Download modeling the impact of edge avoidance on avian nest

Ecological Applications, 12(6), 2002, pp. 1567–1575
q 2002 by the Ecological Society of America
MODELING THE IMPACT OF EDGE AVOIDANCE ON AVIAN NEST
DENSITIES IN HABITAT FRAGMENTS
ERIC K. BOLLINGER1
AND
PAUL V. SWITZER
Department of Biological Sciences, Eastern Illinois University, Charleston, Illinois 61920 USA
Abstract. In fragmented landscapes, many species of birds are absent from, or have
reduced densities in, small habitat fragments. This pattern may result, at least in part,
because birds avoid placing their nests near habitat edges where nest success often is low.
We sought to clarify the role played by edge avoidance in producing these patch size effects.
Using a numerical approach, we modeled nest densities in patches of different sizes and
shapes both for species displaying edge avoidance (i.e., ‘‘edge-sensitive’’ species) and for
those not displaying this characteristic (i.e., ‘‘edge-insensitive’’ species). Edge avoidance
in our model was defined as a reduced probability of nest placement occurring near a habitat
edge. Our model produced the expected result that edge avoidance reduced nest densities
in patches of all sizes compared to densities of edge-insensitive species. Surprisingly,
however, edge avoidance did not reduce nest densities in small patches relative to large
patches, and nest densities actually increased exponentially as patch size decreased for
edge-insensitive species. Also unexpected was the result that nests of edge-sensitive species
were found in the edge habitat at frequencies only slightly below those expected based on
edge area, whereas edge-insensitive species actually had higher than expected nest densities
in edge habitat. However, in our model, edge-sensitive species displayed a greater reduction
in nest densities near edges when their overall patch density was reduced by half, suggesting
that edge avoidance is density dependent. Finally, both types of species showed marked
increases in nest densities in linear habitat patches compared to square patches. These
patterns were a direct result of our settlement rule that required a female’s nest location to
be a minimum distance from other nests. This study suggests that knowledge of the settlement rules used by female birds may be a key to accurately demonstrating the existence
and assessing the potential consequences of edge avoidance. Detailed observations of
marked females immediately following arrival at habitat patches, as well as a comparison
of nest densities, territory sizes, polygyny levels, and use of habitat off territory, would
greatly help our understanding of this interesting and important phenomenon.
Key words: avian nest placement; edge avoidance; habitat fragmentation; habitat interior; habitat
patch; model; nest density; patch size.
INTRODUCTION
Habitat fragmentation breaks up contiguous habitat
into fragments of various sizes and shapes, resulting
in reduced population densities and lowered reproductive success for a variety of avian species (Whitcomb
et al. 1981, Robbins et al. 1989, Robinson 1992, Wiens
1994). Research on birds has suggested that for many
species, the quality of habitat in small or more linear
fragments is lower than that in larger or more circular
fragments (e.g., Gibbs and Faaborg 1990, Brawn and
Robinson 1996, Helzer and Jelinski 1999, Zanette et
al. 2000). Densities of these species (sometimes referred to a ‘‘habitat interior’’ or ‘‘area sensitive’’ species; Whitcomb et al. 1981) frequently decline with
patch size and they are frequently absent from small
patches even when these habitats contain ample area
for several territories (Robbins 1979, Blake and Karr
Manuscript received 20 April 2001; revised 29 January 2002;
accepted 20 February 2002; final version received 29 March
2002.
1 E-mail: [email protected]
1984, Robbins et al. 1989, Herkert 1994). In addition,
reproductive success is often lower in these small
patches (Robinson 1992, Paton 1994). Patch size effects on breeding densities are known for forest- (e.g.,
Whitcomb et al. 1981, Ambuel and Temple 1983, Robbins et al. 1989) and grassland-nesting species (Herkert
1994, Vickery et al. 1994, Bollinger 1995, Helzer and
Jelinski 1999, Johnson and Igl 2001).
For birds, the negative effects of patch size are often
thought to result, at least in part, from the negative
effects of habitat edges (Gates and Gysel 1978, Paton
1994, Winter et al. 2000). As patch size decreases, the
maximum distance from within the patch to the nearest
edge necessarily decreases and the perimeter (edge) to
area ratio increases. Thus, small patches are more
‘‘edge-dominated’’ and phenomena near edges are
more pronounced (Temple 1986, Wiens 1989, Helzer
and Jelinski 1999). Negative impacts of edges on birds
include greater levels of nest predation near edges by
predators using edges as travel lanes and higher degrees
of nest parasitism by cowbirds (Molothrus spp.) using
edges as perch sites and vantage points for displays
1567
1568
Ecological Applications
Vol. 12, No. 6
ERIC K. BOLLINGER AND PAUL V. SWITZER
and nest searching (Gates and Gysel 1978, Brittingham
and Temple 1983, Wilcove 1985, Johnson and Temple
1986, 1990, Robinson 1992, Paton 1994, Winter et al.
2000). Edge avoidance may also occur when birds
avoid placing their territories or nests near habitat edges, leading to reduced nest densities near edges and in
small patches (e.g., Wiens 1969, Johnson and Temple
1990, Helzer 1996, Bock et al. 1999, Sutter et al. 2000,
Winter et al. 2000).
Edge avoidance may occur in birds for many reasons.
First, if higher predation and parasitism levels occur
at edges (as just discussed), individuals nesting near
edges and having these nests fail (or detecting lower
survival of others nearby; Boulinier and Danchin 1997)
may learn to nest elsewhere (e.g., Switzer 1997), as
suggested by reduced site fidelity following nest failure
(Bollinger and Gavin 1989, Bensch and Hasselquist
1991, Roth and Johnson 1993, review in Switzer 1993).
Alternatively, if the negative effects of edges are consistent, selection should favor individuals that nest
away from edges, resulting in an evolved aversion to
edge habitat. Second, vegetation characteristics (Ranney et al. 1981, Mesquita et al. 1999) and food density
(Burke and Nol 1998, Zanette et al. 2000) may differ
between edges and the interior, possibly due to microclimatic differences that occur near some habitat edges
(Kapos 1989). Finally, certain grassland species may
require the more ‘‘open feel’’ of the interior of large
grassland patches to nest in (see Herkert et al. 1999).
The magnitude and expression of edge avoidance
may also depend on conspecific density. For example,
conspecific density and density-dependent changes in
fitness are key components of many classic models of
habitat selection, e.g., ‘‘ideal free’’ and ‘‘ideal despotic’’ distributions (Fretwell and Lucas 1969, Fretwell
1972). In these models, early nesters select the highest
quality habitat patches. As density increases, fitness
declines and later nesters are better off selecting poorer
quality patches, where conspecific density is lower.
Similarly, for habitat-interior species, early nesters
probably choose to nest in central locations in habitat
patches. If patch density remains low, edge avoidance
will be pronounced. Conspecific attraction and aggregation (e.g., Stamps 1988) could exaggerate this pattern. At high conspecific densities, however, females
may be forced to nest near edges, thereby reducing the
apparent magnitude of edge avoidance. Alternatively,
edges might initially function as ecological traps (Gates
and Gysel 1978) attracting nesting females, even at low
densities. Females may then learn to avoid nesting near
edges as a result of past nesting failures.
Clearly, regardless of the cause, behavioral edge
avoidance could affect densities in patches and might
even lead to the ‘‘minimum patch size’’ requirements
frequently reported for some avian species (Robbins
1979, Wenny et al. 1993). We were interested in identifying the extent to which simple edge avoidance could
be responsible for patch size effects (e.g., reduced den-
sities, absence from small patches, etc.). Relatively few
models have explored the mechanisms creating patch
size patterns and the potential consequences of these
mechanisms (but see Temple 1986, Matter 2000, NeyNifle and Mangel 2000, Donovan and Lamberson
2001). In particular, rarely have studies investigated
the potential consequences of edge avoidance on the
density and pattern of nesting in habitat patches. We
described the effects of edge avoidance on nest densities by modeling nest densities in patches of different
sizes and shapes. We compared nest densities both with
and without edge avoidance. Edge avoidance in our
model was defined as a reduced probability of nest
placement occurring near the habitat edge. Additionally, in the absence of edge avoidance, we wanted to
describe the expected patterns of nest densities (and
locations relative to the edge). Specific questions that
we addressed with our model included the following.
(1) How does edge avoidance affect nest densities in
patches of different sizes and shapes? (2) To what extent does edge avoidance reduce nest densities near
habitat edges? (3) How does overall nest density in a
patch affect the magnitude of edge avoidance?
METHODS
Model description
Our simulation model depicted female birds settling
into a patch of habitat of a given size and shape. We
assumed that the habitat surrounding the patch was
unsuitable for nesting, but not necessarily for foraging.
Observations suggest that in at least some species (e.g.,
Bobolink, Dolichonyx oryzivorus, Upland Sandpiper,
Bartramia longicauda), females may use areas unsuitable for nesting in order to forage, or may forage off
territory (Gavin 1984, Helzer and Jelinski 1999; E. K.
Bollinger, personal observation). Other than edge
avoidance/edge effects (which we will describe), habitat quality was assumed to be homogeneous within the
patch and females were assumed to settle randomly
(except that females were not allowed to nest within a
set distance of other females). We assumed that female
numbers were sufficient to ‘‘fill’’ the patch, given the
restrictions of other females’ nest locations and (possibly) edge avoidance.
We chose to model females for four reasons. First,
much of the empirical evidence concerning the negative
effects of habitat fragmentation and edges is based on
nest success in relation to patch size and edge proximity
(e.g., Johnson and Temple 1990, Robinson 1992, Robinson et al. 1995, Winter et al. 2000). Second, it is the
presence of females, nests, and their subsequent fate,
not simply the presence of males (and their territories),
that determines a population’s viability. Third, males
may incorporate areas into their territories, such as
perch sites aiding in territorial surveillance and advertisement, that are not suitable as nest sites (Petit et
al. 1988). Finally, even if male density is constrained
December 2002
EDGE AVOIDANCE MODEL
by a minimum territory size, females may occur at
higher densities by mating polygynously.
To model nesting patterns, we incorporated settlement rules that females might use when choosing nest
sites. Although few data exist for such rules (but see
Muldal et al. 1985), nest sites for many species are
frequently overdispersed within patches (Brown 1969,
Krebs 1971, Newton et al. 1977, Blancher and Robertson 1985, Galbraith 1988, Yasukawa et al. 1992,
Schieck and Hannon 1993). Therefore, we allowed females to settle randomly within the patch, subject to
the constraint that they could not settle within a set
distance of another nest. We assumed that females used
this proximate spacing rule rather than overall nest density or amount of exclusive area in determining where
they should nest and if they should nest within our
homogeneous patches. Lokemoen et al. (1984), for example, found fairly regular spacing among Mallard
(Anas platyrhynchos) nests at very high densities, but
more random spacing at lower densities. The only modification to our settlement rule was to model edge
avoidance.
We did not model nest success. Thus, our model
differed in this respect from many classic models of
habitat selection (e.g., the ‘‘ideal free distribution’’
model of Fretwell and Lucas 1969), which assume that
fitness declines monotonically with density and that
birds base habitat selection decisions, in part, on conspecific density (Fretwell 1972, Rosenzweig 1981,
Sutherland 1996). We did, however, look at the impact
of conspecific density on the magnitude of edge avoidance.
Simulation overview.—Each trial began with a blank
patch containing a set number of equal-sized cells. We
assumed that females settled asynchronously, such that
only one female was settling at a time. For each female,
a possible cell for settlement (i.e., nesting) was drawn
at random. If no other females were already nesting
within the set spacing rule (and if settlement were possible, given the distance to the edge for edge-sensitive
species), then the female would nest in that cell. If these
criteria were not met, then another possible cell for
settlement was chosen at random and the process was
repeated until an available cell was found for that female. The simulation continued for a patch until no
cells remained available for settlement (most trials) or
until the patch was occupied to half of the maximum
saturation (for the density trials). Other details of the
model and modeling process will be given.
The patch.—The model consisted of a rectangular
patch with dimensions xmax, ymax. The patch was subdivided completely into square cells (in which nests
could be located) of equal size so that an (x, y) coordinate could be used to describe the location of an
individual cell or nest. Habitat quality throughout the
entire patch was assumed to be uniform (e.g., see Wootton et al. 1986).
Settlement rules.—Possible cells for settlement were
1569
FIG. 1. Shape of ‘‘gamma function’’ (see Eq. 1) for different values of g. Distance from edge refers to the number
of cells from the outside edge of the patch. Each cell corresponds to 5 m.
randomly drawn from the patch. If no nest was within
the cell or within a radius of dnest from that cell, the
location was available for settlement. If there was no
edge avoidance, the probability of settlement within
that cell, Ps, was equal to 1. If there was an edge effect,
then we defined the distance to the nearest edge as dedge.
The edge had an effect within a certain range emax such
that
1
Ps 5 
1
dedge
e
 max 1 1

if dedge . emax
(1)
g
2
if dedge # emax.
Thus, for g . 0, Ps increased from the edge to emax.
The habitat in a patch # emax is sometimes referred to
in this paper as ‘‘edge habitat.’’ By varying g, we could
vary the form of the edge function to model different
degrees of edge avoidance (Fig. 1). Adding 1 to emax
in the denominator constrained Ps to be ,1 for all dnest
# emax.
We used a random number generator to see if a female would settle in the selected cell, given that cell’s
Ps. If she did not settle, then in order to have Ps remain
valid, that cell was blocked from having any future
females settling in it. In reality, an empirical estimation
of Ps would likely be a result of the summation of the
settlement probabilities of all females for that location
(i.e., the likelihood of finding a nest in that location).
However, giving one female a chance, Ps to settle in
ERIC K. BOLLINGER AND PAUL V. SWITZER
1570
Ecological Applications
Vol. 12, No. 6
tially for ‘‘edge-insensitive’’ species. Also as expected,
edge-sensitive species typically had a lower proportion
of nests in the edge relative to the amount of edge area,
although the magnitude of this effect was small (Fig.
3a). However, edge-insensitive species actually had a
larger proportion of nests near an edge than expected
based on edge area (Fig. 3a). This result is a consequence of our settlement rules; females settling near
an edge needed a smaller amount of potential nesting
area to be vacant because we assumed that no females
settled outside the patch borders in the surrounding
habitat (although it is possible that these edge-nesting
females could use areas outside the patch for foraging).
FIG. 2. Nest density as a function of patch size for edgeinsensitive (i.e., no edge effect) and edge-sensitive (g 5 3)
trials. All patches were square.
that location is functionally equivalent to the summed
probability.
A settlement trial continued until either all of the
available cells were filled or until a female could not
find an available cell within a specified number of attempts (i.e., random draws of cells). We set this maximum number of settlement attempts equal to the number of cells in the patch. This cutoff was mainly incorporated to speed up the trials, however, the maximum was rarely reached and means of 50 trials were
calculated to better estimate the number and placement
of nests within the patch for a given set of parameter
values.
Baseline parameters and variations.—We assumed
that each cell was 5 3 5 m. Our baseline runs were
completed with a dnest 5 75 m, emax 5 50 m, g 5 3 (so
that Ps remained low until close to emax), and a patch
area of 25 ha. Both patch size (1–100 ha) and shape
(square to narrow rectangle) were varied, both including edge avoidance (for ‘‘edge-sensitive species’’) and
omitting edge avoidance (for ‘‘edge-insensitive species’’). We also varied nest density by repeating the
patch size trials at half of the maximal nest density
determined from the first set of trials.
RESULTS
Effects of area
To examine the effects of patch area independently
of patch shape, we varied the size of a square patch
from 1 to 100 ha. At all sizes, edge-sensitive species
had lower nest densities than edge-insensitive species.
This difference, as expected, was most dramatic at the
smallest patch sizes (Fig. 2), where a larger proportion
of the patch is near an edge. Surprisingly, however,
nest densities did not decline in smaller patches for
edge-sensitive species and actually increased exponen-
FIG. 3. Proportion of nests in the edge as a function of
the proportion of the patch that was edge habitat for edgeinsensitive (i.e., no edge effect) and edge-sensitive (g 5 3)
trials. (a) Patches maximally saturated with nests; (b) patches
half saturated with nests. All patches were square, and the
proportion of edge habitat was varied by changing the size
of the patch (as in Fig. 2).
EDGE AVOIDANCE MODEL
December 2002
1571
Sensitivity analyses
We varied dnest, emax, and g to investigate their impact
on the patterns just described. As expected, as dnest and
emax increased, nest density declined and, for emax, a
lower proportion of nests occurred in the edge. As g
increased (making the edge avoidance ‘‘stronger’’), the
number and proportion of nests in the edge both decreased, but the magnitude of the effect was not large
(Fig. 5).
DISCUSSION
FIG. 4. Number of nests in a 25-ha patch as a function
of patch shape for edge-insensitive (i.e., no edge effect) and
edge-sensitive (g 5 3) trials. The log of length/width was
used to clarify the results; as the ratio increases, the patch is
becoming more linear.
Effects of density
To investigate the effects of the overall density of
nests in a patch on patch settlement patterns, we used
patches of identical shape (square) but different areas,
and stopped the trial when the patch was at half of its
maximum saturation (determined from ‘‘Effects of
Area’’ trials, see Fig. 2). As with saturated patches (Fig.
3a), at lower densities, edge-sensitive species had a
lower proportion of nests near the edge relative to the
available area and edge-insensitive species had a higher
proportion of nest near the edge (Fig. 3b). However,
edge-sensitive species exhibited a markedly higher
edge avoidance at the ‘‘half-saturated’’ densities,
whereas the oversettlement in the edge by edge-insensitive species was slightly lower (Fig. 3b).
Our model produced the expected result that nest
densities in habitat patches were lower for edge-sensitive species. Surprisingly, however, edge avoidance
did not reduce nest densities in small patches compared
to larger patches, and nest densities actually increased
exponentially as patch size decreased for edge-insensitive species. Also unexpected was the result that nests
of edge-sensitive species were found in the edge habitat
at frequencies only slightly below that expected based
on edge area. Edge-insensitive species actually had
higher than expected nest densities in edge habitat.
However, when overall patch densities were reduced
by half, the magnitude of edge avoidance was much
more pronounced for edge-sensitive species. Finally,
both types of species showed marked increases in nest
densities in linear habitat patches compared to square
patches.
These results have potentially important implications
for the conservation biology and behavioral ecology of
birds. Determining if edge avoidance, area sensitivity,
and patch size effects exist for a particular species may
be more complex than simply looking for decreased
nest densities near edges (e.g., Bock et al. 1999) or in
small patches (e.g., Ambuel and Temple 1983, Howe
1984, Lynch and Whigham 1984, Herkert 1994, Bol-
Effects of shape
To investigate the effects of shape independent of
patch area, we varied the dimensions of a 25-ha patch.
Interestingly, as the patch became more linear, nest
density actually increased for both edge-sensitive and
edge-insensitive species (Fig. 4). The biggest increase,
however, occurred for edge-insensitive species. Furthermore, the increase occurred when dnest became similar to, or greater than, the width of the patch. The
consequence of these patch dimensions is that females
begin spacing themselves out linearly along the length
of the patch, resulting in each female occupying a
smaller area of nesting habitat while maintaining an
equal spacing. The end result is a higher nest density.
FIG. 5. Number of nests in a 25-ha patch as a function
of g. For g, ‘‘0’’ refers to having no detrimental effect of
edge; ‘‘max’’ refers to having probability of settling Ps 5 0
for all edge habitat. Note the x-axis log scale.
1572
ERIC K. BOLLINGER AND PAUL V. SWITZER
linger 1995). For example, our results suggest that species may maintain relatively constant densities regardless of patch size, yet still be edge sensitive. In addition,
edge-sensitive species may not show markedly lower
nest densities near habitat edges. However, our results
also indicate that when population size for edge-sensitive species is low, edge avoidance will be more pronounced. If reproductive success is higher away from
edges, this pattern will lead to higher per capita fitness
when population size is low. This is good news from
a conservation perspective because this pattern should
help declining populations of edge-sensitive species to
recover. It is similar to a result from Donovan and
Lamberson’s (2001) model, in which they concluded
that by preferentially selecting larger patches, birds
could to some extent counteract the negative effects of
habitat fragmentation. Our results are analogous, but
from the perspective of a single patch. Finally, increased nest densities near edges or higher nest densities in small or more linear patches may not indicate
that a particular species is an ‘‘edge-preferring’’ species, but simply ‘‘edge-insensitive.’’ For example, the
Bobolink, a species that actually avoids wooded edges
(O’Leary and Nyberg 2000), had higher than expected
nest densities near grassland–old field, grassland–rowcrop, and grassland–pasture edges in New York and
Illinois grasslands (E. K. Bollinger and T. A. Gavin,
unpublished data).
The patterns of nest densities generated by our model
are primarily a consequence of our settlement rule that
nests must be a minimum distance apart, i.e., dnest (see
Muldal et al. 1985). For example, nest density increased near the edge for ‘‘edge-insensitive’’ species
because less ‘‘exclusive area’’ per nest is needed near
edges. Because no nests are located outside the patch,
nests can be ‘‘lined up’’ along the edge every dnest (with
a minimum of [p(dnest/2)2]/2 of exclusive area), whereas
nests in the interior occur at lower densities approximating one nest per circular area of ;p(dnest/2)2.
Thus, the validity and applicability of our results rest
largely on the extent to which female birds base their
settlement and nest site selection decisions on distances
to other females’ nests. Certainly, ‘‘conventional wisdom’’ suggests that nests for monogamous, territorial
species (including many forest and grassland interior
species) are often ‘‘overdispersed’’ (see Brown 1969,
Krebs 1971, Newton et al. 1977, Andrén 1991, Yasukawa et al. 1992, Schieck and Hannon 1993), indicating
that females may be using some type of ‘‘distance rule’’
for nest location (Muldal et al. 1985). In addition, for
visual species like most birds, proximate limitations on
the detection of neighbors may be responsible for the
nest and territory spacings (Eason 1992, Eason and
Stamps 1992). Therefore, having increased internest
distances in linear patches (which would be necessary
if females were using an area-based settlement rule)
may not be realistic, based on the proximate mechanisms that females use for spacing. Females may also
Ecological Applications
Vol. 12, No. 6
base their settlement decisions on overall bird/nest densities and resource levels (e.g., Fretwell and Lucas
1969), especially if reproductive success declines with
nest density. Therefore, one could find equal nest densities across patch sizes resulting from either edge
avoidance and the settlement rules used in our model
or from Fretwell and Lucas’ (1969) ‘‘ideal free’’ or
‘‘ideal despotic’’ (Fretwell 1972) distributions. Unfortunately, little empirical evidence exists on the proximate settlement rules used by female birds, and we are
left to infer mechanisms based on existing patterns such
as overdispersion.
What does our model have to say about species that
are found in lower densites in small (or more linear)
patches, or that are found in lower densities near edges
(i.e., habitat-interior species)? It certainly suggests that
‘‘edge avoidance’’ by females in homogeneous patches
using a minimum internest distance as their settlement
rule is insufficient to produce this pattern, except at
reduced overall densities. What else, then, is likely to
be at work?
First, the reduced nest densities found near edges or
in small patches may simply be an artifact of elevated
predation rates in these locations. If predation rates are
high, fewer active nests are available to be found even
if the number of nests actually initiated does not differ
between edge and interior habitats (or small vs. large
patches).
Second, nest densities may be constrained by male
numbers and territory size in monogamous species. If
males are territorial, Stamps and Bueckner (1985) and
Stamps et al. (1987) have shown that territory size
should be larger near habitat edges if adjacent habitat
is unsuitable for intruders, because territory owners
have to expend less energy defending territory borders.
Lowered defense cost per area allows males to expand
their territory size. Thus, if females are unlikely to mate
polygynously, then nest density may decline near edges
because fewer (but larger) male territories occur there.
On the other hand, it has been suggested that female
Red-winged Blackbirds (Agelaius phoeniceus) and
Bobolinks respond more to habitats than to potential
mates and their territory boundaries when selecting
nest sites (Searcy 1979, Wootten et al. 1986).
Third, even if females ‘‘accept’’ the reduced paternal
care that typically accompanies polygynous mating
systems (e.g., Martin 1974), they may not accept the
reduced area per nest that follows directly from our
model’s settlement rules. This is especially likely to be
the case for species with ‘‘all-purpose’’ (or type A)
territories (as in Nice [1941] and Hinde [1956]) in which
most food is gathered inside the borders of the territory.
However, for species gathering a significant fraction of
food outside their territory, this consequence of our
settlement rule may be more realistic. Thus, we might
predict that species with these types of territories (e.g.,
Red-winged Blackbirds and Bobolinks) may be more
likely to be found in higher densities in small patches
December 2002
EDGE AVOIDANCE MODEL
and our more linear patches (if they are edge insensitive), or to be found in similar densities in small and
large patches (if they are edge sensitive). Empirical
evidence suggests that this is not the case for the Bobolink (e.g., Bollinger and Gavin 1992, Helzer and Jelinski 1999), but may be true for Red-winged Blackbirds (Herkert 1994).
Fourth, the presence of some adjacent interior habitat
may affect the settlement of birds within the edge of
that habitat. Even though the vegetation, climate, and
food levels may be similar between the edge of a large
habitat and a similar-sized, but isolated narrow strip of
habitat, birds may be more likely to settle in the edge
of the large habitat than in the narrow strip. Such a
preference may occur because birds may prefer to settle
near other birds (conspecific attraction) that may be in
the interior habitat (Stamps 1988).
Finally, females may tend to settle first in the centers
of patches (or in large patches), either largely by chance
or because these areas are preferred (Bensch and Hasselquist 1991; for opposite pattern, see Lanyon and
Thompson 1986). Small patches and areas near edges
would still be considered acceptable nesting habitat.
Thus, later females would occupy edge habitat (or
smaller patches), ‘‘filling up’’ the patches. However, if
female numbers were low, then only large patches and
interior areas would be occupied. This is supported by
our result that edge avoidance was markedly more pronounced at reduced patch densities. Higher proportions
of unmated males in small patches may also be a manifestation of this phenomenon (Gibbs and Faaborg
1990, Porneluzi et al. 1993).
In conclusion, our relatively simple model has highlighted the potential complexity and need for future
research in the seemingly straightforward phenomenon
of edge avoidance. Nest mapping may be insufficient
to determine the extent to which a species avoids or
prefers nesting near habitat edges. This is especially
true given that high predation rates leave fewer active
nests to be found by investigators. Documenting edge
avoidance is also complicated by its likely dependence
on the overall patch density of nests. Pronounced edge
avoidance may occur at low densities, yet this pattern
may largely disappear at high densities. Furthermore,
interactions between edges and interior may make settlement different between the edges of large patches
and ‘‘edge-like’’ habitat in narrow habitat strips. Finally, we clearly need more information on the settlement rules used by females if we are to accurately
demonstrate the existence and assess the potential consequences of edge avoidance. For example, settlement
patterns of males have been shown to markedly impact
territory size (Knapton and Krebs 1972). Detailed observations of marked females immediately following
their arrival at habitat patches, as well as a comparison
of nest densities, territory sizes, polygyny levels, and
use of habitat off territory, would help our understanding of this interesting and important phenomenon.
1573
ACKNOWLEDGMENTS
We thank Eastern Illinois University’s Council for Faculty
Research for helping to fund our research. Special thanks are
due Yoriko Saeki for running many of the model trials, B.
Luttbeg, J. Switzer, and I. Switzer who helped with some
details of the model, and D. Johnson, J. Brawn, and an anonymous reviewer for comments on the manuscript.
LITERATURE CITED
Ambuel, B., and S. A. Temple. 1983. Area-dependent changes in the bird communities and vegetation of southern Wisconsin forests. Ecology 64:1057–1068.
Andrén, H. 1991. Predation: an overrated factor for overdispersion of birds’ nests? Animal Behaviour 41:1063–
1069.
Bensch, S., and D. Hasselquist. 1991. Territory infidelity in
the polygynous great reed warbler Acrocephalus arundinaceus: the effect of variation in territory attractiveness.
Journal of Animal Ecology 60:857–871.
Blake, J. G., and J. R. Karr. 1984. Species composition of
bird communities and the conservation benefit of large versus small forests. Biological Conservation 30:173–187.
Blancher, P. J., and R. J. Robertson. 1985. Predation in relation to nest spacing of kingbird nests. Auk 102:654–658.
Bock, C. E., J. H. Bock, and B. C. Bennett. 1999. Songbird
abundance in grasslands at a suburban interface on the
Colorado High Plains. Studies in Avian Biology 19:131–
136.
Bollinger, E. K. 1995. Successional changes and habitat selection in hayfield bird communities. Auk 112:720–730.
Bollinger, E. K., and T. A. Gavin. 1989. The effects of site
quality on breeding site fidelity in bobolinks. Auk 106:584–
594.
Bollinger, E. K., and T. A. Gavin. 1992. Eastern bobolink
populations: ecology and conservation in an agricultural
landscape. Pages 497–506 in J. M. Hagan III and D. W.
Johnston, editors. Ecology and conservation of neotropical
migrant landbirds. Smithsonian Institution Press, Washington, D.C., USA.
Boulinier, T., and E. Danchin. 1997. The use of conspecific
reproductive success for breeding patch selection in territorial migratory species. Evolutionary Ecology 11:505–
517.
Brawn, J. D., and S. K. Robinson. 1996. Source–sink population dynamics may complicate the interpretation of longterm census data. Ecology 77:3–12.
Brittingham, M. C., and S. A. Temple. 1983. Have cowbirds
caused forest songbirds to decline? BioScience 33:31–35.
Brown, J. L. 1969. Territorial behavior and population regulation in birds. Wilson Bulletin 81:293–329.
Burke, D. M., and E. Nol. 1998. Influence of food abundance,
nest-site habitat, and forest fragmentation on breeding ovenbirds. Auk 115:96–104.
Donovan, T. M., and R. H. Lamberson. 2001. Area-sensitive
distributions counteract negative effects of habitat fragmentation on breeding birds. Ecology 82:1170–1179.
Eason, P. K. 1992. Optimization of territory shape in heterogeneous habitats: a field study of the red-capped cardinal
(Paroaria gularis). Journal of Animal Ecology 61:411–
424.
Eason, P. K., and J. A. Stamps. 1992. The effect of visibility
on territory size and shape. Behavioral Ecology 3:166–172.
Fretwell, S. D. 1972. Populations in a seasonal environment.
Princeton University Press, Princeton, New Jersey, USA.
Fretwell, S. D., and H. L. Lucas. 1969. On territorial behavior
and other factors influencing habitat distribution in birds.
Part I. Theoretical development. Acta Biotheoretica 19:16–
36.
Galbraith, H. 1988. Effects of agriculture on the breeding
1574
ERIC K. BOLLINGER AND PAUL V. SWITZER
ecology of lapwings Vanellus vanellus. Journal of Applied
Ecology 25:487–503.
Gates, J. E., and L. W. Gysel. 1978. Avian nest dispersion
and fledging success in field–forest ecotones. Ecology 59:
871–883.
Gavin, T. A. 1984. Broodedness in Bobolinks. Auk 101:179–
181.
Gibbs, J. P., and J. Faaborg. 1990. Estimating the viability
of ovenbird and Kentucky warbler populations in forest
fragments. Conservation Biology 4:193–196.
Helzer, C. 1996. The effects of wet meadow fragmentation
on grassland birds. Thesis. University of Nebraska, Lincoln, Nebraska, USA.
Helzer, C., and D. E. Jelinski. 1999. The relative importance
of patch area and perimeter-area ratio to grassland breeding
birds. Ecological Applications 9:1448–1458.
Herkert, J. R. 1994. The effects of habitat fragmentation on
Midwestern grassland bird communities. Ecological Applications 4:461–471.
Herkert, J. R., S. A. Simpson, R. L. Westemeier, T. L. Esker,
and J. W. Walk. 1999. Response of northern harriers and
short-eared owls to grassland management in Illinois. Journal of Wildlife Management 63:517–523.
Hinde, R. A. 1956. The biological significance of territories
of birds. Ibis 98:340–369.
Howe, R. W. 1984. Local dynamics of bird assemblages in
small forest habitat islands in Australia and North America.
Ecology 65:1585–1601.
Johnson, D. W., and L. D. Igl. 2001. Area requirements of
grassland birds: a regional perspective. Auk 118:24–34.
Johnson, R. G., and S. A. Temple. 1986. Assessing habitat
quality for birds nesting in fragmented tallgrass prairies.
Pages 245–249 in J. Verner, M. L. Morrison, and C. J.
Ralph, editors. Wildlife 2000: modeling habitat relationships of terrestrial vertebrates. University of Wisconsin
Press, Madison, Wisconsin, USA.
Johnson, R. G., and S. A. Temple. 1990. Nest predation and
parasitism of tallgrass prairie birds. Journal of Wildlife
Management 54:106–111.
Kapos, V. 1989. Effects of isolation on the water status of
forest patches in the Brazilian Amazon. Journal of Tropical
Ecology 5:173–185.
Knapton, R. W., and J. R. Krebs. 1972. Settlement patterns,
territory size, and breeding density in the song sparrow
(Melospiza melodia). Canadian Journal of Zoology 52:
1413–1420.
Krebs, J. R. 1971. Territory and breeding density in the great
tit, Parus major L. Ecology 52:2–22.
Lanyon, S. M., and C. F. Thompson. 1986. Site fidelity and
habitat quality as determinants of settlement pattern in male
Painted Buntings. Condor 88:206–210.
Lokemoen, J. T., H. F. Duebbert, and D. E. Sharp. 1984. Nest
spacing, habitat selection, and behavior of waterfowl on
Miller Lake Island, North Dakota. Journal of Wildlife Management 48:309–321.
Lynch, J. F., and D. F. Whigham. 1984. Effects of forest
fragmentation on breeding bird communities in Maryland,
USA. Biological Conservation 28:287–324.
Martin, S. G. 1974. Adaptations for polygynous breeding in
the bobolink Dolichonyx oryzivorus. American Zoologist
14:109–119.
Matter, S. F. 2000. The importance of the relationship between population density and habitat area. Oikos 89:613–
619.
Mesquita, R. C. G., P. Delamonica, and W. F. Laurance. 1999.
Effect of surrounding vegetation on edge-related tree mortality in Amazonian forest fragments. Biological Conservation 91:129–134.
Muldal, A., H. L. Gibbs, and R. J. Robertson. 1985. Preferred
Ecological Applications
Vol. 12, No. 6
nest spacing of an obligate cavity-nesting bird, the Tree
Swallow. Condor 87:356–363.
Newton, I., M. Marquiss, D. N. Weir, and D. Moss. 1977.
Spacing of sparrowhawk nesting territories. Journal of Animal Ecology 46:425–441.
Ney-Nifle, M., and M. Mangel. 2000. Habitat loss and changes in the species–area relationship. Conservation Biology
14:893–898.
Nice, M. M. 1941. The role of territory in bird life. American
Midland Naturalist 26:441–487.
O’Leary, C. H., and D. W. Nyberg. 2000. Treelines between
fields reduce the density of grassland birds. Natural Areas
Journal 20:243–249.
Paton, P. W. C. 1994. The effect of edge on avian nest success:
how strong is the evidence? Conservation Biology 8:17–
26.
Petit, K. E., D. R. Petit, and L. J. Petit. 1988. On measuring
vegetation characteristics in bird territories: nest sites vs.
perch sites and the effect of plot size. American Midland
Naturalist 119:209–215.
Porneluzi, P., J. C. Bednarz, L. J. Goodrich, N. Zawada, and
J. Hoover. 1993. Reproductive performance of territorial
ovenbirds occupying forest fragments and a contiguous forest in Pennsylvania. Conservation Biology 7:618–622.
Ranney, J. W., M. C. Bruner, and J. B. Levenson. 1981. The
importance of edge in the structure and dynamics of forest
islands. Pages 67–95 in R. L. Burgess and B. M. Sharpe,
editors. Forest island dynamics in man-dominated landscapes. Springer-Verlag, New York, New York, USA.
Robbins, C. S. 1979. Effect of forest fragmentation on bird
populations. Pages 198–212 in R. M. DeGraaf and K. E.
Evans, editors. Workshop proceedings: management of
north central and northeastern forests for nongame birds.
U.S. Forest Service General Technical Report NC-51.
Robbins, C. S., D. K. Dawson, and B. A. Dowell. 1989.
Habitat area requirements of breeding forest birds of the
middle Atlantic states. Wildlife Monographs Number 103.
Robinson, S. K. 1992. Population dynamics of breeding neotropical migrants in a fragmented Illinois landscape. Pages
408–418 in J. M. Hagan III and D. W. Johnston, editors.
Ecology and conservation of neotropical migrant landbirds.
Smithsonian Institution Press, Washington, D.C., USA.
Robinson, S. K., F. R. Thompson III, T. M. Donovan, D. R.
Whitehead, and J. Faaborg. 1995. Regional forest fragmentation and the nesting success of migratory birds. Science 267:1987–1990.
Rosenzweig, M. L. 1981. A theory of habitat selection. Ecology 62:327–335.
Roth, R. R., and R. K. Johnson. 1993. Long-term dynamics
of a Wood Thrush population breeding in a forest fragment.
Auk 110:37–48.
Schieck, J. O., and S. J. Hannon. 1993. Clutch predation,
cover, and the overdispersion of nests of the Willow Ptarmigan. Ecology 74:743–750.
Searcy, W. A. 1979. Female choice of mates: a general model
for birds and its application to Red-winged Blackbirds
(Agelaius phoeniceus). American Naturalist 114:77–100.
Stamps, J. A. 1988. Conspecific attraction and aggregation
in territorial species. American Naturalist 131:329–347.
Stamps, J. A., and M. Buechner. 1985. The territorial defense
hypothesis and the ecology of insular vertebrates. Quarterly
Review of Biology 60:155–181.
Stamps, J. A., M. Buechner, and V. V. Krishnan. 1987. The
effects of habitat geometry on territorial defense costs: intruder pressure in bounded habitats. American Zoologist
27:307–325.
Sutherland, W. J. 1996. From individual behavior to population biology. Oxford University Press, Oxford, UK.
Sutter, G. C., S. K. Davis, and D. C. Duncan. 2000. Grassland
December 2002
EDGE AVOIDANCE MODEL
songbird abundance along roads and trails in southern Saskatchewan. Journal of Field Ornithology 71:110–116.
Switzer, P. V. 1993. Site fidelity in predictable and unpredictable habitats. Evolutionary Ecology 7:533–555.
Switzer, P. V. 1997. Past reproductive success affects future
habitat selection. Behavioral Ecology and Sociobiology 40:
307–312.
Temple, S. 1986. Predicting impacts of habitat fragmentation
on forest birds: a comparison of two models. Pages 301–
304 in J. Verner, M. L. Morrison, and C. J. Ralph, editors.
Wildlife 2000: modeling habitat relationships of terrestrial
vertebrates. University of Wisconsin Press, Madison, Wisconsin, USA.
Vickery, P. D., M. L. Hunter, Jr., and S. M. Melvin. 1994.
Effects of habitat area on the distribution of grassland birds
in Maine. Conservation Biology 8:1087–1097.
Wenny, D. G., R. L. Clawson, J. Faaborg, and S. L. Sheriff.
1993. Population density, habitat selection and minimum
area requirements of three forest-interior warblers in central
Missouri. Condor 95:968–979.
Whitcomb, R. F., J. F. Lynch, M. K. Klimkiewicz, C. S. Robbins, B. L. Whitcomb, and D. Bystrak. 1981. Effects of
forest fragmentation on avifauna of the eastern deciduous
forest. Pages 125–206 in R. L. Burgess and B. M. Sharpe,
1575
editors. Forest island dynamics in man-dominated landscapes. Springer-Verlag, New York, New York, USA.
Wiens, J. A. 1969. An approach to the study of ecological
relationships among grassland birds. Ornithological Monographs Number 8.
Wiens, J. A. 1989. The ecology of bird communities. Volume
2. Processes and variations. Cambridge University Press,
Cambridge, UK.
Wiens, J. A. 1994. Habitat fragmentation: island v landscape
perspectives on bird conservation. Ibis 137:S97–S104.
Wilcove, D. S. 1985. Nest predation in forest tracts and the
decline of migratory songbirds. Ecology 66:1211–1214.
Winter, M., D. H. Johnson, and J. Faaborg. 2000. Evidence
for edge effects on multiple levels in tallgrass prairie. Condor 102:256–266.
Wootten, J. T., E. K. Bollinger, and C. J. Hibbard. 1986.
Mating systems in homogeneous habitats: the effects of
female uncertainty, knowledge costs, and random settlement. American Naturalist 128:499–412.
Yasukawa, K., R. A. Boley, J. L. McClure, and J. Zanocco.
1992. Nest dispersion in the Red-winged Blackbird. Condor 94:775–777.
Zanette, L., P. Doyle, and S. M. Trémont. 2000. Food shortage
in small fragments: evidence from an area-sensitive passerine. Ecology 81:1654–1666.