Social network analysis of mixed-species flocks: exploring the structure and evolution of interspecific social behaviour

Mixed-species social aggregations are common across taxa. There are two, nonexclusive, hypotheses typically proposed to explain the formation of social groups: increased predator vigilance and greater foraging efficiency. In mixed-species groups, these hypotheses are typically tested with species-level summary measures such as flocking propensity, the assignment of species-level roles, mean body size, and foraging and habitat characteristics. Literature syntheses make it clear that while these hypotheses are important, much about mixed-species groups remains unexplained. We suggest that we can substantially increase our understanding of the evolution and ecology of mixed-species social groups in terms of both traditional and novel hypotheses by shifting the analytical focus to bottom-up approaches common in intraspecific investigations of sociality. Bottom-up approaches to analyses of social structure treat pairwise interactions as the fundamental unit of analysis and social structure as an emergent property rather than relying on a priori assignments of species as units of association. The construction of social networks from pairwise interaction rates allows us to assess the factors that promote group formation on the basis of individuals, a more appropriate level of selection, rather than species groups. We illustrate this approach with data from mixed-species foraging assemblies in tits (Paridae), finding significant effects of dominance on social behaviour within species. This new focus allows us to address questions about active associations among heterospecifics, the role of individuals within mixed-species societies, and the role of environments, which will collectively provide a richer description of the evolution and function of mixed-species societies. Crown Copyright (C) 2012. Published on behalf of The Association for the Study of Animal Behaviour by Elsevier Ltd. All rights reserved.