Adult-adult play in primates: Comparative analyses of its origin, distribution and evolution

Comparative analyses were conducted on a data set derived from the literatu re so as to test several hypotheses which were developed to explain the dis tribution of adult-adult play fighting within the order primates. Ratings f or play occurring in sexual and non-sexual contexts were developed. Three h ypotheses were evaluated: (i) that play occurring in non-sexual social cont exts is a byproduct of its use in sex; (ii) that the occurrence of play is related to its use for social assessment and manipulation, and so is more l ikely to be present in species with reduced familiarity between individuals ; and (iii) that phylogenetic affiliation influences the likelihood that sp ecies within clades engage in play. We used independent contrasts to test t he first two hypotheses, and both were significant, with the presence of pl ay in sexual contexts accounting for 14-16% of the variance of play in nons exual contexts, and reduced social familiarity accounting for 30-40% of the variance in the occurrence of play in non-sexual contexts. To test the thi rd hypothesis, we mapped the occurrence of both types of play onto known ph ylogenies. The overlap was not congruent, indicating that phylogenetic rela tionships did not account for the distribution of play. Given that play in both sexual and non-sexual contexts was more likely to occur in species wit h a social organization involving reduced frequency of contact between the sexes and other social group members, we suggest that the likely adaptive v alue of play fighting is as a tool for social assessment and manipulation. The possible factors that mitigate the use of play fighting for these purpo ses, such as the availability of other forms of communication that could se rve similar functions, are discussed.