«Item type Thesis or dissertation Authors Davis, Nicolas Citation Davis, N., Schaffner, C. M., & Smith, T. E. (2005). Evidence that zoo visitors ...»
The latter relationship was positive and it was not likely attributable to a Type I error as alpha equalled 0.00009. However, the slope of the data points was not steep, and levels were still relatively low as compared to other known social stressors (see Chapter 5), suggesting that although increasing visitor numbers at Chester Zoo were associated with an increase in cortisol, large numbers of visitors are not a highly stressful experience for these spider monkeys. One intervening variable that potentially precludes a more dramatic HPA response to the impact of visitors is the enclosure design. A captive animal will be more able to cope with a potentially negative stimulus, such as exposure to zoo visitors, if it is allowed to respond with active avoidance or escape responses (Carlstead, 1996). The study animals had the choice to hide from visual contact with visitors as the enclosure provided a variety of locations where the spider monkeys could be concealed from view, including tunnels, thick vegetation and grassy mounds (see Chapter 2, Figures 2.4 to 2.7).
Previous research has demonstrated that animals which control their environment, experience less stress than animals with no control (Weiss, 1968).
The majority of subjects, four out of five, demonstrated the statistically significant trend of rising cortisol levels with increasing visitor numbers. One subject however showed the opposite trend. One possible explanation is that visitors did not affect this subject, and inter animal variability is well documented (Boccia, et al., 1995; Moberg, 1985, 2000; Mormede, et al., 2007; see Chapter 1, section 1.4.7).
Alternatively, the HPA response of this subject to visitors might have been masked by additional factors since cortisol is modified by factors in addition to stress such as reproductive status, age and social dynamics (Abbott, et al., 2003; M. R. Clarke, et al., 1996; Gust, et al., 2000; Saltzman, et al., 1998). Being a retrospective study with no specific period of observations it is also possible that potential stressors such as aggressive incidents or reproductive events which were not recorded or observed could have been missed. These potential factors could therefore not be controlled for.
While the study suggests that visitors do have an impact on cortisol levels in the majority of adult spider monkeys the data for all monkeys were probably influenced by some of these potentially confounding variables since they were out of my control. Where possible their impact was accounted for, for example the separation of the male for management reasons was assessed.
This study only looked at the physiological aspect of a stress response which on its own maybe be difficult to interpret (G. Mason & Mendl, 1993). For example, one possible factor on cortisol levels is the effect of locomotion (Coleman, et al., 1998; Mormede, et al., 2007). Rates of activity are known to correlate with cortisol in marmosets (T. E. Smith, et al., 1998), although there appears to be some species differences (Mormede, et al., 2007). As a number of studies show a link between visitor numbers and increased activity (Hosey & Druck, 1987; Hosey, et al., 2009b;
Mitchell, Herring, et al., 1992; Wells, 2005) it is possible that any increases in cortisol in the current study may be due to increases of locomotion associated with the higher visitor numbers rather than a stress response to the high visitor numbers.
Therefore the interpretation of these results would have been aided by the collection of behavioural data in a more integrative approach (Dawkins, 2004). Based purely on anecdotal evidence from observations during this study no obvious changes in activity patterns occurred across varying visitor numbers. In addition, when I assessed the impact of the introduction of a new male on this same group of spider monkeys that involved cortisol and behavioural measures, there was no correlation between locomotion and cortisol levels (Chapter 6). Subtle changes in rates of scratching or where animals position themselves in the enclosure might however reveal more of a response (Carder & Semple, 2008; Maestripieri, et al., 1992).
The assessment of visitor numbers was based on overall visitor attendance at Chester zoo, and therefore some discrepancy between the assessment of visitor numbers and actual numbers at the spider monkey enclosure was possible. A count of visitors at the spider monkey enclosure may have provided a more accurate assessment. However, the position of their enclosure near the main entrance is likely to be correlated with actual numbers at the enclosure. There are additional factors that may influence the impact of visitors beyond that of sheer numbers. The behaviour of visitors, (Birke, 2002; Mitchell, Tromborg, et al., 1992), their viewing position (Chamove, et al., 1988) and the installation of visual screens (Blaney & Wells, 2004) have all been shown to mediate the effect visitors have on non-human primates and could also be taken into account.
An elevated GC level in itself does not necessarily indicate a negative effect on an animal’s welfare and moderate increases in GC are associated with optimized vigilance (Wiepkema & Koolhaas, 1993), enhanced learning, increased alertness and exploration (Chamove & Anderson, 1989). The ability for an individual to respond to short-term stressors could even be seen as beneficial, as the stimulation of the HPA axis would incite positive arousal (Chamove & Moodie, 1990). Long-term exposure to a stressor (chronic stress), however can have serious implications for an animal’s welfare (Moberg, 2000). There are therefore inherent difficulties when interpreting physiological changes and a multidisciplinary approach including behavioural and various physiological data has been recommended to assess welfare of captive animals (G. Mason & Mendl, 1993).
To conclude I quantified levels of urinary cortisol in captive spider monkeys in response to varying visitor numbers. Levels of urinary cortisol increased with rising visitor number suggesting that visitors had a potential negative impact on the monkeys. Although the increases in cortisol were not high when compared to known stressful events they still could, if sustained over long periods, be a concern for the welfare of spider monkeys in zoological parks. That a response was found in a zoo enclosure that is large and complex enough to allow the animals to choose to be out of view of zoo visitors is interesting. These finding have implications for other zoo exhibits where spider monkeys do not have such a choice and should be considered in the design of new enclosures and management practices.
PATTERNS OF INJURY IN ZOO-HOUSED SPIDER
MONKEYS: A PROBLEM WITH MALES?
4.1 Introduction There is variation in the social organisation of non-human primates, ranging from solitary to pair living to multi-male/multi-female communities, the latter commonly characterised by female philopatry and male dispersal (Kappeler & van Schaik, 2002). In addition, most group-living species are characterised by a high degree of group cohesion. Some species (e.g., chimpanzees, Pan troglodytes, and spider monkeys, Ateles spp) live in groups that are characterised by a high degree of fission-fusion dynamics in which individuals travel in small fluid subgroups or parties that change in membership throughout each day (Chapman, Wrangham, & Chapman, 1995; McFarland Symington, 1990). Fission-fusion dynamics are thought to have evolved as a means to reduce intragroup competition over spatially and temporally distributed fruit (Aureli, et al., 2008; Chapman, Fedigan, Fedigan, & Chapman, 1989; McFarland Symington, 1987, 1988, 1990). In addition, fissioning is used as a way to reduce the escalation of aggression in wild spider monkeys (Aureli & Schaffner, unpublished data). A second feature of chimpanzee and spider monkey social organisation is male philopatry and female dispersal, in which males remain in their natal group and females leave to join new groups upon reaching sexual maturity (McFarland Symington, 1990).
In the wild the spider monkeys average group range size is 278 Ha (Di Fiore & Campbell, 2007), with communities varying in size from 15 to 56 individuals (Shimooka, et al., 2008). The demographics of communities are highly varied both across communities and species. There are reports of 1 to 15 adult males, 5 to 18 adult females, 0 to 7 sub-adult males, 0 to 7 sub-adult females and 1 to 10 juveniles (Shimooka, et al., 2008). Male-male social relationships are reported to be the most affiliative as they spend more time together and groom each other more than any other adult age-sex combination (Aureli & Schaffner, 2008). The reported pattern of aggression in wild spider monkeys involves males targeting females most frequently.
Males target females by chasing them to the ground although they are very rarely physically attacked (Campbell, 2003; Fedigan & Baxter, 1984; McFarland Symington, 1987; Slater, Schaffner, & Aureli, 2008; van Roosmalen & Klein, 1988).
However, male-male aggression was unreported until recently in wild communities (Aureli & Schaffner, 2008; van Roosmalen & Klein, 1988), but two recent reports indicate that male-male aggression can be severe and in some cases lethal among intra-community males (Campbell, 2006b; Valero, Schaffner, Vick, Aureli, & Ramos-Fernandez, 2006). Aggression between females however appears to be rare in wild populations (van Roosmalen & Klein, 1988), although there are reports of longterm resident females targeting newer immigrant females (Asensio, Korstjens, Schaffner, & Aureli, 2008).
I observed two different incidents of male-male aggression between the adult male and two juvenile males at Chester Zoo. One case resulted in the death of a juvenile male that was related to the adult male. As a consequence, it was realised that very little information was available about aggressive behaviour in zoo-housed spider monkeys. However, a variety of factors are known to influence aggression in other captive primates, including the presence of human visitors in zoological parks, reduction in their living space, changes in group composition, variation in reproductive and social status and a lack of control over their physical and social environment (Honess & Marin, 2006a; Hosey, 2005; Morgan & Tromborg, 2007).
The aim of this chapter was to develop a questionnaire to investigate the prevalence of aggression in zoo-housed spider monkeys and determine whether there was a relationship between group composition and patterns of aggression. In particular information about the direction, intensity and context of any reported aggressive behaviour among the monkeys was requested. In addition, to investigate the influence of the physical environment on the occurrence and pattern of the aggression, there was a follow up request for information about enclosure dimensions. Based on the patterning of aggression reported from field studies (see above), three sets of predictions were made. Firstly, adult males would be the most frequent actors of aggression and that adult females would be the most frequent targets of minor aggression. Secondly, adult males would be the actors of severe and lethal aggression and that the juvenile males would be the targets. Finally, it was predicted that females would be the least frequent actors of aggression.
4.2.1 Procedure I developed my questionnaire to obtain information on incidents of aggression in zoo housed spider monkeys based on an earlier study that used a similar tool for exploring patterns of aggression in captive lion tamarins (Inglett, et al., 1989). The primary aim was to collate accurate information in a form that would reduce subjectivity and over generalisation and that could then be analysed. Using the International Species Information System (ISIS, 2008) I identified a total of 55 zoological parks world-wide that maintained social groups of at least three adult spider monkeys with a mixed sex composition. A questionnaire was distributed in English in March 2002 by email to the appropriate curators, along with a covering letter, which requested information on any recorded aggressive events that had occurred in the previous five years (Appendix A). In addition, I requested that any available ARKS (Animal record keeping system) or suitable zoo records, which are maintained by the keepers responsible for the spider monkeys, be returned with the questionnaire. The first question requested information about the species of spider monkey. At the time of drafting the questionnaire four species were housed regularly in zoological parks, including A. belzebuth, A. fusciceps, A. geoffroyi, and A.
paniscus. However, A. fusciceps has recently been reclassified and is now recognised as A. geoffroyi rufiventris (Rylands, et al., 2000; see Chapter 2). The questions focused on the frequency, context, direction and intensity of aggressive events; the age and sex of the individuals involved; the group composition at the time of aggressive events and the patterning of spider monkey aggression relative to other species that were housed at the same zoological park.
Age classifications were categorised with adults over six years old (van Roosmalen & Klein, 1988), sub-adults from four to six years, juveniles from two to four years and infants under two years. The resulting age-sex categories were as follows: adult males, adult females, sub-adult males, sub-adult females, juvenile males and juvenile females. The aggressive incidents were classified into three categories of different intensities based on the descriptions provided: “minor”, which included either no observed injuries or superficial injuries; “severe”, which included single, or multiple wounds that required veterinary treatment; and “lethal”, when the individual was killed outright or where the injuries were so serious they necessitated that the individual be euthanized. Thirteen zoological parks also provided information about the size of the spider monkey enclosures. I requested the information as a follow up to the initial questionnaire as social behaviour, including aggressive interactions can be influenced by the area of the enclosure (Caws & Aureli, 2003; Hosey, 2005; Judge & de Waal, 1997; Kummer & Kurt, 1965) and the number of individuals in the social group. This additional information was then used to assess area as a potential factor in the prevalence of aggression.