Sexual segregation in ungulates: individual behaviour and the missing link

Previous 'explanations' of sexual segregation in ungulates establish no more than a prerequisite for habitat segregation because they do not include a model of competitive habitat selection. Here we provide one based on the ideal free distributions of mutually competing, optimally foraging, individual deer. We parameterised our model using field data collected from a population of fallow deer (Dama dama) in a Mediterranean forest. The predictions of the inter-sex competition model were in full agreement with observational data, but those of single sex distributions (conventional theory) were not. The 'conventional' hypothesis, that segregation arises simply from sex differences, predicted no more than moderate (20-40%) levels of segregation, even in optimal conditions. By introducing inter-sex resource competition, the predicted segregation can generally more than double and full segregation becomes possible in some circumstances. The modelling showed segregation to be density-dependent, varying in complicated ways with season and animal density. Sensitivity analysis showed the vulnerability of the 'conventional' understanding of environmental variation and uncertainty. Using our competition model we show that as diet difference increases, direct competition between the sexes declines, so that as males increasingly differ from females, segregation declines and the two sexes are more likely to be found mixed (as long as the chosen food is available to both in the same area). Conversely, small differences among male and female deer are amplified by both food depletion and inter-sex competition to give substantial levels of segregation. The theoretical framework on which our model is built strongly suggests that sexual dimorphism in the context of scramble competition may be the fundamental cause of sexual habitat-segregation among ungulates.