Abstract
Previous research has suggested that chimpanzees (Pan troglodytes) display higher levels of aggression in captivity than in the wild. One of the challenges of captive management, therefore, is to balance the chimpanzees' need for social interaction with managements' desire to minimize wounding and aggression. Various captive studies have examined the effects of individual and social variables on the frequency of wounding aggression, but none have examined these variables simultaneously. We collected retrospective wounding data for severe wounds from 83 captive chimpanzees (36 males, 47 females) from January 1993 to December 2003. The context of the wounding event, including individual age and sex, group age and sex composition, group duration, and portion of the week (weekday vs. weekend) were collected. Logistic regression analysis was performed to determine which variables had a significant effect on the probability of a severe wounding event. The sex and age composition of the group, group duration, and portion of the week had a statistically significant association with wounding. All-male groups (Odds Ratio (OR)=6.738) had the highest risk of wounding aggression, with uni-male groups (OR=3.311) having the next largest. Compared to individuals in all sub-adult groups, individuals in either all-adult (OR=4.516) or mixed-age (OR=3.587) groups had a higher risk of wounding. There was an inverse association between group duration and wounding (OR=0.821). Finally, there was an increased risk of wounding during the work week (OR=1.653). These results suggest that captive management should pay close attention to group composition, as well as levels of human activity, when devising strategies to reduce captive chimpanzee aggression. Zoo Biol 29:351-364, 2010.