| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Behavioral Ecology Vol. 11 No. 6: 624-632
© 2000 International Society for Behavioral Ecology
Living with the enemy: avoidance of hyenas and lions by cheetahs in the Serengeti
Institute of Zoology, Zoological Society of London, Regent's Park, London, NW1 4RY, England
Address correspondence to S. M. Durant. E-mail: s.durant{at}ucl.ac.uk .
Received 8 June 1998; revised 17 December 1999; accepted 24 January 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Predator avoidance is likely to play a strong role in structuring species communities, even where actual mortality due to predation is low. In such systems, mortality may be low because predator avoidance is effective, and if the threat of predation is lifted then entire community structures may be altered. Where competition is intense, then competitor avoidance may have a similar impact on communities. Avoidance behaviors have been documented for a wide range of species, but this is the first attempt to document avoidance behavior within a large carnivore community. Audio playback techniques are used to examine the risk perceived by cheetahs from their two main competitors that are also their main predators, lions and hyenas. The results from these experiments show that cheetahs actively moved away from lion and hyena playback experiments, compared with dummy playbacks where no sound was played. Cheetahs showed no differences in their responses to playbacks dependent on their sex or reproductive status, suggesting they were responding principally to a competition rather than a predation threat. However, cheetahs were much less likely to hunt after competitor playbacks than after dummy playbacks, and this resulted in a lower kill rate after competitor playbacks, demonstrating that the perceived presence of competitors had a noticeable impact on the foraging rate of cheetahs. Furthermore, while cheetahs moved just as far following lion playbacks as after hyena playbacks, they spent significantly more time looking at the loudspeaker and were less likely to make a kill after lion playbacks, suggesting that cheetahs perceive lions to be a greater threat than hyenas.
Key words: anti-predator behavior, carnivores, competition, foraging strategy, kleptoparasitism, playback experiments, predator avoidance, predation risk.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Predator avoidance, whereby prey avoid encounters with predators, is one means by which prey are able to reduce the chance of predation. An avoidance behavior can be defined as any behavioral strategy that enhances the survivorship of prey by reducing the probability that they occur within the foraging range of potential predators (Brodie et al., 1991
). Predator
avoidance is likely to play a strong role in structuring species communities,
even where actual mortality due to predation is low
(Lima and Dill, 1990;
Turner and Mittelbach, 1990).
In such circumstances, mortality may be low precisely because predator
avoidance is effective and, if the threat of predation is lifted, then entire
community structures may be altered. Avoidance of predation has attracted
increasing attention over recent years, however predation is not the only
negative interspecific interaction which a species may want to avoid;
competitor avoidance behaviors may also be important as a mechanism for
shaping communities. Competitor avoidance can be defined similarly to predator
avoidance, except enhancement of survival is through indirect rather than
through direct mechanisms.
Both predator and competitor avoidance are likely to carry costs. For
example, a reduction in activity or an increase in the use of refuges due to a
perceived risk of predation can lead to a reduction in foraging rate
(Kennedy et al., 1994;
Ward et al., 1997). The
relative balance between the costs and benefits of predator avoidance may
differ between different species and between different age and reproductive
classes within species, and can result in different avoidance strategies, even
when the predator is identical (Peckarsky,
1996; Sih, 1992).
In addition, some species may evolve an ability to adapt their avoidance
tactics in response to a perceived predation risk
(Loose and Dawidowicz, 1994;
McIntosh and Townsend, 1994;
Peckarsky, 1996). The
evolution of such a flexible response will depend on the costs associated with
gathering information about a potential predation threat
(Dill, 1986;
Sih, 1987). Selection will
only favor flexibility in predator avoidance tactics when the cost of
gathering information about the predation threat is relatively low, when the
risk of predation fluctuates unpredictably, and when there are reliable cues
for detecting predation risk (Harvell,
1986,
1990). Competitor avoidance is
likely to be subject to weaker selective forces than predator avoidance, since
the cost of not avoiding a competitor will generally be lower than the cost of
not avoiding a predator, as the latter carries a risk of direct mortality.
To date most studies of predator avoidance have concentrated on aquatic and
small mammal communities, while communities of large mammals have been
neglected (but see Bshary and Noe,
1997). This article addresses this gap in our knowledge by
examining avoidance of the two main predators, which are unusually also the
main competitors, of the cheetah (Acinonyx jubatus)—lions
(Panthera leo) and spotted hyenas (Crocuta crocuta). In
cheetahs, offspring survival is strongly affected by lion and hyena predation.
Over 90% of cheetah cubs die before reaching independence, predominantly due
to predation (Laurenson,
1994). Adult cheetahs may also lose their kills to these predators
(Caro, 1994;
Schaller, 1972), and may also
be killed by lions (Durant SM, unpublished data). There are no published
records of adult cheetahs being killed by hyenas, although this possibility
cannot be completely discounted. The impact of lions on cheetah populations
results in a negative relationship between cheetah population size and lion
density both across and within different protected areas
(Durant et al., under review;
Laurenson, 1995b).
Since cheetahs have small jaws and a light build, a mother cannot defend
her cubs or kills against lions and hyenas. However, she can reduce predation
of her cubs by adapting her behavior and adopting an avoidance strategy
(Caro, 1994). Cheetahs make use
of various techniques in order to minimize direct interactions with lions and
hyenas. They reduce visual and audio cues by killing silently by asphyxiation
after a short chase, hunting during the day when many of their competitors are
inactive, and dragging kills immediately into cover to avoid attracting
vultures to carcasses (Caro,
1994). These behaviors minimize contact with competitors and may
reduce rates of kleptoparasitism and cub mortalities. Cheetahs can further
minimize direct contact by making use of any spatial heterogeneity in the
distributions of lions and hyenas and seeking out "competition
refuges"—areas with low densities of their competitors
(Durant, 1998).
The risk of predation and kleptoparasitism to cheetahs is likely to be
influenced by a number of predictable factors, therefore a flexible avoidance
response may have evolved in this species. Cheetahs are more likely to be
noticed by predators when they are active and hunting, when they are also
vulnerable to kleptoparasitism, and so avoidance should be particularly marked
at these times. Moreover, if cub vulnerability is the predominant cause of
avoidance, avoidance should be more marked in females than in males, since the
former are more frequently accompanied by dependent cubs and hence might
perceive predators to be a greater threat. Female cheetahs should also show
greater avoidance when they are accompanied by dependent cubs than when
solitary. Finally, since young cubs are more vulnerable to predation than
older cubs (Caro, 1994), the
strength of avoidance by a mother should depend on the age of her cubs.
Direct observation of causes of mortality of cheetahs are rare
(Caro, 1994), making it
difficult to directly ascertain the extent of the relative threat that hyenas
and lions pose to cheetahs. Moreover, observation of interactions between
cheetahs and their competitors are highly variable in terms of the distance of
nearest approach of the competitors and the time of day at which interactions
occur (Caro, 1994). If more
reproductively successful cheetahs are better at avoiding lions and hyenas
(Durant, 2000), then data
gathered by field observation alone would be biased towards less successful
cheetahs, since these would most often be seen near other competitors. In this
article I employ audio playback techniques to examine avoidance behavior by
cheetahs. Such experiments allow the manipulation and standardization of
naturally rare events.
I tested the following predictions in order to ascertain the extent of avoidance tactics by cheetahs and whether cheetahs adapt avoidance behavior according to the relative risk of predation or competition. First, I tested whether cheetahs avoid hyenas and lions. Second, I tested whether reactions are flexible, and are affected by various cues such as the hunger state of the cheetah in the experiment and the presence of prey. Third, I tested whether responses are dependent on sex or reproductive state. Finally, I tested whether avoidance has a real cost to cheetahs in terms of a reduced foraging rate. Henceforth lions and hyenas are referred to as "competitors," however it should be borne in mind that they are also known predators of cheetah cubs and, in the case of lions, adult cheetahs.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study population
The hypotheses were tested on data from a long-term study population of cheetahs in the Serengeti National Park in Tanzania. The study area covers 2200 km2in the south east of the Serengeti (for full description see Caro, 1994
). Within this
area the habitat ranges from open woodland, dissected with rivers, in the
north and west, through the long grass plains to short grass plains in the
south east. The entire area is scattered with rocky outcrops known as
"kopjes" which provide havens for trees and bushes and are often
the only cover available out on the open plains. The climate is seasonal with
a wet season that starts in November and ends in June
(Sinclair, 1979).
Cheetahs on the Serengeti plains have been studied intensively since the
mid-1970s (Caro, 1994). Each
cheetah can be individually recognized by distinctive spot patterns on its
face and haunches (Caro and Durant,
1991). Cheetahs were located in the early morning and late
afternoon when they are most active, by driving to high points with good
visibility and scanning through binoculars (10 x 50 magnification).
Active cheetahs can be seen up to and occasionally beyond a distance of 3 km.
Upon approach each cheetah was identified and its hunger state estimated by
its belly size according to an increasing 14 point scale
(Caro, 1994). This scale has
proven to be accurate and has a high reliability between different observers
(Caro, 1994). Cubs were aged
when they were first seen, when their age could be estimated to within an
accuracy of 1 month (Caro,
1994).
Playback experiments
Audio playback techniques were used to quantify competitor avoidance
behavior, in a series of controlled experiments. Between May 1993 and
September 1996 lion and hyena vocalizations were played to individual cheetahs
and their reactions recorded. Recordings of lion roars and hyena whoops were
obtained within the study area using either a Panasonic V250, Sony TCD D3 or
Sony TCD D7 digital audio tape recorder linked to a Sennheiser MKH816
microphone. Recordings were made from within 30 m of a single individual
vocalizing lion or hyena. Lion recordings were of adult females, while the sex
of the caller in the hyena recordings was not determined. Playbacks were
played through a Sony XM 4020 Sony amplifier linked to a Sony TCD D3 digital
audio tape recorder and a Martin Audio CT2 studio monitor loudspeaker. The
volume of the playback experiments was standardized across experiments so that
calls were played at natural sound pressure levels. Experiments were conducted
in the early morning between 0630 and 0930 when cheetahs are most active and
are likely to hear lion and hyena vocalizations
(East and Hofer, 1991;
Kruuk, 1972;
Schaller, 1972).
The speaker was placed approximately 200 m from the cheetah and observations made from a vehicle at a distance of 100 m. The speaker was placed upwind of the cheetah to ensure that the sound carried with the wind. Since cheetahs rely principally on sight and hearing, rather than smell, it is unlikely that the fact that they were unable to smell a competitor despite being able to hear one would have influenced their reaction. The location of the loudspeaker and initial position of the cheetah were noted using a Global Positioning system (GPS) which was accurate to within 100m. This allowed verification of the initial distance between the speaker and the cheetah (mean and standard error 232 ± 8m). Wherever possible the speaker was hidden from the cheetah either in long grass, in a dip, behind a termite mound, or behind a bush. All playbacks were conducted in open habitat on the short or long grass plains, or at the woodland edges. Cheetahs in this study were well habituated to vehicles. A total of 45 individuals were used in the analyses presented in this article. Thirty-four of these were females and 11 males. In 55 out of the 79 playbacks made to females, females were accompanied by dependent cubs at the time of the experiment, 14, 20, and 21 of these were to females in dummy, hyena, and lion experiments respectively.
Lion playbacks consisted of a single bout of roars
(McComb et al., 1994), where
hyena playbacks involved two bouts of the same whoop recording separated by a
5 min interval. Since hyenas often whoop when moving
(Kruuk, 1972), playing two
bouts was intended to give the impression that a hyena is stationary in the
area. Whooping is a long distance contact call and, once away from the den, is
most commonly given by solitary hyenas
(Kruuk, 1972), therefore it is
unlikely that a cheetah, upon hearing two whoops, would perceive two hyenas as
being present. Playbacks alternated between two different hyena recordings and
five different lion recordings. A total of 36 hyena playbacks and 37 lion
playbacks were conducted (Table
1). In addition 17 "dummy" playbacks where the
equipment was set up but no sound was played were used for controlled
comparisons. No responses by lions or hyenas were recorded in any of the
experiments reported here.
|
The presence of prey (Thomson's or Grant's gazelle) within 1 km of the cheetah at the start of the experiment was noted. For 1 h after starting to play the first recording, instantaneous scans were taken every 30 s to record whether the cheetah was looking at the speaker and whether it was moving. The number of scans where a cheetah was looking at the speaker or moving was divided by the total number of scans when the cheetah was visible to give a proportion of time spent looking or moving. The latency or time to first movement was also recorded, if no movement occurred, as was the case in 20% of experiments, latency was set to 1 h, the duration of observation after experiments. In addition, for all except nine experiments, data were collected for at least 5 min before the start of each experiment. These data were used in paired t tests to assess changes in behavior before and after the three types of playbacks.
During the 1-h observation period the presence of any hunting activity was noted. A cheetah was defined to stalk if it placed its head below shoulder height, stared fixedly at prey and walked a minimum of two steps towards prey. A cheetah was defined to chase prey if it broke into a fast run after prey. At the end of the hour the cheetah's location was noted. Repeat lion and hyena playbacks of the same type to the same individual were made a maximum of three times with a minimum of 1 month between repeats. Data from dummy playback experiments were further supplemented by location and hunting data obtained on 16 occasions from cheetahs that were first sighted between 0630 and 0930am and subsequently followed for an hour (Table 1). These data were used in analyses of distance and hunting behavior.
Statistical analyses
A combination of paired t tests, ANCOVAs, and generalized linear
models were used for analyses. The initial distance of the loudspeaker, hunger
state, presence of prey and time at the start of the experiment were
controlled for in all analyses except paired t-tests and those
investigating hunting behavior. In the latter analyses these terms were
initially included in the model and were then deleted stepwise by dropping the
least significant terms one by one until all remaining terms had probability
values less than 0.1. This process was also adopted in analyses investigating
restricted data sets, that is those estimating effects of cub presence, cub
age and litter size, and explains small differences in the degrees of freedom
in reported statistics. Paired t tests implicitly controlled for all
these variables.
In paired t tests, whenever an individual was involved in more
than one experiment of the same type, responses were averaged for that
individual. All other analyses, except where indicated, controlled for
individual identity as a categorical variable whenever it exerted a
significant effect. Terms were judged as statistically significant when
probabilities were less than 5%, however all probabilities of less than 10%
are reported. Because a number of statistical analyses are conducted on the
same data set, Bonferroni statistical significance is indicated next to
reported p values in the text and in Tables
3,4,5
by way of the following symbols: I for p <.1;
* for p <.05; ** for p <.01; and
*** for p <.001
(Rice, 1989). It should be
noted that Bonferroni corrections give an overly conservative result
(Hsu and Nelson, 1998;
Samuel-Cahn, 1996).
|
|
|
Three types of model were used:
- Paired t tests and ANCOVAs for analyses of proportionate measures
of cheetah activity (looking at the speaker or moving). Proportions were
converted to a normal distribution before analysis using an arcsin
transformation after first correcting for zero values
(Sokal and Rohlf, 1981).
- ANCOVAs for analyses of the latency of first movement and the straight line
distance from the speaker 1 h after the start of the playback. Since the
logarithm of both latency and distance were distributed normally (goodness of
fit test on the logarithm of latency and distance respectively:
52 = 6.54, ns and 72 =
8.18, ns), these variables were log transformed before analysis.
- A logistic regression model for analyses of the binomial variate of the
presence or absence of hunting behavior during the experiment.
All analyses were conducted using the GENSTAT 5 version 3.1 statistical
package (Payne et al., 1987).
Results from ANCOVAs were tested using the t and F
distributions, while those from logistic regression models were tested using a
2 distribution (Sokal and
Rohlf, 1981).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cheetahs did not respond differently to different recordings of the same competitors in levels of vigilance (measured here as the proportion of time they spent looking at the speaker), movement patterns or hunting behavior. Cheetahs also did not respond significantly differently according to whether the speaker was visible to or was hidden from the cheetah at the start of the experiment. Finally, distance moved and hunting behavior after dummy playbacks did not differ according to whether the equipment was set up or not, allowing the use of these data in supplementing experimental data in subsequent analyses.
Reactions to predators
Cheetahs changed their behavior when a sound of a lion or hyena
vocalization was played, but not when no sound was played
(Table 2). Cheetahs were
significantly more likely to look at the loudspeaker and spent significantly
more time moving during the first 30 min after both lion and hyena playbacks
than during the observation period prior to the experiments
(Table 2). In the second half
hour of observation vigilance was also significantly higher after both lion
and hyena playback experiments than during the pre-playback period, but
cheetahs only spent significantly more time moving after lion experiments.
There was no difference in vigilance or movement before and after dummy
experiments.
|
Comparing across the three different playback types over the entire 1-h observation period after experiments, cheetahs were significantly more likely to look at the loudspeaker after lion or hyena playbacks than after dummy playbacks (Figure 1, Table 3). Cheetahs looked at the loudspeaker most after lion playbacks, and least after dummy playbacks. The latency to first movement did not differ between playback types (Table 4), however cheetahs spent significantly more time moving after lion or hyena playbacks than after dummy playbacks, moving most after a lion playback and least after a dummy playback (Figure 1, Table 4). In addition, at the end of the hour cheetahs were significantly farther from the loudspeaker after lion and hyena playbacks than after dummy playbacks (Figure 1, Table 4), suggesting that they were moving in a consistent direction and moved more quickly after competitor playbacks.
|
Despite an increase in activity, cheetahs were much less likely to hunt
after a lion or hyena playback than after a dummy playback
(Figure 1,
Table 5). Cheetahs were also
less likely to chase and kill prey after competitor playbacks than after dummy
playbacks (Table 5). Of those
cheetahs which did hunt, cheetahs were no less likely to chase prey after
playbacks of competitors (effect of playback type, 2 = 0.80,
ns), but they were less likely to kill prey after playbacks of lions than
after dummy or hyena playbacks (effect of playback type: coefficients for
dummy, hyena and lion playbacks respectively 0.0, 0.94, -9.7,
22 = 11.21, p =.004*). This
suggests that once cheetahs started hunting, they were just as likely to chase
prey after lion playbacks as after dummy playbacks, but they may not have
invested as much energy into the chase, resulting in a lower kill rate, and
indicating an avoidance of investing energetic effort into hunts once
started.
Are reactions strongest to lions or hyenas?
Cheetahs were significantly more likely to look at the loudspeaker after
lion playbacks than after hyena playbacks (effect of playback type,
F1,67 = 5.73, p =.019). However the latency to
first movement did not differ between lion and hyena playbacks (effect of
playback type, F1,68 = 1.71, ns), neither did cheetahs
spend more time moving (effect of playback type: F1,68 =
1.47, ns), nor move farther (effect of playback type,
F1,68 = 0.74, ns). Cheetahs were also no less likely to
hunt or chase prey after a lion playback than after a hyena playback (effect
of playback type on hunting, 12 = 1.25, ns, and on
chasing: 12 = 0.70, ns), however they were less
likely to make a kill after a lion playback (effect of playback
type,12 = 4.05, p =.044).
Factors affecting reaction
Cheetahs were not significantly more vigilant after playback experiments
when they were hungry than when they were well fed
(Table 3). However their
latency to first movement was significantly greater if they were well fed
(Table 4), and they spent
significantly less time moving (Table
4), moved significantly less far
(Table 4), were less likely to
hunt, and were marginally less likely to chase prey, but were not
significantly less likely to make a kill
(Table 5).
If there were no prey present at the start of the experiment, cheetahs were marginally less vigilant (Table 3), spent significantly more time moving (Table 4), but were only marginally significantly farther from the loudspeaker at the end of the experiment than if prey were present (Table 4). They were no more likely to hunt if prey were initially present than if prey were absent (Table 5). This was partly because either prey would move into or cheetahs move out of the area during the 1-h observation, and partly because cheetahs often hunted prey, such as hares or gazelle fawns, which were initially not visible to the observer.
Other variables such as the time of day and the initial distance of the loudspeaker from the cheetah had some effect on responses. Cheetahs were less vigilant and had a greater latency to first movement when experiments were started later in the day (Tables 3 and 4), and were farther from the loudspeaker at the end of the experiment if they were farther from it at the beginning (Table 4). Individual identity did not have a significant effect on vigilance or movement patterns, but did affect hunting behavior (Tables 3,4,5).
Sex and reproductive status
Female cheetahs did not react more strongly to competitor playbacks than
male cheetahs. Although females were significantly more likely to look at the
loudspeaker than males (effect of sex, t71 = 2.61,
p =.011I), this effect was probably a result of a
generally higher level of vigilance for females compared with males, since
there was no significant interaction between sex and playback type (effect of
interaction between sex and playback type: F1,71 = 0.09,
ns). Females were not significantly more likely to move than males (effect of
sex, t72 = 0.09, ns, interaction between sex and playback
type,F1,72 = 1.24, ns), neither did they move farther
(effect of sex, t86 = 1.29, ns, interaction between sex
and playback type: F2,88 = 1.61, ns) and were no more
likely to hunt, chase or kill after playback experiments (effect of sex for
hunt, chase and kill respectively, 12 = 0.02, ns,
12 = 0.01, ns and 12 =
0.00, ns, interaction between sex and playback type,
22 = 0.53, ns, 22 =
0.60, ns and 22 = 0.45, ns). It should be noted
that only two males were involved in hyena and only one in dummy playback
experiments (but location and hunting data were available for a further four
males—see Table 1), while
eight males were involved in lion experiments. Therefore these analyses were
largely dependent on a difference between the sexes in response to lion rather
than hyena playback experiments.
Female cheetahs did not react more strongly to competitor playbacks if they
had cubs than if they were solitary. Solitary females were marginally less
likely to look at the loudspeaker after competitor playbacks than females with
cubs (effect of cub presence, t67 = 1.74, p
=.086), but this effect was independent of playback type (effect of
interaction between cub presence and playback type, F1,66
= 0.15, ns) and therefore probably reflected higher overall levels of
vigilance for cheetah mothers (Laurenson,
1995a). Females with cubs were not significantly more likely to
move after experiments than were females without cubs (effect of cub presence,
t65 = 1.01, ns, interaction between cub presence and
playback type, F1,64 = 0.40, ns), neither did they move
farther (effect of cub presence, t76 = 0.58, ns,
interaction between cub presence and playback type, F2,74
= 0.83, ns). Furthermore, females were not less likely to hunt, chase or kill
prey after playback experiments if they had cubs (effect of cub presence,
12 = 0.71, ns; 12 =
0.09, ns and 12 = 0.04, ns, and interaction between
cub presence and playback type: 22 = 1.78, ns;
22 = 2.39, ns and 22 =
0.32, ns on hunts, chases, and kills respectively).
Reactions of females with cubs did not depend on the cubs' age or number. Females did not look at the speaker more often, spend more time moving, move farther or hunt more frequently as their litter size increased or the age of their cubs decreased, regardless of playback type (Table 6).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
These results demonstrate that cheetahs actively moved away from lions and hyenas once they perceived them to be present through an auditory stimulus. Perception of the presence of a competitor through playbacks also had a measurable impact on foraging rates of cheetahs, since cheetahs were much less likely to hunt after competitor playbacks than after dummy playbacks, resulting in a lower kill rate. Cheetahs appeared to perceive lions as a greater threat than hyenas since they were significantly more vigilant and were less likely to make a kill after lion playbacks than after hyena playbacks, although they did not move significantly farther. These differences are likely to be particularly marked given that two calls were played during hyena experiments while only one call was played during lion experiments.
It could be argued that the results obtained in this study might have been obtained if a cheetah avoids any loud sound regardless of whether it is the sound of a competitor. However, although some might argue that the sounds of ungulates such as wildebeest might be a more appropriate control than that used here, it might have also been argued that cheetahs may have been attracted to these sounds, and hence any differences between controls and predator playbacks were driven by prey attraction rather than competitor avoidance. In fact, a prior study is necessary to find a suitable control, and this was not feasible within the constraints of time and resources in this study. However while it is not possible to completely discount the loud noise hypothesis, it is most likely that the cheetahs were responding to the perceived presence of a competitor, especially given that cheetahs responded differently in terms of vigilance and kill rates to lion compared with hyena playbacks.
The stronger reaction of cheetahs to playbacks of lion calls than to hyena
calls could have one of two explanations. First, cheetahs may perceive lions
to be a greater threat than hyenas. Second, cheetah responses may vary because
lions and hyenas have different hunting strategies, as lions are stalking and
hyenas coursing predators. Unfortunately it is difficult to distinguish
between these two explanations. Previous studies have shown that prey are more
vigilant and have a longer flight distance (i.e., retreat further) to stalking
predators compared with coursing predators
(FitzGibbon and Lazarus,
1995). However whether this is because the stalking predator is a
greater threat than a coursing predator, or is due to a different response
pattern to alternative predator tactics is difficult to determine.
Were cheetahs responding to a perceived predation or a competition threat?
Cheetahs did not adapt their behavior to their reproductive status. This
result is surprising, given the high cub mortality due to predation
(Laurenson, 1994), and could
result from one of three explanations. First, cheetahs may be more at risk
when they have dependent cubs but, if the cost of avoidance is low, selection
may be too weak for cheetahs to adapt their behaviors to the relative
predation risk. Second, cheetahs may themselves be at risk with or without
dependent cubs, and by responding to predators they are avoiding a real threat
to themselves as well as their cubs. Third, cheetahs may continue to respond
to predators even when they have no cubs, because they are still vulnerable to
kleptoparastism. Since this study showed that cheetahs suffered a measurable
reduction in foraging intake when they perceived other competitors to be
present, the first explanation is unlikely. Distinguishing between the last
two explanations depends on determining the relative threat that lions and
hyenas pose to adult cheetahs.
While adult cheetahs have been known to be killed by lions within the
Serengeti (Durant SM, unpublished data), there are no records of predation by
hyenas within this ecosystem. Therefore, while there might be no reason for
cheetahs to adapt their behavioral responses to lions according to the
presence of dependent cubs, there may be a strong reason for cheetahs to adapt
their responses to hyenas, yet this pattern was not reflected in the results
reported here. Alternatively, since cheetahs lose kills to both lions and
hyenas (Caro, 1994), they are
vulnerable to kleptoparastism from both these competitors. Hence there are
probably indirect costs to not responding to the presence of a competitor even
when an individual's survival is not directly threatened. Given this, it seems
likely that cheetahs would benefit from avoiding both these competitors
regardless of their reproductive status, through an avoidance strategy.
Further evidence for this hypothesis is provided by the results obtained in
this study, where cheetahs moved farther from competitor playback experiments
when they were hungry than when they were well fed and were least vulnerable
to kleptoparasitism.
A number of other studies have demonstrated behavioral avoidance of
predators (Chivers and Smith,
1995; Dickman and Doncaster,
1984; Flowers and Graves,
1997; Hileman and Brodie,
1994; Holomuzki and Short,
1988; Kiesecker et al.,
1996; Li and Li,
1979; Loose and Dawidowicz,
1994; Main, 1987;
Peckarsky, 1996;
Semlitsch and Reyer, 1992;
Ward et. al., 1997;
Werner, 1991), however, in
most of these examples avoidance is due to a visual, scent or chemical
stimulus. There are few examples of avoidance due to an auditory stimulus (but
see Bshary and Noe, 1997).
Avoidance of infanticide, a behavior that is directly related to avoidance of
cub predation, has been documented for lions, where female lions are able to
recognize strange male lions that pose an infanticidal threat to her cubs, and
move away from playbacks of their calls
(McComb et al., 1992).
Examples of avoidance of kleptoparastism are rare.
Avoidance through active movement away from a potential predation or
competitive threat is only likely to be worthwhile if predators or competitors
are rare or aggregated (Colegrave,
1997). Thus prey, by moving away from a predator once seen, are
more likely than not to be moving to an area where predators are at lower
densities. this, in effect, means the prey is making use of a spatio-temporal
refuge, in a similar way to a physical refuge fixed in time and space. Both
lions and hyenas have an aggregated distribution
(Durant, 1998), and so a
cheetah which moves away from these competitors once seen has a better than
average chance of moving into an area with lower competitor densities.
Predator avoidance is also only likely to be a useful strategy if
anti-predator tactics are ineffective, since individuals will not choose to
bear the energetic costs of avoidance when they are unlikely to be at any real
risk. Anti-predator tactics are more likely to be employed by social species,
since they are more effective when a large number of individuals are involved
and individual risks of predation can be spread across the group
(Stanford, 1995). In addition,
social species are more likely to be the target of predator attacks because of
the increased conspicuousness of large groups
(Cowlishaw, 1997;
Lima and Dill, 1990). These
principles are also valid when an individual faces a kleptoparastism threat
rather than a predation threat. Cheetahs are generally solitary or in small
groups and are unlikely to benefit from anti-predator behavior, since they are
much smaller than lions and hyenas and are nearly always displaced by these
competitors at kills (Caro,
1994; Durant SM, personal observation).
If a cheetah is spotted by a lion or hyena it will often be approached (Durant SM personal observation). At this point a cheetah can stay put and risk attack, or flee and risk fleeing into the path of another competitor, since both hyenas and lions live in social groups. Both these options are risky, and so it is likely to be better for a cheetah to move away before the threat becomes real, preferably before it is sighted by other competitors. In this study cheetahs generally moved a few hundred m from competitors (Figure 1), distances easily covered by lions and hyenas, but which make the difference between detection or no detection of a resting cheetah by competitors. However, the activity associated with hunting is more easily detectable over these distances, which might explain the reduction in hunting activity after competitor playbacks.
Does avoidance have implications for the distribution of cheetahs within
the ecosystem? A previous study has shown that whenever cheetahs are found
near high densities of lions or hyenas they are less likely to be hunting and
more likely to be moving than at low densities
(Durant, 1998). Furthermore,
both lions and hyenas are found near high densities of gazelle, the main prey
of cheetahs on the Serengeti plains (Caro,
1994; FitzGibbon,
1990a), where cheetahs are more frequently found near low
densities of gazelle, while avoiding areas with no gazelle at all
(Durant, 1998). By avoiding
competitors, cheetahs might move away from areas with high prey densities to
areas of lower prey densities, where they are able to survive because of their
higher hunting success on small groups or isolated individuals
(FitzGibbon, 1990b). The
mobility of cheetahs and their ability to avoid direct competition in an
ever-changing landscape of competitors and prey may be the key to their
coexistence with lions and hyenas.
|
ACKNOWLEDGEMENTS |
|---|
I thank Tanzanian National Parks and the Tanzania Wildlife Research Institute (TAWIRI) for permission to conduct this study. This research was funded by the National Geographic Society, the Royal Society, Frankfurt Zoological Society, and the Institute of Zoology. I also thank M. Borner, S. and S. Tham, J. Ole Kwai, H. van Lawick and his team, and all my colleagues at TAWIRI for logistical support. K. McComb provided lion roar playback tapes, inspiration and help with the experimental design. G. Cowlishaw, T. Jones, R. Pettifor, S. Semple, and two anonymous referees have all improved this manuscript.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Brodie EDJ, Formanowicz DRJ, Brodie EDI, 1991. Predator avoidance and antipredator mechanisms: distinct pathways to survival. Ethol Ecol Evol 3: 73-77.
Bshary R, Noe R, 1997. Anti-predation behaviour of red colobus monkeys in the presence of chimpanzees. Behav Ecol Sociobiol 41: 321-333.
Caro TM, 1994. Cheetahs of the Serengeti Plains: group living in an asocial species. Chicago: University of Chicago Press.
Caro TM, Durant SM, 1991. Use of quantitative analyses
of pelage characteristics to reveal family resemblances in genetically
monomorphic cheetahs. J Hered 82:
8-14.
Chivers DP, Smith RJF, 1995. Chemical recognition of risky habitats is culturally transmitted among fathead minnows, Pimephales promelas (Osteichthyes, Cyprinidae). Ethology 99: 286-296.
Colegrave N, 1997. Can a patchy population structure affect the evolution of competition strategies? Evolution 51: 483-492.
Cowlishaw G, 1997. Trade-offs between foraging and predation risk determine habitat use in a desert baboon population. Anim Behav 53: 667-686.
Dickman CR, Doncaster CP, 1984. Responses of small mammals to red fox (Vulpes vulpes) odour. J Zool Lond 204: 521-531.
Dill LM, 1986. Animal decision-making and its ecological consequences: the future of aquatic ecology and behaviour. Can J Zool 65: 803-811.
Durant SM, 1998. Competition refuges and coexistence: An example from Serengeti carnivores. J Anim Ecol 67: 370-386.
Durant SM, 2000. Predator avoidance, breeding experience and reproductive success in endangered cheetahs (Acinonyx jubatus). Anim Behav 60: 121-130.[Web of Science][Medline]
Durant SM, Kelly MJ, Caro TM, under review. Factors affecting life and death in Serengeti cheetahs: Environment, age and sociality. Behav Ecol.
East ML, Hofer H, 1991. Loud-calling in a female dominated mammalian society. I. Structure and composition of whooping bouts of spotted hyaenas, Crocuta crocuta. Anim Behav 42: 637-649.
FitzGibbon CD, 1990a. Mixed species grouping in Thomson's and Grant's gazelles: the antipredator benefits. Anim Behav 39: 1116-1126.
FitzGibbon CD, 1990b. Why do hunting cheetahs prefer male gazelles? Anim Behav 40: 837-845.
FitzGibbon CD, Lazarus J, 1995. Antipredator behavior of Serengeti ungulates: Individual differences and population consequences. In: Serengeti II: Dynamics, management and conservation of an ecosystem (Sinclair ARE, Arcese P, eds). Chicago: University of Chicago Press; 274-296.
Flowers MA, Graves BM, 1997. Juvenile toads avoid chemical cues from snake predators. Anim Behav 53: 641-646.
Harvell CD, 1986. The ecology and evolution of inducible defenses in a marine bryozoan: cues, costs and consequences. Am Nat 128: 810-823.[Web of Science]
Harvell CD, 1990. The ecology and evolution of inducible defenses. Q Rev Biol 65: 323-340.[Medline]
Hileman KS, Brodie ED, 1994. Survival strategies of the salamander Desmognathus ochrophaeus: interaction of predator-avoidance and anti-predator mechanisms. Anim Behav 47: 1-6.
Holomuzki JR, Short TM, 1988. Habitat use and fish avoidance behaviors by the stream-dwelling isopod Lirceus fontinalis. Oikos 52: 79-86.
Hsu JC, Nelson B, 1998. Multiple comparisons in the general linear model. J Comp Graph Stat 7: 23-41.
Kennedy M, Shave CR, Spencer HG, Gray RD, 1994. Quantifying the effect of predation risk on foraging bullies no need to assume an ifd. Ecology 75: 2220-2226.
Kiesecker JM, Chivers DP, Blaustein AR, 1996. The use of chemical cues in predator recognition by western toad tadpoles. Anim Behav 52: 1237-1245.
Kruuk H, 1972. The spotted hyaena: a study of predation and social behavior. Chicago: University of Chicago.
Laurenson MK, 1994. High juvenile mortality in cheetahs (Acinonyx jubatus) and its consequences for maternal care. J Zool Lond 234: 387-408.
Laurenson MK, 1995a. Behavioural costs and constraints of lactation in free-living cheetahs. Anim Behav 50: 815-826.
Laurenson MK, 1995b. Implications of high offspring mortality for cheetah population dynamics. In: Serengeti II: Dynamics, management and conservation of an ecosystem (Sinclair ARE, Arcese P, eds). Chicago: University of Chicago Press; 385-399.
Li JL, Li HW, 1979. Species-specific factors affecting predator-prey interactions of the copepod Acanthocyclops vernalis with its natural prey. Limnol Oceanogr 24: 613-626.
Lima SL, Dill LM, 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Can J Zool 68: 619-640.
Loose CJ, Dawidowicz P, 1994. Trade-offs in diel vertical migration by zooplankton: the costs of predator avoidance. Ecology 75: 2255-2263.[Web of Science]
Main KL, 1987. Predator avoidance in seagrass meadows: prey behavior, microhabitat selection, and cryptic coloration. Ecology 68: 170-180.[Web of Science]
McComb K, Packer C, Pusey A, 1994. Roaring and numerical assessment in contests between groups of female lions, Panthera leo. Anim Behav 47: 379-387.[Web of Science]
McComb K, Pusey A, Packer C, Grinnell J, 1992. Female lions can identify potentially infanticidal males from their roars. Proc R Soc, Lond B 252: 59-64.
McIntosh AR, Townsend CR, 1994. Interpopulation variation in may-fly antipredator tactics: differential effects of contrasting predatory fish. Ecology 75: 2078-2090.
Payne RW, Lane PW, Ainsley AE, Bicknell KE, Digby PGN, Harding SA, Lech PK, Simpson HR, Todd AD, Verrier PJ, White RP, Gower JC, Tunnicliffe Wilson G, Paterson LJ, 1987. GENSTAT 5 Reference manual. Oxford: Oxford University Press.
Peckarsky BL, 1996. Alternative predator avoidance syndromes of stream-dwelling mayfly larvae. Ecology 77: 1888-1905.
Rice WR, 1989. Analyzing tables of statistical tests. Evolution 43: 223-225.[Web of Science]
Samuel-Cahn E, 1996. Is the Simes improved Bonferroni
procedure conservative? Biometrika 83:
928-933.
Schaller GB, 1972. The Serengeti Lion: A study of predator-prey relations. Chicago: University of Chicago Press.
Semlitsch RD, Reyer H-U, 1992. Modification of anti-predator behaviour in tadpoles environmental conditioning. J Anim Ecol 61: 353-360.
Sih A, 1987. Predators and prey lifestyles: an evolutionary and ecological overview. In: Predation: direct and indirect impacts on aquatic communities (Kerfoot CW, Sih A, eds). Hanover: University Press of New England; 203-224.
Sih A, 1992. Prey uncertainty and the balancing of antipredator and feeding needs. Am Nat 139: 1052-1069.[Web of Science]
Sinclair ARE, 1979. The Serengeti environment. In: Serengeti: Dynamics of an ecosystem (Sinclair ARE, Norton-Griffiths M, eds). Chicago: University of Chicago Press; 31-45.
Sokal RR, Rohlf FJ, 1981. Biometry. New York: W. H. Freeman and Company.
Stanford CB, 1995. The influence of chimpanzee predation on group size and anti-predator behavior in red colobus monkeys. Anim Behav 49: 577-587.
Turner AM, Mittelbach GG, 1990. Predator avoidance and community structure interactions among piscivores, planktivores, and plankton. Ecology 71: 2241-2254.
Ward JF, Macdonald DW, Doncaster CP, 1997. Responses of foraging hedgehogs to badger odour. Anim Behav 53: 709-720.
Werner EE, 1991. Nonlethal effects of a predator on competitive interactions between two anuran larvae. Ecology 72: 1709-1720.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Trawalter, J. A. Richeson, and J. N. Shelton Predicting Behavior During Interracial Interactions: A Stress and Coping Approach Personality and Social Psychology Review, November 1, 2009; 13(4): 243 - 268. [Abstract] [PDF]
|
|
|
|
 |
|
 A. Sansom, J. Lind, and W. Cresswell Individual behavior and survival: the roles of predator avoidance, foraging success, and vigilance Behav. Ecol., November 1, 2009; 20(6): 1168 - 1174. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Dobson Food-web structure and ecosystem services: insights from the Serengeti Phil Trans R Soc B, June 27, 2009; 364(1524): 1665 - 1682. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. Webb, R. M. Pringle, and R. Shine Intraguild predation, thermoregulation, and microhabitat selection by snakes Behav. Ecol., March 1, 2009; 20(2): 271 - 277. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Lind and W. Cresswell Determining the fitness consequences of antipredation behavior Behav. Ecol., September 1, 2005; 16(5): 945 - 956. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Durant, M. Kelly, and T. M. Caro Factors affecting life and death in Serengeti cheetahs: environment, age, and sociality Behav. Ecol., January 1, 2004; 15(1): 11 - 22. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||