The influence of demographic variation on social network stability in wild vervet monkeys

From a cognitive point of view, management and knowledge of social relationships is thought to be very challenging. Because of ecological and demographic constraints, relationships are likely to be prone to variation and hence need constant updating. Social network analysis is a potential tool to quantify the information that needs to be processed. However, despite the growing number of studies on social networks, few have focused on their dynamics and how they evolve across time. Here we present one of the rare studies that tests the influence of demographic variation on social relationships' stability through temporal analysis. Using field data collected on three wild groups of vervet monkeys, Chlor-ocebus aethiops, we first analysed the relationships' stability by running correlations between 3-month periods. Then, we investigated how natural demographic variation changed individual centralities (eigenvector) and strength of dyadic relationships within both grooming and proximity networks over a period of 2 years. In vervets, females are philopatric, while males emigrate from their natal group. Thus, we tested whether changes in demography had more influence on network centrality measures and relationship strength in females and their juveniles than in males. Correlations between periods yielded no evidence that predictability of future relationship quality declined with time from current relationship quality. In addition, male immigration was mostly responsible for increases in the core group members' centrality while male emigration had the opposite effect. Regarding dyadic relationships, we found inconsistent patterns that varied with respect to how age/sex and immigration/disappearance affected the network studied (grooming versus proximity). Our findings support the idea that social networks are dynamic structures that vary through time. Similar analyses on other species are needed to investigate which network features emerge as candidates responsible for variation in the complexity with which individuals keep track of relationships. (C) 2017 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.