How males form coalitions against group rivals and the Pandit/van Schaik coalition model

Abstract

In group living animals, there is pronounced variation in the formation and function of cooperation between males via coalitionary aggression. Pandit, van Schaik and colleagues developed a mathematical model to predict the evolution of different coalition types in group-living male primates, the PvS model. Coalitions are classified into five types dependent on the ranks of the participants and the function of the aggression. The main factor determining the coalition types expected to evolve is contest potential, an estimate of female monopolisability by individual males. We examined the model using groups of Assamese (Macaca assamensis) and Barbary macaques (Macaca sylvanus) to gain a full range of contest potentials. We observed, across groups, 393 coalitions during 3645 h of data collection. We measured contest potential on a species-specific basis dependent on the information males can infer about female reproductive state. By examining coalition formation in different populations and species, but also in the same groups over time, we showed the strengths and weaknesses of the PvS model. We discuss why our results do not fully fit the model’s predictions, including differing costs/benefits of coalition formation, such as delayed benefits via increased status, making rank-changing coalitions viable at mid–low contest potential. Alternative factors not considered by the model include the effect of male social bonds on coalition partner choice and the effect of female mate-choice on coalition target selection. Finally, we suggest possible improvements to the model and provide information on how best to test the current predictions of the PvS model.