Male aggression varies with consortship rate and habitat in a dolphin social network

Abstract

Coalitions and alliances exemplify the core elements of conflict and cooperation in animal societies. Ecological influences on alliance formation are more readily attributed to within-species variation where phylogenetic signals are muted. Remarkably, male Indo-Pacific bottlenose dolphins in Shark Bay, Western Australia, exhibit systematic spatial variation in alliance behavior, not simply within a species or population, but within a single social network. Moving SE-NW along Peron Peninsula in Shark Bay, males ally more often in trios than pairs, consort females more often, and exhibit greater seasonal movements. Ecological models predict more male-male conflict in the north, but sufficient observations of aggression are lacking. However, dolphins often incur marks, in the form of tooth rakes, during conflicts. Here we report that the incidence of new tooth rake marks varies systematically in the predicted pattern, with greater marking in the north, where males form more trios and consort females at a higher rate. While our previous work demonstrated that alliance complexity has an ecological component, we can now infer that ecological variation impacts the level of alliance-related conflict in Shark Bay. To understand ecological influences on animal societies, researchers have focused on differences within species, where confounds due to evolutionary history are minimized. Such differences are usually found among geographically separated populations, but in Shark Bay, Western Australia, male dolphin alliance size and access to females increase along a spatial axis within a single social network. Here we report that aggression levels, evidenced by tooth rake marks, increase along the same axis. Alliances are of particular interest as they represent a complex kind of relationship, often implicated in the evolution of social intelligence. Our discovery of spatial variation in alliance behavior and aggression within a social network provides a unique opportunity to investigate the intersection of cognition, social structure, and ecology.