Nesting with the wasp Ropalidia cincta increases nest success of Red-cheeked Cordonbleu (Uraeginthus bengalus) in Ghana

Associations between birds and social Hymenoptera (ants, wasps, bees) are common in tropical regions and are usually assumed to be commensal relationships that benefit birds but neither help nor harm the arthropods. However, benefits to birds have been documented in only four such associations, and no previous research has rigorously investigated costs or benefits to associated hymenopterans. We followed the nesting cycles of an estrildid finch, the Red-cheeked Cordonbleu (Uraeginthus bengalus), and a common nesting associate, the wasp Ropalidia cincta, during 2002 and 2003 in northern Ghana to compare reproductive success of birds and wasps nesting in association with that of birds and wasps nesting separately. Red-cheeked Cordonbleus and wasps nested together in the same tree 3.7× as often as expected if nesting decisions were made independently, with 74% of bird nests and 74% of wasp colonies occurring in associations. Bird nesting was initiated ∼33 days after founding of an associated wasp colony; bird nests and wasp colonies were, on average, 42 cm apart. In both years, Red-cheeked Cordonbleus in nesting associations with wasps were twice as likely to fledge young as birds nesting in trees without wasps. Reduced predation was apparently a major reason for increased fledging success: we documented four cases of nest predation on 122 Red-cheeked Cordonbleu nests associated with wasps, and 11 cases on 90 nests not associated with wasps. Association with birds did not affect the success of wasp colonies. Although our observational study cannot rule out the possibility that both species coincidentally shared a preference for a habitat feature in limited supply, suitable nest sites did not appear to be limiting (74% of potential nest trees had neither bird nor wasp nests). Reproductive success of Red-cheeked Cordonbleu populations in this region may be limited by the number of available wasp colonies. By designing our study to address four working hypotheses (commensalism, mutualism, parasitism, coincidence of habitat preference), we have provided strong evidence that this relationship is commensal.