«Item type Thesis or dissertation Authors Davis, Nicolas Citation Davis, N., Schaffner, C. M., & Smith, T. E. (2005). Evidence that zoo visitors ...»
The design of the enclosure during group formation can also be important (Brent, et al., 1997; Westergaard, et al., 1999). Aggression and injuries in chimpanzees were lower in facilities where the design allowed for dyadic pairing and gradual introductions, with initial visual contact, followed by limited tactile contact before full contact (Alford, et al., 1995; McDonald, 1994). In rhesus macaques wounding rates were higher if there were no visual or social barriers present during introductions (Westergaard, et al., 1999).
The sex of the individual seems to be an influential factor. Introductions of male chimpanzees were much more likely to be unsuccessful than those involving females (Brent, et al., 1997). This fits in with the social organisation of chimpanzees in which males respond with affiliative behaviour toward newly immigrant females.
Similar patterns were also found in other non human primates (Bernstein, 1991;
Crockett, et al., 1994). Prior familiarity of the individuals may significantly reduce the impact of forming new groups (Schaffner & French, 2004; Schaffner & Smith, 2005). Other factors include previous social housing and age at introduction (Reinhardt, et al., 1995).
6.3 Spider monkeys
The dispersal pattern for spider monkeys is one of the few exceptions in mammals in which female emigration is the rule and males remain in their natal group all of their lives (Aureli & Schaffner, 2008; Greenwood, 1980; see Chapter 2).
However, in zoo populations, to maintain genetic diversity, it is common practice for the males to be transferred across facilities with females remaining in their natal group. The consequences of this unnatural dispersal pattern in spider monkeys across zoological parks has been investigated (see Chapter 4), however, its effect on the behaviour and stress response is unknown and has not previously been assessed.
Based on the social ecology of spider monkeys, particularly regarding female dispersal, and previous studies involving introductions of other primate species in captivity in which abrupt introductions have lead to serious and occasional lethal aggression (Crockett, et al., 1994; Reinhardt, et al., 1995) I made two predictions.
The first was that the introduction of the new male into an established group of female spider monkeys and their offspring would initiate a stress response in the residents and the new male. I further predicted that behavioural indices of stress, such as self-directed behaviour would be more evident immediately after the male was introduced. In this study, the effect of the introduction process and following period was assessed for eight weeks using behavioural observations and urinary cortisol to measure activity of the HPA axis.
The aim of this investigation was to examine the behavioural and adrenocortical responses of all adult and sub adult group members to the introduction of a replacement adult male into an established breeding spider monkey group within a zoo environment. Behavioural observations and the collection of urine commenced two weeks prior to the arrival of the new male and continued for six weeks following his arrival to assess changes in behaviour, proximity and cortisol levels of individuals within the group throughout the introduction period. To reduce the impact a gradual introduction technique was used to allow for a period of familiarisation prior to the full introduction. The new male was housed in a separated area of the enclosure that allowed the opportunity for full visual and tactile contact with the rest of the group (see Chapter 2, Figure 2.4 and 2.5).
6.5.1 Subjects The group size over the study period was made up of 11 individuals, briefly dropping to 10 between the relocation of the original male and the introduction of the new male. The group was made up of one adult male, five adult females, four juveniles and one infant (see Table 6.1).
6.5.2 Procedure 22.214.171.124 Urine collection Urine collection occurred throughout the eight week study period and followed the protocol outlined in Chapter 2.
*Birth date unknown, wild caught, arrived in Chester Zoo in 1982 as adult individual likely 10 years of age at that time 126.96.36.199 Observations The study period was divided into six separate time categories to reflect the ongoing events. Pre-introduction represented the control period before the resident male was moved (27.02.08 – 06.03.08); ‘no male’ represented the time of no adult male in the group i.e. between when the resident male was relocated and before the arrival of the new male, Poy (07.03.08 – 10.03.08); ‘Poy back’ was the time following the arrival of the new male when he was housed separately at the back of the enclosure, but in full visual and potential tactile contact with other members of the group (11.03.08 – 13.03.08); ‘introduction’ was the period when the new male was introduced with the rest of the group, and included a temporary separation following a fight (13.03.08 – 17.03.08); ‘inside’ was the time when the group was housed together, but Poy was not permitted to go to the outside enclosure, therefore there were brief periods for a few hours during several of the observation days when the male was separated into the rear part of the enclosure to allow the rest of the group access to the outside enclosure (17.03.08 – 06.04.08); and finally ‘outside’, which represented full 24hr integration of the male with the group with full access to indoor and outdoor enclosures (07.04.08 – 27.04.08). Each focal animal was observed for a 10 minute focal with continuous recording used in which its behaviour and proximity to other members of the group was recorded (see Table 6.2 and 6.3) (Martin & Bateson, 2007). The data were entered onto a check sheet (see Appendix
D) to be later entered for data analyses. I carried out the observations from one of three time periods, 08:00 – 10:59, 11:00 – 13:59 and 14:00 –17:00. This was done to take account of any differences in behaviour patterns of the spider monkeys during the day that could have been affected by external factors, such as keeper routine and visitor numbers. Each individual was selected at random, but only observed once per time period. The number of focal observations per time period was not fixed. Of the 61 days of the study observations occurred on 49 days. The ethogram follows Schaffner and Aureli (2005) developed for this same group of spider monkeys. Due to the infrequency of agonistic behaviours and embraces they were not included on the ethogram, although were recorded as they occurred.
6.5.3 Analyses Over the study period, urinary cortisol levels and behaviour were investigated using linear mixed models (LMM’s). LMM’s were applied following the procedures laid out in Chapter 5. For this investigation, Maximum likelihood (ML) method was used. The state behaviours and proximity data were converted to proportions and transformed for statistical analyses using ARCSINE square root transformations (Martin & Bateson, 2007). Mean proportions were calculated by taking the proportion for each behaviour from each individual for each time period and then calculating the overall mean proportion from each phase of the study.
Scratches were analysed as mean number per time category. These values were entered as a continuous dependent variable, time as the fixed variable and identity as the random variable in the models. The cortisol values were also entered as a continuous variable. The resident individuals were analysed separately from the data obtained from the resident and new male. When a main effect of time was detected, post hoc comparisons were carried out using Bonferroni’s test. An alpha level of 0.05 was adopted for all inferential statistical analyses.
Table 6.2 Ethogram.
Behaviour Definition Individual/self-directed behaviour
The data for males were presented as descriptive statistics (mean ± standard deviation). Differences in the behaviour of the new male over time or between the new male and the resident male were determined when the mean rate ± standard deviation did not overlap between any two sets of scores.
The first set of analyses examined changes within individuals in the three proximity categories of ‘no contact’, ‘proximity’ (within arms reach) and ‘in contact’ over the six time categories. The second set of analyses examined changes over time in the six different behaviours of ‘feeding’, ‘resting’, ‘locomotion’, ‘auto grooming’, ‘groom other’ and ‘self scratching’. The third and final set of analyses examined changes in levels of cortisol over the study period. Post hoc tests used pairwise comparisons and adopted the Bonferroni’s correction.
6.6.1 Social behaviours Analyses demonstrated that the proportion of time individual residents spent in ‘no contact’ with another individual changed significantly over the six time categories [F (1, 5) = 5.666, p.001, AIC = 229.377, see Figure 6.1A]. Post hoc analysis revealed a significant decrease from ‘Pre-introduction to ‘Poy back’ (p
0.5 0.4 0.3 0.2
.005), and ‘Pre-introduction and inside’ (p 0.005). There was also an increase from ‘Poy back’ to ‘outside’ (p.001) and ‘inside’ and ‘outside’ (p.05). Not surprisingly, the change in the proportion of time females were not in contact appeared to correspond to the pattern of proportion of time they spent in contact with each other, which also changed over time [F (1, 5) = 7.277, P.001, AIC = 274.757, see Figure 6.1B]. Post hoc analysis revealed a significant increase between ‘preintroduction and ‘Poy back’ (p 0.05), and ‘pre-introduction’ and ‘inside’ (p 0.05); and a significant decrease between ‘Poy back’ and ‘outside’ and ‘inside’ and ‘outside’ (both comparison, p.01). There was no significant difference detected over time for proximity [F (1, 5) = 1.80, P = 0.114, AIC = 99.179, Table 6.4].
For the males there was a difference in the proportion of time spent in ‘no contact’ with another individual for the two males (see Table 6.5). The newly introduced male spent a greater proportion of his time in ‘no contact’ than did the previous resident male (Table 6.5). There was also a change in the proportion of time the former resident male spent ‘in contact’ with other group members compared to the new male (Table 6.5) The new male spent almost no time in contact with residents across the duration of the study period, however as time progressed, he spent a greater proportion of his time in proximity to other residents, particularly during the time period when all individuals were confined to the indoor part of the enclosure (Table 6.5).
0.1 0.08 0.06 0.04
For resident females, mean values of behaviour revealed that ‘groom other’ showed a significant change over time [F (1, 5) = 2.422, P.0001, AIC = -49.718, see Figure 6.2], with a significant decrease between the time category ‘inside’ and ‘outside’ (p.05). For the males there was more ‘groom other’ behaviour by Poy during the inside phase, which then reduced in the outside phase (see Table 6.4).
6.6.2 Self directed behaviour For the resident females, ‘self scratching’ showed a significant change over time [F (1, 5) = 5.936, p.001, AIC = 667.353, see Figure 6.3] with significant increases from ‘before’ to ‘Poy back’ (p.005) and ‘no male’ to ‘Poy back’ (p.005), and a decrease from ‘Poy back’ to ‘inside’ (p.005) and ‘outside’ (p.005).
Locomotion approached a significant effect over time, but none of the post hoc comparisons yielded significant differences [F (1, 5) = 2.099, p = 0.07, AIC = see Table 6.4].
For the males, Poy had a higher level of ‘self scratching’ following his arrival as compared to Ric, during the ‘Poy back’ and introduction phases, which then reduced over the study period to levels similar to Ric’s in the pre-introduction phase ***
Figure 6.3 Mean proportion of observation time in female resident monkeys for self-scratching.
Vertical lines depict standard errors of the means. See text for explanation of significant effects.
(Table 6.5). Although frequencies were very small, lower levels of ‘auto grooming’ were recorded for Poy than Ric.
6.6.3 Individual behaviour For female residents feeding [F (1, 5) = 0.311, p = 0.91, AIC = 112.983, see Table 6.4], resting [F (1, 5) = 5.936, p = 0.24, AIC = 175.007, see Table 6.4], and auto grooming [F (1, 5) = 5.936, p = 0.449, AIC = -124.180, see Table 6.4] did not vary over the different phases of the study, although locomotion did approach significance [F (1, 5) = 2.098, p = 0.067, AIC = -130.008, see Table 6.4].
For the males, the mean values of behaviour differed between the values of the previous male (Ric), and new male (Poy) in the amount if time spent ‘feeding’ with Poy spending considerably longer period of time feeding or foraging for food;
and an increase in ‘locomotion’ when Poy had access to outside (see Table 6.5).
6.6.4 Cortisol Cortisol values for resident females revealed significant fluctuations over time [F (1, 5) = 1.928, p =.009, AIC = 388.858, see Figure 6.4], although post hoc comparisons did not reveal significant differences among the different phases of the study. Although the values for the males showed a higher level of cortisol in Poy compared to Ric the difference was not meaningful. However, Poy had a higher cortisol level when he was in the back of the enclosure compared to when he was fully introduced (see Table 6.6).
This study represents a multidisciplinary approach to assessing the stress response in a group of spider monkeys following the changing of a new male in an established group. Over the eight week period there were some changes in behaviour.
Of particular interest is the initial increase and then decrease of the scratching behaviour among the resident females and the newly introduced male.