Fertility and mortality patterns of captive Bornean and Sumatran orangutans: Is there a species difference in life history? more

Available online at www.sciencedirect.com Journal of Human Evolution 54 (2008) 34e42 Fertility and mortality patterns of captive Bornean and Sumatran orangutans: is there a species difference in life history? Holly B. Anderson a, Melissa Emery Thompson a, Cheryl D. Knott a,*, Lori Perkins b a Department of Anthropology, Harvard University, 11 Divinity Avenue, Cambridge, MA 02138, USA b Collections Department, Zoo Atlanta, 800 Cherokee Avenue SE, Atlanta, GA 30315-1440, USA Received 25 August 2006; accepted 22 May 2007 Abstract Across broad taxonomic groups, life history models predict that increased ecological predictability will lead to conservative investment in reproductive effort. Within species, however, organisms are predicted to have increased reproductive rates under improved environmental conditions. It is not clear how these models apply to closely-related species. In this paper, we examine predictions from these models as applied to variability in reproductive rates between the two species of orangutans, Pongo pygmaeus (Bornean) and Pongo abelii (Sumatran). Orangutans exhibit characteristics of a ‘‘slow’’ life history strategy with large bodies, late age at maturity, low reproductive rates, and long lifespan. Recently, researchers proposed that Sumatran orangutans may have an even slower life history than Bornean orangutans as a result of ecological and genetic differences (Wich et al., 2004). We examined this hypothesis by studying important aspects of life history of both species under conditions of relative ecological stability, in captivity. In this large dataset, there were no significant species differences in age of first or last reproduction, completed fertility, perinatal and postnatal mortality, or female longevity. Bornean orangutans in captivity did have significantly longer interbirth intervals, and male Bornean orangutans had higher survival past maturity. Our results do not support the hypothesis that selection has led to decreased reproductive effort under conditions of increased habitat quality in Sumatra (Wich et al., 2004), and instead suggest that phenotypic flexibility may be particularly important in explaining differences between closely related species. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Reproductive parameters; Interbirth intervals; Reproductive ecology Introduction Life history adaptation can be viewed broadly as selection for ‘‘optimal’’ allocation of energy to the competing demands of maintenance, growth, and reproduction (Gadgil and Bossert, 1970). Thus, reproductive effort in mammals shares complex relationships with other variables such as body size, mortality rates, and ecological consistency. Primates exemplify selection for ‘‘slow’’ life histories characterized by late age at maturity, long lifespans, and long interbirth intervals (IBIs) (Charnov and Berrigan, 1993). Orangutans are at one extreme: among * Corresponding author. E-mail addresses: handers@post.harvard.edu (H.B. Anderson), memery@ fas.harvard.edu (M. Emery Thompson), knott@fas.harvard.edu (C.D. Knott), Lori410@mindspring.com (L. Perkins). 0047-2484/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jhevol.2007.05.014 the largest and longest-lived of primates (Wich et al., 2004), they have the longest IBIs of any mammal regardless of size (Galdikas and Wood, 1990; Knott, 2001). Interestingly, field reports suggest that there may be significant variation in reproductive rates between the two orangutan species (Pongo pygmaeus, Bornean orangutan, and P. abelii, Sumatran orangutan), and that such variation may be linked to habitat differences between the two islands (Wich et al., 2004). The available evidence suggests that IBIs in Sumatran orangutans are longer than in Bornean orangutans. Sumatran orangutans at Ketambe have a mean IBI of 111.5 months (9.3 years, n ¼ 23), which is significantly longer than the mean IBI of 92.6 months (7.7 years, n ¼ 23) for Bornean females at Tanjung Puting (Galdikas and Wood, 1990; Tilson et al., 1993; Wich et al., 2004). Although the samples are very small, reported mean IBIs of 98.4 months (8.2 years, H.B. Anderson et al. / Journal of Human Evolution 54 (2008) 34e42 35 n ¼ 8 with estimated birth dates) from Suaq Balimbing in Sumatra (van Noordwijk and van Schaik, 2005) and 86.4 months (7.2 years, n ¼ 8) from Gunung Palung in Borneo (Knott et al., in press) also support this difference. Wich et al. (2004) hypothesized that the IBI difference is one indication of a general divergence in life history strategies that has resulted from differences in ecological stability between the Bornean and Sumatran habitats. Although Bornean and Sumatran orangutans were only recently classified as separate species, genetic studies estimate a divergence time of 1.1 to 5 million years (Xu and Arnason, 1996; Groves, 2001; Warren et al., 2001; Steiper, 2006). Recent ecological work emphasizes differences in habitat quality between the two islands. Consistent with the prevalence of rich volcanic soils on Sumatra, fruit is more readily available and stable in Sumatran study sites (Knott, 1998; Delgado and van Schaik, 2000; Fox et al., 2004; Knott and Kahlenberg, 2006). Thus, habitat quality and life history traits are theorized to form a gradient from west to east, correlating with other features such as cranial capacity, which is largest in P. abelii and smallest in the easternmost subspecies P. pygmaeus morio (Taylor and van Schaik, 2007). Wich et al.’s (2004) hypothesis is consistent with the models of evolved life history differences between species. In comparisons across taxonomic groups, species with relatively rich habitat quality face lower mortality risks and reproduce at a slower pace (Gittleman, 1986; Charnov, 1991; Charnov and Berrigan, 1993). In such species, rather than maximization of reproductive rates, natural selection is expected to favor relatively slow reproduction and increased investment in survival and in individual offspring. However, life history strategies also incorporate a norm of reaction as a means to accommodate random ecological effects (Ricklefs and Wikelski, 2002). Under life history models for short-term allocation of energy, large-bodied species like orangutans are expected to be able to increase their reproductive effort when energy intake is high (van Noordwijk and de Jong, 1986; Zera and Harshman, 2001). This responsiveness is mediated by the organism’s physiology, particularly the endocrine system (Ketterson and Nolan, 1992; Ricklefs and Wikelski, 2002; Ellison, 2003; Knott, 2005). Reproductive ecologists describe strong positive links between energy availability and reproductive function (Wasser and Barash, 1983; Schneider and Wade, 2000). For example, intraspecies comparisons among primates generally reveal increased reproductive rates among captive or nutritionally-advantaged populations relative to populations living in natural ecological conditions (Asquith, 1989; Borries et al., 2001; Knott, 2001; Altmann and Alberts, 2003). This response is particularly elegant for the hominoids, in which reproductive function has a graded relationship with variance in energy balance (Ellison, 2003; Ellison et al., 1993; Knott, 2001, 2005; Emery Thompson, 2005; Emery Thompson and Wrangham, in review). Among wild orangutans at Gunung Palung, Borneo, Knott (1998, 1999) demonstrated that females had higher levels of ovarian hormones and increased conception success when their energy balance was most positive. Thus, while interspecies models of life history predict a risk-aversive strategy of conservative reproductive effort with increased ecological certainty, models of phenotypic flexibility predict the opposite effect. Orangutans provide a valuable model for the study of contrasting life history predictions. In relation to other primates, they conform to predictions for the correlation between ecological stability and slow life histories. However, within populations, they meet predictions from reproductive ecology: energy availability correlates with increased reproductive rates. It is not clear how either model might apply to a contrast between the two closely-related orangutan species. While preliminary evidence on orangutan interbirth intervals indicates slower reproductive rates with increased habitat quality (Wich et al., 2004), we find a number of problems with this argument (Knott et al., in press). First, sample sizes of wild orangutan birth intervals are still relatively small. The best published sample from Borneo (Tanjung Puting) consists of 12 complete (birth to birth) and 11 incomplete (no second birth at time of study) intervals from 11 female orangutans over 17 study years (Galdikas, 1985); the best Sumatran sample (Ketambe) is based on 16 complete and 7 incomplete intervals from 9 females over 32 study years (Wich et al., 2004). While these samples were sufficient to produce a statistically significant difference in complete birth interval length between populations, it is difficult to interpret differences observed over study periods that are short relative to the lifespan of the species. While infant mortality is reportedly low in the wild (Wich et al., 2004, Knott, unpublished data), the relatively solitary nature of these species makes it feasible that missed births and early perinatal deaths could skew estimates based on small samples. In addition, we cannot easily determine whether these are systematic differences between populations and species, or if they represent stochastic variation in habitat quality, demography, or forest disturbance. A second problem is that life history theory predicts correlated differences in other life history traits (Eisenberg, 1981; Harvey and Clutton-Brock, 1985; Harvey and Zammuto, 1985; Read and Harvey, 1989; Charnov, 1991) that have not been observed yet in wild orangutans. No marked difference is evident in age of first birth between species (Pongo pygmaeus: 15.7 years, Tilson et al., 1993; P. abelii: 15.4 years, Wich et al., 2004), though these sample sizes are also very small. The few adult body weights that are available from the wild do not indicate a size difference: 86.3 (80e91) kg (n ¼ 4) and 38.8 (33e45) kg (n ¼ 5) for Bornean males and females, respectively, and 86.2 kg (n ¼ 1) and 38.3 (34e43) kg (n ¼ 5) for Sumatran males and females, respectively (Markham and Groves, 1990). Cranial analyses suggest more diversity in size and shape within the island of Borneo than between species (Groves et al., 1992; Taylor and van Schaik, 2007), as well as a strong influence of ontogenetic response to altitudinal differences (Courtenay et al., 1988). Most troubling, there is no suggestion as yet of a significant difference in either adult or juvenile mortality between species. Orangutans seem to have exceedingly low infant and juvenile mortality compared to other apes (Wich et al., 2004, Knott, unpublished data). While this is consistent with life history predictions 36 H.B. Anderson et al. / Journal of Human Evolution 54 (2008) 34e42 about orangutans compared to other apes, there is no evidence that infant mortality differs enough to drive the evolution of life history differences between orangutan species. Here, we present life history data from captive orangutans in U.S. and international facilities as a means to experimentally control for ecological conditions. Specifically, we compare the ages of females and males at their first and last reproduction, IBIs, total numbers of offspring per female, infant mortality rates, and longevity between the two species. The life history hypothesis (Wich et al., 2004) predicts that, while captive orangutans may experience accelerated reproduction and greater survival overall, significant species differences in life history traits should be maintained; furthermore, differences in birth intervals should be accompanied by differences in correlated life history variables, such as maturation rates and survival. Alternately, the reproductive ecology hypothesis (Knott, 2001) predicts that, assuming no systematic differences in quality of care, captive Bornean and Sumatran orangutans should have accelerated reproductive schedules relative to their wild counterparts and show little or no species difference in life history traits. While we have taken care in our methodology to minimize the influence of captive breeding manipulations on the data reported, we are aware that captive data on reproduction are problematic. However, because of the relatively recent designation as separate species, Bornean and Sumatran orangutans often shared facilities and caregivers and were even interbred. Therefore, we have no expectation that these species experienced different captive environments. Methods All data are from the International Studbook of the Orangutan, which contains information from 313 institutions worldwide on 2,566 captive orangutans, with birthdates between 1901 and 2005. Our data set included information from when the studbook was last published in 1994 (Perkins, 2001), as well as data compiled through 5 December 2005. These data included each individual’s ID number, sex, species, dam and sire ID number, birth date, and death date when known. We also had a list of individuals (both males and females) known to have been sterilized, although it is a reasonable assumption that not all sterilized orangutans had been reported. All orangutans were classified by species as Sumatran, Bornean, hybrid, or unknown. We analyzed records for 881 Sumatran orangutans (446 females, 418 males, and 17 individuals of unknown sex) and 1,035 Bornean orangutans (499 females and 513 males, and 23 individuals of unknown sex). Of these subjects, 39.9% (n ¼ 764) were born in the wild. Wild-born individuals were typically acquired as juveniles, and their birth dates were estimated by zoo staff at the time of acquisition. Therefore, data from orangutans that spent their entire lives in captivity are the most reliable and relevant to our hypotheses. Thus, we excluded wild-born individuals from analyses that required age accuracy (age at first and last reproduction, lifespan, and total number of offspring by age category) but not from IBI analyses. All life history analyses were calculated with ages and birth data as of 5 December 2005. Where feasible, we used KaplanMeier analyses for right-censoring data. Chi-square statistics reported for these analyses represent the outcome of Breslow Generalized Wilcoxon tests (SPSS, Inc.); this is the most conservative test, but can adequately accommodate these sample sizes and censoring proportions (Breslow, 1970; Benedetti et al., 1982). Results for female age at first and last reproduction reflect the ages at the time of the birth. Male age of reproduction refers to the approximate date of fertilization; the difference between the sire’s birth date and the offspring’s birth date was corrected for the 245 day mean gestation period for orangutans (Nowak, 1999). Age at first reproduction Age at first birth was evaluated with a Kaplan-Meier survival analysis. In order to best exclude the influence of primary sterility or the use of contraception on this analysis for females, we determined the latest age of first birth for a captive dam of known age (25.6 years) and excluded females that had not given birth by this age. Similarly, we excluded males that had not sired an offspring by the latest known age of first male reproduction (28.1 years). Age at last reproduction We calculated the mean age at last reproduction with the understanding that these values are likely to systematically underestimate the physiological capacity to reproduce (for both species) due to contraception practices. We included only males and females that had survived to the mean calculated age of first reproduction (16.0 years for females and 18.6 years for males) and had produced at least one offspring. Birth rates We calculated interbirth intervals using Kaplan-Meier analyses, including both open and completed intervals when the first infant survived for at least four years. We determined the maximum completed interval (203.2 months) and excluded the 33 (of 1,240) incomplete intervals that exceeded this length; these females were likely to have been sterile, isolated from mates, or contracepted. As a second test, we calculated the number of offspring born to each dam according to her current age or age at death, and compared fertility of the two species for each decade of life using Mann-Whitney U tests. Although contraception and repeated infant removal may be expected to stretch our range of results in both directions, mean figures provide realistic estimates of orangutan fertility. Infant mortality We separated infant mortality into two categories: (1) prenatal and neonatal mortality, which we defined as all deaths occurring before (stillbirths or abortions) or during the day H.B. Anderson et al. / Journal of Human Evolution 54 (2008) 34e42 37 of birth, and (2) postnatal mortality, which we defined as all deaths of infants after the day of birth and before four years of age. These analyses included all infants born into captivity before 5 December 2001, thus excluding individuals for which it was unknown whether they survived to four years of age. Longevity We evaluated lifespans of captive-born orangutans using Kaplan-Meier survival analyses. We provide four estimates based on analyses of mean lifespan for individuals surviving (1) the first day of life; (2) the first year of life; (3) the period of infant dependency (4 years); and (4) until the mean age of reproductive maturity. Results Ages at first and last reproduction mothers (captive-born ¼ 69.2 months, wild-born ¼ 53.7 months; c2 ¼ 11.62, df ¼ 1, p ¼ 0.001). We therefore treated IBIs from captive-born and wild-born individuals separately for between-species comparisons. Among captive-born individuals, Bornean mothers had mean IBIs of 75.3 months Æ 4.9 SE (ntotal ¼ 108, ncensored ¼ 27), while Sumatran mothers had mean IBIs of 69.2 months Æ 5.4 SE (ntotal ¼ 122, ncensored ¼ 27). The birth intervals were significantly shorter for captive-born Sumatran than for Bornean mothers (c2 ¼ 4.40, df ¼ 1, p ¼ 0.04, Fig. 1a). Among wild-born individuals, Bornean mothers had mean IBIs of 57.8 months Æ 3.7 SE (ntotal ¼ 249, ncensored ¼ 46), while Sumatran mothers had mean IBIs of 53.7 months Æ 2.9 SE (ntotal ¼ 308, ncensored ¼ 46). Thus, there was a smaller absolute difference in IBIs of wild-born mothers than we observed in captive-born mothers, and there was not a statistically significant species difference (c2 ¼ 0.06, df ¼ 1, p ¼ 0.8, Fig. 1b). Fertility Sumatran females had a mean age of first birth of 16.4 years, which was not significantly different from the age of first birth of Bornean females (15.5 yrs., Table 1, c2 ¼ 0.47, df ¼ 1, p ¼ 0.49). Similarly, male age of first reproduction did not differ significantly between species; Sumatran males sired their first offspring at an average age of 19.4 years, while Bornean males first reproduced at 17.8 years (c2 ¼ 1.44, df ¼ 1, p ¼ 0.28). Female age of last reproduction averaged 26.2 years for Sumatran orangutans and 29.0 years for Bornean orangutans (Table 1). Male age of last reproduction averaged 26.3 years for Sumatran orangutans and 24.7 years for Bornean orangutans. Neither difference was statistically significant (females: c2 ¼ 0.26, df ¼ 1, p ¼ 0.61; males: c2 ¼ 0.71, df ¼ 1, p ¼ 0.40). Interbirth intervals Among Bornean mothers, captive-born individuals had significantly longer mean IBIs than wild-born individuals (captive-born ¼ 75.3 months, wild-born ¼ 57.8 months; c2 ¼ 23.77, df ¼ 1, p < 0.001). The same was true for Sumatran Deceased, captive-born Sumatran mothers had a mean completed fertility of 2.8 offspring (n ¼ 34, range ¼ 1e11). Deceased, captive-born Bornean mothers had similar fertility: 2.5 offspring (n ¼ 20, range ¼ 1e6, U ¼ 327.0, z ¼ À0.241, p ¼ 0.81). It is difficult to interpret differences in completed fertility in this manner because there are many mitigating factors; these include mortality or removal of infants and the reproductive span of the mother, taking into account her age at death. To improve this analysis, we compared the fertility of mothers by age decade depending on their current age or age at death (Fig. 2). There was substantial variation within age groups, which reflects captive breeding practices such as contraception and removal of newborn infants. In this analysis, we found no significant differences within age groups between Bornean and Sumatran mothers ( p > 0.28). Infant mortality Of the Bornean orangutans born into captivity, 13.1% (71 of 542) died during the day of birth, or before birth due to Table 1 Comparison of age at first and last reproduction between Bornean and Sumatran males and females using Kaplan-Meier models Bornean orangutans (P. pygmaeus) n of events n censored Mean Estimate (yrs) Std. error 95% Confidence interval Lower bound Age at first reproduction (yrs) Captive-born females 88 Captive-born males 60 Age at last reproduction (yrs) Captive-born females 15 Captive-born males 12 158 191 69 42 15.5 17.8 29.0 24.7 0.56 0.82 1.46 1.17 14.4 16.2 26.1 22.4 Upper bound 16.6 19.4 31.8 27.0 101 67 24 23 183 214 67 37 16.4 19.4 26.2 26.3 0.55 0.75 1.26 1.75 Sumatran orangutans (P. abelii) n of events n censored Mean Estimate (yrs) Std. error 95% Confidence interval Lower bound 15.3 17.9 23.8 22.9 Upper bound 17.4 20.9 28.7 29.7 0.47 1.44 0.26 0.71 1 1 1 1 0.49 0.23 0.61 0.40 Species comparison (Breslow test) Chi-Square df p 38 H.B. Anderson et al. / Journal of Human Evolution 54 (2008) 34e42 Fig. 2. Total number of offspring born to each captive-born dam, classified by age of death or current age. Boxes indicate the range between first and third quartiles; horizontal line designates the median; vertical spans indicate the range of values; and the boxed number is the number of mothers ( p > 0.28). Likewise, the proportion of postnatal deaths was similar between species. Of Bornean infants who survived their first day of life, 90 of 440 (20.5%) died before the age of 4 years. Of their Sumatran counterparts, 22.4% (113 of 504) died before they reached 4 years of age. Again, the distribution of infant deaths between species did not differ from that expected by chance (c2 ¼ 0.42, df ¼ 1, p ¼ 0.52). Longevity Table 2 summarizes data on the mean lifespan by species. The lifespans of captive-born Bornean females tended to be longer than those of Sumatran females, but survival analyses indicated no significant differences ( p ! 0.13; Fig. 3a). Perinatal and postnatal mortality did not cause lifespans to differ significantly between Bornean and Sumatran males. However, Bornean males that survived past maturity had significantly longer lifespans than Sumatran males (c2 ¼ 11.03, df ¼ 1, p ¼ 0.001). The difference in expected mean lifespan at maturity was only 1.3 years, although the models in Fig. 3b indicate that the greater trend was for approximately 10e20% greater survival rates in adult Bornean males. Although many wild-born males were suspected to be older, the eldest Bornean male of known age was 33.7 years, while the eldest Sumatran male of known age was 40.8 years; thus, the possible Bornean male mean lifespan in this model is truncated relative to the Sumatran sample. Discussion We examined the demography of captive orangutan populations to determine whether, under similar ecological Fig. 1. Interbirth intervals of orangutan mothers, when first infant survived for at least 4 years, Kaplan-Meier analysis. (a) captive-born mothers, species comparison: c2 ¼ 4.40, df ¼ 1, p ¼ 0.04; (b) wild-born mothers: c2 ¼ 0.06, df ¼ 1, p ¼ 0.8. Y-axes reflect the cumulative proportion of intervals that remain open (i.e., no second infant born) at the time indicated by the X-axes. Breaks in the curve indicate completed birth intervals, while markers represent censored data (intervals that were incomplete at the time of data analysis). stillbirths or abortions. Of the Sumatran orangutans, 10.6% (63 of 594) died before or during the day of birth. The distribution of prenatal and neonatal infant deaths between Bornean and Sumatran orangutans was not significantly different from chance (c2 ¼ 1.49, df ¼ 1, p ¼ 0.22). H.B. Anderson et al. / Journal of Human Evolution 54 (2008) 34e42 Table 2 Comparison of longevity of Bornean and Sumatran orangutans using Kaplan-Meier models Bornean orangutans (P. pygmaeus) n of n Mean events censored Estimate Std. 95% (yrs) error Confidence interval Lower Upper bound bound Captive-born females Survived at least 1 day Survived at least 1 yr Survived at least 4 yrs Survived to maturity (16.0 Yrs) Captive-born males Survived at least 1 day Survived at least 1 yr Survived at least 4 yrs Survived to maturity (18.6 Yrs) 84 58 47 23 104 65 49 15 153 153 153 153 122 122 122 122 28.0 31.4 33.1 37.1 19.3 23.3 25.4 30.6 1.39 1.39 1.37 1.33 1.01 1.00 0.95 0.74 25.2 28.7 30.4 34.5 17.4 21.4 23.5 29.1 30.7 34.1 35.7 39.7 21.3 25.3 27.2 32.0 118 85 72 33 149 103 84 48 151 151 151 151 109 109 109 109 23.6 26.8 28.3 33.0 18.8 22.8 24.9 29.3 1.04 1.02 0.99 0.92 1.03 1.06 1.06 1.05 Sumatran orangutans (P. abelii) n of n Mean events censored Estimate Std. 95% (yrs) error Confidence interval Lower Upper bound bound 21.5 24.8 26.4 31.2 16.8 20.7 22.8 27.3 25.6 28.8 30.2 34.8 20.8 24.9 27.0 31.4 1.41 1.64 2.31 0.91 0.36 0.85 1.04 11.03 1 1 1 1 1 1 1 1 Species comparison (Breslow test) Chi-Square df p 39 0.24 0.20 0.13 0.34 0.55 0.36 0.31 0.001 conditions, there was any evidence of systematic divergence in life history strategies between Sumatran and Bornean orangutans. We found no statistical species difference in the age of first or last reproduction in males or females, in perinatal and postnatal mortality rates, or in the completed fertility of mothers. Lifespans of the two species were similar under captive conditions, although captive-born Bornean males lived significantly longer past the age of maturity than Sumatran males. In addition, Bornean mothers had longer IBIs than Sumatran mothers; this difference was small, but significant. Thus, our data on orangutans of known age suggest no major differences in life history parameters between Bornean and Sumatran individuals under captive conditions. The small differences that do exist are in the direction opposite to that predicted by Wich et al. (2004) and by the preliminary wild data, suggesting that any genetic change may have led to increased, not decreased, reproductive effort with increased habitat quality. The convergence of life history parameters between captive individuals of the two species supports the role of energy availability as a primary determinant of reproductive capability, as suggested by models of reproductive ecology. However, we cannot rule out the alternate hypothesis that each species has evolved a different norm of reaction in response to ecological change. Under this scenario, life history parameters might converge because Sumatran orangutans experience a more intense response to increased energy availability than do Bornean orangutans. However, this is unlikely because wild studies have suggested that Sumatran orangutans show little reproductive responsiveness to minor variation in food availability (Wich et al., 2006). Curiously, while we found little evidence for species differences, we found that wild-born individuals had faster reproductive rates than captive-born individuals. This difference likely reflects breeding initiative on the part of zoo management. Caregivers have intentionally promoted the reproduction of orangutans born in the wild more than those born in captivity because they provide a greater contribution to the genetic diversity of the captive population (Perkins, pers. comm.). Our results provide important contrasts to data from orangutans living their entire lives in the wild. The 4.5e6 year IBI that characterized the captive population is shorter than even the lowest estimates from wild populations (7.2 yrs., Gunung Palung, Knott et al., in press; 6.1e6.5 yrs, Kutai/Kinabatangan, Wich et al., in press). While wild interbirth interval figures may be inflated by unobserved births, even the shortest birth intervals recorded are over 5 years (Knott, 2001). Faster reproductive rates have been observed in captive versus wild chimpanzees (Tutin, 1994; Knott, 2001). Bonobo and gorilla IBIs, on the other hand, change little in captivity versus the wild (Tutin, 1994; de Lathouwers and van Elsacker, 2005; Robbins et al., 2006). This may indicate that these species have decreased plasticity in life history variation. However, it has been suggested that reduced feeding competition relative to chimpanzees leads bonobos and gorillas to reproduce near their maximum rate in wild populations (de Lathouwers and van Elsacker, 2005). Alternatively, provisioning in wild bonobo poplations (de Lathouwers and van Elsacker, 2005) and captive breeding problems in gorillas (Sievert et al., 1991) might be expected to reduce the observed IBI difference in these species between captivity and the wild. Contrary to expectation, our data on the infant and adult mortality of captive orangutans did not indicate improved survivorship over the wild. Thirty percent of captive offspring died at birth or before the age of 4 years. By comparison, infant mortality reported from Ketambe, in Sumatra, was 11% for females and 3% for males (P. abelii, Wich et al., 2004). At Gunung Palung, Borneo, no infant deaths have been recorded in over 45,000 hours of observation (Knott, unpublished). However, early data from Suaq Balimbing, in Sumatra, suggest infant mortality commensurate with that of the captive population (Wich et al., 2004). Studies of other captive catarrhines, including chimpanzees and gorillas, have illustrated a general 40 H.B. Anderson et al. / Journal of Human Evolution 54 (2008) 34e42 orangutans should also be considered tentative, as all ages of older individuals are estimated with large potential error. Age estimation is particularly difficult for males due to sexual bimaturism and delay of development for as long as 20 years after maturity (Kingsley, 1982; Schurmann and van Hooff, ¨ 1986; Leigh and Shea, 1995; Maggioncalda et al., 1999; Utami et al., 2002; Knott and Kahlenberg, 2006). Indeed, the age structure of males in the Ketambe demography (Wich et al., 2004) is irregular, with more old individuals than might be suggested by the existing population of younger individuals; this suggests that some ages may be overestimated. Pending the development and strengthening of orangutan field sites across a diversity of habitats, we suggest that more data be collected on the range of ecological and life history variability within each orangutan species before final conclusions about between-species differences can be drawn. Finally, while these analyses do not support geneticallydetermined life history differences between orangutan species, they do aid in understanding the unique evolution of human life history. Under conditions of ample food availability and low energy expenditure, orangutan birth intervals (4.5e6 years) are considerably longer than those of even ecologically-stressed human populations (3e3.5 years, Howell, 1979; Hill and Hurtado, 1996). Therefore, it appears that the norms of reaction for reproductive rates are not overlapping in these species. Although these hominoids share characteristics of large body size, late age at maturity, and long lifespan, human evolution has led to a dramatic increase in the ability to devote energy to reproductive effort. As theorists have suggested (Kaplan et al., 2000; Wrangham et al., 2000; Knott, 2001; Aiello and Key, 2002), explanation of this marked shift in life history strategy lies at the heart of understanding the evolution of our species. Acknowledgements We thank the members of the zoological community who have cooperated to produce the International Studbook of the Orangutan. The SSP and studbook are currently maintained ´ at Zoo Atlanta. We thank Susan Anton, David Pilbeam, and three anonymous reviewers for helpful suggestions on the manuscript. References Aiello, L.C., Key, C., 2002. Energetic consequences of being a Homo erectus female. Am. J. Hum. Biol. 14, 551e565. Altmann, J., Alberts, S.C., 2003. Intraspecific variability in fertility and offspring survival in a non-human primate: behavioral control of ecological and social sources. In: Wachter, K. (Ed.), Offspring: Human Fertility Behavior in a Biodemographic Perspective. National Academy Press, Washington, DC, pp. 140e169. Asquith, P.J., 1989. Provisioning and the study of free-ranging primates: history, effects, and prospects. Yearb. Phys. Anthropol. 32, 129e158. Beck, B.B., Power, M.L., 1988. Correlates of sexual and maternal competence in captive gorillas. Zool. Biol. 7, 339e350. Benedetti, J.K., Liu, P.-Y., Sather, H.N., Seinfeld, J., Epton, M.A., 1982. Effective sample size for tests of censored survival data. Biometrika 69, 343e349. Fig. 3. Survivorship plots of captive orangutans that survived to maturity (16.0 yrs. for females, 18.6 yrs. for males), Kaplan-Meier analysis. (a) captive-born females, species comparison: c2 ¼ 0.91, df ¼ 1, p ¼ 0.34; (b) captive-born males: c2 ¼ 11.03, df ¼ 1, p ¼ 0.001). Censored data points reflect the ages of individuals still living at the time of data analysis. tendency for high early-infant mortality that can often be attributed to maternal incompetence (Davenport, 1979; Seal et al., 1985; Beck and Power, 1988; Courtenay and Santow, 1989; Latinen, 1989; Debyser, 1995). Can it really be true that despite provisioning and veterinary supervision, captive animals have decreased survivorship? We propose that in this case, it is the wild data that must be approached with caution. Due to the dispersed social organization of orangutans, researchers are likely to miss deaths that occur within the first months of life in individuals who are followed irregularly. The data on longevity of wild H.B. Anderson et al. / Journal of Human Evolution 54 (2008) 34e42 Borries, C., Koenig, A., Winkler, P., 2001. Variation of life history traits and mating patterns in female langur monkeys (Semnopithecus entellus). Behav. Ecol. Sociobiol. 50, 391e402. Breslow, N., 1970. A generalized Kruskal-Wallis test for comparing K samples subject to unequal patterns of censorship. Biometrika 57, 579e594. Charnov, E.L., 1991. Evolution of life history variation among female mammals. Proc. Natl. Acad. Sci. U. S. A. 88, 1134e1137. Charnov, E.L., Berrigan, D., 1993. Why do female primates have such long lifespans and so few babies? or life in the slow lane. Evol. Anthropol. 1, 191e194. Courtenay, J., Groves, C., Andrews, P., 1988. Inter- or intra-island variation? An assessment of the differences between Bornean and Sumatran orangutans. In: Schwartz, J.H. (Ed.), Orang-utan Biology. Oxford University Press, New York, pp. 19e30. Courtenay, J., Santow, G., 1989. Mortality of wild and captive chimpanzees. Folia Primatol. 52, 167e177. Davenport, R.K., 1979. Some behavioral disturbances in captivity. In: Hamburg, D.A., McCown, E.R. (Eds.), Perspectives on Human Evolution. The Great Apes, vol. V. Benjamin/Cummings Publishing Co., Menlo Park. de Lathouwers, M., van Elsacker, L., 2005. Reproductive parameters of female Pan paniscus and P. troglodytes: quality versus quantity. Int. J. Primatol. 26, 55e71. Debyser, I.W.J., 1995. Catarrhine juvenile mortality in captivity, under seminatural conditions, and in the wild. Int. J. Primatol. 16, 935e969. Delgado, R., van Schaik, C.P., 2000. The behavioural ecology and conservation of the orangutan (Pongo pygmaeus): a tale of two islands. Evol. Anthropol. 9, 201e218. Eisenberg, J.F., 1981. The Mammalian Radiations: An Analysis of Trends in Evolution, Adaptation, and Behavior. University of Chicago Press, Chicago. Ellison, P.T., 2003. Energetics and reproductive effort. Am. J. Hum. Biol. 15, 342e351. Ellison, P.T., Panter-Brick, C., Lipson, S.F., O’Rourke, M.T., 1993. The ecological context of human ovarian function. Hum. Reprod. 8, 2248e2258. Emery Thompson, M., 2005. Endocrinology and ecology of wild female chimpanzee reproduction. Ph.D. Dissertation, Harvard University. Emery Thompson M., Wrangham R.W., in review. Diet and reproduction in wild chimpanzees (Pan troglodytes schweinfurthii) in Kibale National Park, Uganda. Fox, E.A., van Schaik, C.P., Sitompul, A., Wright, D.N., 2004. Intra- and interpopulational differences in orangutan (Pongo pygmaeus) activity and diet: implications for the invention of tool use. Am. J. Phys. Anthropol. 124, 162e174. Gadgil, M., Bossert, W.H., 1970. Life historical consequences of natural selection. Am. Nat. 104, 1e24. Galdikas, B.M.F., 1985. Orangutan sociality at Tanjung Puting. Am. J. Primatol. 9, 101e119. Galdikas, B.M.F., Wood, J.W., 1990. Birth spacing patterns in humans and apes. Am. J. Phys. Anthropol. 83, 185e191. Gittleman, J.L., 1986. Carnivore life history patterns: allometric, phylogenetic, and ecological associations. Am. Nat. 127, 744e771. Groves, C.P., 2001. Primate Taxonomy. Smithsonian Institution Press, Washington, D.C. Groves, C.P., Westwood, C., Shea, B.T., 1992. Unfinished business: mahalanobis and a clockwork orang. J. Hum. Evol. 22, 327e340. Harvey, P.H., Clutton-Brock, T.H., 1985. Life history variation in primates. Evolution 39, 559e581. Harvey, P.H., Zammuto, R.M., 1985. Patterns of mortality and age at first reproduction in natural populations of mammals. Nature 315, 319e320. Hill, K., Hurtado, A.M., 1996. Ache Life History: The Ecology and Demography of a Foraging People. Aldine de Gruyter, New York. Howell, N., 1979. Demography of the Dobe! Kung. Aldine de Gruyter, New York. Kaplan, H., Hill, K.R., Lancaster, J.B., Hurtado, A.M., 2000. A theory of human life history evolution: diet, intelligence, and longevity. Evol. Anthropol. 9, 156e184. Ketterson, E.D., Nolan Jr., V., 1992. Hormones and life histories: an integrative approach. Am. Nat. 140, S33eS62. Kingsley, S., 1982. Causes of non-breeding and the development of secondary sexual characteristics in the male orang-utan: a hormonal study. In: de 41 Boer, L.E.M. (Ed.), The Orang-utan, Its Biology and Conservation. W. Junk, The Hague, pp. 215e229. Knott, C.D., 1998. Changes in orangutan caloric intake, energy balance and ketones in response to fluctuating fruit availability. Int. J. Primatol. 19, 1061e1079. Knott, C.D., 1999. Reproductive physiological and behavioral responses of Orangutans in Borneo to fluctuations in food availability. Ph.D. Dissertation, Harvard University. Knott, C.D., 2001. Ape models of female reproductive ecology. In: Ellison, P.T. (Ed.), Reproductive Ecology and Human Evolution. Aldine, Chicago, pp. 429e463. Knott, C.D., 2005. Energetic responses to food availability in the great apes: implications for hominid evolution. In: Brockman, D.K., van Schaik, C.P. (Eds.), Seasonality in Primates: Studies of Living and Extinct Human and Non-Human Primates. Cambridge University Press, Cambridge, pp. 351e378. Knott, C.D., Emery Thompson, M., Wich, S.A., in press. The ecology of reproduction in wild orangutans. In: Wich, S.A., Utami, S.S., Mitra Setia, T., van Schaik, C.P. (Eds.), Orangutans: Ecology, Evolution, Behaviour and Conservation. Oxford University Press, New York. Knott, C.D., Kahlenberg, S.M., 2006. Orangutans in perspective: forced copulations and female mating resistance. In: Campbell, C.J., Fuentes, A., Mackinnon, K.C., Panger, M.A., Bearder, S.K. (Eds.), Primates in Perspective. Oxford University Press, New York, pp. 290e304. Latinen, K., 1989. Demography of chimpanzees in captivity. In: Heltne, P.G., Marquardt, L.A. (Eds.), Understanding Chimpanzees. Harvard University Press, Cambridge, pp. 354e359. Leigh, S.R., Shea, B.T., 1995. Ontogeny and the evolution of adult body size dimorphism in apes. Am. J. Primatol. 36, 37e60. Maggioncalda, A.N., Sapolsky, R.M., Czekala, N.M., 1999. Reproductive hormone profiles in captive male orangutans: implications for understanding developmental arrest. Am. J. Phys. Anthropol. 109, 19e32. Markham, R., Groves, C.P., 1990. Weights of wild orang-utans. Am. J. Phys. Anthropol. 81, 1e3. Nowak, R.M., 1999. Walker’s Primates of the World. The Johns Hopkins University Press, Baltimore. Perkins, L., 2001. International Studbook of the Orangutan. Zoo Atlanta, Atlanta. Read, A.F., Harvey, P.H., 1989. Life history differences among the eutherian radiations. J. Zool. 219, 329e353. Ricklefs, R.E., Wikelski, M., 2002. The physiology/life-history nexus. Trends Ecol. Evol. 17, 462e468. Robbins, A.M., Robbins, M.M., Gerald-Steklis, N., Steklis, H.D., 2006. Agerelated patterns of reproductive success among female mountain gorillas. Am. J. Phys. Anthropol. 131, 511e521. Schneider, J.E., Wade, G.N., 2000. Inhibition of reproduction in service of energy balance. In: Wallen, K., Schneider, J.E. (Eds.), Reproduction in Context: Social and Environmental Influences on Reproduction. The MIT Press, Cambridge, pp. 35e82. Schurmann, C.L., van Hooff, J.A.R.A.M., 1986. Reproductive strategies of the ¨ orang-utan: new data and a reconsideration of existing sociosexual models. Int. J. Primatol. 7, 265e287. Seal, U.S., Flesness, N., Foose, T., 1985. Neonatal and infant mortality in captive-born great apes. In: Graham, C.E., Bowen, J.A. (Eds.), Clinical Management of Great Apes. Alan R. Liss, New York, pp. 193e203. Sievert, J., et al., 1991. Reproductive intervals in captive female western lowland gorillas with a comparison to wild mountain gorillas. Am. J. Primatol. 24, 227e234. Steiper, M.E., 2006. Population history, biogeography, and taxonomy of orangutans (Genus: Pongo) based on a population genetic meta-analysis of multiple loci. J. Hum. Evol. 50, 509e522. Taylor, A.B., van Schaik, C.P., 2007. Variation in brain size and ecology in Pongo. J. Hum. Evol. 52, 59e71. Tilson, R., Seal, U.S., Soemarna, K., Ramano, W., Sumardja, E., Poniran, S., van Schaik, C.P., Leighton, M., Rijksen, H., Eudey, A.A., 1993. Orangutan population and habitat viability analysis report. Orangutan Population and Habitat Viability Analysis Workshop, Medan, Indonesia. Tutin, C.E.G., 1994. Reproductive success story: variability among chimpanzees and comparisons with gorillas. In: Wrangham, R.W., McGrew, W.C., 42 H.B. Anderson et al. / Journal of Human Evolution 54 (2008) 34e42 Utami, S.S., Mitra Setia, T., van Schaik, C.P. (Eds.), Orangutans: Ecology, Evolution, Behaviour and Conservation. Oxford University Press, Oxford. Wich, S.A., Utami-Atmoko, S.S., Mitra Setia, T., Djoyoduharmo, S., Geurts, M.L., 2006. Dietary and energetical responses of Sumatran orangutans (Pongo abelii) to fruit availability fluctuations. Int. J. Primatol. 27, 1535e1550. Wich, S.A., Utami-Atmoko, S.S., Mitra Setia, T., Rijksen, H.D., Schurmann, C., van Hooff, J.A.R.A.M., van Schaik, C.P., 2004. Life history of wild Sumatran orangutans (Pongo abelii). J. Hum. Evol. 47, 385e398. Wrangham, R.W., Jones, J.H., Laden, G., Pilbeam, D., Conklin-Brittain, N., 2000. The raw and the stolen: cooking and the ecology of human origins. Curr. Anthropol. 40, 567e594. Xu, X., Arnason, U., 1996. The mitochondrial DNA molecule of Sumatran orangutan and a molecular proposal for two (Bornean and Sumatran) species of orangutan. J. Mol. Evol. 43, 431e437. Zera, A.J., Harshman, L.G., 2001. The physiology of life history trade-offs in animals. Ann. Rev. Ecol. Syst. 32, 95e126. de Waal, F.B.M., Heltne, P.G. (Eds.), Chimpanzee Cultures. Harvard University Press, Cambridge, pp. 181e194. Utami, S.S., Goossens, B., Bruford, M.W., de Ruiter, J.R., van Hooff, J.A.R.A.M., 2002. Male bimaturism and reproductive success in Sumatran orangutans. Behav. Ecol. 13, 643e652. van Noordwijk, M.A., de Jong, G., 1986. Acquisition and allocation of resources: their influence on variation in life history tactics. Am. Nat. 128, 137e142. van Noordwijk, M.A., van Schaik, C.P., 2005. The development of ecological competence in Sumatran orangutans. Am. J. Phys. Anthropol. 127, 79e94. Warren, K.S., Verschoor, E.J., Langenhuijzen, S., Heriyanto, R.A., Vigilant, L., Heeney, J.L., 2001. Speciation and intrasubspecific variation of Bornean orangutans, Pongo pygmaeus pygmaeus. Mol. Biol. Evol. 18, 472e480. Wasser, S.K., Barash, D.P., 1983. Reproductive suppression among female mammals: implications for biomedicine and sexual selection theory. Q. Rev. Biol. 58, 513e538. Wich, S.A., de Vries, H., Ancrenaz, M., Perkins, L., Shumaker, R., Suzuki, A., van Schaik, C.P., in press. Orangutan life history variation. In: Wich, S.A.,