Brandon C. Wheeler, Clara J. Scarry, Andreas Koenig

Agonism is common in group-living animals, shaping dominance relationships and ultimately impacting individual fitness. Rates of agonism vary considerably among taxa, however, and explaining this variation has been central in ecological models of female social relationships in primates. Early iterations of these models posited a link to diet, with more frequent agonism predicted in frugivorous species due to the presumed greater contestability of fruits relative to other food types. Although some more recent studies have suggested that dietary categories may be poor predictors of contest competition among primates, to date there have been no broad, cross-taxa comparisons of rates of female-female agonism in relation to diet. This study tests whether dietary variables do indeed predict rates of female agonism and further investigates the role of group size (i.e., number of competitors) and substrate use (i.e., degree of arboreality) on the frequency of agonism. Data from 44 wild, unprovisioned groups, including 3 strepsirhine species, 3 platyrrhines, 5 colobines, 10 cercopithecines, and 2 hominoids were analyzed using phylogenetically controlled and uncontrolled methods. Results indicate that diet does not predict agonistic rates, with trends actually being in the opposite direction than predicted for all taxa except cercopithecines. In contrast, agonistic rates are positively associated with group size and possibly degree of terrestriality. Competitor density and perhaps the risk of fighting, thus, appear more important than general diet in predicting agonism among female primates. We discuss the implications of these results for socio-ecological hypotheses.