An evaluation of the geographic method for recognizing innovations in nature, using zoo orangutans

Innovation and social learning are the raw materials for traditions and culture. Of these two, innovation has received far less scrutiny, largely because of difficulties assessing the innovation status of behaviors. A recent attempt proposes recognition of innovations in natural populations based on assessment of the behavior's properties and its geographic and local prevalence. Here we examine the validity of this approach and the list of 43 potential innovations it generated for wild orangutans by extending the comparison to zoo orangutans. First, we created an inventory of the behavioral repertoire in the zoo population. Four of ten putative innovations recognized in the field and potentially present in captivity did not occur despite appropriate conditions, suggesting they are indeed innovations. Second, we experimentally produced relevant conditions to evaluate whether another five potential innovations could be elicited. Based on their continued absence or on their latencies relative to known behaviors, four of the potential innovations could be assessed as innovations and one as a modification. Because 53% of relevant innovations recognized in the field could be confirmed in this analysis, and another 27% assigned possible innovation status, we conclude that the geographic method for detecting innovation in the wild is valid. However, the experiments also yielded up to 13 additional innovations, suggesting that zoo orangutans are far more innovative than wild ones. We discuss the implications of this latter finding with regard to limiting factors for the expansion of cultural repertoires in wild orangutans.