Zoo Biology 29 : 1–15 (2010)
RESEARCH ARTICLE
Life Span, Reproductive Output, and Reproductive Opportunity in Captive Goeldi’s Monkeys (Callimico goeldii) Kara Nuss1 and Mark Warneke2 1
Committee on Evolutionary Biology, University of Chicago, Chicago, Illinois Chicago Zoological Society, Chicago, Illinois
2
In the absence of long-term field studies, demographic and reproductive records from animals housed in zoos and research laboratories are a valuable tool for the study of life history variables relating to reproduction. In this study, we analyzed studbook records of more than 2,000 individuals born over a 40-year period (1965–2004) to describe life history patterns of captive Goeldi’s monkeys (Callimico goeldii) housed in North America and Europe. Using Kaplan–Meier survival analysis methods, we found the mean life span to be 5.5 years. The rate of infant mortality, defined as death before 30 days, was approximately 30%, with European animals being more likely to survive infancy than North American animals. When individuals surviving at least 1.5 years are considered, lifetime reproductive output averaged 3.5 offspring, yet more than one-third of individuals did not produce any offspring. Using a smaller dataset of individuals with known pairing histories, we developed a measure of opportunity for reproduction (OFR), which represented the total time an individual was known to be housed with a potential mate. For both sexes, we found that the correlation between OFR and number of offspring produced was much higher than the correlation between life span and number of offspring produced. This result highlights the importance of taking into account an individual’s OFR. As a whole, our findings help characterize the life histories of captive Goeldi’s monkeys and emphasize the impact management practices may have on reproductive success. Zoo Biol 29:1–15, 2010. r 2009 Wiley-Liss, Inc.
Keywords: reproductive success; captive breeding; life span Correspondence to: Kara Nuss, Committee on Evolutionary Biology, University of Chicago, Culver Hall– 402, 1025 E. 57th Street, Chicago, IL 60546. E-mail:
[email protected]
Grant sponsor: GAANN; Grant number: P200A060043. Received 21 November 2007; Revised 4 January 2009; Accepted 22 January 2009 DOI 10.1002/zoo.20239 Published online 23 February 2009 in Wiley InterScience (www.interscience.wiley.com).
r 2009 Wiley-Liss, Inc.
2
Nuss and Warneke
INTRODUCTION Analyses of breeding records from captive populations, both in zoos [e.g., Roberts, 1982; Wolf et al., 2000] and breeding centers [e.g., Harvey et al., 2004; Zhang et al., 2004] have become more common, owing in part to an increased effort to use scientific methods to manage captive populations for long-term demographic and genetic stability. A considerable body of literature based on analyses of studbooks and colony records has been amassed for primate species. Many of these studies have examined reproductive performance across species [Baker and Woods, 1992; Ha et al., 2000; Tardif, 1984]. Others have investigated factors affecting individual reproductive output, such as female mate choice [Keddy, 1986], variation in interbirth intervals [Koenig et al., 1990], seasonality of birth [de Vleeschouwer et al., 2003], and maternal age [Smucny et al., 2004]. In some studies, key factors that influence reproductive success were identified [e.g., Fairbanks and McGuire, 1984; Ha et al., 1999], allowing colony managers or zoo keepers to make changes in attempts to increase the future success of the breeding colony. Although the use of colony and zoo records offers many advantages, their use also presents unique challenges, especially in the measurement of individual reproductive success. In sexually reproducing species, individuals must have access to a member of the opposite sex to produce offspring, but in captivity individuals may be housed in nonbreeding situations. Such situations include remaining in their natal group for an extended period, being housed only with members of the same sex, or being housed with a contracepted or sterilized member of the opposite sex. The ability of zoos to decrease reproductive opportunity led Foose [1977] to distinguish between potential fertilities, when animals have unlimited breeding opportunity, and managed fertilities, which take into account the artificial influence of management. The difference between potential and managed fertilities likely varies among species and institutions, making it imperative that the impact of artificial management be considered in any research using studbook data to examine reproductive success. In this article, we used data from the Goeldi’s monkey (Callimico goeldii) international studbook to identify key life history parameters for this species in captivity. The Goeldi’s monkey is a New World primate widely dispersed throughout Bolivia, Peru, Colombia, Ecuador, and Brazil [Christen, 1998; Hershkovitz, 1977]. In captivity, animals are normally housed as breeding pairs or as family groups (a single breeding pair with young offspring). Field studies have reported multiple infants of approximately the same age in a single wild group, suggesting more than one breeding female [Christen, 1998; Masataka, 1981; Pook and Pook, 1981]. Initial attempts at maintaining groups with multiple breeding females in captivity had limited success owing to high levels of aggression between females or failure to rear offspring [Carroll, 1988]. More recently, a European research colony successfully created stable polygynous groups comprising two related females and an unrelated male [Mattle et al., 2004]. In zoos, however, animals are almost always housed as monogamous pairs along with their offspring. A strong inbreeding taboo has been documented and incestuous matings between sons and mothers are extremely rare, as evidenced by the very low conception rate of females housed without an unrelated breeding male [Sodaro and Hanson, 2004]. Thus, paternity can be assigned with near certainty in most situations, allowing for the investigation of male, as well as female, reproductive success. Zoo Biology
Reproduction in Captive Goeldi’s Monkeys
3
Multiple births within 1 year are common for females in captivity, and semiannual birth seasons have also been reported in the wild [Porter, 2001]. Although closely related marmosets, tamarins, and lion tamarins routinely produce twins, C. goeldii almost always give birth to single offspring [Altmann et al., 1988; Hershkovitz, 1977]. Offspring reach sexual maturity at approximately 1 year of age [Dettling and Pryce, 1999; Lorenz, 1972], but they are not normally placed in a breeding situation at that time. Under the current management systems, individuals are usually left in their natal group to gain experience with infants by assisting in the rearing of at least one younger sibling [Sodaro, 2004]. Waiting until animals are at least 2 years of age before breeding also lengthens the generation time, making it easier for population managers to maintain a greater degree of genetic diversity over a longer period of time. The primary goal of this research was to use studbook data to estimate various life history variables for captive Goeldi’s monkeys, specifically life span, rate of infant mortality, and reproductive output. By calculating these parameters and considering differences among regions or time periods, our aim was to identify certain areas where further research might be most beneficial for improving the success of captive breeding programs. A secondary goal was to investigate the effects of management on lifetime reproductive success, which led us to develop a new measure to quantify reproductive opportunity. METHODS Data Collection All analyses were based on data recorded and managed in the International Callimico Studbook, [Warneke, 2005]. The studbook was kept in electronic form using S.P.A.R.K.S (Single Population Analysis and Records Keeping System developed by International Species Information System, Apple Valley, MN). At the time of analysis, all information in the studbook was complete and updated through the end of 2004. To format data for separate analyses, all information from S.P.A.R.K.S. was transferred into Microsoft Access. For most animals, the following information was available from the studbook: birth or capture date, birth type (wild-caught or captive-born), sire and dam, inbreeding coefficient, sex, whether infant was reared by parents or hand-reared by keepers, death date, housing institutions. For some animals, especially in the earliest years, dates of birth, capture or death were estimated or specified only to a month or year. These animals were excluded from any analyses that required exact dates. Distribution of Births The studbook contained records of more than 2,000 captive births occurring at 131 locations in 26 countries. The majority of these institutions are under the guidance of a larger management group. Those in North America are managed collectively by a Species Survival Plan (SSP), formally established in 1992 by the Association of Zoos and Aquariums (AZA). Many animals at European institutions are managed by a European Endangered Species Program (EEP), which was established in 1990 by the European Association of Zoos and Aquariums (EAZA). We analyzed each region separately to account for possible life history variation Zoo Biology
4
Nuss and Warneke
resulting from differences in the management of these two populations. In North America, the population has always been intensively managed with an aim of equalizing the genetic representation of the population’s founders. Animals are selected for breeding pairs on the basis of mean kinship rankings, and pairs that reproduce successfully may eventually be separated to avoid overrepresentation of that genetic lineage in the population. In contrast, the European population has been managed on a more as-needed basis, in which the EEP does not prescribe all pairings, but makes recommendations when asked to do so. European family groups may be left together long after one or both of the breeding individuals have died. It is possible that these differences could influence the amount of breeding opportunity that individuals are given, and in turn might impact reproductive output. The studbook also contained a small number of individuals from countries outside of either SSP or EEP management, from Asia (Japan, Hong Kong, and Singapore), South America (Brazil and Peru), and Africa (South Africa). We have reported birth numbers for animals outside of SSP and EEP management, but did not conduct further analyses given the smaller numbers. We also excluded from further analyses animals housed at a research colony at the University of Zurich, Switzerland, given the differences in management goals between zoos and research colonies. Calculating Life Span and Rate of Infant Mortality To study life span, we included only live births (excluding 66 stillbirths and 37 miscarriages) and limited the dataset to animals under North American and European management. We also excluded individuals with unknown or estimated death dates (41 animals) and those that were transferred to nonaccredited institutions that did not respond to recent studbook update questionnaires and did not subscribe to the International Species Inventory System (ISIS) (51 animals). The remaining dataset contained 644 North American animals and 1,173 European animals. Because data failed to meet assumptions for parametric statistics, nonparametric tests were used. Mann–Whitney U tests were employed to test for differences between regions and between the sexes in life span and reproductive output. For all statistical tests a was set at 0.05. A previous study of elephant studbook data demonstrated that survival analysis methods, which allow for the inclusion of information from animals that are still alive, were often more accurate than simply calculating a mean of known life span values, i.e., including only animals that had already died [Wiese and Willis, 2004]. Here, we employed a survival analysis method, using the Kaplan–Meier estimator of the survivorship function, which allows for still-living individuals to be included as censored data [Hosmer and Lemeshow, 1999]. We reported both the mean life span and median life span (age at which 50% of individuals are still alive) calculated from this analysis. We defined infant mortality as death before 30 days and calculated the rate of infant mortality using the same dataset as above, with all live births included and no distinction made between hand-reared and parent-reared offspring. We then limited the dataset to all individuals that survive beyond 30 days of age and recalculated the mean and median life span, again using survival analysis methods to include still living animals. Zoo Biology
Reproduction in Captive Goeldi’s Monkeys
5
Quantifying Reproductive Output Given that both sexes reach sexual maturity at approximately 1 year [Dettling and Pryce, 1999; Sodaro, 2004] and the gestation period ranges from 144 to 159 days [Carroll et al., 1989; Jurke et al., 1994; Ziegler et al., 1989], very few offspring were born to parents that, at the time of birth, were under the age of 1.5 years. Therefore, when considering reproductive output, we included only animals that survived at least 1.5 years. We also chose to only include animals that were deceased and had lived their entire life span in captivity, so that their entire reproductive history was known. Complete reproductive histories were known for 364 males and 316 females, and we analyzed North American and European animals separately. We were specifically interested in offspring production, and therefore included all recorded offspring regardless of viability. Measuring Opportunity for Reproduction In most cases, the studbook did not contain sufficient information about an individual’s reproductive opportunity; specifically, the amount of time an individual was housed with a potential breeding partner (i.e., a member of the opposite sex that was not contracepted or sterilized) was not recorded. However, between 1977 and 1989 detailed records kept at Brookfield Zoo noted the animals that were housed in breeding pairs and the length of time each pair existed. After excluding any female that was contracepted or any male that had been housed with a contracepted female, there were 51 males and 53 females that were housed in at least one breeding situation throughout that time. We devised a measure to quantify the breeding opportunity for these individuals; opportunity for reproduction (OFR) was calculated by adding the total number of days an individual was housed with a potential mate for each breeding situation in which they were kept. We then divided this result by 365 so that the final OFR value represented the total amount of time in years that an individual was housed in a breeding situation while at Brookfield Zoo. To further investigate reproductive opportunity, we examined the relationship between life span and total reproductive output for these same animals (excluding wild-caught and animals and other individuals with unknown life span). We used Spearman correlation coefficients to describe the strength of the relationships among reproductive output, OFR, and life span.
RESULTS Distribution of Births Between January 1, 1965 and December 31, 2004, there were 2,229 births (including stillbirths and miscarriages) recorded, however, 32 of these had estimated or unknown birth dates and were excluded from further analyses. Figure 1 shows the distribution, by region, of the remaining 2,197 recorded births. The majority of births occurred in institutions under US or European management. Additional births recorded in the studbook included those at Asian, South American, and African zoos and at the University of Zurich’s research colony (births from these locations are labeled as ‘‘other’’ in Figure 1). Zoo Biology
6
Nuss and Warneke
Fig. 1. Distribution of captive C. goeldii births from 1965 to 2004, shown in 5-year intervals. Births occurring in African, Asian, and South American institutions are grouped together in the category ‘‘other.’’
Fig. 2. Survival curves for animals born in North America (black line) and Europe (gray line) generated by Kaplan–Meier survival analysis. Individuals with known life span and those that were still living were included in the analysis (North America: n 5 644, Europe: n 5 1,173). Note the sharp decline in survival that occurs immediately after birth in both regions.
Of the 2,197 recorded births, 1,000 were male and 941 were female. In 256 cases, sex was not recorded (in many instances, because the animal died in infancy and sex was not determined postmortem). There was no significant difference between the number of male and female births (binomial test, P40.05). Life Span and Infant Mortality Figure 2 shows survival curves of animals under North American and European management. As calculated by survival analysis, the mean life span was 5.1970.23 years for North American animals (n 5 644) and 5.6970.20 years for European animals (n 5 1,173). The difference in life span between the two regions was not statistically significant (Log–Rank test, P40.05). The median life span was Zoo Biology
Reproduction in Captive Goeldi’s Monkeys
7
3.41 years for North American animals and 3.84 years for European animals. When only individuals surviving more than 30 days were considered, the mean life span for the combined North American and European populations was 7.7670.18 years (n 5 1,284) and the median life span was 7.01 years. The overall rate of infant mortality, defined as death before 30 days of age, was 29.3%. The rate of infant mortality in North America was 32.8%, compared with 27.5%, in the European population. North American infants were significantly more likely to die before 30 days of age (w2 test, w2 5 5.66, df 5 1, P 5 0.017). However, there was not a significant difference in infant mortality between males and females in either region (w2 test, P40.05). In the US population, 9.5% of individuals were hand-reared, whereas 6.8% of European animals were hand-reared. The rate of infant mortality in hand-reared animals was 43.6%, and hand-reared infants were significantly more likely to die before 30 days than parent-reared infants (w2 test, w2 5 14.31, df 5 1, P 5 0.00016). Reproductive Output We examined the reproductive output for all individuals that survived at least 1.5 years and had lived their entire life span in captivity, excluding the 323 still-living individuals in the initial dataset (the mean reproductive output for the live group was 1.6270.18 offspring). Table 1 shows the number of males and females with complete reproductive histories from both North American and European management, along with the mean number of offspring produced, as well as the percentage that did not reproduce. Although the mean number of offspring for each group was 2.9 or greater, this average was lowered by the considerable portion of individuals that did not reproduce (more than 45% of males and 30% of females). Among individuals that reproduced at least once, mean number of offspring ranged from 5.32 for North American males to 6.24 for European females. There was not a significant difference, for either sex, in the number of offspring produced between the two regions (males: P 5 0.57; females: P 5 0.68). Opportunity for Reproduction The time span for which pairing information was available varied among subjects, but typically, individuals were housed in one or two breeding situations while at Brookfield Zoo. The number of different breeding partners an individual encountered ranged from 1–7 for males and 1–5 for females. During the study TABLE 1. Offspring production by each sex within each region North America Sex Males Females
Europe
n
Mean7SE
No offspring (%)
n
Mean
No offspring (%)
149 134
2.8970.33 3.8670.36
45.64 30.60
215 182
3.4070.32 3.9170.35
45.58 33.52
We included only individuals that had survived at least 1.5 years and had lived their entire life span in captivity so that lifetime reproductive output was known. For each sex and region, the number of subjects (n) is listed along with the mean number (and standard error) of offspring produced. The column ‘‘No offspring’’ represents the percentage of the group that never reproduced over their entire lifespan.
Zoo Biology
8
Nuss and Warneke
Fig. 3. Reproductive opportunity, life span, and reproductive output of animals housed at Brookfield Zoo (1977–1989). Males (n 5 51) represented by ~; females (n 5 53) represented by B. (A) Correlation between the amount of time housed with a mate (OFR) and reproductive output during that time (males, rs 5 0.90, Po0.0001; females, rs 5 0.93, Po0.0001). (B) Correlation between total life span and total reproductive output for the same individuals (males, rs 5 0.54, P 5 0.0002; females, rs 5 0.60, Po0.0001).
period, 28.8% of males and 22.6% of females did not produce any offspring. The mean OFR for these nonreproducing animals was 0.60 years for males and 0.50 years for females. As shown in Figure 3A, number of offspring was highly correlated with OFR (males: rs 5 0.90, Po0.0001; females: rs 5 0.92, Po0.0001). For comparison, Figure 3B shows the relationship between total number of offspring produced and life span for the same individuals (excluding those animals with unknown life span). Although life span and reproductive output were significantly correlated (males: rs 5 0.54, P 5 0.0002; females: rs 5 0.60, Po0.0001), the correlation was not as high as that observed between OFR and reproductive output.
DISCUSSION Distribution of Births This research analyzed 40 years of studbook data to examine demographic and management trends that might have affected reproductive success. As shown in Figure 1, birth numbers in both the US and Europe were low during the first 15 years of breeding (1964–1979). This period corresponded to the initial phases of the captive breeding programs when there were relatively few breeding individuals Zoo Biology
Reproduction in Captive Goeldi’s Monkeys
9
and low rates of breeding success. In North America, this species became a captive-breeding priority (with an original target population size of 200), in the 1980s. That was the most prolific decade for the North American population, with 426 births recorded in the 1980s. The distribution of European births initially followed a similar pattern with birth numbers increasing in the 1980s. However, births continued to increase in the early 1990s before stabilizing for the following 10 years as available housing spaces were filled. The drastic decrease in North American births during 1990–1994 was a reflection of management practices enacted during that time. As the rapid growth of regional populations during the 1980s filled available space, population managers focused on maximizing genetic diversity. In the North American population, they chose to contracept some females with melengestrol acetate implants from 1986 to 1992, but this method was never used in Europe. Unexpectedly, this method, routinely used in other primates, resulted in permanent sterility in many of the treated C. goeldii females [Munson et al., 2005; Murnane et al., 1996]. The unintended effects of this management decision were reflected in the North American population’s demography. Between 1992 and 2000, the annual growth rate (l) was negative, but improved in 2000–2002 when population managers imported animals with low levels of relatedness to the North American population from European institutions, thereby increasing the number of breeding pairs in North America [Long et al., 2006]. The increase in North American births in the final 5 years of available data (2000–2004) represented the stabilization of the captive population near the target goal of 125 individuals. Life Span and Infant Mortality We used survival analysis to most accurately estimate life span of animals in each region (Figure 2). Although animals in Europe lived slightly longer on average than their North American counterparts (5.7 years compared to 5.2 years), the difference between the two regions was not significant. Data on life span are less likely to be reported than other life-history traits [Dyke et al., 1993], but there have been some reports from captive colonies of closely related species. In a research colony of Wied’s black-tufted-ear marmosets (Callithrix kuhlii), the average life span for all individuals that survived at least to weaning (at approximately 9 weeks), was 5.5 years [Ross et al., 2007]. A multicolony study of common marmosets (C. jacchus) reported that females had an average life span of approximately 6 years [Tardif et al., 2003]. Caro et al. [1995] compared female life span and fertility among 13 primate species, including 3 callitrichid species: common marmosets, golden lion tamarins (Leontopithecus rosalia), and saddle-back tamarins (Saguinus fuscicollis). Again the mean life spans in these species were similar to C. goeldii, leading us to conclude that the life span of captive C. goeldii is similar to that seen in captive callitrichids. However, Dyke et al. [1993] used mortality statistics from four callitrichid colonies (C. jacchus, L. rosalia, S. fuscicollis, and S. oedipus) to generate standard life tables for smaller New World primates. Compared to these results, the median life span reported here for captive C. goeldii is higher than many other small New World primates, where the composite 50% survival is less than 1 year of age. It is possible that these species had higher rates of infant mortality that led to a lower expected life span than we calculated for C. goeldii. Zoo Biology
10
Nuss and Warneke
Overall, approximately 30% of captive C. goeldii births did not survive beyond 30 days of age. A greater proportion of North American animals died during infancy compared with European animals, and hand-reared individuals also had a higher proportion of infant mortality compared with parent-reared animals. A previous study by Carroll [1982] analyzed 40 births at Jersey Zoo between 1975 and 1981 and found that 15% of the infants (6 of 40) died before reaching 30 days. The higher rate of infant mortality reported here might be owing to the inclusion of data from the earliest years of captive breeding when husbandry guidelines were still being developed. Furthermore, this dataset included more than 100 institutions, and it is implausible that all had extensive knowledge of callitrichids or previous experience with C. goeldii. It is likely that this factor also contributed to the higher rate of infant mortality in this study. The numerous callitrichid species housed in laboratory settings provide data on infant mortality, allowing for a comparison of C. goeldii with related species. Early studies reported high levels of stillbirth and infant mortality. In a cotton-top tamarin colony, 32% of births were stillborn and an additional 20% died before reaching 1 week [Kilborn et al., 1983]. A common marmoset colony had an 11% stillbirth rate and a 32% infant mortality rate for live-born young [Poole and Evans, 1982]. Tardif et al. [1984b], however, reported a 75% 3-month survival rate for common marmosets and a 60% rate for cotton-top tamarins. The 3-month survival rate for this study of C. goeldii is 64%, falling in between the marmoset and tamarin values. Others have reported similar 3-month survival values: 80% for common marmosets [Rothe et al., 1993] and 69% for cotton-top tamarins [Price and McGrew, 1990]. In a review of platyrrhine juvenile mortality, Debyser [1995] calculated a 31% rate of mortality before 3 months in the callitrichids (all data from captive populations with one exception). In captive environments at least, C. goeldii have an infant mortality rate comparable to callitrichids. It is important to note, however, that most callitrichids routinely produce twins or triplet litters in captivity, whereas C. goeldii almost always have single-offspring births and are probably incapable of rearing both offspring from a twin birth [Altmann et al., 1988]. Therefore, although the rates of infant mortality are similar, C. goeldii females may be giving birth to fewer infants during their reproductive lifetimes, and could therefore be producing fewer offspring that survive to sexual maturity. An analysis of the studbooks of many captive callitrichid populations would help to determine whether significant biological differences contribute to differences in reproductive success between C. goeldii and other callitricihids. From a captive breeding standpoint, decreasing infant mortality could lead to substantial increases in population size and possible conservation of genetic diversity. The first step in decreasing infant mortality would be identifying key factors that increase the risk of infant death before 30 days. A number of studies have used captive breeding records to investigate potential factors contributing to infant survival or mortality. Inbreeding depression, in the form of increased infant mortality, was found to occur in many species of captive primates [Ralls and Ballou, 1982]. Inbreeding depression has been documented in captive C. goeldii, with Lacy et al. [1993] reporting that each 10% increase in inbreeding resulted in a 33% decrease in survival. Previous experience with infants, obtained either by rearing younger siblings or earlier offspring, has often been correlated with increased infant survival. Females with previous experience rearing younger siblings had higher rates Zoo Biology
Reproduction in Captive Goeldi’s Monkeys
11
of offspring survival in common marmosets and cotton-top tamarins [Tardif et al., 1984a], as well as lion tamarins [French et al., 1996]. Increased parity was found to be associated with higher rates of offspring survival in cotton-top tamarins [Price and McGrew, 1990] and in lion tamarins, where multiparous females had significantly higher rates of offspring survival than primiparous females [French et al., 1996]. Other studies have investigated additional factors, but found that they did not significantly affect offspring survival in the given study population. These factors include maternal age in common marmosets, cotton-top tamarins, and saddle-back tamarins [Jaquish et al., 1991], proximity to conspecifics in Geoffroy’s tamarins [Kuhar et al., 2003] and group size in common marmosets [Rothe et al., 1993]. The detailed records contained in the C. goeldii studbook provide an opportunity to thoroughly investigate many factors (e.g., maternal and paternal age, parity, level of inbreeding, environmental conditions) that may potentially affect offspring survival. Research is currently underway to determine the factors and interactions that are most important to offspring survival in captive C. goeldii. Reproductive Output When the dataset was limited only to individuals that survived at least 1.5 years, mean reproductive output ranged from 2.9 offspring for North American males to 3.9 offspring for European females (Table 1). Note that these values reflect total offspring produced and do not take offspring survival into account. Given that approximately 30% of offspring do not survive beyond 30 days, actual reproductive success would be considerably lower in all groups. Surprisingly, more than 30% of females and 45% of males that survived at least 1.5 years produced no offspring during their life span. This variation in reproductive output was in line with the results reported by Pryce and Dettling [1995] who documented similar variation in breeding success among a sub-sample of 10 females that were founding members of the captive population. These results could be better interpreted if more information about individuals’ breeding opportunities were available. For instance, among those individuals that never reproduced, it is unclear as to how much time (if any) they were housed in a potential breeding situation. The contraception of some North American females (and the unintended sterility that resulted) have undoubtedly led to decreased reproductive output among some individuals, although this issue impacted only a relatively small number of individuals for a brief time. As discussed in the following section, the longer an individual is housed with a potential mate, the more offspring they produce. Without detailed records of breeding opportunity, further analyses of reproductive output, such as age-specific fertility, would be difficult to interpret because of the inability to separate the effects of management decisions from actual, biological signals. Opportunity for Reproduction The OFR dataset included individuals housed in breeding pairs held at Brookfield Zoo between 1977 and 1989. For each individual, we calculated an OFR value, representing the amount of time an individual was housed with a potential mate. The correlation between OFR and reproductive output was much higher than the correlation between life span and total reproductive output (Figure 3). For both Zoo Biology
12
Nuss and Warneke
males and females, approximately 90% of the variation in reproductive output was explained by their OFR. This result implies that, at least for the colony at Brookfield Zoo, management decisions had a much greater effect on individual reproductive success than did inherent biological characteristics. The Brookfield colony has always been managed to equalize founder representation in the population, ensuring that rare gene lines were not lost. This management style often meant that C. goeldii pairs that were breeding and rearing their offspring well were the ones whose reproduction was stopped because their success would lead to overrepresentation of their genetic lines in the overall population. Conversely, pairs that had poor reproductive and/or rearing success were given more opportunity to breed in hopes of having their genetic lines better represented. Pairs with known or anticipated problems in rearing offspring were often kept at Brookfield so as not to intentionally transfer problems to other institutions. Such management strategies resulted in Brookfield Zoo actually having lower overall reproductive success compared with other institutions, providing strong evidence that the observed correlation between reproductive output and OFR was not simply owing to selection of the most prolific breeders and exclusion of those individuals that did not readily breed. The high correlation between OFR and reproductive output emphasizes the importance of accounting for an individual’s reproductive opportunity, however this information has rarely been recorded in primate studbooks. Some demographic studies note that reproductive opportunity was not taken into account [e.g., Gage, 1998]. Other studies may exclude some individuals when additional information suggested an interruption in breeding [e.g., Caro et al., 1995] or include only individuals producing at least one offspring, so that all animals had at least some opportunity for reproduction [e.g., Ricklefs et al., 2003]. In some populations, age can be used as a rough estimate of opportunity for reproduction, as it is generally assumed that the longer an individual lives, the more opportunity he or she will have to reproduce [Rhine et al., 2000]. However, when management decisions delay or interrupt breeding, age may be a poor proxy for reproductive opportunity. Rarely, detailed information on reproductive opportunity is available and incorporated into analysis. In a study of reproductive success of a ruffed lemur colony (Varecia variegata), contraception and reduced access to adult males limited reproductive opportunity for a number of females in the study [Weigler et al., 1994]. The authors noted that failing to limit the analysis to females that had actual breeding opportunity would have resulted in an underestimate of reproductive success for the population. In this study, we were able to calculate OFR for a subset of reproductively capable individuals in the captive population of C. goeldii. Based on our finding that OFR was highly correlated with reproductive output, we recommend that future versions of zoo management software include data fields for information regarding breeding opportunity. Individuals that never had the opportunity to breed could then be excluded from any analysis of reproductive success. Furthermore, individuals that were given ample breeding opportunity but failed to reproduce could be identified, so that possible causes of reproductive failure could be explored. This analysis clearly demonstrates that when examining reproductive output in a captive population it is critical to consider an individual’s OFR. Accounting for the many factors that influence a decision to pair animals, and therefore the reproductive Zoo Biology
Reproduction in Captive Goeldi’s Monkeys
13
opportunity of those animals, becomes a critical component of objectively and accurately assessing reproductive success in captive populations. CONCLUSIONS 1. More than 2,000 captive C. goeldii births occurring between 1965 and 2004 were recorded in the international studbook, with the majority of these births taking place at North American and European institutions. 2. Based on survival analysis, mean life span was 5.2 years for North American animals and 5.7 years for European animals. Infant mortality was 32.8% in the North American population and 27.5% in the European population. Both life span and rate of infant mortality were similar to results obtained from studies of captive marmoset and tamarin colonies. 3. Among individuals that survived at least 1.5 years and lived their entire life span in captivity, the mean number of offspring was 3.5, with European females having the highest output. However, more than 30% of females and 45% of males never reproduced. 4. A measure of reproductive opportunity, OFR was significantly correlated with number of offspring produced. This correlation was higher than that between life span and reproductive output. Controlling for reproductive opportunity may be an essential step in any study using captive animals to evaluate reproductive success. ACKNOWLEDGMENTS We thank Allison Walsh and Eric Westhus for assistance formatting the database and Vince Sodaro for access to housing records. We also thank Robert Lacy, Sue Margulis, Jill Mateo, and Melinda Pruett-Jones for their suggestions on earlier versions of this article. Finally, we thank four anonymous reviewers for their comments. While conducting this research, Kara Nuss was supported by GAANN training grant P200A060043, awarded to the Committee on Evolutionary Biology, University of Chicago. REFERENCES Altmann J, Warneke M, Ramer J. 1988. Twinning among Callimico goeldii. Int J Primatol 9:165–168. Baker AJ, Woods F. 1992. Reproduction of the emperor tamarin (Saguinus imperator) in captivity, with comparisons to cotton-top and golden lion tamarins. Am J Primatol 26:1–10. Caro TM, Sellen DW, Parish A, Frank R, Brown DM, Voland E, Borgerhoff Mulder M. 1995. Termination of reproduction in nonhuman and human female primates. Int J Primatol 16:205–220. Carroll JB. 1982. Maintenance of the Goeldi’s monkey Callimico goeldii at Jersey Wildlife Preservation Trust. Int Zoo Yearbook 22: 101–105.
Carroll JB. 1988. The stability of multifemale groups of Goeldi’s monkey Callimico goeldii in captivity. Dodo 25:37–43. Carroll JB, Abbott DH, George LM, Martin RD. 1989. Aspects of urinary oestrogen excretion during the ovarian cycle and pregnancy in Goeldi’s monkey, Callimico goeldii. Folia Primatol 52:201–205. Christen A. 1998. The most enigmatic monkey in the Bolivian rain forest—Callimico goeldii. Neotrop Primates 6:35–37. de Vleeschouwer K, Leus K, Van Elsacker L. 2003. Characteristics of reproductive biology and proximate factors regulating seasonal breeding in captive golden-headed lion tamarins (Leontopithecus chrysomelas). Am J Primatol 60:123–127.
Zoo Biology
14
Nuss and Warneke
Debyser IWJ. 1995. Platyrrhine juvenile mortality in captivity and in the wild. Int J Primatol 16:909–933. Dettling A, Pryce CR. 1999. Hormonal monitoring of age at sexual maturation in female Goeldi’s monkeys (Callimico goeldii) in their family groups. Am J Primatol 48:77–83. Dyke B, Gage TB, Ballou JD, Petto AJ, Tardif SD, Williams LE. 1993. Model life tables for the smaller new world monkeys. Am J Primatol 29:269–285. Fairbanks LA, McGuire MT. 1984. Determinants of fecundity and reproductive success in captive vervet monkeys. Am J Primatol 7:27–38. Foose TJ. 1977. Demographic models for management of captive populations. Int Zoo Yearbook 17:70–76. French JA, Pissinatti A, Coimbra-Filho AF. 1996. Reproduction in captive Lion Tamarins (Leontopithecus): seasonality, infant survival, and sex ratios. Am J Primatol 39:17–33. Gage TB. 1998. The comparative demography of primates: with some comments on the evolution of life histories. Annu Rev Anthropol 27:197–221. Ha JC, Robinette RL, Sackett GP. 1999. Social housing and pregnancy outcome in captive pigtailed macaques. Am J Primatol 47:153–163. Ha JC, Robinette RL, Sackett G. 2000. Demographic analysis of the Washington Regional Primate Research Center pigtailed macaque colony, 1967–1996. Am J Primatol 52:187–198. Harvey NC, Dankovchik JD, Kuehler CM, Levites T, Kasielke S, Kiff L, Wallace MP, Mace ME. 2004. Egg size, fertility, hatchability, and chick survivability in captive California condors (Gymnogyps californianus). Zoo Biol 23:489–500. Hershkovitz P. 1977. Living new world monkeys (Platyrrhini) with an introduction to primates. Chicago: The University of Chicago Press. p 1117. Hosmer DW, Lemeshow S. 1999. Applied survival analysis. New York: Wiley. Jaquish CE, Gage TB, Tardif SD. 1991. Reproductive factors affecting survivorship in captive Callitrichidae. Am J Phys Anthropol 84:291–305. Jurke MH, Pryce CR, Dobeli M, Martin RD. 1994. Non-invasive detection and monitoring of pregnancy and the postpartum period in Goeldi’s monkey (Callimico goeldii) using urinary pregnanediol-3-alpha-glucuronide. Am J Primatol 34:319–331. Keddy AC. 1986. Female mate choice in vervet monkeys (Cercopithecus aethiops sabaeus). Am J Primatol 10:125–134. Kilborn JA, Sehgal P, Johnson LD, Beland M, Bronson RT. 1983. A retrospective study of infant mortality of cotton-top tamarins (Saguinus oedipus) in captive breeding. Lab Anim Sci 33:168–171. Koenig A, Radespiel U, Siess M, Rothe H, Darms K. 1990. Analysis of pairing–parturition intervals and interbirth intervals in a colony
Zoo Biology
of common marmosets (Callithrix jacchus). Zeitschrift fuer Saeugetierkunde 55:308–314. Kuhar CW, Bettinger TL, Sironen AL, Shaw JH, Lasley BL. 2003. Factors affecting reproduction in zoo-housed Geoffroy’s tamarins (Saguinus geoffroyi). Zoo Biol 22:545–559. Lacy RC, Petric A, Warneke M. 1993. Inbreeding and outbreeding in captive populations of wild animal species. In: Thornhill NW, editor. The natural history of inbreeding and outbreeding. Chicago: The University of Chicago Press. p 352–374. Long S, Lynch C, Pruett-Jones M, Warneke M. 2006. Callimico species survival plan. Chicago: Lincoln Park Zoo. p 24. Lorenz R. 1972. Management and reproduction of the goeldi’s monkey Callimico goeldii (Thomas, 1904) Callimiconidae, Primates. Saving the lion marmoset proceedings of the wild animal propagation trust golden lion marmoset conference. Wheeling: Wild Animal Propagation Trust. p 92–109. Masataka N. 1981. A field study of the social behavior of Goeldi’s monkeys (Callimico goeldii) in North Bolivia. I. Group composition, breeding cycle, and infant development. Kyoto University Overseas research reports of new world monkeys. Vol. 2. p 23–32. Mattle F, Benz B, Anzenberger G. 2004. Experimentally induced polygyny in two callitrichid species: comparison between Callithrix and Callimico. Folia Primatol 75:394. Munson L, Moresco A, Galle PP. 2005. Adverse effects of contraceptives. In: Asa CS, Porton IJ, editors. Wildlife contraception: issues, methods, and applications. Baltimore: Johns Hopkins Press. p 66–82. Murnane RD, Zdziarski JM, Walsh TF, Kinsel MJ, Meehan TP, Kovarik P, Briggs M, Raverty S, Phillips LG. 1996. Melengestrol acetate-induced exuberant endometrial decidualization in Goeldi’s marmosets (Callimico goeldii) and squirrel monkeys (Saimiri sciureus). J Zoo Wildl Med 27:315–324. Pook AG, Pook G. 1981. A field study of the socio-ecology of the Goeldi’s monkey (Callimico goeldii) in Northern Bolivia. Folia Primatol 35:288–312. Poole TB, Evans RG. 1982. Reproduction, infant survival and productivity of a colony of common marmosets (Callithrix jacchus jacchus). Lab Anim 16:88–97. Porter LM. 2001. Social organization, reproduction and rearing strategies of Callimico goeldii: new clues from the wild. Folia Primatol 72:69–79. Price EC, McGrew WC. 1990. Cotton-top tamarins (Saguinus (o.) oedipus) in a semi-naturalistic captive colony. Am J Primatol 20:1–12. Pryce CR, Dettling A. 1995. Fundamental and applied research into Goeldi’s monkey breeding: an integration of behavioural, endocrine and life-history techniques. Research and captive
Reproduction in Captive Goeldi’s Monkeys propagation. Fuerth, Germany: Filander Verlag. p 285–291. Ralls K, Ballou J. 1982. Effects of inbreeding on infant mortality in captive primates. Int J Primatol 3:491–505. Rhine RJ, Norton GW, Wasser SK. 2000. Lifetime reproductive success, longevity, and reproductive life history of female yellow baboons (Papio cynocephalus) of Mikumi National Park, Tanzania. Am J Primatol 51:229–241. Ricklefs RE, Scheuerlein A, Cohen A. 2003. Agerelated patterns of fertility in captive populations of birds and mammals. Exp Gerontol 38:741–745. Roberts M. 1982. Demographic trends in a captive population of red pandas (Ailurus fulgens). Zoo Biol 1:119–126. Ross CN, Fite JE, Jensen H, French JA. 2007. Demographic review of a captive colony of callitrichids (Callithrix kuhlii). Am J Primatol 69:234–240. Rothe H, Koenig A, Darms K. 1993. Infant survival and number of helpers in captive groups of common marmosets (Callithrix jacchus). Am J Primatol 30:131–137. Smucny DA, Abbott DH, Mansfield KG, Schultz-Darken NJ, Yamamoto ME, Alencar AI, Tardif SD. 2004. Reproductive output, maternal age, and survivorship in captive common marmoset females (Callithrix jacchus). Am J Primatol 64:107–121. Sodaro V. 2004. Reproduction. In: Sodaro V, editor. Callimico species survival plan husbandry manual. Chicago: Chicago Zoological Society. p 37–39. Sodaro V, Hanson AM. 2004. Social organization and housing of captive Goeldi’s monkeys. In: Sodaro V, editor. Callimico species survival plan
15
husbandry manual. Chicago: Chicago Zoological Society. p 24–36. Tardif SD. 1984. Social influences on sexual maturation of female Saguinus oedipus oedipus. Am J Primatol 6:199–209. Tardif SD, Richter CB, Carson RL. 1984a. Effects of sibling-rearing experience on future reproductive success in two species of Callitrichidae. Am J Primatol 6:377–380. Tardif SD, Richter CB, Carson RL. 1984b. Reproductive performance of three species of Callitrichidae. Lab Anim Sci 34:272–275. Tardif SD, Smucny DA, Abbott DH, Mansfield K, Schultz-Darken N, Yamamoto ME. 2003. Reproduction in captive common marmosets (Callithrix jacchus). Comp Med 53:364–368. Warneke M. 2005. Callimico goeldii: international studbook. Chicago: Chicago Zoological Society. Weigler BJ, Feeser PL, Grant SM, Norris CA. 1994. Epidemiologic assessment of reproductive success and infant mortality in ruffed lemurs (Varecia variegata). Zoo Biol 13:527–536. Wiese RJ, Willis K. 2004. Calculation of longevity and life expectancy in captive elephants. Zoo Biol 23:365–373. Wolf KN, Wildt DE, Vargas A, Marinari PE, Ottinger MA, Howard JG. 2000. Reproductive inefficiency in male black-footed ferrets (Mustela nigripes). Zoo Biol 19:517–528. Zhang G, Swaisgood RR, Zhang H. 2004. Evaluation of behavioral factors influencing reproductive success and failure in captive giant pandas. Zoo Biol 23:15–31. Ziegler TE, Snowdon CT, Warneke M. 1989. Postpartum ovulation and conception in Goeldi’s monkey, Callimico goeldii. Folia Primatol 52:206–210.
Zoo Biology
