2010 NORTHEASTERN NATURALIST 17(1):19–28
Effect of Direct and Indirect Cues of
Predation Risk on the Foraging Behavior of the
White-footed Mouse (Peromyscus leucopus)
Benjamin G. Fanson*
Abstract - Understanding predator-prey dynamics requires an understanding of how
prey assess predation risk. This study tested the effect of microhabitat, moon stages,
and mammalian predator urines (Vulpes vulpes [Red Fox], Mustela vison [Mink], and
Procyon lotor [Raccoon]) on the degree of predation risk perceived by Peromyscus
leucopus (White-footed Mouse). Giving-up densities from artificial food patches
were used to quantify perceived predation risk. White-footed Mice exhibited a strong
preference for cover microhabitat and for the new moon stage. However, the mice
did not significantly alter their foraging behavior in response to the predator urines
compared to a water control. Additionally, mice foraged less on colder nights. The
results suggest that mammalian predator urines may not provide reliable information
on actual predation risk for the White-footed Mice and that the mice extensively use
indirect cues to assess predation risk.
Introduction
Perceived predation risk affects an animal’s decision-making process
across several axes of behavior (e.g., foraging, mating, and parental care
[Caro 2005, Lima and Dill 1990]). During periods of higher perceived
predation risk, animals may alter their behavior in many ways, such as
avoiding riskier areas or decreasing their activity level (Caro 2005, Kats and
Dill 1998). The consequences of these altered behaviors can shape population
and community dynamics (Lima 1998). Several theoretical studies have
explored the ecological effects of anti-predator behavior on community interactions
(Bouskila 2001, Brown 2001, Luttbeg and Schmitz 2000, Luttbeg
and Sih 2004). However, a crucial assumption of these models is that prey
possess the ability to assess and update their estimates of predation risk.
Therefore, a key component in understanding the ecological effects of antipredator
behavior is to understand how prey assess predation risk.
Most prey probably use indirect cues to estimate predation risk. For example,
many animals appear to alter their activity in response to temporal
(e.g., night vs. day) and spatial characteristics (e.g., forest vs. grassland,
under a shrub vs. in the open) that may be associated with different degrees
of predation risk (Brown and Kotler 2004, Caro 2005, Lima 1998). Prey
can also use information from direct cues (e.g., odor or sound of a predator)
for estimating predation risk (Kats and Dill 1998). Indirect cues probably
represent relatively static rules of thumb and provide limited information.
*Department of Biological Sciences, Purdue University, 915 West State Street, West
Lafayette, IN 47907; bfanson@gmail.com.
20 Northeastern Naturalist Vol. 17, No. 1
Direct cues have the potential to provide a variety of information including
predator type, predator state, and likelihood of a predator encounter (Kats
and Dill 1998, Laurila et al. 2004, McNamara et al. 2005).
Rodents have been shown to use indirect cues to assess predation risk.
Studies have shown that microhabitat, habitat, lunar cycle, and time of day
can affect a rodent’s perceived predation risk (Brown and Kotler 2004,
Caro 2005, Lima and Dill 1990). In general, less is known about responses
to direct cues. Detection of mammalian predators is likely to occur mostly
via olfaction (Herman and Valone 2000, Jedrzejewski et al. 1993, Ward et
al. 1997). Nolte et al. (1994) suggested that prey may use odors associated
with sulfurous metabolites from meat digestion as indicators of a predator
presence. However, previous studies have shown mixed responses of prey
to urine and feces of mammalian predators (Hayes et al. 2006, Herman and
Valone 2000, Kats and Dill 1998, Orrock et al. 2004).
In this study, I examined the effects of indirect (microhabitat and moon
stage) and direct (predator urines) cues of predation risk on the foraging
behavior of Peromyscus leucopus (Rafinesque) (White-footed Mouse). By
measuring patch use in artificial foraging patches, I first tested for differences
between microhabitats (under a shrub or in the open) across two lunar
cycles. Then, during a waning moon stage, I measured the patch use of mice
in both microhabitats when exposed to the urine of Vulpes vulpes L. (Red
Fox), Mustela vison Schreber (Mink), and Procyon lotor L. (Raccoon), and
to a water control.
Methods
Study site and study species
The study was conducted during the winter of 2004–2005 at Purdue University’s
Ross Biological Reserve near West Lafayette, IN. This hardwood
forest borders the Wabash River and contains a mix of old growth and secondary
growth patches. The forests are dominated by Quercus spp. (oaks),
Carya spp. (hickories), Acer spp. (maples), and Fagus spp. (beeches). The
undergrowth of the secondary forest is mostly Lonicera maackii (Rupr.)
Herder (Amur Honeysuckle).
White-footed Mice are abundant in the mixed forests of the eastern
United States and were common at the study site. These omnivorous rodents
are prey to various mammalian predators. Mink are a major predator of
White-footed Mice (Korschgen 1958, Mumford and Whitaker 1982), but
are rare at the Ross Reserve. Red Fox are common at this site (B.G. Fenson,
pers. observ.) and can prey heavily on White-footed Mice (Mumford and
Whitaker 1982). Raccoon are also present at this site, but are not primary
predators of White-footed Mice (although occasional predation events have
been documented; Mumford and Whitaker 1982).
Experiments
The use of foraging patches has become a common tool for measuring
perceived risk in animals (Brown and Kotler 2004). Patch-use theory
2010 B.G. Fanson 21
predicts that a forager should leave a patch when the harvest rate diminishes
to the point that it equals the foraging costs (including predation, energetic,
and missed-opportunity costs) (Brown 1988). At this point, the amount of
food left in the patch (called giving-up density; GUD) provides a surrogate
for the quitting harvest rate. Thus, by manipulating only the predation costs,
it is possible to obtain a quantitative measure of perceived predation risk by
comparing GUDs between patches (Brown 1988).
For the foraging patches, I used one-gallon plastic milk containers with
the lids on (14 x 14 x 23 cm). I cut a hole in the bottom and fitted a PVC pipe
(6 cm in diameter by 15 cm in length) into the hole to act as an entrance for
the mice. I then thoroughly mixed 10 g of small popcorn seeds into 1.5 L of
dry sand and added it to the foraging patch. The container was then laid on its
side with the PVC pipe angled to provide a ramp into the container. The milk
containers allowed ambient light in, but protected the sand from moisture.
Mice were accustomed to foraging in these kind of patches, which had been
used in previous studies in this area.
To manipulate indirect cues, I tested for differences in perceived risk
under cover and in open microhabitats during two lunar cycles (early Dec
through late Jan). I collected three days of data for both the waxing and
waning stages (moon disc ≈70% illuminated; 2 sequential days for first
lunar cycle, one for the second cycle), two days of data during the full
moon stage (moon disc >95% illuminated; one day for each lunar cycle),
and three days during the new moon stage (moon disc <5%; sequential
days between the two full moons). For the experimental setup, I haphazardly
placed six stations in secondary forests and another six in primary
forests (>50 m between nearest stations). At each station, I placed one
foraging patch in a cover location (under a pile of honeysuckle branches
and leaves) and another patch ≈1 m away in a more open location. During
the experiment, I mixed seeds into the sand <1 h before sunset and
allowed the mice to forage all night. In the morning, I capped the PVC
pipe to prevent any foraging during the day. The next evening I collected
the remaining seeds and re-charged the patch. The collected seeds were
cleaned and weighed (±0.1 g) to measure GUDs. Finally, minimum nightly
temperatures were recorded.
For the direct cue experiment, I used commercially obtained urines of
three different mammals: Red Fox, Mink, and Raccoon. I selected fox urine
because foxes are a current predator for this population and Mink urine since
Mink are a historically common predator. The Raccoon is abundant throughout
the Ross Reserve and is not a major predator; therefore, the Raccoon
urine represented a urine control. Finally, I used water as the sham control.
Since mice may perceive the risk of the direct predator cues differently
depending on the microhabitat, I also tested for a microhabitat by scent interaction.
The results from the previous experiment revealed avoidance of
patches during the waxing and full moon stages, so I chose to conduct this
experiment during the waning and new moon stages of the lunar stage in
order to get high response rates.
22 Northeastern Naturalist Vol. 17, No. 1
The experimental setup was similar to the indirect cue experiment. I used
the same foraging patch design and the same stations as the first experiment.
I used a replicated Latin square design with night (x4) and station (x4) as
blocking factors and scents as the treatment factor. A set of four stations was
used in each replicate, thus giving three replicates (12 stations). During a
waning moon stage, I applied a different scent each night to the whole station
(both cover and open microhabitats). At each foraging patch, a metal
wire with a hook was placed into the ground near the entrance (<10 cm). On
the hook, I placed a strip of absorbent cloth that had been sprayed (≈0.5 mL)
with the specified scent treatment (the urine had been warmed to ≈38 ºC to
mimic mammalian body temperature). The treatments were applied <1 h
before sunset. The following day I collected the remaining seeds, removed
the cloth strip, and wiped down the wire with isopropyl alcohol (for all treatments).
I then reran the above experiment during the following new moon
stage, resulting in each station receiving two applications of each scent.
Trapping
To confirm the identify of the foragers and the number of individuals at
a station, I live-trapped at the end of each experiment for two nights (total
of four nights). At each station, I baited three Sherman live-traps with millet
and corn seeds, giving a total of 144 trap nights. All stations had at least one
White-footed Mouse individual, with three of the 12 stations having more
than one mouse (up to three). No individual was caught at multiple stations,
and no other species were caught.
Statistical analysis
I performed two separate mixed-model analyses using PROC MIXED
(SAS 9.1; Cary, NC). For the lunar data, I created a mixed linear model
with microhabitat and moon stage as the fixed factors. For the scent data, I
created a mixed model with microhabitat and scent type as the fixed factors.
Additionally, since the ambient temperature can affect the energetic costs
of foraging (Brown 1988, Kotler et al. 1993), I added in nightly minimum
temperatures as a covariate to both models. To test for the potential problem
of carryover effects with the scent data, I initially included an additional factor
that included the previous treatment type (see Littell et al. 2006). After
running the model, this effect was found to be nonsignificant (P = 0.54) and
was removed.
All stations most likely had the same individual(s) foraging in both
patches at a station. Additionally, patches foraged temporally closer should
be more correlated than patches further apart in time. To deal with these covariations,
I included station and night as random factors and then modelled
the residual correlation matrix using the spatial power law (“sp[pow]” in
MIXED). This method is similar to a first-order autoregressive correlation,
but is applicable for unequally spaced data (Littell et al. 2006). I then applied
a Kenward-Roger correction to the denominator degrees of freedom (Littell
et al. 2006).
2010 B.G. Fanson 23
I visually inspected all residuals for normality and homoscedasticity
assumptions, and no transformations were needed for either model. For
any significant factors, I performed post-hoc comparisons of the least
square means, and to control for Type I error, I applied a Tukey-Kramer
adjustment to the P-values. All means are displayed with ± SE, unless
otherwise indicated.
Results
Microhabitat and moon stage
Microhabitat and moon stage appeared to have a strong effect on GUDs
(F2,85.9 = 6.66, P = 0.012; F2,5.99 = 6.42, P = 0.027, respectively); however,
there appears to be no strong interaction between the two variables (F3,115 =
1.77, P = 0.16; Fig. 1). Mice foraged patches in the cover microhabitat to
lower GUDs than in the open microhabitat (5.39 ± 0.74 vs. 6.17 ± 0.74 g),
and there was no strong evidence that this preference or the magnitude of the
difference changed with different moon stages; however, moon stage did affect
the overall pattern of foraging. Mice foraged 100% of the patches during
the new moon stage, 68% of the patches during the waning stage, 50% during
waxing stage, and 29% during the full moon. The new moon stage and waning
stage had the lowest mean GUDs and did not differ significantly from
Figure 1. Effect of moon stage on the giving-up densities of White-footed Mice. Error
bars represent 1 standard error. Different letters indicate a significant difference
in means (P < 0.05).
24 Northeastern Naturalist Vol. 17, No. 1
each other (Δ⎯x = 2.05 ± 1.25 g; t6.91 = 1.64, P = 0.428). However, new moon
did differ significantly from waxing (Δ⎯x = 4.18 ± 1.24 g; t6.57 = 3.40, P = 0.05)
and full moon (Δ⎯x = 6.43 ± 1.70 g; t5.96 = 3.78, P = 0.04) stages. However, after
Tukey adjustments, waning stage did not differ significantly from waxing
(Δ⎯x = 2.14 ± 1.19 g; t5.67 = 1.79, P = 0.36) or full (Δ⎯x = 4.39 ± 1.67 g; t5.45 =
2.64, P = 0.13) stage. Finally, nightly temperature had a significant negative
relationship with GUDs (β = -0.22 ± 0.04 g/ºC; F1,54.9 = 33.15, P < 0.001).
Microhabitat and scents
Overall, 91% of the 168 patches were foraged over the seven nights.
Only two of the 84 stations had no foraging activity. GUD results showed
no strong evidence that the predator urine affected their foraging behavior
(F3,25.4 = 2.05, P = 0.13). Interestingly, the water treatment had the highest
mean GUDs (5.11 ± 0.60 g), followed by Raccoon (4.9 ± 0.60 g), Mink (4.2
± 0.60 g), and fox (3.82 ± 0.60 g) treatment. However, water and fox GUDs
were not significantly different (Δ⎯x = 1.30 ± 0.58 g; t26.3 = 2.25, P = 0.14).
The water treatment mean was very similar to the waning stage mean in the
first experiment (Δ⎯x = 0.44 ± 0.71 g; t14 = 0.62, P = 0.54). The mice did have
lower GUDs in the cover compared to the open microhabitat (3.56 ± 0.52
vs. 5.46 ± 0.52 g, respectively; F1,10.6 = 22.48, P = 0.0007). I found no strong
evidence that microhabitat affected the scent preference of the mice (F3,117 =
0.23, P = 0.88). Similar to the previous experiment, temperature also had a
negative effect on GUDs (β = -0.42 ± 0.09 g/ºC; F1,4.65 = 22.00, P = 0.007).
Discussion
The objective of this study was to examine a forager’s use of indirect and
direct cues of predation risk. Moon stage and microhabitat had strong effects
on the foraging behavior of White-footed Mice. The mice foraged patches
more thoroughly during the new moon stages and when foraging under
vegetation. No strong support existed that mice foraged the microhabitats
differently during different moon stages. In contrast to the strong effects of
the indirect cues, the direct due of predator scent had little effect on foraging
behavior during the waning and new moon phase.
Minimum nightly temperature positively affected the foraging behavior
of the mice. A similar relationship between temperature and GUDs has also
been found with gerbils (Kotler et al. 1993). One likely explanation is that
higher energetic costs of thermoregulation lead to higher foraging costs.
Nightly temperatures ranged from -14 ºC to 10 ºC, resulting in metabolic
rates from ≈12 to ≈6 cm3O2(g*h)-1, respectively (Wickler 1980). Patch-use
theory predicts that as energetic costs increase, an optimal forager should
quit a patch at a higher harvest rate, assuming a constant marginal rate of
substitution (that is, no change in the value of energy for the animal) (Brown
1988). Another possibility for this trend is the potential confounding effect
of cloud cover. I avoided collecting data on completely cloudy nights,
but some nights had a notable amount of cloud cover (up to 80%). Nightly
2010 B.G. Fanson 25
temperature may be positively correlated with cloud cover; therefore, mice
would have experienced reduced energetic costs on warmer nights as well as
reduced predation risk associated with less moon illumination. However, using
meteorological records, I isolated warm clear nights and the residuals for
those days were close to the regression line between GUDs and temperature,
suggesting that this explanation is not likely.
Microhabitat has been frequently shown to affect the foraging behavior
of rodents (Brown 1988, Kotler et al. 1993, Orrock et al. 2004). The common
explanation is that there is differential predation risk between the microhabitats
(Brown 1988, Kotler et al. 1993, Morris 1997). The preference for
cover microhabitat is very common when the main predators of a rodent are
avian and mammalian. In contrast, snakes often hunt in cover microhabitats
and can cause rodents to prefer open microhabitat (Bouskila 1995). At this
study site, both foxes and owls are very common, and no snakes were present
during this wintertime study. Thus, differences in predation risk most likely
explain these preferences.
Similarly, for nocturnal rodents, lunar illumination can be associated
with higher predation risk, increasing the detection abilities of highly visual
predators, such as owls and mammals (Clarke 1983, Kotler et al. 1991). In
this study, White-footed Mice had lower GUDs during the new moon than
during the waxing and full moon stages, probably due to lower predation
risk during new moon stage. Several studies using differing techniques have
found similar behavioral responses between full vs. new moon stages for
White-footed Mice (Bowers and Dooley 1993, Clark and Kaufman 1991,
Jekanoski and Kaufman 1995). However, two studies found no effect of illumination
levels on behavior. Barry and Francq (1982) found evidence that
White-footed Mice did not avoid higher illumination areas. Schmidt (2006)
found White-footed Mice ate similar number of seeds during new and full
moon stages; however, the mice did forage fewer seeds during the full moon
than the new moon if an owl call was played during the night.
For the direct cues, predator urine did not have a significant effect on foraging
behavior. A potential concern of any experiment is that the results are
an artifact of the design. This experiment implemented a Latin square design
that could potentially have carryover effects (that is, previous treatment has
subsequent effects on next treatment). However, two pieces of evidence suggest
little or no carryover effects. First, the water control mean GUD (5.1 g)
in the scent experiment was very similar to the mean GUD during the waning
moon stage (4.7 g) in the previous experiment, suggesting that adding a scent
did not result in all patches being foraged less. Second, I statistically tested
for the effect of the previous treatment on the next night’s GUDs and the
result was not significant. Thus, I believe that carryover effects are probably
not a major concern in this study.
Another potential concern is the power of the design. I conducted a
retrospective power analysis using observed variances and actual sample
sizes (Thomas 1997). For pairwise comparisons with a Tukey-adjusted α, an
26 Northeastern Naturalist Vol. 17, No. 1
effect size of 1.0 g and 2.1 g had 20% and 80% power, respectively. To put
this into perspective, microhabitat had an effect size of 1.9 g, and between
full and new moon, the effect size was 6.6 g. Consequently, I feel that the
design had sufficient power to detect modest to large effects of scents.
The scent results do match a similar study with Peromyscus polionotus
(Wagner) (Oldfield Mouse), in which the mice did not show any response
to mammalian predator odors (Orrock et al. 2004). In that study, they used
Lynx rufus (Schreber) (Bobcat), Vulpes sp. (fox), Canis latrans Say (Coyote),
and Felis pardalis L. (Ocelot) urines for their predators. However, a
potential caveat to the results in both studies is that GUDs were measured
only during waning and new moon stages. Schmidt (2006) found that the
White-footed Mice only altered its foraging behavior in response to owl
vocalizations during the full moon and not the new moon. Therefore, there
may be non-additive effects between indirect and direct cues. In other
words, the effect of direct cues may vary depending on the lunar phase.
This study looked only for such non-additive effects between microhabitat
and scents, which were not found. Thus, it is possible that response to
predator scents may change across moon stages.
The results from this study agree with a recent meta-analysis that
analyzed the effect of indirect and direct cues of predation risk on GUDs
(Verdolin 2006). The meta-analysis revealed that odor (and actual presence
of a predator) had no significant effect on GUDs, but habitat structure had a
large, consistent effect on GUDs. The inconsistency in the effects of predator
presences and odors suggest that probably no simple rule exists for how
prey exploit direct cues of a predator. Using a theoretical model, McNamara
et al. (2005) showed that a bird should change its foraging behavior
depending on the nature of the information provided by an exposure to a
predator. For instance, an encounter with the scent of a predator may indicate
that it is still around and more likely to be encountered (e.g., Accipiter
nisus (L.) [Eurasian Sparrowhawk]; Cresswell 1996), or it may provide no
information on the probability of re-encountering the predator if the predator
moves through the environment haphazardly (e.g., Accipiter striatus
Vieillot [Sharp-shinned Hawk]; Roth and Lima 2007). Thus, studying the
foraging and movement behavior of an animal’s predator should help predict
the usefulness of direct cues to its prey.
Acknowledgments
I wish to thank Jeff Lucas, Adam Boyko, and Mark Nolen for comments and
suggestions; Loren Lohmann for assistance in collecting the data; Ani for consistent
insights on predator behavior; and Kerry Fanson for editorial comments on
the manuscript.
Literature Cited
Barry, R.E., and E.N. Francq. 1982. Illumination preference and visual orientation of
wild-reared mice, Peromyscus leucopus. Animal Behaviour 30:339–344.
2010 B.G. Fanson 27
Bouskila, A. 1995. Interactions between predation risk and competition: A field study
of Kangaroo Rats and snakes. Ecology 76:165–178
Bouskila, A. 2001. A habitat selection game of interactions between rodents and their
predators. Annales Zoologici Fennici 38:55–70.
Bowers, M.A., and J.L. Dooley. 1993. Predation hazard adn seed removal by small
mammals: Microhabitat- versus patch-scale effects. Oecologia 94:24–254.
Brown, J.S. 1988. Patch use as an indicator of habitat preference, predation risk, and
competition. Behavioral Ecology and Sociobiology 22:37–47.
Brown, J.S. 2001. Ecology of fear: Foraging games between predators and prey with
pulsed resources. Annales Zoologici Fennici 38:71–87.
Brown, J.S., and B.P. Kotler. 2004. Hazardous duty pay and the foraging cost of
predation. Ecology Letters 7:999–1014.
Caro, T. 2005. Antipredator Defenses in Birds and Mammals. The University of
Chicago Press, Chicago, IL.
Clark, B.K., and D.W. Kaufman. 1991. Effects of plant litter on foraging and nestingbehavior
of prairie rodents. Journal of Mammalogy 72:502–512.
Clarke, J.A. 1983. Moonlight's influence on predator prey interactions between
Short-eared Owls (Asio flammeus) and Deer Mice (Peromyscus maniculatus).
Behavioral Ecology and Sociobiology 13:205–209.
Cresswell, W. 1996. Surprise as a winter hunting strategy in Sparrowhawks Accipiter
nisus, Peregrines Falco peregrinus and Merlins F. columbarius. Ibis
138:684–692.
Hayes, R.A., H.F. Nahrung, and J.C. Wilson. 2006. The response of native Australian
rodents to predator odours varies seasonally: A by-product of life-history variation?
Animal Behaviour 71:1307–1314.
Herman, C.S., and T.J. Valone. 2000. The effect of mammalian predator scent on the
foraging behavior of Dipodomys merriami. Oikos 91:139–145.
Jedrzejewski, W., L. Rychlik, and B. Jedrzejewski. 1993. Responses of Bank Voles to
odors of 7 species of predators: Experimental-data and their relevance to natural
predator-vole relationships. Oikos 68:251–257.
Jekanoski, R.D., and D.W. Kaufman. 1995. Use of simulated herbaceous canopy by
foraging rodents. American Midland Naturalist 133:304–311.
Kats, L.B., and L.M. Dill. 1998. The scent of death: Chemosensory assessment of
predation risk by prey animals. Ecoscience 5:569–569.
Korschgen, L.J. 1958. December food habits of Mink in Missouri. Journal of Mammalogy
39:521–527.
Kotler, B.P., J.S. Brown, and O. Hasson. 1991. Factors affecting gerbil foraging
behavior and rates of owl predation. Ecology 72:2249–2260.
Kotler, B.P., J.S. Brown, and W.A. Mitchell. 1993. Environmental factors affecting
patch use in 2 species of gerbilline rodents. Journal of Mammalogy
74:614–620.
Laurila, A., J. Kujasalo, and E. Ranta. 2004. Different antipredator behaviour in two
anuran tadpoles: Effects of predator diet. Behavioral Ecology and Sociobiology
40:329–336.
Lima, S.L. 1998. Nonlethal effects in the ecology of predator-prey interactions:
What are the ecological effects of anti-predator decision-making? Bioscience
48:25–34.
Lima, S.L., and L.M. Dill. 1990. Behavioral decisions made under the risk of predation:
A review and prospectus. Canadian Journal of Zoology 68:619–640.
28 Northeastern Naturalist Vol. 17, No. 1
Littell, R.C., G.A. Milliken, W.W. Stroup, R.D. Wofinger, and O. Schabenberger.
2006. SAS for mixed models. SAS Institute, Inc., Cary, N.C.
Luttbeg, B., and O.J. Schmitz. 2000. Predator and prey models with flexible individual
behavior and imperfect information. American Naturalist 155:669–683.
Luttbeg, B., and A. Sih. 2004. Predator and prey habitat selection games: The effects
of how prey balance foraging and predation risk. Israel Journal of Zoology
50:233–254.
McNamara, J.M., Z. Barta, A.I. Houston, and P. Race. 2005. A theoretical investigation
of the effect of predators on foraging behaviour and energy reserves. Proceedings
of the Royal Society B-Biological Sciences 272:929–934.
Morris, D.W. 1997. Optimally foraging Deer Mice in prairie mosaics: A test of habitat
theory and absence of landscape effects. Oikos 80:31–42.
Mumford, R.E., and J.O. Whitaker. 1982. Mammals of Indiana. Indiana University
Press, Bloomington, IN.
Nolte, D.L., J.R. Mason, G. Epple, E. Aronov, and D.L. Campbell. 1994. Why are
predator urines aversive to prey. Journal of Chemical Ecology 20:1505–1516.
Orrock, J.L., B.J. Danielson, and R.J. Brinkerhoff. 2004. Rodent foraging is affected
by indirect, but not by direct, cues of predation risk. Behavioral Ecology
15:433–437.
Roth, T.C., and S.L. Lima. 2007. Use of prey hotspots by an avian predator: Purposeful
unpredictability? American Naturalist 169:264–273.
Schmidt, K.A. 2006. Non-additivity among multiple cues of predation risk: A behaviorally
driven trophic cascade between owls and songbirds. Oikos 113:82–90.
Thomas, L. 1997. Retrospective power analysis. Conservation Biology 11:276–280.
Verdolin, J.L. 2006. Meta-analysis of foraging and predation risk trade-offs in terrestrial
systems. Behavioral Ecology and Sociobiology 60:457–464.
Ward, J.F., D.W. MacDonald, and C.P. Doncaster. 1997. Responses of foraging
Hedgehogs to Badger odour. Animal Behaviour 53:709–720.
Wickler, S.J. 1980. Maximal thermogenic capacity and body temperature of
White-footed Mice (Peromyscus) in summer and winter. Physiological Zoology
53:338–346.