U.S. Department of Agriculture: Animal and Plant Health Inspection Service
An analytical framework for quantifying and testing patterns of temporal dynamics in social networks
Date of this Version
2013
Citation
Animal Behaviour 85 (2013) 83-96; http://dx.doi.org/10.1016/j.anbehav.2012.10.010
Abstract
Change is fundamental to all social systems. Temporal dynamics are critical in understanding how relationships form and change over time but rarely are studied explicitly in animal groups. Social network approaches are useful in describing association patterns and provide promising tools for investigating the dynamics of change in social structure but have rarely been used to quantify how animal associations change over time. In this study, we describe and test a framework for temporal analysis of social structure. We propose an analytical framework of methods that integrates across social scales and comparatively analyses change in social structure across multiple types of social association. These methods enable comparisons in groups that differ in size and are flexible to allow application to weighted and unweighted networks, where ties can be directed or undirected, and relationships can be symmetric or asymmetric. We apply this analytical framework to temporal social network data from experimentally formed captive groups of monk parakeets, Myiopsitta monachus, to both evaluate our analysis methods and characterize the social structure of this species. We compared dynamics of dyadic network formation, ego network formation and global network stabilization patterns across neutral, affiliative and agonistic associations. We found that social structure of captive monk parakeets formed and stabilized over a short period, but patterns differed by social association type. We also found evidence for consistency in the temporal dynamics of formation and stabilization of social structure between replicate social groups. Our analysis methods successfully identified change in social structure that corresponded well with qualitative observations. This framework is likely to be useful in characterizing patterns of temporal dynamics in social structure in longitudinal data in wide variety of social systems and species.