Body mass and growth rates in captive chimpanzees (Pan troglodytes) cared for in African wildlife sanctuaries, zoological institutions, and research facilities

Abstract Captive chimpanzees (Pan troglodytes) mature earlier in body mass and have a greater growth rate compared to wild individuals. However, relatively little is known about how growth parameters compare between chimpanzees living in different captive environments. To investigate, body mass was measured in 298 African sanctuary chimpanzees and was acquired from 1030 zoological and 442 research chimpanzees, using data repositories. An analysis of covariance, adjusting for age, was performed to assess same‐sex body mass differences between adult sanctuary, zoological, and research populations. Piecewise linear regression was performed to estimate sex‐specific growth rates and the age at maturation, which were compared between sexes and across populations using extra‐sum‐of‐squares F tests. Adult body mass was greater in the zoological and resarch populations compared to the sanctuary chimpanzees, in both sexes. Male and female sanctuary chimpanzees were estimated to have a slower rate of growth compared with their zoological and research counterparts. Additionally, male sanctuary chimpanzees were estimated to have an older age at maturation for body mass compared with zoological and research males, whereas the age at maturation was similar across female populations. For both the zoological and research populations, the estimated growth rate was greater in males compared to females. Together, these data contribute to current understanding of growth and maturation in this species and suggest marked differences between the growth patterns of chimpanzees living in different captive environments.


| INTRODUCTION
Obesity is related to a multitude of comorbidities in captive chimpanzees (Pan troglodytes), including hypertension (Andrade et al., 2011;Ely et al., 2013;Videan et al., 2007), insulin resistance (Andrade et al., 2011), cardiovascular disease (Seiler et al., 2009;Strong et al., 2020), metabolic syndrome (Nunamaker et al., 2012;Steinetz et al., 1996) and inflammatory disease (Nehete et al., 2014;Obanda et al., 2014). Accordingly, appropriate management of body mass is an important consideration for the physical health and longevity of this species in captivity (Obanda et al., 2014). Successful management of body mass in captive chimpanzees, however, requires a comprehensive understanding of the normative growth pattern. Such data have only been well-characterized in research populations, which have shown that females are typically lighter and attain body mass maturation earlier than males (Gavan, 1953;Grether & Yerkes, 1940;Hamada et al., 1996;Leigh & Shea, 1996). In contrast, comparatively few reports have examined the body mass of zoological (Vančata & Vančatová, 2002) or sanctuary (Cole et al., 2020;Obanda et al., 2014) populations. Nonetheless, a recent comparison between research chimpanzees and those living in African sanctuaries has identified the latter have a lower body mass and a slower rate of weight gain before maturation of body mass (Cole et al., 2020). However, it is currently unknown how the growth characteristics of zoological chimpanzees compare to that of research or sanctuary populations.
Growth is influenced by numerous factors, including physical activity and diet (Rogol et al., 2000) which vary across captive living environments (i.e., zoological institutions, research facilities, and African sanctuaries). In many of the sanctuaries in Africa, chimpanzees have access to large forested enclosures 10-100 times the size of the largest zoological (Wobber & Hare, 2011) or research enclosure. The smaller enclosure size in both zoological and research facilities may translate into lower physical activity levels, which in turn, could result in an earlier onset of maturation as has previously been documented in humans (Bacil et al., 2015). Further, a staple portion of the zoological and research chimpanzee diet is commercial monkey biscuit, which is of higher caloric density than native vegetation (AZA Ape TAG, 2010) that sanctuary chimpanzees primarily consume. The size and composition (i.e., male to female ratio and hierarchy) of social groupings also vary across the different captive environments, and accordingly, within-group competition for food is likely to vary (Markham & Gesquiere, 2017). Groups in African sanctuaries can contain up to 50 individuals, and as greater group sizes are associated with complex social hierarchies that have increased competition for resources (Markham & Gesquiere, 2017), competition for food is likely to be greater in African sanctuaries compared with zoological and research facilities, where group size is smaller (e.g., Andrade et al., 2011;Birkett & Newton-Fisher, 2011;Nunamaker et al., 2012;Videan et al., 2007). Consequently, the variations in diet, social grouping, and physical activity across captive living environments could influence adult body mass, the growth rate, and/or the timing of body mass maturation (i.e., asymptotic adult body mass). The aims of this study were, therefore, two-fold (i) to compare adult body mass, growth rates, and ages at maturation for body mass between sanctuary, zoological, and research chimpanzees; (ii) to compare these growth parameters between sexes, within each population. It was hypothesized that in comparison to their zoological and research counterparts, sanctuary chimpanzees would be lighter, have a slower rate of growth, and have an older estimated age at body mass maturation. Additionally, it was hypothesized that across all three populations, body mass would be greater, and maturation would be attained at an older age, in males compared with their female counterparts.

| Sanctuary population
Single measurements of body mass were obtained in 298 chimpanzees (P. troglodytes) during routine health checks at three African rehabilitation sanctuaries (Tchimpounga Chimpanzee Rehabilitation Centre, Congo; Chimfunshi Wildlife Orphanage, Zambia; Tacugama Chimpanzee Sanctuary, Sierra Leone; Table 1). The three sanctuaries are members of the Pan African Sanctuary Alliance (PASA) and the chimpanzees were cared for in accordance with the recommendations of the PASA operations manual (Farmer et al., 2009). The majority of the chimpanzees (n = 252) were wild-born orphans confiscated by wildlife authorities, commonly at the age of approximately 1-3 years, although some were older at arrival. The age of these individuals was estimated on arrival by highly experienced sanctuary veterinarians using dental development and records obtained during the confiscation (Cole et al., 2020;Wobber et al., 2010). For those chimpanzees born in captivity (n = 46), their precise age was used. Chimpanzees were housed in semi-free ranging enclosures spanning from 2.5 to 77.0 hectares, in mixed-sex and mixed-age groups of 10-50 individuals. In addition to the native vegetation within the enclosures, the chimpanzees were supplemented routinely throughout the day with seasonal, locally obtained fruits and vegetables. While the subspecies were not known for every chimpanzee, the sanctuary population was likely to be of mixed subspecies; many chimpanzees at Tchimpounga were P. t. troglodytes, whereas the majority at Chimfunshi were thought to be P. t.
schweinfurthii and those at Tacugama to be P. t. verus. Body mass was measured using either a calibrated hanging scale (Salter Brecknell, 235-6S) or Seca electronic weighing scales (Seca, Vogel, and Halke) and was assessed to the nearest 0.1 kg. All procedures and protocols involved in this study have been endorsed by the PASA Advisory Council and Cardiff Metropolitan University, UK, approved by the British and Irish Association of Zoos and Aquariums and ethically approved by the University of British Columbia, Canada. Association of Zoos and Aquariums (WAZA), or which held a WAZA-affiliated association. These data were initially screened for obvious data input errors, and were then checked for outliers using the robust regression and outlier removal (ROUT) method (Q set to 1%) in GraphPad Prism (GraphPad Prism for Windows, version 8.0.1); however, this process did not identify any statistical outliers.

| Zoological population
To correspond with the age range of the African sanctuary population (0-32 years in males and 0-38 years in females), male and female zoological chimpanzees older than 32 years and 38 years of age, respectively, were excluded from the study. This was to ensure that the data sets were age comparable, and therefore any statistical findings were not due to differences in the age range between the populations. A single body mass measurement was randomly selected from each chimpanzee in the database, using the RAND function in Microsoft Excel (2016), to prevent any confounding effects of repeated measures. A total of 409 males and 621 females were included in the final analysis (Table 1). Unfortunately, no detailed information was available regarding the housing or diet of this population. Subspecies information was also not available for many of the individuals, however, the information that was available would suggest that, similar to the sanctuary population, the subspecies was mixed.

| Statistical analysis
Differences in mean adult body mass were assessed within sex across the three populations (i.e., sanctuary, zoological, and research) and between sexes within each population, using a two-way analysis of covariance (ANCOVA) with Dunn-Sidak post hoc analyses, for which group and sex were independent variables and age was the covariate. The size of the effect was estimated using Cohen's d; here, where M 1 = mean of group 1, M 2 = mean of group 2, and s P = pooled standard deviation (SD) between groups 1 and 2 (Cohen, 1988). An effect size of ≤0.2 was deemed a small effect, ≤ 0.5 a medium effect, and ≥0.8 a large effect. ANCOVA was performed using the Statistical Package for the Social Sciences version 26 (SPSS Inc.).
Alpha was set at p < .05, and data were expressed as the mean difference (±SD) and 95% confidence intervals (CIs).
Growth rates and ages at maturation for body mass were estimated using sex-specific piecewise least-squares linear regressions in Graph-Pad Prism. An unconstrained analysis was chosen to model body mass and identify a pair of best fit lines and the breakpoint between these two lines (Altmann & Alberts, 2005). The slope of the regression line to the left of the breakpoint can be used as an estimate of growth rate (Altmann & Alberts, 2005;Huck et al., 2011) and the breakpoint as the estimated age at which maturation of body mass occurs (Leigh, 1994;Leigh & Terranova, 1998). This breakpoint was used to define the adult populations for the ANCOVA described above (i.e., those to the right of the breakpoint were considered adults). The extra-sum-of-squares F-test was used to determine whether growth rate and age at maturation differed across populations, or between sexes within a population.

| Growth rates and ages at maturation
Male sanctuary chimpanzees had a slower rate of body mass growth and attained body mass maturation at an older age compared with their zoological (p < .001 and p = .031, respectively) and research counterparts (p < .001 and p = .014, respectively; Figure 1 and Table 3). In contrast, male zoological and research populations had a similar growth rate and age at maturation (Figure 1 and Table 3). In females, sanctuary chimpanzees also had a slower rate of growth compared with their zoological (p = .018) and research counterparts (p = .007; Figure 1 and Table 3). The rate of growth was similar, however, between female zoological and research chimpanzees (Figure 1 and Table 3). Additionally, the age at maturation did not differ between the three female populations (Figure 1 and Table 3).

| Sex differences in growth parameters
In both the sanctuary and zoological populations, male chimpanzees had a greater adult body mass compared with females (sanctuary: mean difference ± SD = 8.7 ± 10.1 kg, CI = 5.8-11.6 kg, p < .001, d = 0.86; zoological: mean difference ± SD = 7.9 ± 10.0 kg, CI = 6.4-9.5 kg, p < .001, d = 0.79; Table 2). However, there was no sex difference in F I G U R E 1 Comparison of body mass between male (a) and female (b) African sanctuary, zoological, and research populations of chimpanzee (Pan troglodytes). Body mass of sanctuary (represented by dot-dashed lines), zoological (represented by dashed lines), and research chimpanzees (represented by solid lines) were fitted using piecewise least-squares linear regression, with 95% confidence intervals shown (represented by gray area). Individual data points in the sanctuary, zoological, and research populations are represented by triangles, circles, and diamonds, respectively.
T A B L E 3 Growth rate (kg yr −1 ; slope 1) and maturation age (years; breakpoint) of the African sanctuary, zoological, and research chimpanzee (Pan troglodytes) populations, as estimated using piecewise least-squares linear regression adult body mass within the research population (mean difference ± SD = 0.1 ± 9.9 kg, CI = −4.3-4.3 kg, p = .995; Table 2). The rate of growth did not differ between sexes in the sanctuary population; however, males in the zoological and research populations had a greater growth rate compared to their female counterparts (both p < .001; Table 3). The age of body mass maturation was not statistically different between sexes, for any population (Table 3).

| DISCUSSION
The purpose of this study was to compare adult body mass, body mass growth rates, and the ages of body mass maturation between sanctuary, zoological, and research chimpanzees, and to compare these growth parameters between sexes, within each population. The main findings were (1) zoological and research chimpanzees were heavier than those living in sanctuaries; (2) male sanctuary chimpanzees had a slower rate of growth and attained maturation at an older age compared to male zoological and male research chimpanzees; (3) in females, sanctuary chimpanzees also had a slower rate of growth compared with their zoological and research counterparts, however the age at maturation was similar across the female populations; (4) no sex difference was observed for the growth rate in the sanctuary population; whereas, in zoological and research chimpanzees, males had a greater growth rate than females. These data contribute to the current understanding of growth and maturation in this species and suggest that growth patterns may vary between chimpanzees living in different captive environments.

| Differences in adult body mass across captive populations
African sanctuary chimpanzees have previously been reported to have a lower body mass compared to research chimpanzees (Cole et al., 2020). Consistent with these findings, and in agreement with our hypothesis, the present study has also shown that adult body  (Ross, 2014), it is logistically impossible to re-create the size and complexity of the environment that many sanctuary animals experience. Future work should compare physical activity levels between sanctuary and zoological chimpanzees to confirm or refute whether differential opportunity for physical activity influences overall size or rates of growth in different captive populations.
Differences in diet and food availability across captive living environments may also contribute to the greater body mass in zoological and research populations, compared with sanctuary animals.
For example, a staple portion of the zoological and research chimpanzee diet is commercial monkey biscuit (AZA Ape TAG, 2010), which likely provides greater caloric and lower fiber intake than the natural vegetation that sanctuary chimpanzees consume. Additionally, it is possible that portion size (i.e., the amount of food per chimpanzee) differs across the three populations, which could affect body mass.
However, this information was not available across the three populations and so it is not possible to make this direct comparison.
Body mass could also be influenced by the size of the chimpanzee's social group. Larger group sizes are associated with a complex social hierarchy, and lower-ranking individuals may have reduced access to resources compared to more dominant individuals. Accordingly, body mass may be more variable amongst chimpanzees in African sanctuaries, which have much larger group sizes compared to those in zoological and research institutions, where within-group competition is likely lower (Markham & Gesquiere, 2017). Accordingly, it is possible that zoological and research chimpanzees could have a more positive energy balance than sanctuary animals, which may explain the greater adult body mass we have described.

| Differences in growth rate across captive populations
Environmental factors, such as diet (Jarrett et al., 2020) and the energetic costs related to physical activity and foraging (Zihlman et al., 2007), are believed to influence the rate of growth in primates.
As discussed above, both diet and physical activity are likely to differ across captive living environments, which could result in a slower rate of growth in the sanctuary population. However, the influence of environmental factors on growth rate could be further exacerbated in sanctuary chimpanzees by their status as an orphan. Previous research in wild chimpanzees observed a lower muscle mass in orphans compared to nonorphaned individuals (Samuni et al., 2020).
In their study, Samuni et al. (2020) proposed that the compromised growth in orphan chimpanzees could result from a need to allocate energy towards independent travel, foraging, and navigating a complex social hierarchy. In support of this, an exploratory analysis performed in our sanctuary population showed that despite no differences in either the age of body mass maturation or adult body CURRY ET AL.
| 103 mass, the rate of growth was slower in orphans (3.2 kg yr −1 ) compared to those who were sanctuary born (3.6 kg yr −1 ). Whilst this provides useful insight, the sample size of the sanctuary-born cohort used in this exploratory analysis was relatively small and so further research is needed to confirm this finding.  (Martins et al., 2011). These environmental stressors have been shown to affect growth more adversely in males than females (Semproli & Gualdi-Russo, 2007), and could contribute to our findings. A similar sex-dependent relationship has also been observed in humans and rodents, whereby poor nutrition was associated with a greater delay in puberty in males than in females (Kulin et al., 1982;Sánchez-Garrido et al., 2013). However future investigation is required to assess whether a similar sex-dependent relationship is present in chimpanzees.

| Sexual dimorphism
In primates, body mass dimorphism (i.e., that males are heavier than females) can either arise through sex differences in the duration and/ or rate of growth (Setchell et al., 2001). However, Leigh and Shea (1996) have proposed that, in chimpanzees, body mass dimorphism is caused by differences in the rate of growth and no sex differences in growth duration. Whilst our data support this hypothesis in zoological and research populations, no sex differences were observed for growth rate in the sanctuary population. Duration of growth, therefore, may have a comparably greater effect on sexual dimorphism in the sanctuary population. This is supported by the finding that males were estimated to attain body mass maturation approximately one and a half years after females; although this did not reach statistical significance according to conventional analysis. It is possible that this prolonged growth of sanctuary males reflects greater intermale resource competition compared to that in zoological or research institutions, which could result from their larger group size, as has been proposed in other primate species (Leigh & Shea, 1996).

| Study limitations
The piecewise linear regression method adopted in this study was beneficial for identifying the estimated ages at maturation, but it does provide a simplistic view of growth rates by assuming they are constant. Alternative methods, such as pseudo velocity curves (Hamada & Udono, 2002) costs (i.e., gestation and lactation) will influence the growth pattern of female chimpanzees. However, the authors were unable to determine its effects in this study as detailed information is not available across all of our populations.

| CONCLUSION
This study contributes to our current understanding of chimpanzee growth and highlights that growth patterns may vary between chimpanzees living in different captive environments. Chimpanzees in African sanctuaries have a lower body mass than those in zoological and research facilities and a slower growth rate than their research counterparts. Additionally, male sanctuary chimpanzees also had a delayed body mass maturation compared to their zoological and research counterparts, whereas the age of maturation was similar across female populations. These results provide a valuable perspective regarding the influence of living environment on growth and suggest that caution should be observed when extrapolating growth parameters across different captive environments.