Determinants of social structure in a northern population of black-tailed prairie dogs, Cynomys ludovicianus

Animal sociality can vary in response to both biotic and abiotic factors, and variation in the rate and types of interactions that occur between individuals of a group can result in greatly differentiated social structure and ultimately social evolution. However, social structures are often generalized to the species or population level and do not consider potential shifts in structure over time. We used social network analysis to investigate the social structure of black-tailed prairie dogs (hereafter ‘prairie dogs’) and determine the demographic features that may underpin the observed social structures in a population that experiences increased seasonal shifts of environmental conditions compared to populations in the centre of the geographical range. Prairie dogs live in highly social colonies, display an elaborate range of social behaviours and experience very dynamic changes in body condition due to variable environmental conditions. We used exponential random graph models (ERGMs) to examine the effects of body condition, age, sex and spatial arrangement on static and temporal social networks. We expected that spatial arrangement would most strongly influence social structure given the delineated burrow systems of a prairie dog colony. We hypothesized that body condition influences social structure over time, where those in better body condition maintain the social structure. Prairie dogs associated preferentially with individuals near themselves and of the same age. Furthermore, we also determined that body condition correlated positively with maintaining social structure over time, where those in poorer body condition experienced an increase in the loss and gain of relationships compared to those in better body condition that maintained connections. Our study confirms previous conclusions of association based on age and also highlights the importance of considering dynamic individual traits such as body condition in the importance of long-term social structure. These results also show the promise of both static and temporal ERGMs in a biological system and demonstrate the potential for increased temporally driven hypothesis testing using this technique.