Landscape Ecology vol. 2 no.

3 pp 191-199 (1989)
SPB Academic Publishing bv, The Hague

Are small-scale landscape features important factors for field studies of

small mammal dispersal sinks?

Marybeth Buechner
2709 Adrian Drive, Davis, C A 95616, USA

Keywords: landscape ecology, dispersal, patch, emigration, small mammals, dispersal sink

Abstract

Interest in the influence of landscape features on animal movement has been widespread; however, few field
studies of the emigration of small mammals from patches of habitat directly consider the effects of the small-
scale landscape features. The simulation models of Stamps et al. (1987a, b) and Buechner (1987a, b) suggest
that the size of a dispersal sink relative to the size of the source patch, the average distance traveled by dispers-
ers in the sink, the ease with which dispersers cross the edge between the sink and a source patch, and source
patch perimeter:area ratio may all be important influences on emigration rates. A review of field studies of
small mammal dispersal into sinks suggests that in a substantial fraction of such studies the values of these
factors fall within the ranges that the simulation models indicate have the greatest potential effect on emigra-
tion rates. New field studies of dispersal sinks that include a consideration of these factors are necessary in
order to evaluate the magnitude of the impact of these factors on natural populations.

Introduction sideration of general landscape features such as the

number or overall connectivity of habitat patches
The role that landscape features play in the disper- (e.g., Wegner and Merriam 1979; Fahrig et al.
sal of animals has been recognized for many years 1983). However, relatively few empirical studies ad-
(e.g., Stickel 1946; Van Vleck 1968; Brown 1969) dress the possible influences of more detailed land-
and both theoretical and empirical studies have sug- scape features such as the size, shape, or edge
gested that small-scale landscape features can in- permeability of individual habitat patches. The
fluence the movement of animals of many taxa into goal of this paper is to assess the potential impor-
or out of habitat patches (e.g., Kareiva 1983, 1985; tance of certain small-scale landscape features for
Wiens 1985; Wilcove 1985; Forman and Godron studies of small mammal emigration from habitat
1986; Schonewald-Cox and Bayless 1986; Buechner patches into dispersal sinks.
1987a, b; Hansson 1987; Stamps et al. 1987a, b). Stamps et al. (1987a, b) and Buechner (1987a, b)
Ecologists studying small mammals have been espe- used a series of computer simulation models to
cially interested in the functioning of 'dispersal examine the influence of several small-scale land-
sinks', areas which absorb more dispersers than scape features on the proportion of dispersing
they produce, effectively removing animals from animals emigrating from source patches. These
the source population. A number of studies of models simulated a rectangular habitat patch con-
small mammal populations have included a con- tiguous with a second type of habitat which func-
192

Table 1. S u m m a r y of the simulation models of Stamps et al. (1987a, b) and Buechner (1987a, b). Variables shown in italics are considered
in this paper.

Parameter analyzed Range of values examined in Range where changes in this parameter had the greatest
the simulations affect on emigration rate

Source patch 6 - 6 0 0 home ranges per patch Little effect of patch size per se was found

Number of dispersers 0 - 4 dispersers per home range No effect on the proportion of dispersers emigrating from
produced per home range in the source patch the source patch

Source patch 9-100O7o of patch home ranges Low P:A ratio (P:A = .09-0.35)
perimeter:area ratio contiguous with patch edge

Permeability o f edges o f Probability of a disperser at the Low edge permeability (probability of crossing edge =
source patch patch edge moving into sink = 0.0-0.30)
0.0-1.0

Relative size o f sink and Sink size divided by patch size Small sink size (S/H = 0.10-1.0)
source patch = 0.10-17.0

Average distance dispersers Probability that a disperser will Low stopping rates (stopping rate = 0.05-0.30)
move through the sink, stop moving with each unit of
i.e., stopping rate in the area crossed = 0.05-1.00
sink

Sink location with respect Embedded in or surrounding the Little effect of sink location was found
to source patch source patch

tioned as a dispersal sink. Dispersers were generat- Buechner (1987a, b). The models were designed to
ed within the source patch and moved throughout be relevant to species from three vertebrate taxa
the simulated areas; stopping rates for dispersers (reptiles, birds and mammals) and the range of
were defined for both the source patch and the sink. values of parameters examined reflect this broad
The models distinguished between 'dispersal' and approach. The range o f parameters examined and
'emigration'. Dispersers were defined as those in- the results of the models are summarized in Table 1.
dividuals which left their natal home range and The models indicate that four factors may have
moved one or more home range diameters before major influences on the movement of animals from
stopping (dying or settling into a new home range). habitat patches into dispersal sinks: (1) the peri-
Short movements entirely within a habitat patch meter:area ratio of the source patch; (2) the size of
were still considered dispersal; thus, dispersers the dispersal sink relative to the size of the source
might stop moving either within the source patch or patch; (3) the average distance which dispersers can
within the dispersal sink. Only those dispersers travel through the dispersal sink, and (4) the ease
which left the source patch were defined as having with which individuals move across the edges of the
emigrated. Thus, animals leaving their home ranges source patch (Table 1). The simulations also sug-
but stopping within the source patch were opera- gest in which part of their range changes these
tionally defined as having dispersed but not having parameters are expected to have the greatest effect
emigrated. The emigration rate from the source on emigration (Table 1). In this paper, I compare
patch was defined as the proportion of those dis- the ranges of these parameters that occur in field
persers originating within the source patch which studies of small mammals that involve emigration
stopped moving outside of it (-- number of emi- into dispersal sinks with the ranges that the simula-
grants divided by the total number of dispersers). tions suggest are most likely to influence emigration
In all, the effects of seven factors on emigration rates.
rates were examined by Stamps et al. (1987a, b) and
 193

Methods area of good habitat (Table 2; see also brief reviews

in Dobson 1981 and Wolton and Flowerdew 1987)
Many field studies of small mammal populations or by adding attractive resources (e.g., food; Dob-
involve the capture of individuals leaving source son 1981) to an otherwise suboptimal area. Sink
populations and moving into dispersal sinks. I type was assessed straightforwardly from the
reviewed available issues of the journals Acta methods reported by the author(s) of each report.
Zoologica Fennica, American Midland Naturalist, 2. Sink and patch habitat: This category provides
Canadian Journal of Ecology, Ecology, Journal of additional information about the experimental de-
Animal Ecology, Journal of Mammalogy and sign employed in the reviewed studies. I noted the
Oecologia for field studies of rodents in which the general habitat types that occurred within the
movement of animals into dispersal sinks was source patch and within the sink. Although no
measured and from which the relevant parameters direct estimates of the ease with which individuals
could be estimated. In each case, the dispersal sink could move across patch edges were obtainable, the
was an area which more animals moved into than difference in habitat types may provide some infor-
moved out of (as measured by the authors of the mation about the permeability of the edge between
reports). In studies of naturally occurring sinks the source patch and the dispersal sink (as defined
movement was habitat-related. In most, but not all, by Stamps et al. 1987a, b).
cases of artificially produced sinks this occurred 3. Relative sink size: This was estimated from il-
because animals entering the sink were snapped lustrations or written information in each report
trapped by the researchers; in one case (Table 2 #9) and was calculated as the area of the dispersal sink
animals were attracted to the sink by the addition divided by the area of the source patch.
of food. 4. Patch perimeter:area ratio: This ratio, as the
Studies selected for analysis employed some com- term is used in the simulations, refers to the part of
bination of snap and live trapping to measure the the patch that is in contact with the sink (Buechner
demographic characteristics of the populations in- 1987a, b). Stamps et al. (1987a, b) and Buechner
volved, the number (and in some cases the direction (1987a, b) calculated patch perimeter: area ratio as
and length) of dispersal movements, and the popu- the proportion of home ranges within the patch that
lation densities and/or population growth rates in are contiguous with the dispersal sink. Although
source areas and in the dispersal sink. In some numerical estimates of this measure of perimeter:
cases, I analyzed a portion of an extensive study area ratio could not be calculated for any of the
without discussing the rest of the larger work (e.g., studies reviewed, I was able to estimate roughly
Pokki 1981). The most common questions asked by whether the proportion of home ranges within the
the reviewed studies concerned the demographic source patch that adjoined the sink was most likely
characteristics of dispersing animals such as their high (0.65-1.0), moderate (0.35-0.65), or low
sex, age, weight or breeding condition, or the rela- (0.0-0.35). This was done based on information
tionships between the magnitude of dispersal and about the size and shape of the patch, arrangement
the density or growth rate of the source population. of sink and patch habitats and the home range size
The majority of the studies reported dispersal rates, of the animals living in the patch.
however, because experimental designs varied 5. Fate of dispersers: I noted whether dispersing
widely, direct comparisons of the magnitude of dis- animals captured in the sink were removed from the
persal are not possible in this analysis. population or were released to continue their move-
In each of the 17 empirical studies reviewed, I ment. This dichotomy is related to disperser stop-
noted the following information: ping rate as defined in the simulations - the proba-
1. Sink type: This refers to whether the sink was bility that a disperser will stop moving with each
naturally occurring or artificially produced by the unit of distance crossed (Stamps et al. (1987a, b).
investigators. Many investigators produce sinks For example, if all dispersers stop moving as soon
artificially by removing resident animals from an as they enter the sink, the stopping rate in the sink
194

would be 1.0. In field studies, where dispersers may and a half times larger than the source patch (S/H
stop moving because they are captured and re- > 2.5) changes in sink size had relatively little effect
moved by the researchers conducting the study, on emigration (Table 1; Buechner 1987a, b).
stopping rate in the sink may approach 1.0 if dis- Although only four studies of naturally occur-
persers entering the sink are immediately captured ring dispersal sinks were found, even this small
and removed from the population. number of examples indicates that a wide range of
Note that some of these estimates were obtained relative sink size may be found in natural situa-
from figures and descriptions in the papers, rather tions. Three naturally occurring dispersal sinks
than from direct, quantified, reports by the were equal in size to or smaller than the area oc-
authors, and although useful for relative compari- cupied by the source population (Table 2). The
sons, should not be treated as precise values. simulations suggest that in situations like this, even
small changes in sink size could produce substantial
differences in emigration rate. In one of the four
Results studies of naturally occurring sinks an area of poor
habitat much larger than the source patch acted as
Patch perimeter:area ratio a sink (Table 2); in such situations differences in
sink size between studies or between patches were
In the simulations the perimeter:area ratio of a unlikely to confound measures of emigration rate.
patch was strongly positively related to the magni- Sinks created artificially by researchers vary
tude of emigration, and changes in patch perimeter: widely in size with respect to the source patehes
area ratio most strongly influenced emigration (Table 2), however, most artificial sinks in the
when patch perimeter:area ratio was low. (i.e., less reviewed studies were smaller than the source
than 0.35; Table 1). Among the 13 reviewed studies patches. S / H for the reviewed studies of artificial
in which source patches were contiguous with artifi- sinks ranged from .002-1.63. Eleven of the studies
cially created sinks, three were estimated to have of artificial sinks had estimated S / H values less
low perimeter:area ratios and ten to have moderate than or equal to 1.0 (Table 2), the range over which
perimeter:area ratios. The four studies of naturally changes in sink size are expected to have the sub-
occurring sinks encompass a greater range of peri- stantial influence on the magnitude of emigration
meter:area ratio values, with two patches having (Table 1).
low perimeter:area ratios and two having high
perimeter:area ratios (Table 2). Thus, five of 17
reviewed studies appear to have perimeter:area ra- Distance traveled by dispersers through the sink
tios falling in the range where the models suggest
that differences in patch perimeter:area ratios are Stamps et al. (1987a) defined the 'stopping rate' of
most likely to influence emigration rates. dispersers in the sink as the probability per move
(each unit of sink habitat crossed) that a disperser
will stop (see also Buechner 1987c). The inverse of
Sink size this stopping rate thus represents the average dis-
tance moved by dispersers through the sink before
In the models, the relationship between relative stopping. In the models, as stopping rates in-
sink size and emigration from the source patch fol- creased, emigration increased (this occurred be-
lowed a decelerating, asymptotic curve (Buechner cause when stopping rates were high in the sink very
1987a) such that when sinks were smaller than the few dispersers returned to the patch after having en-
source patch ( S / H s 1.0) relatively small changes tered the sink; see Stamps et al. 1987a). Emigration
in sink size produced substantial changes in emigra- rate declined substantially with changes in sink
tion rate. The effect of changes in sink size gradual- stopping distance only when stopping rates were
ly decreased, so that when sinks were more than two low ( < 0.30) (Buechner 1987a, b).
 195

Table 2. Characteristics of the small mammal dispersal sinks in the 17 studies reviewed here.

Species Sink Sink Patch habitat Sink S/H 2 Patch P:A 3 Rel./rem. 4
type 1 habitat

Microtus agrestis, Pokki 1981 nat. woods old field 1.0 high rel.
Microtus pennsylvanicus,
Tamarin 1977 nat. woods old field 1.0 + low rem.
Perornyscus leucopus
Krohne and Baccus 1985 nat. old field forest 0.2 low rel.
Microtus ochrogaster and
M. pennsylvanicus, nat. agricul, bluegrass 20 + high tel.
Verner and Getz 1985 field
Peromyscus art. mature forest mature forest 0.1 low rem.
leucopus Stickel q 946
Microtus pennsylvanicus, art. old field old field 0.01 low rem.
Van Vleck 1968 - 0.07
Sigmodon hispidus and art. coastal coastal prairie 1.0 mod rem.
Reithrodontomys fulvescens prairie
Joule and Cameron 1975
Peromyscus maniculatus art. coastal coastal 0.002 mod rein.
Fairbairn 1978
Spermophilus beecheyi,
Dobson 1979 art. oak savanna oak savanna 0.05 mod rel.
(food added)
Microtus ochrogaster,
Gaines et al. 1979 (a) art. old field old field 1.0 mod rem.
(b) art. burned field old field ? rood rem.
Microtus townsendii,
Beacham 1980 art. grassy field grassy field 0.9 rood rein.
Spermophilus beechyi,
Dobson 1981 art. savanna savanna 1.0 rood rem.
Microtus pennsylvanicus,
Baird and Birney 1982 art. meadow meadow 0.33 mod rem.
- 0.49
Peromyscus leueopus,
Krohne et al. 1984 art. mature forest mature forest 1.63 mod tel.
Microtus pennsylvanicus,
Tamarin et al. 1984 art. woods field 0.20 low rein.
Peromyscus leucopus,
Krohne and Miner 1985 art. mature forest mature forest 1.63 mod rem.
Microtus californicus,
Heske 1987 art. coastal prairie coastal prairie 0.08 mod rel.

i Naturally occurring or artificially produced.

2 Sink size divided by source patch size.
3 Source patch perimeter:area ratio, i.e., the proportion of home ranges in the patch which are contiguous with the edge between the
patch and the sink.
4 Whether dispersers captured in the sink were released to continue their movement or removed from the population.

In the reviewed studies one of the four studies of stopping rate influenced emigration rates directly.
n a t u r a l l y o c c u r r i n g s i n k s a n d 10 o f 13 s t u d i e s o f ar-
tificial sinks employed experimental d e s i g n s in
w h i c h all e m i g r a n t s a r e c a p t u r e d a n d r e m o v e d f r o m Barriers to e m i g r a t i o n
the sink u p o n c a p t u r e , a t e c h n i q u e likely to p r o d u c e
r e l a t i v e l y h i g h s t o p p i n g r a t e s . T h u s , 11 o f t h e 17 The simulations measured the edge permeability of
reviewed studies employ designs which the simula- a p a t c h as t h e p r o b a b i l i t y t h a t a d i s p e r s e r a p p r o a c h -
t i o n s s u g g e s t m a k e it u n l i k e l y t h a t d i f f e r e n c e s i n i n g t h e e d g e o f t h e p a t c h w o u l d c r o s s it a n d m o v e
196

into the sink (Stamps et al. (1987b). In the simula- above parameters for field studies of dispersal, it is
tions the edge permeability of a patch was strongly necessary to relax the 'all-else-being-equal' assump-
positively related to the magnitude of emigration tion implicit in analyzing one variable at a time and
(Table 2; details in Stamps 1987b). In field studies to consider the ways in which the parameters inter-
the similarity of habitat types in the sink and source act to influence emigration. The simulation models
areas may affect the permeability of the edge to showed that the value of stopping rate occurring in
movement; presumably fewer animals will readily a given situation can affect the magnitude of the in-
cross into unfamiliar or suboptimal habitat than fluence of both patch perimeter:area ratio (Stamps
would enter familiar or optimal habitat. In three of et al. 1987a and b) and sink size (Buechner 1987a
the four reviewed studies of naturally occurring and b) on emigration. The influence of patch
sinks and two of the 14 studies of artificial sinks the perimeter:area ratio on emigration is expected to be
sink is an area of poor habitat which dispersers may strongest when dispersers stop as soon as they enter
be reluctant to enter (Table 2). the sink (Stamps 1987b; Buechner 1987a, b). In 11
The permeability of the edge between two areas of the 17 reviewed studies dispersers captured in the
may also be affected by the presence of resident sink were removed from the population (Table 2):
animals. When the sink is an area of good habitat four of these studies also had low patch perimeter:
from which all residents have been removed dis- area ratios. Such designs are especially likely to
persers are expected to enter the cleared area at a have measurements of emigration rate confounded
rate greater than if residents were present in the by small differences in patch perimeter:area ratio.
sink. This has been shown to be the case for some Stopping rates also influence the relationship be-
species and has been labelled the 'vacuum effect' tween sink size and emigration rate. Changes~ in
(e.g., Tamarin et al. 1984 or Krohne and Baccus sink size are not expected to have any effect on
1985). Emigration rates in studies in which the sink emigration rate if all dispersers stop as soon as they
and source patch contain similar habitat and in leave the sink (stopping rate = 1.0; Buechner
which all animals captured in the sink are removed, 1987b). In those studies in which dispersers are cap-
as was done in 9 of the 14 studies o f artificial sinks tured and removed from the population as soon as
reviewed here, may be influenced by this effect. they enter the sink, stopping rates in the sink may
In the simulations the effect of small differences approach 1.0. The relative size of the sink is, thus,
in edge permeability on emigration are expected to most likely to be an important factor in those
be greatest when patch edges are relatively impas- studies in which dispersers are captured after
sable (Stamps et al. 1987b). This implies that varia- having moved some distance through the sink or are
tions in the permeability of edges may be most im- released after capture. Six of the studies reviewed
portant for studies in which the sink and source (including three natural and three artificial sinks)
habitats are very different or where artificial bar- had designs in which dispersers were released after
riers substantially impede movement. Some studies capture (Table 2); of these, four had S / H _< 1.0, a
use semi-permeable barriers such as drift fences situation in which small differences in sink size are
(e.g., Beacham 1980; Johnson and Gaines 1987) or likely to have substantial effects on emigration rate.
mowed strips (e.g., Verner and Getz 1985) sur-
rounding source habitat patches. Unfortunately,
no information is available about the extent to Discussion
which movement across the edge is reduced by these
barriers. The results indicate that the parameters identified
by the simulation models of Stamps et al. (1987a,
b) and Buechner (1987a, b) are of potential impor-
Interaction effects tance for a substantial fraction of field studies of
the movement of small mammals into dispersal
In order to assess the potential importance of the sinks. All but one (# 16, Table 2) of the reviewed
 197

studies had one or more parameters in the sensitive necessary in order to evaluate the magnitude of the
range indicated by the simulation models. Even impact of these factors on natural populations. The
when interaction effects were considered, over half current review suggests some types of field experi-
(8 of 17) of these studies had combined parameter ments that could be used to test for confounding
values in the sensitive range. factors or to validate the implications of the
The designs of many of the studies reviewed here models. Field studies in which all dispersers are cap-
are such that uncontrolled or unmeasured small- tured and removed as soon as they enter the sink
scale landscape features may have influenced emi- could be especially valuable for studying the effects
gration rates, however, a number of problems com- of patch edge permeability and perimeter:area ra-
plicate interpretation of these results. The studies tio. In this case, the effects of patch perimeter:area
reviewed here do not represent a direct test of the ratios could be studied without the additional com-
models. In some cases, parameter values had to be plicating effects of variable sink sizes. Patches of
estimated by indirect means as they were not direct- varying sizes and shapes, embedded within the
ly measured by the researchers. In the reviewed same sink habitat, could be used to estimate the ef-
studies variables other than those analyzed here fects of patch perimeter:area ratio on the propor-
were of primary interest and the role of these varia- tion of potentially dispersing animals within the
bles, which is unknown, complicates any conclu- patch population that actually emigrate. The use of
sions. None of the reviewed studies is concerned drift fences or other man-made barriers could serve
solely with the magnitude of emigration; for the to control for patch edge permeability. While
most part they address questions related to popula- difficult methodologically, it may also be possible
tion regulation or stability. Small-scale landscape to obtain field estimates of patch edge permeability
features are likely to be confounding the results of from removal studies (see Stamps et al. 1987b), by
these studies only to the extent that the magnitude estimating what percentage of the dispersers ap-
of emigration is relevant to the questions they ad- proaching the edge actually cross into the sink.
dress. Studies in which captured dispersers are released
Although the degree to which the results of the may be yield information about the effects of sink
reviewed studies were confounded by small-scale size. Emigration rates could be compared for
landscape features is not clear, it is clear that studies patches of the same size and shape, containing the
of emigration rates, p e r se, will have to be designed same type of habitat and with the same type of
to carefully consider such factors. If future studies edges (e.g., a given type of drift fence) but adjacent
which focus directly on the magnitude of emigra- to sinks of different sizes. In addition, designs
tion from small mammal populations utilize de- which release dispersers could provide estimates of
signs similar to those employed in the reviewed naturally occurring disperser stopping rates (due to
studies, variation in uncontrolled landscape fea- settlement and/or mortality) by monitoring how
tures will make the comparison of results between far dispersers move before stopping (Buechner
studies difficult. In particular, comparisons of the 1987c).
magnitude of emigration into two or more sinks, or Although the models of Stamps et al. (1987a, b)
from two or more patches, could easily have their and Buechner (1987a, b) strongly indicate that
results confounded by uncontrolled landscape small-scale landscape features may be influencing
features. emigration rates in some field studies of the move-
Since methods are not always comparable, and ment of small mammals into dispersal sinks, they
emigration rates are not usually reported, the do not specify how such factors may influence
reviewed studies do not allow a detailed assessment population regulation processes. The simulations
of exactly how landscape features are influencing do not address the role of dispersal in population
emigration rates in the populations being studied. regulation, nor do they incorporate any density-
New field studies of dispersal sinks which include a dependence of dispersal rates (Stamps 1987a, b).
consideration of small-scale landscape factors are This is because the simulations represent simplified
198

'snapshot' views of emigration into dispersal sinks, Dobson, F.S. 1981. An experimental examination of an artificial
rather than modeling continuous dispersal from a dispersal sink. J. Mammal. 62: 74-81.
Fahrig, L., Lefkovitch, L. and Merriam, G. 1983. Population
growing or declining population. The simplifying
stability in a patchy environment. In Analyses of Ecological
assumptions of the simulations do not belie the Modelling. pp. 61-67. Edited by W.K. Lauenroth, G.V.
potential importance of the implications of the Skogeboe and M. Flug. Elsevier, New York.
models; instead they indicate the need for field Fairbairn, D.J. 1978. Dispersal of deer mice, Peromyscus
studies designed to systematically examine the im- maniculatus: Proximal causes and effect on fitness. Oecologia
32: 171-193.
portance of landscape features in real populations.
Forman, R.T.T. and Godron, M. 1986. Landscape ecology.
The details of the relationships between emigration John Wiley and Sons, New York.
rates and population dynamics or stability remain Gaines, M.S., Vivas, A.M. and Baker, C.L. 1979. An ex-
unknown (Lidicker 1985). Until an understanding perimental analysis of dispersal in fluctuating vole popula-
of such relationships is elucidated, sink size and lo- tions: Demographic parameters. Ecology 60: 814-828.
Game, M. 1980. Best shape for nature reserves. Nature 287:
cation, patch perimeter:area ratio and edge perme-
630-632.
ability, and the stopping rate of dispersers remain Hansson, L. 1987. Dispersal routes of small mammals at an
variables of potentially great importance for field abandoned field in central Sweden. Holarctic Ecol. 10:
studies of small mammal populations. 154-159.
Heske, E.J. 1987. Spatial structuring and dispersal in a high den-
sity population of the California vole Microtus californicus.
Holarctic Ecol. 10: 137-148.
Acknowledgments Johnson, M.L.and Gaines, M.S. 1987. The selective basis for
dispersal of the prairie vole, Microtus ochrogaster. Ecology
Drs. Judy Stamps, Thomas Schoener, William 68: 684-694.
Lidicker, James H. Brown, Christine Schonewald- Joule, J. and Cameron, G.N. 1975. Species removal studies. I.
Dispersal strategies of sympatric Sigmodon hispidus and
Cox, and several anonymous reviewers provided
Reithrodontomys fulvescens populations. J. Mamm. 56:
helpful comments on earlier drafts of this manus- 378-396.
cript. The early stages of this work were partially Kareiva, P. 1983. Influence of vegetation texture on herbivore
supported by an NSF predoctoral fellowship to the populations: Resource concentration and herbivore move-
author while at the University of California, Davis. ment. In Variable Plants and Herbivores in Natural and
Managed Systems. pp. 259-289. Edited by R.F. Denno and
M. McClure. Academic Press, New York.
Kareiva, P. 1985. Finding and losing host plants by Phyllotreta:
References Patch size and surrounding habitat. Ecology 66: 1809-1816.
Krobne, D.T. and Baccus, R. 1985. Genetic and-ecological
Baird, D.D. and Birney, E.C. 1982. Characteristics of dispersing structure of a population of Peromyscus leucopus. J. Mam-
meadow voles Microtus pennsylvanicus. Amer. Midl. Nat. mal. 66: 529-537.
107: 262-283. Krohne, D.T., Dubbs, B.A. and Baccus, R. 1984. Analysis of
Beacham, T.D. 1980. Dispersal during population fluctuations dispersal in an unmanipulated population of Peromyscus leu-
of the vole, Mierotus townsend#. J. Anim. Ecol. 49: copus. Amer. Midl. Nat. 112: 146-156.
867-877. Krohne, D.T. and Miner, M.S. 1985. Removal trapping studies
Brown, J.L. 1969. Territorial behavior and population regula- of dispersal in Peromyscus leucopus. Can. J. Zool. 68:
tion in birds. Wil. Bull. 81: 293-329. 71-75.
Buechner, M. 1987a. Conservation in insular parks: Simulation Lidicker, W.Z. 1985. Dispersal. Spec. Publ. Amer. Soc. Mamm.
models of factors affecting the movement of animals across 8: 420-454.
park boundaries. Biol. Cons. 41: 57-76. Pokki, J. 1981. Distribution, demography and dispersal of the
Buechner, M. 1987b. Implications of the behavior of individuals field vole, Microtus agrestis (L.), in the Tvarminne archipela-
for the ecology of populations: Analyses of the effects of go, Finland. Acta Zool. Fenn. 164: 1-48.
spacing, dispersal and dominance behaviors. Unpublished Schonewald-Cox, C.M. and Bayless, J.W. 1986. The boundary
Ph.D. Dissertation, Animal Behavior Graduate Group, model: A geographic analysis of design and conservation of
University of California, Davis. nature reserves. Biol. Conser. 38: 305-322.
Buechner, M. 1987c. A geometric model of vertebrate dispersal: Stamps, J.A., Buechner, M. and Krishnan, V.V. 1987a. The ef-
Tests and implications. Ecology 68(3): 10-318. fects of habitat geometry on territorial defense costs: Intruder
Dobson, F.S. 1979. An experimental study of dispersal in the pressure in bounded populations. Amer. Zool. 129: 533-552.
California ground squirrel. Ecology 60:1103-1109. Stamps, J.A., Buechner, M. and Krishnan, V.V. 1987b. The el-
 199

fects of edge permeability and habitat geometry on emigra- Wegner, J.F. and Merriam H.G. 1979. Movements by birds and
tion from patches of habitat. Amer. Nat. 129:533 552. small mammals between a wood and adjoining farmland
Stickel, L.F. 1946. The source of animals moving into a depopu- habitats. J. Appl. Ecol. 16: 349-358.
lated area. J. Mamm. 27: 301-307. Wiens, J.A. 1985. Vertebrate responses to environmental patch-
Tamarin, R.H. 1977. Dispersal in island and mainland voles. iness in arid and semiarid ecosystems. In The Ecology of
Ecology 58:964 978. Natural Disturbance and Patch Dynamics. pp. 169-193.
Tamarin, R.H., Reich, L.M. and Moyer, C.A. 1984. Meadow Edited by S.T.A. Pickett and P.S. White. Academic Press,
vole cycles within fences. Can. J. Zool. 62: 1796-1804. New York.
Van Vleck, D.B. 1968. Movements of Microtus pennsylvanicus Wilcove, D.S. 1985. Nest predation in forest tracts and the
in relation to depopulated areas. J. Mamm. 49: 92-103. decline of migratory songbirds. Ecology 16:1211-1214.
Verner, L. and Getz, L.L. 1985. Significance of dispersal in fluc- Wolton, R.J. and Flowerdew, J.R. 1985. Spatial distribution
tuating populations of Microtus ochrogaster and M. pennsyl- and movements of wood mice, yellow-necked mice and bank
vanicus. J. Mamm. 66: 338-347. voles. Symp. Zool. Soc. Lond. 55: 249-275.