Temporal dynamics of association patterns in Geoffroy's spider monkeys, Ateles geoffroyi: a multiscale approach using temporal multiplex networks

Collective patterns of sociospatial behaviour are the result of internal and external factors operating at different temporal scales and can give rise to social structures. In species with high degrees of fission–fusion dynamics, individuals can respond to behavioural drivers by joining or departing from subgroups, thereby modulating how often they co-occur in the same subgroups. Their collective patterns of co-occurrence can help us understand how groups respond to their changing environments. We examined association stability (defined as temporal consistency in patterns of co-occurrence in the same subgroup) in a wild group of Geoffroy's spider monkeys, Ateles geoffroyi , over 24 years, to investigate (1) regularities and disruptions that may give insights into the capacity of spider monkey groups (and those of other species with high degrees of fission–fusion dynamics) to persist through time given the social and ecological challenges faced by individuals and (2) how groups with high degrees of fission–fusion dynamics respond to external drivers and the limitations of their responses. Associations were analysed as temporal multiplex networks where each layer represented associations over a given period considering five temporal scales: fortnight, month, season, year and triennium. We used the structural reducibility analysis to detect temporal redundancy (interpreted as consistency) in association patterns and periods of association stability and instability across scales, including a salient drop in stability in the aftermath of two hurricanes. Two association stability metrics based on the similarity between chronologically adjacent layers revealed scale-dependent relations of association stability with demographic and social variables including subgroup size and its variability, certainty in social context and the number of newborn infants. Our results show the usefulness of multilayer networks to analyse social dynamics in species with high degrees of fission–fusion dynamics.