Vol.:(0123456789) 1 3
Oecologia (2022) 198:91–98
https://doi.org/10.1007/s00442-021-05080-w
BEHAVIORAL ECOLOGY –ORIGINAL RESEARCH
Prey tells, large herbivores fear the human ‘super predator’
Daniel A. Crawford1,2 · L. Mike Conner2 · Michael Clinchy3 · Liana Y. Zanette3 · Michael J. Cherry1
Received: 9 July 2021 / Accepted: 13 November 2021 / Published online: 4 January 2022
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2021
Abstract
Fear of the human ‘super predator’ has been demonstrated to so alter the feeding behavior of large carnivores as to cause
trophic cascades. It has yet to be experimentally tested if fear of humans has comparably large efects on the feeding behavior
of large herbivores. We conducted a predator playback experiment exposing white-tailed deer to the vocalizations of humans,
extant or locally extirpated non-human predators (coyotes, cougars, dogs, wolves), or non-predator controls (birds), at supplemental food patches to measure the relative impacts on deer feeding behavior. Deer were more than twice as likely to fee
upon hearing humans than other predators, and hearing humans was matched only by hearing wolves in reducing overall
feeding time gaged by visits to the food patch in the following hour. Combined with previous, site-specifc research linking
deer fecundity to predator abundance, this study reveals that fear of humans has the potential to induce a larger efect on
ungulate reproduction than has ever been reported. By demonstrating that deer most fear the human ‘super predator’, our
results point to the fear humans induce in large ungulates having population- and community-level impacts comparable to
those caused by the fear humans induce in large carnivores.
Keywords Ecology of fear · Behavioral response · Odocoileus virginianus · Perceived predation risk · Playback experiment
Introduction
Humans kill large and medium-sized carnivores at rates far
exceeding those of non-human predators, meriting their label
as an ecological ‘super predator’ (Darimont et al. 2015).
Recent experiments have demonstrated that the unique and
disproportionate predatory pressures humans exact on carnivores are refected in carnivores fearing humans more than
other carnivores (Clinchy et al. 2016; Suraci et al. 2019b),
and correspondingly that human-induced fear (antipredator
responses) can cause cascading efects across trophic levels
at a landscape scale (Smith et al. 2017; Suraci et al. 2019a;
Zanette and Clinchy 2020). Fear efects on large herbivore
foraging behavior afecting community dynamics have similarly been experimentally demonstrated to result from the
fear large carnivores induce in large ungulates (Zanette and
Clinchy 2020), and accumulating evidence indicates that
foraging costs associated with antipredator behaviors may
reduce reproductive rates in large ungulates (Say-Sallaz
et al. 2019; Zanette and Clinchy 2020). A recent global
meta-analysis of mortality in terrestrial vertebrates suggests that humans not only kill carnivores, but large ungulates as well, at greater rates than non-human predators do
(Hill et al. 2019). Thus, large herbivores may be expected
to fear humans more than they fear non-human predators
(i.e., carnivores), in the same way that carnivores have been
shown to fear humans more than they fear other carnivores.
Correspondingly, the magnitude of population- and community-level efects caused by the fear humans induce in large
herbivores may be anticipated to exceed the demonstrated
magnitude of such efects caused by the fear large herbivores
have of large carnivores. To explore these potential population- and community-level impacts, it is frst necessary to
Communicated by Indrikis Krams.
This research was a collaboration among the labs of the Principal
Investigators Cherry, Conner, Zanette.
* Daniel A. Crawford
[email protected]
1 Department of Rangeland and Wildlife Sciences, Caesar
Kleberg Wildlife Research Institute, Texas A&M University
– Kingsville, 700 University Blvd, MSC 218, Kingsville,
TX 78363, USA
2 The Jones Center at Ichauway, 3988 Jones Center Drive,
Newton, GA 39870, USA
3 Department of Biology, Western University, London,
ON N6A 5B7, Canada
92 Oecologia (2022) 198:91–98
1 3
test the relative impacts that fear of humans and non-human
predators have on large herbivore foraging behavior (Clinchy
et al. 2016; Smith et al. 2017; Suraci et al. 2019a, b).
Recent research on white-tailed deer (Odocoileus virginianus) conducted in southwestern Georgia, USA, demonstrated that experimentally removing coyotes (Canis
latrans), the sole non-human predator of adult deer in the
system studied, caused deer to be less fearful, spending less
time vigilant and more time feeding (Cherry et al. 2015).
Correlative evidence in this same system showed that deer
fecundity was signifcantly greater in years when coyote
numbers were low (Cherry et al. 2016a), and the experimentally demonstrated efect fear of coyotes has on deer feeding
time provides a straightforward explanation for the greater
deer fecundity in low coyote years. Other recent studies provide compelling evidence that fear of wolves (Canis lupus)
afects reproduction in elk (Christianson and Creel 2014),
and the fear of cougars (Puma concolor) afects reproduction in mountain goats (Dulude-de Broin et al. 2020) and
bighorn sheep (Bourbeau-Lemieux et al. 2011). Deer, elk,
and bighorn sheep are all hunted in North America, and
data from Darimont et al. (2015) indicate that these species
are killed by humans at a signifcantly greater rate than they
are by large carnivores (Supplementary online material) and
can all thus be expected to fear humans more than they fear
large carnivores.
Prior research in our study area experimentally demonstrated that fear of coyotes afects deer feeding time (Cherry
et al. 2015) and later linked long-term fuctuations in coyote
abundance to deer fecundity (Cherry et al. 2016a, b). Given
these site-specifc fndings, we experimentally tested if fear
of humans has an even greater efect on deer feeding time in
this system, which would infer a greater efect on deer fecundity. As recent colonizers of eastern North America, coyotes
partially fll niches vacated by extirpated large carnivore
species, namely wolves and cougars, and often function as
the dominant non-human predator of deer in the absence of
larger predator species (Cherry et al. 2016b). While coyotes
can and do successfully prey on adult deer (Chitwood et al.
2014), human predation rates of adult deer likely exceed
those of coyotes in southwestern Georgia (Cherry et al.
2016b) or eastern North America. Data suggest that, even
if wolf and cougar populations did persist throughout their
historical ranges, humans would remain the leading predator
of deer in the system (Darimont et al. 2015). Domestic dogs
(Canis lupus familiaris) are the other large carnivore extant
in this system, but these are not used locally to hunt deer,
though feral dogs may kill some deer (Vanak and Gompper
2009).
To test the relative efects fear of humans and fear of
coyotes and large carnivores have on deer feeding behavior,
we experimentally exposed free-living wild deer to auditory
playbacks of the vocalizations of humans, coyotes, wolves,
cougars, dogs and non-threatening controls (birds) at baited
food patches, and recorded the behavioral responses elicited
by each playback. Our experiment followed well-established
protocols comparable to those used in demonstrating the fear
humans induce in carnivores (Clinchy et al. 2016; Smith
et al. 2017; Suraci et al. 2019a, b) including the quantifcation of three metrics pertaining to foraging behavior: (i)
abandonment of the food patch (feeing) upon hearing a
treatment; (ii) reduced time spent feeding upon hearing a
treatment among those deer that did not abandon the food
patch; and (iii) the aggregate efect hearing a treatment had
on the number of times deer were recorded at the food patch
in the hour after frst hearing the treatment. Combined with
our previous work indicating that fear of large carnivores
afects deer reproduction (Cherry et al. 2015, 2016a), our
results provide strong support for fear of humans having
greater population-level impacts on large herbivores. Moreover, the close correspondence between our results and those
demonstrating the fear carnivores have of humans (Clinchy
et al. 2016; Suraci et al. 2019b) suggests the fear humans
induce in large herbivores may have comparable communitylevel impacts (Smith et al. 2017; Suraci et al. 2019a). We
discuss these implications of our results considering recent
meta-analyses of worldwide data pointing to fear of humans
having pervasive efects on the behavior of mammals of
every size and type (Gaynor et al. 2018; Tucker et al. 2018).
Methods
We conducted our experiment at the Jones Center at Ichauway, a 12,000-ha ecological research site located in the lower
coastal plain of southwestern Georgia, USA [described further in Cherry et al. (2015, 2016a)], during May–Jun 2018.
White-tailed deer are the dominant large herbivore and only
native ungulate on-site. Approximately 30% of the estimated
population is harvested by humans annually. To measure
feeding behavior and to ensure that deer would feed in front
of our experimental playback sites, we provided shelled
corn at each of 23 sites. We selected playback sites by generating spatially balanced (separated by > = 1 km) random
points along the study site’s unpaved road network in areas
of the property that were treated with prescribed fre during
Jan–May 2018. At each random point, we identifed the tree
nearest to 50 m along a bearing perpendicular to the road
at that location, and pre-baited the location with ~ 13 L of
shelled corn for 3 days prior to initiation of experimental trials. Once we initiated trials, we visited sites daily to replenish the shelled corn, retrieve data, and maintain electronic
equipment.
We recorded deer responses to playbacks using Automated Behavioral Response (ABR) systems, consisting
of a video-enabled camera trap linked to a playback unit
Oecologia (2022) 198:91–98 93
1 3
triggered by the camera’s activation (Suraci et al. 2017a).
We deployed the playback speaker and camera 5 m from the
center of the provisioned corn and programmed the ABRs to
broadcast playbacks 10 s into each 20 s video for a playback
duration of 10 s. We tested the efects on deer feeding behavior of hearing either humans, coyotes, wolves, cougars, dogs,
or controls (birds). To comprise an optimal, non-threatening
control composed of familiar, benign hetero-specifc animal
vocalizations (Hettena et al. 2014), we used the vocalizations
of three locally abundant species of birds, the Carolina Wren
(Thryothorus ludovicianus), Chuck Will’s Widow (Antrostomus carolinensis), and Barred Owl (Strix varia), broadcast
during diel, crepuscular, and nocturnal hours, respectively.
We designed avian vocalizations to constitute a single treatment (bird sounds) and treated them as such in our analyses
(Zanette et al. 2011; Epperly et al. 2021). We used eight
exemplars of each species with human exemplars each consisting of a single individual speaking conversationally (i.e.,
in a neutral fashion not conveying alarm or threat; following
Clinchy et al. 2016; Smith et al. 2017; Suraci et al. 2019a,
b). We edited all exemplars for consistency in amplitude
and quality using Audacity® (Team 2014) and broadcast the
playbacks at a consistent mean sound pressure level of 70 dB
at 1 m to ensure responses to our stimuli were unrelated to
variability in sound intensity across or within treatments,
and that the sound was loud enough to be audible, but not
startling, for animals within the 15 m detection range of
the camera’s motion sensor (Smith et al. 2017; Suraci et al.
2019b; Zanette and Clinchy 2020; Epperly et al. 2021). We
balanced and randomized treatments across the diel cycle by
programming each ABR at each site, such that the treatment
broadcast if triggered changed every 15 min. Further, we
organized the sequence of treatments for each ABR at each
site to avoid order efects and randomly selected exemplars
within each 15 min to avoid pseudo-replication (following
Epperly et al. 2021; playback schedules used are provided in
Supplementary online material). Because the data collected
by each ABR comprise a stand-alone playback experiment
(complete with treatment and control playbacks), the use of
ABRs at multiple sites represents multiple replicates of the
same experiment (Suraci et al. 2017a).
We categorized videos as independent treatment-specifc
exposures if>60 min elapsed since the last time a deer heard
the same treatment at that site, which is more conservative
than the > 30 min elapsed time between samples used in
most camera trap studies (Suraci et al. 2017b; Epperly et al.
2021). We quantifed three impacts of playback treatments
on foraging behavior. Upon a deer frst hearing a treatment
in an independent exposure video, we recorded (i) whether
the deer fed (i.e., abandoned the food patch) upon hearing the playback. Regardless of group size, if any individual fed, the response was deemed a fight response. If the
subject(s) did not fee, we determined (ii) whether the time
spent feeding during the 10 s playback broadcast was less
than during the 10 s prior to the broadcast. Both measures
were treated as binary responses, i.e., did the deer run or
not, and if it did not, did time spent feeding decrease or not.
If group size was > 1, mean time spent feeding before and
during the playback were used to determine if time spent
feeding decreased. Finally, we quantifed (iii) the aggregate
efect on feeding by tallying the number of videos in which
deer were recorded at the food patch in the hour after frst
hearing a treatment (Smith et al. 2017). While an animal
might fee at any sudden sound, or reduce feeding if it does
not fee, it would then be expected to return to the food patch
to continue feeding at a rate inversely related to how frightening it perceived the sound to be, i.e., the more frightening
the less likely it will return, and if it does or did not fee, the
less likely it is to spend time feeding.
We used logistic generalized linear mixed-efects models
(GLMM), with playback treatment as a fxed efect and site
as a random efect. We evaluated treatment-specifc efects
in independent exposure videos by estimating the probability of deer (i) feeing or staying and, if the deer stayed, (ii)
decreased foraging post-playback relative to pre-playback
broadcast. We used a Poisson GLMM, with playback treatment as a fxed efect and site as a random efect, to estimate
treatment-specifc efects on (iii) the number of videos in
which we detected deer at the food patch in the hour following each independent exposure video. We performed all
GLMMs using the “lme4” package (Bates et al. 2015) in
Program R v 3.6.3 (R Core Team 2020).
Results
We recorded 822 independent exposure videos across
23 sites, with the number of videos being well balanced
among the six treatments (control, n = 159; coyote, 128;
cougar, 138; dog, 126; wolf, 123; human, 148). Upon hearing humans, deer were almost twice as likely to fee (i.e.,
abandon the food patch) as upon frst hearing any species of
large carnivore, the contrast between the reaction to humans
vs. any given large carnivore being signifcant in every
case (Fig. 1a; all p < 0.001). Deer were also signifcantly
more likely to fee from any large carnivore compared to
the control playbacks (all p < 0.001). Among the carnivore
playbacks, deer were equally likely to run whether the large
carnivore was extant (coyote), extinct (cougar, wolf), or a
human commensal (dog). Deer that did not fee were more
than twice as likely to reduce the time they spent feeding
(i) if they heard humans or wolves, compared to hearing
control playbacks (p<0.001), whereas they were not signifcantly more likely to do so if they heard coyotes, cougars,
or dogs (Fig. 1b). Hearing humans had the strongest impact
on (ii) the number of times deer were recorded at the food
94 Oecologia (2022) 198:91–98
1 3
patch in the following hour (Fig. 2b; see Appendices for
additional model results), and hearing any species of large
carnivore caused a signifcant reduction in the number of
times deer were recorded at the food patch in the following
hour in comparison to hearing control playbacks (Fig. 2a;
all p < 0.001).
Discussion
Our results experimentally demonstrate that the fear the
human’ super predator’ inspires in large herbivores can
exceed their fear of non-human predators, leading to significant relative reductions in time spent feeding, which can
be expected to cause fear of humans to have greater population- and community-level consequences than fear of
non-human predators. Fear of humans signifcantly—and
most strongly—impacted (i) abandonment of food patches
and (ii) time spent feeding if deer did not fee, and consistently had signifcantly stronger efects on all three metrics
than did fear of coyotes, the largest wild carnivore in this
system. Hearing coyotes, however, did lead to signifcant
efects on the probability of deer (i) abandoning the food
patch (feeing), and (ii) the aggregate number of times deer
were recorded at the food patch, corroborating our previous experimental demonstration that fear of coyotes afects
deer feeding behavior (Cherry et al. 2015) and, potentially,
fecundity (Cherry et al. 2016a). Human playbacks also
had signifcantly stronger efects on deer feeding than did
dog playbacks, suggesting that deer discriminate between
humans and their large carnivore commensals (dogs) as
demonstrated experimentally in wild carnivores (Clinchy
et al. 2016; Suraci et al. 2019b). Interestingly, the nonhuman predator feared most by white-tailed deer in our
Fig. 1 a Predicted probabilities
of white-tailed deer feeing
upon frst exposure to playback
treatments at 23 experimental
sites during May–June 2018
at the Jones Center at Ichauway, Newton, Georgia, USA.
Control playbacks consisted of
extant avian species (Carolina
Wren, Chuck Will’s Widow,
and Barred Owl) while treatment playbacks included
extant predators (coyote, dog,
and human) as well as locally
extinct predators (cougar, wolf).
Additionally, we predicted the b
the probability of reduced time
feeding after playback treatments relative to before playbacks initiated for white-tailed
deer that did not fee upon frst
exposure to playback treatments
Oecologia (2022) 198:91–98 95
1 3
experiment was not the extant coyote but the extirpated
wolf, precisely as demonstrated in a playback experiment on closely related mule deer (Odocoileus hemionus;
Hettena et al. 2014).
Previous work linking deer fecundity to fear of coyotes
(Cherry et al. 2016a) on our study area revealed one of the
largest efect sizes reported to date regarding fear of large
carnivores afecting large herbivore reproduction (efect
size = – 0.47), according to a recent review on the population- and community-level efects of fear in free-living wildlife (Zanette and Clinchy 2020). If fear efects on fecundity
are assumed to scale directly with the efects that fear has
on feeding (sensu Brown et al. 1999), then given the 27%
reduction in (iii) the number of times deer were recorded at
the food patch caused by fear of humans relative to fear of
coyotes (Fig. 2b), the magnitude of the efect on deer fecundity caused by their fear of humans would be projected to be
– 0.60; larger than any efect on large herbivore reproduction yet attributed to fear of large carnivores (range = – 0.18
to – 0.55; Zanette and Clinchy 2020). We are of course
not proposing that diferences in feeding rate translate so
directly into diferences in fecundity. The actual efect fear
of humans has on ungulate fecundity remains to be experimentally tested, and our purpose here is simply to suggest
the relative magnitude of efect that might be expected to be
found in such an experiment.
Experimental testing, rather than a correlational approach,
is essential to determine the magnitude of the efect fear of
humans has on deer fecundity given the likely complexity
of the efects that humans have on deer. The experiments to
date demonstrating that carnivores fear humans more than
other carnivores have included species, such as badgers
(Meles meles), skunks (Mephitis mephitis) and opossums
(Didelphis virginiana), which despite being evidently very
fearful of humans nonetheless occur in greater abundance
around humans because of the greater availability of food
near humans, from agricultural sources or human food waste
(Macdonald and Newman 2002; Clinchy et al. 2016; Wang
et al. 2015; Suraci et al. 2019a). White-tailed deer abundance in eastern North America has increased dramatically
since European settlement because of the increase in food
from agricultural sources, greater browse availability due to
Fig. 2 Model coefcients and
their associated 95% confdence intervals estimated with
a Poisson generalized linear
mixed-efects model of count
data representing the number
of independent deer detections
occurring in the hour following treatment playbacks at
23 randomly selected sites at
the Jones Center at Ichauway,
Newton, Georgia, USA. a
Control playbacks consisted of
extant avian species (Carolina
Wren, Chuck Will’s Widow, and
Barred Owl). b Efects elicited
by extant predators (coyote,
dog, and human) and locally
extinct predators (cougar, wolf)
are presented on a magnifed
scale for ease of comparison.
The dashed line indicates zero
efect size, and confdence intervals overlapping zero indicate
no efect
96 Oecologia (2022) 198:91–98
1 3
forest clearing, the extirpation of large carnivores, and other
human-caused changes in the environment (Rushing et al.
2020). Our results corroborate the fndings from the large
number of fight initiation distance studies done on deer and
other ungulates in North America (reviewed in Stankowich
2008) by experimentally demonstrating that deer greatly fear
humans. Yet the deer in our study still came back for the corn
we provided as bait. Fear efects on population and community dynamics are expected precisely because of the tradeof animals must make between time spent on feeding and
time spent avoiding predators (Zanette and Clinchy 2020).
Humans have a unique ecology in being uniquely dangerous
(Darimont et al. 2015), which as our experiment and others
demonstrate necessitate animals spending more time avoiding humans, but at the same time, our unique human ecology
includes increasing the quantity and quality of food available
to many species such that this likely reduces the time they
need to spend feeding (Moll et al. 2021). This complexity is
why experimentally manipulating one factor at a time (fear
in this case) appears the most expeditious means of determining the relative magnitude of the impacts we humans are
having on deer and other wildlife.
A growing number of recent experiments have demonstrated that fear of large carnivores can cause trophic cascades by afecting the feeding behavior of large herbivores
(Zanette and Clinchy 2020). In our experiment, the highmagnitude fear responses to human playbacks relative to
other predators suggests that fear of humans may be more
likely to cause trophic cascades than fear of non-human
predators. The potential for fear of humans to cause trophic
cascades has been experimentally demonstrated in large carnivores (Smith et al. 2017; Suraci et al. 2019a; Zanette and
Clinchy 2020). That comparable cascades may be caused by
the fear humans inspire in large herbivores is further supported by our results closely corresponding with those from
recent playback experiments demonstrating that the feeding
behavior of large and medium-sized carnivores is likewise
more strongly impacted by hearing the human ‘super predator’ than hearing carnivores (Clinchy et al. 2016; Suraci
et al. 2019b).
In their review of predator playback experiments, Hettena
et al. (2014) reported that mammals responded to extirpated
felids and canids in 5 of 7 and 6 of 15 experiments, respectively; hence, that extirpated cougars and wolves elicited
responses in deer in our experiment was not unexpected.
Though prey may respond to extirpated predators, Hettena
et al.’s review indicates that preys generally show stronger
responses to predators they have been exposed to both currently and historically. In addition to humans, with which
deer overlap at both evolutionary and ecological time scales,
our comparative experiment included coyotes, a relative
newcomer to eastern North America, as well as extirpated
wolves and cougars. Our results support the notion that
evolutionary and ecological overlap of prey with a predator
lends to a heightened ability to recognize and respond to
auditory cues for that predator as humans elicited the strongest treatment efect for all three metrics. Interestingly, extirpated wolves (i) reduced the probability of deer foraging
following playbacks more and had a stronger (ii) efect on
deer feeding behavior than the only extant, wild carnivore
in the experiment, coyotes. This fnding is in accordance
with empirical results reported by Hettena et al. (2014) from
their playback experiment targeting mule deer. One potential
explanation of the strong efects of wolves relative to coyotes
is the evolutionary overlap between deer and wolves; however, it is also of note that coyotes are smaller canids that,
in eastern North America, typically hunt singularly or in
small packs and more commonly prey on juvenile than adult
deer (Benson et al. 2017; Bragina et al 2019). Conversely,
wolves are larger and are efcient predators of adult deer
(Benson et al. 2017). As such, wolves may be anticipated
to elicit stronger responses in deer than their locally novel
predator, coyotes. This fnding is of particular importance
given increasingly common large carnivore restoration
eforts intended to preserve or restore ecosystem function
via re-establishment of apex predator populations in vacated
portions of their ranges (Corlett 2016). One critical concern
pertaining to such endeavors relates to the potential inability
of naïve prey to respond appropriately upon encountering
re-established predators; however, our results suggest that,
if wolves were to recolonize or were reintroduced to historic
ranges, white-tailed deer populations would be behaviorally
responsive to the large canids, likely due to their evolutionary exposure to wolves (Hettena et al. 2014).
Humans have a unique ecology that includes killing large
and medium carnivores at rates exceeding those of other,
non-human predators (Darimont et al. 2015), and the aforementioned large carnivore restoration eforts are necessitated
by human persecution of carnivores large and small (Estes
et al. 2011; Di Marco 2014; Ripple et al. 2014). While much
research has focused on the efects of human disturbance
on wildlife behavior, a growing body of evidence suggests
that fear of the human ‘super predator’ may often be the
ultimate driver of wildlife responses to such disturbance.
Recent meta-analyses of worldwide data highlight alterations in movement (Tucker et al. 2018) and increases in nocturnality (Gaynor et al. 2018) of terrestrial mammals of all
sizes and types in response to human disturbance. These data
accentuate that fear of the human ‘super predator’ appears
globally pervasive, and our results suggest potentially pervasive efects on ungulate reproduction as well. Previous
experiments have indicated that the fear instilled in large
and medium carnivores by the human ‘super predator’ can
afect how frequently large carnivores kill prey (Smith et al.
2017) and induce trophic cascades (Suraci et al. 2019a).
Our results elucidate the potential for the fear of humans
Oecologia (2022) 198:91–98 97
1 3
to cause population- and community-level efects equal to
or of greater magnitude than those observed in non-human
predator–prey systems and, in concert with previous results
from experimental investigations of deer–carnivore interactions in our study area (Cherry et al. 2015, 2016a, b), corroborate recent studies demonstrating the potential for fear
of the human ‘super predator’ to incur far-reaching ecological consequences (Estes et al. 2011; Darimont et al. 2015;
Suraci et al. 2016, 2019a; Say-Sallez et al. 2019; Zanette
and Clinchy 2020).
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00442-021-05080-w.
Author contribution statement DAC carried out the data collection
and statistical analyses, participated in the design of the study, and
drafted the manuscript; LMC participated in the design of the study and
critically revised the manuscript; MC participated in the design of the
study and helped draft the manuscript; LYZ participated in the design
of the study and critically revised the manuscript; MJC participated
in the design of the study, helped draft the manuscript, and critically
revised the manuscript. All authors gave fnal approval for publication
and agree to be held accountable for the work performed therein.
Funding This work was funded by the Jones Center at Ichauway [Grant
no. JCI2020-06] and Natural Sciences and Engineering Research Council of Canada grants awarded to Liana Y. Zanette.
Data availability Cleaned datasets and code to reproduce reported
results and included fgures are available upon reasonable request to
the corresponding author and will be made publicly available on Dryad
prior to publication.
Declarations
Conflict of interest The authors declare that they have no confict of
interest.
Ethics approval Not applicable.
Consent to participate Not applicable.
References
Bates D, Machler M, Bolker B, Walker S (2015) Fitting linear mixedefects models using lme4. J Stat Softw 67:1–48
Benson JF, Loveless KM, Rutledge LY, Patterson BR (2017) Ungulate
predation and ecological roles of wolves and coyotes in eastern
North America. Ecol Appl 27:718–733
Bourbeau-Lemieux A, Festa-Bianchet M, Gaillard J-M, Pelletier F
(2011) Predator-driven component Allee efects in a wild ungulate. Ecol Lett 14:358–363
Bragina EV, Kays R, Hody A, Moorman CE, DePerno CS, Mills LS
(2019) Efects on white-tailed deer following eastern coyote colonization. J Wildl Manag 83:916–924
Brown JS, Laundre JW, Gurung M (1999) The ecology of fear: optimal foraging, game theory, and trophic interactions. J Mamm
80:385–399
Cherry MJ, Conner LM, Warren RJ (2015) Efects of predation risk
and group dynamics on white-tailed deer foraging behavior in a
longleaf pine savanna. Behav Ecol 26:1091–1099
Cherry MJ, Morgan KE, Rutledge BT, Conner LM, Warren RJ (2016a)
Can coyote predation risk induce reproduction suppression in
white-tailed deer? Ecosphere 7:e01481
Cherry MJ, Warren RJ, Conner LM (2016b) Fear, fre, and behaviorally
mediated trophic cascades in a frequently burned savanna. For
Ecol Manag 368:133–139
Chitwood MC, Lashley MA, Moorman CE, DePerno CS (2014) Confrmation of coyote predation on adult female white-tailed deer in
the southeastern united states. Southeast Nat 13:30–32
Christianson D, Creel S (2014) Ecosystem scale declines in elk recruitment and population growth with wolf colonization: A beforeafter-control-impact approach. PLoS ONE 9:e102330
Clinchy M, Zanette LY, Roberts D, Suraci JP, Buesching CD, Newman
C, Macdonald DW (2016) Fear of the human ‘super predator’ far
exceeds the fear of large carnivores in a model mesocarnivore.
Behav Ecol 27:1826–1832
Corlett RT (2016) Restoration, reintroduction, and rewildling in a
changing world. Trends Ecol Evol 31:453–462
Darimont CT, Fox CH, Bryan HM, Reimchen TE (2015) The unique
ecology of human predators. Science 349:858–860
Di Marco M, Boitani L, Malon D, Hofmann M, Iacucci A, Meijarrd E,
Visconti P, Schipper J, Rondinini C (2014) A retrospective evaluation of the global decline of carnivores and ungulates. Conserv
Biol 28:1109–1118
Dulude-de Broin F, Hamel S, Mastromonaco GF, Côté SD (2020) Predation risk and mountain goat reproduction: evidence for stressinduced breeding suppression in a wild ungulate. Funct Ecol
34:1003–1014
Epperly HK, Clinchy M, Zanette LY, McCleery RA (2021) Fear of
large carnivores is tied to ungulate habitat use: evidence from a
bifactorial experiment. Sci Rep 11(1):1–11
Estes JA, Terborgh J, Brashares JS, Power ME, Berger J, Bond WJ,
Carpenter SR, Essington TE, Holt RD, Jackson JBC et al (2011)
Trophic downgrading of planet Earth. Science 333:301–306
Gaynor KM, Hojnowski CE, Carter NH, Brashares JS (2018) The
infuence of human disturbance on wildlife nocturnality. Science
360:1232–1235
Hettena AM, Munoz N, Blumstein DT (2014) Prey responses to predator’s sounds: a review and empirical study. Ethology 120:427–452
Hill JE, DeVault TL, Belant JL (2019) Cause-specifc mortality of the
world’s terrestrial vertebrates. Glob Ecol Biogeogr 28:680–689
Macdonald DW, Newman C (2002) Population dynamics of badgers
(Meles meles) in Oxfordshire, U.K.: numbers, density and cohort
life histories, and a possible role of climate change in population
growth. J Zool Lond 256:121–138
Moll RJ, Killion AK, Hayward MW, Montgomery RA (2021) A framework for the Eltonian niche of humans. Bioscience. https://doi.org/
10.1093/biosci/biab055
R Core Team (2020) A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. https://
www.R-project.org/
Ripple WJ, Estes JA, Beschta RL, Wilmers CC, Ritchie EG, Hebblewhite M, Berger J, Elmhagen B, Letnic M, Nelson MP et al (2014)
Status and ecological efects of the world’s largest carnivores.
Science 343:151–162
Rushing CS, Rohrbaugh RW, Fiss CJ, Rosenberry CS, Rodewald
AD, Larkin JL (2020) Long-term variation in white-tailed deer
abundance shapes landscape-scale population dynamics of forestbreeding birds. For Ecol Manag 456:117629
Say-Sallaz E, Chamaillé-Jammes S, Fritz H, Valeix M (2019) Nonconsumptive efects of predation in large terrestrial mammals:
mapping our knowledge and revealing the tip of the iceberg. Biol
Conserv 235:36–52
98 Oecologia (2022) 198:91–98
1 3
Smith JA, Suraci JP, Clinchy M, Crawford A, Roberts D, Zanette LY,
Wilmers CC (2017) Fear of the human ‘super predator’ reduces
feeding time in large carnivores. Proc R Soc B 284:20170433
Stankowich T (2008) Ungulate fight responses to human disturbance:
a review and meta-analysis. Biol Cons 141(9):2159–2173
Suraci JP, Clinchy M, Dill LM, Roberts D, Zanette LY (2016) Fear of
large carnivores causes a trophic cascade. Nat Comm 7:1–7
Suraci JP, Clinchy M, Mugerwa B, Delsey M, Macdonald DW, Smith
JA, Wilmers CC, Zanette LY (2017a) A new automated behavioural response system to integrate playback experiments into
camera trap studies. Meth Ecol Evol 8:957–964
Suraci JP, Clinchy M, Zanette LY (2017b) Do large carnivores and
mesocarnivores have redundant impacts on intertidal prey? PLoS
ONE 12:e0170255
Suraci JP, Clinchy M, Zanette LY, Wilmers CC (2019a) Fear of humans
as apex predators has landscape-scale impacts from mountain
lions to mice. Ecol Lett 22:1578–1586
Suraci JP, Smith JA, Clinchy M, Zanette LY, Wilmers CC (2019b)
Humans, but not their dogs, displace pumas from their kills: an
experimental approach. Sci Rep 9:1–8
Team TA (2014) Audacity(R): Free audio editor and recorder
Tucker MA et al (2018) Moving in the anthropocene: Global reductions in terrestrial mammalian movements. Science 359:466–469
Vanak AT, Gompper ME (2009) Dogs (Canis familiaris) as carnivores: their role and function in intraguild competition. Mamm
Rev 39:265–283
Wang Y, Allen ML, Wilmers CC (2015) Mesopredator spatial and
temporal responses to large predators and human development
in the Santa Cruz Mountains of California. Biol Cons 190:23–33
Zanette LY, Clinchy M (2020) Ecology and neurobiology of fear in
free-living wildlife. Annu Rev Ecol Evol Syst 51:297–318
Zanette LY, White AF, Allen MC, Clinchy M (2011) Perceived predation risk reduces the number of ofspring songbirds produce per
year. Science 334:1398–1401