Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Behavioral Ecology The official journal of the ISBE International Society for Behavioral Ecology Behavioral Ecology (2016), 27(4), 1073–1077. doi:10.1093/beheco/arw015 Original Article Zebra reduce predation risk in mixed-species herds by eavesdropping on cues from giraffe Melissa H. Schmitt, Keenan Stears, and Adrian M. Shrader School of Life Sciences, University of KwaZulu-Natal, Private Bag X01, Scottsville 3209, South Africa Received 28 September 2015; revised 19 January 2016; accepted 23 January 2016; Advance Access publication 17 February 2016. Predation risk of individuals moving in multispecies herds may be lower due to the heightened ability of the different species to detect predators (i.e., mixed-species effect). The giraffe is the tallest land mammal, maintains high vigilance levels, and has good eyesight. As a result, heterospecific herd members could reduce their predation risk if they keyed off the giraffe’s antipredator behaviors. However, because giraffe rarely use audible alarm snorts, heterospecifics would need to eavesdrop on cues given off by the giraffe that indicate predator presence (e.g., body posture), to benefit from herding with giraffe. To test this, we compared the vigilance of zebra herding with conspecifics, with those herding with giraffe. Our results indicate that giraffe reduce zebra vigilance in zebra–giraffe herds and that in these herds, giraffe are the primary source of information regarding predation risk. In contrast, when zebra herd with conspecifics, they rely primarily on personal information gleaned from their environment, as opposed to obtaining information from conspecifics about predation risk. Key words: eavesdropping, mixed-species effect, multispecies herds. INTRODUCTION For animals, a key advantage of group living is a reduction in predation risk through dilution and/or collective detection (Hamilton 1971; Schmitt et al. 2014). This reduced predation risks lowers vigilance levels and allows herbivores to devote more time to other activities. However, there are fitness advantages to maintaining personal vigilance scans (Dehn 1990). For example, individuals that detect an approaching predator are more likely to escape compared with those that rely on collective detection (i.e., being made aware of predator presence by another group member detecting the predator; Elgar 1989; FitzGibbon 1989). For some species, risk can be reduced further by herding with a diluting partner (i.e., another species that shares a common predator; zebra and wildebeest are “diluting partners” sensu Schmitt et al. 2014). A diluting partner may reduce predation risk from a shared predator by having different detecting abilities. Such mixed-species effects on risk likely reflect a heightened ability of the mixed group to detect approaching predators compared with single species herds. This mixed-species effect has been shown to reduce personal vigilance in a number of species (Scheel 1993; van der Meer et al. 2012; Schmitt et al. 2014). For example, kudu (Tragelaphus strepsiceros) were able to drink for longer at waterholes Address correspondence to M.H. Schmitt. E-mail: [email protected]. K.S. Coauthor is now at the Department of Ecology, Evolution, and Marine Biology, University of California Santa Barbara, Santa Barbara, CA 9316-9620, USA © The Author 2016. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: [email protected] when herding in mixed-species groups compared with conspecific groups (Périquet et al. 2010). Cattle (Bos taurus), when in the presence of mule deer (Odocoileus hemionus) stimuli showed a “comfort response” whereby they reduce their vigilance levels. This reduction in vigilance allowed the cattle to spend more time foraging (Kluever et al. 2009). Although the benefits of mixed-species herding have been shown (Scheel 1993; Périquet et al. 2010; van der Meer et al. 2012; Schmitt et al. 2014), the extent to which this benefit varies across different coherding species has scarcely been explored. For example, body size can influence potential predation risk, with smaller species having greater predation risk and exhibiting higher vigilance levels than larger species (Sinclair et al. 2003). Therefore, herding with smaller species should be more beneficial than herding with larger species, as long as the species share the same predators (i.e., diluting partners; Schmitt et al. 2014). Alternatively, large-bodied species may be less negatively influenced by visual constraints, such as vegetation, when detecting predators. Therefore, small herbivores could potentially enhance their fitness by herding with larger species whose detecting abilities are less diminished by habitat structure. The association of giraffe (Giraffa camelopardalis) with other herbivores has led to the assumption that giraffe may benefit other coherding species (e.g., Dagg and Foster 1976; Leuthold 1977). This may come from the giraffes’ increased ability to detect predators due to their height (Young and Isbell 1991), good eyesight (Mitchell et al. 2013), and high vigilance levels (Creel et al. 2014). However, there is no scientific evidence to support this assumption. Moreover, giraffe only rarely produce audible alarm snorts 1074 (Schaller 1972; Moss 1975). Thus, eavesdropping on alarm calls cannot be the main mechanism by which coherding species reduce their predation risk. Nevertheless, giraffe do exhibit an easily identifiable staring posture when in the presence of predators (Dagg 1971). This posture is most likely a cue, which is an incidental feature that does not have intended meaning to a receiver (Saleh et al. 2007), unlike a signal (e.g., alarm call/snort), which is believed to elicit a hard-wired response (Beauchamp et al. 1976). As a result, it may be possible that individuals can learn about potential threats by observing the reaction (a cue) of other members in their group (e.g., giraffe), even in the absence of an alarm call. For example, in gray kangaroos (Macropus giganteus), a species that does not produce alarm calls, conspecifics cue off the reactions of other group members to predators to gain information regarding threats (Pays et al. 2013). However, whether one species can react to the posture of another (i.e., a cue; Maynard Smith and Harper 2003) in the absence of an alarm call, has scarcely been explored. Consequently, the aims of this study were to 1) determine whether the presence of giraffe can reduce zebra vigilance because they share the same key predator (i.e., lion; Hayward and Kerley 2005), 2) compare the impact between giraffe and a known “diluting partner” (wildebeest) on zebra vigilance, and 3) quantify the degree to which giraffe alter the manner in which zebra assess risk. We expected that due to the height advantage of giraffe, that zebra would reduce their time spent vigilant and rely more heavily on information gathered from coherding giraffe. Additionally, we predicted that giraffe may have a greater impact on zebra vigilance behavior compared with wildebeest because not only are they a diluting partner, but due to their height advantage, zebra may perceive the quality of information from giraffe to be better. If this was the case, then we expected that zebra would focus more of their vigilance toward giraffe and rely less on personal scanning of the environment. Alternatively, we may find that zebra are unable to decipher nonaudible cues (i.e., the giraffe’s behavioral posture indicating the presence of a predator), and thus, giraffe would not influence zebra vigilance behavior. MATERIAL AND METHODS All aspects of the research design were approved by our institution (Ethics Code: 13/11/Animal). We conducted fieldwork during August–September (late dry season) 2013 and December (wet season) 2014. We collected data from Hluhluwe-iMfolozi Park and Kruger National Park, South Africa. To minimize habitat differences in perceived predation risk, we limited observations to herds feeding in savannas with visibly similar tree cover. However, within a savanna, there can be differences in the tree:grass ratio that could influence perceived predation risk (Stears and Shrader 2015) and ultimately vigilance (Metcalfe 1984). Giraffe tend to avoid woody habitats, particularly with young, and select open savannas where personal vigilance and speed of escape are facilitated (Young and Isbell 1991). Thus, we assumed in our study that giraffe selected similar habitats and small-scale differences in tree:grass ratios would not result in differences in vigilance levels. To control for variation across our study sites, we included a site effect in our statistical analysis. Additionally, we collected data 2 h after first light and 2 h before last light to maintain the same level of perceived predation risk across days (as per Scheel 1993; Schmitt et al. 2014). All observations occurred from a stationary vehicle using binoculars. To avoid potential behavioral changes due to vehicles, we only Behavioral Ecology collected data when no other vehicles were present and when the zebra were more than 20 m from the road. In a study of lion prey preferences, Hayward and Kerley (2005) found that across 48 different lion populations, giraffes were the 4th most preferred prey item, whereas zebra were the 7th most preferred prey item. Because both zebra and giraffe are categorized as favored prey items for lions (Hayward and Kerley 2005), these herbivores gain the benefit of both dilution and detection when herding together (i.e., diluting partners; Schmitt et al. 2014). A study by Schmitt et al. (2014) found that detection was the main factor in reducing vigilance in small and medium herd sizes (2–30 individuals). In this study, we were only interested in the effects of detection in reducing vigilance levels. Therefore, to make vigilance comparisons between herds as a result of detection, we limited data collection to similar small-sized zebra-only and zebra–giraffe herds that we encountered (i.e., 2–14 individuals). Zebra-only herds comprised a group of zebra that fed within 6 body lengths of their nearest conspecific (~12 m), whereas zebra–giraffe herds contained at least 1 giraffe within 12 m of a zebra. We collected data from adult zebra of both sexes, but avoided herds that contained zebra or giraffe juveniles. Giraffe within the herds comprised either adult males and/or females. Vigilance observations started when a focal zebra had its head down and was grazing. We considered zebra being actively vigilant when they lifted their head above grazing height and scanned for predators, or focused their gaze and actively listened (as per Scheel 1993; Schmitt et al. 2014). Due to the placement of their eyes on the side of their head, zebras can use monocular and binocular vision (Barnett 2004). As a result, they can see to their side using monocular vision while they forage (Harman et al. 1999). This ability could potentially allow zebra to be passively vigilant while they feed and thus detect approaching predators. However, because zebra head placement while feeding is pointing toward the ground, we are unable to differentiate between zebra being passively vigilant (looking for potential predators using monocular vision) or searching for another foraging patch. Furthermore, it is unlikely that the level of passive vigilance shown by zebra while foraging will differ between zebra-only and zebra–giraffe herds. Thus, we restricted our data collection to active forms of vigilance (Scheel 1993; Creel and Winnie 2005; Creel et al. 2014). Within a herd, we did not record data from the same individual twice, but rather collected data from >75% of the individuals in the herd. To explore whether the presence of giraffes influenced zebra vigilance, we compared active (head raised) zebra vigilance when herding in zebra-only herds, with zebra herding in zebra–giraffe herds. We observed a zebra for 3 min and recorded: 1) time spent actively vigilant, 2) intensity of each vigilance event (i.e., general or focused scan), and 3) the source of information used to determine predation risk (i.e., herd member or the environment). We defined a general scan as a zebra scanning with its head up without fixing its attention (vision or ears) in a particular direction (zebraonly: n = 153 scans, zebra–giraffe: n = 79 scans). A focused scan comprised a zebra staring in a fixed direction, with its ears pricked, either looking toward another herd mate (i.e., zebra or giraffe) or out toward the environment (zebra-only: n = 325 scans, zebra– giraffe: n = 17 scans). We sampled both herd types in both reserves (iMfolozi: zebra–giraffe: n = 11 herds, 64 individuals, zebra-only: n = 34 herds, 106 individuals; Kruger: zebra–giraffe: n = 7 herds, 13 individuals, zebra-only: n = 17 herds, 63 individuals). With regard to source of information, when herding in a zebraonly herd, a zebra could obtain information about predation risk Schmitt et al. • Giraffe reduce predation risk of zebra by looking at a conspecific, conducting its own scan of the environment, or both. When herding with giraffe, zebra had the same options, but looking at herd mates comprised conspecifics and/ or giraffe. For zebra, binocular vision (oriented in the direction of the muzzle) provides better depth perception compared with their side-orientated monocular vision (Harman et al. 1999). As a result, it is more likely that these herbivores would use binocular vision to actively scan for approaching predators. Thus, we considered zebra being actively vigilant when their head was raised and their muzzle was directed toward a zebra, giraffe, or out to the environment. If zebra looked at 2 sources separately in a scan (e.g., herd mate and the environment), we noted that as both. Additionally, we were able to discern the direction of a zebra’s scan a majority of the time, even when the focal individual was in the center of the herd. However, if a zebra directed a scan toward a zebra and a giraffe, and we could not discern which individual was the focus of the scan, we recorded the scan as toward an unknown target. We used a subset of our total samples to determine information source because we only began collecting these data midway through our data collection (iMfolozi: zebra–giraffe: n = 7 herds, 21 individuals, zebra-only: n = 22 herds, 63 individuals; Kruger: zebra–giraffe: n = 7 herds, 13 individuals, zebra-only: n = 10 herds, 24 individuals). Finally, to explore the relative influence that giraffe have on zebra vigilance behavior, we compared the amount of time zebra spent vigilant when herding with giraffe and when herding with a known diluting partner—wildebeest. We collected zebra vigilance behavior in the same manner as described above. These data come from Schmitt et al. (2014) (iMfolozi: zebra–giraffe: n = 11 herds, 32 individuals; zebra–wildebeest: n = 2 herds, 14 individuals; Kruger: zebra–giraffe: n = 9 herds, 13 individuals, zebra–wildebeest: n = 4 herds, 32 individuals). Data analysis Prior to statistical analysis, we calculated mean individual vigilance per herd (i.e., herds as replicates) to avoid possible pseudoreplication. To test whether time devoted to vigilance was influenced by herding with giraffe, we tested for herd independence to ensure the data could be pooled. There was no herd effect; thus, individual herds were independent. Therefore, we used the mean time a herd was vigilant as our dependent factor. We used an analysis of covariance (ANCOVA) to test whether the mean time spent vigilant by the herds varied with herd type (zebra-only and zebra–giraffe). The covariates such as number of giraffe, total number of ungulates, and season were nonsignificant and thus removed from the final model. To control for a site effect, we used “site” as a covariate in all analyses. Data were box–cox y transformed prior to analysis. To determine the relative impact of giraffe herding with zebra compared with when zebra herd with a known diluting partner, wildebeest, we used an ANCOVA with time spent vigilant as the dependent variable. We used herd type (zebra–giraffe and zebra– wildebeest) as the main effect and number of zebra in a herd as the covariate. Initially, we also included site, season, and total number of ungulates as covariates. However, these were nonsignificant and thus removed from the final model. These data are from Schmitt et al. (2014) and were box–cox y transformed prior to analysis. To test whether the intensity of vigilance events varied between zebra in zebra-only or zebra–giraffe herds, we compared the proportion of general versus focused scans within each herd. We used proportion of each scan type per herd as the dependent variable in an analysis of variance (Anova) and the interaction of herd 1075 type (zebra-only and zebra–giraffe) and scan type (1 = general, 2 = focused) as the main factors. Therefore, each herd had a proportion for a general and a fixed scan. We were able to differentiate between these proportions by including the independent variable, scan type. By including both these proportions for each herd type, we were able to determine if there was any change in the proportional use of each behavior as herd type changed. Covariates site (iMfolozi, Kruger), total number of ungulate individuals in a herd, and season were nonsignificant and removed. Data were transformed using arcsine square root. Finally, to determine the source(s) that zebras used to gather information about perceived predation risk, we used an Anova with proportion of scans (regardless of the scan intensity) directed toward an information source (i.e., herd mate, environment, or herd mate plus environment), per herd as the dependent variable. We then used the interaction of herd type (zebra-only and zebra–giraffe) and information source as the main effects. Covariates site, season, and total number of ungulates were nonsignificant and thus removed. Data were arcsine square root transformed prior to analysis. RESULTS The presence of giraffe lowered the average time a zebra was vigilant by nearly two-thirds (ANCOVA: F2,68 = 15.75, P < 0.0001). Zebra herding with conspecifics spent an average (mean ± standard error) of 27 ± 2 s vigilant, whereas zebra herding with giraffe spent only 10 ± 4 s vigilant per three-min observation period. The covariate number of giraffe was nonsignificant, indicating that the presence rather than number of giraffe reduced zebra vigilance. When we compared the amount of time zebra spent vigilant when herding with giraffe versus with wildebeest, we found that the presence of giraffe lead to a 50% reduction of zebra vigilance compared with the level zebra maintained when herding with wildebeest (ANCOVA: F2,23 = 13.280, P < 0.0001). This reduction indicates that zebra perceive information gleaned from giraffe to be more valuable than that from wildebeest. In addition to altering vigilance time, the intensity of vigilance scans by zebra differed between herds with and without giraffe (F3,130 = 19.31, P < 0.0001). When herding alone, zebra used focused scans proportionally more (0.70 ± 0.039) than general scans (0.29 ± 0.039). However, when herding with giraffe, zebra used general scans proportionally more (0.81 ± 0.064) than focused scans (0.19 ± 0.064). This pattern was consistent across sites. The source of information used by zebra to assess risk depended on whether giraffe were present or not (Figure 1; significant herd type × information source interaction term; F5,132 = 46.46, P < 0.0001). Zebras herding alone actively scanned the environment (0.79 ± 0.04) significantly more than individuals herding with giraffe (0.20 ± 0.06). In contrast, zebra herding with giraffe actively scanned herd mates (0.72 ± 0.06) significantly more than zebras herding alone (0.15 ± 0.04). Neither herd types extensively scanned both herd mates and the environment. Zebra herding with giraffe looked directly at herd mates 73% of their scans, and of these scans directed at their herd mates, zebra looked directly at giraffe 45%, giraffe and zebra 23%, and zebra 5% of the time. In contrast, zebra herding alone only used their herd mates ~20% of the time for information. DISCUSSION Moving in groups is a key way in which animals reduce predation risk (Krause and Ruxton 2002). However, for this to be possible, Behavioral Ecology 1076 Proportion of scans directed towards information sources 0.8 Zebra-Giraffe Herds Zebra-Only Herds 0.6 0.4 0.2 0.0 Herd Mate Environment Herd Mate+Environment Information sources Figure 1 Mean (± standard error) proportional use of information sources by zebra when herding with conspecifics or giraffe. group members must be able to obtain information about potential threats from other group members (Metcalfe 1984; Elgar 1989). A number of studies have shown that information about predators can be conveyed through alarm calls (Seyfarth et al. 1980; Fichtel and Kappeler 2002; Manser et al. 2002), but this is a challenge in species that either do not have auditory alarm calls (Pays et al. 2013), or rarely uses them (e.g., giraffe; Schaller 1972; Moss 1975). Our results show that when one or more giraffe were present in a group of foraging zebra, the time zebra devoted to vigilance dropped by nearly two-thirds compared with when herding with conspecifics. Furthermore, in mixed-species herds, the majority of zebra scans were directed toward giraffe. Therefore, we hypothesize that the mechanisms through which this reduction occurs is likely because zebra are able to interpret the leaked cues (e.g., posture) from the giraffe. As a result, zebra are able to reduce their perceived predation risk even in the absence of alarm calls. Interestingly, the amount of time that zebra devoted to vigilance when herding with giraffe was half as much as when they herded with wildebeest, a known diluting partner (Schmitt et al. 2014). Thus, it seems that zebra perceive these leaked cues from giraffe to be more valuable, and more reliable, than information from a similar-sized diluting partner. Furthermore, in zebra–giraffe herds, zebra relied more on general vigilance scans compared with the focused scans (i.e., staring in a particular direction with ears pricked) used in zebraonly herds. Moreover, there was a near complete switch in behavior from relying on direct personal assessment of risk in zebra-only herds, to primarily using social information (i.e., eavesdropping) on other herd members in zebra–giraffe herds. Because giraffe do not frequently use alarm snorts, in fact, we never heard them make any vocalizations during our study, it is plausible that the zebra actively eavesdropped off the giraffe’s behavior (i.e., cues) to assess predation risk. In zebra–giraffe herds, the majority of social information (~45%) used by zebra involved directing their scans toward giraffe. An additional 23% focused toward both giraffe and conspecifics. However, the zebras’ inspections of conspecifics were more likely to protect against competition (Artiss and Martin 1995) because zebra in zebra-only herds did not focus on conspecifics to reduce predation risk. The switch in information source used by zebra is surprising because in mixed-species herds, zebra theoretically should use both conspecific and heterospecific cues equally to better determine predation risk (Goodale and Kotagama 2008), rather than relying mainly on cues from giraffe. However, this shift can potentially be explained by 2 reasons. First, when compared with zebra and other similar-sized ungulates, the elevated position of giraffe’s eyes (and thus a better view than what shorter species have), coupled with good eyesight (Mitchell et al. 2013), most likely increases the distance over which a giraffe can detect approaching predators. This is similar to other species that rely on elevated conspecific sentinels for antipredator information, such as meerkats (Suricata suricatta) (Santema and Clutton-Brock 2013). Second, the large body size and the characteristic body language displayed by giraffe when they detect the presence of lion (i.e., freeze and stare posture Dagg 1971) facilitate visual eavesdropping on these cues by zebra. To do this, it is likely that zebra focus on the entire body posture of the giraffe and the fact that the giraffe has/have ceased feeding and are focusing their gaze, and thus their head, in 1 direction, indicating the presence of predators. The information gained by zebra from giraffe is perhaps more accurate than information from conspecifics or other coherding species, such as wildebeest, because the height of the giraffe makes it less likely to be obstructed by vegetation. For example, a study on 2 species of shorebirds (Arenaria interpres and Calidris maritime) found that with decreasing visibility there was an increase in individual vigilance (Metcalfe 1984). This was because the shorebirds could not rely as much on antipredator information gathered from other members in the flock because they were obstructed from view, in addition to a greater risk of predation from hidden predators with decreasing visibility (Metcalfe 1984). Thus, the information from giraffe is likely more valuable to zebra than information from conspecifics or similar-sized heterospecifics sharing the same predation risk. This is likely because zebra can gain more information about predators (e.g., direction of approach) from giraffe than they can from potentially obstructed conspecifics within the herd. If our interpretation is correct, this then raises the question, if zebra gain so much by moving with giraffe, then why do not they always herd with them? A potential reason for this is that giraffe may experience lower levels of predation by lions compared with zebra. In general, bigger-bodied animals are less sensitive to predation (Sinclair et al. 2003) and therefore should have lower vigilance levels than smaller-bodied herbivores. For example, Sinclair et al. (2003) found that in the Serengeti, lions hardly preyed on giraffe. However, a meta-analysis of the prey preference of lions that covered 48 different lion populations found that giraffe were determined to be the 4th most preferred prey item, whereas zebra were only the 7th most preferred prey item (Hayward and Kerley 2005). Thus, due to the risk of predation from lions, giraffe should maintain high vigilance levels. Although this level may be less than that of zebras, our results suggest that the quality of the information (i.e., unobstructed view) still benefits zebra. Otherwise, why would they rely so heavily on the giraffe? However, if this is the case, then it does not explain why zebra and giraffe do not always move in mixed herds. A more likely reason for the lack of herding of zebra with giraffe may have to do with information that is gained beyond predator presence. In our study, it was common for zebra to herd with giraffe, but they more frequently herded with wildebeest, impala, or a combination of these 2 herbivores. Specifically, because zebra are grazers, they do not gain foraging information by moving with browsing giraffe. In contrast, when herding with other grazers/mixed feeders (e.g., wildebeest and impala), zebra can not only reduce their predation risk (Schmitt et al. 2014), but also obtain social information Schmitt et al. • Giraffe reduce predation risk of zebra about the availability and location of food (Valone and Templeton 2002; Shrader et al. 2007; Pays et al. 2014). Thus, the combination of these two factors may be more beneficial than the reduced predation risk provided by giraffe, hence explaining the more frequent herding by zebra with grazers and mixed feeders. However, when zebra herded with giraffe they did follow them and continue to feed in close proximity when the giraffe moved off (Schmitt MH, Stears K, personal observation). During our study, we did see giraffe herding with other browsing herbivore species including kudu and impala (Schmitt MH, Stears K, personal observation). However, kudu and impala browse at a much lower height than giraffe (Du Toit 1990). Therefore, similar to zebra, these browsing herbivores only gain a reduction in predation risk and no benefit from social information about the availability and location of food. This could explain why other browsing species do not continuously herd with giraffe. Ultimately, our data suggest that when in zebra-only herds, zebra primarily assess predation risk by directly scrutinizing their outward environment (i.e., personal information). However, when moving in zebra–giraffe herds, they shift and primarily rely on giraffe as a way of detecting predators. The combination of these data suggests that zebra are not reliant on auditory alarm snorts/calls, and can possibly interpret leaked cues from giraffe to reduce their vigilance behavior. Thus, it would seem that for zebra in zebra–giraffe herds that the reliability of the information about predation risk is far greater than the benefits that zebra obtain by scanning the environment themselves. Additionally, the survival and fitness benefits obtained by eavesdropping on giraffe are greater than when herding with conspecifics or similar-sized diluting partners. FUNDING This project was funded through a grant to A.M.S. (IPRR) from the National Research Foundation. We thank R.J. Schmitt and S.J. Holbrook who provided statistical advice and valuable comments on earlier drafts as well as S. Swarbrick for constructive discussions. We would also thank T. Caro and an anonymous reviewer for their constructive comments and suggestions. Handling editor: Johanna Mappes REFERENCES Artiss T, Martin K. 1995. Male vigilance in white-tailed ptarmigan, Lagopus leucurus: mate guarding or predator detection? Anim Behav. 49:1249–1258. Barnett KC. 2004. Equine ophthalmology: an atlas and text. New York: Saunders. Beauchamp GK, Doty RL, Moulton DG, Mugford RA. 1976. The pheromone concept in mammalian chemical communication: a critique. In: Doty RL, editor. Mammalian olfaction, reproductive processes, and behaviour. 1st ed. London: Academic Press. p. 144–157. Creel S, Schuette P, Christianson D. 2014. Effects of predation risk on group size, vigilance, and foraging behavior in an African ungulate community. Behav Ecol. 25:773–784. Creel S, Winnie JA. 2005. Responses of elk herd size to fine-scale spatial and temporal variation in the risk of predation by wolves. Anim Behav. 69:1181–1189. Dagg AI. 1971. Giraffa camelopardalis. Mamm Species. 5:1–8. Dagg AI, Foster JB. 1976. The giraffe: its biology, behaviour and ecology. New York: Van Nostrand Reinhold Company. Dehn M. 1990. Vigilance for predators: detection and dilution effects. Behav Ecol Sociobiol. 26:337–342. Elgar MA. 1989. Predator vigilance and group size in mammals and birds: a critical review of the empirical evidence. Biol Rev Camb Philos Soc. 64:13–33. 1077 Fichtel C, Kappeler PM. 2002. Anti-predator behavior of group-living Malagasy primates: mixed evidence for a referential alarm call system. Behav Ecol Sociobiol. 51:262–275. FitzGibbon CD. 1989. A cost to individuals with reduced vigilance in groups of Thomson’s gazelles hunted by cheetahs. Anim Behav. 37:508–510. Goodale E, Kotagama SW. 2008. Response to conspecific and heterospecific alarm calls in mixed-species bird flocks of a Sri Lankan rainforest. Behav Ecol. 19:887–894. Hamilton WD. 1971. Geometry of the selfish herd. J Theor Biol. 31:295–311. Harman AM, Moore S, Hoskins R, Keller P. 1999. Horse vision and an explanation for the visual behaviour originally explained by the ‘ramp retina’. Equine Vet J. 31:384–390. Hayward MW, Kerley GIH. 2005. Prey preferences of the lion (Panthera leo). J Zool. 267:309–322. Kluever BM, Howery LD, Breck SW, Bergman DL. 2009. Predator and heterospecific stimuli alter behaviour in cattle. Behav Processes. 81:85–91. Krause J, Ruxton GD. 2002. Living in groups. Oxford: Oxford University Press. Leuthold W. 1977. African ungulates. Berlin (Germany): Springer-Verlag. Manser MB, Seyfarth RM, Cheney DL. 2002. Suricate alarm calls signal predator class and urgency. Trends Cogn Sci. 6:55–57. Maynard Smith J, Harper D. 2003. Animal signals. New York: Oxford University Press. van der Meer E, Pays O, Fritz H. 2012. The effect of simulated African wild dog presence on anti-predator behaviour of kudu and impala. Ethology. 118:1018–1027. Metcalfe NB. 1984. The effects of habitat on the vigilance of shorebirds: is visibility important? Anim Behav. 32:981–985. Mitchell G, Roberts DG, Van Sittert SJ, Skinner JD. 2013. Orbit orientation and eye morphometrics in giraffes (Giraffa camelopardalis). Afr Zool. 48:333–339. Moss C. 1975. Portraits in the wild: behavior studies of East African mammals. Boston (MA): Houghton Mifflin. Pays O, Beauchamp G, Carter AJ, Goldizen AW. 2013. Foraging in groups allows collective predator detection in a mammal species without alarm calls. Behav Ecol. 24:1229–1236. Pays O, Ekori A, Fritz H. 2014. On the advantages of mixed-species groups: impalas adjust their vigilance when associated with larger prey herbivores. Ethology. 120:1207–1216. Périquet S, Valeix M, Loveridge AJ, Madzikanda H, Macdonald DW, Fritz H. 2010. Individual vigilance of African herbivores while drinking: the role of immediate predation risk and context. Anim Behav. 79:665–671. Saleh N, Scott A, Bryning G, Chittka L. 2007. Distinguishing signals and cues: bumblebees use general footprints to generate adaptive behaviour at flowers and nest. Arthropod Plant Interact. 1:119–127. Santema P, Clutton-Brock T. 2013. Meerkat helpers increase sentinel behaviour and bipedal vigilance in the presence of pups. Anim Behav. 85:655–661. Schaller GB. 1972. The Serengeti lion: a study of predator-prey relationships. Chicago (IL): University of Chicago Press. Scheel D. 1993. Watching for lions in the grass: the usefulness of scanning and its effects during hunts. Anim Behav. 46:695–704. Schmitt MH, Stears K, Wilmers CC, Shrader AM. 2014. Determining the relative importance of dilution and detection for zebra foraging in mixedspecies herds. Anim Behav. 96:151–158. Seyfarth RM, Cheney DL, Marler P. 1980. Monkey responses to three different alarm calls: evidence of predator classification and semantic communication. Science. 210:801–803. Shrader AM, Kerley GIH, Kotler BP, Brown JS. 2007. Social information, social feeding and competition in group living goats. Behav Ecol. 18:103–107. Sinclair AR, Mduma S, Brashares JS. 2003. Patterns of predation in a diverse predator-prey system. Nature. 425:288–290. Stears K, Shrader AM. 2015. Increases in food availability can tempt oribi antelope into taking greater risks at both large and small spatial scales. Anim Behav. 108:155–164. Du Toit JT. 1990. Feeding-height stratification among African browsing ruminants. Afr J Ecol. 28:55–61. Valone TJ, Templeton JJ. 2002. Public information for the assessment of quality: a widespread social phenomenon. Philos Trans R Soc Lond B Biol Sci. 357:1549–1557. Young TP, Isbell LA. 1991. Sex differences in giraffe feeding ecology: energetic and social constraints. Ethology. 87:79–89.