Universal behaviors as candidate traditions in wild spider monkeys.

Candidate traditions were documented across three communities of wild spider monkeys (Ateles geoffroyi) using an a priori approach to identify behavioral variants and a statistical approach to examine differences in their proportional use. This methodology differs from previous studies of animal traditions, which used retrospective data and relied on the 'exclusion method' to identify candidate traditions. Our a priori approach increased the likelihood that behavior variants with equivalent functions were considered and our statistical approach enabled the proportional use of 'universal' behaviors, i.e., used across all communities, to be examined for the first time in any animal species as candidate traditions. Among universal behaviors we found 14 'community preferred' variants. After considering the extent to which community preferred variants were due to ecological and, to a lesser degree, genetic differences, we concluded that at least six were likely maintained through social learning. Our findings have two main implications: (i) tradition repertoires could be larger than assumed from previous studies using the exclusion method; (ii) the relative use of universal behavior variants can reinforce community membership.