REMOVALS BASED ON LOW WITHIN-MATRILINE RELATEDNESS CAN REDUCE SEVERE AGGRESSION IN CAPTIVE RHESUS MACAQUES |
D. Hannibal1,2, B. Beisner1,2, A. Nathman1, A. Maness1 and B. McCowan1,2
1California National Primate Research Center, UC Davis, Davis, CA 95616, USA, 2Department of Population Health & Reproduction |
 |
| Previous observational research of captive rhesus macaques suggests that higher coefficients of relatedness within matrilines are associated with lower rates of severe within-matriline aggression and more cohesive grooming communities. We tested whether experimental removals targeted to increase the relatedness coefficient of remaining matriline members leads to fewer grooming communities and a lower ratio of severe aggression within matrilines. The study sample was comprised of 100 adult females within 13 matrilines in 1 mixed-age-sex group. The Girvan-Newman method was used to detect communities and modularity (score Q) within matrilines. We used generalized linear regression models to determine the best model predicting the ratio of severe within-matriline aggression. Seven matrilines received experienced removals and results show fewer grooming communities in those matrilines (Beta=-1.33, p=0.04), but not a reduction in modularity (Beta=0.04, p=0.99). The coefficient of relatedness was negatively correlated with number of grooming communities (rho=-0.52, p<0.001) and modularity (rho=-0.58, p<0.001), so these had to be tested separately rather than in one model. The number of grooming communities (Beta=0.11, p=0.03) within matrilines was a better predictor of the ratio of severe aggression than coefficient of relatedness (Beta=0.72; p=0.36) or modularity (Beta=-0.14, p=0.82). These results show that the coefficient of relatedness can be a useful management tool to selectively remove individuals from captive rhesus macaque social groups as long as it results in fewer grooming communities within targeted matrilines. |
|
|