Sexual segregation among feral goats: testing between alternative hypotheses

Sexual segregation occurs widely in ungulates, and four principal hypotheses (predation, forage quality, social preferences and activity patterning) have been suggested as possible explanations. However, no single explanation has so far received widespread support. We used data on the grouping behaviour of feral goats, Capra hircus, on the Isle of Rum, Scotland, U.K. to test between competing hypotheses. We first used Conradt's (1999, Animal Behaviour, 57, 1151–1157) test to show that sexual segregation is not a consequence of sex differences in habitat preference. Since the predation risk hypothesis can be ruled out by both the absence of important predators on Rum and the tendency for those individuals most sensitive to predation (mother–offspring pairs) to be the furthest away from their neighbours, any explanation for sexual segregation must either lie in differences in activity patterning or be the consequence of a social preference for associating with same-sex individuals. We analysed data on the patterns of fission, behavioural synchrony and neighbourhood in relation to party size and composition to show that the activity budget hypothesis is the more likely. However, the data also suggest that the social preferences hypothesis none the less has a residual influence on segregation tendencies of the Rum goats. We conclude that one likely reason for the conflicting results in the literature may be that all four hypotheses in fact apply simultaneously, but that their relative weightings may depend on habitat- and species-specific characteristics.