Abstract # 13309 Event # 139:

Scheduled for Friday, August 23, 2019 03:00 PM-03:15 PM: (Room 326) Oral Presentation


PREDICTORS OF TRAUMA DIFFER FOR HIGH AND LOW-RANKED COMPARED TO MIDDLE-RANKED RHESUS MACAQUES (MACACA MULATTA)

J. J. Vandeleest, B. Beisner, D. L. Hannibal, K. Fujii, K. Balasubramaniam, F. Hsieh and B. McCowan
University of California-Davis, California National Primate Research Center, One Shields Avenue, Davis, CA 95616, USA
line
     Social trauma is a common concern in captive, group-housed rhesus monkeys. A greater understanding of the predictors of trauma can improve welfare for socially housed macaques. We studied a group of rhesus macaques (N=101 adults, 27 male) for 6 weeks to examine whether rank, aggression, status signaling, or conflict policing behavior predicted trauma. Agonistic interactions were recorded using event sampling, and incidence of trauma (e.g., lacerations, punctures) was assessed twice weekly. Use of a new multivariate clustering method (Data Mechanics) revealed patterns not discernable from standard multiple regression. Data Mechanics clustered subjects into 8 groups (2 of high-rank, 5 of middle-rank, and 1 of low-rank animals) and clustered social behavior into 4 factors relating to receiving aggression, giving aggression, conflict policing, and submissive signaling. Results showed two broad patterns. First, trauma incidence was lowest in clusters of high-ranking subjects (9-18%) and highest in low-ranking subjects (75%). Second, among middle-ranking subjects trauma incidence was more complex. The two subject clusters that experienced more trauma than the other three (60-64% vs. 27-39%) also engaged in more conflict policing (Mann-Whitney tests, p<0.05). These results suggest that rank is only a strong predictor of trauma risk at the extremes of the hierarchy, while among middle-ranking animals, participation in risky social interactions, such as conflict policing, explains trauma risk--a pattern not identified in traditional methods.