Dynamic social networks in guppies (Poecilia reticulata)

Abstract

One of the main challenges in the study of social networks in vertebrates is to close the gap between group patterns and dynamics. Usually scan samples or transect data are recorded to provide information about social patterns of animals, but these techniques themselves do not shed much light on the underlying dynamics of such groups. Here we show an approach which captures the fission-fusion dynamics of a fish population in the wild and demonstrates how the gap between pattern and dynamics may be closed. Our analysis revealed that guppies have complex association patterns that are characterised by close strong connections between individuals of similar behavioural type. Intriguingly, the preference for particular social partners is not expressed in the length of associations but in their frequency. Finally, we show that the observed association preferences could have important consequences for transmission processes in animal social networks, thus moving the emphasis of network research from descriptive mechanistic studies to functional and predictive ones.

Social dynamics, network structure, and information diffusion in fish shoals.

“…Simulated outputs from a number of Markov chain models were compared to the observed data to assess goodness-of-fit (Figure 1.4). Wilson et al (2014) found that the best-fitting model had focal individuals selecting shoaling partners with individualspecific probabilities, suggesting active social preferences were at work in this population. These Markov chain models were then used to generate simulated networks whose structure was compared to that of real networks constructed from the observed shoaling partner data.…”

“…If transmission processes occur at a fast rate relative to changes in network structure, then NBDA is a powerful technique. If instead transmission processes and structural changes occur over similar timescales, then a more fine-grained view of social dynamics might be needed Wilson et al, 2014). Furthermore, if acquiring a trait changes an individual's interaction patterns, then a social network constructed prior to when an individual acquired a trait might not accurately reflect its social relationships after trait acquisition.…”

“…Understanding the dynamics and structure of a population's social network provides us with predictive power with respect to these processes and can enhance our understanding of how social organization influences individual behavior (e.g., Croft et al, 2006;Wilson et al, 2014).…”

“…SNA identifies emergent structural properties that arise from complex patterns of social relationships and explores how these properties influence individual behavior occurring within the network and social processes playing out over it (Newman, 2003). SNA also tries to understand the ecological factors and underlying social dynamics that result in the emergence of particular structural elements (e.g., Barabási & Albert, 1999;Wilson et al, 2014). To accomplish these goals, modern SNA has a wide array of tools at its disposal, including: descriptive measures, network models, simulations, and comparative approaches (e.g., Pinter-Wollman et al, 2014).…”

“…Wilson et al (2014) conducted repeated focal follows of wild female guppies (P. reticulata) in their native habitat of Trinidad, periodically recording whether or not the focal fish was shoaling with another female and if so, with who. From these observations, they constructed Markov chain models describing shoaling dynamics, where the shoaling behavior of an individual at time t + 1 depends solely on its behavioral state at time t, and each behavioral state-e.g., shoaling, swimming alone-is associated with a unique set of transition probabilities describing the likelihood of future states.…”