Behaviour 2019
Assessment During Intergroup Contests in Animal Societies
Patrick A. Green, Michael A. Cant. University of Exeter Centre for Ecology and Conservation, Penryn, Cornwall, United Kingdom

A central goal of the study of animal behavior is to explain variation in contest outcomes and processes—who wins, and why. Much of our knowledge of these questions has come from studies of assessment during dyadic, or one-on-one, fights. More recently, there has been growing interest in intergroup contests, those between groups of social-living animals, and what these tell us about the evolution of collective aggression. Here, we describe how dyadic contest theory can be adapted to the intergroup case. We test this theory using over 20 years’ data on a wild population of an intensely territorial social mammal, the banded mongoose (Mungos mungo). We analyzed over 300 intergroup contests to test how a group’s ability to win is predicted by the number, sex, social status (i.e., dominant versus subordinate), and traits (e.g., weight, age) of its members. We also asked whether the average traits of the group or traits of “key individuals” are more important to competitive success. This study reveals how individual attributes, group composition, and information-gathering combine to determine the outcomes of intergroup contests