gape
patterns: Implications for the evolution of nestling similarity in a parasitic
finch. Animal Behaviour, 70, 839–848.
Segelbacher, G., T. Sacher, A. Schwarzenberger, S. Woitsch, F. Bairlein, & T. Cop-
pack. (2008). Eight microsatellite loci characterised in the European blackbird,
Turdus merula. Journal of Ornithology, 149, 131–133.
Servedio, M. R. & M. E. Hauber. (2006). To eject or to abandon? Life history traits
of hosts and parasites interact to influence the fitness payoffs of alternative anti-
parasite strategies. Journal of Evolutionary Biology, 19, 1585–1594.
Servedio, M. R. & R. Lande. (2003). Coevolution of an avian host and its parasitic
cuckoo. Evolution, 57, 1164–1175.
Sherman, P. W., H. K. Reeve, & D. W. Pfenning. (1997). Recognition systems. In
J. R. Krebs & N. B. Davies (eds.), Behavioural Ecology (pp. 69–96). Oxford,
UK: Blackwell Scientific.
Soler, J. J., J. G. Martinez, M. Soler, & A. P. Møller. (1999) Genetic and geo-
graphic variation in rejection behavior of cuckoo eggs by European magpie
populations: An experimental test of rejecter-gene flow. Evolution, 53,
947–956.
Soler, M, M. Martín-Vivaldi, & T. Pérez-Contreras. (2002). Identification of the sex
responsible for recognition and the method of ejection of parasitic eggs in some
potential common cuckoo hosts. Ethology, 108, 1093–1101.
Spottiswoode, C. N. (2010). The evolution of host-specific variation in cuckoo egg-
shell strength. Journal of Evolutionary Biology, 23, 1792–1799.
Spottiswoode, C. N. & M. Stevens. (2010). Visual modeling shows that avian host
parents use multiple visual cues in rejecting parasitic eggs. Proceedings of the
National Academy of Sciences, USA, 107, 8672–8676.
Stoddard, M. C. & M. Stevens. (2010). Pattern mimicry of host eggs by the common
cuckoo, as seen through a bird’s eye. Proceedings of the Royal Society of London,
B, 277, 1387–1393.
Stoddard, M. C. & M. Stevens. (2011). Avian vision and the evolution of egg color
mimicry in the common cuckoo. Evolution, 65, 204–213.
Stokke, B. G., A. Moksnes, & E. Røskaft. (2002). Obligate brood parasites as selec-
tive agents for evolution of egg appearance in passerine birds. Evolution, 56,
199–205.
Stokke, B. G., G. Rudolfsen, A. Moksnes, & E. Røskaft. (2004). Rejection of conspe-
cific eggs in chaffinches: The effect of age and clutch characteristics. Ethology,
110, 459–470.
Stokke, B. G., F. Takasu, A. Moksnes, & E. Røskaft. (2007). The importance of
clutch characteristics for antiparasite adaptations in hosts of avain brood para-
sites. Evolution, 61, 2212–2228.
Svennungsen, T. O. & Ø. H. Holen. (2010). Avian brood parasitism: Information
use and variation in egg-rejection behaviour. Evolution, 64, 1459–1469.
268 ANIMAL BEHAVIOR

Takasu, F., K. Kawasaki, H. Nakamura, J. Cohen, & N. Shigesada. (1993). Modeling

the population dynamics of a cuckoo-host association and the evolution of
host defenses. American Naturalist, 142, 819–839.
Takasu, F., & C. Moskát. (2011). Modeling the consequence of increased host toler-
ance toward avian brood parasitism. Population Ecology, 53, 187–193.
Takasu, F., C. Moskát, A. R. Munoz, S. Imanishi, & H. Nakamura. (2009). Adapta-
tions in the common cuckoo (Cuculus canorus) to host eggs in a multiple-hosts
system of brood parasitism. Biological Journal of the Linnean Society, 98,
291–300.
Thompson, J. N. (1994). The Coevolutionary Process. Chicago: University of Chicago
Press.
Trine, C. L. (2000). Effects of multiple parasitism on wood thrush nesting success. In
J. N. M. Smith, T. Cook, S. I. Rothstein, S. K. Robinson, & S. G. Sealy (eds.),
Ecology and Management of Cowbirds and Their Hosts (pp. 135–144). Austin:
University of Texas Press.
Underwood, T. J. & S. G. Sealy. (2006). Parameters of brown-headed cowbird
Molothrus ater egg discrimination in warbling vireos Vireo gilvus. Journal of
Avian Biology, 37, 457–466.
Vikan, J. R., B. G. Stokke, F. Føssey, C. Jackson, E. Huhta, J. Rutila, A. Moksnes, &
E. Røskaft. (2009). Fixed rejection responses to single and multiple experi-
mental parasitism in two Fringilla hosts of the common cuckoo. Ethology,
115, 840–850.
Weaver, H. B. & C. R. Brown. (2004). Brood parasitism and egg transfer in cave
swallows (Petrochelidon fulva) and cliff swallows (P. pyrrhonota) in South
Texas. Auk, 121, 1122–1129.
White, D. J., L. Ho, & S. G. Freed-Brown. (2009). Counting chicks before they
hatch: Female cowbirds assess temporal quality of nests for parasitism. Psycho-
logical Science, 20, 1140–1145.
Yom-Tov, Y. (1980). Intraspecific nest parasitism in birds. Biological Reviews, 55,
93–108.
Yom-Tov, Y. (2001). An updated list and some comments on the occurrence of intra-
specific nest parasitism in birds. Ibis, 143, 133–143.
 10

How Animals Remember Places and Find

Their Way Around
David F. Sherry

INTRODUCTION
There are very few animals that do not move through space at some point in
their life, and many are in motion more or less constantly during their waking
hours. They leave places that they will later return to. They depart from places
that might be either novel or familiar in search of somewhere else, a somewhere
they may know well or may never have been before. Novelist Salman Rushdie
(1995) wrote “In my family . . . we arrive hoping for somewhere better,” and that
probably describes the lives of many animals. Animals find food only to move on
as they deplete the patch in which they are foraging and face better prospects
elsewhere. Many animals leave the place they were born and set out in search
of suitable habitat where they can establish a territory, find a mate, and repro-
duce. Animals defend home ranges and territories that they know well and will
expand their own territory quickly if a neighbor vacates a desirable area, as well
as respond quickly to intrusions across the boundary into their own territory.
Some animals leave their nest or burrow in search of food, which they then bring
back home before setting out again in search of more. Animals can find them-
selves in unfamiliar surroundings, faced with the problem of getting back home.
Most spectacularly, many animals set out on migratory journeys of thousands of
kilometers to reach more a favorable habitat for breeding or wintering, some-
times with no previous experience of either the destination or the route.
270 ANIMAL BEHAVIOR

In this chapter I will describe research that has been done with food-storing
birds and other animals that illustrates a few of the many ways animals remem-
ber spatial locations and navigate from one place to another, along with
describing some of the brain mechanisms that underlie spatial memory and
orientation. Many different methods of orientation and navigation are used
by animals (see Volume 1, Chapter 12). Some are reasonably well understood,
but for others it remains a mystery exactly how they work. There is, likewise, a
great deal of information about some brain mechanisms of memory and orien-
tation and relatively little about others. Remembering places, orienting, and
navigating are among the most fascinating things that animals do. The fascina-
tion comes from knowing for certain that animals do these things and the
realization that we have only a very incomplete understanding of how they
do it. We know, for example, that some small birds create thousands of scat-
tered food caches in the fall and winter and retrieve them by remembering
these unique spatial locations. We know that some birds leave the temperate
latitudes where they breed, travel thousands of kilometers to the tropics, and
return again to breed in the same place they nested last year. We know that
sea turtles leave the beach where they were born, spend years in the ocean,
and then return again to their natal beach, usually only briefly, to nest and
lay eggs. This chapter will look at some of the things that have been discovered
about orientation and memory for places on the relatively small spatial scale of
animals’ daily lives in searching for stored food, finding the way back to a nest,
and navigating through a familiar home range.

FOOD-STORING BIRDS
A number of birds familiar to even casual observers of animal behavior—
chickadees and tits, jays, and nuthatches—have a secret life. Some of the food
they find they do not eat right away but instead store to consume later.
Although many people are completely unaware of this behavior, it is actually
not difficult to observe if you carefully watch a chickadee taking food from a
backyard feeder in fall or winter. At that time of year, as days grow short and
food becomes more difficult to find, black-capped chickadees (Poecile atrica-
pillus), Carolina chickadees (P. carolinensis), and their European relatives
marsh tits (P. palustris) and willow tits (P. montanus) create thousands of
caches, each containing a single food item. Repeated quick visits by the same
bird to a feeder full of sunflower seeds mean that the bird is storing most of
what it is carrying away because opening and eating a sunflower seed usually
takes a chickadee 50 seconds or more (Boisvert & Sherry, 2000). Intervals
between visits that are shorter than this mean the chickadee has stored the
HOW ANIMALS REMEMBER PLACES AND FIND THEIR WAY AROUND 271

seed, which it can do very quickly, before coming back for more. Chickadees
and tits will cache seeds, pieces of nut, insects, indeed anything they normally
eat, in any small crevice in tree bark, or under leaves, or in moss on the
ground. They will also readily cache food in captivity, and this has led to a
great deal of research on a basic question in animal spatial memory: how do
food-storing animals find their caches again? This research has not only
addressed an old question about whether or not food-storing animals ever find
their caches again, it has also led to new discoveries about spatial memory and
orientation (Shettleworth & Hampton, 1998; Gould et al., 2010; Sherry &
Hoshooley, 2010; Smulders et al., 2010).

Memory and Behavior

Chickadees and tits clearly find their way back to their caches. Allen Stevens
and John Krebs proved this by equipping marsh tits with a small magnet on
the bird’s leg band (Stevens & Krebs, 1986). Birds were offered peanuts
labeled with technetium 99, a radioactive isotope with a half-life of six hours.
When the researchers found a cache site—by searching the bird’s winter
territory with a portable scintillation counter—they placed a small detector
at the cache site, connected to a battery-powered alarm clock. The detector
was sensitive to changes in the local magnetic field, and when the bird with
the magnet on its leg band returned to collect its cached food, it triggered
the detector and stopped the clock. The majority of cache retrieval attempts
by magnet-carrying birds occurred within 12 hours of daylight following cach-
ing, and these attempts were 90 percent successful. Retrieval at longer intervals
was less successful because stored seeds had often been taken by this time by
other birds or by rodents.
It is one thing to show that birds such as marsh tits return and collect the
food they have stored, as do many other food-storing birds and even squirrels
(Jacobs & Liman, 1991). It is another thing to discover how they do it. In
general, the problem the food storers face is how to return to a very specific
location in space that was likely visited only once at the time the cache was
made. Chickadees and tits never, so far as we know, reuse the same cache site
in the wild. Caching usually occurs, however, in an area the animal is familiar
with: a winter home range in the case of chickadees and tits, or a preferred
caching area in the case of another prodigious food storer, the Clark’s nut-
cracker (Nucifraga columbiana) (Gould et al., 2010). But in either case, the
general area is familiar. The task is to relocate, after some period of time, a very
specific (and small) spatial location somewhere in an area of at least several
hectares and sometimes much larger. A number of studies have shown that
an important cue is familiar local landmarks.
272 ANIMAL BEHAVIOR

Landmarks
Stephen Vander Wall let Clark’s nutcrackers store pine nuts in a large
aviary (Vander Wall, 1982). Nutcrackers bury their caches by poking their bill
into the ground, opening the bill to create a hole, and letting seeds that are car-
ried in a pouch under their tongue fall into the hole. The floor of the aviary in
Vander Wall’s experiment had a scattered array of rocks and logs. After a nut-
cracker had placed its caches, Vander Wall moved all of the objects in one half
of the aviary 20 cm in the same direction. When the bird returned to search
for food, its searches were accurate on the side of the aviary where nothing
had changed, but on the other side of the aviary searches were off by 20 cm
in the same direction that the objects had been moved. In the middle, where
birds might be expected to have used as landmarks some objects that were later
moved and some that stayed in place, birds missed their caches by intermedi-
ate distances between 0 and 20 cm. The nutcrackers had learned and remem-
bered where their caches were with respect to landmarks.
In an experiment with black-capped chickadees, Rachel Herz, Liana
Zanette, and I allowed birds to store food in an artificial environment con-
tained inside a plain 2 × 2 × 2 meter room (Herz et al., 1994). An observer
could watch the birds from outside this room through one of four dark Plexi-
glas windows, one in each of the four walls. The bird could be introduced to
the room through a small port in each window. Inside were four symmetrical
artificial trees with branches extending in four directions at right angles. Each
branch had small holes drilled in it where the birds could store seeds. The trees
were designed so that 90° and 180° rotations of a tree clockwise or counter-
clockwise would look the same as the original orientation of the tree. Small
colored cards were placed on the branches near each cache site as local land-
marks, and large colored shapes and posters were placed on the walls of the
room as distal landmarks. After the birds had stored food, they were removed
from the room for four hours. When they came back to search for their caches
the room was in one of four conditions: either local landmarks had been
removed, or distal landmarks had been removed, or all landmarks had
been removed, or no landmarks had been removed. Stored food had also been
removed, so we were confident birds could not see or smell any cached food.
Removal of distal landmarks reduced the birds’ search accuracy, but removal
of local landmarks did not. Removal of both kinds of landmarks also reduced
search accuracy. These results show that the birds rely more on distal land-
marks, which in this environment were a meter or two from the cache sites,
than on local landmarks right at the cache site. Because the port through
which the bird was admitted to the room was selected at random, birds could
not have used this as a cue to the location of caches. Similarly, the window
HOW ANIMALS REMEMBER PLACES AND FIND THEIR WAY AROUND 273

through which the observer watched was selected at random, so any move-
ment or sound from the observer that the birds could detect could not be used
as a cue either.
We performed a second experiment in which only distal landmarks were
present during caching. After birds had cached food we rotated the distal land-
marks by 90° (either clockwise or counterclockwise) and by 180°. Birds
searched at sites that were correct with respect to the new locations of the land-
marks rather than at the sites where they had actually cached. These results,
like Vander Wall’s (1982), show that food-storing birds use landmarks near
their caches to identify cache locations.
In hindsight, this makes a lot of sense. Small local landmarks right at the
cache site are more likely to be displaced or covered with snow or fallen leaves
than larger landmarks at a distance. In addition, small local landmarks need to
be found before they can be used. Finding them may be as difficult as finding
the cache site itself. Large distal landmarks are more useful because they are
less likely to be disturbed or covered and can be seen from a distance. But
how does a large object a meter or so from a cache site indicate a location?
In order for landmarks to work, the animal must remember two things about
the relation between the landmark and the place it is searching for: distance
and direction. An asymmetrical landmark can provide direction information.
Imagine a fallen log with a branch protruding on one side. Suppose the cache
is on the same side of the fallen log as the protruding branch. It might lie
somewhere along a line pointing off at a 90° angle from the branch side of
the log. But where along this line? Direction alone does not specify a location
in space. A landmark can also give distance information. The cache site might
be about 1.5 m from fallen log. But in which direction? Distance alone is not
enough information either. Combining direction and distance, however, pin-
points the cache site. It lies 1.5 m away from fallen log on a line extending at
an angle of 90° from the branch side. A quantity that has both direction and
magnitude is a vector, and so using a landmark requires remembering a vector:
a direction and a distance. This is not to say that animals consciously remem-
ber distance and direction in the way that we do, or analyze the problem the
way we would, but we know animals can accurately relocate places using land-
marks (Collett et al., 2007), and that means they must have some representa-
tion in their brain or nervous system of the vector between the landmark and
the place they are looking for.
In practice, remembering vectors from landmarks probably involves error
in estimating and remembering both the distance and direction, so using
multiple landmarks, each with its own vector, makes landmark use more accu-
rate. The fact that nutcrackers missed their caches in the middle of the aviary
274 ANIMAL BEHAVIOR

in Vander Wall’s experiment by intermediate distances suggests they used

more than one landmark. Intersection of multiple vectors would normally give
a more accurate determination of a place than using just one vector.

A Global Reference System

Landmark use is not the only spatial strategy used by food-storing birds to
relocate caches. Many animals can orient to properties of the environment
that, unlike a landmark, are available practically everywhere on Earth. These
properties of the physical world do not pinpoint a location in the way that
landmarks do but instead provide a reference for orientation. The magnetic
field of the earth is one such physical feature. Another is the regular movement
of the sun through the sky. Food-storing birds (and many other animals)
obtain compass-like information from the position of the sun, and so use of
the sun to determine a bearing is known as sun-compass orientation.
The sun can be used as a compass because it rises on the eastern horizon,
traverses the sky, and sets in the west. Birds apparently ignore the height of
the sun in the sky and instead use a property of celestial objects called the azi-
muth. The azimuth is the point on the horizon that would be intersected by a
line dropped vertically from the object. As the sun travels through the sky, its
azimuth marches along the horizon, starting in the east, moving through the
south (for an observer in the northern hemisphere), and ending up in the west.
If the animal has information about time of day, then the sun’s azimuth can be
a very effective compass. The dawn position of the sun’s azimuth is in the east,
its noon azimuth is in the south (in the northern hemisphere) and its sunset
azimuth is in the west. The path of the sun through the sky and hence the
daily change in its azimuth depends on the season and the observer’s location
of the surface of the earth. In the southern hemisphere, the noon position of
the sun is in the north, but the idea is the same. On the equator at the spring
and autumn equinox, the sun climbs vertically in the sky so that its azimuth is
due east from sunrise until about noon, when its azimuth shifts 180° to due
west as the sun passes through the zenith and then maintains a due west azi-
muth until sunset. A solar azimuth that behaves in this way is not much use
for determining north and south, but over most of the earth’s surface at most
times of year the sun’s azimuth moves progressively along the horizon during
the day, providing reliable compass information for an animal with an internal
clock that can be used to determine time of day.
The sun compass is used by migratory birds and homing pigeons (Columba
livia) to determine bearing, but somewhat surprisingly it is also used by food-
storing birds for orientation on a much smaller spatial scale (Wiltschko &
Balda, 1989). We can test whether an animal is using the position of the sun
HOW ANIMALS REMEMBER PLACES AND FIND THEIR WAY AROUND 275

to obtain compass information by experimentally altering either the position

of the sun or the time of day. Although a Biblical account describes the sun
stopping for a day until Joshua could finish off his enemies, “So the sun stood
still in the midst of heaven, and hasted not to go down about a whole day”
(Joshua 10:13), experimentally altering the position of the sun is obviously
fanciful. The practical test of whether or not an animal is using the sun com-
pass for orientation is to provide misleading information about time of day,
and this is quite easy. For a bird held in captivity, the time the lights come
on and go off each day can be adjusted, imposing a clock shift on the animal.
Such a clock shift is what we experience as jet lag. Eventually our biological
circadian rhythms adjust, as do birds’, at a rate of about one hour per day.
After six days, birds adjust their daily pattern of activity to a six-hour clock
shift.
Wolf Wiltschko and Russell Balda (1989) allowed scrub jays (Aphelocoma
coerulescens) to cache seeds in an outdoor aviary with a clear view of the sky.
Birds could see surrounding landmarks outside the cage. After the birds had
cached food, they were brought indoors and experienced a six-hour clock shift.
A six-hour clock shift should produce an error in compass bearing of 90º. Six
hours is one-quarter of the 24-hour day, and a six-hour clock shift should
therefore cause an equivalent error in the 360º range of possible compass bear-
ings, which would equal 90º. This was, indeed, the effect of the six-hour clock
shift on the scrub jays when they were tested again in the outdoor aviary.
Their cache searching was rotated by approximately 90º.
Sarah Duff, Lesley Brownlie, Mark Sangster, and I repeated this experiment
with a different food-storing bird, the black-capped chickadee, tested outdoors
on the roof of a nine-story building with a clear view of the sky and a clear
view of surrounding landmarks outside their cage. We obtained similar results
(Duff et al., 1998). The outcome of these two experiments is remarkable. It
means that birds that can see local landmarks, which we know they use to
retrieve their caches, still deviate in their search away from the actual sites of
caches and toward the locations indicated by the sun compass. It is as if a
six-hour jet lag caused us to walk past our front door and continue on down
the street, off course by 90º. We wanted to know more about how chickadees
integrate sun-compass and landmark information, so we repeated our experi-
ment but this time tested the birds, following a six-hour clock shift, with their
cage in either the same location as it had been when they made their caches or
in a new site where none of the landmarks were familiar. At the site where they
could see familiar landmarks, the clock shift caused a rotational error as before.
At the new site where none of the landmarks were familiar, birds searched at
random. This means that sun-compass information, which could have been
276 ANIMAL BEHAVIOR

used to search for caches at the new site (with a deviation error of 90º), was
not used at all at the new site, probably because there were no familiar land-
marks. Chickadees are not able to rely exclusively on sun-compass information
to find their caches but instead seem to use it to orient an array of familiar
landmarks, imposing a compass direction on the familiar landscape. A refer-
ence system used in this way assigns a polarity to the familiar array. We may,
for example, be familiar with the landmarks in our local park but also have
some sense that the end of the park with the fountain is north. The layout
of landmarks in the park would be exactly the same if the park was to mysteri-
ously rotate by 180º, but it would be polarized differently, with the fountain
now in the south. This may be how chickadees integrate sun-compass and
landmark information, using local landmarks to find caches while at the same
time polarizing the environment using sun-compass information. Assigning
polarity to familiar areas may help animals move from one familiar area to
another because they can use the polarizing cue to align this mosaic of familiar
areas (Wiltschko & Balda, 1989). If birds are clock shifted they make system-
atic errors because landmarks and the polarizing cue are in conflict. In a novel
location, as in our experiment with chickadees, there are no familiar landmarks
to guide their search for caches. In terms of landmarks, the place they are
searching does not look at all like the place where they hid their food, and hav-
ing a polarizing cue for this unfamiliar landscape does not help.

NAVIGATION BY PATH INTEGRATION

Animals use vectors not only to determine the relation between a landmark
and a place but also in a type of navigation called path integration. Path inte-
gration is a particularly interesting form of navigation because, at least in prin-
ciple, it can be used in the absence of any landmarks or global reference system.
The desert ant Cataglyphis fortis lives in the Sahara and forages in a relatively
featureless landscape for prey, mostly dead invertebrates. Despite a circuitous
outward path, Cataglyphis is able to return to its nest by a straight-line direct
route once it finds food (Müller & Wehner, 1988). It returns home by follow-
ing a vector, a distance and direction from its current location to the start of its
outward journey. This can be proven by interrupting the ant during its home-
ward journey and displacing it. When released, the ant completes its vector,
but because of the displacement travels a path parallel to its initial homeward
route. When it reaches the end point of the vector it searches the immediate
area for a nest that is actually elsewhere (Müller & Wehner, 1988).
How does an ant that has followed a twisting path outward from its nest
determine the correct vector to get back home? The solution is path
HOW ANIMALS REMEMBER PLACES AND FIND THEIR WAY AROUND 277

integration, a method of integrating all of the turns it makes and the length of
its straight paths between turns to keep a running record of the correct dis-
tance and direction home. When Cataglyphis makes a turn it can determine
the magnitude of the turn by the change that occurs in its orientation with
respect to the sun and the pattern of polarized light in the sky. Polarized sky-
light is used by many insects for navigation. Regions of the compound eye that
point skyward are specialized for the detection of polarized light. “Polarized”
referring to light is very different from the meaning of the term used earlier
to describe an environment that is polarized by a compass cue (Pye, 2001).
Light is a wave form oscillating at right angles to the path the light travels.
Normally, light consists of waves oscillating at all possible right angles to the
direction of travel. Polarized light consists of light with oscillation of its wave
form confined to a single plane at a right angle to the direction of travel, the
result of passing through a polarizing filter or reflecting off some kinds of sur-
faces. Sunlight becomes polarized by passing through the atmosphere. The
degree of polarization in a clear sky depends on the position of the sun. In
the direction of the sun, sky light is not polarized, but its polarization increases
away from the sun, reaching a maximum at 90º from the sun. For animals that
can see polarized light, like insects, the pattern of skylight polarization is used
as a global reference system. Sun-compass orientation in insects is often
achieved by orientation to these patterns of polarization in the sky. The proof
of this is that insects like honey bees can determine a compass bearing even if
they cannot see the sun, only a patch of clear sky. On an overcast day, when
they cannot see sky-light polarization either, they cannot use this source of
compass information.
Determining the magnitude of its turns with respect to polarized sky-
light and weighting these turns by the length of the path travelled between
turns gives Cataglyphis enough information to come up with a homeward
vector that is approximately correct. Martin Müller and Rudiger Wehner
(1988) showed that a simple model that weighted turns in this way pro-
duced not only the same homeward bearing the ants followed but also pro-
duced the same deviation errors from the correct homeward route that the
ants made.
Experiments with homeward-bound ants also show that manipulating the
position of the sun is not as fanciful as I said earlier. A century ago, Felix Sant-
schi showed that ants use the position of the sun for orientation by preventing
ants from seeing the actual sun and reflecting an image of the sun back at the
ant with a mirror (Santschi, 1911). The mirror caused an apparent 180º shift
in the location of the sun from the ant’s point of view, and the result was a
180º reversal of the ant’s path.
278 ANIMAL BEHAVIOR

NEUROBIOLOGY OF SPATIAL MEMORY AND ORIENTATION

The various mechanisms of spatial orientation described above—landmark
use, the sun compass, and path integration—are only possible because the
brain and nervous system are collecting information, processing and integrat-
ing this information, and issuing appropriate motor instructions for the ani-
mal to move in a particular direction, turn, search, reverse direction, and
execute all the other components of getting from one place to another. Some-
times the processing also involves retrieving information from memory about
landmarks, previous routes, or the outcome of previous searches. There has
been extensive research on the neural mechanisms of spatial memory and ori-
entation. Much has been learned, but much remains to be discovered. The
next section will discuss some of things that have been discovered about a part
of the brain called the hippocampus. The hippocampus is found in birds and
mammals (including humans), and a region that is probably its evolutionary
precursor is found in the dorsomedial area of the reptile brain. The human
hippocampus serves a variety of cognitive functions, and damage to the hippo-
campus usually results in “anterograde” amnesia. People with this kind of
amnesia have deficits in their ability to remember things that have happened
since their hippocampus was damaged. They may have little or no “retro-
grade” amnesia, which is the inability to remember things that happened
before their hippocampus was damaged. This has led to the general view that
the hippocampus is involved in processing experience in order to remember it,
while the long-term storage of memory takes place elsewhere in the brain.
Whereas the human hippocampus plays a role in memory for many kinds
of information, the hippocampus of animals has been examined primarily for
the role it plays in spatial memory. This is not to say that the hippocampus
of animals serves no other functions—it probably does—but it clearly plays
a major role in spatial memory.

The Hippocampus of Food-storing Birds

Black-capped chickadees with their hippocampus removed continue to
cache food and search for it, but they look in the wrong places (Sherry & Vac-
carino, 1989). They behave as if they remember that they should search for
caches but cannot remember where to look. We also trained chickadees to
search for food they had not stored. Food locations were always in the same
place for some birds. For other birds, food was always found in different pla-
ces, but the locations that contained food were marked with small colored
cards and empty places were marked with cards of a different color. Chicka-
dees with their hippocampus removed could not find food when they had to
HOW ANIMALS REMEMBER PLACES AND FIND THEIR WAY AROUND 279

remember spatial locations but could find food normally when they had to
remember the significance of the colored cards (Sherry & Vaccarino, 1989).
This result is very similar to other results, with birds and mammals, showing
that damage to the hippocampus has a selective effect on memory for spatial
locations and does not disrupt learned associations between a cue and the pres-
ence of food.
Food-storing birds, interestingly, have a hippocampus that is larger than the
hippocampus of birds that do not store food (Krebs et al., 1989; Sherry et al.,
1989). Birds in the chickadee and tit family, the Paridae, and birds in jay and
crow family, the Corvidae, which includes the Clark’s nutcracker, vary a great
deal in body size and hence total brain size. But they have consistently larger
hippocampuses, relative to brain size, than birds in families in which food stor-
ing does not occur (Lucas et al., 2004). It seems that the ecological pressures
favoring food storing in certain groups of birds have led to natural selection
for good spatial memory and orientation abilities, which has in turn modified
the brain regions supporting these cognitive abilities.

Hippocampal Neurogenesis
For many years the adult brain was thought not to produce new brain cells.
The belief was that once adulthood was reached, no new neurons were added
to the adult brain. Although there had been previous indications that this
might be a mistaken view (Altman, 1962), it was research on the song-
control system of birds that eventually proved that neurogenesis can occur
in adulthood (Goldman & Nottebohm, 1983; Nottebohm, 2004). New neu-
rons are also added in adulthood to the hippocampus of food-storing birds
(Barnea & Nottebohm, 1994). Food-storing birds add far more new neurons
to their hippocampus than non-food-storing birds do, and there is a seasonal
pattern in the recruitment of new neurons to the hippocampus of food-
storers that is not found in non-food-storers (Hoshooley & Sherry, 2007).
Although there can be variation in the exact seasonal timing of hippocampal
neurogenesis in food-storers, it tends to be correlated with food storing itself,
occurring at a higher frequency in fall and winter than in summer (Sherry &
Hoshooley, 2010).

Place Cells, Grid Cells, and Head Direction Cells

The way neurons in the hippocampus and elsewhere in the brain represent
space has been extensively examined in research with rodents, usually labora-
tory rats (Rattus norvegicus). Some neurons in the hippocampus are “place”
cells. In general, neurons communicate with other neurons by discharging
280 ANIMAL BEHAVIOR

an electrical signal, the action potential. Place neurons in the hippocampus

fire action potentials when the animal is in a particular place, usually a region
in the enclosure in which they are being observed. Each place cell has its own
place field, a region where that neuron produces action potentials. Elsewhere
in the enclosure, the neuron is quiet. Place fields come in all shapes and sizes,
sometimes in a corner of the enclosure, sometimes along a wall or in open
space in the center of the enclosure. Place cells can have two or more separate
regions where the neuron is active. It does not usually matter from which
direction the animal enters the preferred place or how it got there, only that
it is in the place field of the cell.
In another part of the brain, the entorhinal cortex, there are truly remark-
able cells called grid cells. The entorhinal cortex is a major source of input to
the hippocampus. Grid cells are cells that fire in a regular hexagonal pattern.
That is, when all the locations in which a grid cell fires are plotted on a map
of the animal’s enclosure, the plot forms a regular array. Such an array could
be the basis for a map of space in the brain. In addition, there is further type
of cell, in the subiculum, known as head-direction cells. These cells fire when
the animal is facing in a particular direction. Different head-direction cells
have different preferred head directions. Taken together, place cells, grid cells,
and head-direction cells could be part of a very sophisticated spatial mapping
system in the rat’s brain providing the animal with information about where
it is, what direction it is facing, and where on a regular grid of space it finds
itself. These cells are sensitive to the shape of the animal’s environment, and
place cells change their place fields if the shape of the environment is changed
or its overall size is increased or decreased. Head-direction cells normally
maintain the same head direction when the animal moves from one environ-
ment to another, but if a familiar environment is modified or reoriented,
head-direction cells will respond to these changes with a change in direction.

SEX DIFFERENCES IN SPATIAL ABILITY

It is a widely held belief that men never ask for directions. Accurate or not,
this is one of many claims that the sexes differ in spatial ability. For people, it
is very hard to determine whether men and women differ in spatial ability
because of some attribute that accompanies being male or female, like the sec-
ondary sexual characteristics, or whether it is the result of men and women
being treated differently from the moment they are born. In animals, however,
sex differences in spatial ability are unlikely to be the result of cultural beliefs
about gender roles. Furthermore, studies of animals show that it is sometimes
males and sometimes females that perform better on tests of spatial ability.
HOW ANIMALS REMEMBER PLACES AND FIND THEIR WAY AROUND 281

Meadow voles (Microtus pennsylvanicus) are polygynous. Males defend

home ranges that encompass the home ranges of more than one female. Males
mate with females but provide no parental care for the young. The more
females a male mates with, the more offspring he will leave, and so male
meadow voles are essentially competing with each other by attempting to
include in their home range as many female home ranges as they can. The
male that includes more females in his home range will leave more offspring
that will possess the traits that allowed him to occupy a large home range,
and one of those traits is spatial ability. Pine voles (Microtus pinetorum) and
prairie voles (Microtus ochrogaster) are, in contrast, monogamous. Males and
females live together in the same home range, and both males and females pro-
vide parental care.
Steve Gaulin and his colleagues compared the home-range size of male and
female meadow voles and pine voles and found that, indeed, male meadow
vole home ranges are much larger than those of females, while male and female
pine vole home ranges are about the same size (Gaulin & FitzGerald, 1986).
When tested on mazes in the laboratory, male meadow voles performed better
than female meadow voles, but male and female pine voles did not differ.
Next, they compared male and female home-range size and spatial ability in
meadow voles and the other monogamous species, the prairie vole. As before,
male meadow vole home-range size was much larger than female home-range
size, while the home-range sizes of male and female prairie voles did not differ.
This time, in the laboratory, mazes were cleverly constructed so that once a
vole had run from the start to the end of the maze, it could run back through
the maze again, and the same pattern of turns it used the first time would take
it back to the start. The walls of the maze could be altered to form many differ-
ent mazes with this same property of reversibility. The researchers found that
male meadow voles performed better than female meadow voles, while male
and female prairie voles performed about the same (Gaulin & FitzGerald,
1989). Finally, they compared the size of the hippocampus in males and
females relative to the size of the rest of the brain. The hippocampus was larger
in male meadow voles than in females and about the same size in male and
female pine voles (Jacobs et al., 1990). In the polygynous species, in which
males compete with each other by expanding their home-range size, males
had larger home ranges, better spatial ability, and a larger hippocampus than
females. In the monogamous species, males and females had the same-sized
home range, equal spatial ability, and the same size of hippocampus. As with
food-storing birds, it appears that natural selection for spatial ability has pro-
duced a change in a brain region involved in spatial ability, in this case exerting
its effect selectively on one sex and not the other.
282 ANIMAL BEHAVIOR

Sex differences in spatial ability, however, do not always favor males. It is

the role spatial ability plays in the lives of males and females that influences
spatial ability, not sex per se. Brown-headed cowbirds (Molothrus ater) are
brood parasites. Females lay their eggs in the nests of other birds and leave
the eggs for these “hosts” to incubate. Host parents feed the young cowbird
after it hatches, usually to the detriment of their own young because cowbird
chicks are large, aggressive, and grow rapidly. Female cowbirds lay their eggs
at dawn, or a little before dawn, and so are going directly to a host nest at
the time of egg laying without taking the time to search for a host nest. The
nest must have been discovered at least a day earlier and possibly much earlier.
Some research indicates that female cowbirds may make repeated trips to a
host nest to check on the laying progress of the host female (White et al.,
2009). This is important to female cowbirds because they must add their egg
to the host clutch at the right time. Too early, and the hosts may just abandon
the nest and start another. Too late, and the host young may hatch and leave
the nest before the cowbird egg hatches. Research by David White and col-
leagues shows that in fact female cowbirds prefer to lay in a nest to which
the host female is still adding eggs. A clutch to which the host female is still
adding eggs means that the hosts have not yet started incubation and develop-
ment of the young inside the eggs has not begun. Female cowbirds thus search
out, find, and remember the locations of potential host nests before they need
them. It is not clear how many potential host nests a female cowbird carries in
memory at any one time, but it is certainly a few and possibly many. Male
brown-headed cowbirds do not search for host nests. Males and females feed
together in grain fields and livestock yards in the latter part of the day, but
males do not participate in finding or remembering the locations of host nests.
Female brown-headed cowbirds have a larger hippocampus than males, a
sex difference not found in other closely related nonparasitic blackbirds
(Sherry et al., 1993). Like meadow voles, sex-specific selection has produced
a difference in the brain, but this time favoring females, the sex that probably
makes greater use of the hippocampus for remembering the locations of host
nests or maybe tagging remembered nest locations with additional informa-
tion like the stage of completion of the host clutch.
The group of blackbirds to which brown-headed cowbirds belong has
many members in South and Central America. One, the shiny cowbird
(Molothrus bonariensis), is a generalist parasite like the brown-headed cowbird.
Brown-headed cowbirds place their eggs in the nests of over 200 different spe-
cies, and over 100 of these are known to successfully raise brown-headed cow-
bird chicks. Shiny cowbirds place their eggs in the nests of about the same
number of species, over 70 of which are known to successfully raise shiny
HOW ANIMALS REMEMBER PLACES AND FIND THEIR WAY AROUND 283

cowbird chicks. Female shiny cowbirds search for host nests while males do
not. Another cowbird, the screaming cowbird (Molothrus rufoaxillaris), is a
specialist parasite, using only one species as its host. Male and female scream-
ing cowbirds search together for nests of their host. The host of the screaming
cowbird is a cowbird, the bay-winged cowbird (Agelaioides badius), which is
not a brood parasite at all. Bay-winged cowbirds build their own nests, incu-
bate their own eggs, and raise their own young, along with the occasional
screaming cowbird chick. Female shiny cowbirds have a larger hippocampus
than males, as might be expected from the sex difference in searching for host
nests. Hippocampus size does not differ between the sexes in either screaming
cowbirds, because both sexes search for host nests, or in bay-winged cowbirds,
because neither sex searches for host nests (Reboreda et al., 1996).

TAGGING SPATIAL LOCATIONS WITH OTHER INFORMATION

Spatial location is often just one part of an event. “Where” can be part of a
memory that also includes what happened at that specific place and when.
Memory of this kind is sometimes called episodic or episodic-like memory
because it includes information about what, where, and when. In humans,
episodic memory is accompanied by a sense of having personally experienced
an event and being able to consciously recollect the episode. Psychologist
Endel Tulving said that human episodic memory is “autonoetic,” from two
Greek root words, auto meaning self and noesis meaning thought (Tulving,
1972). Episodic memory in people is autonoetic because it is “self-knowing.”
Whether animals are capable of conscious recollection of their own experience
is a controversial area with many arguments pro and con. Similarly, whether
memory for what, where, and when in animals is really similar to human epi-
sodic memory is unresolved. Animals can, however, clearly remember what,
where, and when together. Nicky Clayton and Anthony Dickinson trained
food-storing scrub jays (Aphelocoma coerulescens) to store two kinds of food
in ice-cube trays (Clayton & Dickinson, 1998). One type of food, waxworms,
was preferred to the other type of food, peanuts. Birds were trained, however,
that after a long interval of 124 hours, waxworms went bad while peanuts were
always good to eat. Birds were allowed to store both types of food but then
were tested after either a short interval of 4 hours or a long interval of
124 hours. If birds could remember what, where, and when, they should be
able to remember whether they had stored waxworms a short time ago or a
long time ago, as well as in which part of the ice-cube tray they had placed
each type of food. The results showed that after 4 hours birds consistently
searched for waxworms in the correct cache sites, but after 124 hours
284 ANIMAL BEHAVIOR

(by which time the waxworms had gone bad) they retrieved peanuts from the
correct cache sites. These tests were conducted with neither type of food
actually present in the caches so the birds could not detect the food and had
to use memory for what they had hidden where. Scrub jays could remember
what (waxworms or peanuts), where (in which part of the ice-cube tray), and
when (4 or 124 hours ago). Similar experiments have now been performed
with other animals, and some, but not all, produce similar results. Magpies
(Pica pica), black-capped chickadees, and rats can all solve the what-where-
when problem (Babb & Crystal, 2005; Feeney et al., 2009; Zinkivskay
et al., 2009), while some primates seem unable to (Hampton et al., 2005).
Research on memory for what, where, and when shows that memory for
spatial location does not exist in a vacuum. Places have rich associations for
us, and for animals, too, they are probably much more than just a point in
space. Some places have food, others none. Some places have greater risk of
exposure to predators than others. Some places are far off and are visited only
rarely, perhaps to obtain a rare nutrient or to lay eggs and then depart again,
like nesting sea turtles. Some places are so familiar animals may not even pay
attention to them. Rodents and bats will crash into obstacles they can easily
detect if these obstacles are placed to block a familiar route that the animals
use regularly (Stamps, 1995).

CONCLUSIONS
We have barely scratched the surface of research on spatial memory, orienta-
tion, and navigation. Our understanding of the brain mechanisms of memory
and spatial ability has grown rapidly in the past decade as new methods of
investigating brain activity in both animals (Zapka et al., 2010) and people
(Maguire et al., 2000) have become available. A growing consensus among
researchers is that animals probably use a great many different means of orien-
tation and navigation, either simultaneously or in a hierarchical fashion
matched to the spatial scale of the orientation problem they are solving
(Bingman & Cheng, 2005). New information is continually becoming avail-
able on old questions, like how animals use landmarks, a question that Niko
Tinbergen, one of the founders of the modern scientific study of animal behav-
ior, investigated in the 1930s (Tinbergen, 1932). New questions are being
raised as we gain greater awareness of the navigational prowess of animals and
the previously unknown sensory systems they use to solve them (Fleissner
et al., 2003; Keary & Bischof, 2012). Some long-distance movement may be
solvable by very simple means: compute a vector and follow it. Some local
movement may involve complex transformations of grid-like representations
HOW ANIMALS REMEMBER PLACES AND FIND THEIR WAY AROUND 285

of the environment, neural representations of place and direction, and computa-

tions of appropriate motor output to initiate movement. The study of spatial abil-
ity, orientation, and navigation is likely to pose fascinating and challenging
problems for researchers and students of animal behavior for many years to come.

ACKNOWLEDGMENTS
I would like to thank the Natural Sciences and Engineering Research
Council of Canada for research support, and the many students and colleagues
with whom I have collaborated in investigating the spatial ability of food-
storers, brood parasites, and other animals.

REFERENCES AND SUGGESTED READING

Altman, J. (1962). Are neurons formed in the brains of adult mammals? Science, 135,
1127–1128.
Babb, S. & J. Crystal. (2005). Discrimination of what, when, and where: Implications
for episodic-like memory in rats. Learning and Memory, 36, 177–189.
Barnea, A. & F. Nottebohm. (1994). Seasonal recruitment of hippocampal neurons
in adult free-ranging black-capped chickadees. Proceedings of the National
Academy of Sciences, USA, 91, 11217–11221.
Bingman, V. P. & K. Cheng. (2005). Mechanisms of animal global navigation: Com-
parative perspectives and enduring challenges. Ethology, Ecology and Evolution,
17, 295–318.
Boisvert, M. J. & D. F. Sherry. (2000). A system for the automated recording of feed-
ing behavior and body weight. Physiology and Behavior, 71, 147–151.
Clayton, N. S. & A. Dickinson. (1998). Episodic-like memory during cache recovery
by scrub jays. Nature, 395, 272–274.
Collett, T. S., P. Graham, & R. A. Harris. (2007). Novel landmark-guided routes in
ants. Journal of Experimental Biology, 210, 2025–2032.
Duff, S. J., L. A. Brownlie, D. F. Sherry, & M. Sangster. (1998). Sun compass orien-
tation and landmark orientation by black-capped chickadees (Parus atricapil-
lus). Journal of Experimental Psychology: Animal Behavior Processes, 24, 243–253.
Feeney, M. C., W. A. Roberts, & D. F. Sherry. (2009). Memory for what, where, and
when in the black-capped chickadee (Poecile atricapillus). Animal Cognition,
12, 767–777.
Fleissner, G., E. Holtkamp-Rötzler, M. Hanzlik, M. Winklhofer, G. Fleissner, N.
Petersen, & W. Wiltschko. (2003). Ultrastructural analysis of a putative mag-
netoreceptor in the beak of homing pigeons. Journal of Comparative Neurology,
458, 350–360.
Gaulin, S. J. C. & R. W. FitzGerald. (1986). Sex differences in spatial ability: An evo-
lutionary hypothesis and test. American Naturalist, 127, 74–88.
286 ANIMAL BEHAVIOR

Gaulin, S. J. C. & R. W. FitzGerald. (1989). Sexual selection for spatial-learning abil-

ity. Animal Behaviour, 37, 322–331.
Goldman, S. A. & F. Nottebohm. (1983). Neuronal production, migration, and dif-
ferentiation in a vocal control nucleus of the adult female canary brain. Pro-
ceedings of the National Academy of Sciences, USA, 80, 2390–2394.
Gould, K. L., D. M. Kelly, & A. Kamil. (2010). What scatter-hoarding animals have
taught us about small-scale navigation. Philosophical Transactions of the Royal
Society London, B, 365, 901–914.
Hampton, R. R., B. M. Hampstead, & E. A. Murray. (2005). Rhesus monkeys
(Macaca mulatta) demonstrate robust memory for what and where, but not
when, in an open-field test of memory. Learning and Motivation, 36, 245–259.
Herz, R. S., L. Zanette, & D. F. Sherry. (1994). Spatial cues for cache retrieval by
black-capped chickadees. Animal Behaviour, 48, 343–351.
Hoshooley, J. S. & D. F. Sherry. (2007). Greater hippocampal neuronal recruitment
in food-storing than in non-food-storing birds. Developmental Neurobiology,
67, 406–414.
Jacobs, L. F., S. J. C. Gaulin, D. F. Sherry, & G. E. Hoffman. (1990). Evolution of
spatial cognition: Sex-specific patterns of spatial behavior predict hippocampal
size. Proceedings of the National Academy of Sciences, USA, 87, 6349–6352.
Jacobs, L. F. & E. R. Liman. (1991). Grey squirrels remember the locations of buried
nuts. Animal Behaviour, 41, 103–110.
Keary, N. & H.-J. Bischof. (2012). Activity changes in zebra finch (Taeniopygia gut-
tata) brain areas evoked by alterations of the earth magnetic field. PLoS
ONE, 7, e38697.
Krebs, J. R., D. F. Sherry, S. D. Healy, V. H. Perry, & A. L. Vaccarino. (1989).
Hippocampal specialization of food-storing birds. Proceedings of the National
Academy of Sciences, USA, 86, 1388–1392.
Lucas, J. R., A. Brodin, S. R. de Kort, & N. S. Clayton. (2004). Does hippocampal
size correlate with the degree of caching specialization? Proceedings of the Royal
Society of London, B, 271, 2423–2429.
Maguire, E. A., D. G. Gadian, I. S. Johnsrude, C. D. Good, J. Ashburner, R. S. J.
Frackowiak, et al. (2000). Navigation-related structural change in the hippo-
campi of taxi drivers. Proceedings of the National Academy of Sciences, USA,
97, 4398–4403.
Müller, M. & R. Wehner. (1988). Path integration in desert ants, Cataglyphis fortis.
Proceedings of the National Academy of Sciences, USA, 85, 5287–5290.
Nottebohm, F. (2004). The road we travelled: Discovery, choreography, and signifi-
cance of brain replaceable neurons. Annals of the New York Academy of Sciences,
1016, 628–658.
Pye, D. (2001). Polarized Light in Science and Nature. Bristol, UK: Institute of Physics
Publishing.
Reboreda, J. C., N. S. Clayton, & A. Kacelnik. (1996). Species and sex differences in
hippocampus size in parasitic and non-parasitic cowbirds. Neuroreport, 7, 505–
508.
HOW ANIMALS REMEMBER PLACES AND FIND THEIR WAY AROUND 287

Rushdie, S. (1995). The Moor’s Last Sigh. Toronto: Vintage.

Santschi, F. (1911). Observations et remarque critique sur le mécanisme de l’orienta-
tion chez les fourmis. Revue Suisse de Zoologie, 19, 305–338.
Sherry, D. F., M. R. L. Forbes, M. Khurgel, & G. O. Ivy. (1993). Females have a
larger hippocampus than males in the brood-parasitic brown-headed cowbird.
Proceedings of the National Academy of Sciences, USA, 90, 7839–7843.
Sherry, D. F. & J. S. Hoshooley. (2010). Seasonal hippocampal plasticity in food-
storing birds. Philosophical Transactions of the Royal Society of London, B, 365,
933–943.
Sherry, D. F. & A. L. Vaccarino. (1989). Hippocampus and memory for food caches
in black-capped chickadees. Behavioral Neuroscience, 103, 308–318.
Sherry, D. F., A. L. Vaccarino, K. Buckenham, & R. S. Herz. (1989). The hippocam-
pal complex of food-storing birds. Brain Behavior and Evolution, 34, 308–317.
Shettleworth, S. J. & R. R. Hampton. (1998). Adaptive specializations of spatial cog-
nition in food-storing birds? Approaches to testing a comparative hypothesis.
In R. P. Balda, I. M. Petterberg, & A. C. Kamil (eds.), Animal Cognition in
Nature (pp. 65–98). San Diego, CA: Academic Press.
Smulders, T. V., K. L. Gould, & L. A. Leaver. (2010). Using ecology to guide the
study of cognitive and neural mechanisms of different aspects of spatial
memory in food-hoarding animals. Philosophical Transactions of the Royal Soci-
ety of London, B, 365, 883–900.
Stamps, J. (1995). Motor learning and the value of familiar space. American Natural-
ist, 146, 41–58.
Stevens, T. A. & J. R. Krebs. (1986). Retrieval of stored seeds by marsh tits Parus pal-
ustris in the field. Ibis, 128, 513–525.
Tinbergen, N. (1932). Über die Orientierung des Bienenwolfes (Philanthus triangu-
lum Fabr.). Zeitscrhift für vergleichende Physiologie, 16, 305–334.
Tulving, E. (1972). Episodic and semantic memory. In E. Tulving & W. Donaldson
(eds.), Organization of Memory (pp. 381–403). San Diego, CA: Academic
Press.
Vander Wall, S. B. (1982). An experimental analysis of cache recovery in Clark’s nut-
cracker. Animal Behaviour, 30, 84–94.
White, D. J., L. Ho, & G. Freed-Brown. (2009). Counting chicks before they hatch:
Female cowbirds can time readiness of a host nest for parasitism. Psychological
Science, 20, 1140–1145.
Wiltschko, W. & R. P. Balda. (1989). Sun compass orientation in seed-caching scrub
jays (Aphelocoma coerulescens). Journal of Comparative Physiology, A, 164,
717–721.
Zapka, M., D. Heyers, M. Liedvogel, E. D. Jarvis, & H. Mouritsen. (2010). Night-
time neuronal activation of cluster N in a day- and night-migrating songbird.
European Journal of Neuroscience, 32, 619–624.
Zinkivskay, A., F. Nazir, & T. V. Smulders. (2009). What-where-when memory in
magpies. Animal Cognition, 12, 119–125.
This page intentionally left blank
 11

Cognitive Repertoire, Cognitive Legacies,

and Collective Reasoning: The Next Frontiers
in Conservation Behavior
Guillermo Paz-y-Miño-C

INTRODUCTION
Animal cognition includes perception, learning, memory, decision making,
and other processes in which animals obtain information about the environ-
ment through their senses, retain it, and act on it (Shettleworth, 2001,
2010). In this chapter, I discuss case studies that have used animal-cognition
principles in conservation. I expand on four conceptual essays about the inter-
face of behavior and conservation, which were previously published in The
Conservation Behaviorist (TCB), a biannual periodical of the Animal Behavior
Society’s Conservation Committee: “Animal Cognition and Its Role in Con-
servation Behavior” (Paz-y-Miño-C, 2005), “Behavioral Unknowns: An
Emerging Challenge for Conservation” (Paz-y-Miño-C, 2003), “Contribution
of Animal Behavior Research to Conservation Biology” (Paz-y-Miño-C,
2006), and “Behavior and Conservation in the Galapagos” (Paz-y-Miño-C,
2007). I recommend the reader, particularly the students and those working
in hands-on conservation, to visit TCB online for theoretical and practical dis-
cussions on behavior and conservation (http://animalbehaviorsociety.org/
Committees/ABSConservation/ConservationBehaviorist). For a great intro-
duction to the maturing field of conservation behavior, I suggest the
290 ANIMAL BEHAVIOR

minitextbook Primer for Conservation Behavior by Blumstein and Fernández-

Juricic (2010). In addition, its predecessor volumes, or “the classics”
(Paz-y-Miño-C, 2011), should be consulted: Behavioral Approaches to
Conservation in the Wild (Clemmons & Buchholz, 1997), Behavioral Ecology
and Conservation Biology (Caro, 1998), Behaviour and Conservation (Gosling
& Sutherland, 2000), and Animal Behavior and Wildlife Conservation (Festa-
Bianchet & Apollonio, 2003). There are dozens of significant article contribu-
tions and reviews on the role of behavior in conservation biology, some of
which are discussed in this chapter and listed in the references. For recent
encyclopedic reviews on conservation and behavior, including its practical
applications to management, see Bekoff (2004), Buchholz and colleagues
(2008), and Swaisgood (2010). For criticisms on the relevance of behavioral
research in conservation, I recommend Tim Caro’s work (2007). And for an
analytical forecast about the future of field-based etho-ecological studies, par-
ticularly behavioral ecology, in the context of the current and global conserva-
tion crisis, the essay by Caro and Sherman (2011) is, indeed, insightful.
My specific goal here is to draw attention to the value of animal-cognition-
based studies that relate to conservation—conducted mostly by conservation
biologists needing to apply cognition principles in their work, remark on the
cognitive concepts intrinsic to each case, and encourage discussion and
research in the interface of animal cognition and conservation behavior.

ANIMAL COGNITION AS A BEHAVIORAL UNKNOWN

In the context of conservation, animal cognition continues to be a “behav-
ioral unknown” (Paz-y-Miño-C, 2003), a concept I proposed based on Nor-
man Myers’s (1995) “environmental unknowns.” Myers defined
environmental unknowns as those problems we had not even identified as
yet but for which we were all accountable. He listed, as examples, climate
change, mass extinctions, and ozone-layer depletion and remarked that society
disregarded their existence until the early 1980s although these “surprise phe-
nomena” had already disrupted ecological and evolutionary processes.
I argued, likewise, that as data became available to biologists and environmen-
tal public awareness grew, new unknowns were emerging and that one such
surprise was the impact of global disruptions on the behavior of animals
(Paz-y-Miño-C, 2003). A crucial question surfaced: what will be the impact
of global environmental problems on aspects of behavior that are still unknown to
scientists, such as cognition, behavioral endocrinology and physiology, communica-
tion, behavioral diversity, or behavioral evolution? I highlighted that behavioral
unknowns were emerging at a time when ethological data were most needed
COGNITIVE REPERTOIRE, COGNITIVE LEGACIES, AND COLLECTIVE REASONING 291

for captive breeding of endangered species, reintroduction programs, and hab-

itat restoration. We knew then—as much as we know now—that by partici-
pating in seed dispersal, pollination, predation, competition, and parasitism,
animals shape communities and ecosystems and that behaviors could be easily
disturbed directly through physiology (metabolic and reproductive processes)
or indirectly through the ecosystem, by changes in the abundance of resources,
predators, parasites, and competitors. Indeed, scarce scientific information had
limited our capacity to anticipate and prevent the negative effects caused by
global environmental problems on all aspects of behavior (i.e., its causation,
development, function, and evolution, based on Tinbergen’s four approaches
to behavioral analysis, see Tinbergen, 1963; for an elegant theoretical connec-
tion between classical Tinbergenian ethology and conservation behavior, see
Buchholz, 2007).
In 2006, I assessed the contribution of behavioral paradigms in conserva-
tion studies by identifying and evaluating 277 articles (from a total of 1,631)
published in the journal Conservation Biology between 1987 and 2002 that
were directly related to animal behavior and conservation (Paz-y-Miño-C,
2006). Four main areas of behavioral research had been commonly addressed
in these studies (Figure 11.1): dispersal and settlement (in 100% of all publi-
cations), reproductive behavior and social organization (80%), species

Figure 11.1. Areas of animal behavior most commonly discussed in conservation biol-
ogy studies. Number of articles per category is indicated next to each bar. Percent-
ages were calculated for each category in respect to the total number of articles in the
survey (N ¼ 277). The arrow highlights cognition, N ¼ 18, or 7 percent, the main
focus of this chapter. (Modified from Paz-y-Miño-C, 2006, with permission from The
Conservation Behaviorist)
292 ANIMAL BEHAVIOR

interactions (55%), and foraging/feeding and pollination (33%). These areas

had helped researchers to understand and alleviate conservation problems in
three major fields (Figure 11.2): extinction of endangered species and biodi-
versity loss (53% of all studies), habitat destruction (30%) and ecosystem
management and restoration (29%). Then and now these areas of research
and fields of application of behavior to conservation continue to be relevant,
although nowadays they are addressed technically and understood conceptu-
ally under modern perspectives (see Caro & Sherman, 2011).
Animal cognition was significantly discussed in only 7 percent of the stud-
ies I examined back in 2006 (Figure 11.1). Today, the impact of human-made
environmental disruptions on the phenotypic expression of cognitive abilities,
such as perception, learning, memory, and decision making, are of much
interest considering the proliferation of comparative-cognition-based research,
which suggests that “animal minds” across taxa are not only complex and
impressive in comparison to humans’ (Dukas, 1998; Shettleworth, 2001,
2010; Baber, 2003; de Waal & Tyack, 2003; Wasserman & Zentall, 2006;
Cheney & Seyfarth, 2007; Dukas & Ratcliffe, 2009), but also that the infor-
mation animals carry in their brains from generation to generation (here I refer
to them as cognitive legacies, distinctive from traditions, that can be the
product of collective reasoning, particularly in social animals but not restricted
to them; see Box 11.1), and that is essential for survival, could be lost abruptly
as habitats become fragmented or polluted, or disappear.
Therefore, some of the crucial questions biologists still need to investigate
are as follows: How many nongenetic and valuable cognitive legacies are lost

Figure 11.2. Conservation problems most commonly discussed in behavioral conserva-

tion studies (N ¼ 277). Values correspond to categories explained in Figure 11.1.
(Modified from Paz-y-Miño-C, 2006, with permission from The Conservation
Behaviorist)
COGNITIVE REPERTOIRE, COGNITIVE LEGACIES, AND COLLECTIVE REASONING 293

Box 11.1
Some cognitive terminology that can be applied to conservation

Cognitive repertoire—The collection of cognitive abilities of a species as

inherited in the gene pool. It implies genetic variability, at the individual
and population levels, to express the capacities in specific environments.
Sudden reductions in population numbers and extinctions can erode the
genetic, multiallelic diversity that sustains a cognitive repertoire.
Cognitive legacy—The nongenetic inherited expression of perceiving,
learning, remembering, and making decisions in the environment. It
includes the common and highly predictable experiences an animal, or
groups of animals, faces or face in nature during development and adult-
hood, as well as the rare or unique events from which animals learn to make
adaptive decisions. Cognitive legacies are comparable to animal traditions
(learned, inherited behavior) but, in the cognitive context, the former refer
to situations where animals learn to learn, reason on how to reason in socio-
sexual contexts, enhance the expression of such natural abilities to perform
cognitively (collective reasoning), and accumulate this collection of meta-
cognition (via self- or collective reflection; for a significant review on compa-
rative metacognition see Terrace & Son, 2009) which is passed on vertically
or horizontally. If this information is lost, due to extinctions of populations
or species, it will be challenging to recover it during captive breeding, habi-
tat restorations, or reintroductions of animals into the wild.
Collective reasoning—Refers to situations in which animals perceive,
learn, remember, and make decisions while interacting with conspecifics in
the environment. Subtle behavioral communication cues among animals
can facilitate collective assessment of the environment while foraging, feed-
ing, competing for resources or mates, avoiding predators, nesting, and dis-
persing to or colonizing habitats. During social facilitation, animals would
enhance their cognitive performance, learning from bouts of collective
remembering about the environment or situations in which they reason
together. Over time, collective reasoning will contribute to building nonge-
netically inherited cognitive legacies.

when individual animals die, or when entire populations of animals, adapted

to specific habitat conditions, become extinct? How significant, for conserva-
tion purposes, is the loss of cognitive legacies stored in the animals’ minds,
which can only be perpetuated nongenetically? After cognitive legacies have
been lost, how can they be restored? Is the loss of animal cognitive information
significantly comparable to the loss of cultural information (products of com-
plex and chronologically cumulative collective reasoning) among humans or
294 ANIMAL BEHAVIOR

simply an analogous scenario from which we can learn about the downfall of
traditions but for which application to conservation is limited? Below, I elabo-
rate on these challenging questions and try to answer them.

ANIMAL COGNITION AND ITS ROLE IN CONSERVATION

Historically, animal cognition (cognitive ethology, sensu stricto as in Kamil,
1998) has contributed to conservation in six main areas (Figure 11.3): training
animals to avoid predators, securing that mate choice in captivity mimics mate
choice in nature, eliminating interspecific sexual and social imprinting,
enhancing social learning and social facilitation, identifying inappropriate
conservation strategies, and identifying and minimizing the negative effects
of maladaptive behaviors (Paz-y-Miño-C, 2005). Let us examine a few practi-
cal examples. Below I include selected case studies in which the connection
between conservation biology and animal cognition is direct; however, I have
deliberately omitted dozens of experimental reports with potential to link con-
servation and cognition.

Training Animals to Avoid Predators

Animals that have been isolated from predators, either throughout their
lifetime or over evolutionary time, may no longer express effective antipreda-
tor behavior (Griffin et al., 2000). In fact, mortality due to predation is the
principal cause of failure in animal reintroduction and translocation pro-
grams—it is easier to teach animals to cope with predators if they have experi-
enced ontogenetic isolation than if they have undergone evolutionary isolation
from predators (Griffin et al., 2000). In the absence of predators, antipredator
behavior may degenerate or be lost. For example, a significant threat to
released California condors (Gymnogyps californianus) is nest predation by
common ravens (Corvus corax). Wild and released condors lack defenses
against ravens, possibly because ravens have become abundant during recent
times and condors have had limited evolutionary exposure to raven predators
(Meretsky et al., 2000).
Learning-theory principles can be used to predict which antipredator
responses can be enhanced or, if possible, fully recovered by training animals
prior to their reintroduction into the wild or translocation into new habitats
(Griffin et al., 2000). Training techniques involve conditioning procedures
in which animals learn that model predators are predictors of aversive events.
Researchers have documented how terrestrial predator-recognition and
defense behaviors are more developed and easier to restore in the pukeko
(Porphyrio porphyrio), an Australian and New Zealand bird that evolved in the
Figure 11.3. Concept map showing simplified connections between cognitive ethology,
or “animal cognition,” and conservation biology. The multidisciplinary field of con-
servation biology is connected to various disciplines; one of them is animal behavior,
which is linked, as well, to subspecializations, such as cognitive ethology and behav-
ioral ecology. The documented contributions of animal cognition to conservation
efforts are depicted.
296 ANIMAL BEHAVIOR

presence of terrestrial marsupial predators, than in the takahe (Porphyrio man-

telli), a large, flightless gallinule that evolved in the absence of predators until
the end of the nineteenth century, when European settlers introduced the
stoat (Mustela erminea) to New Zealand (Bunin & Jamieson, 1996; Moore
et al., 2008).
Animals on islands are extraordinarily tame relative to animals on conti-
nents, and when exotic predators arrive, they trigger a shift in selective forces
on the insular fauna. Feral cats (Felis catus) in the Galápagos Islands, for exam-
ple, are responsible for increased wariness (loss of unusual tame behavior) in
the lava lizards (Tropidurus spp.) as a result of predation pressure (Stone et
al., 1994). Survival of lizards on islands inhabited by cats is attributable to fast
development of predator-avoidance strategies (i.e., associative learning, aver-
sive events) supported by selection toward phenotypes with effective fleeing
responses.

Ensuring That Mate Choice in Captivity Mimics Mate Choice in Nature

Animals select mates in complex natural manners, yet “forced” pairing in
captive breeding programs or “semiforced” pairing in reintroductions of very
small populations into the wild has always been a behavioral and conservation
concern (review in Asa et al., 2011). This is because animals in nature choose
mates, in most cases, freely. Not only their mating systems in the wild (e.g.,
promiscuous, monogamous, polyandrous, polygynous, and their variants)
influence mate choice and consequent reproductive success, but also their
compatibility between pairs or groups of reproducing individuals (but see Tre-
genza & Wedell, 2000). These behavioral and cognitive harmonies further
translate into adequate parental care (i.e., higher offspring survival), plus
future cooperation of offspring and coalition formation in adulthood. The
“right” choice during mate choice can indeed have direct and indirect long-
term fitness consequences.
Playback and decoy techniques, often used to artificially attract birds and
encourage them to recolonize habitats, have been successful in nine species,
including terns, vireos, flycatchers, warblers, sparrows, and bobolinks
(Ahlering et al., 2010); this conservation strategy stimulates recruitment, the
formation of breeding pairs, and nesting, and it takes advantage of the cogni-
tive predisposition of birds to mimic, from conspecifics, both habitat selection
and mate choice.
Captive female Chinook salmon (Oncorhynchus tshawytscha) seem to prefer
major histocompatibility complex (MHC) dissimilar mates, which may
improve offspring resistance to pathogens (Garner et al., 2010); by allowing
females to choose mates freely and disassortatively, optimal MHC
COGNITIVE REPERTOIRE, COGNITIVE LEGACIES, AND COLLECTIVE REASONING 297

compatibility might result in their descendants. But evidence that genetic

compatibility drives mate choice in other taxa is scarce (Tregenza & Wedell,
2000; Miller et al., 2009).

Eliminating Interspecific Sexual and Social Imprinting

Puppet rearing, a technique aimed to reduce sexual and filial imprinting on
human caretakers, has been used in bird species. Maladaptive imprinting is
most likely to occur during early stages in an animal’s behavioral development.
Studies indicate that rearing common ravens with a puppet does not affect
social behaviors prior to release, dispersal from the release area, or interaction
with wild birds after release. Ravens raised with a puppet, however, are more
fearful of caretakers and more vigilant prior to release than ravens reared with-
out a puppet and in full view of humans.
These effects on behavior have translated to changes in survival after release
of captive-reared young (Valutis & Marzluff, 1999). Likewise, aversive condi-
tioning of puppet-reared California condors in later releases has reduced
initial tendencies to approach humans and human structures (i.e., collisions
with electric wires and towers; Meretsky et al., 2000, but see Holden, 2002).

Enhancing Social Learning and Social Facilitation

Animals that live in stable social groups have substantial cognitive abilities
and usually interact with conspecifics and the environment in complex man-
ners. Orangutans (Pongo pygmaeus), for example, have extensive parental care
and prolonged infant and juvenile periods in which animals are extremely
dependent; their appropriate social and sexual responses are learned within
the context of the social group (Yeager, 1997). The ability to find and prepare
food is generally acquired through a combination of observational and exper-
imental learning. Most nonhuman primates reared in captivity, particularly
those that have been reared or housed in social isolation, exhibit behavioral
abnormalities (e.g., repetitive behaviors, rocking, self-abuse) not observed in
wild populations. Lack of stimulation and social interaction can result in
developmental retardation (Yeager, 1997).
Some reintroduction programs have taken these premises into considera-
tion. For example, female and male wild-born orphan chimpanzees (Pan trog-
lodytes troglodytes) have been successfully released into the Conkouati Reserve,
Republic of Congo, by identifying behavioral traits that the animals required
prior to translocation from diverse localities into the release area, for instance,
the release of mostly adolescent females mimicked the natural pattern of
movement between communities. The complexity of chimpanzee social
298 ANIMAL BEHAVIOR

behavior also required that the released animals have the full species-specific
repertoire (e.g., greetings, grooming, and agonistic behaviors needed to estab-
lish and reaffirm relationships) and show no abnormal or inappropriate
patterns of behavior (Tutin et al., 2001).

Identifying Inappropriate Conservation Strategies

Mate choice, sexual selection, and dominance associated with mate
acquisition depend on cognitive processes such as perception, learning,
memory, and decision making. For example, rhino dehorning, a controversial
practice designed to remove the incentive for poachers to kill the hornless
animals, may have had mate-choice implications in both the one-horned
rhinoceros (Rhinoceros unicornis) and the white rhinoceros (Ceratotherium
simmum), two sexually dimorphic species in which males have larger horns
than females (Berger & Cunningham, 1998). Horn size is likely a consequence
of sexual selection, and it is positively associated with dominance in males.
Because dehorning generates unnatural horn-size asymmetries that affect combat
outcomes between males, dehorning may have had fitness consequences due to
female assessment (perception) of male quality. Note that dehorning was discon-
tinued as a conservation practice due to high neonate mortality; females use
horns in neonate defense (Berger & Cunningham, 1998).

Identifying and Minimizing the Negative Effects of Maladaptive Behaviors

Some behaviors become maladaptive when selective pressures change, usu-
ally because of human intervention, and animals are no longer able to assess
(perception) the change or respond appropriately to it (e.g., traffic collisions
or collisions with overhead wires, ingestion of anthropogenic debris, disorien-
tation caused by artificial lights); maladaptive responses can also include
cognitive-based alterations in habitat selection, movement and dispersal deci-
sions, foraging strategies, and sociosexual behavior, including maladaptive
mate choice. For important reviews on animal behavioral responses to
human-induced environmental change, see Barber and colleagues (2009)
and Tuomainen and Candolin (2011). If a species can survive long enough
and the behavior has heritable variation, these maladaptive traits can disap-
pear through natural selection (Reed, 1999), and the species may persist with
modified cognitive capabilities of adaptive value in the new environment (for
specific examples, see Box 11.2).
To minimize the negative effect of these maladaptive responses, various
techniques have been applied or suggested; all take into consideration animal
perception, learning, information processing, and decision making. Examples
COGNITIVE REPERTOIRE, COGNITIVE LEGACIES, AND COLLECTIVE REASONING 299

Box 11.2
Examples of maladaptive behavior

Mortality due to traffic collisions in Florida scrub-jays (Aphelocoma coerules-

cens) is particularly high in immigrant (naïve) birds that colonize and estab-
lish territories in habitats along roads (Mumme et al., 2000).
Excessive tameness and curiosity shown by released California condors
(Gymnogyps californianus) toward humans and urbanized areas have apparently
contributed to mortality due to collisions with overhead wires (Meretsky et al.,
2000); but recent studies with kori bustards (Ardeotis kori), blue cranes
(Anthropoides paradisea), and white storks (Ciconia ciconia) suggest that such
collisions might result from the birds’ inability to see forward (frontal blind
areas) as consequence of the lateral position of the eyes (Martin & Shaw,
2010). Traffic collisions are responsible for increased mortality in at least 10
species of European ungulates (Groot-Bruinderink & Hazebroek, 1996).
Simple human traffic has induced diurnal animals and their predators to
become nocturnal in Sumatran rain forests, where poachers are active
during daylight, thus disrupting the animals’ circadian rhythms (Griffiths
& Van Schaik, 1993).
Road construction in central Amazonian Brazil affects movement pat-
terns of understory birds in mixed-species flocks; the birds are unable to
adapt to the formation of abrupt edges inside their territories and avoid
crossing the roads to visit areas where foraging was frequent before the clear-
ing (Develey & Stouffer, 2001).
Extra-pair siring in blue tits (Parus caeruleus) is enhanced by artificial
night lights, which alter reproductive behavior by influencing dawn singing,
egg laying and mate fidelity (Kempenaers et al., 2010); under natural condi-
tions, early dawn songs broadcasts male quality, but in urban areas with night
lights, time-altered male vocalizations render unreliable information to females,
who respond maladaptively by choosing earlier singers (for an extensive list of
references on the effects of artificial lighting in behavior, see Longcore, 2010).
Mate-attraction acoustic signals and territory defense, as well as alert
vocalizations, or sound-based signals, during predator-related danger, are
masked by anthropogenic noise pollution, to which fish, frogs, birds, ceta-
ceans, bats, and ground squirrels have responded maladaptively by not flee-
ing away to calmer areas but instead trading acoustic- or sound-based
habitat assessment with sometimes less effective visual cues, thus increasing
vigilance behavior at the expense of foraging, feeding, or mating times (see
reviews in Slabbekoorn & Ripmeester, 2008; Laiolo, 2010).
Foraging behavior (time spent with prey at kill sites) and survival of
Amur tigers (Panthera tigris altica) have been affected by road construction
300 ANIMAL BEHAVIOR

and human disturbance on and near the Sikhote-Alin State Biosphere,

Zapovednik, Russia. Tigers disturbed at kills spend less time at kill sites
and consumed less meat from each kill than undisturbed tigers do. Adult
and cub mortality is greatest in areas with primary roads (Kerley et al.,
2002).
Ingestion of anthropogenic debris by terrestrial and aquatic species has
been documented extensively. California condors exhibit high mortality
caused by lead poisoning resulting from ingestion of bullet fragments in car-
casses (Meretsky et al., 2000). Reduced nutrient gain from diets diluted by
consumption of debris is a common problem among posthatchling and
juvenile loggerhead sea turtles (McCauley & Bjomdal, 1999).
By lacking the ability to sense heavy-metal pollution in water, Eurasian
otters (Lutra lutra) continue to inhabit and feed in abruptly degraded envi-
ronments where lead and arsenic concentration is high; because colonization
and persistence in habitats is socially facilitated by the physical presence of
two or more residents, otters aggregate and inevitably sicken at polluted sites
(Stamps & Swaisgood, 2007; Delibes et al., 2009). Stamps and Swaisgood
(2007) have discussed how animals can prefer and settle in habitats whose
appearance resembles that of natal environments, a phenomenon called natal
habitat preference induction, which, as illustrated by the Eurasian otter exam-
ple, can cognitively encourage them to behave maladaptively.
Elephant kin-based associations, which rely not only on close relatedness
and hierarchical matriarchal groups but also on collective learning, memory,
and reasoning to cognitively map the environment and track locations of
family members and predators, is disrupted by poaching (Archie & Chiyo,
2012; Byrne et al., 2009). Socially empathic elephant groups loose cohesive-
ness when the most experienced and “knowledgeable” matriarchs die, and
the remaining wanderers respond maladaptively, in cases increasing crop
raidings and conflicts with humans (Archie & Chiyo, 2012).

include construction of under-road passages for terrestrial vertebrates, estab-

lishment of buffer zones to reduce contact between animals and people
(i.e., vehicles, motorized tour boats, tourists), and intermittently lighted sig-
nals triggered by passing animals (Klein et al., 1995; Rodgers & Smith,
1995; Groot-Bruinderink & Hazebroek, 1996; Galicia & Baldassarre, 1997;
Rodgers & Schwikert, 2002; Martin & Shaw, 2010).

THE NEXT FRONTIER

Earlier in this chapter, I posited a question that deserves further examina-
tion: Is the loss of animal cognitive information—due to, for example,
COGNITIVE REPERTOIRE, COGNITIVE LEGACIES, AND COLLECTIVE REASONING 301

extinctions and abrupt habitat change—significantly comparable to the loss of

cultural information among humans, or is it simply an analogous scenario
from which we can learn about the downfall of traditions but whose applica-
tion to conservation is limited? We should first highlight that the cognitive
repertoire of humans is huge; it includes not only profound consciousness
and awareness about self and the cosmos but also numerous cognitive
capacities that have been documented experimentally: object permanence,
learning-set formation, numerical concept, tool use, visual perspective tak-
ing, behavior reading, transitive reasoning, spatial memory, discrimina-
tion reversal, tactical deception, and insight learning, among others (see
Box 11.3). Second, we should acknowledge that modern studies have demon-
strated that nonhuman animals also possess versions of such capacities (or that
humans possess capacities that are versions of those of animals!), which are
taxa-specific in skillfulness and vary vastly taxonomically, thus suggesting that
the evolution of animal cognition has been driven by both phylogeny and con-
vergence under comparable ecological scenarios (Balda et al., 1998; Dukas
1998; Shettleworth, 2001, 2010; Bekoff et al., 2002; Cheney & Seyfarth,
2007; Clayton et al., 2007; Dukas & Ratcliffe, 2009; but see Seed et al.,
2009). Third, we must therefore assume that, analogously to humans, nonhu-
man animals are equipped genetically and phenotypically to perform cogni-
tively in evolving environments (cognitive repertoires; Box 11.1) but that,
contrary to humans, sudden population extinctions might affect nonhuman
animals uniquely due to the latter’s inability to preserve non-genetically-
transmitted information actively (cognitive legacies; Box 11.1). And our
capacity, as behaviorists or conservationists, to teach animals artificially what
they normally learn by themselves, or from each other, via interactions in soci-
osexual contexts (collective reasoning; Box 11.1), is indeed limited. Therefore,
taxa whose survival in the wild depends significantly on learning behaviors
must be prone to perish when confronting environmental devastation.
After cognitive legacies of wild animals have been lost, are they gone for-
ever, or can they be restored? Animals have evolved mechanisms to discrimi-
nate, recognize, and evaluate habitats, resources, and conspecifics; they track
the position, social behavior, and foraging success of individuals; group them
by age, sex, reproductive status, genetic relatedness (e.g., kin versus nonkin),
and dominance rank; as well as infer relationships among individuals in socio-
sexual contexts (Balda et al., 1998; Dukas 1998; Shettleworth, 2001, 2010;
Paz-y-Miño-C et al., 2004; Cheney & Seyfarth, 2007; Clayton et al., 2007;
Dukas & Ratcliffe, 2009; Seed et al., 2009). These cognitive abilities, which
are context dependent and phenotypically plastic, develop from birth to adult-
hood, and their nongenetic components are passed on vertically from parents
302 ANIMAL BEHAVIOR

Box 11.3
Cognitive abilities documented in humans and other taxa

Object permanence—Awareness that objects continue to exist even when

they are not physically present before the animal; object permanence implies
that the animal recalls the existence of an object despite its momentary or
prolonged absence. Documented in apes, lemurs, capuchin monkeys, mar-
mosets, dogs, cats, parrots, and corvids (review in Hoffman et al., 2011).
Learning-set formation—Sometimes referred to as the capacity of learn-
ing to learn, or to be able to solve a cognitive problem and use this experi-
ence to solve subsequent problems of increasing difficulty and in fewer
trials or faster time; cognitively experienced animals learn to take shortcuts
or skip steps when solving gradually more complex cognitive challenges.
Documented broadly in vertebrates and invertebrates, including ants and
octopus (Graham et al., 2010; Anderson & Mather, 2010).
Numerosity and numerical concept—The ability to understand quan-
tity of items or objects; it includes the capacity to discriminate between
one, two, few, or many items in a set or to count them readily; numerical
concept does include the symbolic and abstract representation of quantity
in the brain as well as the concept of zero, which some nonhuman animals
possess. Documented in apes, rhesus monkeys, cetaceans, squirrel monkeys,
raccoons, rodents, pigeons, parrots, and salamanders (review in Beran,
2008).
Tool use—Using or purposely modifying inanimate objects outside the
animal’s own anatomy to procure or modify resources or to alter the envi-
ronment with the purpose of benefiting from the alteration. Documented
broadly in vertebrates and in some invertebrates, including octopus (review
in Seed & Byrne, 2010).
Visual perspective taking—The ability to use an animal’s own visual
perspective and the visual perspectives of others to assess, understand, and
predict the behavior of con- and heterospecifics; it allows the animal to
mentally place itself in someone else’s place and make spatial cognitive deci-
sions based on such perspective. Documented in apes, Old and New World
monkeys, dogs, and corvids (for reviews, including gaze-direction-based
cognition, see Rosati & Hare, 2009; Dally et al., 2010; Udell et al., 2011;
MacLean & Hare, 2012).
Behavior reading—The ability to assess and predict future behaviors of
a con- or heterospecific based on the latter’s—or demonstrator’s—“body
language” or behavioral patterns; behavior reading can occur as a simple
mimic of another animal’s behavior (sometimes referred to as imitation of
COGNITIVE REPERTOIRE, COGNITIVE LEGACIES, AND COLLECTIVE REASONING 303

actions, copying, or response facilitation by priming) or as a more elaborate

assessment—in the mind of the observer—of the potential intentions of the
demonstrator’s actions. Documented broadly in vertebrates, particularly in
apes, baboons, Old and New World monkeys, dogs, and rodents (review
in Bates & Byrne, 2010).
Transitive reasoning—The ability of an animal to infer a relationship
between two or more items or objects based on previous but incomplete
information about the items or objects; for example, the capacity to infer
that if animal A is dominant to animal B, and B is dominant to C, it follows
that when an observer watches animals A and C together, and if the rela-
tionship is linear (A > B > C), the observer should infer that A will dominate
C. Documented in apes, lemurs, squirrel monkeys, rodents, pigeons, par-
rots, corvids, roosters, and cichlids (review in Zentall et al., 2008).
Spatial memory—The ability to discriminate between and remember
specific and different locations of objects in relation to each other, the
observer, and the environment; spatial memory includes an animal’s spatial
orientation and can include a cognitive mental map used by the animal to
navigate, recall, and track the location of objects in the environment. Doc-
umented in apes, squirrel monkeys, marmosets, dogs, rodents, pigeons,
chicks, corvids, and hymenoptera (reviews in Gibson& Kamil, 2009; Gra-
ham et al., 2010; Tommasi et al., 2012).
Discrimination reversal—The ability to exchange the attribute or value
of an item with another item; it includes, for example, the ability to dis-
criminate first between object A (rewarded when the animal chooses it)
and B (not rewarded) and to later reverse the attribute or value of A (no
longer rewarded when the animal chooses it) with that of B (now rewarded).
Documented in apes, rhesus and capuchin monkeys, rodents, pigeons, pass-
erine birds, reptiles, octopus, bees, and butterflies (reviews in Rodrigues et
al., 2010; Gaalema, 2011; Rayburn-Reeves et al., 2011).
Tactical deception—The ability to purposely confuse, deceive, or mis-
guide a con- or heterospecific; it includes, for example, stereotypic behaviors
like pretending to have an injury to divert a predator’s attention (e.g., “bro-
ken wing” in birds) or sophisticated pilfering, or concealed access to and use
of resources. Documented in apes, baboons, and corvids (Byrne, 2010).
Insight learning—The ability to understand the intricacies of a problem
and solve it without previous trial and error; it implies that the animal exam-
ines the components of a problem in its brain, solves it mentally, and then
proceeds to act in an apparently “all-of-a-sudden solution” to physically
resolve the task. Documented in apes, keas, and corvids (Shettleworth,
2009; Seed & Byrne, 2010).
304 ANIMAL BEHAVIOR

to offspring or shared horizontally among cohorts of overlapping generations

(Schmajuk, 1997; Bekoff et al., 2002). Therefore, it will be challenging
for behaviorists and conservationists to restore the loss of such complex cogni-
tive legacies, particularly when they are learned, practiced, and enhanced
collectively during sociosexual encounters among animals (see Box 11.1).
Preserving large and functional ecosystems might be the only viable strategy
to perpetuate animal cognitive legacies and, in some cases, animal cultures
(sensu stricto as in Byrne & Bates, 2010; Whitehead, 2010; van Schaik &
Burkart, 2011).

CONCLUDING REMARKS
The six areas discussed above and shown in Figure 11.3, where wildlife
managers have used principles of cognitive ethology in conservation efforts (
i.e., the training of animals to avoid predators, ensuring that mate choice in
captivity mimics mate choice in nature, eliminating interspecific sexual and
social imprinting, enhancing of social learning and social facilitation, identify-
ing inappropriate conservation strategies, and detecting and treating the nega-
tive effects of maladaptive behaviors), demonstrate that animal cognition
principles not only have the potential to contribute to conservation efforts
but have indeed played a unique role in animal rehabilitation and captive
breeding for reintroductions. Some of the paradigms currently explored by
cognitive ethologists could further influence conservation efforts in significant
ways, for example, phenotypic plasticity in learning and animal adaptability to
changing environments (i.e., animals learn to assess new landscapes, new prey,
new predators), visual and vocal communication (i.e., signaling, dialect forma-
tion in increasingly isolated metapopulations), spatial orientation and naviga-
tion (memory in food-storing birds), foraging and search-image formation in
constantly evolving prey or habitat (e.g., virtual behavioral ecology), hierarchy
learning, social learning, and social facilitation (Balda et al., 1998; Shettle-
worth, 2001, 2010; Bond & Kamil, 2002; Seed et al., 2009). The extent to
which these capabilities are being influenced or disrupted by current habitat
fragmentation and degradation, or even by larger-scale environmental prob-
lems (i.e., climate change, ozone depletion, pollution), deserves closer analysis.
Animal cognition theory, therefore, can continue to assist us in identifying,
understanding, and guiding our conservation practices to restore mildly
affected or partially lost cognitive legacies. However, wildlife managers and
conservation biologists ought to be alert that recovering cognitive legacies of
species already extinct in the wild, but saved in captivity or reserves, will be a
significant challenge.
COGNITIVE REPERTOIRE, COGNITIVE LEGACIES, AND COLLECTIVE REASONING 305

ACKNOWLEDGMENTS
I thank Ken Yasukawa for inviting me to write this chapter. My research
collaborator in numerous other projects, Avelina Espinosa, provided signifi-
cant feedback during the preparation of the manuscript. Many thanks to the
members of the Animal Behavior Society Conservation Committee for spon-
soring, over a decade, my role as editor of The Conservation Behaviorist
(TCB): Debra M. Shier (associate editor of TCB), Richard Buchholz, Daniel
T. Blumstein, Ronald R. Swaisgood, Bruce A. Schulte, Colleen Cassady
St. Clair, Mark L. Wildhaber, Esteban Fernández-Juricic, Allison C. Alberts,
James Ha, J. Cully Nordby, Elizabeth V. Lonsdorf, John Eadie, M. Elsbeth
(Misty) McPhee, and Jessica Ward. Students in my laboratory at the Univer-
sity of Massachusetts–Dartmouth, Elsa Yeung, Elizabeth Spinney, Rachael
Bonoan, and Felecia Clodius, contributed with discussion and very useful
references to document this chapter.

REFERENCES AND SUGGESTED READING

Ahlering, M. A., D. Arlt, M. G. Betts, R. J. Fletcher, J. J. Nocera, & M. P. Ward.
(2010). Research needs and recommendations for the use of conspecific-
attraction methods in the conservation of migratory songbirds. Condor, 112,
252–264.
Anderson, R. C. & J. A. Mather. (2010). It’s all in the cues: Octopuses (Enteroctopus
dofleini) learn to open jars. Ferrantia, 59, 8–13.
Archie, E. A. & P. I. Chiyo. (2012). Elephant behaviour and conservation: Social rela-
tionships, the effects of poaching, and genetic tools for management. Molecular
Ecology, 21, 765–778.
Asa, C. S., K. Traylor-Holzer, & R. C. Lacy. (2011). Can conservation-breeding pro-
grammes be improved by incorporating mate choice? International Zoo Year-
book, 45, 203–212.
Baber, C. (2003). Cognition and Tool Use. Boca Raton, FL: CRC Press.
Balda, R. P., I. M. Pepperberg, & A. C. Kamil (eds.). (1998). Animal Cognition in
Nature. San Diego, CA: Academic Press.
Barber, J. R., K. R. Crooks, & K. M. Fristrup. (2009). The costs of chronic noise expo-
sure for terrestrial organisms. Trends in Ecology and Evolution, 25, 180–189.
Bates, L. A. & R. W. Byrne. (2010). Imitation: What animal imitation tells us about
animal cognition. Wiley Interdisciplinary Reviews: Cognitive Science, 1, 685–695.
Bekoff, M. (ed.). (2004). Encyclopedia of Animal Behavior. Westport, CT: Greenwood
Press.
Bekoff, M., C. Allen, & G. M. Burghardt (eds.). (2002). The Cognitive Animal.
Cambridge, MA: MIT Press.
306 ANIMAL BEHAVIOR

Beran, M. J. (2008). The evolutionary and developmental foundations of mathemat-

ics. PLoS Biology, 6, e-19. doi:10.1371/journal.pbio.0060019.
Berger, J. & C. Cunningham. (1998). Natural variation in horn size and social domi-
nance and their importance to the conservation of black rhinoceros. Conserva-
tion Biology, 12, 708–711.
Blumstein, D. T. & E. Fernández-Juricic. (2010). A Primer of Conservation Behavior.
Sunderland, MA: Sinauer Associates.
Bond, A. B. & A. C. Kamil. (2002). Visual predators select for crypticity and poly-
morphism in virtual prey. Nature, 415, 609–613.
Buchholz, R. (2007). Behavioural biology and effective and relevant conservation
tool. Trends in Ecology and Evolution, 22, 401–407.
Buchholz, R., P. Yamnik, C. N. Pulaski, & C. A. Campbell. (2008). Conservation
and behaviour. In Encyclopedia of Life Sciences (pp. 1–9). Chichester, UK: John
Wiley & Sons.
Bunin, J. S. & I. G. Jamieson. (1996). Response to a model predator of New
Zealand’s endangered takahe and its closest relative, the pukeko. Conservation
Biology, 10, 1463–1466.
Byrne, R. W. (2010). Deception: Competition by misleading behavior. In M. D.
Breed & J. Moore (eds.), Encyclopedia of Animal Behavior (pp. 461–465).
Oxford, UK: Academic Press.
Byrne, R. W. & L. A. Bates. (2010). Primate social cognition: Uniquely primate,
uniquely social, or just unique. Neuron, 65, 815–830.
Byrne, R. W., L. A. Bates, & C. J. Moss. (2009). Elephant cognition in primate per-
spective. Cognitive and Behavior Reviews, 4, 1–15.
Caro, T. (1998). Behavioral Ecology and Conservation Biology. New York: Oxford
University Press.
Caro, T. (2007). Behavior and conservation: A bridge too far? Trends in Ecology and
Evolution, 22, 394–400.
Caro, T. & P. W. Sherman. (2011). Endangered species and a threatened discipline:
Behavioural ecology. Trends in Ecology and Evolution, 26, 111–118.
Cheney, D. L. & R. M. Seyfarth. (2007). Baboon Metaphysics. Chicago: University of
Chicago Press.
Clayton, N. S., J. M. Dally, & N. J. Emery. (2007). Social cognition by food-catching
corvids: The western scrub-jay as a natural psychologist. Philosophical Transac-
tions of the Royal Society, B, 362, 507–522.
Clemmons, J. E. & R. Buchholz. (1997). Behavioral Approaches to Conservation in the
Wild. New York: Cambridge University Press.
Dally, J. M., N. J. Emery, & N. S. Clayton. (2010). Avian theory of mind and
counter espionage by food-caching western scrub-jays (Aphelocoma californica).
European Journal of Developmental Psychology, 7, 17–37.
Delibes, M., S. Cabezas, B. Jiménez, & M. J. González. (2009). Animal decisions and
conservation: The recolonization of a severely polluted river by the Eurasian
otter. Animal Conservation, 12, 400–407.
COGNITIVE REPERTOIRE, COGNITIVE LEGACIES, AND COLLECTIVE REASONING 307

Develey, P. F. & P. C. Stouffer. (2001). Effects of roads on movements by understory

birds in mixed-species flocks in central Amazonian Brazil. Conservation Biology,
15, 1416–1422.
de Waal, F. B. M. & P. L. Tyack (eds.). (2003). Animal Social Complexity. Cam-
bridge, MA: Harvard University Press.
Dukas, R. (ed.). (1998). Cognitive Ecology. Chicago: University of Chicago Press.
Dukas, R. & J. M. Ratcliffe (eds.). (2009). Cognitive Ecology II. Chicago: University
of Chicago Press.
Festa-Bianchet, M. & M. Apollonio (eds.). (2003). Animal Behavior and Wildlife
Conservation. Washington, DC: Island Press.
Gaalema, D. E. (2011). Visual discrimination and reversal learning in rough-necked
monitor lizards (Varanus rudicollis). Journal of Comparative Psychology, 125,
246–249.
Galicia, E. & G. A. Baldassarre. (1997). Effects of motorized tourboats on the behav-
ior of nonbreeding American flamingos in Yucatan, Mexico. Conservation Biol-
ogy, 11, 1159–1165.
Garner, S. R., R. N. Bortoluzzi, D. D. Heath, & B. D. Neff. (2010). Sexual conflict
inhibits female mate choice for major histocompatibility complex dissimilarity
in Chinook salmon. Proceedings of the Royal Society of London, B, 277,
885–894.
Gibson, B. & A. Kamil. (2009). The synthetic approach to the study of spatial
memory: Have we properly addressed Tinbergen’s “four questions”? Behaviou-
ral Processes, 80, 278–287.
Gosling, L. M. & W. J. Sutherland (eds.). (2000). Behaviour and Conservation. New
York: Cambridge University Press.
Graham, P., A. Philippides, & B. Baddeley. (2010). Animal cognition: Multi-modal
interactions in ant learning. Current Biology, 20, R639–R640.
Griffin, A. S., D. T. Blumstein, & C. S. Evans. (2000). Training captive-bred or
translocated animals to avoid predators. Conservation Biology, 14, 1317–1326.
Griffiths, M. & C. P. van Schaik. (1993). The impact of human traffic on the abun-
dance and activity periods of Sumatran rain forest wildlife. Conservation Biol-
ogy, 7, 623–626.
Groot-Bruinderink, G. W. T. A. & E. Hazebroek. (1996). Ungulate traffic collisions
in Europe. Conservation Biology, 10, 1059–1067.
Hoffman, A., V. Rüttler, & A. Nieder. (2011). Ontogeny of object permanence and
object tracking in the carrion crow, Corvus corone. Animal Behaviour, 82,
359–367.
Holden, C. (2002). Condor failure. Science, 298, 1169.
Kamil, A. C. (1998). On the proper definition of cognitive ethology. In R. P. Balda, I.
M. Pepperberg, & A. C. Kamil (eds.), Animal Cognition in Nature (pp. 1–28).
San Diego, CA: Academic Press.
Kempenaers, B., P. Borgstöm, P. Loës, E. Schlicht, & M. Valcu. (2010). Artificial
night lighting affects dawn song, extra-pair siring success, and lay date in song-
birds. Current Biology, 20, 1735–1739.
308 ANIMAL BEHAVIOR

Kerley, L. L., J. M. Goodrich, D. G. Miquelle, E. N. Smirnov, H. B. Quigley, & M.

G. Hornocker. (2002). Effects of roads and human disturbance on Amur
tigers. Conservation Biology, 16, 97–108.
Klein, M. L., S. R. Humphrey, & H. F. Percival. (1995). Effects of ecotourism on dis-
tribution of waterbirds in a wildlife refuge. Conservation Biology, 9, 1454–
1465.
Laiolo, P. (2010). The emerging significance of bioacoustics in animal species conser-
vation. Biological Conservation, 143, 1635–1645.
Longcore, T. (2010). Sensory ecology: Night lights alter reproductive behavior of blue
tits. Current Biology, 20, R893–R895.
MacLean, E. L. & B. Hare. (2012). Bonobos and chimpanzees infer the target of
another’s attention. Animal Behaviour, 83, 345–353.
Martin, G. R. & J. M. Shaw. (2010). Bird collisions with power lines: Failing to see
the way ahead? Biological Conservation, 143, 2695–2702.
McCauley, S. J. & K. A. Bjorndal. (1999). Conservation implications of dietary dilu-
tion from debris ingestion: Sublethal effects in post-hatchling loggerhead sea
turtles. Conservation Biology, 13, 925–929.
Meretsky, V. J., N. F. R. Snyder, S. R. Beissinger, D. A. Clendenen, & J. W. Wiley.
(2000). Demography of the California condor: Implications for reestablish-
ment. Conservation Biology, 14, 957–967.
Miller, H. C., J. A. Moore, N. J. Nelson, & C. H. Daugherty. (2009). Influence of
major histocompatibility complex phenotype on mating success in a free-
ranging reptile population. Proceedings of the Royal Society of London, B, 276,
1695–1704.
Moore, J. A., B. D., Bell, & W. L. Linklater. (2008). The debate on behavior in con-
servation: New Zealand integrates theory with practice. BioScience, 58, 454–
459.
Mumme, R. L., S. J. Schoech, G. E. Woolfenden, & J. W. Fitzpatrick. (2000). Life
and death in the fast lane: Demographic consequences of road mortality in
the Florida scrub-jay. Conservation Biology, 14, 501–512.
Myers, N. (1995). Environmental unknowns. Science, 269, 358–360.
Paz-y-Miño-C, G. (2003). Behavioral unknowns: An emerging challenge for conser-
vation. The Conservation Behaviorist, 1, 2.
Paz-y-Miño-C, G. (2005). Animal cognition and its role in conservation behavior.
The Conservation Behaviorist, 3, 4–5,9.
Paz-y-Miño-C, G. (2006). Contribution of animal behavior research to conservation
biology. The Conservation Behaviorist, 4, 5.
Paz-y-Miño-C, G. (2007). Behavior and conservation in the Galápagos. The Conser-
vation Behaviorist, 5, 4–5.
Paz-y-Miño-C, G. (2011). Did conservation behavior need a “primer”? It arrived! The
Conservation Behaviorist, 9, 4.
Paz-y-Miño-C, G., A. B. Bond, A. C. Kamil, & R. P. Balda. (2004). Pinyon jays use
transitive inference to predict social dominance. Nature, 430, 778–781.
COGNITIVE REPERTOIRE, COGNITIVE LEGACIES, AND COLLECTIVE REASONING 309

Rayburn-Reeves, R. M., M. Molet, & T. R. Zentall. (2011). Simultaneous discrimi-

nation reversal learning in pigeons and humans: Anticipatory and perseverative
errors. Learning and Behavior, 39, 125–137.
Reed, J. M. (1999). The role of behavior in recent avian extinctions and endanger-
ments. Conservation Biology, 13, 232–241.
Rodgers, J. A. & S. T. Schwikert. (2002). Buffer-zone distances to protect foraging
and loafing waterbirds from disturbance by personal watercraft and outboard-
powered boats. Conservation Biology, 16, 216–224.
Rodgers, J. A. & H. T. Smith. (1995). Set-back distances to protect nesting bird col-
onies from human disturbance in Florida. Conservation Biology, 9, 89–99.
Rodrigues, D., B. W. Goodner, & M. R. Weiss. (2010). Reversal learning and
risk-averse foraging behavior in the monarch butterfly, Danaus plexippus
(Lepidoptera: Nymphalidae). Ethology, 116, 270–280.
Rosati, A. G. & B. Hare. (2009). Looking past the model species: Diversity in gaze-
following skills across primates. Current Opinion in Neurobiology, 19, 45–51.
Schmajuk, N. A. (1997). Animal Learning and Cognition. New York: Cambridge
University Press.
Seed, A. & R. Byrne. (2010). Animal tool-use. Current Biology, 20, R1032–R1039.
Seed, A., N. Emery, & N. Clayton. (2009). Intelligence in corvids and apes: A case of
convergent evolution? Ethology, 115, 401–420.
Shettleworth, S. J. (2001). Animal cognition and animal behaviour. Animal Behav-
iour, 61, 277–286.
Shettleworth, S. J. (2009). Animal cognition: Deconstructing avian insight. Current
Biology, 19, R1039–R1040.
Shettleworth, S. J. (2010). Cognition, Evolution, and Behavior. New York: Oxford
University Press.
Slabbekoorn, H. & E. A. P. Ripmeester. (2008). Birdsong and anthropogenic noise:
Implications and applications for conservation. Molecular Ecology, 17, 72–83.
Stamps, J. A. & R. R. Swaisgood. (2007). Someplace like home: Experience, habitat
selection and conservation biology. Applied Animal Behaviour Science, 102,
392–409.
Stone, P. A., H. L. Snell, & H. M. Snell. (1994). Behavioral diversity as biological
diversity: Introduced cats and lava lizard wariness. Conservation Biology, 8,
569–573.
Swaisgood, R. (2010). Conservation and animal behavior. In M. D. Breed & J.
Moore (eds.), Encyclopedia of Animal Behavior (pp. 359–365). Oxford, UK:
Academic Press.
Terrace, H. S. & L. K. Son. (2009). Comparative metacognition. Current Opinion in
Neurobiology, 19, 67–74.
Tinbergen, N. (1963). On aims and methods of ethology. Zeitschrift für Tierpsychol-
gie, 20, 410–433.
Tommasi, L., C. Chiandetti, T. Pecchia, A. Sovrano, & G. Vallortigara. (2012).
From natural geometry to spatial cognition. Neuroscience and Biobehavioral
Reviews, 36, 799–824.
310 ANIMAL BEHAVIOR

Tregenza, T. & N. Wedell. (2000). Genetic compatibility, mate choice and patterns
of parentage: Invited review. Molecular Ecology, 9, 1013–1027.
Tuomainen, U. & U. Candolin. (2011). Behavioural responses to human-induced
environmental change. Biological Reviews, 86, 640–657.
Tutin, C. E. G., M. Ancrenaz, J. Paredes, M. Vacher-Vallas, C. Vidal, B. Goossens,
M. W. Bruford, & A. Jamart. (2001). Conservation biology framework for
the release of wild-born orphaned chimpanzees into the Conkouati Reserve,
Congo. Conservation Biology, 15, 1247–1257.
Udell, M. A. R., N. R. Dorey, & C. D. L. Winne. (2011). Can your dog read your
mind? Understanding the causes of canine perspective taking. Learning and
Behavior, 39, 289–302.
Valutis, L. L. & J. M. Marzluff. (1999). The appropriateness of puppet-rearing birds
for reintroduction. Conservation Biology, 13, 584–591.
van Schaik, C. P. & J. M. Burkart. (2011). Social learning and evolution: The cultural
intelligence hypothesis. Philosophical Transactions of the Royal Society of
London, B, 366, 1008–1016.
Wasserman, E. A. & T. R. Zentall (eds.). (2006). Comparative Cognition. New York:
Oxford University Press.
Whitehead, H. (2010). Conserving and managing animals that learn socially and
share cultures. Learning and Behavior, 38, 329–336.
Yeager, C. P. (1997). Orangutan rehabilitation in Tanjung Putin National Park,
Indonesia. Conservation Biology, 11, 802–805.
Zentall, T. R., E. A. Wasserman, O. F. Lazareva, R. K. R. Thompson, & M. J.
Rattermann. (2008). Concept learning in animals. Comparative Cognition
and Behavior Reviews, 3, 13–45.
 12

Beauty in the Eye of the Beholder:

Potentially Adaptive Individual Differences
in Human Mate Preferences
Lisa M. DeBruine, Benedict C. Jones, S. Craig Roberts, Anthony C. Little,
and Christopher D. Watkins

INTRODUCTION
Mate choice in humans is often portrayed as more nuanced and individualistic
than mate choice in nonhuman animals, but recent research on individual
differences in human mate choice has drawn much inspiration from studies
of animals as diverse as stickleback, quail, and mice. In this chapter, we will
cover four main areas in which theory and findings from studies of individual
differences in nonhuman animals’ mate choices have informed research on the
human animal. First we will explain how an individual’s own condition influ-
ences the quality of the mates he or she prefers. Next we will describe how
adaptations that function to avoid close inbreeding or extreme outbreeding
can result in mate preferences that are individually tuned through experience
with kin. Then we will cover how mate preferences can be contingent on
aspects of the environment, creating systematic differences between individ-
uals and even populations. Finally, we will address the role of genetic variation
in the major histocompatibility complex (MHC) in shaping mate choice in
complex ways. In conclusion, we will highlight the ways in which seemingly
312 ANIMAL BEHAVIOR

uniquely human patterns of complexity in mate choice can be explained by

the same principles used to explain individual differences in mate choice in
other animals.

CONDITION DEPENDENCE AND MATE CHOICE

Theory
One potential source of individual differences in partner preferences and
choice is the quality, or attractiveness, of the phenotype that individuals find
themselves inhabiting. Certainly quality has been found to influence the
expression of physical traits linked to sexual selection (Møller & Petrie,
2002; Hedrick, 2005; Delcourt & Rundle, 2011). Alongside evidence for
such condition-dependent traits, there is also evidence for condition-
dependent mate preferences and choice in humans (e.g., Little et al., 2001;
Penton-Voak et al., 2003; Vukovic et al., 2010) and nonhuman animals
(e.g., Bakker et al., 1999; Mazzi et al., 2003). There are several potential
mechanisms via which condition and attractiveness could impact on prefer-
ences and mate choice. We briefly explore three explanations here, covering
different levels of biological explanation: genetic, energetic, and behavioral.
Genetic basis—One possible genetic explanation for condition-dependent
mate choice is that those who are the daughters of males who lack good-gene
markers do not inherit a preference for such markers because both the genes
that develop the markers in males and those that determine the preference in
females are closely linked on the genome. In this way daughters of high-
quality fathers inherit and daughters of low-quality females do not inherit
both types of gene, as well as the condition of their fathers, linking condition
and preference (Bakker et al., 1999). Thus heritable condition can determine
preferences due to a classic assumption of Fisherian models of sexual
selection—linkage between genes for advertising traits and genes for the pref-
erences for such traits (Fisher, 1930). An equivalent argument can be made for
male offspring.
Energetic basis—Energetic costs may also play a role in condition-
dependent mate choice. Because searching for and choosing a mate carries an
energy cost, those individuals in better condition may better afford this cost
and so inspect more potential mates in order to maximize mate quality and
hence are likely to set a higher threshold of minimal attractiveness acceptable
in a mate (López, 1999). In general it might be expected that low-quality
females may be least likely to have energy to spare for mate choice, and so
energetic cost to choice is a plausible explanation for condition-dependent
BEAUTY IN THE EYE OF THE BEHOLDER 313

preferences. Low-condition individuals may simply be unable to be choosy

due to the physiological costs involved in making choices.
Behavioral basis—The third explanation invoked here involves a behavioral
response to an individual’s own condition or attractiveness, as condition influ-
ences competitiveness for mates and thus impacts mating strategy. It is poten-
tially adaptive for females of low mate value in order to avoid the costs of
decreased parental investment or potential desertion from high-quality part-
ners (Little et al., 2001). Such reasoning arises from notions that high-
quality males are least likely to invest in or even most likely to desert their part-
ners (Gangestad & Simpson, 2000). High-quality females are most likely to be
able to extract investment from high-quality partners. In this way the mate
value of a woman may determine her preferences for quality due to differing
costs and benefits of mating with high-quality males between females of differ-
ent mate values. In this view, condition, quality, or attractiveness are all terms
that relate to an individual’s mate value or, perhaps more descriptively, “mar-
ket value.” Market value in mate choice is useful as it refers to how much
demand there is for a particular individual as a mate within a population
(Pawlowski & Dunbar, 1999).

Evidence from Nonhuman Animals

Evidence for condition-dependent preferences and choice has been
observed in several nonhuman animal species. Several studies have examined
how condition relates to how attentive an individual is to phenotypic markers
of quality in potential mates. Work by Manfred Milinski and Theo Bakker
(1990) has shown that in threespined stickleback (Gasterosteus aculeatus), para-
sites reduce the intensity of the red coloration around males’ throats. This col-
oration is important to female mate choice, and females preferentially mate
with males with more intense color. When cues to color were removed (by
using green light to make the differences in red coloration invisible), males
that were previously preferred were chosen at levels around chance, although
males’ courtship displays remained the same. Thus females use the intensity
of red coloration to avoid parasitized males and therefore select males with
good immune systems. Importantly, the phenotypic quality of an individual
affects its preferences for members of the opposite gender. Bakker, Reto Kun-
zler, and Dominique Mazzi (1999) allowed female threespined sticklebacks to
choose between computer animations of courting males. The two males dif-
fered only in the intensity of red coloration (red versus orange) around the
throat and so represented an “attractive” versus “unattractive” phenotype. A
significant correlation was found between female condition (measured as body
314 ANIMAL BEHAVIOR

weight or body size) and mate choice. Females who were in better condition
(relatively heavier) showed a greater preference for the red-throated male,
and those in worse condition showed a preference for the orange male. As
these females were raised in laboratory conditions and isolated from males
before becoming reproductively active, these preferences are not dependent
on experience of competition in mating. Further evidence for condition-
dependent preferences is also seen in sticklebacks. Symmetry is a trait that
has long been associated with quality, and female sticklebacks in poor physical
condition do not show as strong a preference for symmetry as those in better
condition (Mazzi et al., 2003).
Another example of condition-dependent mate choice comes from another
fish species. The parasite Gyrodactylus turnbulli reduces the sexual display and
color intensity of male guppies (Poecilia reticulata), which makes them less
attractive to females (Houde & Torio, 1992). Silvia López (1999) examined
how this same parasite influences female mate-choice decisions. Infected and
uninfected females were presented with a choice of two males, one attractive
(high display rate, high color intensity), one unattractive (low display rate,
low color intensity). When presented with these two males, uninfected
females were significantly more likely to choose the attractive male over the
less attractive male. By contrast, infected females were less discriminative in
their choice of mates and showed no preference for the more attractive male
over the less attractive male. Females were raised in laboratory conditions in
same-gender groups and were virgin when presented with males to choose
between, thereby minimizing the possibility that learning played any role in
the acquisition of these preferences. As well as the decreased discrimination,
activity during choice trials also decreased with increasing parasite load, imply-
ing that perhaps such females are unable to engage in energetically costly
choice (López, 1999).
Beyond fish species, condition-dependent choice is also seen in amphibians
and birds. For example, in spadefoot toads (Pelobates fuscus), males in better
condition produce longer calls (Pfennig & Tinsley, 2002). Parasitized females
do not show a preference for male call length whereas unparasitized females
prefer longer calls, which are indicative of a male in good condition (Pfennig
& Tinsley, 2002).
In birds, zebra finches (Taeniopygia guttata) provide an interesting example
of flexibility in mating strategy according to relative attractiveness. Females
made attractive with red leg bands were found to spend less time carrying
out parental duties than both those typical of their sex and of unattractive
females (those with green bands that are unattractive to males) but still had
BEAUTY IN THE EYE OF THE BEHOLDER 315

higher reproductive success, possibly because mates of attractive females spent

more time than that typical of their sex carrying out parental duties (Burley,
1986). Female zebra finches thus apparently reacted to their higher attractive-
ness, adjusting their strategy, and this change in attractiveness also induced
their partners to behave differently. Further, experimental manipulation of
female quality, via rearing in small or large broods, has shown that high-
quality females prefer high-quality males’ songs, while low-quality females pre-
fer low-quality males’ songs (Holveck & Riebel, 2010). In this species,
condition-dependent mate choice is evident and also potentially linked to
behavioral flexibility in assessing individual mate value.
The three theoretical reasons for condition-dependent mate choice outlined
earlier have then all received some support from different researchers. The
genetic explanation is favored by the researchers studying stickleback prefer-
ences (Bakker et al., 1999) because of evidence that daughters of high-
quality fathers inherit and daughters of low-quality females do not inherit
both types of gene, as well as the condition of their fathers (Bakker, 1993).
Work on how parasites influence choosiness in guppies supports the energetic
explanation (López, 1999). Finally, the flexibility in mating behavior in zebra
finches potentially supports the behavioral market-value explanation (Burley,
1986).

Evidence from Humans

In humans, research has focused on various measures of condition or qual-
ity as well as proposed markers of genetic quality. Several studies have focused
on preferences for face traits such as symmetry of masculinity or femininity,
traits that are proposed to indicate mate quality (see Little, Jones, &
DeBruine, 2011, for review). For example, Anthony Little and colleagues
(2001) found that women who thought they were physically attractive pre-
ferred more masculine faces and had greater preferences for symmetry than
those women who thought they were less physically attractive. Such
condition-dependent choice for male facial masculinity in women has also
been shown for actual cues of attractiveness via other-rated facial attractiveness
and measured waist-to-hip ratio (Penton-Voak et al., 2003; Smith et al.,
2009). More attractive women prefer more masculine-faced men than do less
attractive women.
Similar findings have been found when focusing on preferences for facial
health. Condition can encompass physical (e.g., health) and psychological fac-
tors (e.g., stress, anxiety, and depression). Benedict Jones and colleagues
(2005) found that women with attractive and healthy (low) waist-to-hip ratios
316 ANIMAL BEHAVIOR

or who scored low on psychological stress measures expressed greater attrac-

tion to healthy male faces than did women with relatively high waist-to-hip
ratios or who scored relatively high on psychological measures. Further, exam-
ining the role of behavioral quality influencing preferences in men, Jones and
colleagues (2007) have shown that men scoring high on sensation seeking—
men who possess good physical condition and so can afford to pursue high-
risk activities—demonstrate stronger preferences for female face femininity
than their peers who exhibit less risk-taking behavior.
Beyond faces, analogous findings are seen in women’s preferences for mas-
culine male voices. In humans, fundamental frequency (an acoustic correlate
of voice pitch) is sexually dimorphic, with men displaying lower voice pitch
than women, and women tend to prefer men’s voices with low pitch (Feinberg
et al., 2005). As in faces, self-rated attractiveness is positively related to wom-
en’s preferences for masculinized men’s voices (voices with lower pitch)
(Vukovic et al., 2010). Finally, again examining face preferences, Little and
Helena Mannion (2006) found that viewing highly attractive women
decreased women’s self-rated attractiveness and also found that masculinity
preferences were stronger after viewing unattractive women than after viewing
attractive women. This latter finding suggests that it is women’s subjective
impressions of their own market value (i.e., their self-rated attractiveness) that
are particularly important for these individual differences in masculinity pref-
erences (Little & Mannion 2006).
The three theoretical reasons for condition-dependent mate choice outlined
earlier have not all been addressed by researchers studying humans. The
genetic and energetic explanations remain plausible. The behavioral explana-
tion for human preferences put forward by Little and colleagues (2001) is very
much linked to notions of the perceived and real market value of an individual
and thus open to change in line with population change. Demonstrating that
masculinity preferences can be affected by manipulation of self-rated opinion
(Little & Mannion, 2006) highlights flexibility in condition-dependent mate
choice and is in line with the behavioral, market-value-dependent explanation.
The flexibility in mating behavior in zebra finches potentially supports the
behavioral market-value explanation (Burley, 1986), linking human and non-
human explanations of condition-dependent mating preferences. If condition-
dependent preferences in humans reflect market value, then stronger prefer-
ences for high-quality mates among individuals who are of high quality may
be adaptive if high-quality individuals are better able to attract or retain
high-quality mates and thus are better able to offset the possible negative con-
sequences of choosing a high-quality partner (Jones et al., 2005; Little et al.,
2001; Little & Mannion, 2006; Penton-Voak et al., 2003).
BEAUTY IN THE EYE OF THE BEHOLDER 317

KINSHIP AND MATE CHOICE

Theory
Another source of individual differences in mate choice is kin-biased mat-
ing. While kin-biased mating results from species-wide adaptations, the effect
produces individual differences in mate choice because each individual’s fam-
ily template is learned from a different family group.
Two lines of reasoning predict that selection will favor the ability to bias
behavior based on kinship under some circumstances. Inclusive fitness theory
(Hamilton, 1964) demonstrates the adaptive value of allocating effort toward
others as a function of the probability they share copies of your genes that
are identical by descent. However, recognition and categorization of kin are
also important in mate choice because of the well-established costs of close
inbreeding and extreme outbreeding (e.g., mating with a member of the
wrong species; Bateson, 1980). In addition to inclusive fitness theory and opti-
mal outbreeding theory, the phenomenon of parental imprinting may also
lead to mate choice that is biased by kin characteristics. While this chapter is
focused on mate choice, it is instructive here to also consider altruistic behav-
ior that is modulated by kinship cues, as the dissociations between social and
sexual behavior are some of the strongest evidence that responses to potential
kinship cues are adaptive (DeBruine et al., 2008).

Evidence from Nonhuman Animals

Altruism—Selection should favor behaviors that result in investment in
others being modulated by relatedness. Examples of nepotism abound in
nature, including insects (Greenberg, 1979), amphibians (Harris et al.,
2003; Pfennig et al., 1994), fish (Olsén, 1999), birds (Bukacinski et al.,
2000; van der Jeugd et al., 2002), and mammals (Alberts, 1999; Heth et al.,
2003). Nepotism is expressed in many different ways, such as alarm calling
in the presence of relatives (Hauber & Sherman, 1998; Sherman, 1977),
kin-biased dominance interactions (Silk, 2002), and cooperative breeding
(Griffin & West, 2003).
Optimal mate choice—Selection should also favor behaviors that lead to
decreased mating with close relatives and too-distantly related individuals in
order to obtain an optimal level of outbreeding. While matings between
closely related individuals can be deleterious, matings between too-distantly
related individuals can also carry costs such as the disruption of coadapted
gene complexes or suppression of genes adapted for specific environments
(Bateson, 1983). The most extreme cost of outbreeding is hybrid sterility;
traits functioning to prevent cross-species matings are likely to increase fitness.
318 ANIMAL BEHAVIOR

Organisms can regulate the genetic relatedness of mates by recognizing fea-

tures characteristic of close kin and using this information when making deci-
sions about mating partners.
Another benefit of pairing with similar individuals may come not from
genetics but behavior, via increases in partnership stability. Work on birds also
suggests that behavioral compatibility may be linked to reproductive success in
that birds that were more similar were more likely to have a greater number of
offspring than those that were dissimilar (Spoon et al., 2006).
Imprinting—There is a large body of literature on nonhuman animal
imprinting (see Mateo, 2004, for a review). Many studies of nonhuman spe-
cies have examined the effects of early exposure to parental characteristics on
later mate preferences, a phenomenon usually described as sexual imprinting.
Positive visual imprinting (an attraction to visible parental characteristics) has
been demonstrated in both birds (quail: Bateson, 1980; zebra finches: Vos,
1995) and ungulates (sheep and goats; Kendrick et al., 1998), and there is
even suggestive evidence that it occurs in primates (Fujita, 1993).

Evidence from Humans

Altruism—While mammalian mothers have almost 100 percent confidence
in their maternity, the same is not always true for fathers. Investment in young
that varies with cues indicating their probability of genetic relatedness is likely
to have been favored by natural selection (Daly & Wilson, 1982). Consistent
with this, Steven Gaulin and Alice Schlegel (1980) linked paternal confidence
to investment in a cross-cultural sample of 186 preindustrial societies. Addi-
tionally, matrilateral kin (on the mother’s side), who have relatively high cer-
tainty of relatedness, invest more in children than patrilateral kin (father’s
side), who have greater reason to doubt their relatedness (Euler & Weitzel,
1999; Gaulin et al., 1997).
Some experimental evidence suggests that facial resemblance is used as a
cue of kinship that influences prosocial decision making. DeBruine (2002)
demonstrated that people are more likely to trust those who resemble them-
selves when playing an interactive investment game common to experimental
economic research. This “trust game” gave the first player a choice between
evenly splitting a small sum of money between self and the second player or
entrusting a larger sum of money to the second player, who could divide it
equally or selfishly. When the pictured game partner had been subtly manip-
ulated to resemble the experimental participant, he or she was more likely to
trust the partner.
Self-resemblance has also been shown to affect behavior in a group-based
economic game, the “public goods game” (Krupp et al., 2008). In this game,
BEAUTY IN THE EYE OF THE BEHOLDER 319

each individual in a group of four players is given an amount of money and

can decide to donate any proportion of this to a group pot, after which the
total amount in the pot is multiplied and shared equally among all members
of the group, regardless of their initial contribution. Donations to the group
pot, a measure of group cooperation, increased as the number of self-
resembling faces in the group increased.
Optimal mate choice—In humans, mating between close relatives is associ-
ated with the risk of autosomal recessive genetic disorders and miscarriage
(Bittles, 2001), although a positive association between consanguinity
(“blood relation”) and fertility has been found due to other factors associated
with consanguineous marriages, such as earlier age at first reproduction and
longer duration of marriage (Bittles et al., 2002). Similarly, a study of all
recorded marriages in Iceland between 1800 and 1965 showed that third-
and fourth-cousin marriages had significantly higher reproductive success than
either more closely related couples or more distantly related couples (Helgason
et al., 2008).
In humans, a specialized mechanism for avoidance of inbreeding among
close kin has been postulated in the form of the Westermarck effect (Fessler
& Navarrette, 2004; Lieberman et al., 2003; Westermarck, 1921; Wolf,
1995). This refers to the lack of sexual attraction between people who were
closely associated as young children. In most circumstances, such people are
likely to be close genetic relatives; thus, the Westermarck effect functions to
prevent inbreeding. Westermarck’s hypothesis has received empirical support
from a series of ethnographic studies where male and female nonsiblings are
raised together in a way similar to real siblings (Shepher, 1971; Wolf, 1993).
Across these studies, children growing up together avoided later sexual interac-
tion, even when in arranged marriages, despite not being genetically related to
one another.
Debra Lieberman and colleagues (2007) later proposed an additional cue
for kin detection that is separate from coresidence duration: maternal perina-
tal association (MPA). MPA refers to the association between a mother and
her newborn baby. For example, an older sibling witnessing the close relation-
ship between his or her mother and a new baby allows the older sibling to
experience the cue of MPA for the younger sibling. This mechanism provides
a very reliable way for older siblings to detect younger siblings but does not
allow for younger siblings to detect older siblings, whose MPA occurred before
the younger sibling’s birth. Lieberman and colleagues (2007) suggested that
humans assess relatedness through the support of several systems of kin detec-
tion, including MPA and coresidence duration. In support of this hypothesis,
Lieberman and colleagues (2007) found that sibling altruism, aversions to
320 ANIMAL BEHAVIOR

third-party sibling incest, and aversions to personally engaging in sibling incest

were higher for older siblings, who did experience MPA, than for younger sib-
lings, who did not experience MPA. Also, these altruism and incest-avoidance
measures correlated with coresidence duration for younger siblings, who do
not experience MPA, but not for older siblings, who do experience MPA.
Lieberman and colleagues (2007) concluded that the cue of MPA overrides
the cue of coresidence duration.
The use of MPA as a cue of kinship has been supported by several sub-
sequent studies. While Arthur Wolf (1968) found that the age of the bride
at first association predicted marital failure and infertility in sim-pua marriages
(a Chinese tradition of arranged marriages), Lieberman (2009) examined these
data further and found that only the age at first association of the younger
marriage partner predicted fertility, consistent with the idea that older siblings
do not use coresidence information because they rely on MPA. Additionally,
Gwenaël Kaminski and colleagues (2010) proposed that later-borns would
be able to perceive facial resemblance, a putative cue of kinship (reviewed in
DeBruine et al., 2008), more effectively than first-borns if later-borns need
to use alternative methods of kin recognition in the absence of MPA. In two
facial-resemblance detection tasks, later-borns performed better than first-
borns (Kaminski et al., 2010), supporting the idea that later-borns rely more
heavily on other mechanisms to infer relatedness.
Experimental manipulations of facial resemblance have also been used to
investigate mate preferences. For example, images of same-sex faces manipu-
lated to resemble a viewer were judged as more attractive by that research par-
ticipant than by others (DeBruine, 2004). In contrast, resemblance had a
significant but much smaller effect on the attractiveness of other-sex faces, in
line with the prediction that the costs of inbreeding will temper preferences
for self-resemblance in other-sex faces. This finding is consistent with that of
Penton-Voak, Perrett, and Pierce (1999), who found small and inconsistent
preferences for self-resembling face shapes in opposite-sex stimuli.
Following this result, Lisa DeBruine (2005) showed that self-resemblance
increased attributions of trustworthiness to other-sex face images but had a
smaller or negative effect on attributions of attractiveness for long-term and
short-term relationships. As the effect of self-resemblance on the same faces
was different in the three contexts of trustworthiness, long-term relationship
attractiveness, and short-term relationship attractiveness, this result provides
the most convincing evidence that responses to facial self-resemblance are
context-dependent. These context-dependent effects are difficult to interpret
in terms of the mere exposure effect (Zajonc et al., 1987). Indeed, Gavin
Buckingham and colleagues (2006) found that mere visual experience with
BEAUTY IN THE EYE OF THE BEHOLDER 321

male faces caused equivalent increases in perceptions of trustworthiness and

attractiveness of similar faces for both male and female participants.
These findings demonstrate that facial resemblance can affect attributions
and behavior toward others. Divergent effects of facial resemblance in the
domains of prosocial attributions and mate choice clearly refute the hypothesis
that responses to facial resemblance are nonadaptive byproducts of perceptual
phenomena involved in face processing (e.g., mere exposure) and support the
existence of specialized adaptations for kin recognition by facial phenotype
matching.
More recently, studies using experimentally produced facial resemblance
have investigated the role of family composition on inbreeding-relevant
responses. DeBruine, Jones, Watkins, and colleagues (2011) tested
156 women, who judged the trustworthiness and attractiveness of male and
female faces that had been experimentally altered to resemble their own face.
Self-resemblance biases were calculated by subtracting a matched control par-
ticipant’s judgments of these same faces from the experimental participant’s
judgments. Having opposite-sex siblings influenced inbreeding-relevant per-
ceptions of facial resemblance (i.e., male attractiveness) but not prosocial per-
ceptions (i.e., male trustworthiness, female attractiveness, and female
trustworthiness). Women with brothers were less attracted to self-
resembling, unfamiliar male faces than were women without brothers, while
both groups found self-resemblance to be equally trustworthy for the same
faces. This effect was stronger in women with younger, rather than older,
brothers, consistent with the proposal that only younger siblings exhibit the
highly reliable kinship cue of MPA (Lieberman et al., 2007). These findings
suggest that experience with opposite-sex siblings can directly influence
inbreeding-avoidance mechanisms and demonstrate a striking functional dis-
sociation between the mechanisms that regulate inbreeding and the mecha-
nisms that regulate prosocial behavior toward kin.
Imprinting—Similarity between partners and parents also appears apparent
in humans. Two studies have examined paternal and partner age in women,
with small but consistently positive correlations between these variables, indi-
cating that the daughters of older men subsequently tend to choose older part-
ners (Wilson & Barrett, 1987; Zei et al., 1983). Following from this work
showing links between parental age and actual partner age, David Perrett
and colleagues (2002) investigated whether parental age impacted preferences
for faces of different ages. If offspring are attracted to parental characteristics,
individuals born to old parents should be more attracted to older faces than
individuals born to young parents. Using computer-graphic faces Perrett and
colleagues (2002) did indeed find that women born to old parents were
322 ANIMAL BEHAVIOR

relatively less impressed by youth and more positive to age cues in male faces
than women with young parents. For men judging female faces, preferences
appeared to be influenced only by the opposite-sex parent, the mother.
Eye color and hair color are also stable traits that can be observed in parents
across childhood. Wilson and Barrett (1987) showed a trend for women to
choose partners whose eye color resembled their father’s, though this finding
was confounded with own eye color. Following this study, Little, Penton-
Voak, Burt, and Perrett (2003) have shown that there are generally positive
correlations between self and partner’s hair and eye color but that in regression
these relationships are explained by a positive relationship between parental
traits and partner traits. Such effects were somewhat specific to opposite-sex
parental traits, suggesting a focused mechanism potentially akin to imprinting.
Further evidence for familial imprinting in humans comes from studies of
preferences for parental traits. For example, romantic partners and opposite-
sex parents tend to be similar in measured facial proportions (Bereczkei et
al., 2009), ethnicity (Jedlicka, 1980), and general facial appearance (Bereczkei
et al., 2004). Additionally, the extent to which romantic partners or preferred
faces resemble opposite-sex parents is positively correlated with their reported
emotional closeness to the opposite-sex parent (Bereczkei et al., 2002, 2004;
Wiszewska et al., 2007).
Because children resemble their parents (Brédart & French, 1999; Bressan
& Dal Martello, 2002; Bressan & Grassi, 2004; Nesse et al., 1990; Oda et
al., 2002, familial imprinting could also influence preferences for self-
resemblance discussed above (e.g., DeBruine, 2002, 2004, 2005; DeBruine,
Jones, Little, et al., 2011). Christopher Watkins and colleagues (2011)
showed that women’s preferences for self-resemblance in opposite-sex, but
not same-sex, anonymous faces was correlated with the women’s rated emo-
tional closeness to their father but not their mother. These findings link the
research on human kin recognition and imprinting, implicating familial
imprinting in responses to self-resemblance, at least for sexual judgments of
opposite-sex individuals.

ENVIRONMENT AND MATE CHOICE

Theory
Recent formulations of sexual selection theory predict that mate choice may
be affected by environmental factors, such as predation risk, resource quality,
and the environment-contingent costs of sampling mates (see Jennions &
Petrie, 1997, for a review). For example, female threespined sticklebacks show
weaker preferences for males displaying cues of good physical condition if they
BEAUTY IN THE EYE OF THE BEHOLDER 323

are required to swim against a stronger current between mates (Milinski &
Bakker, 1992). This effect of current strength on female choosiness may occur
because the environmental factor (the current) alters the costs of sampling
additional mates (Milinski & Bakker, 1992). Another example of environ-
mental factors influencing mate choice in nonhuman species comes from stud-
ies of Uganda kob (Kobus kob). These studies suggest that the extent to which
females are attracted to males in territories within leks that have good visibility
of the surrounding area, rather than using phenotypic conditions to guide
their choices, is at least partly dependent on the risk of predation (Deutsch
& Weeks, 1992). These are just two of many examples of environmental factors
modulating either the extent to which preferences for physical cues are expressed
or the nature of physical cues that are preferred that have been uncovered in stud-
ies of mate choice in nonhuman species. Importantly, these types of facultative
preferences can help explain geographical differences in mate choices in nonhu-
man animals (Jennions & Petrie, 1997) and can stimulate studies of geographic
variation in human mate preferences (e.g., Gangestad & Buss, 1993). This sec-
tion of our chapter will discuss the evidence that at least three types of environ-
mental factors (pathogen stress, resource scarcity, and the sex ratio of the local
population) have potentially important effects on women’s mate preferences.

Evidence from Humans and Nonhuman Animals

In some nonhuman species, such as peacocks, masculine physical character-
istics are positively correlated with measures of males’ physical health
(Jennions & Petrie, 1997). More recently, similar correlations between physi-
cal masculinity and men’s health have also been reported in several studies. For
example, men rated by others as looking particularly masculine show lower
levels of markers of oxidative stress in their urine samples, a sign that they
are in good physical condition, than do their relatively feminine peers
(Gangestad et al., 2010). Additionally, other studies have reported that mea-
sures of men’s facial masculinity are positively correlated with estimates of
their global health derived from analyses of their medical records (Rhodes et
al., 2003) and the number and duration of respiratory illnesses that they report
having experienced (Thornhill & Gangestad, 2006). Given that facial mascu-
linity appears to be a somewhat valid cue to men’s health, researchers have
suggested that it may be a trait that women value particularly highly in poten-
tial mates in contexts where pathogens pose a particularly serious threat to sur-
vival (DeBruine, Jones, Crawford, et al., 2010; Penton-Voak et al., 2004).
Consequently, several studies have recently investigated the possibility that
measures of actual or perceived pathogen stress might predict population-
level and individual-level variation in women’s preferences for masculine men.
324 ANIMAL BEHAVIOR

In a country-level analysis of the average masculinity preferences of women

in 30 countries, DeBruine, Jones, Crawford, and colleagues (2010) observed a
relatively strong correlation between a national health index derived from
principle component analysis of a range of World Health Organization statis-
tics (e.g., mortality rates, life expectancies, and the impact of communicable
disease) and the extent to which women preferred masculine shapes cues over
feminine shape cues in images of young men’s faces. This correlation indicated
that women tended to show stronger preferences for masculine men in coun-
tries with poorer health and was not an artifact of the possible effects of differ-
ences in the countries’ wealth or women’s sociosexuality (i.e., their interest in
pursuing short-term sexual relationships). Moreover, an analysis of regional
variation in masculinity preferences in the United States revealed a similar cor-
relation between women’s preferences for masculine men and a state health
index derived from U.S. Centers for Disease Control statistics (DeBruine,
Jones, Little, et al., 2011). Both of these findings were replicated when the
data sets were reanalyzed using Fincher and Thornhill’s Combined Parasite
Stress Index (Fincher & Thornhill, 2012), a more direct measure of pathogen
stress, suggesting that the original findings were not an artifact of the specific
health measures that had been used (DeBruine et al., 2012). Although some
researchers have recently proposed that violence-related factors, such as homi-
cide rates, might be a better predictor of regional differences in women’s mas-
culinity preferences than is health (Brooks et al., 2011), the correlation
between homicide rates and masculinity preferences in the sample of 30 coun-
tries was very fragile and was not significant when controlling for the possible
effects of regional differences in wealth (DeBruine, Jones, Little, et al., 2011).
Additionally, homicide rates did not predict regional differences in masculin-
ity preferences among U.S. states (DeBruine, Jones, Little, et al., 2011). These
latter findings raise doubts about Brooks and colleagues’ (2011) claim.
Collectively, the findings described above suggest that regional differences
in the extent to which pathogens pose a risk to health and survival predict
regional differences in women’s masculinity preferences. These findings com-
plement individual-level analyses of women’s masculinity preferences, which
have revealed positive correlations between the strength of women’s disgust
responses to scenarios depicting pathogens and women’s preferences for mas-
culine men (DeBruine, Jones, Tybur, et al., 2010). However, while such cor-
relational analyses of women’s mate preferences can be instructive, they do not
allow strong conclusions to be made about the possible causal link between
masculinity preferences and pathogens. Consequently, more recent work has
used priming paradigms to experimentally manipulate women’s exposure to
pathogen-related cues. For example, Little, DeBrune, and Jones (2011) found
BEAUTY IN THE EYE OF THE BEHOLDER 325

that viewing images depicting likely sources of pathogens (e.g., bowls of body
fluids) increased women’s preferences for masculine characteristics in men’s,
but not women’s, faces. By contrast, viewing control images did not. Consis-
tent with these findings, other work has found that priming women’s concerns
about pathogens by having them complete a questionnaire assessing percep-
tions of their own vulnerability to disease tended to increase their preferences
for putative masculine traits in potential mates (Lee & Zietsch, 2011). These
findings suggest that concerns about pathogens can directly influence women’s
preferences for masculine men and raise the possibility that regional differences
in women’s masculinity preferences might reflect differences in the extent to
which those women have recently been exposed to pathogen cues.
While the work described above focused on investigating the role of patho-
gens in masculinity preferences, other work has focused more on the possible
effects of resource scarcity. In female mate choice in many nonhuman species,
material resources take precedence over phenotypic cues when the environ-
ment is unpredictable and resources are scarce (e.g., Alatalo et al., 1986;
Lifjeld & Slagsvold, 1988). However, feminine characteristics in men are asso-
ciated with their willingness to share resources with others (Price et al., 2011),
and feminine men are perceived as emotionally warm and caring (e.g., Perrett
et al., 1998). It is then, perhaps, unsurprising that recent work on human
mate preferences suggests that women might value cues of men’s willingness
to invest and share resources with others to a greater extent under conditions
where resources are scarce. For example, Anthony Lee and Brendan Zietsch
(2011) recently found that women whose concerns about resource scarcity
were primed by having them complete a questionnaire about their material re-
sources demonstrated stronger preferences for feminine characteristics in
potential mates than did women in control conditions. This finding comple-
ments previous work suggesting that priming women’s concerns about re-
source scarcity by having them imagine themselves in harsh environments in
which resources were scarce tended to increase their preferences for feminine
characteristics in the faces of hypothetical long-term mates (Little et al.,
2007). These findings suggest that women increase their preferences for men
displaying cues of prosociality in order to compensate for scarcity of material
resources. However, other work suggests that similar compensatory effects
occur when participants are primed with scenarios in which emotional re-
sources (e.g., emotional support) are scarce; Watkins and colleagues (in press)
recently demonstrated that priming participants’ concerns about lack of emo-
tional support tended to increase their preferences for feminine men and
women. Collectively, these findings underline the potential importance of re-
source scarcity in human mate preferences.
326 ANIMAL BEHAVIOR

The work described above suggests that resource scarcity increases women’s
preferences for cues of prosocial behavior in men. These findings differ some-
what from those observed in some nonhuman species, potentially due to the
greater role that biparental care typically plays in human mating systems. Are
there environmental factors that influence human behavior in ways that are
more similar to the effects observed in nonhuman species? Perhaps one of
the clearest examples of environmental factors having comparable effects on
mating-related behaviors in nonhuman species and mating-related perceptions
in women comes from research that has explored the effects of cues to the sex
ratio of the local population.
The potential costs of competing for mates in good physical condition are
reduced considerably when potential mates are plentiful and competitors for
mates are relatively scarce (e.g., Pedersen, 1991). This may allow individuals
to increase their preferences for cues that are associated with high quality in
potential mates and require them to engage in less within-sex competition in
order to secure mates (e.g., Pedersen, 1991). Altered mate preferences as a
result of experimentally manipulated operational sex ratios have been
reported for several different species. For example, female guppies show
stronger preferences for attractive male color characteristics when the sex ratio
is biased towards males than when it is biased towards females (Jirotkul,
1999). In field crickets (Gryllus pennsylvanicus), female mate preferences also
show greater selectivity when the sex ratio is biased towards males (Souroukis
& Murray, 1994). In many species, the intensity of within-sex competition
is also affected by altering the sex ratio. For example, biasing the sex ratio
towards own-sex individuals causes more intense within-sex competition in
Japanese medaka (Oryzias latipes) (Clark & Grant, 2010), guppies (Jirotkul,
1999), red-spotted newts (Notophthalmus viridescens) (Verrell, 1983), and
amphipods (Dick & Elwood, 1996). Importantly, these effects complement
findings from studies that investigated relationships between naturally occur-
ring variation in sex ratios and indices of either mate preferences or within-
sex competition for mates.
While the findings described above indicate that sex ratio influences mate
preferences and within-sex competition in many nonhuman species, correla-
tional studies of naturally occurring variation in human sex ratios suggest that
they may also be important for human behavior. For example, Thomas Pollet
and Daniel Nettle (2008) found that women in regions of the United States
with male-biased sex ratios demonstrated stronger preferences for high-
socioeconomic-status (i.e., attractive; Hume & Montgomerie, 2001) men
than did women in regions of the United States with female-biased sex ratios.
More recently, Watkins and colleagues (2012) investigated whether women’s
BEAUTY IN THE EYE OF THE BEHOLDER 327

responses to facial symmetry, a putative cue of mate quality in both men and
women (see, e.g., Gangestad & Thornhill, 2003), could be altered by exposing
them to male-biased or female-biased slideshows. Viewing the slideshows
tended to increase preferences for symmetry in the sex that was depicted as
being in the majority and tended to decrease preferences for symmetry in the
sex that was depicted as being in the minority. These findings suggest that
increasing the apparent proportion of a given sex in the local population
increased the salience of facial cues of quality in that sex, which may support
adaptive appraisals of both potential mates’ and competitors’ quality. A second
experiment suggested that this effect of sex ratio was independent of the degree
of variation in the attractiveness of individuals depicted in the slideshows.
These findings suggest that symmetry preferences in humans are influenced
by cues to the sex ratio of the local population in ways that complement the
facultative responses that have been observed in many other species.
Together, the findings discussed in this section suggest that environmental
factors, such as pathogens, resource scarcity, and sex ratio, can influence
humans’ perceptions of potential mates (and, in some cases, competitors for
mates). Moreover, these effects are often very similar to those seen in other
species and appear to be consistent with formulations of sexual selection
theory that predict effects of environmental factors on mate choice. Encourag-
ingly, and following examples set by researchers studying mating-related
behaviors in nonhuman species, the use of experimental, rather than correla-
tional, methods has often allowed researchers to make strong claims about
the causal role environmental factors can play in human mate preferences.
Indeed, we suggest here that such insights are critical if we are to develop a
more complete understanding of human mating behavior.

MHC AND MATE CHOICE

Theory
In addition to the kinds of effect described above, individual differences in
mate-choice decision making can arise as a result of genetic variation among
potential mates. Particular research effort has focused on genes in the major
histocompatibility complex (MHC). The MHC is found in all vertebrates
and contains genes that encode cell-surface glycoproteins responsible for bind-
ing antigens from pathogens or parasites and presenting them to lymphocytes.
They are thus of central importance in the body’s immune response. Over
evolutionary time, an “arms race” between pathogens and the body’s immune
defense has led to hypervariability among the MHC genes. The most variable
gene in the human genome, for example, is HLA-B, which has well over 2,000
328 ANIMAL BEHAVIOR

known variants, or alleles, while three other human MHC genes (HLA-A,
HLA-C, HLA-DRB1) each have over 1,000 known alleles (these numbers
continue to rise steadily due to advances in rapid sequencing techniques).
MHC variability is also thought to be generated and maintained because
MHC genes are codominantly expressed such that, for any given gene, alleles
inherited from both the mother and the father are functional, leading to
heterozygote advantage (individuals with two different alleles [heterozygotes]
have higher fitness than those with two copies of the same allele [homozy-
gotes]) in the face of infectious assault (Penn et al., 2002; McClelland et al.,
2003).
In recent years, a huge body of evidence has been amassed to demonstrate
that MHC polymorphism is also at least partly generated or maintained
through sexual selection (for reviews, see, e.g., Ziegler et al., 2005; Milinski,
2006). Individuals who choose mates that share no or few MHC alleles will
produce offspring who are more likely to be MHC heterozygous, and because
such offspring will themselves be more resistant to infection, selection could
favor the evolution of MHC-based mate preferences. Furthermore, individuals
with MHC alleles that are relatively rare in the population would then be
more often selected as mates, promoting MHC diversification through
frequency-dependent selection.
Although the benefit of producing MHC-heterozygous offspring clearly has
potential to generate MHC-disassortative mate preferences (preference for
mates with MHC alleles different from one’s own), MHC could also be
involved in shaping mate preferences in another way, by enabling discrimina-
tion of overall genomic dissimilarity in potential mates (Apanius et al., 1997;
Kempenaers, 2007). Here, the extremely polymorphic nature of the MHC
provides a fine-grained cue of pairwise genetic dissimilarity between a choos-
ing individual and a range of potential mates, but the benefits of mating disas-
sortatively are reaped through heterozygosity at genes outside the MHC. In
other words, the MHC could provide a mechanism for avoiding inbreeding
in a general sense.
Regardless of whether the outcome is MHC heterozygosity in offspring or
genome-wide inbreeding avoidance, choosing a mate with few MHC alleles
in common requires some form of phenotypic manifestation of MHC geno-
type in order for discrimination of dissimilar mates to occur. Across a wide
range of vertebrate taxa, this is known to occur through odor. Exactly how
MHC genotype comes to shape an individual’s odor profile is as yet largely
unknown, but it probably works through one of two closely related mecha-
nisms, each of which involves bacterial action (for a full discussion of possible
mechanisms, see Penn & Potts, 1998b). In the first, the commensal
BEAUTY IN THE EYE OF THE BEHOLDER 329

microflora breaks down peptides, which include the individual’s MHC pepti-
des, producing an MHC-characteristic mixture of volatile odor compounds.
Another possibility is that the population structure of the commensal micro-
flora themselves is shaped by the individual’s MHC genotype and thus produ-
ces a characteristic odor. Whatever the mechanism, the mate preferences
expressed in adulthood appear to be learned during odor exposure in the peri-
natal environment, as cross-fostering alters subsequent adult choice so that
preferred mates tend to be dissimilar from the foster family rather than the
chooser’s own MHC genotype (Penn & Potts, 1998a).

Evidence from Nonhuman Animals

MHC-disassortative mate choice—MHC-dependent mate preferences were
first discovered in laboratory mice (Mus musculus). In MHC-congenic strains,
which are genetically identical except for differences at one or more MHC
genes, females tend to prefer males from different strains over males of their
own strain (Yamazaki et al., 1976), and this discrimination is mediated by uri-
nary odor (Yamaguchi et al., 1981; Yamazaki, Beauchamp, Bard, et al., 1990).
Similar results are found in mice carrying single MHC gene mutations
(Yamazaki, Beauchamp, Bard, et al., 1990; Yamazaki et al., 1991; Bard et
al., 2000) and other rodents, including rats (Rattus norvegicus) (Brown et al.,
1987; Singh et al., 1987) and voles (Myodes glareolus) (Radwan et al., 2008).
At least in mice, where most work has been done, perception of genetic differ-
ences in urine is mediated by varying proportions of volatile carboxylic acids
(Singer et al., 1997; Kwak et al., 2009, 2010), which can explain how dis-
crimination can occur at a distance, although involatile peptides, which
require much closer approach, may also play a role (Leinders-Zufall et al.,
2004). Odors of MHC-disparate mouse strains trigger distinctive patterns of
neuronal activation in the main olfactory bulb (Schaefer et al., 2001). Urinary
scent marks from different MHC-congenic strains support distinctive commun-
ities of commensal microflora (Lanyon et al., 2007), and at least some research
using axenic mice (obtained by caesarean section under germ-free conditions)
indicates that MHC-odor discrimination is abolished in the absence of bacteria
(but see Yamazaki, Beauchamp, Imai, et al., 1990; Schellinck et al., 1995).
The finding that there was a surfeit of MHC-homozygous offspring in
seminatural house mouse populations compared with what would be expected
if mating occurred at random, and that this was not due to differences in pre-
weaning mortality (Potts et al., 1991), provided the first significant evidence
that MHC-disassortative mating in mice might have real and measurable
effects outside the controlled conditions of the lab. Since this finding,
MHC-correlated mate preferences have been demonstrated in many and
330 ANIMAL BEHAVIOR

diverse species, across reptiles (Olsson et al., 2003; Miller et al., 2009),
amphibians (Bos et al., 2009), fish (Landry et al., 2001; Reusch et al., 2001;
Aeschlimann et al., 2003; Skarstein et al., 2005; Forsberg et al., 2007), birds
(Freeman-Gallant et al., 2003, 2006), and other mammals including primates
(Setchell et al., 2010; Setchell & Huchard, 2010).
Complexity in MHC-correlated choice—It should also be noted that some
studies report no effects of MHC on mating patterns (Wedekind et al.,
2004; Westerdahl, 2004; Huchard et al., 2010), and many report effects that
are not as straightforward as some of the findings discussed so far. In at least
some cases, absence of evidence for MHC-correlated choice might be
explained by females being prevented from expressing preference due to spe-
cific aspects of the mating system such as levels of male intrasexual competi-
tion over mates (Paterson & Pemberton, 1997; Garner et al., 2010). There
are several other possible reasons, the most important of which we discuss in
this section.
First, assortment at the MHC is only one among a wide range of selective
forces acting on mate choice—see, for example, the other sections of this chap-
ter. Even if we consider only indirect genetic benefits, a female’s preference for
a genetically compatible or complementary (e.g., MHC-dissimilar) male is
weighed against evidence of his “good genes” expressed through other pheno-
typic characters (Mays & Hill, 2004; Neff & Pitcher, 2005). These two pro-
cesses have the potential to deliver nonadditive and additive genetic benefits
in offspring fitness, respectively; in other words, the benefits gained from mat-
ing with a specific male vary across different females in the case of MHC but
are relatively constant for different females mating with the same male carrying
good genes. Again, the first investigation of how females balance these two
critical influences on offspring fitness comes from mice. In addition to
MHC-correlated preferences, females gain fitness benefits through mating
with dominant males, assessing their relative social status using another set of
volatile compounds also present in urinary scent marks (reviewed in Gosling
& Roberts, 2001). When these traits were experimentally covaried, MHC dis-
similarity predicted female preferences when variation among available males
was low for the good gene trait or high for MHC dissimilarity (Roberts &
Gosling, 2003). Evidence from birds (Oh & Badyaev, 2006) suggests that
these conditions of relative variation in availability of “good” and “compatible”
genes among males in a population can occur across seasons and lead to the
same patterns of choice as found by Craig Roberts and Morris Gosling in
the lab.
Second, MHC-dependent mate choice is not necessarily explained simply
by a search for overall dissimilarity. For example, choice may be underpinned
BEAUTY IN THE EYE OF THE BEHOLDER 331

under some conditions by dissimilarity for specific MHC alleles and not
others (Ekblom et al., 2004). Alternatively, growing evidence suggests that
an intermediate, rather than extreme, level of dissimilarity might be preferable
in mates (Penn & Potts, 1999; Aeschlimann et al., 2003; Forsberg et al.,
2007; Roberts, 2009) because intermediate levels of MHC diversity in off-
spring lead to highest immunocompetence (Nowak et al., 1992; Wegner,
Kalbe, et al., 2003; Wegner, Reusch, et al., 2003).
Third, MHC-disassortative patterns may not be observed because females
may simply lack viable alternatives (c.f. Blomqvist et al., 2002). An illuminat-
ing illustration of how patterns of MHC-disassortative mating may be not
immediately apparent, even where they could be important, occurs in socially
monogamous species. In birds, for example, female choice is largely based on a
male’s territory quality, which dictates likely quality of offspring provisioning
(Alatalo et al., 1986). A few females might also be able to complement this
choice using cues of genetic similarity (Mulard et al., 2009), but many will
not. However, if they also engage in extra-pair copulations (EPCs), they have
a “second chance” to obtain genetic diversity in their offspring (Petrie &
Kempenaers, 1998; Jennions & Petrie, 2000; Kempenaers, 2007). In line
with this idea, studies have shown that females who are paired with relatively
genetically similar social partners (i.e., the male who helps feed the chicks)
are more likely to seek EPCs (Blomqvist et al., 2002; Eimes et al., 2005;
Tarvin et al., 2005; Fossoy et al., 2008) and by so doing gain increased off-
spring heterozygosity with enhanced cellular immunity (Fossoy et al., 2008)
and higher hatching success, growth rates, and survival (Mulard et al.,
2009). Furthermore, females who engage in EPCs with geographically more
distant males appear to gain increased heterozygosity, suggesting they actively
seek out genetically dissimilar males (Foerster et al., 2003; Stapleton et al.,
2007). Many of these studies do not specifically examine genetic dissimilarity
at the MHC between social and extra-pair partners, though there is every rea-
son to suspect that the reported effects could be driven by MHC. Indeed,
when MHC is specifically examined, very similar patterns are described, for
example in savannah sparrows (Passerculus sandwichensis) (Freeman-Gallant
et al., 2003). Similar patterns also occur in some monogamous mammals
(Cohas et al., 2008; Schwensow et al., 2008).
Finally, preferences discussed so far reflect potential indirect fitness benefits
gained through increased viability of progeny. However, the MHC could also
have effects on mate choice independently of genetic dissimilarity between
parents. Although heterozygosity is not strictly heritable, it could itself be
attractive in a potential mate (especially in males) and could provide direct
benefits to individuals mating with heterozygotes. This is known as the
332 ANIMAL BEHAVIOR

“good-genes-as-heterozygosity hypothesis” (Brown, 1997, 1999). As a conse-

quence of immune advantage, MHC-heterozygous males may be able to com-
pete more effectively over resources or territories that females need, to provide
better paternal care, or to produce more attractive secondary sexual traits,
which can be expressed in several sensory modalities (Foerster et al., 2003;
Seddon et al., 2004; Ilmonen et al., 2009). Preference for heterozygous males
has been found in some mammals (Schwensow et al., 2008; Ilmonen et al.,
2009)

Evidence from Humans

We have shown in previous sections that individual differences in human
mate preferences often follow similar patterns as those seen in animals, and
the same is true of MHC-correlated preferences. In the first experiment of
its kind, Claus Wedekind and colleagues (1995) tested women’s assessment
of odors of MHC-similar and MHC-dissimilar men. To capture the men’s
odor in such a way that preserved their privacy and enabled a carefully con-
trolled design, Wedekind and colleagues collected t-shirts worn by the men
over two nights. Men undertook to refrain from using artificial fragrances
and to avoid strong-smelling foods during this time, and they slept alone.
Individual women then rated shirts of six men (preselected so that three were
MHC-similar and three were MHC-dissimilar to the woman). In order to
avoid possible variation in olfactory sensitivity over the menstrual cycle,
women were tested during the fertile phase. Wedekind and colleagues found
that these women gave higher ratings of odor pleasantness and sexiness to
the odors of MHC-dissimilar men, just as we would expect based on the
mouse research. Furthermore, they were more likely to report that MHC-
dissimilar odors reminded them of their current (or a previous) partner, indi-
cating that these laboratory-measured preferences might reflect real-life
choices.
There have been several further studies of MHC-correlated odor prefer-
ences in the intervening years. These, and those investigating other ways of
assessing MHC-based human mate choice such as face preferences or actual
spousal MHC gene sharing, have been reviewed by Jan Havlicek and Roberts
(2009). In light of the various factors that muddy the water when attempting
to detect evidence of MHC-related choice in animals (discussed above), in
addition to the other influences on human mate choice (including those
described elsewhere in this chapter), it should be no surprise that the evidence
is somewhat mixed, but Havlicek and Roberts (2009) conclude that there
appears to be “a real phenomenon” (p. 509) at work. For example, three of
six studies report MHC-disassortative preferences in women’s assessment of
BEAUTY IN THE EYE OF THE BEHOLDER 333

men’s odors (another reports a preference for an intermediate level of dissimi-

larity; two found no effect). Furthermore, two of three tests in which men
assessed women’s odors found disassortative preferences (the other found no
effect). Women also appear to find the odor of heterozygous men more attrac-
tive, perhaps especially in the nonfertile phase, which would be consistent
with a direct benefit of pairing with a heterozygous man (Thornhill et al.,
2003).
Beyond odor preferences, a small number of studies have investigated
whether the face might carry MHC-correlated cues. In a visual analogue of
Wedekind and colleagues’ (1995) odor experiment, Roberts, Little, Gosling,
Jones, and colleagues (2005) found that women who rated the faces of three
MHC-similar and three MHC-dissimilar men showed a preference for
MHC similarity, especially when thinking about long-term partners.
Although this may be a surprising result in view of the rest of the MHC liter-
ature, it is consistent with generally assortative facial preferences (see “Kinship
and Mate Choice,” above) and could provide a mechanism, in conjunction
with disassortative odor preferences, to achieving intermediate or optimal lev-
els of dissimilarity in a partner (Roberts, Little, Gosling, Jones, et al., 2005).
Faces of MHC-heterozygous men are also perceived as more attractive, and
their skin as more healthy, than homozygotes (Roberts, Little, Gosling,
Perrett, et al., 2005; Lie et al., 2008). Several other studies have examined
whether actual couples share fewer MHC alleles than expected if they paired
randomly. Including those reviewed by Havlicek and Roberts (2009) and a
further study by Raphaëlle Chaix and colleagues (2008), 3 of 11 report evi-
dence for couples sharing fewer MHC alleles than expected under random
mating (the rest report no difference from random pairing, except one sug-
gesting an assortative effect, but this case is almost certainly explained by pair-
ing within ethnic groups).
Thus the evidence for MHC-correlated pairing in humans, particularly
outside the lab, is far from overwhelming, but, where effects have been found,
they do usually indicate a disassortative pattern. However, as we have seen in
the animal literature, absence of evidence in individual studies may be at least
partly due to other effects that dilute an underlying preference. Such effects are
just as likely in humans. Two examples can serve to illustrate this point.
The first comes from one of the studies that failed to find evidence for
MHC-disassortative choice based on allele-sharing between spouses (Garver-
Apgar et al., 2006). Despite this result, this study is interesting because it also
included an analysis of relationship satisfaction and sexual behavior within the
48 couples sampled. In those couples who shared a relatively high number of
MHC alleles, women reported lower sexual satisfaction with their partner,
334 ANIMAL BEHAVIOR

higher desire for extra-pair partners, and higher incidence of having extra-pair
sex compared with women relatively MHC dissimilar to their partner. The
resemblance between these results and those previously described in other
socially monogamous species is palpable.
The second example takes us back to the initial experiment by Wedekind
and colleagues (1995). While women near to ovulation preferred the odors
of MHC-dissimilar men, another group of women who were using oral con-
traception (the “pill”) showed the opposite preference. The pill suppresses
the hormonal changes experienced during ovulation and produces hormonal
levels more like those seen during pregnancy. Thus, a preference for MHC-
similar male odors in women using oral contraceptives may be analogous to
preferences for MHC similarity in mice during pregnancy, a time when
females are not actively mate-searching and may benefit from being near kin
(Manning et al., 1992). Although there might have been confounding differ-
ences among women who used or chose not to use the pill, a subsequent
within-subjects study showed that initiation of pill use altered preferences
towards MHC similarity (Roberts et al., 2008). As pill use is widespread in
women of reproductive age, this disruptive effect on women’s preference could
potentially explain the relatively small proportion of studies finding MHC-
disassortative mating among real couples (Havlicek & Roberts, 2009) and
might increase the risk of choosing an otherwise less preferred partner (Roberts
et al., 2008). Indeed, in a recent twist to this tale, we found that in a large sam-
ple of well-established couples (who have had at least one child with their part-
ner), women who met their partner while using the pill reported lower sexual
satisfaction than those who had not been, and that their sexual dissatisfaction
increased over the course of the relationship. In those couples who had
divorced, women were more likely than expected to have initiated the divorce
if they were using the pill at the start of the relationship (Roberts et al., 2012).
Together with Christine Garver-Apgar and colleagues’ (2006) study, and in
common with patterns of behavior seen in birds and other monogamous
mammals, these results indicate that the MHC appears to influence the qual-
ity and outcome of the partnership, in addition to its effects in determining a
large proportion of individual variability in mate preference.

CONCLUSION
Here we have reviewed four of the most prominent theories that help to
explain individual differences in mate choice, highlighting the similarities of
evidence in both human and nonhuman animals. It is clear that neither carica-
ture of nonhuman animal or human mate choice is correct; each nonhuman
BEAUTY IN THE EYE OF THE BEHOLDER 335

animal species does not have a single, ideal mate preference, and individual
differences in human mate preferences are not mysterious, unquantifiable
results of culture. Indeed, mate preferences in the human animal can be stud-
ied in many of the same ways as mate preferences in nonhuman animals, and
patterns of human mate choice follow many of the same principles as nonhu-
man animal mate choice.

REFERENCES AND SUGGESTED READING

Aeschlimann, P. B., M. A. Haberli, T. B. H. Reusch, T. Boehm, & M. Milinski.
(2003). Female sticklebacks Gasterosteus aculeatus use self-reference to optimize
MHC allele number during mate selection. Behavioral Ecology and Sociobiology,
54, 119–126.
Alatalo, R. V., A. Lundberg, & C. Glynn (1986). Female pied flycatchers choose
territory quality and not male characteristics. Nature, 323, 152–153.
Alberts, S. C. (1999). Paternal kin discrimination in wild baboons. Proceedings of the
Royal Society of London, B, 266, 1501–1506.
Apanius, V., D. J. Penn, P. Slev, L. R. Ruff, & W. K. Potts. (1997). The nature of
selection on the major histocompatibility complex. Critical Reviews in Immu-
nology, 17, 179–224.
Aparicio, J. M., P. J. Cordero, & J. P. Veiga. (2001). A test of the hypothesis of mate
choice based on heterozygosity in the spotless starling. Animal Behaviour, 62,
1001–1006.
Bakker, T. C. M. (1993). Positive genetic correlation between female preference and
preferred male ornament in sticklebacks. Nature, 363, 255–257.
Bakker, T. C. M., R. Künzler, & D. Mazzi. (1999). Condition-related mate-choice in
sticklebacks. Nature, 401, 234.
Bard, J., K. Yamazaki, M. Curran, E. A. Boyse, & G. K. Beauchamp. (2000). Effect
of B2m gene disruption on MHC-determined odortypes. Immunogenetics,
51, 514–518.
Bateson, P. (1980). Optimal outbreeding and the development of sexual preferences
in Japanese Quail. Zietschrift fur Tierpsychologie, 53, 231–244.
Bateson, P. (1983). Optimal outbreeding. In P. Bateson (ed.), Mate Choice (pp. 257–
278). Cambridge, UK: Cambridge University Press.
Bereczkei, T., P. Gyuris, P. Koves, & L. Bernáth. (2002). Homogamy, genetic simi-
larity, and imprinting: Parental influence on mate choice preferences. Personal-
ity and Individual Differences, 33, 677–690.
Bereczkei, T., P. Gyuris, & G. E. Weisfeld. (2004). Sexual imprinting in human mate
choice. Proceedings of the Royal Society of London, B, 271, 1129–1134.
Bereczkei, T., G. Hegedus, & G. Hajnal. (2009). Facialmetric similarities mediate
mate choice: Sexual imprinting on opposite-sex parents. Proceedings of the
Royal Society of London, B, 276, 91–98.
336 ANIMAL BEHAVIOR

Bittles, A. H. (2001). Consanguinity and its relevance to clinical genetics. Clinical

Genetics, 60, 89–98.
Bittles, A. H., J. C. Grant, S. G. Sullivan, & R. Hussain. (2002). Does inbreeding
lead to decreased human fertility? Annals of Human Biology, 29, 111–130.
Blomqvist, D., M. Andersson, C. Küpper, I. C. Cuthill, J. Kis, R. B. Lanctot, B. K. Sandercock,
T. Székely, K. Wallander & B. Kempenaers. (2002). Genetic similarity between mates
and extra-pair parentage in three species of shorebirds. Nature, 419, 613–615.
Bos, D. H., R. N. Williams, D. Gopurenko, Z. Bulut, & J. A. Dewoody. (2009).
Condition-dependent mate choice and a reproductive disadvantage for
MHC-divergent male tiger salamanders. Molecular Ecology, 18, 3307–3315.
Brédart, S. & R. M. French. (1999). Do babies resemble their fathers more than their
mothers? A failure to replicate Christenfeld and Hill (1995). Evolution and
Human Behavior, 20, 129–135.
Bressan, P. & M. F. Dal Martello. (2002). Talis pater, talis filius: Perceived resem-
blance and the belief in genetic relatedness. Psychological Science, 13, 213–218.
Bressan, P. & M. Grassi. (2004). Parental resemblance in one-year-olds and the
Gaussian curve. Evolution and Human Behavior, 25, 133–141.
Brooks, R., I. M. L. Scott, A. A. Maklakov, M. M. Kasumovic, A. P. Clark, & I. S.
Penton-Voak. (2011). National income inequality predicts women’s prefer-
ences for masculinized faces better than health does. Proceedings of the Royal
Society of London, B, 278, 810–812.
Brown, J. L. (1997). A theory of mate choice based on heterozygosity. Behavioral
Ecology, 8, 60–65.
Brown, J. L. (1999). The new heterozygosity theory of mate choice and the MHC.
Genetica, 104, 215–221.
Brown, R. E., P. B. Singh, & B. Roser. (1987). The major histocompatibility com-
plex and the chemosensory recognition of individuality in rats. Physiology and
Behavior, 40, 65–73.
Buckingham, G., L. M. DeBruine, A. C. Little, L. L. M. Welling, C. A. Conway,
B. P. Tiddeman, & B. C. Jones. (2006). Visual adaptation to masculine and
feminine faces influences generalized preferences and perceptions of trustwor-
thiness. Evolution and Human Behavior, 27, 381–389.
Bukacinski, D., M. Bukacinski, & T. Lubjuhn. (2000). Adoption of chicks and the
level of relatedness in common gull, Larus canus, colonies. Animal Behaviour,
59, 289–299.
Burley, N. (1986). Sexual selection for aesthetic traits in species with biparental care.
American Naturalist, 127, 415–445.
Chaix, R., C. Cao, & P. Donnelly. (2008). Is mate choice in humans MHC-
dependent? PLoS Genetics, 4, e1000184. doi:10.1371/journal.pgen.1000184.
Clark, L. & J. W. A. Grant. (2010). Intrasexual competition and courtship in female
and male Japanese medaka, Oryzias latipes: Effects of operational sex ratio and
density. Animal Behaviour, 80, 707–712.
Cohas, A., N. G. Yoccoz, C. Bonenfant, B. Goossens, C. Genton, M. Galan, B. Kem-
penaers, & D. Allaine. (2008). The genetic similarity between pair members
BEAUTY IN THE EYE OF THE BEHOLDER 337

influences the frequency of extrapair paternity in alpine marmots. Animal

Behaviour, 76, 87–95.
Daly, M. & M. I. Wilson. (1982). Whom are newborn babies said to resemble? Ethol-
ogy and Sociobiology, 3, 69–78.
DeBruine, L. M. (2002). Facial resemblance enhances trust. Proceedings of the Royal
Society of London, B, 269, 1307–1312.
DeBruine, L. M. (2004). Resemblance to self increases the appeal of child faces to
both men and women. Evolution and Human Behavior, 25, 142–154.
DeBruine, L. M. (2005). Trustworthy but not lust-worthy: Context-specific effects of
facial resemblance. Proceedings of the Royal Society of London, B, 272, 919–922.
DeBruine, L. M., B. C. Jones, J. R. Crawford, L. L. M. Welling, & A. C. Little.
(2010). The health of a nation predicts their mate preferences: Cross-cultural
variation in women’s preferences for masculinized male faces. Proceedings of
the Royal Society of London, B, 277, 2405–2410.
DeBruine, L. M., B. C. Jones, A. C. Little, J. R. Crawford, & L. L. M. Welling.
(2011). Further evidence for regional variation in women’s masculinity prefer-
ences. Proceedings of the Royal Society of London, B, 278, 813–814.
DeBruine, L. M., B. C. Jones, A. C. Little, & D. I. Perrett. (2008). Social perception
of facial resemblance in humans. Archives of Sexual Behavior, 37, 64–77.
DeBruine, L. M., B. C. Jones, J. M. Tybur, D. Lieberman, & V. Griskevicius.
(2010). Women’s preferences for masculinity in male faces are predicted by
pathogen disgust, but not by moral or sexual disgust. Evolution and Human
Behavior, 31, 69–74.
DeBruine, L. M., B. C. Jones, C. D. Watkins, S. C. Roberts, A. C. Little, F. G.
Smith, & M. C. Quist. (2011). Opposite-sex siblings decrease attraction, but
not prosocial attributions, to self-resembling opposite-sex faces. Proceedings of
the National Academy of Sciences, USA, 108, 11710–11714.
DeBruine, L. M., A. C. Little, & B. C. Jones. (2012). Extending parasite-stress theory
to variation in human mate preferences (commentary on Fincher &
Thornhill). Behavioral and Brain Sciences, 35, 26–27.
Delcourt, M. & H. D. Rundle. (2011). Condition dependence of a multicomponent
sexual display trait in Drosophila serrata. American Naturalist, 177, 812–823.
Deutsch, J. C. & P. Weeks. (1992). Uganda kob prefer high-visibility leks and terri-
tories. Behavioural Ecology, 3, 223–233.
Dick, J. T. A. & R. W. Elwood. (1996). Effects of natural variation in sex ratio and
habitat structure on mate-guarding decisions in amphipods (Crustacea). Behav-
iour, 133, 985–996.
Eimes, J. A., P. G. Parker, J. L. Brown, & E. R. Brown. (2005). Extrapair fertilization
and genetic similarity of social mates in the Mexican jay. Behavioral Ecology,
16, 456–460.
Ekblom, R., S. Saether, M. Grahn, P. Fiske, J. Kalas, & J. Hoglund. (2004). Major
histocompatibility complex variation and mate choice in a lekking bird, the
great snipe (Gallinago media). Molecular Ecology, 13, 3821–3828.
338 ANIMAL BEHAVIOR

Euler, H. A. & B. Weitzel. (1999). Grandparental caregiving and intergenerational

relations reflect reproductive strategies. In J. M. G. van der Dennen, D. Smil-
lie, & D. R. Wilson (eds.), The Darwinian Heritage and Sociobiology: Human
Evolution, Behavior, and Intelligence (pp. 243–252). Westport, CT: Praeger
Publishers.
Feinberg, D. R., B. C. Jones, A. C. Little, D. M. Burt, & D. I. Perrett. (2005).
Manipulations of fundamental and formant frequencies influence the attrac-
tiveness of human male voices. Animal Behaviour, 69, 561–568.
Fessler, D. M. T. & C. D. Navarrete. (2004). Third-party attitudes toward sibling
incest: Evidence of Westermarck’s hypotheses, Evolution of Human Behaviour,
25, 277–294.
Fincher, C. L. & R. Thornhill. (2012). Parasite-stress promotes in-group assortative
sociality: The cases of strong family ties and heightened religiosity. Behavioral
and Brain Sciences, 35, 1–19.
Fisher, R. A. (1930). The Genetical Theory of Natural Selection. Oxford, UK: Claren-
don Press.
Foerster, K., K. Delhey, A. Johnsen, J. T. Lifjeld, & B. Kempenaers. (2003). Females
increase offspring heterozygosity and fitness through extra-pair matings.
Nature, 425, 714–717.
Forsberg, L. A., J. Dannewitz, E. Petersson, & M. Grahn. (2007). Influence of
genetic dissimilarity in the reproductive success and mate choice of brown
trout: Females fishing for optimal MHC dissimilarity. Journal of Evolutionary
Biology, 20, 1859–1869.
Fossoy, F., A. Johnsen, & J. T. Lifjeld. (2008). Multiple genetic benefits of female
promiscuity in a socially monogamous passerine. Evolution, 62, 145–156.
Freeman-Gallant, C. R., M. Meguerdichian, N. T. Wheelwright, & S. V. Sollecito.
(2003). Social pairing and female mating fidelity predicted by restriction frag-
ment length polymorphism similarity at the major histocompatibility complex
in a songbird. Molecular Ecology, 12, 3077–3083.
Freeman-Gallant, C. R., N. T. Wheelwright, K. E. Meiklejohn, & S. V. Sollecito.
(2006). Genetic similarity, extrapair paternity, and offspring quality in Savan-
nah sparrows (Passerculus sandwichensis). Behavioral Ecology, 17, 952–958.
Fujita, K. (1993). Development of visual preference for closely related species by
infant and juvenile macaques with restricted social experience. Primates, 34,
141–150.
Gangestad, S. W. & D. M. Buss. (1993). Pathogen prevalence and human mate pref-
erences. Ethology and Sociobiology, 14, 89–96.
Gangestad, S. W., L. Merriman, & M. Emery Thompson. (2010). Men’s oxidative
stress, fluctuating asymmetry and physical attractiveness. Animal Behaviour,
80, 1005–1013.
Gangestad, S. W. & J. A. Simpson. (2000). The evolution of human mating: Trade-
offs and strategic pluralism. Behavioural and Brain Sciences, 23, 573–644.
Gangestad, S. W. & R. Thornhill. (2003). Facial masculinity and fluctuating asym-
metry. Evolution and Human Behavior, 24, 231–241.
BEAUTY IN THE EYE OF THE BEHOLDER 339

Garner, S. R., R. N. Bortoluzzi, D. D. Heath, & B. D. Neff. (2010). Sexual conflict

inhibits female mate choice for major histocompatibility complex dissimilarity
in Chinook salmon. Proceedings of the Royal Society of London, B, 277, 885–894.
Garver-Apgar, C. E., S. W. Gangestad, R. Thornhill, R. D. Miller, & J. J. Olp.
(2006). Major histocompatibility complex alleles, sexual responsivity, and
unfaithfulness in romantic couples. Psychological Science, 17, 830–835.
Gaulin, S. J. C., D. H. McBurney, & S. L. Brakeman-Wartell. (1997). Matrilateral
biases in the investment of aunts and uncles: A consequence and measure of
paternity uncertainty. Human Nature, 8, 139–151.
Gaulin, S. J. C. & A. Schlegel. (1980). Paternal confidence and paternal investment:
A cross cultural test of a sociobiological hypothesis. Ethology and Sociobiology,
1, 301–309.
Gosling, L. M. & S. C. Roberts. (2001). Scent-marking by male mammals: Cheat-
proof signals to competitors and mates. Advances in the Study of Behavior, 30,
169–217.
Greenberg, L. (1979). Genetic component of bee odor in kin recognition. Science,
206, 1095–1097.
Griffin, A. S. & S. A. West. (2003). Kin discrimination and the benefit of helping in
cooperatively breeding vertebrates. Science, 302, 634–636.
Hamilton, W. D. (1964). The genetical evolution of social behaviour, I. Journal of
Theoretical Biology, 7, 1–16.
Harris, R. N., T. J. Vess, J. I. Hammond, & C. J. Lindermuth. (2003). Context-
dependent kin discrimination in larval four-toed salamanders Hemidactylium
scutatum (Caudata: Plethodontidae). Herpetologica, 59, 164–177.
Hauber, M. E. & P. W. Sherman. (1998). Nepotism and marmoset alarm calling.
Animal Behaviour, 53, 1049–1052.
Havlicek, J. & S. C. Roberts. (2009). MHC-correlated mate choice in humans:
A review. Psychoneuroendocrinology, 34, 497–512.
Hedrick, A. (2005). Environmental condition-dependent effects on a heritable, pre-
ferred male trait. Animal Behaviour, 70, 1121–1124.
Helgason, A., S. Pálsson, D. F. Guõbjartsson, P. Kristjánsson, & K. Stefánsson.
(2008). An association between the kinship and fertility of human couples. Sci-
ence, 319, 813–816.
Heth, G., J. Todrank, N. Busquet, & C. Baudoin. (2003). Genetic relatedness assess-
ment through individual odour similarities in mice. Biological Journal of the
Linnean Society, 78, 595–603.
Holveck, M.-J. & K. Riebel. (2010). Low-quality females prefer low-quality males
when choosing a mate. Proceedings of the Royal Society of London, B, 277,
153–160.
Houde, A. E. & A. J. Torio. (1992). Effect of parasitic infection on male colour pat-
tern and female choice in guppies. Behavioural Ecology, 3, 346–351.
Huchard E., L. A. Knapp, J. L. Wang, M. Raymond, & G. C. Cowlishaw. (2010).
MHC, mate choice and heterozygote advantage in a wild social primate.
Molecular Ecology, 19, 2545–2561.
340 ANIMAL BEHAVIOR

Hume, D. K. & R. Montgomerie. (2001). Facial attractiveness signals different

aspects of “quality” in women and men. Evolution and Human Behavior, 22,
93–112.
Ilmonen, P., G. Stundner, M. Thoss, & D. J. Penn. (2009). Females prefer the scent
of outbred males: Good-genes-as-heterozygosity? BMC Evolutionary Biology, 9,
104.
Jedlicka, D. (1980). A test of the psychoanalytic theory of mate selection. Journal of
Social Psychology, 112, 295–299.
Jennions, M. D., A. P. Møller, & M. Petrie. (2001). Sexually selected traits and adult
survival: A meta-analysis. Quarterly Review of Biology, 76, 3–36.
Jennions, M. D. & M. Petrie. (1997). Variation in mate choice and mating prefer-
ences: A review of causes and consequences. Biological Review of the Cambridge
Philosophical Society, 72, 283–327.
Jennions, M. D. & M. Petrie. (2000). Why do females mate multiply? A review of the
genetic benefits. Biological Reviews, 75, 21–64.
Jirotkul, M. (1999). Operational sex ratio influences female preference and male-male
competition in guppies. Animal Behaviour, 58, 287–294.
Jones, B. C., L. M. DeBruine, A. C. Little, C. A. Conway, L. L. M. Welling, & F.
Smith. (2007). Sensation seeking and men’s face preferences. Evolution and
Human Behavior, 28, 439–446.
Jones, B. C., A. C. Little, L. Boothroyd, D. R. Feinberg, R. E. Cornwell, L. M.
DeBruine, S. C. Roberts, I. S. Penton-Voak, M. J. L. Smith, F. R. Moore,
H. P. Davis, & D. I. Perrett. (2005). Women’s physical and psychological
condition independently predict their preference for apparent health in faces.
Evolution and Human Behavior, 26, 451–457.
Kaminski, G., A. Dridi, C. Graff, & E. Gentaz. (2009). Human ability to detect kin-
ship in strangers’ faces: Effects of the degree of relatedness, Proceedings of the
Royal Society of London, B, 276, 3193–3200.
Kaminski, G., F. Ravary, C. Graff, & E. Gentaz. (2010). Firstborns’ disadvantage in
kinship detection. Psychological Science, 21, 1746–1750.
Kempenaers, B. (2007). Mate choice and genetic quality: A review of the heterozygos-
ity theory. Advances in the Study of Behavior, 37, 189–278.
Kendrick, K. M., M. R. Hinton, & K. Atkins. (1998). Mothers determine sexual
preferences. Nature, 395, 229–230.
Krupp, D. B., L. M. DeBruine, & P. Barclay. (2008). A cue of kinship promotes co-
operation for the public good. Evolution and Human Behavior, 29, 49–55.
Kwak, J., M. C. Opiekun, K. Matsumura, G. Preti, K. Yamazaki, & G. K. Beau-
champ. (2009). Major histocompatibility complex-regulated odortypes:
ePptide-free urinary volatile signals. Physiology and Behavior, 96, 184–188.
Kwak, J., A. Willse, G. Preti, K. Yamazaki, & G. K. Beauchamp. (2010). In search of
the chemical basis for MHC odourtypes. Proceedings of the Royal Society of Lon-
don, B, 277, 2417–2425.
Landry, C., D. Garant, P. Duchesne, & L. Bernatchez. (2001). “Good genes as heter-
ozygosity”: The major histocompatibility complex and mate choice in Atlantic
BEAUTY IN THE EYE OF THE BEHOLDER 341

salmon (Salmo salar). Proceedings of the Royal Society of London, B, 268, 1279–
1285.
Lanyon, C., S. Rushton, A. O’Donnell, M. Goodfellow, A. Ward, M. Petrie, S. Jen-
sen, L. M. Gosling, & D. J. Penn. (2007). Murine scent mark microbial com-
munities are genetically determined. FEMS Microbiology Ecology, 59, 576–583.
Lee, A. J. & B. P. Zietsch. (2011). Experimental evidence that women’s mate prefer-
ences are directly influenced by cues of pathogen prevalence and resource scar-
city. Biology Letters, 7, 892–895.
Leinders-Zufall, T., P. Brennan, P. Widmayer, P. Chandramani, A. Maul-Pavicic, M.
Jager, X. H. Li, H. Breer, F. Zufall, & T. Boehm. (2004). MHC class I peptides
as chemosensory signals in the vomeronasal organ. Science, 306, 1033–1037.
Lie, H. C., G. Rhodes, & L. W. Simmons. (2008). Genetic diversity revealed in
human faces. Evolution, 62, 2473–2486.
Lieberman, D. (2009). Rethinking the Taiwanese minor marriage data: Evidence the
mind uses multiple kinship cues to regulate inbreeding avoidance, Evolution
and Human Behavior, 30, 153–160.
Lieberman, D., J. Tooby, & L. Cosmides. (2003). Does morality have a biological
basis? An empirical test of the factors governing moral sentiments relating to
incest. Proceedings of the Royal Society of London, B, 270, 819–826.
Lieberman, D., J. Tooby, & L. Cosmides. (2007). The architecture of human kin
detection. Nature, 225, 727–731.
Lifjeld, J. T. & T. Slagsvold. (1988). Female pied flycatchers Ficedula hypoleuca
choose male characteristics in homogeneous habitats. Behavioral Ecology and
Sociobiology, 22, 27–36.
Little, A. C., D. M. Burt, I. S. Penton-Voak, & D. I. Perrett. (2001). Self-perceived
attractiveness influences human female preferences for sexual dimorphism
and symmetry in male faces. Proceedings of the Royal Society of London, B,
268, 39–44.
Little, A. C. D. L., Cohen, B. C. Jones, & J. Belsky. (2007). Human preferences for
facial masculinity change with relationship type and environmental harshness.
Behavioral Ecology and Sociobiology, 61, 967–973.
Little, A. C., L. M. DeBruine, & B. C. Jones. (2011). Exposure to visual cues of
pathogen contagion changes preferences for masculinity and symmetry in
opposite-sex faces. Proceedings of the Royal Society of London, B, 278, 2032–
2039.
Little, A. C., B. C. Jones, & L. M. DeBruine. (2011). Facial attractiveness: Evolution-
ary based research. Philosophical Transactions of the Royal Society of London, B,
366, 1638–1659.
Little, A. C. & H. Mannion. (2006). Viewing attractive or unattractive same-sex indi-
viduals changes self-rated attractiveness and face preferences in women. Animal
Behaviour, 72, 981–987.
Little, A. C., I. S. Penton-Voak, D. M. Burt, & D. I. Perrett. (2003). Investigating an
imprinting-like phenomenon in humans: Partners and opposite-sex parents
have similar hair and eye colour. Evolution and Human Behavior, 24, 43–51.
342 ANIMAL BEHAVIOR

López, S. (1999). Parasitized female guppies do not prefer showy males. Animal
Behaviour, 57, 1129–1134.
Manning, C. J., E. K. Wakeland, & W. K. Potts. (1992). Communal nesting patterns
in mice implicate MHC genes in kin recognition. Nature, 360, 581–583.
Mateo, J. M. (2004). Recognition systems and biological organization: The percep-
tion component of social recognition. Annals Zoologica Fennici, 41, 729–745.
Mays, H. L. J. & G. E. Hill. (2004). Choosing mates: Good genes versus genes that
are a good fit. Trends in Ecology and Evolution, 19, 555–559.
Mazzi, D., R. Kunzler, & T. C. M. Bakker. (2003). Female preference for symmetry
in computer-animated three-spined sticklebacks, Gasterosteus aculeatus. Behav-
ioral Ecology and Sociobiology, 54, 156–161.
McClelland, E. E., D. J. Penn, & W. K. Potts. (2003). Major histocompatibility
complex heterozygote superiority during coinfection. Infection and Immunity,
71, 2079–2086.
Milinski, M. (2006). The major histocompatibility complex, sexual selection, and
mate choice. Annual Review of Ecology Evolution and Systematics, 37, 159–186.
Milinski, M. & T. C. M. Bakker. (1990). Female sticklebacks use male coloration in
sticklebacks and therefore avoid parasitised males. Nature, 344, 330–333.
Milinski, M. & T. C. M. Bakker. (1992). Costs influence sequential mate choice in
sticklebacks, Gasterosteus aculeatus. Proceedings of the Royal Society of London,
B, 250, 229–233.
Miller, H. C., J. A. Moore, N. J. Nelson, & C. H. Daugherty. (2009). Influence of
major histocompatibility complex genotype on mating success in a free-
ranging reptile population. Proceedings of the Royal Society of London, B, 276,
1695–1704.
Møller, A. P. & M. Petrie. (2002). Condition dependence, multiple sexual signals,
and immunocompetence in peacocks. Behavioral Ecology, 13, 248–253.
Mulard, H., E. Danchin, S. L. Talbot, A. M. Ramey, S. A. Hatch, J. F. White, F.
Helfenstein, & R. H. Wagner. (2009). Evidence that pairing with genetically
similar mates is maladaptive in a monogamous bird. BMC Evolutionary Biol-
ogy, 9, 147. doi:10.1186/1471-2148-9-147.
Neff, B. D. & T. E. Pitcher. (2005). Genetic quality and sexual selection: An inte-
grated framework for good genes and compatible genes. Molecular Ecology,
14, 19–38.
Nesse, R. M., A. Silverman, & A. Bortz. (1990). Sex differences in ability to recognize
family resemblance. Ethology and Sociobiology, 11, 11–21.
Nowak, M. A., K. Tarczy-Hornoch, & J. M. Austyn. (1992). The optimal number of
major histocompatibility complex molecules in an individual. Proceedings of the
National Academy of Sciences, USA, 89, 10896–10899.
Oda, R., A. Matsumoto-Oda, & O. Kurashima. (2002). Facial resemblance of Japa-
nese children to their parents. Journal of Ethology, 20, 81–85.
Oh, K. P. & A. V. Badyaev. (2006). Adaptive genetic complementarity in mate
choice coexists with selection for elaborate sexual traits. Proceedings of the Royal
Society of London, B, 273, 1913–1919.
BEAUTY IN THE EYE OF THE BEHOLDER 343

Olsén, H. (1999). Present knowledge of kin discrimination in salmonids. Genetica,

104, 295–299.
Olsson, M., T. Madsen, J. Nordby, E. Wapstra, B. Ujvari, & H. Wittsell. (2003).
Major histocompatibility complex and mate choice in sand lizards. Proceedings
of the Royal Society of London, B, 270, S254–S256.
Paterson, S. & J. M. Pemberton. (1997). No evidence for major histocompatibility
complex-dependent mating patterns in a free-living ruminant population. Pro-
ceedings of the Royal Society of London, B, 264, 1813–1819.
Pawlowski, B. & R. I. M. Dunbar. (1999). Impact of market value on human mate
choice. Proceedings of the royal Society of London, B, 266, 281–285.
Pedersen, F. A. (1991). Secular trends in human sex ratios: Their influence on indi-
vidual and family behavior. Human Nature, 2, 271–291.
Penn, D. J., K. Damjanovich, & W. K. Potts. (2002). MHC heterozygosity confers a
selective advantage against multiple-strain infections. Proceedings of the
National Academy of Sciences, USA, 99, 11260–11264.
Penn, D. J. & W. K. Potts. (1998a). MHC-disassortative mating preferences reversed
by cross-fostering. Proceedings of the Royal Society of London, B, 265, 1299–
1306.
Penn, D. J. & W. K. Potts. (1998b). How do major histocompatibility complex genes
influence odor and mating preferences? Advances in Immunology, 69, 411–436.
Penn, D. J. & W. K. Potts. (1999). The evolution of mating preferences and major
histocompatibility complex genes. American Naturalist, 153, 145–164.
Penton-Voak, I. S., A. Jacobson, & R. L. Trivers. (2004). Populational differences in
attractiveness judgements of male and female faces: Comparing British and
Jamaican samples. Evolution and Human Behavior, 25, 355–370.
Penton-Voak, I. S., A. C. Little, B. C. Jones, D. M. Burt, B. P. Tiddeman, & D. I.
Perrett. (2003). Female condition influences preferences for sexual dimor-
phism in faces of male humans (Homo sapiens). Journal of Comparative Psychol-
ogy, 117, 264–271.
Penton-Voak, I. S., D. I. Perrett, & J. W. Peirce. (1999). Computer graphic studies of
the role of facial similarity in judgments of attractiveness. Current Psychology,
18, 104–117.
Perrett, D. I., I. S. Penton-Voak. A. C. Little, B. P. Tiddeman, D. M. Burt, N.
Schmidt, et al. (2002). Facial attractiveness judgements reflect learning of
parental age characteristics. Proceedings of the Royal Society of London, B, 269,
873–880.
Perrett, D. I., K. Lee, I. S. Penton-Voak, D. A. Rowland, S. Yoshikawa, D. M. Burt,
S. P. Henzi, D. L. Castles, & S. Akamatsu. (1998). Effects of sexual dimor-
phism on facial attractiveness. Nature, 394, 884–887.
Petrie, M. & B. Kempenaers. (1998). Extra-pair paternity in birds: Explaining varia-
tion between species and populations. Trends in Ecology and Evolution, 13,
52–58.
Pfennig, D. W., P. W. Sherman, & J. P. Collins. (1994). Kin recognition and canni-
balism in polyphenic salamanders. Behavioral Ecology, 5, 225–232.
344 ANIMAL BEHAVIOR

Pfennig, K. S. & R. C. Tinsley. (2002). Different mate preferences by parasitized and

unparasitized females potentially reduces sexual selection. Journal of Evolution-
ary Biology, 15, 399–406.
Pollet, T. V. & D. Nettle. (2008). Driving a hard bargain: Sex ratio and male mar-
riage success in a historical US population. Biology Letters, 4, 31–33.
Potts, W. K., C. J. Manning, & E. K. Wakeland. (1991). Mating patterns in semina-
tural populations of mice influenced by MHC genotype. Nature, 352,
619–621.
Price, M. E., J. Kang, J. Dunn, S. Hopkins. (2011). Muscularity and attractiveness as
predictors of human egalitarianism. Personality and Individual Differences, 50,
636–640.
Promerova, M., M. Vinkler, J. Bryja, R. Polakova, J. Schnitzer, P. Munclinger, & T.
Albrecht. (2011). Occurrence of extra-pair paternity is connected to social
male’s MHC-variability in the scarlet rosefinch Carpodacus erythrinus. Journal
of Avian Biology, 42, 5–10.
Radwan, J., A. Tkacz, & A. Kloch. (2008). MHC and preferences for male odour in
the bank vole. Ethology, 114, 827–833.
Reusch, T. B. H., M. A. Haberli, P. B. Aeschlimann, & M. Milinski. (2001). Female
sticklebacks count alleles in a strategy of sexual selection explaining MHC
polymorphism. Nature, 414, 300–302.
Rhodes, G., J. Chan, L. A. Zebrowitz, & L. W. Simmons. (2003). Does sexual
dimorphism in human faces signal health? Proceedings of the Royal Society of
London, B, 270, S93–S95.
Richardson, D., J. Komdeur, T. Burke, & T. von Schantz. (2005). MHC-based pat-
terns of social and extra-pair mate choice in the Seychelles warbler. Proceedings
of the Royal Society of London, B, 272, 759–767.
Roberts, S. C. (2009). Complexity and context of MHC-correlated mating prefer-
ences in wild populations. Molecular Ecology, 18, 3121–3123.
Roberts, S. C. & L. M. Gosling. (2003). Genetic similarity and quality interact in
mate choice decisions by female mice. Nature Genetics, 35, 103–106.
Roberts, S. C., L. M. Gosling, V. Carter, & M. Petrie. (2008). MHC-correlated
odour preferences in humans and the use of oral contraceptives. Proceedings
of the Royal Society of London, B, 275, 2715–2722.
Roberts, S. C., K. Klapilová, A. C. Little, R. P. Burriss, B. C. Jones, L. M. DeBruine,
et al.. (2012). Relationship satisfaction and outcome in women who meet their
partner while using oral contraception. Proceedings of the Royal Society of
London, B, 279, 1430–1436.
Roberts, S. C., A. C. Little, L. M. Gosling, B. C. Jones, D. I. Perrett, V. Carter, & M.
Petrie. (2005). MHC-assortative facial preferences in humans. Biology Letters,
1, 400–403.
Roberts, S. C., A. C. Little, L. M. Gosling, D. I. Perrett, V. Carter, B. C. Jones, I. S.
Penton-Voak, & M. Petrie. (2005). MHC-heterozygosity and human facial
attractiveness. Evolution & Human Behavior, 26, 213–226.
BEAUTY IN THE EYE OF THE BEHOLDER 345

Ryder, T. B., W. P. Tori, J. G. Blake, B. A. Loiselle, & P. G. Parker. (2010). Mate

choice for genetic quality: A test of the heterozygosity and compatibility
hypotheses in a lek-breeding bird. Behavioral Ecology, 21, 203–210.
Schaefer, M. L., D. A. Young, & D. Restrepo. (2001). Olfactory fingerprints for
major histocompatibility complex-determined body odors. Journal of Neurosci-
ence, 21, 2481–2487.
Schellinck, H. M., E. Rooney, & R. E. Brown. (1995). Odors of individuality of
germ-free mice are not discriminated by rats in a habituation-dishabituation
procedure. Physiology & Behavior, 57, 1005–1008.
Schwensow, N., J. Fietz, K. Dausmann, & S. Sommer. (2008). MHC-associated mat-
ing strategies and the importance of overall genetic diversity in an obligate pair-
living primate. Evolutionary Ecology, 22, 617–636.
Seddon, N., W. Amos, R. A. Mulder, & J. A. Tobias. (2004). Male heterozygosity
predicts territory size, song structure and reproductive success in a co-
operatively breeding bird. Proceedings of the Royal Society of London, B, 271,
1823–1829.
Setchell, J. M., M. J. E. Charpentier, K. M. Abbott, E. J. Wickings, & L. A. Knapp.
(2010). Opposites attract: MHC-associated mate choice in a polygynous pri-
mate. Journal of Evolutionary Biology, 23, 136–148.
Setchell, J. M. & E. Huchard. (2010). The hidden benefits of sex: Evidence for
MHC-associated mate choice in primate societies. Bioessays, 32, 940–948.
Shepher, J. (1971). Mate selection among second generation kibbutz adolescents and
adults: Incest avoidance and negative imprinting. Archives of Sexual Behavior,
1, 293–307.
Sherman, P. W. (1977). Nepotism and the evolution of alarm calls. Science, 197,
1246–1253.
Silk, J. B. (2002). Kin selection in primate groups. International Journal of Primatol-
ogy, 23, 849–875.
Singer, A. G., G. K. Beauchamp, & K. Yamazaki. (1997). Volatile signals of the major
histocompatibility complex in male mouse urine. Proceedings of the National
Academy of Sciences, USA, 94, 2210–2214.
Singh, P. B., R. E. Brown, & B. Roser. (1987). MHC antigens in urine as olfactory
recognition cues. Nature, 327, 161–164.
Skarstein, F., I. Folstad, S. Liljedal, & M. Grahn. (2005). MHC and fertilization suc-
cess in the Arctic charr (Salvelinus alpinus). Behavioral Ecology and Sociobiology,
57, 374–380.
Smith, F. G., B. C. Jones, L. L. W. Welling, A. C. Little, J. Vukovic, J. C. Main, & L.
M. DeBruine. (2009). Waist-hip ratio predicts women’s preferences for mas-
culine male faces, but not perceptions of men’s trustworthiness. Personality
and Individual Differences, 47, 476–480.
Souroukis, K. & A. M. Murray. (1994). Female mating behavior in the field cricket,
Gryllus pennsylvanicus (Orthoptera: Gryllidae) at different operational sex
ratios. Journal of Insect Behaviour, 8, 269–279.
346 ANIMAL BEHAVIOR

Spoon, T. R., J. R. Millam, & D. H. Owings. (2006). The importance of mate

behavioural compatibility in parenting and reproductive success by cockatiels,
Nymphicus hollandicus. Animal Behaviour, 71, 315–326.
Stapleton, M. K., O. Kleven, J. T. Lifjeld, & R. J. Robertson. (2007). Female tree
swallows (Tachycineta bicolor) increase offspring heterozygosity through extrap-
air mating. Behavioral Ecology and Sociobiology, 61, 1725–1733.
Tarvin, K. A., M. S. Webster, E. M. Tuttle, & S. Pruett-Jones. (2005). Genetic sim-
ilarity of social mates predicts the level of extrapair paternity in splendid fairy-
wrens. Animal Behaviour, 70, 945–955.
Thornhill, R. & S. W. Gangestad. (2006). Facial sexual dimorphism, developmental
stability, and susceptibility to disease in men and women. Evolution and
Human Behavior, 27, 131–144.
Thornhill, R., S. W. Gangestad, R. Miller, G. Scheyd, J. K. McCullough, & M.
Franklin. (2003). Major histocompatibility genes, symmetry and body scent
attractiveness in men and women. Behavioral Ecology, 14, 668–678.
van der Jeugd, H. P., I. T. van der Veen, & K. Larsson. (2002). Kin clustering in bar-
nacle geese: Familiarity or phenotype matching? Behavioral Ecology, 13, 786–
790.
Verrell, P. A. (1983). The influence of ambient sex ratio and intramale competition
on the sexual behaviour of the redspotted newt, Notophthalmus viridescens
(Amphibia: Urodela: Salamandridae). Behavioral Ecology and Sociobiology,
13, 307–313.
Vos, D. R. (1995). Sexual imprinting in zebra-finch females: Do females develop a
preference for males that look like their father? Ethology, 99, 252–262.
Vukovic, J., B. C. Jones, L. DeBruine, D. R. Feinberg, F. G. Smith, A. C. Little, L. L.
M. Welling, & J. Main. (2010). Women’s own voice pitch predicts their pref-
erences for masculinity in men’s voices. Behavioral Ecology, 21, 767–772.
Watkins, C. D., L. M. DeBruine, A. C. Little, & B. C. Jones. (2012). Social support
influences preferences for feminine facial cues in potential social partners.
Experimental Psychology, 59, 340–347.
Watkins, C. D., L. M. DeBruine, F. G. Smith, B. C. Jones, J. Vukovic, & P. J.
Fraccaro. (2011). Like father, like self: Emotional closeness to father predicts
women’s preferences for self-resemblance in opposite-sex faces. Evolution and
Human Behavior, 32, 70–75.
Watkins, C. D., B. C. Jones, A. C. Little, L. M. DeBruine, & D. R. Feinberg. (2012).
Cues to the sex ratio of the local population influence women’s preferences for
facial symmetry. Animal Behaviour, 83, 545–553.
Wedekind, C., T. Seebeck, F. Bettens, & A. J. Paepke. (1995). MHC-dependent
mate preferences in humans. Proceedings of the Royal Society of London, B,
260, 245–249.
Wedekind, C., M. Walker, J. Portmann, B. Cenni, R. Muller, & T. Binz. (2004).
MHC-linked susceptibility to a bacterial infection, but no MHC-linked cryp-
tic female choice in whitefish. Journal of Evolutionary Biology, 17, 11–18.
BEAUTY IN THE EYE OF THE BEHOLDER 347

Wegner, K. M., M. Kalbe, J. Kurtz, T. B. H. Reusch, & M. Milinski. (2003). Parasite

selection for immunogenetic optimality. Science, 301, 1343–1343.
Wegner, K. M., T. B. H. Reusch, & M. Kalbe. (2003). Multiple parasites are driving
major histocompatibility complex polymorphism in the wild. Journal of Evolu-
tionary Biology, 16, 224–232.
Westerdahl, H. (2004). No evidence of an MHC-based female mating preference in
great reed warblers. Molecular Ecology, 13, 2465–2470.
Westermarck, E. A. (1921). The History of Human Marriage. London: Macmillan.
Wilson, G. D. & P. T. Barrett. (1987). Parental characteristics and partner choice: Some
evidence for Oedipal imprinting. Journal of Biosocial Science, 19, 157–161.
Wiszewska, A., B. Pawlowski, & L. G. Boothroyd. (2007). Father-daughter relation-
ship as a moderator of sexual imprinting: A facialmetric study. Evolution and
Human Behavior, 28, 248–252.
Wolf, A. P. (1968). Childhood association, sexual attraction, and the incest taboo: A
Chinese case. American Anthropologist, 68, 883–898.
Wolf, A. P. (1993). Westermarck redivivus. Annual Review of Anthropology, 22, 157–175.
Wolf, A. P. (1995). Sexual Attraction and Childhood Association: A Chinese Brief for
Edward Westermarck. Stanford, CA: Stanford University Press.
Yamaguchi, M., K. Yamazaki, G. K. Beauchamp, J. Bard, L. Thomas, & E. A. Boyse.
(1981). Distinctive urinary odors governed by the major histocompatibility
locus of the mouse. Proceedings of the National Academy of Sciences, USA, 78,
5817–5820.
Yamazaki, K., G. K. Beauchamp, J. Bard, & E. A. Boyse. (1990). Single MHC gene
mutations alter urine odour constitution in mice. In D. W. MacDonald, D.
Müller-Scharze, & S. E. Natynczuk (eds.), Chemical Signals in Vertebrates 5
(pp. 255–259). New York: Plenum Press.
Yamazaki, K., G. K. Beauchamp, Y. Imai, J. Bard, S. P. Phelan, L. Thomas, et al.
(1990). Odortypes determined by the major histocompatibility complex in
germ-free mice. Proceedings of the National Academy of Sciences, USA, 87,
8413–8416.
Yamazaki, K., G. K. Beauchamp, F. W. Shen, J. Bard, & E. A. Boyse. (1991). A dis-
tinctive change in odortype determined by H-2D/L mutation. Immuno-
genetics, 34, 129–131.
Yamazaki, K., E. A. Boyse, V. Miké, H. T. Thaler, B. J. Mathieson, J. Abbott, J.
Boyse, Z. A. Zayas, & L. Thomas. (1976). Control of mating preferences in
mice by genes in the major histocompatibility complex. Journal of Experimen-
tal Medicine, 144, 1324–1335.
Zajonc, R. B., P. K. Adelmann, S. T. Murphy, & P. M. Niendenthal. (1987). Con-
vergence in the physical appearance of spouses. Motivation and Emotion, 11,
335–346.
Zei, G., P. Astolfi, & S. D. Jayakar. (1983). Correlation between father’s age and hus-
band’s age: A case of imprinting. Journal of Biosocial Science, 15, 116–117.
Ziegler, A., H. Kentenich, & B. Uchanska-Ziegler. (2005). Female choice and the
MHC. Trends in Immunology, 26, 496–502.
This page intentionally left blank
 Glossary

Action potential—A brief change in the electrical potential of the cell membrane of
a neuron. Action potentials are propagated along the axons of neurons to convey sig-
nals from one part of the nervous system to another.
Active space—The area over which a signal can be detected by a receiver.
Active time—How long a signal is remembered by the receiver.
Agonistic behaviors—Any social behavior related to fighting.
Alarm call—A signal produced by an animal in response to danger.
Altruistic behavior—A behavior that is costly to the individual performing it and
beneficial to another individual, where costs and benefits are measured in terms of
effects on expected lifetime reproductive success.
Amplexus—A form of mating in which a male amphibian grasps a female with his
front legs to enable him to fertilize the female’s eggs.
Anadromous—Referring to fishes that spend most of their lives in the sea and
migrate to fresh water to breed.
Animal cognition—The study of the mental capacities of nonhuman animals.
Audience effect—A behavior only being produced in the presence of an appropriate
receiver.
Azimuth—An angular measurement from a fixed point, such as north, that
describes a location on the horizon.
Behavioral ecology—An approach to investigations of how different ecological
circumstances affect animal behavior and the evolution of traits with functions that
fit these different environments.
350 GLOSSARY

Behavioral strategy—One of two or more options subject to natural selection.

Behavioral syndrome—Correlation of rank-order differences among individuals
through time or across situations.
Behavior reading—Ability to assess and predict future behaviors of a con- or hetero-
specific based on the latter’s (or demonstrator’s) “body language” or behavioral patterns.
Bet hedging—An evolutionary strategy to optimize fitness in varying and unpredict-
able environments that is similar to betting on multiple outcomes, in which mean
fitness is sacrificed to decrease variation in fitness.
Biomedicine—The branch of medical science that applies biological principles to
clinical practice.
Broadband—An auditory signal with a wide range of frequencies.
Brood parasite—An animal that places its eggs or young with another individual to
raise.
Brood parasitism—Use of a host of the same species (intraspecific brood parasit-
ism) or a different species (interspecific brood parasitism) to care for the parasite’s
offspring. Occurs in birds, fish, and insects.
Brood reduction—Frequency-dependent mortality of dependent young.
Cache—A hiding place for food; food-storing behavior of animals.
Categorical perception—Continuous stimulus variation that is perceived as being
in more than one category where discrimination of two stimuli that differ in the
same magnitude is stronger when the stimuli are in different categories than when
they are in the same category.
Cephalothorax—The fused head and thorax of arthropods.
Circadian rhythm—An endogenous activity rhythm with a period of about 24 hours.
Cladistics—A system of classification based on the evolutionary relationships and
history of groups of organisms rather than purely on shared features.
Clock shift—A light/dark cycle out of phase with the earth’s 24-hour light/dark
cycle.
Coefficient of relatedness—A measure for the level of consanguinity (degree to
which they are descended from a common ancestor) between two given individuals.
Coevolutionary arms race—Two species (e.g., prey and predator, or host and para-
site) interact on an ongoing basis, each forcing the other to evolve better arms
(predator or parasite) or better defenses (prey or host).
Cognitive ethology—Application of Tinbergen’s four questions about the causation,
development, function, and evolution of behavior to the cognitive and mental abil-
ities of animals.
Cognitive legacy—The nongenetic inherited expression of perceiving, learning,
remembering, and making decisions in the environment.
GLOSSARY 351

Cognitive repertoire—The collection of cognitive abilities of a species as inherited

in the gene pool.
Collective reasoning—Situations in which animals perceive, learn, remember, and
make decisions while interacting with conspecifics in the environment.
Commensal microflora—A symbiosis between bacteria present in the large intestine
and the human host in which the bacteria benefit but the human host is neither
helped nor harmed.
Common garden—An experimental design in which two or more species are trans-
planted from different environments into a common environment.
Community—A group of actually or potentially interacting species living in the
same place.
Conditioning—A means of modifying behavior in which a consequence makes a
behavior likely or unlikely to occur again.
Confusion effect—A large number of prey escaping from predation at the same time
making it difficult for the predator to pursue a specific individual prey animal.
Consanguinity—Relationship by blood or by a common ancestor.
Conservation behavior—The application of knowledge of animal behavior to solve
wildlife conservation problems.
Conspecific—A member of the same species.
Contingency—Dependence of an event on the fulfillment of a condition.
Convergent phenotypes—Similar phenotypes that have evolved independently in
two or more species that are not closely related.
Converging evolution—Species with very different evolutionary backgrounds that
may have diverged long ago converging to similar behavioral traits due to similarities in
social systems or environments that lead to similar solutions to adaptive problems.
Cooperative breeding—A social system in which individuals contribute care to off-
spring that are not their own at the expense of their own reproduction.
Core (offspring)—Offspring in a brood that the parents normally raise fully.
Counterbalanced—An experimental design in which different groups of subjects
receive treatments in different orders to control for order effects.
Countersinging—Singing in response to the singing of another individual.
Crop milk—A secretion from the lining of the crop of pigeons and doves with
which the parents feed their young by regurgitation.
Cryptic—Difficult to see.
Cryptic species—Two or more morphologically indistinguishable biological groups
that are incapable of interbreeding.
Crystallized song—The final stage of song development in which the song or song
types are produced with little variation (are stereotyped).
352 GLOSSARY

Deception (deceptive signal)—In a functional sense, the signaler benefits by causing

a receiver to alter its behavior in a costly way. In a cognitive sense, the signaler attrib-
utes mental states to itself and others, acts intentionally to change the behavior of the
responder, and creates a mental representation of an external event in the receiver’s
mind, which results in actions that are based on that receiver’s mental representation.
Definitive plumage—The plumage of a bird that, once attained, does not change
substantially in color or pattern for the rest of the bird’s life.
Developmental constraint—A bias on the development of behavior or a limitation
on phenotypic variability caused by the structure, character, composition, or dynam-
ics of the developmental system.
Developmental psychology—The branch of psychology that studies the social and
mental development of children.
Diel—Denoting a period of 24 hours.
Dilution effect—Reduction in the likelihood that a predator will eat a particular
individual because a number of individuals are present.
Diploid—An organism or cell with two sets of chromosomes—usually, one set from
the mother and another set from the father.
Disassortative mate preference—Choice of a mate with a dissimilar phenotype.
Discriminability—The ease with which a receiver can determine a signal’s message.
Discrimination reversal—Ability to exchange the attribute or value of an item with
another item.
Display—A stereotyped pattern of behavior that functions in communication with
others.
Diverging evolution—Two populations becoming increasingly different from one
another, eventually forming separate species.
Dominance—The effect of one allele at a locus being partially or entirely hidden by
another allele at the same locus. A social relationship in which one individual has
power or priority of access over another.
Dominance hierarchy—A ranking of individuals in a social group based on their
dominance relationships.
Dominant frequency— The frequency of a sound with the most energy.
Donation behavior—Voluntary sharing of a resource (e.g., food) with another
individual.
Downregulation—The process by which a cell decreases the quantity of a cellular
component in response to an external variable.
Eavesdropping—Observing the contests, behavior, or signaling of others.
Ecological speciation—The evolution of reproductive isolation between populations
as a result of ecologically based divergent natural selection.
GLOSSARY 353

Ecology—The study of the relations of organisms to one another and to their physi-
cal surroundings.
Ecosystem—An interacting community of living organisms and the nonliving com-
ponents of their environment.
Ectothermic—An organism that controls body temperature through external means.
Egg rejection—Removal of an egg from the nest.
Endothermic—An organism that generates heat to maintain its body temperature.
Entorhinal cortex—A part of the temporal cortex that receives input from all sen-
sory systems and provides major input to the hippocampus.
Episodic-like memory—In animals, memory for events that includes what occurred,
where, and when.
Episodic memory—In humans, autobiographical memory for personally experi-
enced past events. Episodic memory differs from semantic memory, for the meaning
of words or concepts, and procedural memory, for skills and habits.
Epistasis—The effects of one gene being modified by one or several other (modi-
fier) genes.
Ethologist—A scientist who studies the behavior of animals from a biological
perspective.
Ethospecies—A species whose genetic isolation depends on its behavior.
Eventual variety—A song organization in which one song type is repeated many
times before switching to another song type, which is then repeated many times.
Evolutionary biology—The study of the origin and descent with modification of
species and their changes over time.
Experimental learning—Learning by trial and error.
Extra-pair copulation/mating (EPC)—Sexual activity with a partner other than
the mate.
Extra-pair siring—Production of offspring by males other than a female’s mate.
Facultative—An organism that does something (e.g., siblicide) as an option.
Fitness—Ability to survive and reproduce.
Fluctuating asymmetry (FA)—An organism’s deviation from perfect bilateral
symmetry.
Food (sharing) call—A signal that attracts a mate, offspring, or members of the
social group to a food source.
Food web—A depiction of feeding connections (what species eats and what species
gets eaten) in an ecological community.
Foraging—Finding and collecting resources.
Frugivory—Eating fruit.
354 GLOSSARY

Functionally referential—Signals that evoke highly specific receiver responses in

the absence of contextual cues.
Fundamental frequency—The lowest frequency of a sound.
Genetics—The study of heredity and the variation of inherited characteristics.
Genome expression analysis—The simultaneous measurement of the activity of
thousands of genes to produce a broad understanding of cellular function.
Graded—Something that varies in intensity or frequency or both, thereby transmit-
ting quantitative information.
Graph theory—The study of mathematical structures (graphs) used to model con-
nections (edges) between objects (nodes) from a collection of objects.
Grid cells—Neurons in the entorhinal cortex that fire when the animal is in any of
a number of spatial locations that are distributed in a regular grid-like pattern.
Habituation—The decline of a behavior with the repeated presentation of an elicit-
ing stimulus.
Haploid—A cell or organism with a set of unpaired chromosomes.
Harmonics—A series of tones whose frequencies are integral multiples of the funda-
mental frequency of a sound.
Hatching asynchrony—A pattern in which eggs in a brood do not hatch together.
Head-direction cells—Neurons in a brain region called the subiculum that fire
when the animal is facing in a particular direction.
Heritable—(1) Able to be inherited. (2) The extent to which a phenotype is influ-
enced by the genotype.
Heritable variation—The proportion of observable differences in a trait among
individuals in a population that is due to genetic differences.
Heterospecific—A member of a different species.
Heterozygote—An individual having two different alleles of a particular gene or
genes.
Heterozygote advantage—The genotype with two different alleles for a gene has a
higher relative fitness than either of the genotypes with two copies of the same allele.
Hippocampus—A part of the brain in birds and mammals involved in memory and
spatial orientation.
Home range—The geographic area to which an animal normally confines its activity.
Homology—Sharing a common ancestry.
Homozygote—An individual having two copies of the same allele of a particular
gene or genes.
Illuminance spectrophotometry and spectral analysis—A method of measuring
the components of light.
GLOSSARY 355

Immediate early genes (IEGs)—Genes that are activated temporarily and rapidly in
response to a wide variety of cellular stimuli.
Immediate variety—A song organization in which a bird changes song types after
every song.
Imprinting—A learning process that occurs early in the life of a social animal in
which a specific behavior pattern is established through association with a parent or
other role model.
Inbreeding—Mating of closely related individuals. Production of offspring from
parents more closely related than the average of a population.
Inclusive fitness—The ability of an individual to contribute its genes to the next
generation, taking into account the shared genes passed on by the individual’s close
relatives.
Incremental forgetting—An individual forgetting in a sequence of gradual steps
something it has previously learned.
Information centrality—How often a node lies along edges (paths) that connect
other pairs of individuals within a social network.
Insight learning—Ability to understand the intricacies of a problem and solve it
without previous trial and error.
Intersexual selection—Differential reproductive success that results from individ-
uals of one sex choosing mates from members of the other sex.
Interspecific—Occurring between members of different species.
Intrasexual selection—Differential reproductive success that results from members
of one sex competing with each other for opportunities to mate with members of
the other sex.
Intraspecific—Occurring between members of the same species.
Kin selection—A mechanism of evolution in which characteristics favor the survival
of close relatives of an individual.
Lateral plates—Rigid, bony plates along the side of a fish’s body.
Learning-set formation—Ability to “learn to learn,” or to be able to solve a cogni-
tive problem and use this experience to solve subsequent problems of increasing dif-
ficulty in fewer trials or faster time.
Lek—An assembly area where animals display and court potential mates.
Linear algebra—A branch of mathematics that uses a system of linear equations of
several unknowns.
Maculation—Spotting pattern.
Major histocompatibility complex (MHC)—A cell-surface protein molecule that
helps the vertebrate immune system recognize foreign substances and that is
encoded by a large gene family.
356 GLOSSARY

Marginal (offspring)—“Extra” offspring in a brood that parents do not normally

raise fully.
Mate choice—Process by which individuals select social or copulation partners.
Mate fidelity—Being faithful to a reproductive partner.
Maternal perinatal association (MPA)—The association between a mother and
her newborn baby.
Mating success—The number of mates acquired or matings achieved.
Matrilateral—Related through the mother.
Meiosis—Special cell division that produces (haploid) gametes.
Metapopulation—A group of spatially separated populations of the same species
that interact at some level.
Microarray—A two-dimensional arrangement of miniaturized chemical-reaction
areas a glass slide or silicon thin film that is used to test DNA fragments, antibodies,
or proteins.
Mimetic—Relating to mimicry (imitation or resemblance).
Mobbing—An antipredator behavior in which prey individuals collectively attack or
harass a predator.
Mobbing call—A vocalization that elicits mobbing behavior or recruits individuals
to mob a predator.
Monogamy—A form of social behavior in which an individual has only one mate at
any one time.
Morphological—Relating to the form and structure of organisms.
Multimodal signal—A signal that uses more than one sensory modality.
Multiple messages—Signals that convey different information in different modalities.
Natural selection—The nonrandom process by which traits become either more or
less common in a population as a function of differential survival or reproduction of
their bearers.
Navigation—Ability to use current global position to advance to a destination.
Neurogenesis—The production of new neurons.
Numerosity (numerical concept)—Ability to understand the quantity of items or
objects, including the symbolic and abstract representation of quantity as well as the
concept of zero.
Nuptial coloration—The colors and patterns of color expressed by individuals seek-
ing mates.
Object permanence—To be aware that objects continue to exist even when they are
not physically present before the animal.
Obligate—Capable of functioning or surviving only by performing a particular
behavior (e.g., siblicide).
GLOSSARY 357

Observational learning—A type of learning that occurs by observing, retaining,

and replicating behavior performed by others.
Operational sex ratio—The relative number of sexually competing males that are
ready to mate versus sexually competing females that are ready to mate.
Orientation—The determination of an animal’s position relative to a goal.
Pair bond—An enduring preferential association between two sexually mature
adults that is characterized by selective affiliation, contact, and copulation of the
partners.
Parasitic castration—An internal parasite suppressing development of the gonads
and subsequent expression of adult secondary sex characteristics.
Parasitoid—An insect whose larvae live as parasites that eventually kill their hosts.
Parental care—Any behavior by a parent that enhances the reproductive success
(fitness) of offspring.
Parental investment—Any behavior that increases offspring fitness but decreases
the parents’ ability to invest in other offspring.
Paternal care—Care for offspring provided by the father.
Path integration—A method of orientation in which an animal’s motion, including
turns and straight paths, is integrated to maintain a updated record of the correct
distance and direction to the start of the path.
Patrilateral—Related through the father.
Pedipalps—The second pair of head appendages of spiders, horseshoe crabs, scor-
pions, and mites.
Phenotypic evolution—The descent with modification of the physical, chemical,
and behavioral characteristics of a species.
Phenotypic gambit—The concept that study of the function of behavior does not
require knowledge of underlying genetic or molecular mechanisms.
Philopatric—Returning faithfully to or remaining near a particular site or area such
as the home nesting colony.
Phonotaxis—Movement oriented with respect to a sound source.
Physiological zero—The temperature at which bird eggs will begin to develop
(about 30oC).
Physiology—The study of the physical and chemical phenomena involved in the
functions and activities of organisms, organs, tissues, and cells.
Plastic song—Vocalizations during the development of song that have distinct sylla-
bles and recognizable elements, some of which will eventually be used in adult song.
Playback experiment—A method in which recordings of an animal’s vocalizations
are broadcast from a speaker and the response of the subject is observed.
Pleiotropy—One gene influencing multiple phenotypic traits.
358 GLOSSARY

Polarized light—Light with oscillation of its wave form confined to a single plane
at a right angle to the direction of travel of the wave.
Polyandry—A form of social behavior in which a female has more than one mate;
females mate multiply, resulting in stronger sexual selection on females.
Polygyny—A form of social behavior in which males have more than one mate;
males mate multiply, resulting in stronger sexual selection on males.
Polyphenism—Multiple discrete forms arising from a single genotype as a result of
different environmental conditions.
Predation pressure—The extent to which predation affects a population of a prey
species.
Predator—An animal that preys upon other animals.
Premating reproductive isolation—The inability of different species to breed with
each other as a result of mechanisms that occur prior to mating.
Prey—An animal that is preyed upon by other animals.
Promiscuity—A form of social behavior in which a member of one sex mates with
more than one member of the other sex but the relationship is ephemeral and dis-
solves after mating.
Prosocial—Relating to behavior that is positive, helpful, and intended to promote
social acceptance and friendship.
Prosoma—The anterior portion of the body of spiders that lacks segmentation.
Proximate—The reason a behavior is performed in terms of its immediate trigger
or mechanisms that control its performance.
Quantitative trait loci (QTL) mapping—Identifying the locations of genes that
affect a quantitative trait.
Radiation—Diversification of a species or single ancestral type into several forms.
Radioactive isotope—An isotope (a form of a chemical element with the same
number of protons in the nucleus but a different of neutrons) that has an unstable
nucleus that emits radiation.
Recognition template—A pattern or memory used to guide the development of dis-
crimination among objects.
Repertoire match—A response targeted to a particular individual in which a
responder sings a song type in the target individual’s repertoire.
Representational—A signal that generates a mental representation of a specific
object in the mind of the listener.
Reproductive inhibition—Individuals being prevented from reproducing by more
dominant individuals.
Seismic communication—Communication through vibrations generated by percus-
sion on the substrate.
GLOSSARY 359

Semelparous—Waiting until near the end of the lifespan to reproduce and then
having little or no postreproductive lifespan.
Sensitive period—A portion of the lifespan during which an organism can respond
to exogenous stimuli that are needed for the development of a particular behavior.
Sensory bias—A preference for particular traits that did not evolve as a result of the
sexual selection caused by that preference.
Separation calls—Vocalizations given in response to being apart from one or more
members of a social group.
Sequencing—Identifying the arrangement of subunits (e.g., nucleotide residues or
amino acids) in a larger molecule (e.g., DNA or protein).
Sexual imprinting—A mating preference in adulthood that was learned through
experiences when very young.
Sexual selection—Differential reproductive success caused by competition for mates.
Siblicide—The killing of a sibling or siblings as a behavior pattern typical in various
animal groups.
Signal—A characteristic that has evolved by natural selection to convey information
to other organisms.
Sign stimulus—The stimulus that releases the stereotyped actions of animals.
Site fidelity—The degree to which an animal returns to a specific site.
Social facilitation—Improvement in the performance of a behavior when in the
presence of others.
Social learning—Facilitation of the acquisition of adaptive patterns of behavior by
observing the behavior of others.
Social network analysis—A set of mathematical and visual methods for exploring
complicated connections among entities.
Sociosexuality —An individual’s willingness to engage in sexual activity without
having emotional ties to the sex partner.
Song repertoire—The number of different song types produced by a single bird.
Song type—A distinct rendition of the species-typical song.
Sound spectrogram—A visual representation of sound in which the horizontal
dimension corresponds to time (reading from left to right) and the vertical dimen-
sion corresponds to frequency (or pitch), with higher sounds shown higher on the
display.
Spatial memory—Ability to discriminate between and remember specific and
different locations of objects in relation to each other, the observer, and the
environment.
Sun compass—Orientation using the position of the sun and the time of day.
Sympatric—Two species existing in the same geographic area and regularly encoun-
tering each another.
360 GLOSSARY

Syntax—The rules by which elements are placed into a sequence.

Tactical deception—Ability to purposely confuse, deceive, or misguide a con- or
heterospecific.
Teaching—Behavior that facilitates learning of a naïve individual at some cost to
the teaching individual.
Territory—A fixed area containing specific resources such as food, nest sites, or mates
from which one or more individuals exclude other members of the same species.
Tool use—To use or purposely modify inanimate objects outside the animal’s own
anatomy to procure or modify resources or to alter the environment with the pur-
pose of benefiting from the alteration.
Trade-off—A situation that involves losing one aspect of something in return for
gaining another aspect.
Tradition—A behavior that is performed repeatedly over a period, that is shared
among two or more members of a group, and that depends in part on socially aided
learning for its generation in new practitioners.
Transcriptome sequencing—Identifying the arrangements of nucleotide residues in
the RNA molecules produced by one or a population of cells.
Transitive reasoning—Ability to infer a relationship between two or more items or
objects based on previous but incomplete information about the items or objects.
Triad census—A method of quantifying the relations among animals by considering
three at a time.
Type I error—In statistics, the probability of rejecting a true null hypothesis. Also,
alarm calling to the approach of an innocuous species.
Type II error—In statistics, the probability of accepting a false null hypothesis.
Also, failing to respond to a dangerous predator.
Type match—A response targeted to a particular individual in which a responder
sings the same song type as the target individual.
Ultimate—The reason a particular behavior evolved in terms of its effect on repro-
ductive success.
Unintended receivers—Individuals that eavesdrop on signals that did not evolve to
communicate with those individuals.
Upregulation—The process by which a cell increases the quantity of a cellular com-
ponent in response to an external variable.
Valid—Verifiably correct.
Vector—A mathematical quantity that has both magnitude and direction.
Visual perspective taking—Ability to use an animal’s own visual perspective and
the visual perspectives of others to assess, understand, and predict the behavior of
con- and heterospecifics.
Vocal repertoire—All of the subtle variations in male song.
 About the Editor and Contributors

EDITOR
Editor KEN YASUKAWA is Mead Family Professor of Biology at Beloit Col-
lege and Honorary Fellow in Zoology at the University of Wisconsin at Mad-
ison. Using Niko Tinbergen as his model, he has studied the behavior and
ecology of birds and has focused primarily on the red-winged blackbird since
1973.

CONTRIBUTORS
ÇAĞLAR AKÇAY is a Postdoctoral Fellow at the Laboratory of Ornithology
at Cornell University in Ithaca, New York. He studies kin recognition and
vocal signaling in birds.

MICHAEL D. BEECHER is Professor of Psychology and Biology at the

University of Washington in Seattle, Washington. He studies the function
and development of song sparrow song using a combination of field and
laboratory methods.

ALISON M. BELL is an Assistant Professor in the School of Integrative

Biology at the University of Illinois. Her research focuses on understanding
why individual animals behave differently from each other, or why individuals
of the same species have different personalities. She uses the threespined
362 ABOUT THE EDITOR AND CONTRIBUTORS

stickleback fish as a model system to study the proximate mechanisms under-

lying personality and the ultimate (evolutionary) consequences of personality.

XIMENA E. BERNAL is an Assistant Professor in the Department of Biologi-

cal Sciences at Purdue University and a Research Associate at the Smithsonian
Tropical Research Institute in Panama. She is interested in animal communi-
cation and in particular eavesdropping behavior.

DAVID CLARK, a Professor of Biology at Alma College, Alma, Michigan, has

dedicated his research to animal communication and the evolution of visual dis-
plays. His studies have focused mainly on the dimorphic jumping spider but
include lizards, toads, turtles, and wolf spiders. In some of his most recognized
research, he presented realistic, computer-animated images of jumping spiders
displayed on a handheld television screen to live female jumping spiders. Clark’s
work has been published in various scientific journals, including Animal Behav-
iour and the Proceedings of the National Academy of Sciences, and was cited in
Natural History. His work has been featured in National Geographic, BBC Wild-
life, the Chicago Tribune, the New York Times, and the Chronicle of Higher Edu-
cation. Clark has taught classes in animal behavior, biology of spiders, ecology,
evolution, general zoology, invertebrate biology, and scientific methods. His
research projects with Alma College undergraduates have examined spider mate
choice, anole lizard visual communication using robotics, and eavesdropping in
male wolf spiders. Clark obtained his B.S. and M.S. degrees from Central
Michigan University and his Ph.D. from the University of Cincinnati.

LISA M. DEBRUINE is a Reader in the Institute of Neuroscience and

Psychology at the University of Glasgow. She is interested in how humans
use facial resemblance to tell who their kin are and how people respond to cues
of kinship in different circumstances.

CHRISTOPHER S. EVANS was Professor of Brain, Behaviour, and Evolu-

tion at Macquarie University. Chris was interested in the structure and func-
tion of signals, using both controlled experiments and observational studies
under natural conditions. This two-pronged strategy combined code-
breaking exercises designed to provide insights into mechanism with studies
of social groups that documented variation in reproductive success. The goal
is to link cognition with function.

MARK E. HAUBER is Professor of Psychology at Hunter College and Head

of Biopsychology and Behavioral Neuroscience at the Graduate Center of the
ABOUT THE EDITOR AND CONTRIBUTORS 363

City University of New York. His research focuses on evolutionary hypotheses

and ecological contexts of animal behavior. He is interested in the behavioral
and neural basis of social recognition systems and focuses on studying how
animals tell their friends apart from their enemies.

BENEDICT C. JONES is a Professor in the Institute of Neuroscience and

Psychology at the University of Glasgow. He is interested in how people
respond to facial cues.

ANTHONY C. LITTLE is a Royal Society University Research Fellow in the

Department of Psychology at the University of Stirling. His research has
focused on whether an evolutionary approach can help answer the question
of what makes certain faces attractive and why certain traits may be valued.

JASON LOW is Senior Lecturer in Psychology at Victoria University of

Wellington. His research focuses on the nature, development, and limits of
human children and adults’ implicit and explicit theory-of-mind systems to
ascribe complex mental states to predict behavior. He is also interested in social
and numerical cognition in animals and focuses on studying how New Zealand
robins encode numerosity in the wild and interpret human behavioral cues.

DAVID B. MCDONALD is a Professor in the Department of Zoology and

Physiology at the University of Wyoming. He studies behavioral and evolution-
ary ecology using a combination of social network models and detailed behavioral
studies of marked individuals. Social network models enable Dave to use a male
manakin’s past social interactions to predict his fate five years later.

DOUGLAS W. MOCK is George Lynn Cross Research Professor Emeritus of

Biology at the University of Oklahoma. He is generally interested in evolu-
tionary and ecological aspects of behavior and is particularly interested in com-
bining direct behavioral observations with experimental testing of
evolutionary hypotheses under field conditions.

RACHEL A. PAGE is a Staff Scientist at the Smithsonian Tropical Research

Institute in Panama. She is interested in predator-prey interactions, with a
focus on bat foraging behavior.

GUILLERMO PAZ-Y-MIÑO-C is based at the University of Massachusetts

at Dartmouth. He has broad interests in evolution, animal behavior, and con-
servation biology. His research focuses on evolutionary social cognition, the
364 ABOUT THE EDITOR AND CONTRIBUTORS

role of memory in kin recognition, and the application of behavioral para-

digms in conservation biology. He is also interested in science education, par-
ticularly the communication of evolutionary theory to the public and the
conservation of biological diversity.

S. CRAIG ROBERTS is a Senior Lecturer in the Department of Psychology

at the University of Stirling. His research interests are in evolutionary psychol-
ogy, behavioral ecology, and evolutionary medicine.

MICHAEL J. RYAN is Clark Hubbs Regents Professor in Zoology, Section of

Integrative Biology, University of Texas, and a Research Associate at the
Smithsonian Tropical Research Institute in Panama. His main interests are
in animal communication and sexual selection.

DAVID F. SHERRY is Professor of Psychology and Biology at the University

of Western Ontario and a cofounder of the Advanced Facility for Avian
Research. His research examines the evolution and neurobiology of memory
and spatial orientation.

CAROLYNN L. SMITH is an Associate Lecturer in the Department of Biol-

ogy at Macquarie University in Sydney, Australia. Her research focuses on
communication and cognition with an emphasis on multimodal signals.

CHARLES T. SNOWDON is Hilldale Professor of Psychology and Zoology

and Director of the Letters and Science Honors Program at the University of
Wisconsin. He has studied the behavior of cotton-topped tamarins and related
species for over 35 years.

GEORGE UETZ is a Professor of Biological Sciences at the University of

Cincinnati. He studies animal behavior and ecology, with the goal of under-
standing how the environment influences the evolution of behavior. He is
interested in questions about both the proximate mechanisms and selective
forces involved in communication and social behaviors of spiders.

CHRISTOPHER D. WATKINS is a Lecturer in the Division of Psychology

in the School of Social and Health Sciences at the University of Abertay Dun-
dee. His research focuses on how cues of dominance are perceived during
social interactions.
 About the Editorial Board

Daniel T. Blumstein, Department of Ecology and Evolutionary Biology,

University of California–Los Angeles, Los Angeles, California
Dorothy L. Cheney, Department of Biology, University of Pennsylvania,
Philadelphia, Pennsylvania
David L. Clark, Department of Biology, Alma College, Alma, Michigan
Ellen S. Davis, Department of Biological Sciences, University of Wisconsin–
Whitewater, Whitewater, Wisconsin
Bennett G. Galef Jr., Department of Psychology, Neuroscience, and
Behaviour, McMaster University, Hamilton, Ontario, Canada
James C. Ha, Department of Psychology, University of Washington, Seattle,
Washington
Kim L. Hoke, Department of Biology, Colorado State University, Ft. Collins,
Colorado
Robert L. Jeanne, Department of Entomology, University of Wisconsin–
Madison, Madison, Wisconsin
Patricia B. McConnell, Department of Zoology, University of Wisconsin–
Madison, Madison, Wisconsin
Douglas W. Mock, Department of Biology, University of Oklahoma,
Norman, Oklahoma
366 ABOUT THE EDITORIAL BOARD

Michael Noonan, Department of Biology, Canisius College, Buffalo, New

York
Sarah R. Partan, School of Cognitive Science, Hampshire College, Amherst,
Massachusetts
Robert M. Seyfarth, Department of Psychology, University of Pennsylvania,
Philadelphia, Pennsylvania
Zuleyma Tang-Martínez, Department of Biology, University of Missouri–St.
Louis, St. Louis, Missouri
 Index

Action potential, 280 American toads, 82, 84

Active space, 131 Amnesia, 278
Active time, 131 Amphibian papilla (AP), 142, 144
Adaptive suicide, 168–69 Amphipods, 326
Aerial predators, 191–93 Amplexus, 128–29
Affect-based signals, 180 Amur tigers, 296
Affiliative behavior, 224 Anadromous, 2
Aggression: plumage and, 113–14; Animal Behavior and Wildlife Conservation
siblicide, 163–65; song learning and, (Festa-Bianchet and Apollonio), 290
56–57; in tamarins, 230; tidbitting, 195. Animal Behavior Society, 289
See also Boldness-aggressiveness Animal cognition: as a behavioral unknown,
behavioral syndrome 290–94; definition of, 289; eliminating
Aggression drive, 2 interspecific sexual and social imprinting,
Agonistic behaviors, 298 297–98; enhancing social learning and
Akçay, Çağlar, 52 facilitation, 298; ensuring mate choice in
Alarm calls: acoustic structure of, 185; in captivity mimics mate choice in nature,
Belding’s ground squirrels, 183; 297; further research, 300–301, 304;
deception, 189; as not reflexive, 187; identifying and minimizing maladaptive
in prime-probe experiments, 188; behaviors, 299; identifying inappropriate
in ring-tailed lemurs, 180–81; in conservation strategies, 298–99; role in
tamarins, 233–34, 235; in vervet conservation, 294–99; terminology, 293;
monkeys, 176, 179 training animals to avoid predators,
Alberts, Jeffrey, 227 294–97
Alert signals, 295 “Animal Cognition and Its Role in
Altruism, 112, 317, 318 Conservation Behavior”
Amboseli National Park, 179 (Paz-y-Miño-C), 289
American Society of Primatologists, 217 Anterograde amnesia, 278
368 INDEX

Antilophia, 102 “Behavioral Unknowns: An Emerging

Antipredator behavior, 294–97 Challenge for Conservation”
Anuran acoustic processing, 142 (Paz-y-Miño-C), 289
Anuran communication, 142 “Behavior and Conservation in the
AP (Amphibian papilla), 142, 144 Galapagos” (Paz-y-Miño-C), 289
Assessment/management perspective, 199 Behaviorism, 197
Attenborough, David, 104 Behavior reading, 301, 302–3
Audience effect, 182–83 Behaviour and Conservation (Gosling and
Auditory adaptations: of frog-biting Sutherland), 290
midges, 145; of frog-eating bats, Belding’s ground squirrels, 183
144–45; introduction to, 141–42; Bergstrom, Carl, 190
of Túngara frogs, 142–44 Bet-hedging strategy, 54
Autonoetic memory, 283 Bewick’s wrens, 51
Aversion conditioning, 298 Big bang reproduction, 110
Avian brood parasitism: background, Biomedicine, 22
250–51; definition of, 249; Birds, food-storing: global reference system,
discrimination between own and foreign 274–76; hippocampal neurogenesis, 279;
eggs, 251–54; egg-manipulation hippocampus of, 278–79; introduction
experiments, 258–59; overview of, to, 270–71; landmarks, 272–74; memory
249–50; recognition responses of hosts of and behavior, 271; neurobiology of spatial
brown-headed cowbirds vs. common memory and orientation, 278–80
cuckoos, 254–58; summary, 259 Birds, siblicide in: adaptive suicide, 168–69;
Azimuth, 274 aggression, 163–65; ecological crunch,
159; family as center of cooperation and
Baboons, 178 conflict, 154–56; fitness consequences,
Background noise, 122 165–68; history of study of, 171;
Bakker, Theo, 313 introduction to, 153–54; nonaggressive
Balda, Russell, 275 forms of, 169–70; overt aggression,
Basilar papilla (BP), 142, 144 163–65; parental complicity, 170–71;
Bats, 124–25, 134–38, 144–45 parental control and manipulation of
Baugh, Alexander, 146 sibling symmetries, 160–62; parental
Baylis, Jeffrey, 197–98 overproduction, 156–59
Bay-winged cowbirds, 283 Birds of paradise, 99
Begging behavior, 169–70 Black-capped chickadees, 272–73, 275–76,
Behavioral Approaches to Conservation in the 278–79, 284
Wild (Clemmons and Buchholz), 290 Black eagles, 157, 166
Behavioral basis for mate choice, 313 Black stork, 161
Behavioral ecologists, work of, 105–9 Blue cranes, 300
Behavioral ecology, 23, 105, 165, 218 Blue-footed boobies, 167
Behavioral Ecology and Conservation Biology Blue tits, 300
(Caro), 290 Boldness-aggressiveness behavioral syndrome:
Behavioral strategy, 109 as adaptive under predation risk, 12–17;
Behavioral syndromes: as constraints over proximate mechanisms, 17–22;
developmental time, 4–9; as constraints in sticklebacks, 2–4
over evolutionary time, 9–12; Bonobos, 212
introduction to, 3–4; stability of, 4–9 Boobies, 167
Behavioral unknowns, 290–91 Borkent, Art, 139
INDEX 369

BP (Basilar papilla), 142, 144 facilitation, 298; ensuring mate choice in

Broadband, 135 captivity mimics mate choice in nature,
Broadband pulse, 192 297; further research, 300–301, 304;
Brood parasites, 155, 250, 282. See also Avian identifying and minimizing maladaptive
brood parasitism behaviors, 299; identifying inappropriate
Brood reduction, 153 conservation strategies, 298–99; role in
Brown fat, 227 conservation, 294–99; terminology,
Brown-headed cowbirds, 254–58, 282–83 293; training animals to avoid predators,
Brownlie, Lesley, 275 294–97
Brown pelicans, 169 Cognitive ethology, 294, 295
Brush-legged wolf spiders, 64 Cognitive legacies, 292, 293
Buckingham, Gavin, 320–21 Cognitive repertoire, 293
“Butterfly flight,” 104 Cognitive skills, 235–38, 302–3
Buzz weent, 114–15 Collective reasoning, 293
Coloration, 252, 255
Caches, 270–71 Combined Parasite Stress Index, 324
California condors, 294, 296, 298, 300 Commensal microflora, 328–29
Cannibalism, 159 Common garden experiments, 52
Capranica, Robert, 141–42 Common ravens, 294, 298
Captive breeding, 211, 213–15, 238 Communication: auditory adaptations,
Capuchin monkeys, 235 141–42; auditory adaptations of frog-
Caro, Tim, 290 biting midges, 145; auditory adaptations
Categorical perception, 130 of frog-eating bats, 144–45; auditory
Cattle egrets, 169 adaptations of Túngara frogs, 142–44;
Cephalothorax, 70 female túngara frog as intended receiver,
Chickadees, 270–71, 272–73, 129–34; frog-biting midges as unintended
275–76, 278–79, 284 receiver, 138–41; frog chorus as
Chickens, 190 communication, 124–25; frog-eating
Chimpanzees, 212, 213, 237–38, 298 bat as unintended receiver, 134–38;
Chinook salmon, 297 introduction to, 121–23; male túngara
Chiroxiphia, 102–3 frog as sender, 125–29; other eaves-
Chucks, 125–27, 131–34, 137, 140, 142, droppers, 145–47; song sparrows, 42–50;
143–44 summary, 147; between tamarins, 231–35;
Circadian rhythms, 275 unintended receiver-signaler interactions,
Cladistic evidence, 139 123–24. See also Referential signals
Clark, Dave, 69, 73, 78, 92 Communication, investigation of with video
Clark’s nutcrackers, 271–72 digitization and playback: communication
Clayton, Nicky, 283 and sexual selection in S. ocreata with
Clock shift, 275–76 video playback, 78–85; current video
Clutch learning, 253 playback studies, 85–90; development of
Coast range sculpins, 8 video playback as experimental technique,
Coefficient of relatedness, 155–56 76–78; discovery and new research
Coevolutionary arms race, 250 opportunities in M. Inclemens, 73–76;
Cognition, animal: as a behavioral unknown, ethnospecies in the genus Schizocosa,
290–94; definition of, 289; eliminating 64–67; introduction to, 63–64; Maevia
interspecific sexual and social imprinting, inclemens, 69–73; sexual selection in
297–98; enhancing social learning and S. ocreata, 67–69; summary, 90–92
370 INDEX

Communities, 291 Core members, 157, 168

Compass, 274–75 Cortisol, 17, 18–19, 223
Competition, in manakins, 113–15 Cotton-top tamarins. See Tamarins
Complex calls, 130–38, 140 Counterbalance, 39
Complexity in MHC-correlated choice, 330 Countersinging, 34
Complex social networks, 115–17 Courtship behavior: jumping spiders, 70,
Complex visual environments, and video 72–73; manakins, 103–5; video
playback studies, 85–87 digitization and playback, 73–76; wolf
Concealed ovulation, 219–20 spiders, 66–67. See also Schizocosa ocreata
Concept maps, 295 Courtship strategy, and cooperative
Condition dependence and mate choice: breeding, 230
evidence from humans, 315–16; Cousin marriages, 319
evidence from nonhuman animals, Couvade syndrome, 224
313–15; theory, 312–13 Cowbirds, 254–58, 282–83
Conditioning procedures, 294 Crabs, 146
Confusion effect, 193 Cronin, Katherine, 237–38
Consanguinity, 319 Crop milk, 222
Conservation and human behavior. See Cryptic species, 65, 183
Tamarins Crystallized song, 53–54, 54
Conservation behavior: animal cognition as a Cuckoos, 252–58, 256, 258–59
behavioral unknown, 290–94; animal
cognition’s role in, 294–99; definition of, Darwin, Charles, 105, 175, 178
289–90; eliminating interspecific sexual Dawkins, Richard, 199
and social imprinting, 297–98; enhancing “Dear enemy” relationship, 57; introduction
social learning and facilitation, 298; to, 34–35; playback experiment, 36–38;
ensuring mate choice in captivity mimics song in song sparrows, 35–36; song
mate choice in nature, 297; further sparrow neighbors as, 38–42
research, 300–301, 304; identifying and DeBruine, Lisa, 320–21, 324, 324–25
minimizing maladaptive behaviors, 299; Deception, in referential signals, 182,
identifying inappropriate conservation 188–90, 195, 197
strategies, 298–99; introduction to, Deceptive signaling, 50
289–90; summary, 304; training animals Decoy techniques, 297
to avoid predators, 294–97 Definitive plumage, 113
The Conservation Behaviorist (TCB), 289 Denterlein, Gerri, 67
Conservation biology, 295 Descartes, René, 176
Conservation Biology, 291 Desert ants, 276–77
Conspecific calls, 125, 129–30 Desertion, 160
Conspecifics, 3, 51, 121, 183 Developmental constraint, 9
Contingencies, 198 Developmental psychology, 8
“Contribution of Animal Behavior Diana monkeys, 188
Research to Conservation Biology” Dickinson, Anthony, 283
(Paz-y-Miño-C), 289 Diel shifts, 123
Convergent phenotypes, 3 Dilution effect, 183
Converging evolution, 212 Dimorphism, 70, 73
Cooperation, in manakins, 111–13 Diploid organisms, 154
Cooperative breeders, 210, 212, Disassortive mate preferences, 328
218, 225–30, 237–38 Discordancy, 253
INDEX 371

Discovery Park, 33 Eventual variety, 35

Discriminability, 184 Evolutionary biology, 22
Discrimination reversal, 301, 303 Experimental learning, 298
Disney’s Animal Kingdom, 217, 239 Extra-pair copulations (EPCs), 331
Dispersal and settlement, 291 Extra-pair mating, 34
Displays, 251 Extra-pair siring, 300
Distal landmarks, 272–73 Eye color, 322
Diurnal animals, 300 Eyewitness evidence, 106–7
Diverging evolution, 212
Dogs, 184, 197–98 Facial health, 315–16
Dominance, 106, 114–15, 298–99 Facial resemblance, 320–21
Dominance hierarchy, 68, 108–9, Facial symmetry, 327
163–64, 190 Facultative brood parasitism, 250
Dominant frequency, 125 Facultative siblicide, 166–68
Donation behavior, 238 False alarm calls, 189
Downregulation, 26 Family, 154–56, 160–62
Duff, Sarah, 275 Feral cats, 296–97
Field crickets, 326
Eagles, 179 Field observation, 107–8, 217–18
Eastern phoebes, 255, 257–58 Fincher and Thornhill’s Combined Parasite
Eavesdropping, 40–42, 88–90, 122–24, 139, Stress Index, 324
142, 145–47, 189 Fitness, 17, 101, 155, 168–69
Ecological speciation, 2 Florida scrub-jays, 300
Ecology, 22 Fluctuating asymmetry, 79–81
Ecosystems, 291 Food calls, 179–83, 185, 187,
Ectothermy, 159 193–97, 232–33
Egg discrimination, 251–54 Food sharing, 236–38
Egg manipulation experiments, 258–59 Food-sharing calls, 237
Egg rejection, 250 Food shortages, 160–61
Elasticity analysis, 110 Food-storing birds: global reference system,
Elephants, 296 274–76; hippocampal neurogenesis, 279;
Endangered species, 210–11 hippocampus of, 278–79; introduction
Endothermy, 159 to, 270–71; landmarks, 272–74; memory
Energetic basis for mate choice, 312–13 and behavior, 271; neurobiology of spatial
Entorhinal cortex, 280 memory and orientation, 278–80
Environmental unknowns, 290 Food webs, 116
Environment and mate choice: evidence from Foraging, 34
humans and nonhuman animals, 323–27; Foraging behavior, 296
theory, 322–23 Foraging/feeding, 291–92
EPCs (Extra-pair copulations), 331 Foraging skills, 240
Episodic/episodiclike memory, 283 Fork-tailed drongos, 189
Epistatic interactions, 24 Fowl: referential alarm signals for aerial
Estrogen, 227 predators, 191–93; referential alarm
Ethologists, 1 signals for ground predators, 191;
Ethospecies, 65 referential food signals, 193–97;
Eurasian otters, 296 referential signals in, 190–97
European ungulates, 300 French, Jeffrey, 216
372 INDEX

Freshwater fish, 2–3 Havlicek, Jan, 332

Frog chorus: auditory adaptations, 141–42; Head bobs, 193
communication, 124–25; female túngara Head direction cells, 279–80
frog as intended receiver, 129–34; Hebets, Eileen, 81, 91
frog-biting midges as unintended receiver, Helpers, in cooperative breeding, 226–31
138–41; frog-eating bat as unintended Heritable differences, 23
receiver, 134–38; male túngara frog as Heritable variation, 299
sender, 125–29 Herz, Rachel, 272
Frugivory, 101–2 Heterospecific calls, 125, 129–30
Functionally referential signals, 179–80, 181, Heterospecifics, 183
186–88 Heterospecific songs, 51
Fundamental frequency, 125, 316 Heterozygote advantage, 328
Heterozygotes, 328
Gallistel, Charles, 188 Hierarchical signaling hypothesis, 49
Garver-Apgar, Christine, 334 Hinde, Robert, 199
Gaulin, Steve, 281, 318 Hippocampal neurogenesis, 279
Genetic basis for mate choice, 312 Hippocampus of food-storing birds, 278–79,
Genetics, 24 281, 282–83
Genome expression analysis, 24–27 Hoke, Kim, 143
Genomics, 22 Home ranges, 269, 271
Genus Schizocosa, 64–67 Homing pigeons, 274
Gerhardt, Carl, 123, 142 Homology, 24
Gestures, 180 Homozygotes, 328
Glucocorticoids, 223–24 Hormones, 215–16, 222–25, 227
“Good-genes-as-heterozygosity Horn size, 299
hypothesis,” 332 Host-based recognition systems
Graded signaling system, 45 approach, 251
Grafen, Alan, 23 Huber, Mark, 123
Graph theory, 116 Huddle, 159
Great blue herons, 164 Human behavior, study of tamarins
Great egrets, 8, 163, 168 and, 240–41
Great reed warbler, 253–54, 256 Human mate preferences, potentially
Grid cells, 279–80 adaptive individual differences: condition
Grooming behavior, 220, 229–30 dependence and mate choice, 312–16;
Ground predators, 191, 198 environment and mate choice, 322–27;
Guppies, 104, 314, 326 introduction to, 311–12; kinship and
mate choice, 317–22; MHC and mate
Habituation, 121, 199, 199–200, 215 choice, 327–34; summary, 334–35
Hair color, 322 Huntingford, Felicity, 3, 4
Haldane, J. B. S., 154
Hampshire College, 92 Illuminance spectrophotometry, 85
Haploid organisms, 154 Immediate early genes (IEGs), 143
Harland, Duane, 91 Immediate variety, 35
Harlow, Harry, 225 Imprinting, 294, 318, 321–22
Harmonics, 125 Inbreeding, 219, 319
Hatching asynchrony, 161–62 Inclusive fitness, 154, 171, 317
Hauser, Marc, 234 Incremental forgetting, 121–22
INDEX 373

Indiana University, 227 Lee, Anthony, 325

Indirect fitness, 165 Lek mating system, 99–100, 102,
Infant attachment, 225–26 112, 114–15
Infanticide, 170 Leks, 323
Infant mortality, 221–22 Leopards, 179
Information centrality, 116 Lieberman, Debra, 319–20
Insight learning, 301, 303 Linear algebra, 116
Intersexual selection, 67–68 Little, Anthony, 315, 316, 324–25
Interspecific brood parasitism, 250 Localization, 135–36
Interspecific sexual and social imprinting, Local landmarks, 271, 272–73, 275
297–98 Loggerhead sea turtles, 296
Intrasexual selection, 67–68 Long bobs, 193
Intraspecific brood parasitism, 250 Long call, 234
López, Silvia, 314
Jackson, Robert, 91 Lorenz, Konrad, 1–2
Jakob, Beth, 91 Luteinizing hormone, 227
Japanese medaka, 326
Jones, Benedict, 315–16, 321, 324, 324–25 Maculation, 252
Jumping spiders, 69–73 Maevia inclemens: discovery and new research
opportunities in, 73–76; two distinct male
Kaminski, Gwenaël, 320 morphs, 69–73; video digitization and
Keddy-Hector, Anne, 122 playback with, 73–76; video playback, 77
Kin-bias dominance interactions, 317 Magpies, 284
Kin selection, 105, 112, 230 Major histocompatibility complex (MHC):
Kinship and mate choice: evidence from Chinook salmon, 297; evidence from
humans, 318–22; evidence from humans, 332–34; evidence from
nonhuman animals, 317–18; theory, 317 nonhuman animals, 329–32; mate choice,
Kittiwakes, 167 327–29; overview of, 311–12; theory,
Kolluru, Gita, 123 327–29. See also MHC-dependent mate
Kookaburras, 164 choice; MHC-disassortative mate
Kori bustards, 300 preferences
Krebs, John, 199, 271 Maladaptive behaviors, 296, 297–98,
Kunzler, Reto, 313 299, 300
Male-male lifelong partnerships in tropical
Laboratory rats, 279–80 birds: competition, 113–15; complex
Laboratory vs. field studies, 217–18 social networks, 115–17; cooperation,
Lachmann, Michael, 190 111–13; courtship display, 103–5;
Lactation, 214 distribution and taxonomy, 102–3;
Landmarks, 272–74, 276 frugivory, 101–2; introduction to,
Language, 175–76, 240 99–101; mating, 109–11; natural
Larnyx morphology, 125–27 history, 101–5; summary, 117–18
Lateral plates, 3 Male parental care, 212, 214,
Lava lizards, 297 218–19, 222–25
Lead poisoning, 296 Manakins: competition, 113–15; complex
Learning preference, 54 social networks, 115–17; cooperation,
Learning-set formation, 301, 302 111–13; courtship display, 103–5;
Learning-theory principles, 294 distribution and taxonomy, 102–3;
374 INDEX

frugivory, 101–2; introduction to, Mimetic eggs, 250, 254, 257

99–101; mating, 109–11; natural history, Mobbing, 182
101–5; summary, 117–18 Mobbing calls, 232
Mannion, Helena, 316 Modern learning theory, 198
Marginals, 157–59, 161, 168 Monoamines, 17, 18–19
Marler, Peter, 52, 53 Monogamy, 34, 281, 297
Marmosets, 209–10, 238 Morjan, Carrie, 78
Marmots, 181 Morphology, 2, 99
Marsh wrens, 51 Morton, Eugene, 199
Masculine voices, 316 Moskát, Csaba, 253, 256–57, 258
Masculinity, and health, 323–27 MPA (Maternal perinatal association),
Mate-attraction signals, 295 319–20, 321
Mate choice, 298–99 Müller, Martin, 277
Mate competition, 88 Multimodal signals, 68–69, 183–84, 193
Mate fidelity, 225 Multiple messages, 184
Maternal perinatal association (MPA), Music, 234–35
319–20, 321 Myers, Norman, 290
Maternal styles, 226
Mate selection, 297 Natural selection, 101, 279, 299
Mating, manakins, 109–11 Nature Conservancy, 239
Mating calls, 125–29, 130–31. See also Navigation. See Orientation and navigation
Túngara frogs Nazca boobies, 167
Mating success, 101 Nelson, Douglas, 53
Matrilateral kin, 318 Nepotism, 317
Mazzi, Dominique, 313 Nettle, Daniel, 326
McClintock, Will, 81, 82 Neurobiology of spatial memory and
McConnell, Patricia, 197–98 orientation, 278–80
McDermott, Josh, 234 Neuroendocrine mechanisms, 17
McKeever, Sturgis, 139 Neurogenesis, 279
Meadow voles, 281 Nice, Margaret, 33
Meerkats, 189 Nobel Prize, 1
Meiosis, 154 Nocturnal animals, 300
Metapopulations, 304 Noise, in communication, 121–22
Mexican tetra, 142 Noise pollution, 296
MHC (Major histocompatibility complex): Nominal representation, 188
Chinook salmon, 297; evidence from Noninvasive hormone assays, 215–16, 238
humans, 332–34; evidence from Noninvasive research methods, 215–16
nonhuman animals, 329–32; mate choice, Nonredundant messages, 184
327–29; overview of, 311–12; theory, Nordby, Cully, 52
327–29 Norepinephrine, 18–20
MHC-dependent mate choice, 330–31, 332 Novel males, 228
MHC-disassortative mate Numerical concept, 301, 302
preferences, 328–34 Numerosity, 302
Microarrays, 24, 25 Nuptial coloration, 2
Midges, 124–25, 138–41, 145
Migratory birds, 274 Object permanence, 301, 302
Milinski, Manfred, 313 Obligate brood parasitism, 250
INDEX 375

Obligate siblicide, 166 Pedipalps, 70

Observational learning, 298 Perrett, David, 321–22
Odor preferences, 332–34 Persons, Matt, 83–84
Odors in communication, 235 Peters, Susan, 52
Ohio University, 65 Phenotypic evolution, 2
One-horned rhinoceros, 299 Phenotypic gambit, 23–24
On the Expression of the Emotions in Man and Philopatric species, 167
the Animals (Darwin), 175 Phonotaxis, 125
Operational sex ratios, 326 Physiological zero, 161
Opossums, 146 Physiology, 22
Optimal mate choice, 317–18, 319 Pigs, 164–65
Optimal outbreeding theory, 317 Pine voles, 281
Oral contraceptives, 334 Place cells, 279–80
Orangutans, 298 Plasticity of discrimination threshold, 257
Orientation and navigation: food-storing Plastic song, 53–54, 54
birds, 270–76; introduction to, 269–70; Playback experiments: dear enemy
methods of, 270; navigation by path relationships in song sparrows, 41; mate
integration, 276–77; neurobiology of choice in captivity, 297; overview of,
spatial memory and orientation, 278–80; 36–38; referential alarm signals for
sex differences in spatial ability, 280–83; ground predators, 191; referential food
summary, 284–85; tagging spatial signals, 194–95; song repertoire, 55; song
locations with other information, 283–84 sparrow code, 45; structural variation,
Overproduction, parental, 156–59 200; tamarin communication, 231; on
Owings, Daniel, 199 vervet monkeys, 186
Oxytocin, 220–21, 222 Pleiotropy, 23
Plumage, 113–14
Pacific salmon, 110 POC (Parent-offspring conflict), 155
Pair bonds, 219, 220 Polarized light, 277
Parasites, and mate choice, 314 Pollet, Thomas, 326
Parasitic castration, 65 Pollination, 291–92
Parasitoids, 123 Polyandry, 297
Parental care: in captivity, 297; experience, Polygyny, 99, 281, 297
239; learning, 221–26; in orangutans, Polymorphism, 328
298; in song sparrows, 34; sticklebacks, 2. Polyphenism, 70
See also Male parental care Potts, Wayne, 112
Parental complicity in siblicide, 170–71 Prairie voles, 281
Parental control, of sibling Predation pressure, 3
asymmetries, 160–62 Predation risk, 13
Parental fitness, 157–59 Predefinitive plumage, 113–14
Parental investment, 156 Premating reproductive isolation, 67
Parental overproduction, 156–59 Prey-cue/prey-quality associations, 137–38
Parent-offspring conflict (POC), 155 Prime-probe experiments, 188
Partan, Sarah, 91–92 Primer for Conservation Behavior (Blumstein
Paternal care, 101 and Fernándex-Juricic), 290
Path integration, 276–77 Prolactin, 222–23, 224
Patrilateral kin, 318 Promiscuity, 297
Peanuts, 283–84 Prosocial decision making, 318
376 INDEX

Prosoma, 70 Reproductive inhibition, 229

Prothonotary warblers, 255 Resource scarcity, 325–26
Proximate mechanisms: behavioral syndrome, Resource-tracking strategy, 158
12; in song sparrows, 35; underlying Retrograde amnesia, 278
boldness-aggressiveness behavioral Rhesus macaque, 225
syndrome, 17–22 Rhinoceros, 299
“Public goods game,” 318–19 Ring-tailed lemurs, 180–81
Public information, 135 Risk management tactics, 182–83
Pukeko, 294, 296 Roberts, Andy, 82
Pulsatile vocalizations, 193 Rovner, Jerry, 65
Puppet rearing, 297–98 Rowell, Thelma, 211, 213
Putty-nosed monkeys, 190 Rushdie, Salman, 269
Pygmy marmoset, 210–11 Ryan, Michael, 122, 143, 144, 146

Quantitative trait loci (QTL) mapping, 24 Sangster, Mark, 275

QuickTime movie format, 76 Santschi, Felix, 277
Savage, Anne, 217, 239
Radiation, 3, 9 Savannah sparrows, 331
Radical behaviorism, 197–99 Scheffer, Sonja, 67–68
Radical structuralism, 197–98 Schizocosa ocreata and rovneri:
Radioactive isotopes, 271 communication and sexual selection in,
Rand, Stanley, 146 67–69; communication and sexual
Rats, 284 selection in with video playback, 78–85;
Recognition responses, 254–58 complex visual environments, 85–87;
Recognition systems, 251–54, 255–56 introduction to, 64–67; social
Recognition templates, 250, 253 contexts, 87–90
Red-spotted newts, 326 Schlegel, Alice, 318
Redundant signals, 183–84 Schwartz, Joshua, 142
Referential alarm signals, 180–82 Screaming cowbirds, 283
Referential signals: for aerial predators, Scrub jays, 275, 283–84, 300
191–93; deception, 182, 188–90, 195, Seismic communication, 68–69
197; evidence for, 178–80; as evidence of Selective attrition, 53
cognitive processing and information, Self-resemblance, 318–19, 320–21
187–88; factors in, 180–82; in fowl, Self-song, 46
190–97; for ground predators, 191; Semelparity, 110
habituation, 199–200; identification of, Sensitive period, 51
186–87; as insight or illusion, 175–78; Sensitivity analysis, 110
radial behaviorism, 197–99; referential Sensory biases, 104
food signals, 193–97; signal structure, Separation calls, 220
182–85; species demonstrating, 176–78; Sequencing, 24
as window into animal minds, 187–90 Serotonin, 20–21
Reintroduction, 239–40, 298 Sex differences in spatial ability, 280–83
Rendall, Drew, 197–98 Sexual behavior of tamarins, 219–20
Repertoire matches, 43, 45–46, 47 Sexual dimorphism, 235
Representational signals, 188, 195 Sexual imprinting, 318
Reproductive behavior, 2, 64–67, 291 Sexual selection: communication and,
Reproductive failure, 227 125–29, 130–31; and communication in
INDEX 377

S. ocreata, 67–69; frugivory, 101; Social contexts, and video playback

inappropriate conservation strategies, studies, 87–90
298–99; physical traits, 312 Social dynamics, as shaper of communication,
Sexual signals, 122, 123 33–57; communication of song sparrows,
Seyfarth, Robert, 176, 199 42–50; dear enemies relationship with
Sherman, Paul, 183 territorial neighbors, 34–42; introduction
Shiny cowbirds, 282–83 to, 33–34; playback experiment, 36–38;
Shizuka, Dai, 108–9 remaining questions (wording ??), 55–57;
Short bobs, 193 song communication system and process
Siblicide in birds: adaptive suicide, 168–69; of song learning, 50–55; song in song
aggression, 163–65; ecological crunch, sparrows, 35–36; song sparrow code,
159; family as center of cooperation and 45–50; song sparrow neighbors as dear
conflict, 154–56; fitness consequences, enemies, 38–42; summary, 57
165–68; history of study of, 171; Social facilitation, 294, 298
introduction to, 153–54; nonaggressive Social imprinting, 294
forms of, 169–70; overt aggression, Social learning, 235–38, 294; enhancement
163–65; parental complicity, 170–71; of, 298
parental control and manipulation of Social network analysis, 106, 115–17
sibling symmetries, 160–62; parental Social networks: manakins, 115–17
overproduction, 156–59 Social organization, 291
Sibling asymmetries, parental control and Sociobiology (Wilson), 105
manipulation of, 160–62 Sociosexuality, 324
Signal evolution, 200 Soft song, 37, 47, 50
Signal matching, 89 Song birds, 165
Signals, 50 Song communication system, 50–55
Signals, referential: for aerial predators, Song learning in song sparrows: introduction
191–93; deception, 182, 188–90, 195, to, 50–51; remaining questions, 55–57;
197; evidence for, 178–80; as evidence of song communication system and process
cognitive processing and information, of song learning, 50–55; summary, 57;
187–88; factors in, 180–82; in fowl, and territory establishment, 53; in western
190–97; for ground predators, 191; song sparrows, 51–55. See also Song
habituation, 199–200; identification of, sparrows
186–87; as insight or illusion, 175–78; Song repertoires, 35–36, 52, 54, 56, 103
radial behaviorism, 197–99; referential Song sparrows: code, 45–50; communication,
food signals, 193–97; signal structure, 42–50; neighbors as dear enemies, 38–42;
182–85; species demonstrating, parental care in, 34; proximate
176–78; as window into animal mechanisms in, 35; song in, 35–36. See
minds, 187–90 also Song learning in song sparrows
Signal structure, 182–85 Song tutors, 36, 52–57
Sign stimuli, 2 Song types, 35–36
Singing conventions, 43, 45 Sound spectrogram, 186, 231
Site fidelity, 112 South American bullfrogs, 146
“Six degrees of separation,” 117 Spadefoot toads, 314
Skinner, B. F., 197 Spatial ability, sex differences in, 280–83
Slud, Paul, 104 Spatial locations, tagging with other
Smith, Liz, 81 information, 283–84
Snakes, 179 Spatial memory, 271, 278–80, 301, 303
378 INDEX

Spatial orientation, 278–80 relationship formation and maintenance,

Species interactions, 291–92 218–21; study of, 210–13; successful
Spectral analysis, 85 breeding, 213–15; summary,
Spiders, 146 238–41; understanding human
Stability of behavior, 4–6 behavior, 240–41
Stamps, Judy, 5–6 T. cirrhosus, 134–38, 144–45
Stevens, Allen, 271 Teaching, 236–38
Sticklebacks: behavioral syndromes as Teie, David, 234–35
constraints over developmental time, 4–9; Terrestrial predators, 191, 198
behavioral syndromes as constraints over Territoriality: communication in song
evolutionary time, 9–12; boldness- sparrows, 42–50; dear enemy
aggressiveness behavioral syndrome, 2–4; relationships in song sparrows,
boldness-aggressiveness syndrome as 34, 37–42
adaptation under predation risk, 12–17; Territorial neighbors, relationships
condition dependence and mate choice, with, 34–42
313–15; environment and mate choice, Territory, 269
322–23; genomics and animal behavior, Territory defense, 295
22; introduction to, 1; overview of, 1–3; Territory establishment, and song learning,
phenotypic gambit, 23–24; proximate 53, 56
mechanisms underlying boldness- Testosterone, 223
aggressiveness behavioral syndrome, Theory of mind, 188
17–22; whole-genome expression, 24–27 Threespine sticklebacks. See Sticklebacks
Stoddard, Philip, 38 Tidbitting, 193–95
Stratton, Gail, 67 Tigers, 296
Stress, 227 Tinbergen, Niko, 1, 284, 291
Stress responsiveness, 18, 20 Tits, 270–71
Struhsaker, Thomas, 176 Tool use, 301, 302
Sun compass, 274–75, 275–76 Trade-offs, 82, 169, 181
Sun-compass orientation, 274, 275–76, 277 Traditions, 292
Swamp sparrows, 51 Transcriptome sequencing, 25
Swordtail fish, 142 Transitive reasoning, 301, 303
Symmetry, preferences for, 315, 327 Triad census, 109
Sympatricity, 129–30 Trials of Life series, 104
Sympatric species, 181 Trivers, Robert, 155
Syntax, 231 “Trust game,” 318
Székely, Tamás, 257 Tulving, Endel, 283
Túngara frogs: auditory adaptations of,
Tactical deception, 301, 303 142–44; female túngara frog as intended
Takahe, 296 receiver, 129–34; frog-biting midges as
Tamarins: cognitive skills, 235–38; unintended receiver, 138–41; frog-eating
communication, 231–35; conservation, bat as unintended receiver, 134–38; male
238–40; description of, 209–10; findings, túngara frog as sender, 125–29; sensory
218–38; helper motivation, 230–31; bias, 104
helper recruitment and maintenance, Tuttle, Merlin, 146
226–30; laboratory vs. field studies, Type I error, 181
217–18; noninvasive research methods, Type II error, 181
215–16; parental development, 221–26; Type matches, 43, 45–46, 47–48, 50
INDEX 379

Uetz, George, 63, 64, 82 Vocal repertoire, 103, 212–13, 231

Uganda kob, 323 Von Frisch, Karl, 1–2, 176
Ultimate mechanism: behavioral syndrome,
12; in song sparrows, 35 Waist-to-hip ratio, 315–16
Undersignalers, 49–50 Watkins, Christopher, 321, 322
Unintended receivers, 122 Waxworms, 283–84
Unintended receiver-signaler Weber’s law, 133, 137
interactions, 123–24 Wedekind, Claus, 332–34
University of Canterbury, 91 Wehner, Rudiger, 277
University of Cincinnati, 64 Westermarck effect, 319
University of Massachusetts, 91 Wheelwright, Nat, 99, 101
University of Nebraska, 91 White, David, 282
Upregulation, 26 White-crowned sparrows, 51
Urgency-base alarm signaling, 181, 193 White rhinoceros, 299
U.S. Centers for Disease Control, 324 White storks, 300
Whole-genome expression,
Validity, 216 24–27
Vander Wall, Stephen, 272, 273–74 Wilczynski, Walter, 143
Vectors, 273, 276 Wild lion tamarins, 237
Vervet monkeys, 176, 179 Wilson, Edward O., 105
Video playback: communication and sexual Wiltschko, Wolf, 275
selection in S. ocreata with video playback, Wing-waving, 37, 50
78–85; complex visual environments, Wisconsin National Primate
85–87; courtship behavior and, 73–76; Center, 210, 216
current studies, 85–90; development of as Wolf, Arthur, 320
experimental technique, 76–78; Wolf spiders, 64–67
identification of functionally referential Wood storks, 160
signals, 186–87; social contexts, 87–90; World Health Organization, 324
tidbitting, 195. See also Communication,
investigation of with video digitization Zanette, Liana, 272
and playback Zebra finches, 314–15, 316
Virtual evolution, 77–78, 82 Ziegler, Toni, 216, 220
Visual cues, 296 Zietsch, Brendan, 325
Visual perspective taking, 301, 302 Zuk, Marlene, 123