Behavioral Ecology
doi:10.1093/beheco/arq108
Advance Access publication 00 Month 2010
Large body and small brain and group sizes are
associated with predator preferences for
mammalian prey
Susanne Shultza,b and Laura V. Finlaysonb
Institute of Cognitive and Evolutionary Anthropology, University of Oxford, 64 Banbury Road, Oxford
OX2 6PN, UK and bSchool of Biological Sciences, University of Liverpool, Liverpool L69 7ZB, UK
a
Predation is a major force in shaping biological communities, both over ecological and evolutionary timescales. In response to
predation pressure, prey have evolved characteristics designed to mitigate predation pressure. We evaluated predator foraging
biases in relation to prey characteristics across 16 vertebrate communities. We show that although predator biases vary, some prey
traits are consistently associated with predator diet composition. Within their acceptable prey size range, predators show positive
bias toward larger bodied prey, small-brained prey (controlling for body size), small group size, and terrestriality. Thus, whether
predator foraging decisions are passive or active, predator choice exerts differential pressure on prey species according to prey
characteristics. Predator biases also were positively associated with early age at maturity, supporting the role of mortality in driving
life-history characteristics. These results support several theoretical models of predation including its role as a selective force
driving evolutionary changes in life history, brain size and sociality, optimal diet theory, and antiapostatic predation. Key words:
encephalization, foraging ratio, predation, sociality, selectivity, wild dogs. [Behav Ecol 21:1073–1079 (2010)]
redation is a key driver of biological systems over both ecological and evolutionary time. Over ecological timescales,
predation plays roles in regulating population dynamics
(e.g., Krebs et al. 1995; Korpimäki and Norrdahl 1998) and
community richness and composition (Hairston et al. 1960;
Paine 1980). Over evolutionary timescales, predation pressure
selects for major morphological and behavioral adaptations
(Dawkins and Krebs 1979; Caro 2005). Thus, prey characteristics associated with differential predation pressure should
both be important in shaping community structure as well
as ultimately influencing long-term evolutionary trends in
prey morphology and behavior. Apart from selection for antipredator traits, extrinsic mortality drives life-history strategies,
such that high mortality rates are associated with early maturity and high reproductive rates (Charnov 1993). Thus, behavioral and morphological characteristics that impact on
realized mortality will have knock on consequences for species
life-history strategies.
Prey employ a wide arsenal of morphological, chemical,
life-history and behavioral traits in response to predation,
which are assumed to effectively reduce mortality rates.
Behavioral strategies that prey use to mitigate predation risk
include increasing vigilance, refuge use, altering activity patterns, and trade-offs in resource patch decisions (reviewed in
Lima 1998). Aggregation also reduces predation risk, owing
to the benefits of dilution, confusion, early warning, and
coordinated defense (Pulliam 1973). A vast literature has
documented grouping responses to predation risk (Lima
1998). Individuals in smaller groups increase investment
in antipredator behavior, and group size has been shown
P
Address correspondence to S. Shultz. E-mail: susanne.shultz
@anthro.ox.ac.uk.
Current Address: L. V. Finlayson is now at Fern Cottage Veterinary
Surgery
Received 5 April 2010; revised 2 June 2010; accepted 3
June 2010.
The Author 2010. Published by Oxford University Press on behalf of
the International Society for Behavioral Ecology. All rights reserved.
For permissions, please e-mail: [email protected]
to change in response to high predation risk (e.g., Whitfield
1988).
A more general facultative trait that could allow prey to reduce predator capture success is behavioral flexibility. In the
wider literature, brain size has often been used as a proxy
for cognitive capacity (and hence of behavioral flexibility).
Behavioral flexibility and innovativeness are usually presented
in ecological terms: larger brained individuals/species will be
better able to employ novel resources (e.g., Lefebvre et al.
1997) and respond to ecological change (Shultz et al. 2005;
Sol et al. 2005) and lower mortality rates (Sol et al. 2007).
However, behavioral flexibility may also provide prey species
with adaptive and unpredictable anti-predator responses. In
fact, Jerison (1973) originally proposed the evolutionary trend
toward larger relative brain size across ungulates and carnivores to be the result of a predator–prey coevolutionary arms
race. An inverse correlation between relative prey brain size
and predator diet composition has been demonstrated in
a limited selection of Afrotropical vertebrate predator communities (Shultz and Dunbar 2006).
Predators, however, are not passive participants in their relationship with prey and should show biases for profitable
and vulnerable prey. These choices can either be active,
where predators preferentially search for specific prey items,
or passive, where biases result from predators being unable
to locate, capture, or process certain types of prey items
(Sih and Christensen 2001). In either case, the result of biases
in predator foraging will result in differential predation pressure on prey individuals. In addition to the prey strategies
summarized earlier, vulnerability can be affected by condition,
spatial positioning, and variation in morphology/behavior
across age/stage classes (Lima 1998; Quinn and Cresswell
2004). An example of spatial positioning is habitat use by both
predators and prey. A terrestrial predator will encounter more
terrestrial prey than arboreal species, resulting in a passive
bias due to encounter frequency. A few studies have demonstrated an association between prey vulnerability and individual
predation risk (e.g., Cresswell 1994, see also Lima 1998; Quinn
and Cresswell 2004). In addition to vulnerability, optimal diet
Behavioral Ecology
1074
theory (ODT) also predicts that predators should select prey
based on their profitability because predators have to expend
energy capturing prey; they should prefer those that offer the
highest net returns (Emlen 1966; Krebs et al. 1977).
Despite the wide-ranging literature that documents prey behavior and decision making under predation risk, the evidence
that evaluates predator foraging decisions in relation to prey
characteristics is more limited (Lima 2002). A few studies have
documented predator attack rates on prey groups of different
sizes in experimental systems, but the connection between attack and capture rates does not appear to be straightforward
(Turchin and Karieva 1989; Krause and Godin 1995).
Although captive studies are informative about predator foraging decisions, they are unable to accurately reflect the choices
and conditions predators face in wild populations (Lima 1998).
The data relating predator foraging to prey behavior from wild
populations are more limited (e.g., Lindstrom 1989; Cresswell
1994; Hebblewhite and Pletscher 2002). However, most of these
studies focus on single communities or predator species. Some
studies have evaluated predator foraging decisions using preferences derived from the ratio of species in the diet to that in
the environment. Götmark and Post (1996) used sparrowhawk
foraging ratios to demonstrate a link between prey exposure
and habitat use and predation risk in passerines. Götmark
and Andersson (2005) later demonstrated that predation rates
are inversely related to breeding aggregation size in great tits.
Shultz et al. (2004) used a community-level approach to evaluate relative predation risk over a suite of predator and prey
species and showed that high predation risk is associated with
terrestriality and small group size.
Apart from differences in prey defense and vulnerability,
a wide literature suggests that foraging predators select prey
based on their encounter frequency or relative abundance
(Pulliam 1974; Van Orsdol 1984). Searching predators are
expected to show apostatic predation or biases toward the
most abundant prey types (Murdoch 1969). Apostatic predation has been proposed as a primary mechanism for promoting the maintenance of polymorphisms and species richness
in prey communities (e.g., Sherratt and Harvey 1993).
Although there still is not widespread evidence for apostatic
predation in natural populations, positive frequencydependent selection has been demonstrated in a number of
experimental and natural systems (Allen and Greenwood
1988; Bond and Kamil 1998; Bond 2007). In contrast, in
antiapostatic predation, predators preferentially prey on rare
or distinctive prey types (Pielowski 1961; Bond 1983). Much of
the evidence supporting frequency-dependent selection by
predators comes from foraging experiments using passive or
immobile prey; however, predator choice has been shown to
be very different when predators are utilizing mobile prey
(Sih and Christensen 2001).
Terrestrial mammalian predators offer an ideal system to
examine commonalities across population in how prey characteristics relate to predator foraging biases. Within populations,
considerable work has been done on foraging behavior. For example, wild dogs (Lycaon pictus) show seasonal biases toward
temporally abundant prey and large-bodied prey (Kruger et al.
1999; Pole et al. 2004), lions show preferences toward small
groups and profitable prey (Scheel 1993), and tigers selectivity
choose large-bodied prey (Karanth and Sunquist 1995). Such
studies do much to illuminate our understanding of the shortterm interactions between predators and prey. However, there
has, to date, been no systematic evaluation of cross-community
patterns of predator foraging choices.
Here, we utilize a range of studies that have been completed
on vertebrate predators to extend the approach used in Shultz
and Dunbar (2006). We provide the first broad evaluation of
the relationship between prey behavioral and ecological char-
acteristics and predator diet composition at the population
level. We compiled diet and community composition data
from terrestrial communities. We predicted the following:
1. According to ODT, predators should prefer larger prey
across their suite of prey species.
2. Prey in smaller groups should be preferred to those in
larger groups.
3. Predators should show apostatic selection and show
biases toward abundant prey.
4. Terrestrial predator diets will be biased toward terrestrial
prey.
5. If brain size is a proxy for behavioral flexibility, relatively
large-brained prey should be less preferred than smallbrained prey.
6. If life history is a derivative of mortality rates, predator
biases should be associated with faster life-history strategies (i.e., earlier age at first reproduction).
MATERIALS AND METHODS
The predator diet composition data were collected from reliable sources providing concurrent diet and community composition estimates (see Table 1 for data and sources).
Although there are many more studies than we have included
on mammalian predator foraging behavior, we were only able
to include those studies that had a contemporary estimate of
prey density. The communities included savannah, temperate
rainforest, and alpine habitats, encompassing data from 3
canids (dhole [Cuon alpines]; wild dog [L. pictus], and wolf
[Canis lupus]), 6 felids (leopard [Panthero pardus], golden
cat [Felis aurata], tiger [P. tigris tigris], jaguar [P. onca], puma
[F. concolor], and ocelot [F. pardalis]), and chimpanzees (Pan
troglodytes). As we were unable to collate information on population density for all possible prey species, the data set contains a subsample of total available prey species. Likewise, it is
not possible to determine which species that do not occur in
the diet are potential prey and which are not. We assume that
the omitted species do not share characteristics that would
bias the analyses. For analyses, we aggregated predators into
type: chimpanzee, canid, forest felids, and tigers. In total, we
compiled data from 16 terrestrial communities, incorporating
9 species of predators, 47 species of prey, resulting in a total of
131 data points.
The prey characteristics we evaluated were body size, brain
size, group size, population density, habitat use, and strata use
(see Supplementary Materials for data and sources). We also
evaluated the relationship between predator biases and prey
life history using age at first reproduction. All data were
checked for normality and transformed where necessary: In
the analyses, we used log10 (body size), log10 (brain Size),
ln (geometric mean of reported group size range), and
log10 (population density).
To evaluate the relationship between predator biases and prey
brain size, we have used a different approach to that in Shultz
and Dunbar (2006). Rather than calculating a relative brain size
via regression residuals, we have incorporated both brain and
body size in models to evaluate whether there is an independent
effect of each variable on predator foraging biases. The rationale behind this is that the calculation of residuals is fraught
with criticism, as a least squares regression underestimates
slopes and the complement of species used in the regression
biases the resulting residuals (also potentially introducing phylogenetic confounds) (see Freckleton 2002). Additionally, calculating brain size residuals over a wide group of distantly
related taxa will inevitably introduce scaling and ontogenetic
confounds (Deacon 1990). However, for comparison, we also
used residual brain size calculated from a major axis regression
and reran the analyses (see Supplementary Materials).
Shultz and Finlayson • Prey characteristics and predator diet choice
Table 1
Sources studies and species
v2
Predator
Site
Source
Chimpanzee
Mahale
Kibale
Tai
Nishida et al. (1992)
**
Mitani and Watts (1999) **
Boesch and Boesch**
Achermann (2000)
Forest felids
Jaguar, puma, ocelot Manu
Leopard
Nagarahole
Bandipur
Serengeti
Tai
Leopard, golden cat Ituri
Tiger
Emmons (1987)
Karanth and Sunquist
(1995)
Johnsingh (1983, 1992)
Schaller (1972)
Zuberbühler and Jenny
(2002)
Hart et al. (1996)
Bandipur
Nagarahole
Johnsingh (1983)
Karanth and Sunquist
(1995)
Pench
Biswas and Sankar
(2002)
Ranthambore Bagchi et al. (2003)
Canids
Dhole
Bandipur
Nagarahole
Wolf
Poland BN
Banff
Alps
Appenines
Selous
Serengeti
Savé
Wild dog
Johnsingh (1983)
Karanth and Sunquist
(1995)
Jedrzejewski et al.
(1992)
Huggard (1993)
Gazzola et al. (2005)
Merrigi et al. (1996)
Creel and Creel (2002)
Schaller (1972)
Pole et al. (2004)
**
**
NS
**
**
**
**
**
where u represents the number of each prey item in the diet, p
represents the proportion of each prey item in the environment, U is the total number of prey individuals recorded, and
n is the total number of prey species.
To evaluate the association between selectivity and prey characteristics, we then used a 2-pronged analytical approach. First,
we correlated prey characteristics with selectivity across each
predator type. Second, we used stepwise information criteria
(Akaike’s Information Criterion [AIC]) to select the bestperforming model relating to predator diet composition
(step.aic command in R package MASS). We ran the model
selection procedure twice, once with and once without
interactions.
Although it is possible that related predators show similar
preferences for prey characteristics, the nature of the data
(with nested prey within predator types) is not amenable to
conventional phylogenetic analyses. Additionally, we assume
that predator diet choices are determined by prey characteristics, not by the degree of relatedness between prey species,
effectively rendering prey-independent data points.
**
**
**
*
*
*
**
NS
**
**
NS, not significant. v2 term indicates whether prey selection differed
significantly from random.
*P , 0.05, **P , 0.01
If a predator is truly opportunistic, the most abundant prey
will also be the most abundant item in the diet. Thus, to determine whether predators are utilizing prey disproportionately to
their abundance, we calculated a standardized forage ratio. Prey
occurrence in the diet and abundance in the environment were
used to calculate selectivity (bias) for each prey species, then
standardized to allow for value distribution between 0 and 1 using the following calculation (from Manly et al. 1993):
ŵi
Standard forage ratio : Bi ¼ P
;
n
ŵi
i¼1
oi
pi
1075
where ŵi ¼
and ŵi ¼ forage ratio for species i
oi ¼ Proportion of species i in diet
pi ¼ proportion of species i available in the environment
An equal proportion of prey in the diet and the community
indicates no bias toward or against a species. To normalize the
standardized ratios, they were arcsin and square root transformed for all analyses. Where we refer to preferences, we
do not infer active choice by the predators; we simply indicate
a positive bias.
To determine whether predators’ diets were more selective
than random, we used equation of Manly et al. (1993):
n X
ui
x2 ¼ 2
ui ln
;
U
pi
i¼1
RESULTS
We first correlated predator selectivity with prey characteristics
(Table 2). Residual brain size was negatively correlated across
all predator types except for tigers. Likewise, there was a negative correlation between selectivity and age at first reproduction for all predators except tigers. Body size was positively
correlated with selectivity for forest felids and tigers. Both of
these predator types also showed a bias toward rare prey species. All predators except for chimpanzees were biased toward
terrestrial prey. Canids and tigers both showed a bias toward
prey in smaller groups.
We then used AIC stepwise model selection to identify the
best-performing model (Table 3). For the main effects model,
body size, brain size, group size, and strata use were retained
in the best-performing model. Predator selectivity was positively associated with body size and negatively associated with
brain size and group size. Predators showed an overall bias
toward terrestrial prey. For the model selection including interactions, the best-performing model retained all the main
effects of the first model but additionally included interaction
terms between predator type and each prey characteristics
(Tables 4 and 5). The interaction between predator type
and brain size and group size are shown in Figure 1. Chimpanzees are strongly biased toward small brain prey, whereas
tigers are less influenced by prey brain size. Chimpanzees
showed the least effect of group size on selectivity (the overall
effect across predators was negative, the apparently positive
effect in Figure 1a is relative to the canid baseline).
We reran the model selection procedure with relative brain
size and obtained very similar results (see Supplementary Materials). We also include correlations and minimum adequate
models for individual predators in the Supplementary Materials.
Prey behavior
In order to understand the variation in biases across predators,
we evaluated the influence of habitat type on the relationship
between prey characteristics and predator biases (Table 6). For
open habitat species, group size and population density were
negatively correlated with predator bias. In closed habitats,
predator bias was positively associated with body size and negatively associated with brain size. For mixed habitats, predator
bias also was positively associated with body size and negatively
associated with brain size and group size. Prey behavior varied
across habitats. In open habitats, large-bodied animals gathered in larger groups than small-bodied prey in the same
Behavioral Ecology
1076
Table 2
Pearson correlation coefficients between selectivity and prey characteristics
Predator
Canid (n ¼ 42)
Chimp (n ¼ 21)
Felid (n ¼ 48)
Tiger (n ¼ 19)
All predators (n ¼ 130)
Log density
MA residual brain size
Log brain
Log body size
Ln group size
Strata
Age first reproduction
20.138
20.340*
20.147
0.062
20.387*
20.462**
20.320*
20.339
20.622**
20.169
0.244
20.045
20.067
20.398
20.325*
20.363*
0.112
0.300*
20.202
20.601**
20.495**
20.653**
0.047
0.784**
0.665**
20.478*
20.543*
20.197
20.323**
20.354**
0.214*
0.381**
20.233**
20.483**
20.420**
MA, major axis. A negative correlation with strata use (1 ¼ terrestrial and 2 ¼ arboreal) indicates a bias toward terrestrial prey.
*P , 0.05, **P , 0.01.
environment (Pearson’s r ¼ 0.51, n ¼ 15, P ¼ 0.03). In contrast, in closed habitats, small-bodied species were in larger
groups (r ¼ 20.24, n ¼ 79, P ¼ 0.03). Overall group size was
highest in open habitats, intermediate in mixed habitats, and
lowest in closed habitats (F2,127 ¼ 17.24, P , 0.001; Figure 2).
We compared predator biases with female age at first reproduction to see whether those species that are being particularly
heavily hit by predation tend to have ‘‘faster’’ life-history strategies. There was a significant negative correlation between age
at maturity and predator bias (Table 2) such that ‘‘preferred’’
prey reach sexual maturity earlier than nonpreferred prey.
DISCUSSION
Our results suggest that some prey characteristics are consistently associated with predator biases, including relatively large
body size, small brain size, terrestriality, and small group size.
The prey characteristics associated with biases in diets varied
across predators, leading to an interaction between prey traits
and predation risk across predator species. Thus, to fully understand the predation pressure exerted on a prey species,
it is necessary to consider the impact of multiple predators
(Sih et al. 1998; Shultz et al. 2004) as the cumulative predation rate from all predators will determine the overall mortality rate experienced by a prey species.
A few caveats need to be made with the analyses presented.
First, the data that we have compiled are by their very nature
noisy as there is measurement error in both diet composition
and population density data. We would, however, expect that
the noise introduced by such measurement error would obscure relationships between prey characteristics and predator
foraging biases rather than creating them. Second, we have
used a small subset of prey characteristics. We do not doubt
Table 3
Best-performing model without interactions, with parameter
estimates
Parameter
Standard
Estimate error
Best modela
Body size
0.274
Brain size
20.409
Group size
20.036
Strata
20.109
Global
model
T
P
AIC
D AIC
252.65
0.079
0.134
0.015
0.058
3.46
23.04
22.42
21.88
0.0007
0.003
0.017
0.062
245.42
9.88
7.24
3.95
1.62
7.23
AIC change is model performance when each variable is sequentially
dropped from the best-performing model.
a
Dropped factors: predator type, site, age at first reproduction, and
population density.
that there are additional characteristics such as flight speed,
morphological traits such as defenses, or coloration that impact on predator choices. However, the traits we have used here
explain considerable variation in predator diet composition.
Third, we acknowledge that there is variation within populations, both between individual predators and across time.
However, as this study is the first attempt to construct a
cross-population evaluation of the relationship between prey
characteristics and predator diet composition, we believe
that it provides valuable insight into evolutionary trends in
predator–prey systems.
Most predators were biased toward larger bodied prey within
their suite of prey species. This does not necessarily lead to the
largest prey suffering the highest predation rates, as each predator exploited a different suite of prey body sizes. This bias
toward large-bodied prey is consistent with ODT as the profitability of a prey individual increases with body size (Emlen
1966). That our data support ODT is interesting as across
studies, support for ODT where prey are mobile has been
equivocal (Sih and Christensen 2001). Our data also suggest
that group-hunting predators (wolves and wild dogs) were
able to exploit the largest and most profitable prey classes,
which suggests that cooperative hunting may allow (or force)
predators to utilize more profitable prey. However, these cooperative hunting species did not show biases toward the largest prey items within their range of utilized prey species.
The consistent negative association between group size
and selectivity indicates that group living provides antipredator benefits at the population level. Prey have limited
opportunities to adjust life history or morphology in response
to predation pressure. However, they can facultatively adjust
aggregation behavior. The relationship between group size
and bias was also habitat specific, with group size having the
largest impact on selectivity in open habitats. However, we argue that the best way to understand the impact of group size on
predator foraging is likely to compare capture success across
Table 4
Best-performing model with interaction terms
Parameter
Factors
AIC
Best-performing
modela
Strata, predator, body size,
brain size, group size,
predator 3 brain size,
predator 3 group size,
predator 3 body size, strata
3 predator
265.1
Global model
257.05
Change
—
8.05
a
Dropped variables: site, population density, habitat, and age of first
reproduction.
Shultz and Finlayson • Prey characteristics and predator diet choice
Table 5
Coefficients and effects on model performance for main factors in
best performing model with interactions
Parameter
Standard
Coefficient error
Brain size
20.40
Body size
0.17
Predator
Strata
20.35
Predator 3 brain size
Predator 3 body size
Predator 3 group size
Group size
20.07
Strata 3 predator
T
Table 6
Minimum adequate models for predator biases across the 3 habitat
types
P
D
value AIC
0.23
0.15
21.73 0.09
1.14 0.26
0.15
22.35 0.02
0.02
1077
22.75 0.01
20.4
17.48
12.44
9.98
9.52
7.05
4.8
4.08
2.12
Change in AIC calculated by sequentially dropping each main factor
and associated interaction terms from model.
different group size within species. This has been shown in
a few experimental systems (Krause and Godin 1995) and wild
populations (Lindstrom 1989; Cresswell 1994). Given the
cross-population nature of these analyses, it was not possible
to evaluate the impact of prey group size within populations.
Charnov (1993) proposed that life-history strategies, and
reproductive rates in particular, are driven by mortality rates.
Under high mortality, individuals should mature faster and
reproduce earlier than where mortality rates are low. Our
results are consistent with this proposal as predator biases
were associated with early age at maturity. It is likely that
predators are differentially preying on juveniles as they are
more vulnerable. Thus, species may mature early to
minimize the time spent in vulnerable age classes. However,
we do not have information on age/stage biases in these
predators.
Large relative brain size was consistently associated with
negative biases by predators. Thus, these results support
Jerison’s (1973) conjecture that the evolutionary trend toward larger brains in prey species may be the result of a co-
Figure 1
The relationship between (a) brain size and (b) group size across
predator types and selectivity. Overall species, there are negative
associations between both group size and brain size and selectivity,
but the strength of the effect varies across predators. Canids are the
baseline for both analyses (tigers: dark solid line, chimpanzees: light
solid line, canids: large dashes, and felids: small dashes).
Habitat
Factor
Degrees
of
freedom
Open
Site
Predator
Group size
Population density
Site
Predator
Body size
Brain size
Group size
Predator
Body size
Brain size
3, 8
1, 8
1, 8
1, 8
11, 19
2, 19
1, 19
1, 19
1, 19
3, 61
1, 61
1, 61
Mixed
Closed
B
20.20
20.19
0.46
20.51
20.05
0.52
20.67
F
P
0.40
0.63
14.75
7.01
1.96
0.43
12.29
7.66
5.55
0.17
25.56
13.01
0.76
0.45
0.005
0.03
0.10
0.66
0.002
0.012
0.03
0.92
,0.001
0.001
evolutionary drive between predators and prey. Given that
brain and body size appear to be strongly associated with
predator diet choices, why do not all prey have relatively
small body size and large brain size? The variation in body
size biases shown by different predators indicates that there
is unlikely to be an optimal body size that minimizes overall
predation rates (except when body size is large enough to
fall outside all/most predator body size ranges). Growing
a large brain is a trickier proposition. Not only is it metabolically costly to invest in cognitive architecture but large brain
size is also associated with a slowing down of maturation and
first reproduction (Western 1979; Eisenberg and Wilson
1981; Isler and van Schaik 2009). Thus, life-history trade-offs,
phylogenetic constraint, and multiple adaptive peaks result
in some species opting out of a costly strategy and instead
adopting alternative ‘‘fast’’ life-history strategies. In other
words, species that have ‘‘cheap’’ behavioral characteristics
are better off investing in maturing quickly and maximizing
reproductive rates rather than maturing late and emphasizing survival. That small-brained species are associated with
positive predator bias, which should consequently result in
higher predation/mortality rates, is consistent with Sol et al.
(2007), who demonstrated that large-brained bird species
suffer lower mortality rates than smaller brained species.
Thus, predation, in addition to other environmental factors,
could be a mechanism for the differential mortality they
reported.
Although there were relationships between prey characteristics and selectivity, the strength of the relationships varied
across predators. For example, although tigers showed significant biases toward terrestrial, large-bodied, and low-density
prey, they did not appear to show any bias toward small-brained
prey. The positive correlation between total brain size and selectivity is an artifact of body size effects as there was no relationship between residual brain size and selectivity. Conversely,
canids, chimpanzees, and felids all showed biases toward smallbrained prey. Canids also appear to be biased by prey group
size and strata use and felids by population density and strata,
whereas chimpanzees appear to be less influenced by other
characteristics. Chimpanzees are very different from the other
predators in that they are not obligate carnivores; vertebrate
prey compose a seasonally variable (and minor) component
of their overall diet. Thus, they may make decisions about prey
choice using different criteria than the other predators. What
these results do tell us is that predators key into prey characteristics differently, and any given antipredator strategy may not
be effective against all predators.
Behavioral Ecology
1078
tor may not be effective against all. Large brain size is associated with negative predator biases, and this may be especially
so for cooperatively hunting predators. Finally, we suggest that
trade-offs between longevity and reproductive rates may be
partially driven by brain size, whereby cognitive capacity may
help prey evade predators, but it comes at a cost of low reproductive rates. We also suggest that population or community perspectives are key to understanding the impact of
predation on evolutionary and ecological trends.
SUPPLEMENTARY MATERIAL
Supplementary material can be found at http://www.beheco
.oxfordjournals.org/.
REFERENCES
Figure 2
Mean prey group sizes across open, mixed, and closed habitats.
Brain size was negatively associated with predator biases in
mixed and closed habitats, whereas group size was negatively
associated with predator biases in open habitats. In complex
mosaic habitats, a variety of potential escape strategies (e.g.,
refuge use, flight, crypsis) is likely to be of more benefit than
in open habitats where the options would be limited to fleeing
or safety in numbers. In open habitats, small-bodied species
tended to gather in smaller groups than larger bodied species,
even though group size was negatively associated with predator
biases. There are several potential explanations for this behavior. The first is that the smaller bodied gazelles use mixedspecies associations to augment group size (Fitzgibbon
1990b); thus, the average number of conspecific individuals
within a group may not accurately reflect total group size. The
second is that smaller bodied species are cryptic or restrict
their habitat use to areas closer to cover, especially around
calving where their young use ‘‘prone’’ responses by hiding
in vegetation (Fitzgibbon 1990a). In contrast, larger bodied
species are more likely to use open spaces and adopt communal defense against predators (especially in terms of protecting young) (Jarman 1974; Caro et al. 2004). For overall
models, terrestrial strata use was associated with higher predator foraging biases than arboreal strata use. However, changing strata use to avoid the risks of terrestrial predators is an
option for few of the prey species in our database (mainly
primates).
Given the vast literature on predator-switching behavior, we
predicted that predators might show biases toward the most
common prey species as they would have the highest encounter rates. Although the most common prey species are the most
abundant in the predator diets, we calculated selection based
on relative abundance in the diet. Using this metric, our results
did not demonstrate a consistent bias toward or against prey
species relative to their abundance. In open habitats, predators
appeared to show antiapostatic biases toward rarer species.
This is the first study to tackle cross-population patterns of
mammalian predator selectivity. We provide evidence that
a number of factors assumed to be related to predation avoidance (large group size, habitat use, and body size) are related
to predator foraging behavior. However, we also demonstrate
that predators vary in the characteristics that are associated
with prey choice, meaning that to understand prey investment
in antipredator behavior, it is necessary to consider the entire
predator community; a strategy that works against one preda-
Allen JA, Greenwood JJD. 1988. Frequency-dependent selection by
predators. Philos Trans R Soc B Biol Sci. 319:485–503.
Bagchi S, Goval SP, Sankar K. 2003. Prey abundance and prey selection by tigers (Panthera tigris) in a semi-arid, dry deciduous forest in
western India. J Zool. 260:285–290.
Biswas S, Sankar K. 2002. Prey abundance and food habit of tigers
(Panthera tigris tigris) in Pench National Park, Madhya Pradesh,
India. J Zool. 256:411–420.
Boesch C, Boesch-Achermann H. 2000. The chimpanzees of the
taı̈ forest: behavioural ecology and evolution. Oxford: Oxford
University Press.
Bond AB. 1983. Visual search and selection of natural stimuli in the
pigeon: attention threshold hypothesis. J Exp Psychol. 9:292–396.
Bond AB. 2007. The evolution of color polymorphism: crypticity,
searching images, and apostatic selection. Ann Rev Ecol Evol Syst.
38:489–514.
Bond AB, Kamil AC. 1998. Apostatic selection by blue jays produces
balanced polymorphisms in virtual prey. Nature. 395:594–596.
Caro TM. 2005. Antipredator defenses in birds and mammals.
Chicago (IL): University of Chicago Press.
Caro TM, Graham CM, Stoner CJ, Vargas JK. 2004. Adaptive significance
of antipredator behaviour in artiodactyls. Anim Behav. 67:205–228.
Charnov EL. 1993. Life history invariants. Oxford: Oxford University
Press.
Creel S, Creel NM. 2002. The African wild dog: behaviour, ecology
and conservation. Princeton (NJ): Princeton University Press.
Cresswell W. 1994. Flocking is an effective anti-predation strategy in
redshanks, Tringa totanus. Anim Behav. 47:433–442.
Dawkins R, Krebs JR. 1979. Arms races between and within species.
Proc R Soc B Biol Sci. 205:489–511.
Deacon TW. 1990. Fallacies of progression in theories of brain-size
evolution. Int J Primatol. 11:193–236.
Eisenberg JF, Wilson DE. 1981. Relative brain size and demographic
strategies in didelphid marsupials. Am Nat. 118:1–15.
Emlen JM. 1966. The role of time and energy in food preference. Am
Nat. 100:611–617.
Emmons LH. 1987. Comparative feeding ecology of felids in a neotropical rainforest. Behav Ecol Sociobiol. 20:271.
Fitzgibbon CD. 1990a. Anti-predator strategies of immature Thomson’s gazelle: hiding and the prone response. Anim Behav. 40:
846–855.
Fitzgibbon CD. 1990b. Mixed-species grouping in Thomson’s gazelles:
the anti-predator benefits. Anim Behav. 39:1116–1126.
Freckleton RP. 2002. One the misuse of residuals in ecology: regression of residuals versus multiple regression. J Anim Ecol. 71:
542–545.
Gazzola A, Bertelli I, Avanzinelli E, Tolosano A, Bertotto P, Apollonio M.
2005. Predation by wolves (Canis lupus) on wild and domestic ungulates of the western Alps, Italy. J Zool. 266:205–213.
Götmark F, Andersson M. 2005. Predation by sparrowhawks decreases
with increased breeding density in a songbird, the great tit. Oecologica. 142:177–183.
Götmark F, Post G. 1996. Prey selection by sparrowhawks, Accipiter nisus:
relative predation risk for breeding passerine birds in relation to their
size, ecology and behaviour. Philos Trans R Soc Biol Sci. 351:
1559–1577.
Shultz and Finlayson • Prey characteristics and predator diet choice
Hairston NG, Smith FE, Slobodkin L. 1960. Community structure,
population control and competition. Am Nat. 94:421–426.
Hart JA, Katembo M, Punga K. 1996. Diet, prey selection and ecological relations of leopard and golden cat in the Ituri Forest, Zaire. Afr
J Ecol. 34:364–379.
Hebblewhite M, Pletscher DH. 2002. Effects of elk group size on predation by wolves. Can J Zool. 80:800–809.
Huggard DJ. 1993. Prey selectivity of wolves in Banff National Park. I.
Prey species. Can J Zool. 71:130–139.
Isler K, van Schaik CP. 2009. Why are there so few smart mammals
(but so many smart birds). Biol Lett. 5:125–129.
Jedrzejewski W, Jedrzejewska B, Okarma H, Ruprecht AL. 1992. Wolf
predation and snow cover as mortality factors in the ungulate community of the Bialowieza National Park, Poland. Oecologia. 90:27–36.
Jerison HJ. 1973. Evolution of the brain and intelligence. London:
Academic Press.
Johnsingh AJT. 1983. Large mammalian prey-predators in Bandipur.
J Bombay Nat Hist Soc. 80:1–57.
Johnsingh AJT. 1992. Prey selection in three sympatric large carnivores in Bandipur. Mammalia. 56:517–526.
Karanth KU, Sunquist ME. 1995. Prey selection by tiger, leopard and
dhole in tropical forests. J Anim Ecol. 64:439–450.
Korpimäki E, Norrdahl K. 1998. Experimental reduction of predators
reverses the crash phase of small-rodent cycles. Ecology. 79:2448–2455.
Krause J, Godin JG. 1995. Predator capture biases for attacking particular prey group sizes- consequences for predator hunting success
and prey predation risk. Anim Behav. 50:465–473.
Krebs CJ, Boutin S, Boonstra R, Sinclair ARE, Smith JNM, Dale MRT,
Martin K, Turkington R. 1995. Impact of food and predation on the
snowshoe hare cycle. Science. 269:1112–1115.
Krebs JR, Erichsen J, Webber M, Charnov EL. 1977. Optimal prey
selection in the great tit (Parus major). Anim Behav. 25:30–38.
Krüger SC, Lawes MJ, Maddock AH. 1999. Diet choice and capture
success of wild dog (Lycaon pictus) in Hluhluwe-Umfolozi Park,
South Africa. J Zool. 248:543–551.
Lefebvre L, Whittle P, Lascaris E, Finkelstein A. 1997. Feeding innovations and forebrain size in birds. Anim Behav. 53:549–560.
Lima SL. 1998. Stress and decision making under the risk of predation: recent developments from behavioral, reproductive, and ecological perspectives. Adv Study Behav. 27:215–290.
Lima SL. 2002. Putting predators back into behavioral predator–prey
interactions. Trends Ecol Evol. 17:70–75.
Lindstrom A. 1989. Finch flock size and risk of hawk predation at
a migratory stopover site. Auk. 106:225–232.
Manly BFJ, McDonald LL, Thomas DL. 1993. Resource selection by
animals: statistical design and analysis for field studies. London:
Chapman Hall.
Merrigi A, Brangi A, Matteucci C, Sacchi O. 1996. The feeding habits
of wolves in relation to large prey availability on Northern Italy.
Ecography. 19:287–295.
Mitani JC, Watts DP. 1999. Demographic influences on the hunting
behaviour of chimpanzees. Am J Phys Anthropol. 109:439–454.
Murdoch WW. 1969. Switching in general predators: experiments on
predator specificity and stability of prey populations. Ecol Monogr.
39:335–354.
1079
Nishida T, Hasegawa T, Hayaki H, Takahata Y, Uehara S. 1992. Meatsharing as a coalition strategy by an alpha male chimpanzee? In:
Nishida T, McGrew WC, Marler P, Pickford M, de Waal FBM, editors.
Topics in primatology, volume 1: human origins. Tokyo, Japan: University of Tokyo Press. p. 159–174.
Paine RT. 1980. Food webs: linkage, interaction strength and community infrastructure. J Anim Ecol. 49:667–685.
Pielowski Z. 1961. Uber den Unifikationse influss der selectiven Nahrungswahl des Habichts (Accipiter gentilis) auf Haustaben. Ekol Pol A.
9:183–194.
Pole A, Gordon IJ, Gorman ML, MacAskill M. 2004. Prey selection by
African wild dogs (Lycaon pictus) in southern Zimbabwe. J Zool
(Lond). 262:207–215.
Pulliam HR. 1973. On the advantages of flocking. J Theor Biol. 38:
419–422.
Pulliam HR. 1974. On the theory of optimal diets. Am Nat. 108:59–74.
Quinn JL, Cresswell W. 2004. Predator hunting behaviour and prey
vulnerability. J Anim Ecol. 73:143–154.
Schaller GB. 1972. The Serengeti lion: a study of predator-prey relations. Chicago (IL): Chicago University Press.
Scheel D. 1993. Profitability, encounter rates, and prey choice of
African lions. Behav Ecol. 4:9–97.
Sherratt TN, Harvey IF. 1993. Frequency-dependent food selection by
arthropods: a review. Biol J Linn Soc. 48:167–186.
Shultz S, Bradbury R, Evans K, Gregory R, Blackburn T. 2005. Brain
size and resource specialisation predict long-term population
trends in British birds. Proc R Soc Lond B Biol Sci. 272:
2305–2311.
Shultz S, Dunbar RIM. 2006. Chimpanzee and felid diet composition
is influenced by prey brain size. Biol Lett. 2:505–508.
Shultz S, Noe R, McGraw WS, Dunbar RIM. 2004. A community-level
evaluation of the impact of prey behavioural and ecological characteristics on predator diet composition. Proc R Soc B Biol Sci. 271:
725–732.
Sih A, Christensen B. 2001. Optimal diet theory: when does it work,
and when and why does it fail? Anim Behav. 61:379–390.
Sih A, Englund G, Wooster D. 1998. Emergent impacts of multiple
predators on prey. Trends Ecol Evol. 13:350–355.
Sol D, Duncan RP, Blackburn T, Cassey P, LeFebvre L. 2005. Big
brains, enhanced cognition, and response of birds to novel environments. Proc Natl Acad Sci U S A. 102:5460–5465.
Sol D, Székely T, Liker A, Lefebvre L. 2007. Big brained birds survive
better in nature. Proc R Soc Lond B Biol Sci. 274:763–769.
Turchin P, Karieva P. 1989. Aggregation in Aphis varians: an effective
strategy for reducing predation risk. Ecology. 70:1008–1016.
Van Orsdol KG. 1984. Foraging behaviour and hunting success of
lions in Queen Elizabeth National Park, Uganda. Afr J Ecol. 22:
79–99.
Western D. 1979. Size, life history and ecology in mammals. Afr J Ecol.
17:185–194.
Whitfield DP. 1988. Sparrowhawks Accipiter nisus affect the spacing
behaviour of wintering Turnstone Arenaria interpres and Redshank
Tringa tetanus. Ibis. 130:284–287.
Zuberbühler K, Jenny D. 2002. Leopard predation and human evolution. J Hum Evol. 43:873–886.