Do capuchin monkeys engage in calculated reciprocity?

Most research on reciprocal cooperation aims to either investigate the conditions that allow its evolution or to document actual cases of reciprocity, while the study of its proximate mechanisms is often neglected. Here, we report on an experiment aimed at testing one of the proximate mechanisms that have been proposed to underlie reciprocal cooperation. We tested 'calculated reciprocity', according to which animals behave cooperatively towards conspecifics driven by the expectation of a return of the favour. We examined food sharing in pairs of tufted capuchin monkeys, Sapajus spp., tested in three different experimental conditions. In the first condition (Reciprocity), we gave the subject some less preferred food that could be shared with the partner and, in a following phase, we gave the partner some preferred food that could be shared with the subject. In the second condition (Control), we gave the subject some less preferred food that could be shared with the partner, but the partner did not have the opportunity to reciprocate. In the third condition (Rewarded Cooperation), we gave the subject some less preferred food that could be shared with the partner, and then we gave the subject a quantity of preferred food proportional to the amount shared with the partner during the test. There was no difference between the three experimental conditions in the amount of food transferred from the subject to the partner. However, in the Rewarded Cooperation condition, capuchin monkeys increased the amount of food shared over successive test sessions. Although it remains undemonstrated whether the subjects understood the differences between the three experimental conditions, our results suggest that the expectation of reciprocation did not motivate capuchin monkeys to share their food. Capuchin monkeys seemed, however, able to learn that sharing can lead to a reward.