Conflict over reproduction in a cooperative mammal: an experimental approach

Animals that live in cooperative societies face inherent conflicts of interest which must be resolved in some way for the society to function. The most fundamental conflict concerns the distribution of reproduction. Animal societies vary enormously in the degree of 'reproductive skew', which measures the evenness with which reproduction is distributed among group members. A large number of evolutionary models have been proposed to account for why societies vary in this way. These models are based on different principles but nevertheless make similar or overlapping predictions, with the consequence that even the best attempts to test the models have up to now been uninformative. The best way to distinguish the models and provide a much richer understanding of reproductive conflict is to manipulate the degree of skew experimentally. We will carry out the first experiment of this kind large-scale on a free-living population of banded mongooses that we have studied for the last eleven years. This species is unusual because nearly all adult females (up to ten) in each group give birth together in the same den, and offspring are cared for by adults of both sexes. We have shown in a pilot study that we can use short-acting contraceptive injections to block pregnancy in breeding females for a single breeding attempt, thereby temporarily increasing the degree of skew while keeping group size and other variables constant. The behavioural responses to the manipulation, combined with genetic analyses of maternity, reproductive success and the kin structure of the group, will enable us to measure the costs of co-breeding, determine which individuals exert control over the distribution of reproduction, and provide information on the behavioural mechanisms employed in reproductive conflict. Moreover, our experiment will test for the first time the direction of causality between variation in the distribution of reproduction and variation in individual social behaviour, a important but untested assumption of models of skew and cooperation.