Assessment of Reproductive Competence

Behavioral Assessment of Reproductive Competence in Captive Giant Pandas

Principal Investigators: David M. Powell, Lin Hong, Devra Kleiman, Smithsonian National Zoological Park

Collaborators: Zhang Hemin, Zhang Guiquan (China Conservation and Research Center for the Giant Panda), Zhang Zhihe, Yu Jianqu (Chengdu Research Base of Giant Panda Breeding), Timothy Ng (Ocean Park, Hong Kong), Zhang Jin Guo (Beijing Zoo), Rebecca Snyder (Zoo Atlanta)

Studies of poor reproductive performance in captive animals are often hampered by a small sample size and the fact that results are often not applicable across institutions. Multi-institutional studies of zoo animals have increased in frequency in the last 5-7 years as a means to acquire large enough sample sizes for statistical analysis and generation of powerful results. These studies have addressed the factors contributing to poor reproduction, mortality, and abnormal or stereotypic behaviors. For example, cheetah that are characterized as “tense” or “fearful” animals by their keepers and standardized behavioral tests are often poor breeders (Wielebnowski, 1999). Similar results have been found for maned wolves (Carlstead, 1998); however, fear scores and thus reproductive failure in female maned wolves can be reduced if multiple den boxes are provided in their enclosures. For black rhino, large enclosures with a low percentage of perimeter accessible to the public are correlated with improved reproductive success and lower mortality, respectively (Carlstead et al., 1999).

Historically, giant pandas have bred poorly in captivity, particularly outside of China. Behavioral problems have often been cited as a cause for the poor reproductive performance. Many males do not show normal reproductive or courtship behavior. Most males show either no interest in estrous females or are overly aggressive to them (Lindburg et al., 1997). A recent analysis of the newest giant panda studbook (Zhong and Gipps, 2001) indicates that only 30% of breeding age males in captivity have sired offspring either through artificial insemination or natural mating, and some of those males have only sired offspring via artificial insemination. Female pandas may periodically show weak or no estrus behavior at all, though they have ovulated, thus making breeding introductions difficult to time.

The goal of this study was to assess a large number of male and female pandas in Chinese facilities and determine what characteristics of exhibitry, husbandry, breeding management, and animal temperament correlate with normal courtship and reproductive behavior.

Research Method

We assessed breeding age pandas at the China Conservation and Research Center for the Giant Panda, the Chengdu Research Base of Giant Panda Breeding, Ocean Park Hong Kong, and the Beijing Zoo during the summer of 2001. We conducted standardized behavioral tests in which each animal was exposed to a novel object and scent to assess temperament or reactivity. We asked keepers to score the reproductive performance of the pandas they cared for using a 1 to 5 scale of the frequency that the panda exhibited each of a list of courtship and reproductive behaviors. We also conducted assessments of exhibit spaces, characterizing them in terms of area, substrates, furniture, proximity to other pandas/species, etc. We also asked keepers to fill out surveys relating to husbandry practices and breeding management protocols.

We were able to assess 11 males and 28 females during the study. To date we have only analyzed the data for the males. Because we still have a small sample size, only correlational analyses were conducted.

Findings

We immediately learned that some males were used for breeding while other males were not. The only behavioral difference we identified between these males was that breeding males bleated more frequently; however, this could be due to their being exposed to females. There were no differences in exhibits or husbandry between breeders and non-breeding males. We did find however that breeding males were on average 2 years older than non-breeding males, so they perhaps had more experience with females. There was no difference in socio-sexual performance between captive and wild-born males; however, wild-born males were significantly older than captive-born males.

Among breeding males, we found that activity levels, attempts at social interaction, and scent marking behaviors were all positively correlated with access to female olfactory cues prior to the breeding season. In addition, some scent marking behavior was positively correlated with access to olfactory cues from other males. We found that copulatory behavior was positively correlated with access to climbable trees in the enclosure, % of exhibit area with high (>2m) vegetation, and with how often the keeper touches the panda. Copulatory behavior was negatively correlated with the % of exhibit perimeter that was accessible to the public.

We can therefore suggest that pandas have extensive opportunities for olfactory contact within and between the sexes in the months prior to the breeding season. Males and females should have the opportunity to interact across a barrier (i.e through metal bars or mesh) prior to and during the breeding season. Keepers should establish calm and comfortable relationships with the animals. Exhibits should be designed to provide cover and refugia from the public and should include ample opportunities for climbing.