African ungulates and their drinking problems: hunting and predation risks
constrain access to water
, Marion Valeix
, Hervé Fritz
, Hillary Madzikanda
, Steeve D. Côté
Université de Lyon, CNRS Université Claude Bernard Lyon, Laboratoire Biométrie et Biologie Evolutive, Bât Gregor Mendel, Villeurbanne, France
Département de Biologie, Pavillon Alexandre-Vachon, Université Laval, Québec, Canada
CIRAD eEMVT, Integrated Wildlife Management Research Unit, Campus International de Baillarguet, Montpellier, France
Zimbabwe Parks and Wildlife Management Authority, Causeway, Harare, Zimbabwe
Received 27 April 2011
Initial acceptance 23 May 2011
Final acceptance 28 September 2011
Available online 16 November 2011
MS. number: A11-00345R
ecology of fear
Prey make several behavioural adjustments to minimize the risk of predation by their natural predators.
When hunted, however, they may have to adjust their behaviour further or differently to cope with this
additional mortality risk. Here, we investigated whether African large ungulates would adjust their
behaviour in response to hunting risk (i.e. risk of being shot by human hunters). We predicted that they
would shift their use of surface water, a key and scarce resource in African savannas, from day hours to
night hours to reduce the risk of encountering human hunters. In Hwange National Park, Zimbabwe, we
monitored waterholes to record the temporal drinking niche of three nonhunted ungulates (i.e. impala,
Aepyceros melampus, greater kudu, Tragelaphus strepsiceros, sable antelope, Hippotragus niger). We also
monitored waterholes in hunting areas in the vicinity of Hwange National Park. In Hwange National Park,
the three species avoided waterholes at night, when the risk of natural predation was higher. Conversely,
in the hunting areas, all three species visited waterholes more often at night. Impala and greater kudu,
however, were less prone to switch towards night-time use of waterholes in hunting areas compared to
sable antelope, although all three species were exposed to similar hunting risk. Our results suggest that
hunting may force African ungulates to shift their visits at waterholes from day hours towards night
hours, but that the magnitude of this shift may be constrained by the predation risk imposed by large
nocturnal carnivores. We conclude that species preyed upon by natural predators adjust their
antipredator behaviour in response to the additional risk of predation imposed by hunting.
Ó2011 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
With ongoing increase of human disturbance on wildlife
habitats, there is growing concern about how disturbance stimuli
may affect animal behaviour (Geist 1970; Caro 1999a). Animals
respond to human disturbance similarly to predation risk, that is,
by adjusting their behaviour in order to reduce the disturbance, for
example, by increasing ﬂight initiation distance or vigilance, or by
shifting habitats (reviewed in Stankowich & Blumstein 2005). There
may be medium- to long-term costs associated with behavioural
adjustments (e.g. lower resource intake rate, lower diet quality, or
decreased mating opportunities; Lima & Dill 1990; Lima 1998). The
repercussions of these nonlethal effects of predation or disturbance
(sensu Lima 1998) might therefore alter the ﬁtness of individuals,
affect population dynamics (Creel et al. 2007; Creel & Christianson
2008), and ultimately community structure and ecosystem
functioning (Brown et al. 1999; Ripple & Beschta 2004).
Hunting may be considered an extreme type of human distur-
bance as individuals from harvested populations experience
a direct risk of mortality. Hunting disturbance disrupts normal
activities, alters diurnal activity rhythms and increases ﬂight
initiation distances in many waterbird species (reviewed in Madsen
& Fox 1995). In large ungulates, studies have reported increased
ﬂight initiation distance in response to hunting disturbance
(reviewed in Stankowich 2008). However, much less is known
about long-term behavioural responses such as shifts in habitat use
(Swenson 1982) and feeding sites (Benhaiem et al. 2008), or
changes in activity patterns (Kufeld et al. 1988).
Here we address an example of long-term behavioural adjust-
ment to hunting in large game ungulates. We assessed whether
ungulates adjust their temporal use of key and scarce resources in
response to hunting risk (i.e. risk of being shot by human hunters).
A common behavioural response of prey to predation risk is to
avoid risky areas where predators are abundant (Ripple & Beschta
2004; Creel et al. 2005; Valeix et al. 2009a). In certain circum-
stances, however, spatial avoidance of predators may not be
possible, particularly when prey have no choice but to use risky
*Correspondence: W.-G. Crosmary, Département de Biologie, Pavillon
Alexandre-Vachon, 1045 avenue de la Médecine, Université Laval, Québec, Québec
G1V 0A6, Canada.
E-mail address: firstname.lastname@example.org (W.-G. Crosmary).
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/anbehav
0003-3472/$38.00 Ó2011 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
Animal Behaviour 83 (2012) 145e153
areas to access rare and scattered resources where predators might
look for them. Under such situations, prey may shift their niche
along the time axis to reduce temporal overlap with predator
activity (reviewed in Kronfeld-Schor & Dayan 2003). While such
temporal adjustments have been illustrated in some predatoreprey
studies (Kotler et al. 1991; Fenn & Macdonald 1995; Roth & Lima
2007; Valeix et al. 2009b), much less has been reported on
hunted species (but see Kilgo et al. 1998; Sunde et al. 2009),
particularly on how hunting risk may affect resource utilization.
Semiarid African savannas are excellent ecosystems to investigate
how hunting risk may constrain temporal access of ungulates to
scarce water resources. First, hunting is very common in Africa, more
particularly sport hunting (Lindsey et al. 2007), or recreational
hunting practised by paying tourists. Yet, only few studies have
reported behavioural adjustments in response to hunting risk in
African ungulates (Caro 1999b; Matson et al. 2005; Setsaas et al.
2007) compared to ungulates from the northern hemisphere
(e.g. reindeer, Rangifer tarandus; reviewed in Reimers & Colman
2006). Second, surface water is limited in African semiarid
savannas. Hence, even though the risk of encountering predators
(Valeixet al. 2009a, b) and human hunters (e.g. Harrison & Bates 1991,
for ungulates in deserts of the Middle East) may be high, most
ungulates have to come to waterholes almost daily to satisfy their
drinking requirements. Third, the richness of African ungulate
communities provides the opportunity for original comparative
studies. While behavioural changes in harvested populations may be
used as indicators of species sensitivity to human hunting, Gill et al.
(2001) stressed that behavioural indicators might not always
adequately reﬂect animal response to human disturbance. For
instance, animals of lower condition, or that undergo higher
constraints on resource requirements, are less likely to adjust their
behaviour in response to human disturbance (e.g. Beale & Monaghan
2004). Likewise, behavioural response to hunting risk may expose
ungulate species to their natural predators (e.g. Kilgo et al. 1998).
Species that are therefore more vulnerable to natural predators might
be less likely to adjust their behaviour to human hunting risk.
Multispecies studies are rare and greatly needed to understand how
speciﬁc ecological constraints may inﬂuence behavioural responses
to human hunting.
We investigated how different African ungulate species (impala,
Aepyceros melampus, greater kudu, Tragelaphus strepsiceros,sable
antelope, Hippotragus niger) adjust their temporal use of waterholes
in response to hunting risk. We hypothesized that ungulates would
shift their temporal use of waterholes to night-time in hunting areas
to decrease temporal overlap with human hunters, in comparison to
protected areas where ungulates drink mainly during daytime
(Valeix et al. 2007a). We expected this shift to be partial, notably
because of the constraint exerted by predation risk at night, as the
main predators in our study area are mostly nocturnal (i.e. lion,
Panthera leo:Schaller 1972; spotted hyaena, Crocuta crocuta:Kruuk
1972). Furthermore, we predicted the magnitude of the shift to
vary according to species vulnerability to natural predation.
The species more often preyed upon (i.e. greater kudu) should
perform a shift of lower magnitude than the other species.
Hwange National Park (nonHA) in northwestern Zimbabwe
E) is approximately 15 000 km
. Sport hunting is
allowed and practised in governmental safari areas and private
concessions around nonHA: Matetsi Safari Area South (MSA South,
189 0 k m
) and Gwayi Intensive Conservation Area South (Gwayi ICA
South, ca. 880 km
)(Fig. 1). The hunting mode practised in these
hunting areas is sport hunting, and thus harvest usually represents
a small fraction of total population size, contrary to culling or
uncontrolled poaching. Harvest rates are traditionally set at about 2%
of population size for most ungulates in Zimbabwean hunting areas
(Cumming 1989). During the study period, we estimated per capita
hunting risk based on the ratio of hunting quotas over population
size estimates for the three species. We found, as did Cumming
(1989), that percentages of harvested individuals were about 2% of
population size (i.e. 2.5%, 1.7% and 1.4% for impala, greater kudu and
sable antelope, respectively). This implies similar per capita risk of
human predation for all ungulate species. These Hunting Areas (HA)
act as buffers between the National Park and human settlements
where subsistence agriculture is the principal activity and wildlife is
only present at low densities (Dunham 2002).
Vegetation is typical of southern African dystrophic wooded and
bushed savannas with patches of grasslands, dominated mainly by
Colophospermum mopane,Combretum spp., Acacia spp., Baikiaea
plurijuga and Terminalia sericea (Rogers 1993). The rainy season
ranges from November to April, and long-term annual rainfall
averages 613 mm (Chamaillé-Jammes et al. 2006). Surface water is
mainly found in natural and artiﬁcially ﬁlled waterholes, and in
some rare rivers. During the dry season, natural waterholes and
rivers dry up and most of the surface water available to animals is
found in artiﬁcially ﬁlled waterholes and river pools.
No sport hunting occurs in nonHA, but the Zimbabwe Parks and
Wildlife Management Authority (ZPWMA) allocates quotas to its
staff for food rations, mainly on elephants, Loxodonta africana, and
buffalos, Syncerus caffer, while other ungulates have not been
allocated since 2007. In the surrounding HA, hunting season ranges
from March to December (since the mid-1970s). Ungulates are
strictly hunted during daytime, from dawn (around 0600 hours) to
dusk (around1900 hours). Hunters are not allowed to hunt from
vehicles, so all hunts are performed on foot. Overall, the abundance
of large carnivores appears to be comparable between the study
HAs and their adjacent nonHA areas (Elliot 2007;Table 1). Hence,
the main difference between HA and nonHA is the predation risk
due to hunting by humans. We thus disposed of a contrasted
system, where the risk to ungulates in nonHA was exclusively
exerted by natural predators and the risk to ungulates in HA
originated both from natural predators and human hunters (further
information on the different study areas is provided in Table 1).
Our study area covered, in the peripheral HA, the hunting
concessions from which we had obtained authorization to conduct
wildlife surveys during the study period: Unit 3 in MSA South
(ca. 360 km
) and four hunting properties in Gwayi ICA South
(ca. 300 km
)(Fig. 1). In nonHA, we chose to cover the blocks that
were adjacent to those hunting concessions: Main camp
(ca. 1300 km
) and Robins (ca. 1000 km
) in the northern part of
Hwange National Park (Fig. 1). Therefore, the hunting concessions
and their adjacent blocks in nonHA presented similar vegetation
types and environmental conditions (i.e. rainfall, temperature, soil
characteristics; Ganzin et al. 2008; Peace Parks Foundation 2009).
We monitored waterholes in nonHA and surrounding HA in
2007 and 2008 during the hot dry season (AugusteOctober). In HA,
we selected waterholes on the basis of availability of standing water
and signs of recent ungulates presence (i.e. fresh spoors and faeces).
We monitored nine waterholes once in 2007 and seven waterholes
once in 2008. Only two waterholes were surveyed both years (i.e.13
different waterholes). In nonHA, each waterhole is surveyed once
a year (since the 1970s) by Wildlife and Environment Zimbabwe
(WEZ) and ZPWMA. Accounting for vegetation similarities with HA
as well as water availability, we randomly selected, from the
W.-G. Crosmary et al. / Animal Behaviour 83 (2012) 145e153146
existing database, 19 waterholes in 2007 and 13 waterholes in 2008
(i.e. 19 different waterholes).
Monitoring consisted in 24 h counts during full moon periods to
document temporal drinking niches of ungulates at waterholes. Full
moon nights guarantee good visibility conditions to carry out
observations, particularly in the areas surrounding waterholes,
which are characterized by the absence of vegetation. Observations
were conducted with binoculars or spotting scopes (15e45) from
platforms, tree hides or a car parked about 100 m away from the
waterholes to avoid disturbing animals. We recorded the time at
which ungulate groups arrived at waterholes, as well as species,
group size and presence/absence of young. We accounted for group
size and presence of young as cofactors in the decision to access
waterholes during daytime or at night, as both group size (‘many
eyes effect’:Pulliam 1973;‘dilution effect’:Hamilton 1971; Dehn
1990) and presence of young (Berger 1991; Burger & Gochfeld
1994) are known to inﬂuence antipredator behaviours.
We focused on two browsers: impala and greater kudu, and one
woodland grazer: sable antelope. These ungulates are strictly
dependent on surface water and they visit waterholes regularly,
mostly during daytime in nonHA (Valeix et al. 2007a). In the study
area, the main predators are lions and spotted hyaenas (Table 1).
Lions are ambush predators and many lion kills (w40%) occur close
to a waterhole (Valeix et al. 2009a, 2011). Hyaenas are cursorial
predators that commonly hunt in waterhole areas (Salnicki et al.
2001; N. Drouet-Hoguet, personal observations). Both carnivore
species are nocturnal and their visits to waterholes occur mainly at
night (70% and 80% of the lion visits recorded during the
Site characteristics and densities of ungulates and large predators in nonhunting
areas (NonHA) in Hwange National Park, Zimbabwe, and in hunting areas (HA) in the
vicinity of Hwange National Park
Robins Gwayi ICA
) 1300 1000 300 360
1.23 1.3 4.34 2.5
Road length (km) 325 400 172 263
Ungulate density (indiv./km
Impala 1.43 (0.47) 5.45 (0.31) e3.0 (0.15)
Greater kudu 1.59 (0.24) 1.52 (0.36) e2.5 (0.20)
Sable antelope 0.22 (0.44) ee1.0 (0.45)
Lions (call-up surveys,
2.8e5.5 *2e3.9 5.3e7.7
Lions (spoor transects) 2.6 *2.2 4.5
Hyaena (call-up surveys,
11.3e22.1 *5.5e10.9 6.5e11.7
Hyaena (spoors/100 km
) 27.2 *11.6 26.5
Leopard (spoors/100 km
) 1.2 e4.2 1.3
Wild dog (spoors/100 km
) 0.4 e1.6 0.2
Cheetah (spoors/100 km
) 0.1 e0.5 1.5
Coefﬁcients of variation associated with ungulate density estimates are given in
parentheses and were calculated by distance sampling (Buckland et al. 2001).
Density estimates were not available but were comparable to those at the Main
Camp (A. J. Loveridge & N. Elliot, personal communication).
For call-up surveys, the lower estimate corresponds to the 2.5 km response
range, and the higher estimate corresponds to the 3.2 km response range
Figure 1. Hwange National Park (nonHA) and adjacent hunting areas (Gwayi ICA South and MSA south, HA) in Zimbabwe. Study area is delineated by thick black lines.
W.-G. Crosmary et al. / Animal Behaviour 83 (2012) 145e153 147
Impala - nonHA
Kudu - nonHA
Sable - nonHA
Kudu - HA
Sable - HA
Impala - HA
25 16 9 16 25
36 25 16 94
96,25 4 2,25 1
Figure 2. Temporal visits at waterholes by groups of impala, greater kudu and sable antelope in 2007 and 2008 during daytime (0600e1900 hours) and night-time (1900e0600
hours) under two contrasted situations: natural predation and no sport hunting (nonHA, Hwange National Park, Zimbabwe) versus natural predation and sport hunting (HA). Each
bar length represents the number of groups that visited waterholes duringeach hour; the area of each bar is thus proportional to the frequency ofgroup visits. The black line in each
graph indicates mean arrival time, with 95% conﬁdence interval.
W.-G. Crosmary et al. / Animal Behaviour 83 (2012) 145e153148
monitoring occurred at night in nonHA and HA, respectively, and
90% and 100% of the hyaena visits occurred at night in nonHA and
HA, respectively; see also Supplementary Material, Fig. S1). Lions
were not hunted in HAs at the time of the study because of
a moratorium on lion hunting that took place in the study area
between 2005 and 2008, and hyaenas are very rarely targeted by
Impalas are more abundant than greater kudus and sable ante-
lopes in our study area (ca. 8800 impalas, 4700 greater kudus and
1200 sable antelopes), but they represent less than 5% of lions’diet
(Loveridge et al. 2007). The four-fold difference in abundance
between greater kudu and sable antelope is reﬂected in the diet of
lions, with greater kudu representing up to 24% and sable antelope
less than 5% (Loveridge et al. 2007). This is consistent with other
African systems where lions prey upon greater kudu and sable
antelope in accordance with the abundance of these prey species,
whereas impala is avoided (reviewed in Hayward & Kerley 2005).
However, in the nearby Kruger National Park, South Africa, 61.3% of
encounters between impala and lion lead to hunts (Funston et al.
2001). Therefore, despite impala not being a preferred prey species
for lion, impalas probably do not consider lions to be a low-level
threat. Hyaenas show no prey species preference (Hayward 2006).
Impala and greater kudu, which are among the commonest prey
species consumed by hyaenas, are taken in proportion to their
relative abundance. The sable antelope, however, is more likely to be
avoided. In our study area, greater kudu, impala and sable antelope
represent 8e12 %, 5e10% and less than 5% of hyaenas’diet, respec-
tively (Drouet-Hoguet 2007). However, about 80% of hyaenas’food is
acquired by scavenging (Drouet-Hoguet 2007), whereas about 16% is
obtained by hunting. Impalas represent 56% of the prey species
hunted, and greater kudus represent 19%, whereas sable antelopes
are not hunted (Drouet-Hoguet 2007). For the secondary large
carnivore species of the study area (i.e. leopard, Panthera pardus,wild
dog, Lycaon pictus, cheetah, Acinonyx jubatus), there is no equivalent
information on diet composition. Nevertheless, reviews on these
large carnivore diets (Hayward et al. 2006a, b, c)indicatethat
leopards preferentially prey upon impala, consume greater kudu in
proportion to their relative abundance, and tend to avoid sable
antelope. Wild dogs preferentially prey upon greater kudu and
impala (see Rasmussen 2009 for our study area), but they consume
sable antelope in proportion to their relative abundance. Cheetahs
preferentially prey upon impala, consume greater kudu in
accordance with their relative abundance and tend to avoid sable
antelope. Overall, these diet studies suggest that greater kudu, and
impala to a lesser extent, are more vulnerable to their natural
predators than are sable antelope.
The magnitude of behavioural adjustments in hunted
populations is likely to increase with hunting pressure (Caro 1999b;
Matson et al. 2005). During road counts carried out in HA in the late
dry season in 2007 and 2008, we encountered four groups of impala/
100 km, three groups of greater kudu/100 km and 1.4 groups of sable
antelope/100 km. In the meantime, the realized hunting quotas (i.e.
actual numbers of animals shot) were 1.2 times higher for impala
than for greater kudu, and four times higher for impala than for sable
antelope (ZPWMA, unpublished data). Therefore, considering the
dilution effect, the human hunting risk perceived by groups of
ungulates was comparable among the three species. Thus, the three
species differed mainly in their exposure to natural predators, and
we therefore expected impala and greater kudu to be less prone to
shift their visits at waterholes during night-time than sable
antelopes when hunted because of their higher exposure to natural
predators then. We observed 226 groups of impala (131 in nonHA, 95
in HA), 188 groups of greater kudu (149 in nonHA, 39 in HA) and 63
groups of sable antelope (41 in nonHA, 22 in HA).
Drinking temporal niche at waterholes was approximated by
the arrival time of groups at waterholes. We plotted a frequency
distribution of group observations at waterholes by hour using
Oriana 2.0 software for circular data (Fisher 1993). We used the
group as the statistical unit. Mean arrival times per species were
compared with the WatsoneWilliams test for circular means
(Fisher 1993). To test whether waterholes were more frequently
visited at night in HA than in nonHA, we compared the respective
temporal distributions of visits at waterholes between the two
zones. To determine the expected number of visits in HA during
daytime (0600e1900 hours) and at night (1900e0600 hours), we
used the distribution of visits at waterholes in nonHA where there
was no sport hunting. We then compared the expected distribution
with the observed distribution in HA using a contingency table
(following Zar 1984), with the Pearson’s chi-square test with Yates’
continuity correction, and the Fisher’s exact test for frequencies
that were smaller than ﬁve.
For each species, we investigated the effects of zone (nonHA or
HA), group size and presence/absence of young on the probability
of arriving at a waterhole at night with a logistic regression model.
The respective effects of group size and presence of young,
however, may be confounded for the three species because these
two factors were correlated (Wilcoxon signed-ranks test with
continuity correction: impala: V¼20301, r¼0.10, N¼201,
P<0.001, N¼201; greater kudu: V¼17020, r¼0.19, N¼184,
P<0.001; sable antelope: V¼177 0 , r¼0.64, N¼59, P<0.001).
We included all possible interactions between ﬁxed effects but
none was statistically signiﬁcant. To compare the probabilities of
visiting waterholes at night between species, we used a logistic
regression with species and zone as factors, and their interaction.
For all logistic models, we used the Akaike Information Criterion
(AIC) to select the best model (Burnham & Anderson 2002). The
model with the lowest AIC value (i.e. the best compromise between
accuracy and precision) was retained. When
AIC between two
models was less than two, we selected the simplest model
according to the parsimony rules (Burnham & Anderson 2002). All
statistical analyses were generalized linear mixed models, with
year and waterhole ﬁtted as random effects, using R lme4 package
Groups of all three species visited waterholes preferentially
during daytime, generally avoiding night hours (Fig. 2). However,
avoidance of waterholes during night-time was less marked in HA,
where proportionally more groups came at night, than in nonHA.
In nonHA, 6% of groups of impala, less than 1% of groups of greater
kudu and 10% of groups of sable antelope visited waterholes at night.
Conversely in HA, 24% of groups of impala, 20% of groups of greater
kudu and 50% of groups of sable antelope visited waterholes at night.
These differences between day and night were signiﬁcant for the
three species (impala, Pearson’s chi-square test:
P<0.001; greater kudu, Fisher’s exact test: N¼188, P¼0.001; sable
antelope, Fisher’s exact test: N¼63, P¼0.001; Fig. 3).
Average arrival times at waterholes did not differ between
nonHA and HA for impala (WatsoneWilliams test: F
P¼0.2; Fig. 2) or greater kudu (F
¼0.15, P¼0.7; Fig. 2). Sable
antelope arrived at waterholes later in HA than in nonHA
(WatsoneWilliams test: F
¼18.68, P<0.0001; average
arrival time in nonHA ¼1140 hours; average arrival time in
HA ¼1856 hours). Circular variances associated with average
arrival times were higher in HA than in nonHA (circular
W.-G. Crosmary et al. / Animal Behaviour 83 (2012) 145e153 149
standard deviation: impala: nonHA ¼65
; greater kudu:
; sable antelope: nonHA ¼56
The distribution of arrival times was thus more scattered in HA than
in nonHA. Temporal distributions of arrival times were more
concentrated during daytime in nonHA than in HA, particularly
during the hottest hours of the day (between 1000 and 1600 hours):
for impala, 49.6% of groups were observed during the hottest hours
of the day in nonHA, as opposed to 35.9% in HA; for greater kudu, it
was 68.5% in nonHA versus 31.5% in HA; and for sable antelope, 51%
in nonHA versus 30% in HA.
Logistic regressions conﬁrmed that the probability of visiting
waterholes at night was higher in HA than in nonHA for the three
species (estimate SE: impala: 2.3 1.0; greater kudu: 3.7 1.1;
sable antelope: 7.0 2.2; Table 2). Moreover, the probability of
visiting waterholes at night was lower for groups of impala with
young (estimate SE ¼2.6 1.2), but not for groups of greater
kudu with young (Table 2). For sable antelope, although the
presence of young was retained in the selected model (Table 2), its
effect on the probability of visiting waterholes at night was highly
variable (estimate SE ¼20.4 2322.1). Group size was not
retained in the selected models and, therefore, did not affect the
probability of visiting waterholes at night for any of the three
species (Table 2).
Sable antelope had a higher probability of visiting waterholes at
night compared to impala and greater kudu in HA, but not in nonHA
(Fig. 4). In HA, impala and kudu had a similar probability of visiting
waterholes during night-time, whereas in nonHA, greater kudu had
a lower chance of visiting waterholes at night than did the two
Prey may shift their niche along the time axis to reduce temporal
overlap with predator activity (reviewed in: Kronfeld-Schor & Dayan
2003; Caro 2005). Human hunters, similarly to natural predators,
may also induce a temporal adjustment of the niche in hunted
populations (e.g. Kilgo et al. 1998). Here, we observed African
ungulates’use of waterholes in semiarid savannas to examine how
% Groups observed at waterholes
Impala Kudu Sable
nonHA nonHAHA HAHA
Figure 3. Effect of sport hunting on attendance of groups of impala, greater kudu and
sable antelope at waterholes during daytime (grey bars; 0600e1900 hours) versus
night-time (black bars; 1900e0600 hours) in 2007e2008. Zones with natural
predation and no sport hunting (nonHA, Hwange National Park, Zimbabwe) versus
natural predation and sport hunting (HA) are compared. Numbers above bars indicate
number of groups present at waterholes in each zone. Statistical tests are from
comparisons of frequencies within contingency tables, but results are presented as
percentages for clarity. *P<0.001.
Summary of the logistic regression modelling the probability of visiting waterholes
at night as a function of zone (hunting area, HA, versus nonhunting area, nonHA),
group size and status (i.e. presence/absence of young) for groups of impala, greater
kudu and sable antelope in Hwange National Park (Zimbabwe) and the surrounding
hunting areas in 2007 and 2008
Null 22 127.1 7.8
Zone 24 125.2 5.9
Status 24 121.6 2.3
Group size 24 128.7 9.4
Zone Dstatus 25 119.3 0
Zone þgroup size 25 127 7.7
Group size þstatus 25 123.3 4
Zone þgroup size þstatus 26 120.5 1.2
Null 22 55.4 5.4
Zone 24 51 1
Status 24 56.5 6.5
Group size 24 55.6 5.6
Zone þstatus 25 50 0
Zone þgroup size 25 50.8 0.8
Group size þstatus 25 57.5 7.5
Zone þgroup size þstatus 26 51.6 1.6
Null 15 58 21
Zone 16 49.7 12.7
Status 16 48.1 11.1
Group size 16 55.2 18.2
Zone Dstatus 17 37 0
Zone þgroup size 17 42.7 5.7
Group size þstatus 17 50.6 13.6
Zone þgroup size þstatus 18 38.9 1.9
The selected models (lowest AIC) are shown in bold font.
Number of parameters. It includes the random factor ‘waterholes’.
Probability of visiting a waterhole at night
Figure 4. Probabilities of visiting waterholes at night (1900e0600 hours) for impala,
greater kudu and sable antelope in 2007 and 2008. Zones with natural predation and
no sport hunting (nonHA, Hwange National Park, Zimbabwe) versus natural predation
and sport hunting (HA) are compared. Estimates from logistic regression models and
their conﬁdence intervals (backtransformed) are shown. Estimates with different
letters were signiﬁcantly different at P<0.05.
W.-G. Crosmary et al. / Animal Behaviour 83 (2012) 145e153150
they adjust their temporal niche in response to both their risk of
natural predation by large carnivores mostly at night and their risk of
being hunted by human hunters during the day. Groups of impala,
greater kudu and sable antelope visited waterholes more often at
night in hunting areas than in nonhunting areas (Hwange National
Park), where visits at waterholes were almost exclusively diurnal.
The temporal pattern of access to water was concentrated around
the hottest hours of the day in the protected area, particularly for
greater kudu, while it was scattered throughout the day in hunting
areas. More interestingly, we observed that groups of impala and
greater kudu were less prone to switch their access to water towards
night-time in hunting areas compared to groups of sable antelope,
although they were exposed to similar hunting risk.
The diurnal use of waterholes by ungulates in African savannas
is largely dictated by the circadian rhythms and physiology of these
ungulates (Finch 1972; Ayeni 1977; Valeix et al. 2007b). It may also
be a tactic to decrease the risk of encountering the main natural
predators of these ecosystems (Valeix et al. 2009b), which visit
waterholes essentially at night to drink or to hunt their prey. The
activity of ungulates at waterholes was almost exclusively diurnal
in the protected area, similar to that observed in other protected
areas (Hwange National Park: Weir & Davison 1965; Valeix et al.
2007a; Tsavo National Park, Kenya: Ayeni 1975), whereas a large
proportion of the groups (24% for impala, 20% for greater kudu and
50% for sable antelope) visited waterholes at night in hunting areas.
We hypothesize that this signiﬁcant temporal shift towards
nocturnal use of waterholes in hunting areas is a behavioural
adjustment of ungulates to the risk of being hunted by
humans during daytime. The increase of nocturnal activities
(e.g. movements, foraging, social interactions) as a way to escape
human disturbance occurring during daytime has been largely
documented in birds (e.g. brent goose, Branta bernicla bernicla:
Riddington et al. 1996; sanderling, Calidris alba:Burger & Gochfeld
1991 ), but less so in ungulates (but see: white-tailed deer, Odocoi-
leus virginianus:Kilgo et al. 1998; red deer, Cervus elaphus:Sunde
et al. 2009).
The signiﬁcant temporal shift towards nocturnal use of
waterholes exposes ungulates to their natural predators. Hence,
ungulates in the hunting areas face a trade-off between the risk of
natural predation and the risk of being hunted. Such trade-off is
poorly understood because most studies on hunting are carried out
on ungulates of temperate ecosystems (moose, Alces alces, elk,
Cervus canadensis:Altmann 1958; roe deer, Capreolus capreolus:de
Boer et al. 2004; Benhaiem et al. 2008; fallow deer, Dama dama:de
Boer et al. 2004; caribou and reindeer, R. tarandus: reviewed in
Reimers & Colman 2006), where the diversity and densities of
predators have been largely reduced by human activities (Berger
1999; Andersen et al. 2006). In comparison, African savannas still
host a rich guild of large carnivores. Our results showed greater
circular variances associated with mean arrival time at waterholes
for the three species in hunting areas, whereas the great majorityof
observations were made during the hottest hours of the day
(1000e1600 hours) in the protected area. In African savannas, the
hottest hours of the day coincide with the period when large
carnivores are less active (Kruuk 1972; Schaller 1972).
Consequently, ungulates drinking at waterholes during this period
may beneﬁt from a reduced risk of natural predation. Conversely,
visits of ungulates at waterholes were more randomly distributed
over the day in hunting areas. As a consequence, their whereabouts
were less predictable to human hunters. Random antipredator
behaviour of prey may also be an efﬁcient tactic against predators;
for instance, Scannell et al. (2010) recently demonstrated that prey
that scan at random are more difﬁcult to catch by predators. The
decrease in the predictability of the temporal pattern of visits at
waterholes may be an alternative tactic adopted by ungulates to
partly reduce hunting risk during daytime without completely
shifting towards night-time, thus limiting increased exposure to
The magnitude of the temporal shift in the use of waterholes
was expected to vary according to species vulnerability to natural
predation and exposure to hunting. Because the three species
differed mainly in their exposure to natural predation, we predicted
that greater kudu, and to a lesser extent impala, should be less
prone to shift their visits at waterholes during night-time than
sable antelope. Interestingly, sable antelope showed the greatest
shift to night use of waterholes in hunting areas, whereas greater
kudu and impala showed the smallest shift. Such results are
consistent with the observation that greater kudu is the second
most frequent prey of lions (after buffalo) (Loveridge et al. 2007),
and that greater kudu and impala represent a greater share of
hyaena diet in Hwange National Park than do sable antelope
(Drouet-Hoguet 2007), while sable antelope has never been iden-
tiﬁed as a main prey for one of the large carnivores in Hwange
National Park. Previous studies in this ecosystem have shown that
greater kudus are particularly responsive to the immediate risk of
predation by lions (Valeix et al. 2009b; Périquet et al. 2010). Hence,
it is not surprising that greater kudu and impala showed less
shifting in their temporal niche at waterholes towards the night in
hunting areas, and that sable antelope, which appears less sensitive
to natural predation, shifted the most to night use of waterholes in
hunting areas. Our results indicate that the higher the risk of
natural predation, the lower the probability of shifting to night use
of waterholes to avoid human hunters.
Note that we have limited our conclusions to nights during
which visibility was good, since we carried out all of our observa-
tions during full moon nights. Lower visibility condition during
other nights may change prey perception of natural predation risk.
Ayeni (1975) observed that many herbivores in Tsavo National Park,
Kenya, drank later into the night on moonlit nights than on
moonless nights, probably because it is then much more difﬁcult to
detect approaching predators. Under moonless or cloudy nights,
the situation of ungulates would be therefore even more compli-
cated than during full moon nights because natural predation risk is
probably enhanced because of lower visibility.
The propensity to shift to nocturnal use of waterholes was also
inﬂuenced by presence of young in impala, and in sable antelope to
a lesser extent, with the probability of coming to waterholes at
night being lower for groups with young than for groups without
young. Because young are highly vulnerable to predators, groups
with young normally adjust their behaviour to compensate for
increased vulnerability (increased vigilance: Burger & Gochfeld
1994; avoidance of risky areas and risky hours: Hamel & Côté
2007). In greater kudu, however, the presence of young in groups
had no signiﬁcant inﬂuence. Since greater kudu is the most
vulnerable prey species, it may not be less dangerous for groups
without young to come at night at waterholes than for groups with
young. This however remains to be investigated. We found no effect
of group size on the probability of coming to waterholes at night.
Perceived predation risk, however, was expected to be lower in
larger groups because more individuals are present to scan the
surroundings for approaching predators (‘many-eyes effect’,
Pulliam 1973) and because of the dilution effect of large groups
(Hamilton 1971; Dehn 1990). At night, because of the lower light
conditions (even under full moon light), more individuals may not
sufﬁciently offset the increased risk of predation in large groups
because of the presence of young. Moreover, the risk for groups
with young to be targeted by natural predators might be higher
than the risk to be targeted by human hunters, the latter usually
seeking trophy males, which are often solitary or found in small
W.-G. Crosmary et al. / Animal Behaviour 83 (2012) 145e153 151
We cannot fully exclude other potential alternative explanations
to the nocturnal shift observed in hunting areas (e.g. differences in
environmental conditions, tourism and natural predation). The
studied hunting areas and Hwange National Park were adjacent and
showed similar soil composition, vegetation structure, temperature
and rainfall characteristics (Ganzin et al. 2008; Peace Parks
Foundation 2009). Moreover, the diurnal preferential use of water-
holes by African ungulates has been shown in other ecosystems
(e.g. Tsavo National Park, Kenya: Ayeni 1975; Etosha National Park,
Namibia: Preez & Grobler 1977). Therefore, differences in environ-
mental conditions are unlikely to account for the difference in the
temporal pattern of visits at waterholes observed in our study
between Hwange National Park and the adjacent peripheral hunting
areas. Tourism was generally low in Zimbabwe during our study
(2007e2008) because of the economic and political crisis, and there
were very few tourists in Hwange National Park and the peripheral
hunting areas. Finally, densities of natural predators, proportion of
waterholes visited by these natural predators, and the temporal
pattern of these visits were similar between Hwange National Park
and the peripheral hunting areas (Table 1,Supplementary Material,
Fig. S1). Hence, we interpret the shift to night use of waterholes in
hunting areas as a response to the risk of being hunted by human
hunters during the day, which is the main difference between the
national park and the peripheral hunting areas.
Our study illustrates how sport hunting inﬂuences the ecology
of ungulates in a system where ungulates have to face a trade-off
between the risk of natural predation by carnivores and the risk
of being hunted, the former being higher at night and the latter
higher during the day. To our knowledge, this is one of the ﬁrst
attempts to explore such trade-off, and one of the rare studies that
contrasts different species. The three study species were charac-
terized by a gradient in the risk of natural predation, leading to
a gradient in the shift made to night use of waterholes in hunting
areas. The species more often preyed upon (i.e. greater kudu) per-
formed the shift of the lowest magnitude, whereas the species less
preyed upon (i.e. sable antelope) performed the shift of the highest
magnitude. Because our study did not measure the potential costs
associated with the shift of visits to waterholes towards night-time
in hunting areas, we cannot assume that this behavioural response
to hunting risk has any impact on individual ﬁtness. To better
understand how sport hunting and induced behavioural adjust-
ments inﬂuence ungulate population dynamics, there is a need to
investigate whether the differences that we observed in ungulate
behaviour between hunting-free areas and hunting areas actually
lead to measurable costs in terms of physiology and survival
(e.g. greater exposure to natural predators, alteration of thermo-
regulation, or changes in time budget).
We acknowledge the Director General of the Zimbabwe Parks
and Wildlife Management Authority for providing the opportunity
to carry out this research and for permission to publish this
manuscript. W.G.C. was supported by scholarships from the French
Agence Nationale de Recherche and the Natural Sciences and
Engineering Research Council of Canada Discovery Grant (to S.D.C.),
and by the Soutien aux cotutelles internationales de thèses de
l’Université Claude Bernard Lyon 1. This research was carried out
within the framework of the HERD project (Hwange Environmental
Research Development), as well as with the Research Platform
Production and Conservation in Partnership RP-PCP, funded by
the French Ministère des Affaires Etrangères, Ambassade de
France au Zimbabwe, CIRAD, CNRS and the IFB Global Change and
Biodiversity. We thank S. Le Bel, CIRAD representative in Zimbabwe,
and S. Chamaillé-Jammes and T. Tarakini for insightful comments
on previous drafts of the manuscript. We are indebted to Wildlife
Environment Zimbabwe for providing waterhole census data, as
well as rangers, students and volunteers who participated in ﬁeld
work. Finally, we sincerely thank Tim Caro and two anonymous
referees for their very fruitful comments on the manuscript.
Supplementary material for this article is available, in the online
version, at doi:10.1016/j.anbehav.2011.10.019.
Altmann, M. 1958. The ﬂight distance in free ranging big game. Journal of Wildlife
Andersen,R., Linnell, J. D. C. & Solberg, E. J. 2006. The future roleof large carnivores in
terrestrial trophic interactions: the northern temperate view. In: Large Herbivore
Ecology, Ecosystem Dynamics and Conservation (Ed. by K. Danell, P. Duncan,
R. Bergström & J. Pastor), pp. 413e424. Cambridge: Cambridge University Press.
Ayeni, J. S. O. 1975. Utilization of waterholes in Tsavo National Park (East). East
African Wildlife Journal,13, 305e323.
Ayeni, J. S. O. 1977. Waterholes in Tsavo National Park, Kenya. Journal of Applied
Beale, C. M. & Monaghan, P. 2004. Behavioural responses to human disturbance:
a matter of choice? Animal Behaviour,68, 1065e1069.
Benhaiem, S., Delon, M., Lourtet, B., Cargnelutti, B., Aulagnier, S.,
Hewison, A. J. M., Morellet, N. & Verheyden, H. 2008. Hunting increases
vigilance levels in roe deer and modiﬁes feeding site selection. Animal Behav-
Berger, J. 1991. Pregnancy incentive and predation constraints in habitat shifts:
experimental and ﬁeld evidence for wild bighorn sheep. Animal Behaviour,41,
Berger, J. 1999. Anthropogenic extinction of top carnivores and interspeciﬁc animal
behaviour: implications of the rapid decoupling of a web involving wolves,
bears, moose and ravens. Proceedings of the Royal Society B,266, 2261e2267.
de Boer, H. Y., van Breukelen, L., Hootsmans, M. J. M. & van Wieren, S. E. 2004.
Flight distance in roe deer (Capreolus capreolus) and fallow deer (Dama dama)
as related to hunting and other factors. Wildlife Biology,10,35e41.
Brown, J. S., Laundré, J. W. & Gurung, M. 1999. The ecology of fear: optimal
foraging, game theory, and trophic interactions. Journal of Mammalogy,80,
Buckland, S. T., Anderson, D. R., Burnham, K. P., Laake, J. L., Borchers, D. L. &
Thomas, L. 2001. Introduction to Distance Sampling. Estimating Abundance of
Biological Populations. Oxford: Oxford University Press.
Burger, J. & Gochfeld, M. 1991. Human activity inﬂuence and nocturnal foraging of
sanderlings (Calidris alba). Condor,93, 259e265.
Burger, J. & Gochfeld, M. 1994. Vigilance in African mammals: differences among
mothers, other females and males. Behaviour,131,154e274.
Burnham, K. P. & Anderson, D. R. 2002. Model Selection and Multimodel Inference:
a Practical Information-theoretical Approach. New York: Springer.
Caro, T. 1999a. The behavioureconservation interface. Trends in Ecology & Evolution,
Caro, T. 1999b. Demography and behaviour of African mammals subject to
exploitation. Biological Conservation,91,91e97.
Caro, T. 2005. Antipredator Defenses in Birds and Mammals. Chicago: Chicago
Chamaillé-Jammes, S., Fritz, H. & Murindagomo, F. 2006. Spatial patterns of the
NDVIerainfall relationship at the seasonal and interannual time scales in an
African savanna. International Journal of Remote Sensing,27, 5185e5200.
Creel, S. & Christianson, D. 2008. Relationships between direct predation and risk
effects. Trends in Ecology & Evolution,23,194e201.
Creel, S., Winnie, J. A., Maxwell, B., Hamlin, K. & Creel, M. 2005. Elk alter habitat
selection as an antipredator response to wolves. Ecology,86, 3387e3397.
Creel, S., Christianson, D., Liley, S. & Winnie, J. 2007. Predation risk affects
reproductive physiology and demography of elk. Science,315, 960.
Cumming, D. H. M. 1989. Commercial and safari hunting in Zimbabwe. In: Wildlife
Production Systems: Economic Utilisation of Wildlife (Ed.by R. J. Hudson, K. R. Drew
& L. M. Bodkin), pp. 147e169. Cambridge: Cambridge University Press.
Dehn, M. M. 1990. Vigilance for predators: detection and dilution effects. Behavioral
Ecology and Sociobiology,26,337e342.
Drouet-Hoguet, N. 2007. Inﬂuenc e des activités anthropogéniques sur le régime
alimentaire et la réponse numérique de la hyène tachetée en savane arborée dys-
trophique dominée par l’éléphant. Ph.D. thesis, University Claude Bernard Lyon 1.
Dunham, K. M. 2002. Aerial Census of Elephants and Other Large Herbivores in North-
west Matabeleland, Zimbabwe: 2001. WWFeSARPO Occasional Paper Series.
Elliot, N. 2007. A comparison of two methods for estimating population densities of
lion (Panthera leo) and other large carnivores using spoor transects and call-up
stations. M.S. thesis, Oxford University.
Fenn, M. G. P. & Macdonald, D. W.1995. Use of middens by red foxes: risk reverses
rhythms of rats. Journal of Mammalogy,76,130e136.
W.-G. Crosmary et al. / Animal Behaviour 83 (2012) 145e153152
Finch, V. A. 1972. The effects of solar radiation, of temperature regulation, and heat
balance in 2 East African antelopes, the eland and the hartebeest. American
Journal of Physiology,222,1374e1379.
Fisher, N. I. 1993. Statistical Analysis of Circular Data. Cambridge: Cambridge
Funston, P. J., Mills, M. G. M. & Biggs, H. C. 2001. Factors affecting the hunting
success of male and female lions in the Kruger National Park. Journal of Zoology,
Ganzin, N., Crosmary, W.-G. & Fritz, H. 2008. A Simpliﬁed Vegetation Map of
Hwange National Park and Surrounding Areas Using Landsat ETMþSatellite
Imagery. Hwange Environmental Research and Development Programme, Report
2008. Harare: Zimbabwe Parks and Wildlife Management Authority.
Geist, V. 1970. A behavioral approach to the management of wild ungulates. In:
Scientiﬁc Management of Animal and Plant Communities for Conservation.
Eleventh Symposium, British Ecological Society (Ed. by E. Duffey & A. S. Watt), pp.
413e424. Oxford: Blackwell Scientiﬁc.
Gill, J. A., Norris, K. & Sutherland, W. J. 2001. Why behavioural responses may not
reﬂect the population consequences of human disturbance. Biological
Hamel, S. & Côté, S. D. 2007. Habitat use patterns in relation to escape terrain: are
alpine ungulate females trading-off better foraging sites for safety? Canadian
Journal of Zoology,85, 933e943.
Hamilton, W. D. 1971. Geometry for the selﬁsh herd. Journal of Theoretical Biology,
Harrison, D. L. & Bates, P. J. J. 1991. The Mammals of Arabia. 2nd edn. Seven Oaks:
Harrison Zoological Museum.
Hayward, M. W. 2006. Prey preferences of the spotted hyaena (Crocuta crocuta) and
degree of dietary overlap with the lion (Panthera leo). Journal of Zoology,270,
Hayward, M. W. & Kerley, G. I. H. 2005. Prey preferences of the lion (Panthera leo).
Journal of Zoology,267, 309e322.
Hayward, M. W., Henschel, P., O’Brien, J., Hofmeyr, M., Balme, G. & Kerley, G. I. H.
2006a. Prey preferences of the leopard (Panthera pardus). Journal of Zoology,
Hayward, M. W., O’brien, J., Hofmeyr, M. & Kerley, G. I. H. 2006b. 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,270,615e627.
Hayward, M. W., O’brien, J., Hofmeyr, M. & Kerley, G. I. H. 2006c. Prey preferences
of the African wild dog Lycaon pictus (Canidae: Carnivora): ecological require-
ments for conservation. Journal of Mammalogy,87,1122e113 1.
Kilgo, J. C., Labisky, R. F. & Fritzen, D. E. 1998. Inﬂuences of hunting on the
behavior of white-tailed deer: implications for conservation of the Florida
panther. Conservation Biology,12,1359e1364.
Kotler, B. P., Brown, J. S. & Hasson, O. 1991. Factors affecting gerbil foraging
behavior and rates of owl predation. Ecology,72, 2249e2260.
Kronfeld-Schor, N. & Dayan, T. 2003. Partitioning of time as an ecological resource.
Annual Review of Ecology, Evolution, and Systematics,34,153e181.
Kufeld, R., Bowden, D. & Schrupp, D. 1988. Inﬂuence of hunting on movements of
female mule deer. Journal of Range Management,41,70e72.
Kruuk, H. 1972. The Spotted Hyena: a Study of Predation and Social Behaviour.
Chicago: Chicago University Press.
Lima, S. L. 1998. Nonlethal effects in the ecology of predatoreprey interactions.
Lima, S. L. & Dill, L. M. 1990. Behavioural decisions made under the risk of
predation: a review and synthesis. Canadian Journal of Zoology,68,619e640.
Lindsey, P. A., Roulet, P. A. & Romañach, S. S. 2007. Economic and conservation
signiﬁcance of the trophy hunting industry in sub-Saharan Africa. Biological
Loveridge, A. J., Davidson, Z., Hunt, J. E., Valeix, M., Elliot, N. & Stapelkamp, B.
2007. Hwange Lion Project Annual Report 2007. Harare: Zimbabwe Parks and
Wildlife Management Authority.
Madsen, J. & Fox, A. D. 1995. Impacts of hunting on waterbirds: a review. Wildlife
Matson, T., Goldizen, A. & Putland, D. 2005. Factors affecting the vigilance and
ﬂight behaviour of impalas: research article. South African Journal of Wildlife
Peace Parks Foundation 2009. Integrated Development Plan: Zimbabwean Compo-
nent of the KAZA TFCA. Harare: Zimbabwe Parks and Wildlife Management
&Fritz,H.2010. Individual vigilance of African herbivores while drinking:
the role of immediate predation risk and context. Animal Behaviour,79,
Preez, J. S. & Grobler, I. D. 1977. Drinking times and behaviour at waterholes of
some game species in the Etosha National Park. Madoqua,10,61e69.
Pulliam, H. R. 1973. On the advantages of ﬂocking. Journal of Theoretical Biology,38,
Rasmussen, G. S. A. 2009. Anthropogenic factors inﬂuencing biological processes of
the painted dog Lycaon pictus. Ph.D. thesis, Oxford University.
Reimers, E. & Colman, J. E. 2006. Reindeer and caribou (Rangifer tarandus)
response towards human activities. Rangifer,26,5
Riddington, R., Hassall, M., Lane, S. J., Turner, P. A. & Walters, R. 1996. The impact
of disturbance on the behaviour and energy budgets of brent geese (Branta b.
bernicla). Bird Study,43, 269e279.
Ripple, W. & Beschta, R. 2004. Wolves and the ecology of fear: can predation risk
structure ecosystems? BioScience,54, 755e766.
Rogers, C. M. L. 1993. A Woody Vegetation Survey of Hwange National Park. Harare:
Department of National Parks and Wildlife Management.
Roth, T. C. & Lima, S. L. 2007. The predatory behavior of wintering Accipiter hawks:
temporal patterns in activity of predators and prey. Oecologia,152,169e178.
Salnicki, J., Teichmann, M., Wilson, V. J. & Murindagomo, F. 2001. Spotted
hyaenas Crocuta crocuta prey on new-born elephant calves in Hwange National
Park, Zimbabwe. Koedoe,44,79e83.
Scannell, J., Roberts, G. & Lazarus, J. 2010. Prey scan at random to evade observant
predators. Proceedings of the Royal Society B,268,541e547.
Schaller, G. B. 1972. The Serengeti lion: a Study of PredatorePrey Relations. Chicago:
Chicago University Press.
Setsaas, T., Holmern, T., Mwakalebe, G., Stokke, S. & Røskaft, E. 2007. How does
human exploitation affect impala populations in protected and partially
protected areas? A case study from the Serengeti Ecosystem, Tanzania. Biolog-
ical Conservation,136, 563e570.
Stankowich, T. 2008. Ungulate ﬂight responses to human disturbance: a review
and meta-analysis. Biological Conservation,141,2159e2173.
Stankowich,T.&Blumstein,D.T.2005. Fear in animals: a meta-analysis and
review of risk assess ment. Proceedings of the Royal Society B,272,
Sunde, P., Olesen, C. R., Madsen, T. L. & Haugaard, L. 2009. Behavioural responses
of GPS-collared female red deer (Cervus elaphus) to driven hunts. Wildlife
Swenson, J. E. 1982. Effects of hunting on habitat use by mule deer on mixed-grass
prairie in Montana. Wildlife Society Bulletin,10,115e120.
Valeix, M., Chamaillé-Jammes, S. & Fritz, H. 2007a. Interference competition and
temporal niche shifts: elephants and herbivore communities at waterholes.
Valeix, M., Fritz, H., Matsika, R., Matsvimbo, F. & Madzikanda, H. 2007b. The role of
water abundance, thermoregulation, perceived predation risk and interference
competition in water access by African herbivores. African Journal of Ecology,46,
Valeix, M., Fritz, H., Loveridge, A. J., Davidson, Z., Hunt, J. E., Murindagomo, F. &
Macdonald, D. W. 2009a. Does the risk of encountering lions inﬂuence African
herbivore behaviour at waterholes? Behavioral Ecology and Sociobiology,63,
Valeix, M., Loveridge, A., Chamaillé-Jammes, S., Davidson, Z., Murindagomo, F.,
Fritz, H. & Macdonald, D. W. 2009b. Behavioral adjustments of African
herbivores to predation risk by lions: spatiotemporal variations inﬂuence
habitat use. Ecology,90,23e30.
Valeix, M., Chamaillé-Jammes, S., Loveridge, A. J., Davidson, Z., Hunt, J. E.,
Madzikanda, H. & Macdonald, D. W. 2011. Understanding patch departure rules
for large carnivores: lion movements support a patch-disturbance hypothesis.
American Naturalist, doi:10.1086/660824, published online 27 June 2011.
Weir, J. & Davison, E. 1965. Daily occurrence of African game animals at waterholes
during dry weather. African Zoology,1,353e368.
Zar, J. H. 1984. Biostatistical Analysis. London: Prentice Hall.
W.-G. Crosmary et al. / Animal Behaviour 83 (2012) 145e153 153