Herbivores employ a suite of antipredator behaviours to minimize risk from ambush and cursorial predators

Abstract

Prey species may adjust their use of antipredator behaviours to counter the hunting strategies (e.g. ambush versus cursorial) and the level of risk imposed by different predators. Studies of suites of behaviours across well-defined contrasts of predation risk and type are rare, however. Here we explored the degree to which six herbivore species adjusted their antipredator behaviours to two predator treatments (lion, Panthera leo, versus cheetah, Acinonyx jubatus, and wild dogs, Lycaon pictus). We focused on prey behaviour (vigilance, grouping, temporal use) at waterholes. We predicted that if the hunting strategy of the predator was the key driver of antipredator behaviour, ambushing lions would elicit a greater response than cursorial cheetah and wild dogs. Alternatively, if predator preference was the main driver, then we expected prey species to adjust their antipredator behaviours in response to the predators that specifically target them (i.e. preferred prey of the different predators). Overall, we found that the herbivores maintained greater vigilance, generally moved in larger groups and used waterholes less at dawn, at dusk or at night (when lions are active) when exposed to the potential threat of ambushing lions. However, some species within the accessible prey range of cheetah and/or wild dogs (i.e. red hartebeest, warthog, gemsbok) moved in larger groups when exposed to these predators. Yet, the magnitude of the differences in group size for these herbivores were small. Thus, we suggest that, overall, the potential threat of ambushing lions was the main driver of antipredator behaviour around waterholes, probably determined by prey weight preference and the possibility of being ambushed.

Full text

Herbivores employ a suite of antipredator behaviours to minimize risk
from ambush and cursorial predators
Douglas F. Makin
a
,
*
, Simon Chamaill
e-Jammes
b
, Adrian M. Shrader
a
a
School of Life Sciences, University of KwaZulu-Natal, Pietermaritzburg, South Africa
b
Centre d'Ecologie Fonctionnelle et Evolutive, UMR 5175, CNRS, Universit
e de Montpellier, Universit
e Paul-Val
ery MontpelliereEPHE, Montpellier, France
article info
Article history:
Received 7 November 2016
Initial acceptance 11 January 2017
Final acceptance 2 March 2017
MS. number: 16-00973R
Keywords:
group size
hunting strategies
predatoreprey interactions
prey preferences
temporal activity
vigilance
Prey species may adjust their use of antipredator behaviours to counter the hunting strategies (e.g.
ambush versus cursorial) and the level of risk imposed by different predators. Studies of suites of be-
haviours across well-defined contrasts of predation risk and type are rare, however. Here we explored
the degree to which six herbivore species adjusted their antipredator behaviours to two predator
treatments (lion, Panthera leo, versus cheetah, Acinonyx jubatus, and wild dogs, Lycaon pictus). We
focused on prey behaviour (vigilance, grouping, temporal use) at waterholes. We predicted that if the
hunting strategy of the predator was the key driver of antipredator behaviour, ambushing lions would
elicit a greater response than cursorial cheetah and wild dogs. Alternatively, if predator preference was
the main driver, then we expected prey species to adjust their antipredator behaviours in response to the
predators that specifically target them (i.e. preferred prey of the different predators). Overall, we found
that the herbivores maintained greater vigilance, generally moved in larger groups and used waterholes
less at dawn, at dusk or at night (when lions are active) when exposed to the potential threat of
ambushing lions. However, some species within the accessible prey range of cheetah and/or wild dogs
(i.e. red hartebeest, warthog, gemsbok) moved in larger groups when exposed to these predators. Yet, the
magnitude of the differences in group size for these herbivores were small. Thus, we suggest that, overall,
the potential threat of ambushing lions was the main driver of antipredator behaviour around water-
holes, probably determined by prey weight preference and the possibility of being ambushed.
©2017 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
Prey possess a whole suite of behaviours that they may employ
to reduce predation risk (Caro, 2005; Lima &Dill, 1990). In partic-
ular, vigilance and grouping are flexible behaviours that can be
used to reduce risk, although they come with associated costs. For
example, increased vigilance allows individuals to detect attacks
earlier, providing a greater chance of escaping (Lima &Bednekoff,
1999), but often reduces food intake rate (Fortin, Boyce, Merrill, &
Fryxell, 2004). Living in larger groups allows individuals to poten-
tially benefit from dilution, collective vigilance and/or deterrence
effects (Beauchamp, 2003; Schmitt, Stears, Wilmers, &Shrader,
2014), but could increase intragroup competition (Krause &
Ruxton, 2002). Because of these costs, prey are not expected to
always display a full suite of antipredator behaviours, but rather to
finely adjust antipredator behaviours to the level of risk, by prior-
itizing certain behaviours over others (e.g. vigilance, grouping,
temporal shifts; Creel, Schuette, &Christianson, 2014).
Predation risk varies both temporally and spatially across the
landscape. This translates into a ‘landscape of fear’(Laundr
e,
Hern
andez, &Altendorf, 2001) that is shaped by differences in
the prey's perception of the likelihood of meeting a specific pred-
ator (e.g. predator density, similar landscape use between predator
and prey, shared time of activity), and of the likelihood of being
killed when attacked (i.e. ‘threat’of the predator). However, as not
all predators are the same, prey species probably adjust the extent
to which they utilize different antipredator behaviours (e.g. vigi-
lance levels, group size) in response to different predators or
predator combinations.
One factor that probably greatly influences antipredator stra-
tegies is the hunting strategy of a predator. For instance, large
mammalian predators are usually classified as either cursorial or
stalking/ambush predators. Cursorial predators roam over large
areas looking for prey, and then approach prey rapidly and silently
when found (Creel &Creel, 2002; Pomilia, McNutt, &Jordan, 2015).
As a result, their distribution in the landscape is generally unpre-
dictable, and thus prey tend not to associate specific places with
predation risk from these species (see discussion in Preisser,
Orrock, &Schmitz, 2007). In contrast, ambush predators rely on
*Correspondence: D. F. Makin, School of Life Sciences, University of KwaZulu-
Natal, Private Bag X01, Scottsville 3209, Pietermaritzburg, South Africa.
E-mail address: doug.makin4@gmail.com (D. F. Makin).
Contents lists available at ScienceDirect
Animal Behaviour
journal homepage: www.elsevier.com/locate/anbehav
http://dx.doi.org/10.1016/j.anbehav.2017.03.024
0003-3472/©2017 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
Animal Behaviour 127 (2017) 225e231

places where the likelihood of meeting prey is high, relying on
small-scale vegetation cover, rather than speed, to approach prey
(Preisser et al., 2007). Thus, areas attracting prey usually also attract
ambush predators, and thus prey should increase their antipredator
behaviour when using these areas (Valeix, Fritz, et al., 2009). For
example, within the arid and semiarid environments that we
studied here, water sources attract both large mammalian herbi-
vores and their ambush predators such as lions, Panthera leo (Ogutu
et al., 2014; Thaker et al., 2011; Valeix et al., 2010; de Boer et al.,
2010).
In addition to a predator's hunting strategy, prey species prob-
ably also consider the degree of threat posed by a specific predator.
Predators tend to target prey species within specific body size
ranges (for lion: Clements, Tambling, Hayward, &Kerley, 2014;
Hayward, 2006; Hayward, Hayward, Tambling, &Kerley, 2011).
Thus, some predators will be more of a threat than others. For
example, lions are more likely to attack a 290e340 kg zebra, Equus
quagga, than a 40e70 kg impala, Aepyceros melampus (Hayward &
Kerley, 2005). As a result, prey species should increase the extent
to which they utilize specific behaviours (e.g. increase vigilance
levels) in response to their primary predators, compared to more
peripheral predators. Yet, an overarching factor that greatly in-
fluences predation risk is the overlap in the activity patterns of
predators and prey (i.e. whether they are nocturnal or diurnal;
Kronfeld-Schor &Dayan, 2003). To minimize contact with preda-
tors, prey species can shift their temporal use of the landscape to
periods when predators are least active. For example, in Hwange
National Park, Zimbabwe, most ungulate species appear to avoid
coming to drink at night when lions are in the vicinity of the
waterholes (Valeix, Fritz, et al., 2009).
Here we explored the degree to which prey species adjust their
antipredator strategies in response to different predators. We
focused our observations at waterholes in a semiarid ecosystem as
a model of key interaction areas between predators and prey, and
studied the antipredator behaviour (grouping, vigilance, time of
use) of six large herbivore species (i.e. eland, Taurotragus oryx;
gemsbok, Oryx gazella; plains zebra, red hartebeest, Alcelaphus
buselaphus caama; warthog, Phacochoerus africanus; blue wilde-
beest, Connochaetes taurinus) at these waterholes. We did this in
two sections of the same reserve that were separated by fences, one
with only lions (ambush predators), the other with cheetah, Aci-
nonyx jubatus, and wild dogs, Lycaon pictus (both cursorial preda-
tors) and no lions.
In many ecosystems, lions select and kill in areas close to water
(Ogutu et al., 2014; Thaker et al., 2011; Valeix et al., 2010; de Boer
et al., 2010). Cheetah and wild dog may also do this, but their
presence near waterholes might be less predictable as their
cursorial hunting strategies probably increase their use of areas
away from water sources, more so than lions (e.g. Ndaimani,
Tagwireyi, Sebele, &Madzikanda, 2016). Thus, we predicted that
if hunting strategy was a key driver of prey a ntipredator behaviour,
lions would elicit a greater antipredator response from prey spe-
cies than the less spatially predictable cheetah and wild dogs. This
could be through all the prey species changing their antipredator
behaviours (e.g. increased vigilance and larger groups) and/or
adjusting their temporal activity patterns more in response to li-
ons than to cheetah and wild dogs. Alternatively, if antipredator
behaviours of prey species are driven more by prey preferences of
predators, then we would expect individual prey species to change
their antipredator behaviours more in response to the predators
that specifically target them (i.e. prey falling within the predator's
preferred prey weight range) than if the prey species falls outside
the predator's prey weight range. This could then result in species-
specific differences both within and between the predator
sections.
METHODS
Ethical Note
The University of KwaZulu-Natal approved all aspects of the
research design (Ethics code: 058/14/Animal).
Data Collection
We conducted our study in Tswalu Kalahari Reserve (Tswalu
hereafter) in the Northern Cape, South Africa (S 27

13
0
30
00
and E
022

28
0
40
00
) from October 2013 to April 2015. The fenced reserve
encompasses 1000 km
2
of restored farmland (Cromhout, 2007)
located in the southern Kalahari (Roxburgh, 2008). Tswalu has a
mean annual rainfall of 250 mm, with an extended dry season
lasting from May to September/October when there is less than
10 mm rainfall (Roxburgh, 2008). Large mammalian herbivores
found in the reserve include kudu, Tragelaphus strepsiceros,
springbok, Antidorcas marsupialis, gemsbok, eland, sable, Hippo-
tragus niger, zebra, red hartebeest, warthog and wildebeest.
Tswalu is divided into two adjacent sections which support
different large predator populations, but are separated by about
50 m comprising a road and two predator fences. The western
section of the reserve (200 km
2
) contains lion (N¼24), while the
eastern section (800 km
2
) contains populations of cheetah (N~10)
and wild dog (N¼14). Habitat types across both sections are
similar, made up of Digitaria polyphylla-dominated hills, Stipagrostis
uniplumis-dominated plains and valleys and Anthephora pubescens-
dominated sand dunes (see Van Rooyan, 1999). Likewise, both
sections have a similar mean annual rainfall (mm), with
326 ±40 mm falling within the western section compared to
345 ±42 mm within the eastern section recorded over a 9-year
period. We limited data collection to the herbivore species that
occurred in both sections of the reserve. These included eland,
gemsbok, zebra, red hartebeest, warthog and wildebeest. The her-
bivores living in the two sections face different levels of predation
risk due to the hunting strategy employed, their activity patterns
and the prey weight preferences of the different predator species
(Hayward &Slotow, 2009; Hayward et al., 2007). Lion are stalk and
ambush predators that are predominantly active at night, while
cheetah and wild dogs are mostly diurnal and hunt by chasing
down their prey (Hayward &Somers, 2009). Comparing prey
weight preferences from a multisite analysis, Clements et al. (2014)
determined that lion have an accessible prey weight class range of
32e632 kg and therefore all six herbivores species monitored in
our study fall within their prey weight range. However, they tend to
prefer prey weights of 92e632 kg (Clements et al., 2014) with
wildebeest and zebra often preferentially targeted over other prey
(Sinclair, Mduma, &Brashares, 2003). In contrast, cheetah and wild
dogs have smaller accessible prey weight ranges of 14e135 kg (with
a peak weight mode of 36 kg; Hayward, Hofmeyr, O'Brien, &Kerley,
2006) and 10e289 kg (peak weight modes of 16e32 kg and
120e40 kg; Hayward, Hofmeyr, et al., 2006), respectively. There-
fore, only warthog and red hartebeest fall within the accessible
range of cheetah, while all the herbivores, except eland, fall within
the accessible prey range of wild dogs (Clements et al., 2014).
However, although warthog fall within the accessible prey range of
both cheetah and wild dogs, they are generally avoided (Clements
et al., 2014;Hayward, Hofmeyr, et al., 2006). Despite discrep-
ancies in prey weight range preferences, cheetah and wild dogs
have the highest recorded dietary overlap (73.5%; Hayward &
Kerley, 2008) of the large African predator guild and therefore
present a significant cumulative predation risk to shared prey
species. Within Tswalu, lion prey upon wildebeest and gemsbok
(Roxburgh, 2008), while cheetah prey on red hartebeest and
D. F. Makin et al. / Animal Behaviour 127 (2017) 225e231226

springbok, and wild dogs prey on kudu, red hartebeest and impala
(Makin, n.d).
Throughout the study, all three predators were observed uti-
lizing waterholes. Moreover, lions were active at waterholes pre-
dominantly at night and during crepuscular periods (80% of
observations). In contrast, cheetah and wild dogs were more
diurnal, visiting waterholes during the crepuscular periods and
during the day (65% and 70% of observations, respectively).
To assess the antipredator strategies used by the different prey
species in response to the different predators, we deployed Bush-
nell video camera traps with heat-motion sensors at five water-
holes in the cheetah and wild dog section and three waterholes in
the lion section. Camera traps were attached to trees 1 m above the
ground. This ensured that each camera's field of view extended
from the ground up to over 2.5 m. Camera traps were placed so that
they had a clear view of the entire waterhole. This enabled all in-
dividuals visiting the waterholes to be recorded. Only videos
showing clearly discernible individuals were included in the data
analysis.
We limited the chances of collecting data from the same indi-
vidual multiple times within a single recording event, by first
noting when individuals left the field of view. We then waited
30 min before collecting data from groups of the same species
comprising the same number of individuals (i.e. potentially the
same group) that entered the field of view (Linkie &Ridout, 2011;
Tambling et al., 2015). Previous studies have suggested that
30 min represents a sufficient trade-off between recording the
same individual multiple times and missing new individuals
(Rovero, Jones, &Sanderson, 2005; Tambling et al., 2015).
We analysed the video camera data recording for (1) herbi-
vores species, (2) time of day (day: 0600e1700; crepuscular:
0400e0600 and 1700e1900; night: 1900e0400), (3) typical group
size (i.e. the reflection of the animal's rather than the human
observer's experience within the group; calculated as G¼(PNi
2
)/
(N), where the sum of the squares of all individuals in all groups
(Ni
2
) is divided by the total number of individuals seen (N);
Jarman, 1974; Reiczigel, Lang, R
ozsa, &T
othm
er
esz, 2008), (4)
predator section (lion versus cheetah and wild dogs), and (5)
proportion of time individual herbivores within groups were
vigilant at waterholes. We followed the approach of P
eriquet et al.
(2010) where vigilance was monitored for a focal animal within
the centre of each group. We did this as central individuals are
less likely to be killed than individuals on the periphery; thus, any
increase in vigilance by central individuals is likely to reflect an
increase in vigilance for all individuals within the group. We
considered an animal to be vigilant when it stood in an upright
position, head alert and actively scanning with ears held forward.
All individuals recorded were adult members of the group. As
females with juveniles will maintain higher levels of vigilance to
protect dependent offspring we focused on the vigilant responses
of females without juveniles (P
eriquet et al., 2010). Additionally,
the study was conducted following a severe drought year and
therefore there was little recruitment into the different herbivore
populations during this period, with most breeding groups con-
sisting of only adults and subadults from the previous year (Makin
Pers. Obs.).
We recorded the proportion of time each individual spent
vigilant at waterholes over a 10 min period or over the entire time
herbivore groups were drinking at a waterhole if it was less than
10 min. We defined both these time periods as an observation.
Vigilance was recorded for individuals within groups that were
close to the waterhole (i.e. drinking or standing on the water's
edge). Within the lion section, we recorded 85 wildebeest, 23
eland, 91 gemsbok, 147 zebra, 36 red hartebeest and 88 warthog
observations, while in the cheetah and wild dog section we
recorded 182 wildebeest, 76 eland, 222 gemsbok, 78 zebra, 187 red
hartebeest and 275 warthog observations.
Statistical Analysis
For each herbivore species, we compared the effect of predator
section (i.e. lion versus cheetah and wild dog) and the interaction
between predator section and herbivore species on changes in the
antipredator behaviours of typical group size and proportion of
time spent vigilant using generalized linear models (GLM) with
Poisson and binomial errors, respectively. To keep the model sim-
ple, we did not include group size as a predictor of the proportion of
time spent vigilant. Preliminary analyses showed that there was no
relationship for five of the six species (all P>0.10), with only red
hartebeest showing a slight positive relationship between indi-
vidual vigilance and increasing group size (z¼3.86, P<0.01), but
this was of a very small magnitude (lion section: y¼0.03xþ0.33;
cheetah and wild dog section: y¼0.03xþ0.11). For each herbivore
species in each predator section, we visually displayed the diel
distribution of visits to waterholes using kernel-based density
plots. In addition, we tested for the statistical significance of dif-
ferences between predator sections by fitting a GLM with Poisson-
distributed errors with the number of herbivore observations
recorded at a waterhole within each time period (night, crepus-
cular, day) as the response variable, and time period and predator
section as explanatory variables, including interactions between
variables. Warthog were not recorded visiting waterholes at night
in the lion section and therefore could not be compared for this
period. All analysis was performed using R 3.21 (R Core Team, 2014)
using the lme4 package (Bates, Maechler, &Bolker, 2012), MASS
package (Venables &Ripley, 2002) and the multcomp package
(Hothorn, Bretz, &Westfall, 2008).
RESULTS
Typical group sizes varied significantly between herbivore
species (x
2
10
¼732.7, P<0.01), between the predator sections
(x
2
6
¼51.5, P<0.01) and for the interaction between herbivore
species and the different predator sections (x
2
5
¼51.4, P<0.01).
Overall, group size did not differ between the sections for eland
(z¼1.7 2 , P¼0.08; Fig. 1a). Zebra (z¼2.12, P¼0.03) and wilde-
beest (z¼5.05, P<0.01) maintained larger groups in response to
lion than to cheetah and wild dogs. In contrast, gemsbok (z¼2.18,
P¼0.03), red hartebeest (z¼2.37, P¼0.018) and warthog
(z¼3.45, P<0.01) maintained slightly larger groups in response
to cheetah and wild dogs than in response to lion (Fig. 1a).
All the herbivore species tended to be more vigilant at water-
holes within the lion section compared to within the cheetah and
wild dog section (Fig. 1b). Differences, however, were only statis-
tically significant for gemsbok (z¼2.52, P¼0.01), red hartebeest
(z¼3.54, P<0.01) and warthog (z¼2.88, P<0.01) groups, and not
for eland (z¼1.69, P¼0.09), zebra (z¼1.92, P¼0.06) or wilde-
beest (z¼0.94, P¼0.35; Fig. 1b).
Gemsbok, zebra and wildebeest were predominantly diurnal at
waterholes in both predator sections (Fig. 2). However, when we
compared the differences in temporal waterhole use (day, crepus-
cular, night) for the same species across sections, we found statis-
tically significant differences in the waterhole use of gemsbok
(z¼2.58, P¼0.01), zebra (z¼2.48, P¼0.02) and wildebeest
(z¼2.14, P¼0.03). Specifically, in the lion section, these herbi-
vores visited the waterholes less during the night (z¼10.98,
z¼5.161, z¼3.63, all P<0.01, respectively) and during
crepuscular periods (z¼3.58, z¼3.59, z¼3.35, all P<0.01,
respectively) than they did in the cheetah and wild dog section.
There were no statistically significant differences in temporal use of
D. F. Makin et al. / Animal Behaviour 127 (2017) 225e231 227

waterholes for eland (night: z¼0.12, P¼1.00; crepuscular:
z¼0.12, P¼1.0 0; d a y: z¼0.36, P¼0.998), red hartebeest
(night: z¼0.69, P¼0.982; crepuscular: z¼0.33, P¼0.999; day:
z¼0.29, P¼0.997) and warthog (crepuscular: z¼0.03, P¼1.0 0 ;
day: z¼0.26, P¼0.998; Fig. 2).
DISCUSSION
In response to predators, prey species can adjust their behaviour
in several ways to reduce risk (Caro, 2005; Creel et al., 2014; Lima &
Dill, 1990). However, as not all predators impose the same threat,
the behavioural strategies utilized by prey are likely to vary in
response to different predator hunting modes (i.e. ambush versus
cursorial), overlap in activity patterns (i.e. nocturnal versus diurnal)
and their prey preferences. We found that the antipredator
behavioural strategies of six herbivore species differed between the
lion and cheetah and wild dog sections. Overall, lions had the
greatest effect suggesting that the threat of this ambush predator
around waterholes was a key driver of the observed antipredator
behavioural adjustments of most of the herbivores.
It is possible, however, that the differences in antipredator be-
haviours we recorded were driven by landscape differences be-
tween the sections, although the two sections were only separated
by about 50 m, had identical history of land use, similar climates/
rainfall and similar topography (Cromhout, 2007; Van Rooyan,
1999). As a result, we believe that it is more likely that the
behavioural differences between the two sections were driven by
differences in predation risk posed by the two sets of predators.
Across African landscapes, lions are one of the most dangerous
predators that herbivores can encounter. The combination of their
large body size and group-hunting tactics mean that they can
successfully kill a number of species ranging from warthogs to large
herbivores including buffalo, Syncerus caffer (550e700 kg), giraffe,
Giraffa spp. (700e1400 kg) and in some cases even elephants,
Loxodonta africana (up to 7 years old; 700e900 kg; Hanks, 1972;
Loveridge, Hunt, Murindagomo, &Macdonald, 2006). Moreover,
lion actively select habitats close to waterholes and therefore pre-
sent a significant risk to herbivores aggregated around these water
sources (Valeix, Fritz, et al., 2009; de Boer et al., 2010).
Comparing differences in the herbivore species' antipredator
behaviours between predator sections, we found that most
*****
*
*
*
Lion Cheetah/wild dogs
25
20
15
10
5
0
Typical group sizes
0.6
0.5
0.4
0.3
0.2
0.1
0
(a)
(b)
Proportion of time spent vigilant
Eland Gemsbok Plains zebra Red hartebeest Warthog Wildebeest
Eland Gemsbok Plains zebra Red hartebeest Wartho
g
Wildebeest
Figure 1. (a) Typical group sizes and (b) mean proportion of time herbivore groups were vigilant at waterholes in the two predator sections (lion versus cheetah and wild dogs).
Bars represent SE. *P<0.05.
D. F. Makin et al. / Animal Behaviour 127 (2017) 225e231228

herbivore species adjusted their behaviours so as to minimize the
risk of attack from ambushing lions. This was evident in that all the
herbivore species maintained greater vigilance in the lion section
(significantly so for gemsbok, red hartebeest and warthog) than the
cheetah and wild dog section. This could also be partly because
vigilance may not be so necessary in the face of cursorial predators
that often ‘test’herds for vulnerable animals (Creel &Creel, 2002).
Moreover, all herbivores preferred to utilize waterholes during
midday when lions tended to be less active (Tambling et al., 2015;
Valeix, Loveridge, et al., 2009). The fact that zebra, wildebeest and
gemsbok (all preferred prey of lions) reduced their night-time us-
age of waterholes in the lion section indicates that these species
adjusted their activity patterns to reduce contact with lions. In
addition, both zebra and wildebeest moved in larger groups in the
lion section than in the cheetah and wild dog section. Thus, both
zebra and wildebeest increased the use of their range of anti-
predator behaviours against their main predator, lions. This was
similar to Valeix, Fritz, et al. (2009) and Valeix, Loveridge, et al.
(2009) who found that in Hwange National Park, wildebeest and
zebra increased their group sizes with the long-term risk of
encountering lion around waterholes.
As all the herbivore species in our study fall within the prey
weight range of lions, it is difficult to tease apart which factors are
driving the observed behavioural differences between the herbi-
vores in the two predator sections. However, as the main species
making these adjustments (i.e. zebra and wildebeest) are generally
preferred prey species of lions, we suggest that it is probably the
combination of prey preference of the lions (Sinclair et al., 2003)
and heightened predation risk at the waterholes (i.e. possibility of
being ambushed) that lead to adjustments to these and the other
species' antipredator strategies (de Boer et al., 2010).
In contrast to the general response towards lions, we found that
red hartebeest, warthog (accessible and avoided prey of cheetah and
wild dog, respectively; Hayward, O'Brien, Hofmeyr, &Kerley, 2006;
Marker, Dickman, Wilkinson, Schumann, &Fabiano, 20 07)and
gemsbok (within the prey range of wild dogs; Hayward, Hofmeyr,
et al., 2006) moved in larger groups at waterholes in the cheetah
and wild dog section. This suggests that these herbivores were
responding to the combined threat from cheetah and wild dog. Yet,
all three of these herbivore species also fall within the prey range of
lions. A potential reason for why these herbivores maintained larger
groups in the presence of cheetah and wild dogs is that it is possible
that the combined risk from these predators was greater than the
risk from lions alone. This may have been due to more frequent
contact with cheetah and wild dogs than lions. Within the lion
section, there were only two prides of lions. In contrast, in the
cheetah and wild dog section there were a minimum of 10 cheetahs,
each moving separately (D. F. Makin, personal observation), and one
pack of wild dogs (i.e. 11 potential encounters with predators).
Moreover, as cheetahand wild dogs are predominantly active during
the day (Hayward &Somers, 2009), and thus there is a greater
overlap in the activity patterns of these predators and their prey, it is
possible that by moving in larger groups these herbivores reduced
the combined risk from both predators (Clements et al., 2014).
Despite this, the differences in group size for all six herbivore
species were small with typical group sizes differing by only a few
individuals between the predator sections. This suggests that group
size may in fact not be a major adaptive response to increased
0.25
0.2
0.15
0.1
0.05
0
0.4
0.3
0.2
0.1
0
Cheetah and wild dogs
Lions
Cheetah and wild dogs
Lions
Cheetah and wild dogs
Lions
Cheetah and wild dogs
Lions
Cheetah and wild dogs
Lions
Cheetah and wild dogs
Lions
0000 0300 0600 0900 1200 1500 1800 2100 2400
0000 0300 0600 0900 1200 1500 1800 2100 2400
0000 0300 0600 0900 1200 1500 1800 2100 2400 0000 0300 0600 0900 1200 1500 1800 2100 2400
0000 0300 0600 0900 1200 1500 1800 2100 2400
0000 0300 0600 0900 1200 1500 1800 2100 2400
Density
0.3
0.25
0.2
0.15
0.1
0.05
0
0.4
0.3
0.2
0.1
0
0.4
0.5
0.3
0.2
0.1
0
0.3
0.25
0.2
0.15
0.1
0.05
0
Time (hours)
(a)
(b)
(c)
(d)
(e)
(f)
Figure 2. Density kernel plots estimating the daily activity patterns of (a) eland, (b) gemsbok, (c) zebra, (d) red hartebeest, (e) warthog and (f) wildebeest at waterholes in the two
predator sections (cheetah and wild dog: dotted black line; lions: solid black line) and time periods (night: dark grey; crepuscular: light grey; day: white).
D. F. Makin et al. / Animal Behaviour 127 (2017) 225e231 229

predation risk from predators at waterholes in Tswalu. If this is the
case, then this suggests that all the herbivores in our study adjusted
their antipredator behaviours more in response to the potential
threat from the two prides of ambushing lions than to the cursorial
cheetah and wild dogs. Additional support for this comes from the
finding that herbivores preferred by cheetah and wild dogs did not
shift to utilizing water holes during the night when these predators
were less active (Ford et al., 2015; Hilborn, Pettorelli, Orme, &
Durant, 2012; but see Cozzi et al., 2012). One possible reason for
this is that there may have been costs that prevented these herbi-
vores from doing this, but it is unclear what these costs may be.
In conclusion, we found that the herbivores tended to display
stronger antipredator behaviour in response to lions (i.e. greater
vigilance, larger groups and temporal shift in water hole usage)
than when living with cheetah and wild dogs. This suggests that the
cursorial hunting strategy of cheetah and wild dogs imposed lower
perceived risk around waterholes than the stalk and ambush
strategy adopted by lions. Our study is one of the few that has
directly addressed the effect of hunting mode on prey behaviour,
using a powerful semiexperimental design. Moreover, our results
support the common assertion that ambush predators are likely to
induce stronger nonconsumptive effects on prey than cursorial
predators (Middleton et al., 2013; Preisser et al., 2007).
Yet, as predation risk varies across the landscape (Laundr
e,
Hern
andez, &Ripple, 2010; Shrader, Brown, Kerley, &Kotler, 2008),
behavioural strategies utilized to reduce this risk probably also vary
spatially. As waterholes represent key interaction areas between
predators and prey, the suite of behaviours utilized by each species
we recorded probably reflect those best suited against ambush
predators.However, as the possibility of ambush probablydeclines as
herbivores move away from waterholes, herbivores possibly adjust
their antipredator behaviours to reduce the use of behaviours that
decrease risk from ambush predators and increase those that are
better suited against roaming cursorial predators. Observations in
landscapes with multiple predators using contrasting hunting stra-
tegies will be required to test this hypothesis. However, in such a
landscape, Thaker et al. (2011) found that all prey species tended to
avoid the activity areas of ambush, but not of cursorial, predators.
They also found that prey responded more to habitat cues than to
actual predator distribution. See Schmitz (2007) for further discus-
sion on additive or substitutive effects in multipredator systems.
Despite focusing on a number of antipredator behaviours, there
are others we did not consider, for instance multiscale habitat use
(e.g. Padi
e et al., 2015) or reactive responses (e.g. Courbin et al.,
2016). Our study, however, highlights an important point, namely
that ecologists (including ourselves) need to move beyond focusing
on a limited set of behaviours (e.g. just vigilance) when studying
prey species' responses to predation risk. This will be difficult, but is
required, as our results highlight that animals do not reduce risk by
simply adjusting one or two behaviours, but rather exploit and
combine an array of antipredator behaviours.
Acknowledgments
We thank the Oppenheimer family and the Tswalu foundation
for allowing us to conduct the study in Tswalu. Funding for this
research was provided by UKZN, NRF (Research Grant 77582, AMS),
GreenMatter (DM) and the Tswalu Foundation. Two referees pro-
vided constructive comments on the manuscript.
References
Bates, D., Maechler, M., & Bolker, B. (2012). lme4: Linear mixed-effects models using
S4 classes (2011). R package version 0.999375-42 http://CRAN.R-project.org/
package¼lme4.
Beauchamp, G. (2003). Group-size effects on vigilance: A search for mechanisms.
Behavioural Processes, 63(3), 111e121.
Caro, T. (2005). Antipredator defenses in birds and mammals. Chicago, IL: University
of Chicago Press.
Clements, H. S., Tambling, C. J., Hayward, M. W., & Kerley, G. I. (2014). An objective
approach to determining the weight ranges of prey preferred by and accessible
to the five large African carnivores. PloS One, 9(7), e101054.
Courbin, N., Loveridge, A. J., Macdonald, D. W., Fritz, H., Valeix, M., Makuwe, M. T.,
et al. (2016). Reactive responses of zebras to lion encounters shape their
predator-prey space game at large scale. Oikos, 125, 829e838.
Cozzi, G., Broekhuis, F., McNutt, J. W., Turnbull, L. A., Macdonald, D. W., & Schmid, B.
(2012). Fear of the dark or dinner by moonlight? Reduced temporal partitioning
among Africa's large carnivores. Ecology, 93(12), 2590e2599.
Creel, S., & Creel, N. M. (2002). The African wild dog: Behavior, ecology, and conser-
vation. Princeton, NJ: Princeton University Press.
Creel, S., Schuette, P., & Christianson, D. (2014). Effects of predation risk on group
size, vigilance, and foraging behavior in an African ungulate community.
Behavioral Ecology, 25(4), 773e784.
Cromhout, M. (2007). The ecology of the African buffalo in the eastern Kalahari region,
South Africa [M.Sc. thesis]. Pretoria, South Africa: University of Pretoria.
de Boer, W. F., Vis, M. J., de Knegt, H. J., Rowles, C., Kohi, E. M., van Langevelde, F.,
et al. (2010). Spatial distribution of lion kills determined by the water de-
pendency of prey species. Journal of Mammalogy, 91(5), 1280e1286.
Ford, A. T., Goheen, J. R., Augustine, D. J., Kinnaird, M. F., O'Brien, T. G., Palmer, T. M.,
et al. (2015). Recovery of African wild dogs suppresses prey but does not trigger
a trophic cascade. Ecology, 96(10), 2705e2714.
Fortin, D., Boyce, M. S., Merrill, E. H., & Fryxell, J. M. (2004). Foraging costs of vig-
ilance in large mammalian herbivores. Oikos, 107(1), 172e180.
Hanks, J. (1972). Growth of the African elephant (Loxodonta africana). African Journal
of Ecology, 10(4), 251e272.
Hayward, M. (2006). Prey preferences of the spotted hyaena (Crocuta crocuta) and
degree of dietary overlap with the lion (Panthera leo). Journal of Zoology, 270(4),
606e614.
Hayward, M. W.,Hayward, G. J., Tambling, C. J., & Kerley, G. I. (2011).Do lions Panthera
leo actively select prey or do prey preferences simply reflect chance responses
via evolutionary adaptations to optimal foraging? PloS One, 6(9), e23607.
Hayward, M. W., Hofmeyr, M., O'Brien, J., & Kerley, G. I. H. (2006). Prey preferences
of the cheetah (Acinonyx jubatus) (Felidae: Carnivora): Morphological limita-
tions or the need to capture rapidly consumable prey before kleptoparasites
arrive? Journal of Zoology, 270(4), 615e627. http://dx.doi.org/10.1111/j.1469-
7998.2006.00184.x.
Hayward, M. W., & Kerley, G. I. (2005). Prey preferences of the lion (Panthera leo).
Journal of Zoology, 267(03), 309e322.
Hayward, M. W., & Kerley, G. I. (2008). Prey preferences and dietary overlap
amongst Africa's large predators. South African Journal of Wildlife Research,
38(2), 93e108.
Hayward, M. W., Kerley, G. I., Adendorff, J., Moolman, L. C., O'brien, J., Sholto-
Douglas, A., et al. (20 07). The reintroduction of large carnivores to the Eastern
Cape, South Africa: An assessment. Oryx, 41(02), 205e214.
Hayward, M. W., O'Brien, J., Hofmeyr, M., & Kerley, G. I. (2006). Prey preferences of
the African wild dog Lycaon pictus (Canidae: Carnivora): Ecological re-
quirements for conservation. Journal of Mammalogy, 87(6), 1122e113 1.
Hayward, M. W., & Slotow, R. (2009). Temporal partitioning of activity in large
African carnivores: Tests of multiple hypotheses. South African Journal of Wildlife
Research, 39(2), 109e125.
Hayward, M. W., & Somers, M. (2009). Reintroduction of top-order predators. Chi-
chester, U.K.: J. Wiley.
Hilborn, A., Pettorelli, N., Orme, C. D. L., & Durant, S. M. (2012). Stalk and chase: How
hunt stages affect hunting success in Serengeti cheetah. Animal Behaviour,
84(3), 701e706.
Hothorn, T., Bretz, F., & Westfall, P. (2008). Simultaneous inference in general
parametric models. Biometrical Journal, 50(3), 346e363.
Jarman, P. (1974). The social organisation of antelope in relation to their ecology.
Behaviour, 48(1), 215e267.
Krause, J., & Ruxton, G. D. (2002). Living in groups. Oxford, U.K: Oxford University
Press.
Kronfeld-Schor, N., & Dayan, T. (2003). Partitioning of time as an ecological
resource. Annual Review of Ecology, Evolution, and Systematics, 34,153
e181.
Laundr
e, J. W., Hern
andez, L., & Altendorf, K. B. (2001). Wolves, elk, and bison:
Reestablishing the “landscape of fear”in Yellowstone National Park, USA. Ca-
nadian Journal of Zoology, 79(8), 1401e1409.
Laundr
e, J. W., Hern
andez, L., & Ripple, W. J. (2010). The landscape of fear: Ecological
implications of being afraid. Open Ecology Journal, 3,1e7.
Lima, S. L., & Bednekoff, P. A. (1999). Temporal variation in danger drives anti-
predator behavior: The predation risk allocation hypothesis. American Natu-
ralist, 153(6), 649e659.
Lima, S. L., & Dill, L. M. (1990). Behavioral decisions made under the risk of pre-
dation: A review and prospectus. Canadian Journal of Zoology, 68(4), 619e640.
Linkie, M., & Ridout, M. (2011). Assessing tigereprey interactions in Sumatran
rainforests. Journal of Zoology, 284(3), 224e229.
Loveridge, A. J., Hunt, J. E., Murindagomo, F., & Macdonald, D. W. (2006). Influence of
drought on predation of elephant (Loxodonta africana) calves by lions (Panthera
leo) in an African wooded savannah. Journal of Zoology, 270(3), 523e530.
Makin, D. F.. (n.d.) [Cheetah and wild dog prey kill data, Tswalu Kalahari Reserve].
Unpublished raw data.
D. F. Makin et al. / Animal Behaviour 127 (2017) 225e231230

Marker, L., Dickman, A., Wilkinson, C., Schumann, B., & Fabiano, E. (2007). The
namibian cheetah: Status report. Cat News, 3,4e13.
Middleton, A. D., Kauffman, M. J., McWhirter, D. E., Jimenez, M. D., Cook, R. C.,
Cook, J. G., et al. (2013). Linking anti-predator behaviour to prey demography
reveals limited risk effects of an actively hunting large carnivore. Ecology Letters,
16, 1023e1030.
Ndaimani, H., Tagwireyi, P., Sebele, L., & Madzikanda, H. (2016). An ecological
paradox: The African wild dog (Lycaon Pictus) is not attracted to water points
when water is scarce in Hwange National Park, Zimbabwe. PloS One, 11(1),
e0146263.
Ogutu, J. O., Reid, R. S., Piepho, H.-P., Hobbs, N. T., Rainy, M. E., Kruska, R. L., et al.
(2014). Large herbivore responses to surface water and land use in an East
African savanna: Implications for conservation and human-wildlife conflicts.
Biodiversity and Conservation, 23(3), 573e596.
Padi
e, S., Morellet, N., Hewison, A., Martin, J. L., Bonnot, N., Cargnelutti, B., et al.
(2015). Roe deer at risk: Teasing apart habitat selection and landscape con-
straints in risk exposure at multiple scales. Oikos, 124(11), 1536e1546.
P
eriquet, S., Valeix, M., Loveridge, A. J., Madzikanda, H., Macdonald, D. W., & Fritz, H.
(2010). Individual vigilance of African herbivores while drinking: The role of
immediate predation risk and context. Animal Behaviour, 79(3), 665e671.
Pomilia, M. A., McNutt, J. W., & Jordan, N. R. (2015). Ecological predictors of African
wild dog ranging patterns in northern Botswana. Journal of Mammalogy,1e10.
Preisser, E. L., Orrock, J. L., & Schmitz, O. J. (2007). Predator hunting mode and
habitat domain alter nonconsumptive effects in predatoreprey interactions.
Ecology, 88(11), 2744e2751.
Reiczigel, J., Lang, Z., R
ozsa, L., & T
othm
er
esz, B. (2008). Measures of sociality: Two
different views of group size. Animal Behaviour, 75(2), 715e721.
Rovero, F., Jones, T., & Sanderson, J. (2005). Notes on Abbott's duiker (Cephalophus
spadix True 1890) and other forest antelopes of Mwanihana Forest, Udzungwa
Mountains, Tanzania, as revealed by camera-trapping and direct observations.
Tropical Zoology, 18(1), 13e23.
Roxburgh, D. J. (2008). Prey and range use of lions on Tswalu Kalahari Reserve. [M.Sc.
thesis]. Pretoria, South Africa: University of Pretoria.
Schmitt, M. H., Stears, K., Wilmers, C. C., & Shrader, A. M. (2014). Determining the
relative importance of dilution and detection for zebra foraging in mixed-
species herds. Animal Behaviour, 96,151e158.
Schmitz, O. J. (2007). Predator diversity and trophic interactions. Ecology, 88,
2415e2426.
Shrader, A. M., Brown, J. S., Kerley, G. I. H., & Kotler, B. P. (2008). Do free-ranging
domestic goats show ‘landscapes of fear’? Patch use in response to habitat
features and predator cues. Journal of Arid Environments, 72(10), 1811e1819.
http://dx.doi.org/10.1016/j.jaridenv.2008.05.004.
Sinclair, A., Mduma, S., & Brashares, J. S. (2003). Patterns of predation in a diverse
predatoreprey system. Nature, 425(6955), 288e290.
Tambling, C. J., Minnie, L., Meyer, J., Freeman, E. W., Santymire, R. M., Adendorff, J.,
et al. (2015). Temporal shifts in activity of prey following large predator rein-
troductions. Behavioral Ecology and Sociobiology, 69(7), 1153e1161.
Team, R. C. (2014). R: A language and environment for statistical computing. Vienna,
Austria: R Foundation for Statistical Computing. http://www.RStudio.com.
Thaker, M., Vanak, A. T., Owen, C. R., Ogden, M. B., Niemann, S. M., & Slotow, R.
(2011). Minimizing predation risk in a landscape of multiple predators: Effects
on the spatial distribution of African ungulates. Ecology, 92(2), 398e407.
Valeix, M., Fritz, H., Loveridge, A. J., Davidson, Z., Hunt, J. E., Murindagomo, F., et al.
(2009). Does the risk of encountering lions influence African herbivore
behaviour at waterholes? Behavioral Ecology and Sociobiology, 63(10),
1483e1494.
Valeix, M., Loveridge, A., Chamaill
e-Jammes, S., Davidson, Z., Murindagomo, F.,
Fritz, H., et al. (2009). Behavioral adjustments of African herbivores to predation
risk by lions: Spatiotemporal variations influence habitat use. Ecology, 90(1),
23e30.
Valeix, M., Loveridge, A. J., Davidson, Z., Madzikanda, H., Fritz, H., &
Macdonald, D. W. (2010). How key habitat features influence large terrestrial
carnivore movements: Waterholes and African lions in a semi-arid savanna of
north-western Zimbabwe. Landscape Ecology, 25(3), 337e351.
Van Rooyan, N. (1999). The vegetation types and veld condition of Tswalu Private
Desert Reserve. Unpublished Report to the Management of Tswalu Kalahari
Reserve.
Venables, W. N., & Ripley, B. D. (2002). Random and mixed effects. In
W. N. Venables, & B. D. Ripley (Eds.), Modern Applied Statistics with S (pp.
271e
300). New York, NY: Springer.
D. F. Makin et al. / Animal Behaviour 127 (2017) 225e231 231