Group-living animals are faced with the challenge of sharing space and local resources amongst group members who may be either relatives or non-relatives. Individuals may reduce the inclusive fitness costs they incur from competing with relatives by either reducing their levels of aggression towards kin or by maintaining physical separation between kin. In this field study, we used the group-living cichlid Neolamprologus multifasciatus to examine whether within-group aggression is reduced among group members that are kin and whether kin occupy different regions of their group’s territory to reduce kin competition over space and local resources. We determined the kinship relationships among cohabiting adults via microsatellite genotyping and then combined these with spatial and behavioural analyses of groups in the wild. We found that aggressive contests between group members declined in frequency with spatial separation between their shelters. Female kin did not engage in aggressive contests with one another, whereas non-kin females did, despite the fact these females lived at similar distances from one another on their groups’ territories. Contests within male-male and male-female dyads did not clearly correlate with kinship. Non-kin male-male and male-female dyads lived at more variable distances from one another on their territories than their corresponding kin dyads. Together, our study indicates that contests among group members can be mediated by relatedness in a sex-dependent manner. We also suggest that spatial relationships can play an important role in determining the extent to which group members compete with one another.

We sampled wild groups of Neolamprologus multifasciatus in the wild and recorded the group members' within-group social/contest behaviours and spatial positioning on the territory. We then used microsatellite genotyping to estimate pairwise relatedness among all group members and tested how relatedness correlated with within-group contests and spatial separation.

We used the statistical software R to conduct all our analyses.