Infanticide as an evolutionarily stable strategy

An evolutionarily stable strategies analysis of infanticide, expressed as two pure strategies, shows that the equilibrial frequency of infanticide depends on only two parameters: the relative advantage of infanticide when others are also infanticidal and the relative advantage of a non-infanticidal strategy if others are non-infanticidal. The possible equilibrial conditions of strictly infanticidal, strictly non-infanticidal, stably polymorphic or history-dependent populations correspond to various earlier models of infanticide. Thus, these earlier models may be subsumed as special cases of the ESS analysis. This permits the comparison of alternative theories of infanticide in specific cases of infanticide in natural populations. In particular, it is argued that if infanticide confers a relative advantage on the perpetrator when other population members are non-infanticidal, this is sufficient to reject a 'maladaptive' explanation for infanticide. The potential existence of history-dependent equilibria suggests there will be extreme difficulty with non-experimental methods that attempt to test the adaptiveness of infanticidal behaviour.