Abundance and social organisation of male Asian elephants in northwestern India

Context Complex intrasexual associations are ubiquitous in social vertebrates. Such associations elucidate key sociobiology insights and inform endangered species management, particularly in human-dominated landscapes. In Asian elephants (Elephas maximus), male associations are widespread, eliciting a growing interest amid escalating negative human–elephant interactions, which often involve a preponderance of males. The robust estimation of elephant densities in such key areas is a key necessity for wildlife managers. Aims We aimed to understand the socio-ecological factors that govern the association patterns among male elephants in the Himalayan region of India. We also aimed to estimate the male density and abundance using robust statistical approaches. Methods To estimate the abundance of male elephants, we used a systematic polygon search method within grid cells to locate and individually identify elephants using morphological markers. The recapture history of the individual elephants was analysed using a spatially explicit capture–recapture (SECR) model to estimate abundance. Clear male photographs were used to record age, body condition and physiological state, along with the group type (mixed-sex, all-male and solitary). Using multinomial mixed-effect regression models, we assessed the relative influence of age, body condition, physiological state and habitat type in determining the group type in which the male occurred. Key results The density of adult males was estimated at 0.05 km−2 (s.e. = 0.01), with abundance of 40 (s.e. = 1.4) individuals in the study area. Younger males (juveniles and sub-adults) had a higher propensity to associate with mixed-sex herds than to be solitary or associate with all-male groups. The all-male groups were observed primarily in open habitats such as grasslands and near human habitation, suggesting potential habitat-driven association patterns. Conclusion We demonstrate that male associations in elephants are dynamic, influenced by both intrinsic factors such as age and physiological state, and extrinsic factors such as habitat. We demonstrated the practical utility of SECR for estimating elephant populations. Implications The demonstrated SECR approach has the potential for use at large spatial scales for robust estimation of population state variables. We recommend further research on stable all-male groups focusing on behavioural patterns and crop-raiding tendencies to better inform conflict mitigation strategies.