Costs and benefits in extreme nesting associations: do sociable weavers benefit from hosting African pygmy falcons?

Avian nesting associations are a prominent feature of breeding bird communities. Protective associations between a predator and prey species represent a scenario where typically antagonistic interacting species may confer benefits to each species. The outcomes of these interactions are likely to be context‐dependent and influenced by biotic and abiotic conditions. African Pygmy Falcons (Polihierax semitorquatus) are obligate nest associates of Sociable Weavers (Philetairus socius), using weaver colonies to breed and roost. As a result, the escalated rate of biotic interactions between associates may enhance rates of adaptation, speciation, and coevolution. Falcons occasionally prey on weavers but have the potential to defend colonies from nest predators. We used observational and experimental tests to determine if falcons deter snakes from accessing weaver colonies and if this increased nest survival for weavers that ‘host’ falcons in their colonies. We observed a reduction in the number of snakes at colonies hosting falcons and an increase in all colonies when weavers were breeding. Falcons were also more aggressive towards a snake stimulus than a control but only when they were breeding. However, weaver nest survival did not increase at colonies hosting falcons. Falcon defence likely reduces weaver nest predation by snakes; however, this is likely offset by nest predation by falcons. Additionally, we compared the breeding success of falcons whose breeding attempts overlapped with weaver breeding to those that did not. Weaver breeding did not explain falcon breeding success but did lead to an increase in falcon nest predation by snakes, and the likelihood that all chicks from a successful brood fledged, suggesting an ‘all or nothing’ scenario when weavers are breeding. In conclusion we show that both weavers and falcons incur costs and benefits of their close association (i.e., snake predation or food availability) and the net effects are likely to depend on the exact conditions in a particular breeding season. The intricacies of communalistic relationships continue to evade objective testing, and long‐term demographic monitoring may offer better proof of the net benefits for each species.