Implementing social network analysis to understand the socioecology of wildlife co‐occurrence and joint interactions with humans in anthropogenic environments

Abstract

1. Human population expansion into wildlife habitats has increased interest in the behavioural ecology of human-wildlife interactions. To date, however, the socioecological factors that determine whether, when or where wild animals take risks by interacting with humans and anthropogenic factors still remains unclear.2. We adopt a comparative approach to address this gap, using social network analysis (SNA). SNA, increasingly implemented to determine human impact on wildlife ecology, can be a powerful tool to understand how animal socioecology influences the spatiotemporal distribution of human-wildlife interactions.3. For 10 groups of rhesus, long-tailed and bonnet macaques (Macaca spp.) living in anthropogenically impacted environments in Asia, we collected data on humanmacaque interactions, animal demographics, and macaque-macaque agonistic and affiliative social interactions. We constructed 'human co-interaction networks' based on associations between macaques that interacted with humans within the same time and spatial locations, and social networks based on macaque-macaque allogrooming behaviour, affiliative behaviours of short duration (agonistic support, lip-smacking, silent bare-teeth displays and non-sexual mounting) and proximity.4. Pre-network permutation tests revealed that, within all macaque groups, specific individuals jointly took risks by repeatedly, consistently co-interacting with humans within and across time and space. GLMMs revealed that macaques' tendencies to co-interact with humans was positively predicted by their tendencies to engage in short-duration affiliative interactions and tolerance of conspecifics, although the

Impact of joint interactions with humans and social interactions with conspecifics on the risk of zooanthroponotic outbreaks among wildlife populations

“…To capture patterns of macaque-macaque social interactions, we constructed social 'grooming networks' that linked macaques based on the proportions of time they spent engaging in grooming their conspecifics, given the extensive previous literature on the importance of grooming patterns in defining primate social structure 44,45 . In a previous study, we revealed that macaques' grooming relationships did not predict their tendencies to co-interact with humans 43 . Through this finding, we established a premise to expect that the patterning and distribution of such co-interactions with humans can be different, somewhat independent pathways for disease transmission than their social structure defined through the patterning and distribution of macaque grooming of conspecifics 43 .…”

“…In a previous study, we revealed that macaques' grooming relationships did not predict their tendencies to co-interact with humans 43 . Through this finding, we established a premise to expect that the patterning and distribution of such co-interactions with humans can be different, somewhat independent pathways for disease transmission than their social structure defined through the patterning and distribution of macaque grooming of conspecifics 43 .…”

“…In doing so, we evaluated the relative vulnerability of these wildlife populations to human-induced disease outbreaks, through both their joint aggregations around and co-interactions with humans, and through their social interactions with conspecifics that underlie their social structure. To capture patterns of macaques' co-interactions with humans, we constructed networks of macaques' (nodes) tendencies to jointly co-occur around and engage with humans (edges) within the same time and location in the context of anthropogenic spaces 43 . To capture patterns of macaque-macaque social interactions, we constructed social 'grooming networks' that linked macaques based on the proportions of time they spent engaging in grooming their conspecifics, given the extensive previous literature on the importance of grooming patterns in defining primate social structure 44,45 .…”

“…Since females and highranking individuals form the core of macaque grooming networks 40,41 , we expected that across species, outbreak sizes through grooming networks would be higher when the first-infected individuals were females (versus males) and higher-ranking (versus lower-ranking) individuals. On the other hand, given the exploratory and increased risk-taking behavior of males resulting in their being more well-connected in co-interaction networks compared to females 43,47 , we expected that outbreak sizes through co-interaction networks would be higher when the first-infected individuals are males (versus females). We also explored whether inter-individual differences in their overall anthropogenic exposure (measured as the total frequencies of interactions with humans, and foraging on anthropogenic foods) impacted model-predicted outbreak sizes through both types of networks.…”

Measuring dominance certainty and assessing its impact on individual and societal health in a nonhuman primate model: a network approach

“…From 2016 to the present day, Perc has been cited in a total of 29 publications (a few are reviewed below). Among just studies focusing on nonhuman primates, studies implementing Perc have ranged from testing phylogenetic and socioecological frameworks of primate social structure [66]; to assessing the links between dominance certainty, health and (by extension) well-being of captive macaques [35,38]; to determining the effects of rank on the socio-ecology of behavioural flexibility and risk-taking behaviours by wild primates through their interacting with humans and other anthropogenic factors [70][71][72]. Outside of nonhuman primates, Perc has also been used to generate metrics of transitivity for studies of sociality in ducks [73].…”