Choice of companions in social play in piglets

Play activity differences between sexes have been predicted in sexually dimorphic species. A non-random composition of pairs of players in social play has also been suggested. In order to test these proposals in piglets, two experiments were set up on a focal-animal basis.In the first trial, social play as observed in two sets of 16 weaned piglets, from 4 litters of 8 piglets, each with an equal number of males and females. A total of 133 and 101 plays, respectively, were recorded and analysed using a log-linear model including the sex and litter of each player and their interactions. The main results show: (a) a greater play activity in males than in females; (b) more play between partners of the same sex than expected assuming statistical independence; (c) more play between partners born in the same litter than expected assuming statistical independence.The role of the pheromones segregated by submaxillary salivary glands in the sexual recognition evidenced in the first experiment was tested in the second trial. Two sets of 16 piglets, coming from 16 different litters, were observed; one of them as control and the other with 8 males whose glands had been removed and 8 normal females. The analysis of 230 plays recorded confirms the results of the first trial concerning sex and does not show significant differences either in the total number of observed plays or in the preference for partners of the same sex between sets, implying a recognition mechanism other than that of these pheromones.The observed sexual dimorphism in play activity as well as the sex and littermate preferences are in agreement with the hypothesis which assigns adaptive values to these social plays.