Reproductive skew and the evolution of group dissolution tactics: a synthesis of concession and restraint models

Reproductive skew theory provides a compelling explanation for the partitioning of reproduction among individuals within animal societies. One constructive criticism of the theory is that there are too many models, all of which have different assumptions and predictions, which makes it difficult to know what to test. Here we begin the process of tackling this problem, by re-examining the assumptions and predictions of basic concession and restraint models, two transactional models that are often tested as alternatives. Concession models assume that the dominant has complete control over the allocation of reproduction but may yield some of the group's reproduction to prevent the subordinate from voluntarily departing. Restraint models assume that the subordinate has complete control over the allocation of reproduction but may not claim all of the group's reproduction to prevent the dominant from forcibly evicting it. We show that the group dissolution tactics that individuals use (forcible eviction or voluntary departure) need not be an assumption of the model, but rather they can be predicted using Hamilton's rule and the standard variables of skew models. We reveal that the assumption that one individual (dominant or subordinate) has complete control over the allocation of reproduction is an idea common to both models, and we resolve this semantic difference by calling this individual ‘the allocator’. We show that, regardless of the group dissolution tactics that individuals adopt, the allocator's share of the reproduction always increases as relatedness increases, as group productivity increases, and as constraints on leaving to breed elsewhere intensify. We conclude that concession and restraint type models make qualitatively similar predictions, and should not be tested as alternatives. In summary, this study makes the transactional framework of reproductive skew more general, by eliminating restrictive assumptions, and more amenable to testing in the field, by clarifying assumptions and predictions.