The Biology Of Bloodsucking In Insects 2ed

Lehane Mj download

https://ebookbell.com/product/the-biology-of-bloodsucking-in-
insects-2ed-lehane-mj-2045336

Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.

The Biology Of Bloodsucking In Insects Second Edition 2nd Edition M J

Lehane

https://ebookbell.com/product/the-biology-of-bloodsucking-in-insects-
second-edition-2nd-edition-m-j-lehane-11081348

The Biology Of Beauty The Science Behind Human Attractiveness Rachelle

M Smith

https://ebookbell.com/product/the-biology-of-beauty-the-science-
behind-human-attractiveness-rachelle-m-smith-46474544

The Biology Of Science Fiction Cinema Mark C Glassy

https://ebookbell.com/product/the-biology-of-science-fiction-cinema-
mark-c-glassy-49052494

The Biology Of Sharks And Rays A Peter Klimley

https://ebookbell.com/product/the-biology-of-sharks-and-rays-a-peter-
klimley-50416384
The Biology Of Arid Soils

https://ebookbell.com/product/the-biology-of-arid-soils-50929506

The Biology Of Sharks And Rays A Peter Klimley Steven Oerding

https://ebookbell.com/product/the-biology-of-sharks-and-rays-a-peter-
klimley-steven-oerding-51442170

The Biology Of Chameleons Krystal A Tolley Editor Anthony Herrel

Editor

https://ebookbell.com/product/the-biology-of-chameleons-krystal-a-
tolley-editor-anthony-herrel-editor-51819814

The Biology Of Race Rev Ed Reprint 2020 James C King

https://ebookbell.com/product/the-biology-of-race-rev-ed-
reprint-2020-james-c-king-51821478

The Biology Of The Cycads Knut Norstog Trevor J Nicholls

https://ebookbell.com/product/the-biology-of-the-cycads-knut-norstog-
trevor-j-nicholls-51938546
This page intentionally left blank
The Biology of Blood-Sucking in Insects
Second Edition

Blood-sucking insects transmit many of the most debilitating dis-

eases in humans, including malaria, sleeping sickness, filaria-
sis, leishmaniasis, dengue, typhus and plague. In addition, these
insects cause major economic losses in agriculture both by direct
damage to livestock and as a result of the veterinary diseases, such
as the various trypanosomiases, that they transmit. The second
edition of The Biology of Blood-Sucking in Insects is a unique, topic-
led commentary on the biological themes that are common in the
lives of blood-sucking insects. To do this effectively it concentrates
on those aspects of the biology of these fascinating insects that
have been clearly modified in some way to suit the blood-sucking
habit. The book opens with a brief outline of the medical, social and
economic impact of blood-sucking insects. Further chapters cover
the evolution of the blood-sucking habit, feeding preferences, host
location, the ingestion of blood and the various physiological adap-
tations for dealing with the blood meal. Discussions on host–insect
interactions and the transmission of parasites by blood-sucking
insects are followed by the final chapter, which is designed as a use-
ful quick-reference section covering the different groups of insects
referred to in the text.
For this second edition, The Biology of Blood-Sucking in Insects has
been fully updated since the first edition was published in 1991. It
is written in a clear, concise fashion and is well illustrated through-
out with a variety of specially prepared line illustrations and pho-
tographs. The text provides a summary of knowledge about this
important group of insects and will be of interest to advanced
undergraduate and to postgraduate students in medical and vet-
erinary parasitology and entomology.

M i k e L e h a n e is Professor of Molecular Entomology and

Parasitology in the Liverpool School of Tropical Medicine.
The Biology of
Blood-Sucking in Insects
SECOND EDITION

M. J. Lehane
Liverpool School of Tropical Medicine
  
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University Press

The Edinburgh Building, Cambridge  , UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521836081

© M. J. Lehane 2005

This book is in copyright. Subject to statutory exception and to the provision of

relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
First published in print format 2005

- ---- eBook (MyiLibrary)

- --- eBook (MyiLibrary)

- ---- hardback

- --- hardback

- ---- paperback

- --- paperback

Cambridge University Press has no responsibility for the persistence or accuracy of

s for external or third-party internet websites referred to in this book, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
 Contents

List of tables page vii

List of boxes x
Preface xi
Acknowledgements xiii

1 The importance of blood-sucking insects 1

2 The evolution of the blood-sucking habit 7
2.1 Prolonged close association with vertebrates 7
2.2 Morphological pre-adaptation for piercing 13
3 Feeding preferences of blood-sucking insects 15
3.1 Host choice 15
3.2 Host choice and species complexes 24
4 Location of the host 27
4.1 A behavioural framework for host location 27
4.2 Appetitive searching 29
4.3 Activation and orientation 32
4.4 Attraction 49
4.5 Movement between hosts 52
5 Ingestion of the blood meal 56
5.1 Probing stimulants 56
5.2 Mouthparts 57
5.3 Vertebrate haemostasis 64
5.4 Host pain 68
5.5 Insect anti-haemostatic and anti-pain factors
in saliva 69
5.6 Phagostimulants 76
5.7 Blood intake 78
6 Managing the blood meal 84
6.1 Midgut anatomy 84
6.2 The blood meal 87
6.3 Gonotrophic concordance 96
vi Contents

6.4 Nutrition 98
6.5 Host hormones in the blood meal 103
6.6 Partitioning of resources from the blood meal 106
6.7 Autogeny 109
7 Host–insect interactions 116
7.1 Insect distribution on the surface of the host 117
7.2 Morphological specializations for life on the host 121
7.3 Host immune responses and insect salivary
secretions 126
7.4 Behavioural defences of the host 134
7.5 Density-dependent effects on feeding success 142
8 Transmission of parasites by blood-sucking insects 150
8.1 Transmission routes 150
8.2 Specificity in vector–parasite relationships 163
8.3 Origin of vector–parasite relationships 167
8.4 Parasite strategies for contacting a vector 170
8.5 Parasite strategies for contacting a vertebrate host 177
8.6 Vector pathology caused by parasites 179
8.7 Vector immune mechanisms 184
9 The blood-sucking insect groups 202
9.1 Insect classification 202
9.2 Phthiraptera 204
9.3 Hemiptera 208
9.4 Siphonaptera 213
9.5 Diptera 219
9.6 Other groups 257

References 259
Index 312
 Tables

1.1 An outline of the early investigations that laid the

foundations of medical and veterinary entomology. page 2
1.2 Rounded estimates for the prevalence of disease, the
number at risk and the disability adjusted life years
(DALYs) for major vector-borne diseases. 3
1.3 Estimated losses in agricultural production caused by
blood-sucking insects. 4
4.1 Generalized opportunities and constraints on host location
by blood-sucking insects feeding during the day or night. 32
4.2 Different blood-sucking insects respond in different ways
to spectral information. 45
5.1 Adaptations of mouthpart components for different
purposes in various haematophagous insect groups. 58
5.2 Blood-sucking insects produce a wide range of
anti-haemostatic factors in their salivary secretions. This
table gives some examples with a range of different
activities. 71
5.3 The size of red corpuscles varies widely in different
animals. Given that many blood-sucking insects have
mouthparts with a terminal diameter of around 10µm, this
may be a factor affecting the feeding efficiency of
blood-sucking insects on different host species. 80
6.1 The size of the red blood meal and the time taken in its
digestion are affected by a range of factors including
ambient temperature, age of the insect, mating status,
stage of the gonotrophic cycle, previous feeding history,
and source of the blood meal. The figures given here are a
rough guideline to the ‘average’ meal size and time for
digestion in a variety of haematophagous insects. 88
6.2 The major constituents of the blood are reasonably
uniform in most host animals. The exception is the high
levels of nucleic acids in the blood of birds and reptiles
because of their nucleated red blood cells. Proteins are far
and away the most abundant nutrients in blood, and
viii List of tables

nutrients are unevenly distributed between whole blood

(B), red blood cells alone (E) and plasma alone (P). 90
6.3 Symbionts are common in insects relying on blood as the
sole food source throughout their lives. An outline is given
of their anatomical locations and the means of transmission
from one generation to the next in different insect groups. 99
6.4 Three types of female Aedes taeniorhynchus have been
identified in terms of egg development: autogenous
females (1); females that are autogenous if mated (2); and
anautogenous forms (3). This pattern is influenced by the
feeding success of the larval stage, as illustrated in this
table. 113
6.5 Some mosquitoes can use sugar meals (10% sucrose in this
experiment) to increase the number of autogenously
produced eggs. 114
7.1 The choice of feeding site of Aedes triseriatus on eastern
chipmunks and grey squirrels is influenced by length and
density of body hair. The different feeding patterns on the
two hosts reflects the differences in hair cover between
them. 119
7.2 The anti-mosquito behaviour of a range of ciconiiform
birds, showing that different host species display various
types and degrees of defensive behaviour against
blood-sucking insects. 137
8.1 Some of the most important associations of
disease-causing organisms carried to humans and other
animals by blood-sucking insects: (a) viruses, (b) rickettsia
and bacteria, (c) protozoa and (d) nematodes. 151
8.2 Blood-sucking insects commonly take meals that are only a
small proportion of the total blood present in the host
animal (the ratio between total blood in the host and size
of the insect’s blood meal is given). This minimizes the
chances of the insect ingesting any individual parasite
during feeding. One strategy adopted by insect-borne
parasites to overcome this problem is to produce large
numbers of infective stages which circulate in the blood of
the host. 171
8.3 The microfilariae of many filarial worms display a
pronounced periodicity, with microfilarial numbers in the
peripheral blood coinciding with the peak biting time of
locally abundant vector species. 173
 List of tables ix

8.4 Tsetse flies infected with trypanosomes feed more readily

and probe more often than uninfected flies, thereby
increasing the chances of parasite transmission. 178
8.5 Comparison of the rate of formation of the peritrophic
matrix among various mosquito species. 189
8.6 The melanization response to subsequent challenge of
infected and uninfected Aedes aegypti, as shown by the
intrathoracic injection of specific microfilariae (mff) which
normally induce a strong melanization reaction. 200
9.1 The groups of insect. Those groups containing
blood-sucking insects are shown in bold. 203
9.2 The geographical distribution of triatomine species, which
have become highly adapted to the domestic-peridomestic
environment of man and so represent a particular threat as
vectors of Chagas’ disease. 211
9.3 The divisions of the order Diptera and the major families
in each division. Families containing blood-sucking species
are in bold type. 221
 Boxes

3.1 The importance of rates of mosquitoes biting humans for

the transmission of malaria. page 20
3.2 Identification of the source of a blood meal. 21
7.1 Histopathology of the various stages in the sequence of
host response to insect bites. 130
8.1 Four blood cell types characterized in Aedes aegypti are
compared to haemocytes described in previous studies on
a variety of insects. 197
 Preface

Blood-sucking insects are the vectors of many of the most debilitating par-
asites of humans and their domesticated animals. In addition they are of
considerable direct cost to the agricultural industry through losses in milk
and meat yields, and through damage to hides, wool and other products.
So, not surprisingly, many books of medical and veterinary entomology
have been written. Most of these texts are organized taxonomically, giving
details of the life cycles, bionomics, relationships to disease and economic
importance of each of the insect groups in turn. I have taken a different
approach. This book is topic-led and aims to discuss the biological themes
common to the lives of blood-sucking insects. To do this I have concen-
trated on those aspects of the biology of these fascinating insects that have
been clearly modified in some way to suit the blood-sucking habit. For
example, I have discussed feeding and digestion in some detail because
feeding on blood presents insects with special problems, but I have not
discussed respiration because it is not affected in any particular way by
haematophagy. To reflect this better I have made a slight adjustment to
the title of the book in this second edition. Naturally there is a subjective
element in the choice of topics for discussion and the weight given to each.
I hope that I have not let my enthusiasm for the particular subjects get the
better of me on too many occasions and that the subject material achieves
an overall balance. The major changes in this second edition most often
reflect the revolutionary influence that molecular biology has had on the
subject in the past 12 years.
Although the book is not designed as a conventional text of medical
and veterinary entomology, in Chapter 9 I have given a brief outline of
each of the blood-sucking insect groups. This chapter is intended as a
quick introduction for those entirely new to the subject, or as a refresher
on particular groups for those already familiar with the divisions of blood-
sucking insects. There are several introductory textbooks of medical and
veterinary entomology available to those requiring more information.
The book is primarily intended for advanced undergraduate and for
postgraduate students, but because it looks at topics that cut across the
normal research boundaries of physiology and ecology, behaviour and cell
biology, I hope it may also be useful for more established scientists who
xii Preface

want to look outside their own specialism. I have tried to distil this broad
spectrum of information, much of which is not readily available to the non-
specialist, into a brief synthesis. For those who want to look further into a
particular area I have included some of the references I found most useful in
writing the text, and these will provide an entry into the literature. Clearly
the subjects covered by the book encompass a vast number of publications
and I am sure to have missed many important and interesting references for
which I apologize in advance both to the reader and my fellow scientists.
Many of the topics discussed in the different chapters are interrelated. To
avoid repetition, and still give the broadest picture possible, I have given
cross-references in the text which I hope the reader will find useful.
From a comparative point of view it is an unfortunate fact that most of the
work on blood-sucking insects has been carried out on a few species. Conse-
quently, tsetse flies and mosquitoes pop up on every other page. In many
instances it remains to be seen how widely the lessons we have learned
from these well-studied models can be applied. Where possible I have
tried to point to general patterns that fit whole groups of blood-sucking
insects. To help me in this I have divided the blood-sucking insects into
three convenient but artificial categories: temporary ectoparasites, perma-
nent ectoparasites and periodic ectoparasites. These categories are based
solely on the behaviour biology of the blood-feeding stadia in the lives of
these insects. Temporary ectoparasites are considered to be those largely
free-living insects, such as the tabanids, mosquitoes, blood-feeding bugs
and blackflies, that visit the host only long enough to take a blood meal.
I also include insects such as the tsetse here, even though the male may
be found in swarms closely associated with the host for large parts of its
life. Permanent ectoparasites are considered to be those insects that live
almost constantly on the host, such as lice, the sheep ked and tungid fleas.
Finally, periodic ectoparasites are considered to be those insects that spend
considerably longer on the host than is required merely to obtain a blood
meal, but that nevertheless spend a significant amount of time away from
the host. Insects that fall into this category include many of the fleas and
Pupipara. These categories are no more than a useful generalization in the
text; I make no claims for their rigour and I realize that it could be argued in
several instances that an insect will sit as easily in one category as another.
 Acknowledgements

I gladly and gratefully acknowledge the help I have received from many
people during the writing of this book or the previous edition, particularly
P. Billingsley, A. Blackwell, J. Brady, H. Briegel, I. Burgess, E. Bursell,
R. Dillon, J. D. Edman, R. Galun, A. G. Gatehouse, M. Gaunt, M. Gillies,
R. H. Gooding, M. Greaves, C. Green, M. Hafner, J. Hogsette, H. Hurd,
A. M. Jordan, K. C. Kim, J. Kingsolver, M. Klowden, A. M. Lackie, B. R.
Laurence, E. Levashina, A. G. Marshall, P. Mellor, D. Molyneux, P. Mor-
rison, W. A. Nelson, G. O’Meara, G. S. Paulson, G. Port, N. A. Ratcliffe,
J. M. Ribeiro, P. Rossignol, M. Rothschild, W. Rudin, C. J. Schofield, M. W.
Service, J. J. B. Smith, W. Takken, S. Torr, G. A. Vale and J. Waage. I also
thank Paula Hynes, Maria Turton, Paula Dwyer and Dafydd Roberts for
help with the illustrations. Finally, most thanks go to my family, particu-
larly Stella, without whose encouragement, support and practical help this
book would never have been finished.
 1
The importance of blood-sucking
insects

Insects are the pre-eminent form of metazoan life on land. The class Insecta
contains over three-quarters of a million described species. Estimates for
the total number of extant species vary between 1 and 10 million, and it
has been calculated that as many as 1019 individual insects are alive at any
given instant (McGavin, 2001). That gives about 200 million for each man,
woman and child on Earth! It is estimated that there are 14 000 species of
insects from five orders that feed on blood (Adams, 1999) but, thankfully,
only 300 to 400 species regularly attract our attention. These blood-sucking
insects are of immense importance to humanity.
Humans evolved in a world already stocked with blood-sucking insects.
From their earliest days insects would have annoyed them with their bites
and sickened them with the parasites they transmitted. As humans evolved
from hunters to herders, blood-sucking insects had a further impact on their
wellbeing by lowering the productivity of their animals. It is reasonable to
assume that, because of their annoyance value, humanity has been in battle
with blood-sucking insects from the very beginning. In recent years this
battle has intensified because of an increasing intolerance of the discomfort
they cause, our fuller understanding of their role in disease transmission
and the demand for greater agricultural productivity. But despite consid-
erable advances in our knowledge of the insects and improvements in the
weapons we have to use against them, there is still no sign of an eventual
winner in this age-old battle.
Many keen observers of nature suspected that insects were in some way
involved with many of the febrile illnesses of humans and their animals
well before confirmatory scientific evidence was available. The explorer
Alexander von Humboldt recorded such a belief amongst the tribes of the
Orinoco region of South America. The great German bacteriologist Robert
Koch reported the belief of the tribes of the Usambara Mountains of East
Africa that the mosquitoes they encountered when they descended to the
plains were the cause of malaria (Nuttal, 1899). Sir Richard Burton, in his
travels in East Africa, recorded the similar belief of Somaliland tribes that
mosquitoes were responsible for febrile illnesses (Burton, 1860). Many of
the peoples living near the tsetse fly belts of East and West Africa associated
tsetse flies with sleeping sickness of humans and nagana of animals. In our
2 The importance of blood-sucking insects

Table 1.1 An outline of the early investigations that laid the foundations of
medical and veterinary entomology.

Date Source Subject

1878 Manson Development of Wuchereria bancrofti in a mosquito

1893 Smith and Kilbourne Babesia bigemina, the causative agent of Texas cattle
fever, transmitted by the tick, Boophilus annulatus
1895 Bruce Transmission of nagana by tsetse fly
1897 Ross Malaria parasites seen to develop in mosquitoes
1898 Ross Transmission of avian malaria by mosquitoes
1898 Simond Transmission of plague from rat to rat by fleas
1899 Grassi, Bignami and Anopheles spp. are the vectors of human malaria
Bastianelli
1900 Reed et al. Transmission of yellow fever by the mosquito Aedes
aegypti
1902 Graham Transmission of dengue by mosquitoes
1903 Bruce and Nabarro Sleeping sickness in humans transmitted by tsetse fly
1903 Marchoux and Transmission of fowl spirochaetes, Borrelia conserina,
Salimbeni by the tick Argus persicus
1907 Mackie Spirochaete causing relapsing fever transmitted by
lice
1909 Chagas Trypansoma cruzi, causative agent of Chagas’ disease,
transmitted by reduviid bugs

own western tradition North American stock ranchers held the belief that
Texas cattle fever was transmitted by ticks (in the class Arachnida, not
Insecta) well before this was confirmed experimentally.
The fact that insects are vectors of disease was only confirmed scientif-
ically at the end of the nineteenth century. The key discovery was made
in 1877 (reported in 1878) by a Scottish doctor, Patrick Manson, work-
ing for the customs and excise service in China. He found that larval
stages of the filarial worm, Wuchereria bancrofti, developed in the body of a
mosquito, Culex pipiens quinquefasciatus (Manson, 1878). This was the start
of an avalanche of investigations that laid the foundations of medical and
veterinary entomology. Some of the key discoveries of this era are outlined
in Table 1.1. The main insects involved in the transmission of all the most
important vector-transmitted diseases (Table 1.2) are now well known.
The list of diseases transmitted is an impressive one and includes the med-
ical scourges malaria, sleeping sickness, leishmaniasis, river blindness, ele-
phantiasis, yellow fever and dengue, and the veterinary diseases nagana,
surra, souma, bluetongue, African horse sickness and Rift Valley fever
 The importance of blood-sucking insects 3

Table 1.2 Rounded estimates for the prevalence of disease, the number at risk
and the disability adjusted life years (DALYs) for major vector-borne diseases.
Figures in millions (M). (DALYs were introduced in the World Bank
Development report of 1990 as an estimate of the burden a disease causes to the
health of the population. They are often used for comparative purposes and for
use in prioritization.)

Major
Disease Prevalence At risk DALYs distribution Major vectors

Malaria 273M 2100M 42M Tropics and Anopheline

subtropics mosquitoes
Onchocerciasis 18M 120M 1M Tropical Africa, Blackflies
(river blindness) Yemen, Latin (Simulium spp.)
America
Lymphatic 120M 1100M 5.6M Africa, Asia and Various
filariasis South America mosquitoes
(elephantiasis)
African 0.5M 50M 2M Sub-Saharan Tsetse flies
trypanosomiasis Africa
Chagas’ disease 16–18M 120M 0.7M Central and Triatomine bugs
South America
Leishmaniasis 12M 350M 2M Africa, Asia and
Latin America Sandflies
Dengue 50M 3000M 0.5M Asia, Africa and Various
Americas mosquitoes

Data largely from World Health Organization web pages as of 11 December 2002: http://
www.who.int/tdr/media/image.html.

(the piroplasms being tick-borne). Gauging the extent of these diseases

is much more problematical, even for human disease. One reason is that
health statistics are a moving target, particularly for those diseases such as
yellow fever that occur as epidemics. But the greatest problem is that the
heartland of these vector-borne diseases is in the under-developed world
where, for a variety of reasons, accurate statistical data are often difficult or
impossible to gather. For this reason, figures given for the extent of a dis-
ease are often not based entirely on hard data, but are an estimate founded
largely upon the experience of an expert. Table 1.2 gives an estimate of
some of the major vector-transmitted diseases of humans, but obviously,
as just indicated, care needs to be taken in the interpretation of the figures.
Blood-sucking insects cause very serious losses to agriculture (Table 1.3).
One way this happens is through the transmission of parasites. The
4 The importance of blood-sucking insects

Table 1.3 Estimated losses in agricultural production caused by blood-sucking

insects.

Animal mainly Estimated losses Geographical

Insect Year affected (millions US$) region

Haematobia irritans 1991 Cattle 800 USA

(horn fly)
Stomoxys calcitrans 1965 Cattle 142 USA
(stable fly)
Tabanids 1965 Cattle 40 USA
Mosquitoes 1965 Cattle 25 USA
Melophagus ovinus 1965 Sheep 9.4 USA
(sheep ked)
Lice 1965 Cattle 47 USA
Sheep 47
Swine 3
Goats 0.8
Tsetse fly 1999 Cattle 4500 Sub-Saharan
Africa
Insects, ticks, mites 1994 3000 USA

Information from: Budd, 1999; Geden and Hogsette, 1994; Kunz et al., 1991; Steelman,
1976.

most celebrated case is trypanosomiasis, transmitted by tsetse flies across

9 million km2 of Africa (Hursey, 2001), and estimated to cause agricul-
tural losses of about US$4.5 billion a year (Budd, 1999). The counter argu-
ment has also been proposed that the tsetse has prevented desertifica-
tion of large areas of land by overgrazing, and has been the saviour of
Africa’s game animals. The debate has been clearly outlined by Jordan
(1986). Other examples of spectacular losses caused by insect-transmitted
disease are the death in 1960 of 200 000 to 300 000 horses in Turkey, Cyprus
and India caused by African horse sickness, transmitted by Culicoides spp.
(Huq, 1961; Shahan and Giltner, 1945); and the estimated deaths in the USA,
between 1930 and 1945, of up to 300 000 equines from Western and East-
ern equine encephalitis transmitted by mosquitoes (Shahan and Giltner,
1945).
In the developed countries it is usually direct losses caused by insects
themselves that are of greatest concern. In exceptional circumstances the
insects may be present in such numbers that stock are killed; for example,
16 000 animals died in Romania in 1923 and 13 900 in Yugoslavia in 1934
because of outbreaks of the blackfly Simulium colombaschense (Baranov,
 The importance of blood-sucking insects 5

1935; Ciurea and Dinulescu, 1924). More usually losses are caused not
by death but by distress to the animal. Good examples are the reduc-
tions in milk yields, weight gains or feed efficiencies that are commonly
caused by the painful bites of the tabanids and biting flies. Estimated
losses in the USA have been calculated (Steelman, 1976). More recent esti-
mates suggest insects, ticks and mites cost the US livestock producer in
excess of $3 billion annually (Geden and Hogsette, 1994). The horn fly
is perhaps the major pest in the USA, with an estimated loss in excess of
$800 million annually (Kunz et al., 1991). Losses are caused by reduced feed
conversion efficiency, reduced weight gains and decreased milk produc-
tion and are the result of blood loss, annoyance, irritation and behavioural
defensive responses on the part of the host.
The sheer annoyance that blood-sucking insects cause to us can easily be
overshadowed by their importance in medical and veterinary medicine. In
some parts of the world, at certain times of the year, there may be so many
blood-sucking insects that any activity outside is difficult or impossible
without protective clothing. For example, the biting activity of the midge
Culicoides impunctatus is thought to cause a 20 per cent loss in working hours
in the forestry industry in Scotland during the summer months (Hendry
and Godwin, 1988). Such disruption is common during the summer blooms
of insects at many of the higher latitudes, and also in many of the wetter
areas of the tropics. These levels of annoyance are still rare for most people,
and for this reason the concept of nuisance insects is much more difficult
to grasp than that of a vector or an agricultural pest causing economic
damage. Perhaps the best way to view annoyance caused by insects is
as a tolerance threshold. It can then be viewed as a variable with widely
separated upper and lower limits; a handful of mosquitoes may be a minor
inconvenience to the beggar in the street but intolerable to the prince in the
palace.
I suggest that, in the developed world at least, we are increasingly
intolerant of nuisance insects. There are several underlying reasons: the
increased awareness in the general population of the importance of insects
in the spread of disease (sometimes over-exaggerated); the growing stress
placed on hygiene and cleanliness; and increasing urbanization, so that for
many people blood-sucking insects are not the familiar, everyday things
that they were once to our grandparents working in a rural economy.
This reduction in our tolerance of nuisance insects causes problems. The
extended leisure time and mobility of many people in the developed world
means that they spend more time in increasingly distant places. The coun-
tries involved are often anxious to promote and develop their tourist indus-
tries, and this has led to pressure to control nuisance insects. This can
be seen in places such as the Camargue in southern France, the Scottish
6 The importance of blood-sucking insects

Highlands (Blackwell, 2000), the Bahamas, New Zealand (Blackwell and

Page, 2003), Florida and many parts of the Caribbean (Linley and Davies,
1971). In addition, population growth has put increased pressure on
marginal land which in the past may have been left alone because of nui-
sance insect problems. Development of this land for leisure, commerce or
housing with no insect control input can be disastrous for the developer,
user or purchaser.
 2
The evolution of the blood-
sucking habit

It is believed that haematophagy arose independently at least six times

among the arthropods of the Jurassic and Cretaceous periods (145–65 mil-
lion years ago) (Balashov, 1984; Ribeiro, 1995). The very patchy nature of
the insect fossil record means that discussion of the evolution of the blood-
sucking habit has until now relied heavily on detective work, with the
major clues lying in the diversity of forms and lifestyles seen in modern-day
insects, and in some cases in the details of their relationships with verte-
brates. From careful interpretation of this evidence quite credible accounts
of the likely evolution of the blood-sucking habit can be made. From this
starting point it has been convincingly argued that the evolution of the
blood-sucking habit in insects has occurred on several occasions, in each
case along one of two main routes (Waage, 1979), and these are discussed
below. Insect molecular systematics is beginning to emerge from its ‘Tower
of Babel’ stage (Caterino et al., 2000) and it will make a major contribution
in defining the detail of the evolutionary routes taken by haematophagous
insects (Esseghir et al., 1997; Hafner et al., 1994; Lanzaro et al., 1998; Mans
et al., 2002; Sallum et al., 2002). The proposed population bottleneck suf-
fered by phlebotomines in the late Pleistocene and the subsequent radia-
tion of the species out from the eastern Mediterranean sub-region is a good
example of what we can expect (Esseghir et al., 1997).

2.1 Prolonged close association with vertebrates

In the first route it is suggested that haematophagous forms may have
developed subsequent to a prolonged association between vertebrates and
insects that had no specializations immediately suiting them to the blood-
sucking way of life. The most common association of this type is likely
to have centred around the attraction of insects to the nest or burrow of
the vertebrate host. Insects may have been attracted to the nest for several
reasons. The humid, warm environment would have been very favourable
to a great many insects. In some circumstances, such as the location of the
nest in a semi-arid or arid area, the protected habitat offered by the nest
may have been essential to the insects’ survival. For many insects the nest
would also have proved attractive for the abundant supply of food to be
8 The evolution of the blood-sucking habit

found there. Certainly many current day insects such as the psocids are
attracted to the high concentrations of organic matter to be found in nests.
Indeed, psocids may become so intimately associated with this habitat that
they develop a phoretic association with birds and mammals, climbing into
fur and feathers, to be translocated from one nest site to another (Mockford,
1967; Mockford, 1971; Pearman, 1960).
Initially feeding on dung, fungus or other organic debris, the insects
attracted to the nest would also have encountered considerable quanti-
ties of sloughed skin, hair or feathers. The regular, accidental ingestion of
this sloughed body covering probably led to the selection of individuals
possessing physiological systems capable of the efficient use of this mate-
rial. Behavioural adaptations may then have permitted occasional feeding
direct from the host itself. It is easy to see how this may have gone hand
in hand with the adoption of a phoretic habit. Morphological and further
behavioural adaptations would have allowed the insect to remain with
the host for longer periods with increasingly efficient feeding on skin and
feathers.
The mouthparts developed for this lifestyle, in which the insect feeds pri-
marily on skin and feathers, were almost certainly of the chewing type, such
as those seen in the present-day Mallophaga. While these mouthparts are
not primarily designed to pierce skin some mallophagans do feed on blood.
Menacanthus stramineus, a present-day mallophagan, feeds at the base of
feathers or on the skin of the chicken. The insect often breaks through
to the dermis, giving it access to blood on which it will feed (Emmerson
et al., 1973). Blood has a higher nutritional value than skin and is far eas-
ier to digest. This is reflected in the increased fecundity of blood-feeding
Anoplura compared to skin-feeding Mallophaga (Marshall, 1981). Once
blood was regularly encountered by insects, it is likely that its high nutri-
tional value favoured the development of a group of insects that regularly
exploited blood as a resource. This would have developed progressively,
through physiological, behavioural and morphological adaptations, first
to facultative haematophagy and eventually, in some insects, to obligate
haematophagy. One way in which the progression from skin feeding to
blood feeding may have occurred is seen in members of the mallophagan
suborder the Rhynchophthirina, such as the elephant louse, Haematomyzus
elephantis. This insect possesses typical mallophagan biting-type mouth-
parts (Ferris, 1931; Mukerji and Sen-Sarma, 1955) which are not primarily
adapted for obtaining blood. By holding the mouthparts at the end of an
extended rostrum (Fig. 2.1) the insect manages to use them to penetrate the
thick epidermal skin layers of the host to get to the blood in the dermis.
It is thought that haematophagous lice developed from an original nest-
dwelling, free-living ancestor (Kim, 1985) along the pathway described
above. We do not know when the change occurred from free-living nest
2.1 Prolonged close association with vertebrates 9

Figure 2.1 Despite having the chewing mouthparts typical of mallophagans,

Haematomyzus hopkinsi is unusual in feeding on blood. The chewing
mouthparts are held on the end of an unusual, elongated rostrum, which may
well be an adaptation helping the insect reach the blood-containing dermis
through the thick skin of its wart-hog host. (Courtesy of Vince Smith)
10 The evolution of the blood-sucking habit

dweller to parasite, but it may well go right back to the appearance of nest-
ing or communal living in land-dwelling vertebrates, which is thought to
have happened during the Mesozoic (225–65 million years ago). So lice may
have predated the emergence of mammals and birds and been parasitic on
their reptilian ancestors (Hopkins, 1949; Rothschild and Clay, 1952). It is
highly likely that ancestral forms were parasitic on primordial mammals
and that from there they radiated along the lines of mammalian evolu-
tion. Helping drive this rapid speciation of the permanent ectoparasites
was the reproductive isolation they suffered from being confined on spe-
cific vertebrate hosts, which may well have enhanced the effects of classical
geographic reproductive isolation. Co-evolution of the host and permanent
(and to a lesser extent temporary) ectoparasites probably led to rapid spe-
ciation in lice and other ectoparasitic forms. The evidence for co-speciation
in lice is strong. Sequence analysis of mitochondrial cytochrome oxidase I
genes suggests co-speciation in the pocket gophers Orthogeomys, Geomys
and Thomomys and their chewing lice (Fig. 2.2) (Hafner et al., 1994).
Co-speciation also predicts temporal congruence between chewing lice and
gopher speciation. This is borne out by analysis of the molecular data, in
which the synonymous substitution rate is approximately an order of mag-
nitude greater in the lice compared to the gophers. This roughly parallels
the differences in generation times of the two groups, suggesting equal
rates of mutation per generation. While the case for pocket gophers and
their chewing lice is strong, the extent to which co-speciation is generally
the case is unclear. Classical taxonomy, which has tended to group species
with origins on the same host, may be misleading. Molecular studies are
showing this is a dangerous practice and that not all species have stuck to
the co-evolutionary model mentioned above (Johnson et al., 2002a; Johnson
et al., 2002b).
Some beetles also appear to be developing along the evolutionary high-
way described above. Several hundred species have been reported from
nests and burrows (Barrera and Machado-Allison, 1965; Medvedev and
Skylar, 1974). Most of these are probably free-living, feeding on the high
levels of organic debris to be found at these sites. Some of these beetles have
developed a phoretic association with the mammal which allows them to
transfer efficiently between nest sites. Many of these phoretic forms also
feed on the host by scraping skin and hair, and some have progressed to the
stage when they will occasionally take blood (Barrera, 1966; Wood, 1964).
The prolonged association of the insect with the vertebrate, which is
the cornerstone of this first route for the evolution of the blood-feeding
habit, may not always have relied on encounters in the nest habitat. Free-
living ancestral forms with few, if any, clear adaptations for the blood-
sucking way of life may have also developed prolonged associations with
 2.1 Prolonged close association with vertebrates 11

Figure 2.2 Phylogenies of pocket gophers and their chewing lice based on
nucleotide sequence data (Hafner et al., 1994). The figure shows composite
trees based on multiple methods of phylogenetic analysis. Branch lengths are
proportional to inferred amounts of genetic change. Pocket gopher genera
are Orthogeomys, Zygogeomys, Pappogeomys, Cratogeomys, Geomys and
Thomomys. Geomys bursarius is represented by two subspecies (a = G. b. halli;
b = G. b. majusculus). Chewing louse genera are Geomydoecus and
Thomomydoecus. The program COMPONENT was used to document
significant similarity in branching structure between these trees. Because the
host and parasite trees were based on DNA sequences from the same gene
(cytochrome c oxidase subunit I), rates of DNA evolution could be compared
in the two groups. Based on these data, Hafner et al. (1994) estimated that
chewing lice were evolving approximately ten times faster than pocket
gophers in this gene region, which is in line with the predictions if
co-evolution is occurring (see p. 10). But see Page et al. (1996).

the vertebrate at some point distant from the nest. This type of association
may have had several different underlying reasons, such as attraction to
feed on vertebrate secretions, or the use of the vertebrate as a basking or
swarming site. But probably the most important factor was the use of the
host’s dung as a larval habitat.
The vertebrate may live in a harsh environment where dung rapidly
dries up or it may bury its dung. In either case, closely associating with
12 The evolution of the blood-sucking habit

the vertebrate would help the insect in locating usable dung. Non-blood-
feeding associations arising for these reasons are seen between some
scarabaeid beetles and vertebrates, but competition to be the first to exploit
dung was probably the commonest reason driving the insects to an ever
closer association with the vertebrate. Dung is a limited resource and there
is often intense competition to utilize it as a larval site. Consequently, there
is pressure on the insect to be the first to introduce its eggs into the newly
deposited dung. One way for the insect to achieve this is to form an increas-
ingly close association with the vertebrate providing the dung. Such a close
association can be seen with the horn fly, Haematobia irritans, which as an
adult is permanently associated with large vertebrates, only leaving them
to oviposit. It will lay eggs in dung within 15 s of defecation by the verte-
brate and, within its distribution range, is almost always the first colonizer
of freshly deposited dung (Mohr, 1943). If such an insect can feed on the
vertebrate, it minimizes the time it will have to spend away from the host
and therefore maximizes the advantage to be gained from the close asso-
ciation, and this may have led the ancestral adult female to feed on body
secretions, open wounds and sores. As was argued above, because blood
is such a nutritious substance, once it was encountered in the diet of the
insect, selection is likely to have favoured the progressive development
of physical and behavioural adaptations leading to full haematophagy.
Selection will have led to the progressive development of mouthparts,
allowing the insect to dislodge scabs, open up old sores and eventually to
penetrate unbroken skin. The evolution of organisms capable of break-
ing the skin and obtaining a blood meal may very well have been an
accelerated process. The wounds produced by the first blood-suckers will
have provided regular blood-feeding opportunities for other organisms
that could not break the skin by their own efforts. Once given access to
blood these too may well have followed the evolutionary pathway outlined
above.
The spasmodic appearance of blood feeding among male insects is a
difficult issue to explain, but the close or permanent association of females
with the vertebrate as discussed above is one factor that may explain it in
some insects. Under these circumstances it may become advantageous for
the male to become associated with the vertebrate because of the greater
likelihood of success in finding a mate. This may then lead to blood feeding
in the male, as it would minimize the time spent away from its host and
therefore maximize its chances of successful mating. Similarly, if blood-
sucking females are irregularly and/or widely dispersed in a habitat, then
the male may gain a mating advantage by staying with the vertebrate and
waiting for the female to arrive to feed. This is seen in the ‘following swarm’
of the tsetse flies but also in the mosquito Aedes aegypti, the males of which
do not take blood (Teesdale, 1955).
 2.2 Morphological pre-adaptation for piercing 13

2.2 Morphological pre-adaptation for piercing

The second route for the evolution of the blood-sucking habit suggests
that blood feeding developed in some insect lineages from ancestral insects
that were morphologically pre-adapted for piercing surfaces (Beklemishev,
1957; Downes, 1970; Waage, 1979). Entomophagous insects are strong can-
didates for such a conversion. The Rhagionidae are a good example: most
of the group are predacious on other insects, but a few species have turned
to blood feeding. How could this changeover have come about? Ento-
mophagous insects would have been attracted to nests and burrows by
the accumulation of insects to be found there and so would have encoun-
tered vertebrates. Away from the nests they would have been attracted to
vertebrates by the accumulation of insects around them, or the vertebrates
may have regularly congregated in the wet areas that are the breeding
sites for many of the ‘lower’ Diptera. The vertebrates involved may have
been permanently resident amphibians or reptiles, or larger vertebrates
that regularly visited such sites for drinking or bathing purposes. In each
of these cases it is easy to see how entomophagous insects could have made
repeated and possibly prolonged contact with vertebrates. These predatory
insects would have physiological and morphological adaptations (such as
efficient protein-digesting enzymes and piercing mouthparts) facilitating
the switch to haematophagy. Haematophagy in these individuals was at
first probably an occasional, chance event which led to full haematophagy
through continued close association with the vertebrate host. It is thought
that haematophagy developed along these lines in the ancestors of the
blood-feeding bugs and in blood-feeding rhagionids and possibly in some
blood-feeding Diptera. There is some doubt about which came first, lar-
val feeding on nest debris or adult feeding on other insects in the nest
habitat. Fleas may also have evolved along this pathway from free-living
mecopteran stock (Hinton, 1958; Tillyard, 1935). The Mecoptera, or scor-
pion flies, contain a modern-day group, the Boreidae, which are apterous
and are capable of jumping. They live in moss and feed on insects. Simi-
lar insects may well have been the ancestors of the fleas, a view receiving
support from molecular systematics (Whiting, 2002).
The lifestyles of several present-day insects support the idea that ento-
mophagous insects gave rise to some blood-sucking insect groups (Waage,
1979). Many personal experiences in Britain (as many as three in one day)
with the flower bug Anthocoris nemorum show that this insect is willing
and able to pierce human skin. This insect is entomophagous, living on
and around flowers where it pounces on small insects visiting the flow-
ers to feed. While its probings of my skin cause a sharp pain, I have yet
to find one that has obviously ingested any blood; however, it still estab-
lishes the fact that entomophagous insects will often show an interest in
14 The evolution of the blood-sucking habit

vertebrates as potential sources of a meal. The hemipteran bug Lyctocoris

campestris is also an entomophagous species, but it takes matters further.
It can live in birds’ nests, where it feeds on other insects, but it will also
take blood meals from vertebrates (Stys and Daniel, 1957). Evidence from
blood-sucking insects themselves also points to the close links between
entomophagy and the blood-feeding habit. The mosquitoes Aedes aegypti
and Culex tarsalis will take body fluids from insect larvae presented to them
under laboratory conditions (Harris et al., 1969). Indeed, this form of feed-
ing is so successful that these mosquitoes go on to produce viable eggs
as a result, which opens up the intriguing possibility that this may occur
naturally in the field.
It has also been argued that haematophagy may have arisen in some
insect groups (including the mosquitoes; Mattingley (1965)) from plant-
feeding ancestors. This is certainly a possibility as many plant-feeding
insects possess piercing and sucking mouthparts that would pre-adapt
them for haematophagy. This is seen in the moth Calpe eustrigata, which is
one of a group of noctuiids that possess an unusually modified, sharp
proboscis used in most species for the penetration of fruit rinds. But
C. eustrigata uses it to penetrate vertebrate skin for the purposes of blood-
feeding. It is probable that plant-feeding ancestors of modern-day blood
feeders would have developed haematophagy only if they were in a posi-
tion of continual association with the vertebrate host. This may have
occurred through mechanisms similar to those already outlined. Attrac-
tion to free-living vertebrates may have occurred in order to feed on bodily
secretions or to use dung as a larval medium. Or insects may have been
attracted to nests to feed on fruits or seeds stored there by the vertebrate.
In this context it is interesting to note that the hemipteran bugs are excep-
tional in using cathepsin-like digestive proteinases. That is consistent with
a proposed evolutionary path for bugs from sap-sucking (Billingsley and
Downe, 1988; Houseman et al., 1985; Terra, 1988) or seed-feeding ances-
tors. Sap feeders, not needing proteases, may have lost their trypsins. If
they then moved to blood feeding they would need to reacquire proteolytic
activity. Seeds are often sources of powerful anti-serine protease molecules
produced to protect the seed from insects. Seed feeders may have moved
to cathepsins to avoid these inhibitors. It is argued that having lost their
trypsins they had to make use of the cathepsins contained in the lysosomes
of all cells, re-routing them for extracellular digestion. Molecular evidence
suggests Triatominae from the Americas and Asia are monophyletic with
an origin in northern areas of South America, in Central America, or in the
southern region of North America about 95 million years ago (Gaunt and
Miles, 2002; Lyman et al., 1999).
 3
Feeding preferences of blood-
sucking insects

3.1 Host choice

Blood-sucking insects feed from a range of different host animals; because
of the bites we receive we are acutely aware of the fact that many of them
feed from humans, but many other animals are also exploited, including
other mammals, birds, reptiles, amphibians and fish, and even insects,
arachnids and annelids (Hocking, 1971). Any one insect does not feed
equally well from all of these potential resources; it displays host choice. For
some insects, particularly some permanent ectoparasites, the host choice
may be very specific. Occasionally, for example for human lice, just a sin-
gle species. For other blood-sucking insects host choice is clearly not as
restricted as introduced exotic hosts (e.g. those in zoos) quickly become
incorporated into the diet of local blood-sucking insects.
Let us consider what is meant by host choice. In its main sense it denotes
the species of host animal or animals from which blood-sucking insects
obtain their blood meals. But host choice can go beyond particular species
of host chosen. Insects often choose to feed on particular individuals from
among preferred species, which may well have implications for disease
transmission (Burkot, 1988; Kelly, 2001; McCall and Kelly, 2002).
Although most blood-sucking insects in their undisturbed, natural sur-
roundings show a preference for feeding from a particular group or species,
or even a primary cohort of their chosen species, the degree of host speci-
ficity shown varies greatly from one type of insect to the next. Some are
entirely dependent on a single species of host while others are willing to
feed from a wide range of hosts. As a rough rule of thumb, it has often been
said that there is a direct relationship between the locomotory capabilities
of an insect and the number of different hosts that it utilizes. Thus, the
permanent ectoparasites, with their limited capacity for movement away
from the host, contain most cases of precise dependence. For example,
the louse Haematomyzus elephantis is confined to elephants, H. hopkinsi to
the wart-hog (Clay, 1963) and Pediculus capitis to humans. Considering the
more mobile periodic ectoparasites, we can find many examples such as
the ‘human’ flea, Pulex irritans, that have a narrow preferred range (bad-
gers and foxes) on which most insects are found, but that can occasionally
16 Feeding preferences of blood-sucking insects

be discovered on a wide range of other hosts (in this case humans, pigs
and other large mammals) (Marshall, 1981). When considering the largely
free-living, temporary ectoparasites we find that many have catholic tastes.
The mosquito Culex salinarius, for example, has a wide range of potential
hosts that includes birds (45 per cent of the blood meals tested), equines
(17 per cent) and canines (15 per cent) (Cupp and Stokes, 1976). This
mosquito will demonstrate its cosmopolitan tastes further if it is disturbed
during the meal by moving willingly from one host species to another to
complete this single, full blood meal. This is a regular and natural occur-
rence for this insect. In one investigation 13 per cent of the blood meals
tested were found to be from a mixture of hosts (Cupp and Stokes, 1976).
The above rule of thumb works well in a broad sense but often breaks
down when particular insects are considered, as can be seen with the help
of two examples. The amblyceran louse, Menacanthus eurysternus, has poor
locomotory abilities off the host and according to our rule of thumb it
should be limited to a small number of hosts. In fact it has been recorded
from 123 different species of bird (Price, 1975)! In contrast, each of four
closely related species of winged streblid were found to be specific to a dif-
ferent genus of bat and this relationship held true even in caves containing
three of the bat genera in crowded conditions (Maa and Marshall, 1981).
Before looking at some of the factors that underpin host choice, let us
make some general comments of particular relevance to temporary ectopar-
asites. The commonest hosts are probably large, social herbivores, which
present an abundant and easily visible food source for many ectoparasites.
They are also reliable food sources because normally they move only slowly
from one pasture to another. Carnivores are less abundant than their prey,
often solitary, less visible and less predictable, occupying a large home
range. For these reasons carnivores are less likely to be a primary element
in the host choice of a blood-sucking insect. Also large vertebrates are likely
to be preferred over smaller ones. This is because small vertebrates, which
suffer greater losses from the attention of a blood-sucking insect, use their
agility to develop efficient defences.
Moving from these generalizations to consider the specific choices made
by particular insects is a more complex issue. Sometimes a reasoned case
for the evolution of a particular pattern of host choice can be made, but
on most occasions we are in the dark. The choice may be determined
by a large number of factors (probably acting in combination), including
behavioural, physiological, morphological, ecological, geographical, tem-
poral and genetic considerations. To give an idea of the complexity of the
issue, let us look at a small number of examples, showing how each of these
factors may affect host choice.
Although it is not always the case (Canyon et al., 1998; Charlwood
et al., 1995; Prior and Torr, 2002), host defensive behaviour can have a
 3.1 Host choice 17

considerable effect on the number of blood-sucking insects successfully

feeding on that host (Torr et al., 2001). The anti-mosquito behaviour dis-
played by a range of ciconiiform birds is a good demonstration of the pro-
tective effects host defensive behaviour can have. At roost some of these
birds perform up to 3000 defensive movements per hour! But the green
heron and the crowned night heron are far less active, performing as few
as 650 defensive movements per hour: as a consequence they receive far
more bites than their more active relatives (Webber and Edman, 1972). If no
other factors were in operation it is easy to see how natural selection would
bear on such a case and would lead a blood-sucking insect to select and
eventually specialize in feeding from the easiest available target. The fact
that under natural conditions mosquitoes still feed from a wide range of
ciconiiform birds, including those displaying the highest levels of defensive
behaviours, is a testament to the complexity of the factors that determine
host choice. None the less, grooming and host defensive behaviour in gen-
eral are likely to be highly efficient agents of natural selection. For humans
the concept of defensive behaviour needs to be extended to include such
practices as sleeping under bednets, screening houses and the use of repel-
lents (modern preparations containing active ingredients such as di-ethyl
toluamide (DEET) or traditional concoctions such as mixtures of cow dung,
urine and ash) (MacCormack, 1984).
The intensity of defensive behaviour often correlates with the num-
ber of blood-sucking insects attacking the host. Thus is it possible that
the seasonal changes in the numbers of blood-sucking insects may play
a part in the seasonal changes in host choice because of changes in the
amount of host defensive behaviour (Edman and Spielman, 1988). Such
seasonal changes in host choice are important in the transmission of some
zoonoses to humans. For example, in the USA the arbovirus eastern equine
encephalitis is enzootic in birds and the mosquito Culiseta melanura is its
vector. During the spring and early summer, when mosquito numbers
are comparatively low, this insect feeds almost exclusively on passerine
birds. Later in the season mosquito numbers increase and so does defen-
sive behaviour by the birds. At this time mosquito host choice becomes
more catholic and includes other bird groups and mammals. It is also in
this later part of the season that epizootics of the virus occur in birds and
horses and epidemics occur in humans. All this is enabled by the greater
range of hosts chosen by the mosquitoes, driven by changing levels of host
defensive behaviour. Mathematical models of disease transmission have
been produced that incorporate the assumption that different individual
levels of host defensive behaviour lead to a non-homogeneous pattern of
host choice within a host species (Kelly and Thompson, 2000).
I suggest that physiological factors are mainly important in determining
host choice in the sense that once the insect has become associated with a
18 Feeding preferences of blood-sucking insects

Figure 3.1 The tsetse fly, Glossina morsitans morsitans, shows greater
longevity when it feeds on rabbits (
) rather than goats () (Jordan and
Curtis, 1972).

narrow range of hosts then specialization in its physiology will occur that
may limit the range of other hosts it can exploit (Krasnov et al., 2003). This
is because natural selection will ensure that all of the insect’s systems will
become tuned to the exploitation of the resources of its major host, which
may restrict the insect’s ability to deal with unusual situations. In some
cases choice of the wrong host can, for physiological reasons, lead to the
death of the insect. For example, a blood meal from the guinea-pig may
form oxyhaemoglobin crystals which will rupture the intestine of several
blood-sucking insects (Krynski et al., 1952).
The physiological consequences of moving on to an unusual host can
have a less obvious, but a nevertheless damaging, effect on the success
of the insect. Several experimental studies have shown that the fecun-
dity of a blood-sucking insect depends on the host on which the insect
feeds. Reduced fecundity can be generated by a reduced rate of devel-
opment of the insect, reduced longevity (Fig. 3.1), a skewed sex ratio, or
reduced food intake or rate of digestion (Chang and Judson, 1977; Nelson
et al., 1975; Rothschild, 1975). We can look at the tsetse fly as an example.
Glossina austeni fed on rabbit blood showed a consistently higher fecundity
than flies of the same species fed on goat blood (Jordan and Curtis, 1968).
In a further series of experiments it was shown that both G. austeni and
 3.1 Host choice 19

G. morsitans morsitans fed on pig blood produced the heaviest puparia,

closely followed by flies fed on goat blood, while the puparia produced
from flies fed on cow blood were considerably lighter (Mews et al., 1976).
Clearly under natural conditions it would be a selective advantage to the
insect to develop mechanisms that avoided such a restriction on its repro-
ductive success by careful choice of hosts. It is also of interest to note that
the source of the blood meal or its quality may also influence the course
of an infection in the insect. If tsetse flies feed on goats or cattle they will
develop a much higher infection rate with Trypanosoma vivax than if they
feed on mice (Maudlin et al., 1984). Similar results are also found in more
natural vector combinations (Nguu et al., 1996).
As with the physiological restrictions on host choice just discussed, mor-
phological factors are also important as a limitation on the range of avail-
able hosts once a certain degree of specialization for a particular host has
taken place. With permanent and periodic ectoparasites especially, restric-
tion in host range usually involves morphological specialization of the
insect’s mouthparts, ovipositor or locomotory attachment apparatus. Spe-
cializations of any of these can easily restrict the host range of the insect
concerned; indeed, in many cases specialization may even limit the area of
the body of the preferred host that the insect can utilize. This is discussed
in more detail in Section 7.1.
It has also been suggested that morphological characteristics of the host
are important in determining host choice. One example is the frequently
reported preference of Anopheles gambiae for human adults rather than chil-
dren. It is suggested that the choice is a direct consequence of the differ-
ence in size of these two potential hosts, the number of bites received by
an individual being in direct proportion to the surface area that individual
contributes to the total immediately available to the mosquito (Port et al.,
1980). Interestingly this is not the case for adult cows and calves; here it is
differences in levels of defensive behaviour that determine that adults are
bitten more than calves by tsetse flies (Torr et al., 2001).
A clear example of ecological influence on host choice is furnished by the
fleas. The immature stages of most fleas live on the detritus in the homes of
the host animal. In other words, host choice for fleas is largely determined
by the ecology of the host, most fleas being restricted to hosts with long-
term homes, or at least seasonal lairs that are used each year. Because they
so commonly use long-term homes, mammals are the main hosts for fleas.
Rodents are unrivalled as home builders and as a consequence are the
most afflicted mammals, harbouring 74 per cent of all known flea species
(Marshall, 1981). Humans are also habitual home dwellers and because of
this we are the only primates that are regularly flea-ridden. Interestingly
the 5000-year-old late-Stone-Age Tyrolean ice man was infested with Pulex
irritans (Spindler, 2001).
20 Feeding preferences of blood-sucking insects

A second example of ecological influence on host choice, of considerable

importance in the health of Europeans, can be seen in the changing feeding
patterns of the anopheline mosquitoes of Europe. Until the middle of the
nineteenth century benign tertian malaria (caused by Plasmodium vivax)
was endemic in Europe as far north as Scandinavia, with P. malariae as
far north as Holland and the UK, and P. falciparum extensive in southern
Europe. At this time malaria began to retreat from Europe and trans-
mission of the disease now rarely occurs. Why did this happen? Ronald
Ross described the full cycle of malaria in 1898 and soon after this, in
1901, it was realized that there were areas of Europe that abounded with
anopheline mosquitoes, and into which malaria carriers routinely passed,
but in which the disease was not transmitted (Falleroni, 1927). Several
theories were put forward to account for this apparent anomaly, but it
was not until after the introduction of the precipitin test, for the identifi-
cation of mosquito blood meals (see Pant et al., 1987), that relevant evidence

Box 3.1 The importance of human-biting rates of mosquitoes for the

transmission of malaria

The vectorial capacity C of a population of mosquitoes transmitting

malaria can be described by the equation
Pn
C = m × a2 ×
−ln(P )
where m is the number of mosquitoes per person and a is the propor-
tion of these mosquitoes that bite humans (biting once to acquire the
infection and transmitting it on the occasion of a subsequent meal on a
human, hence a2 ) – the quantity a is usually estimated using techniques
such as those described in Box 3.2.
The value P n/−ln(P ) describes the expectation of the infective life of
the vector population (Garrett-Jones and Shidrawi, 1969).
If we take the example described in the text we can see that in those
parts of Europe where malaria was being transmitted a = 1/8, whereas
in those areas where mosquitoes were present but no malaria was being
transmitted a = 1/400.
If we assume that m and P n/−ln(P ) are constants, then the equation
tells us that humans in the malarious areas are 2500 times more likely to
contract malaria than those in the non-malarious areas. The difference
in this example is entirely due to the different proportions of mosquitoes
biting humans in the two areas. A good account of these mathematical
models is given by Dye (1992).
Source: From Burkot (1988)
 3.1 Host choice 21

Box 3.2 Identification of the source of a blood meal

The ability to identify the source of an insect’s blood meal accurately

is a powerful tool for the investigation of a wide range of questions
relating to the life of blood-sucking insects. For example, it allows us
to investigate the nature of host choice and it can help us to understand
whether parasites manipulate the host-seeking behaviour of their vectors
(Koella et al., 1998b). It may help in monitoring the efficacy of various
control or surveillance tools such as repellents, bednets and traps and to
estimate the degree of coverage needed for emerging malaria vaccines
(Mukabana et al., 2002a).
Until recently blood meal identification has been achieved through
immunological means (Pant et al., 1987). Increasingly molecular tech-
niques are taking over (Mukabana et al., 2002a). These are usually ‘finger-
printing’ techniques that look at the number of tandem repeats (VNTRs)
and/or short tandem repeats (STRs), both of which occur abundantly in
nuclear DNA in most eukaryote genomes. The number of tandem repeats
at these loci is highly variable with alleles differing in length by an integral
number of repeat units. As well as being highly polymorphic these loci are
stably inherited and so are informative genetic markers. The markers can
be easily accessed because they can be amplified in vitro by polymerase
chain reaction and the size of the products easily determined. They have
been used to identify successfully the blood meal source of mosquitoes
(Koella et al., 1998a; Mukabana et al., 2002a; Mukabana et al., 2002b),
crab lice (Lord et al., 1998) and tsetse flies (Torr et al., 2001). Not only
can they identify the species that was the source of the blood meal, they
can be so specific that they can identify the particular individual within
a species from which the blood meal was taken.

began to accumulate. The key factor was that anopheline mosquitoes in

the malaria-free areas were 400 times as likely to feed on animals other
than humans, while in malarious areas 1 in 8 feeds were from humans
(Hackett and Missiroli, 1931). The importance of these rates of biting for
the transmission of malaria is explained in Box 3.1. Immunological tests
are now being replaced by more sensitive and specific molecular tech-
niques for the identification of vector blood meals (Mukabana et al., 2002a)
(Box 3.2).
Why should the mosquitoes of one area bite humans so regularly and
mosquitoes of the same species in another area bite them so infrequently?
Although it was suggested that the mosquitoes were actually changing
in their relative attractions for humans and domesticated animals (par-
ticularly pigs, horses and cattle), no such change could be demonstrated
experimentally. It is now widely accepted that the switch from humans
22 Feeding preferences of blood-sucking insects

to animals was a response to the changing proportions of human and

non-human hosts available. At this time animal husbandry practices in
Europe were changing so that far more animals were being kept and were
available to the mosquitoes as food sources. A contributory factor was
the improving standard of living for the human population, which led
to larger and brighter housing that was much less attractive as a rest-
ing site for the mosquito. The mosquitoes were diverted to the darker
animal sheds around the house, considerably reducing the likelihood of
their biting humans (Harrison, 1978; Takken et al., 1999). Also of impor-
tance was the decreasing birth rate among the population, reducing the
numbers of human hosts available, a factor accentuated by the increased
mechanization of agriculture which also led to a reduction in the popu-
lations in rural communities. So we can see that host choice is not only
a behavioural, decision-making process on the part of the insect, but can
also be strongly influenced by the relative availability of different host
species.
Geographical considerations are also important in host choice because
if there is no overlap in the range of insect and potential host then the host
is not available as a food source. Global climate change may be affecting
these overlaps (Sutherst et al., 1998). Geographic overlap of host and insect
can work on a variety of scales. The effect is obvious from a global view-
point, but not so obvious when more localized issues are considered. The
tsetse fly can be used as an example, one that has considerable economic
importance. On the largest geographical scale we can see that tsetse flies,
apart from a small fly belt in Saudi Arabia (Elsen et al., 1990), are only found
in Africa, where they are widely distributed from the southern borders of
the Sahara to Mozambique. Within this region they transmit trypanosomes
causing sleeping sickness in humans and nagana in animals. The game ani-
mals of Africa have some degree of immunity to nagana, but normally it
quickly kills introduced domesticated animals. This disease and its trans-
mission by the tsetse fly have had a tremendous influence on the history
of tropical Africa. The combination largely prevented the invasion from
the north of armies dependent on horses. It also impeded Africa’s devel-
opment by limiting the use of draft animals and prevented (and indeed
still prevents) the use of large tracts of land for ranching. In other words
large herbivores entering the tsetse’s geographical range, including horses
and cattle, become available as hosts and are often chosen as food sources.
But geographical overlap is also important in tsetse fly biology on a much
smaller scale. Long before Bruce (1895) demonstrated that the tsetse fly
transmits the trypanosomes that cause illness and death, the herdsmen of
southern Africa had learned by bitter experience that the fly was deadly.
This fact became obvious to them because the tsetse flies of this region
are not evenly spread throughout the land, but are restricted to certain
 3.1 Host choice 23

areas known as fly belts. The herdsmen learned that if they avoided these
belts, and the flies that lived there, then their stock did not acquire nagana
(McKelvey, 1973).
Geographical effects on host choice can be seen on an even smaller scale.
The tree hole mosquitoes Aedes triseriatus and Ae. hendersoni are sibling
species living sympatrically in woodlands in the eastern and midwestern
USA. It might be expected that two such closely related species living in
the same woodland would be feeding on similar hosts, but geographi-
cal separation again affects host choice. Aedes triseriatus feeds mainly at
ground level and consequently feeds on ground-dwelling animals such as
deer and chipmunk. In contrast, Ae. hendersoni feeds mainly in the canopy
of these woodland trees on animals such as tree squirrels (Nasci, 1982).
An important example, because it has consequences for malaria transmis-
sion, occurs with the malaria vectors within the Anopheles gambiae complex
(see Section 3.2). It has often been reported that An. arabiensis is much more
zoophilic that An. gambiensis s.s. but in a carefully designed experiment that
controlled for equal accessibility to hosts, which were presented outside
houses, it was found that both species showed statistically similar levels of
human- and cattle-biting activity (Diatta et al., 1998). So an unwillingness
to bite indoors may be the basis of previous reports of zoophily in An. ara-
biensis rather than a real difference in host preference. Clearly geographical
overlap on even the smallest scale is very important in determining host
choice; indeed, host abundance and proximity may be the ultimate arbiter
of host choice in many instances.
As we have just seen, availability of hosts is often a prime factor in
determining host choice. For the host to be available there must not only
be geographic overlap between the insect and host but also a temporal
overlap because most temporary ectoparasites feed only during a well-
defined period of the day. Let us look at two contrasting ways in which
temporal overlap could occur. Some hosts show such efficient defensive
behaviour that temporary ectoparasites can feed only when the host is at
rest. The degree of overlap between the activity period of the insect and the
resting period of such a host will then be important in determining host
choice. In contrast, animals living in deep burrows are unlikely to be hosts
for exophilic insects when at rest, but may be hosts if their activity period
coincides with that of the insect.
Seasonal variation in the choice of host has been recognized in some
blood-sucking insects, and it can have serious consequences for dis-
ease transmission to humans. In North America the arbovirus St Louis
encephalitis is transmitted during the summer months by the mosquitoes
Culex nigripalpus and C. tarsalis. Both these mosquitoes show a marked sea-
sonal change in their feeding patterns, switching from bird feeding in the
winter and spring to mammal feeding in the summer, when arbovirus
24 Feeding preferences of blood-sucking insects

transmission occurs (Edman, 1974; Edman and Taylor, 1968; Tempelis

and Washino, 1967) (see discussion above concerning eastern equine
encephalitis).
Evidence is beginning to appear to suggest that memory can play a role
in host choice. Thus many blood-fed Anopheles arabiensis that had been
artificially transported returned to the houses where they had previously
obtained a blood meal. Some probably flew more than 400 m to do so,
which implies a considerable spatial memory (McCall and Kelly, 2002;
McCall et al., 2001). In addition, as well as returning to successful feeding
sites, vectors may choose their next host on the basis of previous success-
ful feeding experiences. Such imprinting was seen in Culex spp. in field
cages where the tendency was for individual mosquitoes to choose the
same host on which they had previously fed successfully (Mwandawiro
et al., 2000). Significantly this preference was not retained by the offspring,
which is what would be expected if this is indeed learned behaviour.
Such learned behaviour in host selection, if it proves to be widespread,
can have a considerable impact on disease transmission (Kelly and
Thompson, 2000; McCall and Kelly, 2002). For example, a major component
of host selection within a group of hosts may be the level of individual host
defensive behaviour (see above and Chapter 7). Sick hosts tend to display
fewer defensive behaviours and are fed upon most often (Day and Edman,
1983). If this were absolute then no uninfected hosts would ever become
infected. Clearly that is not the case, but nevertheless this aspect of host
selection may be a factor determining the rate of disease transmission in a
population.
It has often been suggested that there are genetically determined
behavioural traits that produce different feeding patterns in different pop-
ulations of a single insect species (rather than the well-recognized phe-
nomenon of different feeding patterns seen among members of species
complexes – see Section 3.2). Most of the support for this is anecdotal but
convincing evidence has been reported for Aedes simpsoni and Ae. aegypti
(Mukwaya, 1977). These species showed direct evidence of both zoophilic
and anthropophilic populations in the laboratory and in the field. It was
established that single species were involved when no crossing or hybrid
sterility appeared during crossing and backcrossing experiments of the
populations studied.

3.2 Host choice and species complexes

For most blood-sucking insects the spectrum of host choice differs with
changing place and season and given different proportions of available
hosts. Selective adaptation to such changes in local circumstance is the fac-
tor driving speciation. This process is continual and as different species are
 3.2 Host choice and species complexes 25

in the process of emerging they are not always easily distinguishable by

their morphological characteristics. Indeed, several blood-sucking insect
‘species’, including Anopheles gambiae sensu lato and Simulium damnosum
s.l., are known to be species complexes made up of reproductively iso-
lated sibling species, having differing biological characteristics. These dif-
ferences commonly include host choice. Let us consider in more detail the
An. gambiae complex, in which the differences and the distribution of the
different members of the complex can be very important in disease control
terms (Besansky, 1999; Coetzee et al., 2000; Coluzzi et al., 2002).
Anopheles gambiae s.l. is an enormously important malaria vector
throughout Africa south of the Sahara. A major filariasis vector, it also
transmits the arboviral disease O’nyong-nyong. As such an important vec-
tor it has received considerable attention from entomologists. Before the
realization that An. gambiae s.l. was a complex of species, the conflicting
nature of the biological data gathered was a source of puzzlement and
controversy. It is now known that the An. gambiae complex is made up of
seven very similar sibling species and even within those divisions varia-
tions are known to occur (Gentile et al., 2002; Lanzaro et al., 1998; Powell
et al., 1999). Pre-copulatory isolation barriers are believed to minimize sib-
ling hybridization (Favia et al., 1997) which would result in sterile male
progeny. These barriers are probably the major factors driving speciation
processes within the complex, but the precise nature of the barriers is not
known. Four of the seven sibling species breed in fresh water: An. gambiae
sensu stricto, which is distributed throughout sub-Sahelian Africa, partic-
ularly the more humid regions; An. arabiensis, which is also found widely
distributed in Africa but shows some preference for drier areas; and An.
quadriannulatus A and B, which show a more restricted range, being found
only in Ethiopia, Zanzibar and parts of southern Africa. A fifth species,
An. bwambae, is only known from mineral springs in the Semliki forest of
Uganda; it is a malaria vector, but because of its narrow geographical range
is of minor importance. Finally there are two species that breed in brackish
water: An. melas, which is found along the west African coast, and its east
African equivalent, An. merus.
All four freshwater species and An. bwambae are morphologically indis-
tinguishable and require cytotaxonomic, biochemical or molecular meth-
ods for identification (Coluzzi et al., 2002; Munstermann and Conn, 1997).
The saltwater species can be identified from the careful use of morpholog-
ical criteria. Because of their differing biological characteristics, the seven
sibling species have very different vectorial capacities. The principal vec-
tor of human malaria and one of the two major vectors of filariasis in the
complex is An. gambiae s.s., which is primarily an endophilic species bit-
ing humans. In many parts of this mosquito’s range humans are the major
available host and, in these circumstances, blood meal identifications often
26 Feeding preferences of blood-sucking insects

reveal that virtually 100 per cent of the meals are of human origin (Davidson
and Draper, 1953). Interestingly, when cattle or other animals are housed
next to, or in, the dwelling place, the proportion of feeds from humans can
fall to 50 per cent or less (Diatta et al., 1998; Killeen et al., 2001; White, 1974;
White and Rosen, 1973), which brings to mind the European malaria story
told in Section 3.1. In complete contrast, An. quadriannulatus is a strongly
zoophilic species and is not a vector of human disease. Over most of its
range it is exophilic, but in the highlands of Ethiopia, probably as an adap-
tation to the cold nights, it is endophilic. The endophilic form may rarely
feed on humans but does not transmit human disease. The fundamental
difference in host choice between An. gambiae s.s. and An. quadriannulatus
has an innate olfactory basis (Dekker et al., 2001). Anopheles arabiensis falls
between the two: ecologically and behaviourally it is an extremely plas-
tic species; across its geographic range, exophilic and endophilic, anthro-
pophilic and zoophilic forms are to be found. Generally, An. arabiensis is
exophilic and zoophilic, but in the absence of other hosts it can live quite
happily on humans and may rest inside their dwellings. It is a malaria
vector but is less efficient than An. gambiae s.s. This can be seen in the
sporozoite rates (the proportion of the mosquito population carrying sporo-
zoites) of An. arabiensis which are commonly about 1/15 of those of An.
gambiae s.s. (White et al., 1972). The higher sporozoite rates in An. gambiae
s.s. are not a reflection of different susceptibilities to malaria parasites – the
two species are equally susceptible – but arise because An. gambiae s.s. lives
longer and feeds more often on humans (White, 1974). Despite being an
important vector of malaria, An. arabiensis is not a major filariasis vector as
it tends to occur seasonally which precludes it from maintaining the high
transmission rates required for the establishment of endemic foci of this
disease (White, 1974). The two saltwater species feed primarily on non-
human hosts but in the absence of these they can survive quite happily on
humans. Of the two species, An. melas feeds more readily and regularly
on humans and is therefore the more important vector of both malaria and
filariasis, particularly in coastal areas where alternative hosts to humans
are scarce. So, even with seven very closely related insects, innate host
preference can vary greatly.
 4
Location of the host

The difficulty that hungry blood-sucking insects have in locating their next
blood meal depends upon the closeness of their association with the host.
At one extreme we have the permanent ectoparasites which are in the
happy position of having food continually ‘on tap’. Only by accident will
they find themselves more than a few millimetres from the skin of the
host and the blood that it holds. At the other extreme are those temporary
ectoparasites, such as blackflies and tabanids, that do not remain perma-
nently in the vicinity of the host. When these insects are hungry their first
problem is to locate the host, often a difficult and complex behavioural
task. These differences in lifestyle are reflected in the number of antennal
receptors different types of blood-sucking insect possess (Chapman, 1982).
Not surprisingly the more independent, host-seeking insects possess the
most receptors. Thus, lice have only 10 to 20 antennal receptors and fleas
about 50, but the stablefly, which spends most of its time at some distance
from the host, has nearly 5000 antennal receptors. Considering two bugs,
we see that Cimex lectularius has only 56 antennal receptors compared to
2900 on the more adventurous Triatoma infestans.
The level of reliance on blood is also an important factor in host loca-
tion. So for obligate haematophages such as the tsetse fly and triatomine
bugs regular host location is absolutely essential. In contrast, facultative
haematophages can often overcome periods when they cannot find a host
by feeding on other foods such as nectar – mosquitoes and stableflies are
examples.
Most of the detailed information on host finding is restricted to a small
number of temporary ectoparasites and the discussion that follows will
concentrate largely on these.

4.1 A behavioural framework for host location

The location of the host is an integrated, but flexible, behavioural pack-
age that gathers momentum as the host is tracked down. The behaviour
patterns involved are not arranged in a strict hierarchy, that is they do
not occur in a strict sequence with behaviour one always being followed
by behaviour two, followed by three, and so on. This versatility allows a
28 Location of the host

flexible response on the part of the insect to the differing circumstances in

which it will encounter hosts. However, it is probable that insects mostly
encounter host-derived stimuli in a particular sequence. The insect often
makes use of this predictability by permitting the current behavioural pat-
tern in the host location sequence to lower the response threshold for subse-
quent host-related stimuli. For example, an insect that would not normally
respond to a certain visual stimulus may respond strongly if it has just been
exposed to an increase in carbon dioxide levels. In this way a behavioural
momentum is built up during host finding. This behavioural momentum
is further enhanced by the wide range of increasingly strong host stimuli
that the insect encounters as host location proceeds (Sutcliffe, 1987).
From observations of blood-sucking insects both in the laboratory and
the field, and from the clear evidence on the discrimination and selectivity
they can show, we can predict that a variety of host signals are used in host
finding. Information on what signals are used and the processes involved
is still far from complete. In general, visual and olfactory stimuli, aided
by anemotactic and optomotor responses, are the most important signals
when the insect is still at some distance from the host. Nearer to the host
different stimuli become important, particularly humidity and heat.
For the purposes of explanation, the various behaviour patterns
involved in host location can be conveniently divided into three phases
(in reality the whole process is a continuum with one behaviour pattern
dovetailing into the next) (Sutcliffe, 1987):
(a) Appetitive searching – driven by hunger the insect indulges in non-
oriented behaviour likely to bring it into contact with stimuli
derived from a potential host. This usually occurs at specific times
of the day regulated by the insect’s internally programmed activity
cycle.
(b) Activation and orientation – upon receipt of host stimuli (activation)
the insect switches from behaviour patterns driven from within
(appetitive searching) to oriented host location behaviour driven
by host stimuli. The insect uses these host-derived stimuli to track
down the host. These stimuli are of increasing variety and strength
as the insect and host come closer together.
(c) Attraction – the final phase, in which host stimuli are used to bring
the insect into the host’s immediate vicinity, and in which the deci-
sion of whether or not to contact the potential host is made.
Categorization of host location in this way has another benefit. It clearly
indicates that host location is a series of behavioural events, rather than just
one. This may help explain the conflicting reports on insect behavioural
responses to certain host stimuli, and at the same time give us a clear warn-
ing of the need for great care in the design of behavioural experiments and
 4.2 Appetitive searching 29

in the interpretation of results from them. The effect of reflected ultraviolet

(UV) light on the tsetse fly can be used to illustrate this point. Reflected UV
light is a deterrent to the fly during the orientation phase of host finding,
but will increase the number of flies landing during the attraction phase. It
would be easy to confuse the two effects in a poorly designed experiment.
Throughout the rest of the chapter, to avoid confusion over which phase
of host location is being discussed, I will use the words appetitive searching,
activation, orientation and attraction only to refer to the different phases
of host finding outlined above.

4.2 Appetitive searching

Blood-sucking insects usually have a delay period between their emer-
gence from the egg, or previous developmental stage, and their first blood
meal. The reasons for this are not clear, especially as other activities, such as
mating and dispersal, commonly occur during this period. One reason for
the delay may be that, after adult emergence, the reproductive system of
many female blood-sucking insects undergoes a maturation period lasting
several days. Blood meals taken before maturation has passed a certain
point do not add to the reproductive success of the insect, but visiting the
host to get them will greatly increase the insect’s chances of being dam-
aged or killed. For example, in the cat flea, Ctenocephalides felis felis, the
reproductive system takes about four days after emergence to mature and
a further two days after the blood meal to produce eggs. It will be selec-
tively advantageous for the insect to remain off the host (and unfed) for the
first four days after emergence, as this will minimize the danger from the
host’s grooming activity without affecting the insect’s reproductive success
(Osbrink and Rust, 1985). Another reason for the delay in blood feeding
may lie in the progressive thickening and hardening of the cuticle that takes
place during the teneral period. This thickening may mean the mouthparts
are insufficiently hard to permit efficient skin penetration. This occurs in
mosquitoes, which for approximately the first 24 hours after emergence
cannot pierce skin. As the post-emergence delay period progresses, or as
the time since the last blood meal lengthens, the insect becomes increas-
ingly hungry and more likely to begin host seeking. Bouts of activity will
be mainly restricted to particular times of the day (Gibson and Torr, 1999)
because activity in blood-sucking insects, as in other animals, occurs in set
patterns during each 24-hour (circadian) cycle. The timing of these activity
bouts is internally programmed and the time of day at which they occur
is characteristic for each species (Fig. 4.1). The patterns are not inviolable;
species commonly show variations in periodicity when collected from dif-
ferent habitats or at different times of the year. As hunger increases these
periods of activity intensify (Fig. 4.2) and also occupy longer periods of time
30 Location of the host

Figure 4.1 Insects will commonly show just a single peak of activity in a day
(Lewis and Taylor, 1965), but there are plenty of exceptions. The tsetse fly, for
instance, shows two endogenously controlled peaks in activity, one at dawn
and one at dusk (Brady, 1975). The key behaviours in the life of the fly most
frequently take place at these times. Some of these are shown in the graph,
expressed as percentages of mean daily response: A. field biting activity; B.
spontaneous flight in actographs; C. optokinetic responsiveness; D. olfactory
responsiveness (Brady, 1975).

(Brady, 1972). Host location during the day or night each has its particular
advantages and disadvantages (Table 4.1).
Much of the activity seen in the hungry insect is appetitive behaviour,
that is behaviour that maximizes the chances of the insect contacting a
signal derived from a host animal. The simplest appetitive behaviour
pattern is to sit still and wait for a host stimulus to arrive. This may
sound a very chancy business, but providing the insect chooses the resting
site carefully, it could be a sound strategy, combining maximum energy
 4.2 Appetitive searching 31

Figure 4.2 Hunger will lead to an increase in the overall activity of the insect.
This can be demonstrated by monitoring various behavioural responses. Here
the changing, kinetic responsiveness of male tsetse flies to a moving ‘target’
(stripe speed) is recorded over a five-day period. As the fly becomes hungrier
it shows increased flight activity in response to a given stimulus. It is also
interesting to note that the pattern of response stays the same throughout
the period, showing that flies are ‘tuned’ to particular patterns of movement
(Brady, 1972).

conservation with a strong chance of encountering a host. This strategy is

likely to be employed to a greater or lesser degree by virtually all blood-
sucking insects, but it is likely to be used more by forest-dwelling species
than by those living in more open terrains (Sutcliffe, 1986). Insects almost
certainly also engage in active appetitive behaviour. The evidence for this
is that non-activating, non-orientating, non-attracting sampling devices,
such as suction traps sunk below ground or electric nets, will often catch
large numbers of hungry female insects that are impregnated but not ready
for egg laying. As these insects are often trapped at peak biting times, it is
quite possible that they are engaged in non-oriented appetitive activity.
We know virtually nothing of the search patterns used by insects under-
taking active appetitive behaviour. Based on the reasonable assumption
that the insect will attempt to optimize its chances of encountering a host
32 Location of the host

Table 4.1 Generalized opportunities and constraints on host location by

blood-sucking insects feeding during the day or night.

Day Night

Disadvantages 1. Greater risk of desiccation 1. Poor visual clues (especially

colour)
2. Greater wind turbulence 2. Low wind speed and hence
poor directional clues in
host-odour plumes
3. Greater risk from predators 3. Greater background levels of
atmospheric carbon dioxide
4. Host mobile (disadvantage for 4. Host less mobile, so sit-and-
odour-responding insects?) wait strategies less feasible
5. Greater risk from defensive
behaviour of active host

Advantages 1. Good visual clues 1. Less risk of desiccation

2. Higher wind speeds providing 2. Host more likely to be at rest
good directional clues in odour so reduced risk from host
plumes defensive behaviour
3. Reduced background levels of 3. Less risk from predators
atmospheric carbon dioxide
4. Host mobile, making a 4. Less atmospheric turbulence
sit-and-wait strategy feasible and hence more continuous
odour plumes

Gibson and Torr, 1999.

while minimizing its energy expenditure, theoretical work has been car-
ried out to determine such patterns for flying insects. In a wind blowing
consistently from one direction the optimum strategy is for the insect to
fly across the wind, allowing the maximum number of air streams to be
monitored for a particular energy expenditure (Linsenmair, 1973). In the
field, winds often veer rapidly from one direction to another. If they veer
by more than 30◦ from the mean, then downwind flight becomes the most
energy-efficient method of sampling the maximum number of airstreams
(Sabelis and Schippers, 1984). Whether these theoretical conclusions relate
to the appetitive behaviours displayed by blood-sucking insects in the field
remains to be seen.

4.3 Activation and orientation

Activation occurs when the insect comes into contact with a suitable sig-
nal from a potential host animal. Such a signal may simply change the
 4.3 Activation and orientation 33

behavioural awareness of the insect without causing any observable activ-

ity on the insect’s part. In such a case, orienting behaviour would be
released by a subsequent stimulus, for which the insect has now been
primed. Alternatively, the stimulus may directly cause the insect to switch
over from endogenously driven appetitive searching to oriented host loca-
tion behaviour. The insect then uses the information contained in host-
derived signals to orientate towards the host. A range of stimuli are used
by insects in activation and orientation.

Olfaction
There seems to be an olfactory component to host finding in virtually all
blood-sucking insects (Takken, 1996). It might be assumed that it is of
most importance to forest-dwelling insects, because direct visual contact
with the host is most restricted for them. But the evidence from tsetse
flies suggests that the species showing the clearest response to odours are
those (Glossina morsitans, G. pallidipes and G. longipennis) living in fairly
open situations. The forest-dwelling palpalis group show little response at
a distance. This is probably explained by the fate of the odour plume in
the two situations, with a plume at the flight height of insects acting as
a much more continuous, relatively linear guide for insects in an open
situation compared to a forest (David et al., 1982; Elkinton et al., 1987).
Field experiments on plume structure show that an insect following an
odour plume could be flying >90◦ away from the host for up to 25 per cent
of the time when following an odour plume passing through vegetation,
even at 5 m from the source (Brady et al., 1989). So odour is probably most
important in orientation for night-feeding forms living in relatively open
situations.
Olfactory stimuli implicated in host location to date include carbon diox-
ide, lactic acid, ammonia, acetone, butanone, fatty acids, indole, 6-methyl-
5-hepten-2-one and phenolic components of urine (Geier et al., 1999;
Klowden et al., 1990; Knols et al., 1997; Meijerink et al., 2000). For tsetse
flies the most potent chemicals affecting behaviour are carbon dioxide,
acetone, octenol, butanone and various phenols (Gibson and Torr, 1999),
but when these are dispensed in the field at natural dose rates they attract
only about 50 per cent of the tsetse a natural host would attract (Hargrove
et al., 1995; Torr et al., 1995). This suggests that other kairomones remain to
be identified.
It is generally accepted that carbon dioxide can be involved in both
the activation and orientation of virtually all blood-sucking insects. It is
normally present in the atmosphere at between about 0.03 per cent and
0.05 per cent, occasionally rising to 0.1 per cent in dense vegetation at night.
It is secreted by the skin of hosts, but the major emissions occur in exhaled
breath which, in humans, contains about 4.5 per cent carbon dioxide. So
34 Location of the host

Figure 4.3 The effect of varying the emission rate of carbon dioxide on its
drawing power for mosquitoes has been measured by various authors under
field conditions. The lower edge of the ‘attraction range’ shown in the figure
is the furthest trapping point at which an effect of the carbon dioxide was
noted. The upper edge of the zone is the nearest trap at which no effect of
the carbon dioxide was seen (Gillies, 1980).

the carbon dioxide in an odour plume produced from a solitary human

would remain above background concentration until the exhaled breath
had been diluted by a factor of about 100 (Gillies, 1980). Field recordings
of tsetse receptor responses to odours support the idea that odour plumes
break up into filaments and packets under field conditions, with the odour
concentration and frequency of packets decreasing with distance from the
host (Voskamp et al., 1998). Predicting how dilution of the plume will occur
is no easy task. Packets of relatively undiluted odour are likely to travel for
quite some distance downwind, and the distance from the host at which
activation and orientation are still likely to occur will vary according to local
meteorological conditions, with windspeed being particularly important
(Brady et al., 1995; Griffiths and Brady, 1995). Perhaps the easiest way of
dealing with the problem is to look directly at the effects of odour in the field
(Fig. 4.3). The range over which a host animal can activate and orientate
an insect, on the basis of odour alone, has been calculated for some blood-
sucking insects. An ox draws the tabanid Philoliche zonata at 80 m, but this
Other documents randomly have
different content
already disappeared into the valley beyond the enemy's position. By
the time that we reached the top of the high ground, Bruce-
Hamilton's leading battalions were nearly a mile further on, and the
tail of Broadwood's brigade was vanishing in a high cloud of dust to
the eastward. The City Imperial Volunteers, who had lost a few men
in the attack, were resting on the hill after their advance, and eating
their biscuits. Several dead Boers had been found lying among the
rocks, and a burial party was at work digging a grave for these and
for four of our own men who had fallen close by. There were also a
few prisoners--Transvaalers for the most part--who had surrendered
when the troops fixed bayonets. Four miles away to the north-east
the trees and houses of Ventersburg rose from a grassy hollow.
The General decided to bivouac in the valley beyond the
enemy's position, and to set his pickets upon the hills to the
northward. He also sent an officer with a flag of truce into
Ventersburg to demand the surrender of the town, and directed
Broadwood to detach a regiment and some Mounted Infantry to
occupy it, should the enemy comply. In case they should desire to
hold the town the 5-inch guns were brought into position on the
captured heights.
Hoping to secure some supplies, particularly bottled beer, before
everything should be requisitioned by the army, I rode forward after
the flag of truce had gone in and waited where I could see what
followed. When, about an hour later, a cavalry force began to
advance from the direction of Broadwood upon the town, I knew
that all was well, and trotted on to join them. My road led me within
a few hundred yards of the town, but, luckily for me, I did not enter
it alone, and hurried to join the troops. All of a sudden the ominous
patter of rifle shots broke the stillness of the evening, and, turning
to whence the sound came, I saw a score of Boers standing on the
sky-line about a mile away and firing at the advancing Cavalry, or,
perhaps, for I was much nearer, at me. The next minute there
galloped out of the town about a score of Dutchmen, who fled in the
direction of their friends on the western sky-line. Had I ridden
straight into the town I should have run into these people's jaws. I
lost no time in joining the Cavalry, and entered the streets with the
squadron of Blues. It was a miserable little place, not to be
compared with Winburg. There were a few good stores and a small
hotel, where I found what I sought; but the whole town was very
dirty and squalid. Thirty or forty troopers of Roberts's Horse were
firing at the fugitive burghers from the edge of the buildings and
gardens, while a score of reckless fellows were galloping after them
in excited pursuit. The Boers on the hill kept up a brisk fire to help
their comrades in, and not a few of the bullets kicked up the dust in
the village streets, without in the least disturbing the women and
children who crowded together to look at the war, in blissful
ignorance of their danger. When some of these people were told that
they would perhaps be killed if they came out of their houses while
the fighting was going on, they clutched their children and sought
shelter with an energy at which, since, after all, nobody was hurt, it
was pardonable to laugh.
Night put an end to all skirmishing, and under its cover the
Boers retreated--the greater part to Kroonstadt, which, be it
remembered, they meant to hold to the death; but a considerable
proportion to the east, where they collected with the commandos
under Christian de Wet. Broadwood's brigade had captured about a
dozen waggons and thirty prisoners. In all there were fifty-two
unwounded and seven wounded Boers in our hands at the end of
the day. The casualties in Hamilton's force were under fifty. Tucker
and Pole-Carew may have lost the same number between them.
French, who encountered the most stubborn resistance, had a little
over 120. But, in any case, the passage of the Sand River in this
long straggling action was cheaply won at a cost of under 250
officers and men.
All our beasts were so exhausted by the labour of dragging the
waggons through the steep and rocky drift of the Sand, and by the
long pull up the hills on the opposite side, that few of the regiments
got their baggage that night, and hence it was impossible to make
an early start next morning. But it was known that the Field-Marshal
meant to reach Kroonstadt on the next day, and as all the
information at our disposal indicated that the Boers were
entrenching a strong position along a line of wooded bluffs called
the Boschrand, just south of the town, every minute of halt was
grudged.
We moved at eleven o'clock, heading direct for Kroonstadt, and
persevered for two hours after the sun had set, making in all nearly
seventeen miles. The country to our left was flat and open, and as
we converged upon the main army we could see, like red clouds
with the sunset behind them, the long parallel lines of dust, which
marked the marches of the Seventh and Eleventh Divisions; and we
knew besides, that, beyond both columns and west of the railroad,
French was driving his weary squadrons forward upon another wide
swoop. The army drew together in the expectation of a great action.
But for all our marching we could never make up the extra distance
we had to cover in coming diagonally from the flank, and as
darkness fell we realised that the Seventh Division was drawing
across our front, and that Pole-Carew with the guard was striding
along ahead of us all. That night Lord Roberts slept at America
Siding, scarcely six miles from the Boschrand position.
Ian Hamilton marched on again at dawn, transport and convoys
struggling along miles behind, and the fine-drawn yet eager Infantry
close upon the heels of the Cavalry screen. At times we listened for
the sound of guns, for if the enemy stood, the Field-Marshal must
come into contact with them by eight o'clock. And when, after nine
o'clock, no cannonade was heard, the rumour ran through the army
that the Boers had fled without giving battle, the pace slacked off,
and the Infantry began to feel the effects of their exertions.
At eleven a message from Lord Roberts reached General
Broadwood to say that it did not matter by which road Hamilton's
column marched in, as the enemy was not holding his positions.
Thereupon I determined, since there was to be no battle, to see the
capture of Kroonstadt, and being mounted on a fresh pony I had
bought at Winburg, a beautiful and tireless little beast, by an English
blood sire out of a Basuto mare, I soon left the Cavalry behind,
caught up the rear of Tucker's transport, pushed on four or five
miles along the line of march of his division, struck the tail of the
Eleventh Division, and finally overtook the head of the Infantry
columns about three miles from the town.
Lord Roberts entered Kroonstadt at about mid-day with all his
staff. The Eleventh Division, including the Guards' Brigade, marched
past him in the market square, and then, passing through the town,
went into bivouac on the northern side. The rest of the army halted
south of Kroonstadt. Gordon's Cavalry Brigade a mile from the town;
the Seventh Division and Ian Hamilton's force three miles away, in a
wide valley among the scrub-covered, trench-rimmed hills the Boers
had not dared defend. French, whose turning movement had again
been obstinately opposed, reached the railway line north of the town
too late to intercept any rolling stock. Indeed, Major Hunter Weston,
a daring and enterprising engineer, arrived at the bridge he had
hoped to blow up only to find that it had been blown up by the
enemy.
Thus, by one long spring from Bloemfontein, Kroonstadt, the
new capital of the Free State, was captured. It has the reputation of
being one of the prettiest places in the Republic, but even when
allowances are made for the circumstances under which we saw it, it
does not seem that its fame is just. The town looked a little larger
than Winburg, though not nearly so clean and well-kept, and the
whole place was smothered in reddish dust, and dried up by the sun.
The Boers retreated northward along the railway, in spite of all
President Steyn's exhortations, which included the public
sjambokking of several unwilling burghers, and did not stop except
to wreck the permanent way until they reached Rhenoster kopjes.
The President, with the members of the Executive Council and the
seat of Government--which needs to have a good pair of legs
beneath it in times like these--withdrew to Lindley, whither, for
various reasons, it soon became desirable to follow them.

CHAPTER XI
LINDLEY
 Heilbron: May 22, 1900.

Having arrived thus prosperously at Kroonstadt, Lord Roberts

determined to halt until his supplies were replenished and the
railway line from Bloemfontein in working order. Moreover, in the
expectation of a general action outside the town, he had
concentrated all his troops and had drawn the Army of the Right
Flank close in to the main force. Before he advanced again towards
the enemy's position on the Rhenoster River, he wished to extend his
front widely, as he had done in the previous operation. The scheme
of advance by converging columns required a pause after each
concentration before the movement could be repeated; so that while
the Field-Marshal himself remained stationary his energetic
Lieutenant was again on the move.
General Ian Hamilton, with the same troops as before and an
addition of four 'pom-poms,' started from his camp outside
Kroonstadt on the 15th, and after a short march encamped on the
eastern side of the town preparatory to moving on Lindley, whither
President Steyn had withdrawn. The question of supplies was a very
troublesome one, and it was no light matter to thrust out fifty miles
into a hostile country with only three and a half days' food and
forage in hand. Suppose anything should happen to the convoys
which were to follow. Meat in plenty could be found everywhere, but
the stores of flour and other farinaceous goods which the farm-
houses might contain were insufficient and precarious. Even the
benefits of the abundant meat supply were to some extent
discounted by the scarcity of wood, for it is not much satisfaction to
a soldier to be provided with a leg of mutton if he has no means of
cooking. The deficiencies were hardly made good by the arrival of a
small convoy, the greater part of which consisted of disinfectants for
standing camps, and the rest--so valuable in a grass country--of
compressed hay.
Nevertheless, being determined, and trusting, not without
reason, in his supply officer, Captain Atcherley, Hamilton started on
the 16th, and the Infantry bivouacked eighteen miles from
Kroonstadt on the Lindley road--it would perhaps be less misleading
to write track. The Cavalry brigade with one corps of Mounted
Infantry under Broadwood were pushed ten miles further on, and
seized a fine iron bridge, not marked on any map, which spans an
important spruit at Kaalfontein. Here trustworthy information was
received that a large force of Boers with guns was retreating before
Rundle's column (Eighth Division) northwards upon Lindley, and
deeming it important to occupy the town before they arrived,
Hamilton ordered the Cavalry to hurry on and take possession of the
heights to the north of it. It was a double march when ordinary
marches were long. The result, however, justified the effort.
Broadwood 'surprised'--the word is taken from the Boer accounts--
Lindley on the 17th. Scarcely fifty Boers were at hand to defend it. A
waggon with 60,000*l.* in specie barely escaped from the clutches
of the Cavalry. After a brief skirmish the town surrendered. The
British loss was three men wounded. Broadwood then retired as
directed by his chief to the commanding hill to the north to bivouac.
This hill may for convenience be called 'Lindley Hill' in the
subsequent narrative.
 The Infantry and baggage also made a long march on the 17th,
but as the road was obstructed by several bad spruits or dongas,
they were still fourteen miles from Lindley when night closed in.
Even then the transport was toiling on the road, and a large part of
it did not come in, and then in an exhausted condition, until after
midnight. I wonder how many people in England realise what a
spruit is, and how it affects military operations. Those who live in
highly developed countries, where the surface of the earth has been
shaped to our convenience by the patient labour of many years, are
accustomed to find the road running serenely forward across the
valleys, and they scarcely notice the bridges and culverts over which
it passes. All is different in South Africa. The long column of
transport trails across the plain. The veldt in front looks smooth and
easy going. Presently, however, there is a block. What is the matter?
Let us ride forward to see: and so onward to where the single string
of waggons merges in a vast crowd of transport, twenty rows
abreast, mule carts, Cape carts, ox waggons, ambulances, and
artillery, all waiting impatiently, jostling each other, while drivers and
conductors swear and squabble. Here is the spruit--a great chasm in
the ground, fifty feet deep, a hundred yards from side to side. The
banks are precipitous and impassable at all points except where the
narrow single track winds steeply and unevenly down. The bottom is
a quagmire, and though the engineers are doing their best to level
and improve the roadway, it is still a combination of the Earl's Court
water chute and the Slough of Despond. One by one, after a hot
dispute for precedence, the waggons advance. The brakes must be
screwed up to their tightest grip lest the ponderous vehicles rush
forward down the slope and overwhelm their oxen. Even with this
precaution the descent of each is a crash, a scramble, and a bump.
At the bottom like a feather-bed lies the quagmire. Here one waggon
in every three sticks. The mules give in after one effort--unworthy
hybrids. The oxen strain with greater perseverance. But in the end it
is the man who has to do the hauling. Forthwith come fatigue
parties of weary men--it has been a long march already to soldiers
fully equipped. Drag ropes are affixed, and so with sweat, blood, and
stretching sinew, long whips cracking and whistling, white men
heaving and natives yelping encouragement, another waggon comes
safely through. And there are seven miles of transport!
On the morning of the 18th the Infantry were about to move
off, when a patter of rifle shots to the north of the road reminded us
of the presence of the enemy. A foraging party of Major Rimington's
Guides had ridden up to a farm, which stood in full view of the camp
and flew (or was it hoisted afterwards?) a white flag. Arrived there,
they were received by a volley from five Boers in hiding near.
Conceive the impudence of these people: five Boers, within a mile of
eight thousand British and a powerful Cavalry force, fire on a
foraging party! Luckily no harm done; Cavalry gallop out angrily;
Boers vanish among remoter kopjes. 'But,' said the General, 'what
about my convoys?'
So it was arranged that Smith-Dorrien should be left where he
was (twelve miles west of Lindley) with his own brigade, one battery,
and a corps of Mounted Infantry to help in the expected convoy, and
should cut off the corner and rejoin the column at the end of its first
march towards Heilbron. Ian Hamilton with the rest of the troops
then moved on to Lindley. The march lay through the same class of
country hitherto traversed--a pleasant grassy upland which, if not
abundantly supplied with water by nature, promised a rich reward to
man, should he take the trouble to construct even the simplest
irrigation works. Spruits ran in all directions, and only required an
ordinary dam, like the bunds the peasants build in India, to jewel
each valley with a gleaming vivifying lake. The husbanding of water
would repair the scarcity of wood, and the tenth year might see the
naked grass clothed and adorned with foliage. But at present the
country-side is so sparsely populated that the energies of its
inhabitants could not produce much effect upon the landscape. The
unamiable characteristic of the Boer, to shun the sight of his
neighbour's barn, has scattered the farms so widely that little
patches of tillage are only here and there to be seen, and the
intervening miles lie neglected, often not more than twenty acres of
a six thousand acre property being brought into cultivation, which
seems rather a pity.
The fair face of the land under its smiling sky was not unmarked
by the footprints of war. In the dry weather the careless habits of
the soldiers were the constant cause of grass fires. The half-burnt
match, tossed idly aside after a pipe was lighted, or an unguarded
spark from a cooking fire, kindled at once an extensive conflagration.
The strong winds drove the devouring blaze swiftly forward across
the veldt, clouding the landscape by day with dense fumes of smoke
and scarring the scene by night with vivid streaks of flame. So
frequent were these grass fires that they became a serious nuisance,
wasting in an hour many acres of grazing, proclaiming the
movement and marking the track of the army, stifling the marching
columns with pungent odours, destroying the field telegraph, and
only extinguished by the heavy dews of the early morning. But in
spite of repeated injunctions in the daily orders, the accidents--for
which, indeed, there was every excuse--continued, and the plains of
brownish grass were everywhere disfigured with ugly patches of
black ashes which, as the fires burnt outwards, would spread and
spread, like stains of blood soaking through khaki.
At length the track, which had been winding among the smooth
undulations, rounded an unusually steep hillock of kopje character,
and we saw before us at the distance of a mile the pretty little town
of Lindley. The Cavalry bivouacs covered the nearer slopes of the
high hill to the northward. The houses--white walls and blue-grey
roofs of iron--were tucked away at the bottom of a regular cup, and
partly hidden by the dark green Australian trees. We rode first of all
to Broadwood's headquarters, following the ground wire which led
thither. Arrived there we learned the news. Boer laagers and Boer
patrols had been found scattered about the country to the south-
east and north-east. There was occasional firing along the picket
line. The town had upon most searching requisition yielded nearly
two days' supply, and, most important of all, Piet De Wet, brother of
the famous Christian, had sent in a message offering to surrender
with such of his men as would follow his example, if he were
permitted to return to his farm. Broadwood had at once given the
required assurance, and Hamilton on his arrival had wired to Lord
Roberts fully endorsing the views of his subordinate, and requesting
that the agreement might be confirmed. The answer came back with
the utmost despatch, and was to the effect that surrender must be
unconditional. De Wet, it was remarked, was excluded from the
favourable terms of the Proclamation to the Burghers of the Orange
Free State, by the fact that he had commanded part of the
Republican forces. He could not therefore be permitted to return to
his farm. I need not say with what astonishment this decision was
received. The messenger carrying the favourable answer was luckily
overtaken before he had passed through our picket line and the
official letter was substituted. Piet De Wet, who awaited the reply at
a farm-house some ten miles from Lindley, found himself presented
with the alternative of continuing the war or going to St. Helena, or
perhaps Ceylon; and as events have shown he preferred the former
course to our loss in life, honour, and money.
In the afternoon I rode into Lindley to buy various stores in
which my waggon was deficient. It is a typical South African town,
with a large central market square and four or five broad unpaved
streets radiating therefrom. There is a small clean-looking hotel, a
substantial gaol, a church and a schoolhouse. But the two largest
buildings are the general stores. These places are the depôts
whence the farmers for many miles around draw all their necessaries
and comforts. Owned and kept by Englishmen or Scotchmen, they
are built on the most approved style. Each is divided into five or six
large well-stocked departments. The variety of their goods is
remarkable. You may buy a piano, a kitchen range, a slouch hat, a
bottle of hair wash, or a box of sardines over the same counter. The
two stores are the rival Whiteley's of the country-side; and the
diverse tastes to which they cater prove at once the number of their
customers, and the wealth which even the indolent Boer may win
easily from his fertile soil.
Personally I sought potatoes, and after patient inquiry I was
directed to a man who had by general repute twelve sacks. He was
an Englishman, and delighted to see the British bayonets at last.
'You can't think,' he said, 'how we have looked forward to this day.'
I asked him whether the Dutch had ill-used him during the war.
'No, not really ill-used us; but when we refused to go out and
fight they began commandeering our property, horses and carts at
first and latterly food and clothing. Besides, it has been dreadful to
have to listen to all their lies and, of course, we had to keep our
tongues between our teeth.'
It was evident that he hated the Boers among whom his lot had
been cast with great earnestness. This instinctive dislike which the
British settler so often displays for his Dutch neighbour is a
perplexing and not a very hopeful feature of the South African
problem. Presently we reached his house (where the potatoes were
stored). Above the doorway hung a Union Jack. I said--
'I advise you to take that down.'
'Why?' he asked, full of astonishment.
'The British are going to keep the country, aren't they?'
'This column is not going to stay here for ever.'
'But,' with an anxious look, 'surely they will leave some soldiers
behind to protect us, to hold the town.'
I told him I thought it unlikely. Ours was a fighting column.
Other troops would come up presently for garrison duty. But there
would probably be an interval of at least a week. Little did I foresee
the rough fighting which would rage round Lindley for the next three
months. He looked very much disconcerted; not altogether without
reason.
'It's very hard on us,' he said after a pause. 'What will happen
when the Boers come back? They're just over the hill now.'
 'That's why I should take the flag down if I were you. If you
don't fight, keep your politics till the war is over!' He looked very
disappointed, and I think was asking himself how much his
enthusiasm had compromised him. After we had settled the potato
question to his satisfaction and I had sent the sack away upon my
pack pony, he perked up. 'Come and see my garden,' he said, and
nothing loth I went. It was not above a hundred yards square, but
its contents proclaimed his energy and the possibilities of the soil.
He explained how he had dammed a marshy sluit in the side of the
hills to the eastward. 'Plenty of water at all seasons: this pipe you
see, only a question of piping: as much water as ever I want: twenty
gardens: grow anything you like, potatoes mostly, cabbages (they
were beauties), tomatoes and onions, a vine of sweet white grapes,
a bed of strawberries over there--anything: it only wants water, and
there's plenty of that if you take the trouble to get it.'
The signs of industry impressed me. 'How long,' I asked, 'have
you been here?'
'Eight years last February,' he replied; 'see those trees?'
He pointed to a long row of leafy trees about twenty feet high,
which gave a cool shade and whose green colour pleased the eye
after looking at so much brown grass. I nodded.
'I planted those myself when I came: they grow quickly, don't
they? Only a question of water, and that is only a question of work.'
Then I left him and returned to the camp with my potatoes and
some information thrown in.
The next morning before breakfast-time there was firing in the
picket line south of Lindley. The patter of shots sounded across the
valley, and upon the opposite slopes the British patrols could be seen
galloping about like agitated ants. I was at the moment with General
Hamilton. He watched the distant skirmish from his tent door for a
little while in silence. Then he said:
'The scouts and the Kaffirs report laagers of the enemy over
there, and over there, and over there' (he pointed to the different
quarters). 'Now either I must attack them to-day or they will attack
me to-morrow. If I attack them to-day, I weary my troops; and if I
don't we shall have to fight an awkward rear-guard action to get out
of this place to-morrow.'
He did not say at the time which course he meant to follow, but
I felt quite sure he would not take his troops back very far to the
south or south-east to chastise impalpable laagers. We were running
on schedule time and had to make our connections with the main
army, to securing whose smooth and undisturbed march all our
efforts must be directed. So I was not surprised when the day
passed without any movement on our part.
Very early on the 20th the brigades were astir, and as soon as
the light was strong Broadwood's Cavalry began to stream away
over the northern ridges. The guns and the greater part of the
Infantry followed them without delay, so that by seven o'clock the
great column of transport was winding round the corner of Lindley
Hill on the road to Heilbron. The fact that parties of the enemy had
been observed on all sides except the west, made the operation of
disentangling the force from Lindley difficult and dangerous.
Broadwood's duty was to clear the way in front. Legge's corps of
Mounted Infantry guarded the right flank: and Ian Hamilton himself
watched the movement of the rear guard, which consisted of the
Derbyshire Regiment, Bainbridge's corps of Mounted Infantry and, as
a special precaution, the 82nd Field Battery.
The full light of day had no sooner revealed the march of the
troops than the watching Boers began to feel and press the picket
line: and an intermittent musketry spread gradually along the whole
three quarter circle round Lindley. At eight o'clock our troops
evacuated the town itself, at nine, the convoy being nearly round
Lindley Hill, the pickets commenced to draw in. This was a signal for
decided increase in the firing. No sooner were the outposts clear of
the town than the Boers in twos and threes galloped into it and
began to fire from the houses. All kinds of worthy old gentlemen,
moreover, who had received us civilly enough the day before,
produced rifles from various hiding-places and shot at us from off
their verandahs. Indeed, so quickly did the town revert to the
enemy's hands that Somers Somerset, the despatch rider of the
'Times,' was within an ace of being caught. He had arrived late the
night before, and having found a comfortable bed at the hotel went
to sleep without asking questions. The next thing he remembers is
the landlord rushing into his room and crying in great excitement
that the Boers were in the town. He scrambled into his clothes and,
jumping on his horse galloped through the streets and was not fired
at till he was more than a quarter of a mile away. History does not
record whether among such disturbing events he retained his
presence of mind sufficiently to settle his hotel bill.
The General and his staff had watched the beginnings of the
action from the now deserted camping ground, a dirty waste, littered
with rubbish and dotted with the melancholy figures of derelict
horses and mules. So soon as the retiring pickets drew north of the
town, he mounted and made his way to the top of Lindley Hill. From
this commanding table-top the whole scene of action, indeed the
whole surrounding country, was visible. At our feet beyond the
abandoned bivouac lay the houses of Lindley giving forth a regular
rattle of musketry. On either side, east and west, rose two prominent
kopjes held by companies of Mounted Infantry briskly engaged. The
tail of the transport serpent was twisting away into safety round the
base of our hill. Far away on the broad expanse of down parties of
Dutch horsemen cantered swiftly forward; and along a road beyond
the eastern kopje rose a steady trickle of mounted men. They
moved in true Boer fashion--little independent groups of four and
five, now and then a troop of ten or a dozen, here and there a
solitary horseman riding back against the general flow. At no
particular moment were more than thirty to be seen on the mile of
dusty road. Yet to an experienced eye the movement seemed full of
dangerous significance. One became conscious of a growing
accumulation of force somewhere among the hills to the eastward.
The General, who had served on the Indian frontier, understood
rear-guard actions, and his face was grave, as I had not seen it
when larger operations were toward; and at this moment the boom
of a heavy gun told us that the advanced troops were also engaged.
The Boers knew what they wanted. There was an air of decision
about their movements which boded no good to rear or right flank
guard. Gallopers were sent off, one to warn the right corps of
Mounted Infantry, another to bid the main body of the force go dead
slow, another to the threatened eastern kopje to learn the state of
affairs there. The rear-guard battery was brought up on to the table-
top, and came into action. This was, I think, the key of the situation.
The battery planted on Lindley Hill, and casting its shells now in one
direction, now in another, compelled the assailants to keep their
distance, and helped the pickets into safety and new positions
further back. It called to mind some famous knight of history or
romance holding an angry rabble back beyond the sweep of his long
sword, while his comrades made good their retreat. Under this good
protection the pickets, having dutifully held their positions until the
convoy was well on its road, scampered in, and the battery itself
began to think about retiring. But the trickle of Boers along the
eastern roadway had not stopped. Seven or eight hundred men must
have passed already; and those that now came galloped as if they
had some very tangible objective. 'Look out, the right flank!'
But now, the rear guard having disengaged itself from Lindley
town, the General's place was with his main body, and we set off to
trot and gallop the seven miles that intervened between the head
and tail of our force. The firing in front had ceased before we came
up. Indeed, the affair had not been of any importance. About seven
hundred Boers with three or four guns had obstructed the advance
near the Rhenoster River; had even checked the Cavalry screen;
Tenth Hussars had two officers wounded; a dozen other casualties in
the Brigade; Infantry and guns wanted to clear the way. A Cavalry
brigade is not a kopje-smashing machine. 'Never mind, here come
the cow-guns. Now we shall see.' Indeed, as soon as the head of the
21st Brigade began to deploy, the five-inch guns and a field battery
opened on the enemy, who thereupon fled incontinently across the
river, pursued by the fire of the guns and of the Cavalry 'pom-poms.'
We were just congratulating ourselves upon the success of
these curious operations--curious because the drill books do not
contemplate both sides fighting rear-guard actions at the same time-
-when half a dozen riderless horses galloped in from somewhere
miles away on the right flank. Evidently sharp fighting was
proceeding there; the flow of Boers had meant mischief. The
peaceful landscape told no tale. No sound of musketry, nor sign of
action could be distinguished. Indeed, in this scattered warfare one
part of a force may easily be destroyed without the rest even
knowing that a shot has been fired. 'Why scatter them?' asks the
armchair strategist. 'Because if you don't scatter, and haven't got
soldiers who are good enough to act when scattered, you will all get
destroyed in a lump together.'
The General sent directions to the rear guard to communicate
with the flank guard; kept another corps of Mounted Infantry handy
to support either if necessary, and turned his attention to getting his
brigades across the Rhenoster River. While this was proceeding the
head of Smith-Dorrien's column, which had marched prosperously
from their bivouac near Kaalfontein, came into view, and the Army of
the Right Flank stood again united, a fact which suggests some
consideration of its functions in the general scheme of Lord Roberts's
advance.
After Kroonstadt had been captured the republican forces on the
railway retreated to the line of the Rhenoster. Half a mile to the
north of this river there rises abruptly from the smooth plain a long
line of rocky hills, and in this strong position the Boers had
determined to make a stubborn stand. Any force advancing along
the railway would indeed have found it a difficult and costly business
to cross the river and dislodge an enemy so posted. Other low hills
trending away to either flank would have made any turning
movement an exceedingly extended and probably a useless
operation, for the enemy being on the inside of the circle would have
been able to confront the attack wherever it might fall. But the
Rhenoster River, as the reader will see by a glance at the map, rises
considerably south of the point where it intersects the railway; and
so soon as Ian Hamilton's force was across it, the Boers holding the
kopjes position were in considerable danger of being cut off. The
effect of our crossing the Rhenoster between Lindley and Heilbron
should therefore be to clear the march of the main army. All fell out
as Lord Roberts had expected; although the Boers had made great
preparations to defend Rhenoster, had constructed strong
entrenchments and made sidings to detrain their heavy guns, they
evacuated the whole position without a shot being fired, compelled
by the movement of a column forty miles away to their left flank.
All who understood the scope and cohesion of the operations
were delighted at the prospect of getting across the Rhenoster River.
The General was determined, rear and flank guard actions
notwithstanding, to have his army and transport over that night: and
two practicable crossings having been found, Infantry, Cavalry, guns
and baggage began to push across. The last was now increased by
the arrival of Smith-Dorrien, who brought with him a much needed
convoy with sufficient supplies to carry us on to Heilbron and a
march beyond. It was midnight before all the waggons were across;
but though this cruel day of march and sun tore the hearts out of
the transport animals, and the flocks of sheep were so weary they
could scarcely be driven along, we knew that the exertions had not
been made in vain.
 Late in the evening came the news from the right flank guard.
They had waited, fearing to expose the rear guard to a flank attack.
The rear guard had made good its retreat. A gap had sprung up
between the two bodies. The vigilant Boers had pounced in and
stampeded the horses of one Mounted Infantry company. A sharp,
fierce fight followed; rear guard hearing the fusillade swung in to
help. Ultimately the Boers were checked sufficiently to enable rear
and flank guards to cut inwards together and draw off: but it was by
general agreement of participants a very unpleasant affair. The
officer commanding the company whose horses were stampeded
had particularly interesting experiences. The Boers galloped right in
among his men, and a confused scrimmage followed: officer was
running towards stampeded horses; on the way he passed a
burgher; 'Surrender,' cried the Dutchman. 'No,' retorted the officer--
an Irishman--(with suitable emphasis) and ran on, whereupon
burgher dismounted and began shooting; had four shots and missed
every one. Meanwhile officer reached shelter of a convenient rock,
turned in just indignation, fitted his Mauser pistol together and fired
back. The burgher, finding his enemy behind cover, and himself in
the open--by no means the situation for a patriot--jumped on his
horse, and would have galloped away but that the officer managed
to hit him in the leg with his pistol, and so he dropped, according to
the account of an eye-witness, 'like a shot rook.'
The local advantage, however, rested with the Boers, who hit or
captured the greater part of the squadron, including twenty
wounded. Concerning these latter, Piet De Wet sent in a flag of truce
during the night offering to hand them over if ambulances were
sent, and several wounded Boers whom we had taken were given
up. This was accordingly done. Our total losses during the 20th were
about sixty, some of whom were officers. The Boers admitted a loss
of twenty killed and wounded, and it may easily have been more.
The army bivouacked on the north bank of the Rhenoster within two
marches of the town of Heilbron, upon which it was now designed to
move.

CHAPTER XII
CONCERNING A BOER CONVOY

Heilbron: May 22.

Heilbron lies in a deep valley. About it on every side rolls the grassy
upland country of the Free State, one smooth grey-green surge
beyond another, like the after-swell of a great gale at sea; and here
in the trough of the waves, hidden almost entirely from view, is the
town itself, white stone houses amid dark trees, all clustering at the
foot of a tall church spire. It is a quiet, sleepy little place, with a few
good buildings and pretty rose gardens, half-a-dozen large stores, a
hotel, and a branch line of its own.
For a few days it had been capital of the Free State. The
President, his secretaries, and his councillors arrived one morning
from Lindley, bringing the 'seat of government' with them in a Cape
cart. For nearly a week Heilbron remained the chief town. Then, as
suddenly as it had come, the will-o'-the-wisp dignity departed, and
Steyn, secretaries, councillors, and Cape cart, hurried away to the
eastward, leaving behind them rumours of advancing hosts--and (to
this I can testify) three bottles of excellent champagne. That was on
Sunday night. The inhabitants watched and wondered all the next
day.
On the Tuesday morning, shortly after the sun had risen,
Christian De Wet appeared with sixty waggons, five guns, and a
thousand burghers, very weary, having trekked all night from the
direction of Kroonstadt, and glad to find a place of rest and
refreshment. 'What of the English?' inquired the new-comers, and
the Heilbron folk replied that the English were coming, and so was
Christmas, and that the country to the southward was all clear for
ten miles. Thereat the war-worn commando outspanned their oxen
and settled themselves to coffee. Forty minutes later the leading
patrols of Broadwood's Brigade began to appear on the hills to the
south of the town.
Looked at from any point of view, the British force was a
formidable array: Household Cavalry, 12th Lancers and 10th
Hussars, with P and Q Batteries Royal Horse Artillery (you must mind
your P's and Q's with them), two 'pom-poms,' and two galloping
Maxims; and, hurrying up behind them, Light Horse, Mounted
Infantry, Nineteenth and Twenty-first Brigades, thirty field-guns,
more 'pom-poms,' two great 5-in. ox-drawn siege pieces ('cow guns'
as the army calls them), and Ian Hamilton. It was an army
formidable to any foe; but to those who now stared upwards from
the little town and saw the dark, swift-moving masses on the hills--
an avalanche of armed men and destructive engines about to fall on
them--terrible beyond words.
 'And then,' as the poet observes, 'there was mounting in hot
haste,' saddling up of weary ponies, frantic inspanning of hungry
oxen cheated of their well-earned rest and feed, cracking of long
whips, kicking of frightened Kaffirs; and so pell-mell out of the town
and away to the northward hurried the commando of Christian De
Wet.
The Cavalry halted on the hills for a while, the General being
desirous of obtaining the formal surrender of Heilbron, and so
preventing street-fighting or bombardment. An officer--Lieutenant M.
Spender-Clay, of the 2nd Life Guards--was despatched with a flag of
truce and a trumpeter; message most urgent, answer to be given
within twenty minutes, or Heaven knows what would happen; but all
these things take time. Flags of truce (prescribe the customs of war)
must approach the enemy's picket line at a walk; a mile and a half at
a walk--twenty minutes; add twenty for the answer, ten for the
return journey, and nearly an hour is gone. So we wait impatiently
watching the two solitary figures with a white speck above them
draw nearer and nearer to the Boer lines; 'and,' says the brigadier,
'bring two guns up and have the ranges taken.'
There was just a chance that while all were thus intent on the
town, the convoy and commando might have escaped unharmed, for
it happened that the northern road runs for some distance eastward
along the bottom of the valley, concealed from view. But the clouds
of dust betrayed them.
'Hullo! what the deuce is that?' cried an officer.
'What?' said everyone else.
'Why, that! Look at the dust. There they go. It's a Boer convoy.
Gone away.'
 And with this holloa the chase began. Never have I seen
anything in war so like a fox hunt. At first the scent was uncertain,
and the pace was slow with many checks.
Before us rose a long smooth slope of grass, and along the crest
the figures of horsemen could be plainly seen. The tail of the
waggon train was just disappearing. But who should say how many
rifles lined that ridge? Besides, there were several barbed-wire
fences, which, as anyone knows, will spoil the best country.
Broadwood began giving all kinds of orders--Household Cavalry
to advance slowly in the centre; 12th Lancers to slip forward on the
right, skirting the town, and try to look behind the ridge, and with
them a battery of horse guns; 10th Hussars, to make a cast to the
left, and the rest of the guns to walk forward steadily.
Slowly at first, and silently besides; but soon the hounds gave
tongue. Pop, pop, pop--the advanced squadron--Blues--had found
something to fire at, and something that fired back, too; pip-pop,
pip-pop came the double reports of the Boer rifles. Bang--the
artillery opened on the crest-line with shrapnel, and at the first few
shells it was evident that the enemy would not abide the attack. The
horsemen vanished over the sky-line.
The leading squadron pushed cautiously forward--every
movement at a walk, so far. Infantry brigadiers and others, inclined
to impatience, ground their teeth, and thinking there would be no
sport that day, went home criticising the master. The leading
squadron reached the crest, and we could see them dismount and
begin to fire.
We were over the first big fence, and now the scent improved.
Beyond the first ridge was another, and behind this, much nearer
now, dust clouds high and thick. The General galloped forward
himself to the newly-captured position and took a comprehensive
view. 'Tell the brigade to come here at once--sharp.'
A galloper shot away to the rear. Behind arose the rattle of
trotting batteries. The excitement grew. Already the patrols were
skirting the second ridge. The Boer musketry, fitful for a few
minutes, died away. They were abandoning their second position.
'Forward, then.' And forward we went accordingly at a healthy trot.
In front of the jingling squadrons two little galloping Maxims
darted out, and almost before the ridge was ours they were
spluttering angrily at the retreating enemy, so that four burghers, as
I saw myself, departed amid a perfect hail of bullets, which
peppered the ground on all sides.
But now the whole hunt swung northward towards a line of
rather ugly-looking heights. Broadwood looked at them sourly. 'Four
guns to watch those hills, in case they bring artillery against us from
them.' Scarcely were the words spoken, when there was a flash and
a brown blurr on the side of one of the hills, and with a rasping snarl
a shell passed overhead and burst among the advancing Cavalry.
The four guns were on the target without a moment's delay.
The Boer artillerists managed to fire five shots, and then the
place grew too hot for them--indeed, after Natal, I may write, even
for them. They had to expose themselves a great deal to remove
their gun, and the limber and its six horses showed very plainly on
the hillside, so that we all hoped to smash a wheel or kill a horse,
and thus capture a real prize. But at the critical moment our 'pom-
poms' disgraced themselves. They knew the range, they saw the
target. They fired four shots; the aim was not bad. But four shots--
four miserable shots! Just pom-pom, pom-pom. That was all.
Whereas, if the Boers had had such a chance, they would have
rattled through the whole belt, and sent eighteen or twenty shells in
a regular shower. So we all saw with pain how a weapon, which is so
terrible in the hands of the enemy, may become feeble and
ineffective when used on our side by our own gunners.
After the menace of the Boer artillery was removed from our
right flank, the advance became still more rapid. Batteries and
squadrons were urged into a gallop. Broadwood himself hurried
forward. We topped a final rise.
Then at last we viewed the vermin. There, crawling up the
opposite slope, clear cut on a white roadway, was a long line of
waggons--ox waggons and mule waggons--and behind everything a
small cart drawn by two horses. All were struggling with frantic
energy to escape from their pursuers. But in vain.
The batteries spun round and unlimbered. Eager gunners ran
forward with ammunition, and some with belts for the 'pom-poms.'
There was a momentary pause while ranges were taken and sights
aligned, and then----! Shell after shell crashed among the convoys.
Some exploded on the ground, others, bursting in the air, whipped
up the dust all round mules and men. The 'pom-poms,' roused at
last from their apathy by this delicious target and some pointed
observations of the General, thudded out strings of little bombs. For
a few minutes the waggons persevered manfully. Then one by one
they came to a standstill. The drivers fled to the nearest shelter, and
the animals strayed off the road or stood quiet in stolid ignorance of
their danger.
 And now at this culminating moment I must, with all apologies
to 'Brooksby,' change the metaphor, because the end of the chase
was scarcely like a fox hunt. The guns had killed the quarry, and the
Cavalry dashed forward to secure it. It was a fine bag--to wit, fifteen
laden waggons and seventeen prisoners. Such was the affair of
Heilbron, and it was none the less joyous and exciting because, so
far as we could learn, no man on either side was killed, and only one
trooper and five horses wounded. Then we turned homewards.
On the way back to the town I found, near a fine farmhouse
with deep verandahs and a pretty garden, Boer ambulance waggons,
two German doctors, and a dozen bearded men. They inquired the
issue of the pursuit; how many prisoners had we taken? We replied
by other questions. 'How much longer will the war last?'
'It is not a war any more,' said one of the Red Cross men. 'The
poor devils haven't got a chance against your numbers.'
'Nevertheless,' interposed another, 'they will fight to the end.'
I looked towards the last speaker. He was evidently of a
different class to the rest.
'Are you,' I asked, 'connected with the ambulance?'
'No, I am the military chaplain to the Dutch forces.'
'And you think the Free State will continue to resist?'
'We will go down fighting. What else is there to do? History and
Europe will do us justice.'
'It is easy for you to say that, who do not fight; but what of the
poor farmers and peasants you have dragged into this war? They do
not tell us that they wish to fight. They think they have been made a
catspaw for the Transvaal.'
 'Ah,' he rejoined, warmly, 'they have no business to say that
now. They did not say so before the war. They wanted to fight. It
was a solemn pledge. We were bound to help the Transvaalers; what
would have happened to us after they were conquered?'
'But, surely you, and men like you, knew the strength of the
antagonist you challenged. Why did you urge these simple people to
their ruin?'
'We had had enough of English methods here. We knew our
independence was threatened. It had to come. We did not deceive
them. We told them. I told my flock often that it would not be child's
play.'
'Didn't you tell them it was hopeless?'
'It was not hopeless,' he said. 'There were many chances.'
'All gone now.'
'Not quite all. Besides, chances or no chances, we must go
down fighting.'
'You preach a strange gospel of peace!'
'And you English,' he rejoined, 'have strange ideas of liberty.'
So we parted, without more words; and I rode on my way into
the town. Heilbron had one memory for me, and it was one which
was now to be revived. In the hotel--a regular country inn--I found
various British subjects who had been assisting the Boer
ambulances--possibly with rifles. It is not my purpose to discuss here
the propriety of their conduct. They had been placed in situations
which do not come to men in quiet times, and for the rest they were
mean-spirited creatures.
While the Republican cause seemed triumphant they had
worked for the Dutch, had doubtless spoken of 'damned rooineks,'
and used other similar phrases; so soon as the Imperial arms
predominated they had changed their note; had refused to go on
commando in any capacity, proclaimed that Britons never should be
slaves, and dared the crumbling organism of Federal government to
do its worst.
We talked about the fighting in Natal which they had seen from
the other side. The Acton Homes affair cropped up. You will
remember that we of the irregular brigade plumed ourselves
immensely on this ambuscading of the Boers--the one undoubted
score we ever made against them on the Tugela.
'Yes,' purred my renegades, 'you caught the damned Dutchmen
fairly then. We were delighted, but of course we dared not show it.'
(Pause.) 'That was where De Mentz was killed.'
De Mentz! The name recalled a vivid scene--the old field-cornet
lying forward, grey and grim, in a pool of blood and a litter of empty
cartridge cases, with his wife's letter clasped firmly in his stiffening
fingers. He had 'gone down fighting;' had had no doubts what
course to steer. I knew when I saw his face that he had thought the
whole thing out. Now they told me that there had been no man in all
Heilbron more bitterly intent on the war, and that his letter in the
'Volksstem,' calling on the Afrikanders to drive the English scum from
the land, had produced a deep impression.
'Let them,' thus it ran, 'bring 50,000 men, or 80,000 men, or
even'--it was a wild possibility--'100,000, yet we will overcome
them.' But they brought more than 200,000, so all his calculations
were disproved, and he himself was killed with the responsibility on
his shoulders of leading his men into an ambush which, with
ordinary precautions, might have been avoided. Such are war's
revenges. His widow, a very poor woman, lived next door to the
hotel, nursing her son who had been shot through the lungs during
the same action. Let us hope he will recover, for he had a gallant
sire.

CHAPTER XIII
ACTION OF JOHANNESBURG

Johannesburg: June 1.

On the 24th of May, Ian Hamilton's force, marching west from

Heilbron, struck the railway and joined Lord Roberts's main column.
The long marches, unbroken by a day's rest, the short rations to
which the troops had been restricted, and the increasing exhaustion
of horses and transport animals seemed to demand a halt. But a
more imperious voice cried 'Forward!' and at daylight the travel-
stained brigades set forth, boots worn to tatters, gun horses dying at
the wheel, and convoys struggling after in vain pursuit--'Forward to
the Vaal.'
And now the Army of the Right Flank became the Army of the
Left; for Hamilton was directed to move across the railway line and
march on the drift of the river near Boschbank. Thus, for the first
time it was possible to see the greater part of the invading force at
once.
French, indeed, was already at Parys, but the Seventh and
Eleventh Divisions, the Lancer brigade, the corps troops, the heavy
artillery, and Hamilton's four brigades were all spread about the
spacious plain, and made a strange picture; long brown columns of
Infantry, black squares of batteries, sprays of Cavalry flung out far to
the front and flanks, 30,000 fighting men together, behind them
interminable streams of waggons, and, in their midst, like the pillar
of cloud that led the hosts of Israel, the war balloon, full blown, on
its travelling car.
We crossed the Vaal on the 26th prosperously and peacefully.
Broadwood, with his Cavalry, had secured the passage during the
previous night, and the Infantry arriving found the opposite slopes in
British hands. Moreover, the Engineers, under the indefatigable
Boileau, assisted by the strong arms of the Blues and Life Guards,
had cut a fine broad road up and down the steep river banks.
Once across we looked again for the halt. Twenty-four hours'
rest meant convoys with full rations and forage for the horses. But in
the morning there came a swift messenger from the Field-Marshal:
main army crossing at Vereeniging, demoralisation of the enemy
increasing, only one span of the railway bridge blown up, perhaps
Johannesburg within three days--at any rate, 'try,' never mind the
strain of nerve and muscle or the scarcity of food.
Forward again. That day Hamilton marched his men eighteen
miles--('ten miles,' say the text-books on war, 'is a good march for a
division with baggage,' and our force, carrying its own supplies, had
ten times the baggage of a European division!)--and succeeded
besides in dragging his weary transport with him. By good fortune
the Cavalry discovered a little forage--small stacks of curious fluffy
grass called manna, and certainly heaven-sent--on which the horses
subsisted and did not actually starve. All day the soldiers pressed on,
and the sun was low before the bivouac was reached. Nothing
untoward disturbed the march, and only a splutter of musketry along
the western flank guard relieved its dulness.
At first, after we had crossed the Vaal, the surface of the
country was smooth and grassy, like the Orange River Colony, but as
the column advanced northwards the ground became broken--at
once more dangerous and more picturesque. Dim blue hills rose up
on the horizon, the rolling swells of pasture grew sharper and less
even, patches of wood or scrub interrupted the level lines of the
plain, and polished rocks of conglomerate or auriferous quartz
showed through the grass, like the bones beneath the skin of the
cavalry horses. We were approaching the Rand.
On the evening of the 27th, Hamilton's advance guard came in
touch with French, who, with one Mounted Infantry and two Cavalry
brigades, was moving echeloned forward on our left in the same
relation to us as were we to the main army.
The information about the enemy was that, encouraged by the
defensive promise of the ground, he was holding a strong position
either on the Klip Riviersburg, or along the line of the gold mines
crowning the main Rand reef. On the 28th, in expectation of an
action next day, Hamilton made but a short march. French, on the
other hand, pushed on to reconnoitre, and if possible--for the
Cavalry were very ambitious--to pierce the lines that lay ahead.
I rode with General Broadwood, whose brigade covered the
advance of Hamilton's column. The troops had now entered a region
of hills which on every side threatened the march and limited the
view.
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

ebookbell.com