Socially informed dispersal in a territorial cooperative breeder

Abstract

Dispersal is a key process governing the dynamics of socially and spatially structured populations and involves three distinct stages: emigration, transience and settlement. At each stage, individuals have to make movement decisions, which are influenced by social, environmental and individual factors. Yet, a comprehensive understanding of the drivers that influence such decisions is still lacking, particularly for the transient stage during which free-living individuals are inherently difficult to follow. Social circumstances such as the likelihood of encountering conspecifics can be expected to strongly affects decision-making during dispersal, particularly in territorial species where encounters with resident conspecifics are antagonistic. Here, we analysed the movement trajectories of 47 dispersing coalitions of Kalahari meerkats Suricata suricatta through a landscape occupied by constantly monitored resident groups, while simultaneously taking into account environmental and individual characteristics. We used GPS locations collected on resident groups to create a georeferenced social landscape representing the likelihood of encountering resident groups. We used a step-selection function to infer the effect of social, environmental and individual covariates on habitat selection during dispersal. Finally, we created a temporal mismatch between the social landscape and the dispersal event of interest to identify the temporal scale at which dispersers perceive the social landscape. Including information about the social landscape considerably improved our representation of the dispersal trajectory compared to analyses that only accounted for environmental variables. The latter were only marginally selected or avoided by dispersers. Before leaving their natal territory, dispersers selected areas frequently used by their natal group. In contrast, after leaving their natal territory, they selectively used areas where they were less likely to encounter unrelated groups. This pattern was particularly marked in larger dispersing coalitions and when unrelated males were part of the dispersing coalition. Our results suggest that, in socially and spatially structured species, dispersers gather and process social information during dispersal, and that reducing risk of aggression from unrelated resident groups outweighs benefits derived from conspecific attraction. Finally, our work underlines the intimate link between the social structure of a population and dispersal, which affect each other reciprocally.

Density‐dependent dispersal strategies in a cooperative breeder

“…Thus, staying at home and gaining inclusive fitness through cooperation can be an adaptive life-history strategy when vacancies are unavailable (Kokko and Lundberg 2001). In meerkats, this hypothesis is further supported by the fact that evicted females were more likely to return to the natal group at high population densities, and by previous findings, showing that dispersing meerkats avoided areas where conspecifics were found (Cozzi et al 2018). We found higher emigration and settlement rates after periods with low rainfall.…”

“…In meerkats, this hypothesis is further supported by the fact that evicted females were more likely to return to the natal group at high population densities, and by previous findings, showing that dispersing meerkats avoided areas where conspecifics were found (Cozzi et al. ). We found higher emigration and settlement rates after periods with low rainfall.…”

“…A detailed description of the methods and smoothing parameter estimators can be found in Cozzi et al. (). All parameter calculations and statistical analyses were done in R (R Core Team ).…”

“…We defined the size of the study area as the combination of 95% kernel home ranges of all resident groups (Calenge 2006). A detailed description of the methods and smoothing parameter estimators can be found in Cozzi et al (2018). All parameter calculations and statistical analyses were done in R (R Core Team 2013).…”

Estimating interactions between individuals from concurrent animal movements

“…Step‐selection functions are a popular tool to identify animals’ selective behaviour towards landscape features or in relation to their previous visits and experiences (Latombe, Fortin, & Parrott, ; Oliveira‐Santos, Forester, Piovezan, Tomas, & Fernandez, ; Thurfjell, Ciuti, & Boyce, ). Recently, SSFs have further been used to investigate the influence of the conspecific social environment on dispersing meerkats, Suricata suricatta (Cozzi, Maag, Börger, Clutton‐Brock, & Ozgul, ), where the social environment was static and estimated as kernel utilization density from pooled low‐resolution positions. In our approach, we use the same temporal resolution for all individuals and relate individual movements to each other in a dynamic framework.…”