Effects of Predictable Scheduling of Meals on Giant Panda Behavior

Although wild giant pandas spend a considerable amount of time foraging, processing, and consuming a diet comprised almost exclusively of bamboo, captive giant pandas are typically fed highly concentrated foods as well as limited amounts of bamboo and other browse in scheduled meals throughout the day. Giant pandas have been observed to exhibit increased vigilance toward keepers, aggression toward conspecifics, pacing, and other stereotypic behaviors during the period just prior to the feeding of these predictable meals. Thus, the predictable and restricted number of meals routinely provided to captive giant pandas may be a source of disturbance or competition among conspecifics that elicits stereotypic behavior. This study first quantified feeding anticipatory behavior, and then modified the predictability of meals to determine if that altered anticipatory behavior.

Subjects were nine adult giant pandas (2 males, 7 females) housed at the Chengdu Research Base. In the manipulation or treatment phase of the study, the males were simultaneously given 1) their twice daily concentrated meals on a more unpredictable schedule, and 2) more frequent provisioning of fresh bamboo (the volume of food was held constant). In light of possible pregnancies among females, the Research Base’s team of veterinarians and scientists decided that the females’ concentrated diet should be held constant. Thus, in the treatment phase of the study, the females were given only more frequent provisioning of fresh bamboo (amount held constant). No modifications were made to the delivery of their concentrated diet.

The females spent significantly more time engaged in several behaviors in the 30-minute periods prior to feeding of a concentrated meal when compared to nonfeeding periods, including door-directed/human-oriented behavior, stereotypic behavior, and non-stereotypic locomotion. Across all data collection categories, no significant differences were found between study phases for the above-mentioned behaviors of interest. Stereotypic behavior decreased during the treatment phase but the finding only approached significance.

We did not find significant differences in behaviors of interest between experimental phases or observation periods in the males’ data. These findings may be attributable in part to the low power inherent in such a small sample size. However, some visual trends indicating the males displayed feeding anticipatory activity were apparent. Furthermore, visual trends indicated that certain types of stereotypic activity increased during the treatment phase, particularly for one male.

When moved to a more unpredictable feeding schedule, the males exhibited a trend towards increased stereotypic activity. This indicates that the disruption of the heretofore predictable feeding schedule may have been a source of distress, particularly for one male. It is important to note that the animals may still have been going through an initial learning phase while data were being collected. Whether this schedule shift may or may not have been beneficial in the long run is yet unclear. Further, an animal’s reaction to such a schedule change may be dependent on the manner in which this change is implemented. It is possible that a more gradual schedule change will produce different behavioral changes. These results underscore the importance of carefully evaluating the behavior of captive giant pandas under various feeding regimes.

The trend toward decreased stereotypic behavior in the females in the more-frequent bamboo condition indicated that this was a beneficial feeding manipulation. As a result, personnel at the Chengdu Research Base have increased the number of times bamboo is provisioned throughout the day from three to six or more feedings for all pandas. This illustrates that systematic study of management practices can provide information that convinces animal managers to modify and improve husbandry routines to promote well-being.