Facebook for Geese: The Causes and Consequences of Non-random Social Associations in a Group Forager

Abstract

The application of social network analysis in animals has facilitated research into dynamic fission-fusion social systems. These have important implications for the evolution of individual social behaviour, and for population-level processes such as information transfer and disease dynamics. This thesis explores the assumptions behind using networks to study animal social systems in projects using individual-marking or biologging. It then applies these methods to study social structure in a study population of a long-distance migrant, the light-bellied brent goose. It provides new insights about the causes and consequences of social structure, and individual social strategies, in a fission-fusion social system in the context of a migratory cycle. We show that social networks have a strong spatial structure, but with additional non-randomness once these spatial constraints have been accounted for. However, individual social associations are seasonally dynamic. These social structures, and their seasonal dynamics, are highly stable between years. Furthermore, non-random associations have important implications for foraging success. Individuals foraging in more familiar flocks are able to spend more time feeding, and less time involved in aggressive interactions or vigilant. This results in social network position influencing the ability of some individuals to gain body condition during spring staging and leave for breeding grounds in better condition. These results highlight the importance of understanding social networks when investigating individual time-budgets in social foragers. They also emphasise the importance of establishing the link between individual status and social network position before drawing any conclusions about the role of social network position in explaining differences in fitness between individuals in fission-fusion social systems.