Spatial refugia and the coexistence of a diurnal raptor with its intraguild owl predator

Summary Intraguild predation (IGP) has been studied extensively in predatory fish, amphibians and mammals, but less so in top avian predators. We studied the impact of IGP by large eagle owlsBubo buboLinnaeus (body mass ∼1·5–4 kg) on diurnal black kitesMilvus migransBoddaert (∼0·5–1 kg) in eight populations located in the Italian Alps. Eagle owls preyed on both adult and nestling kites, especially when kites nested within 2 km of an owl nest. Food overlap was low as kites preyed mainly on fish and owls on mammals. Eagle owls were absent from two study areas, bred very close to another area and were present at medium–high densities in the other five. Within‐population effects varied and were most severe in areas with high owl densities. Overall, kites responded to predation risk through predator spatial avoidance, being concentrated in interstitial predation‐refugia bordering the core home range of owl pairs. Kite productivity declined steeply with increasing predation risk; no nestling fledged within 1 km of an owl nest. Brood predation was higher in an area with medium owl density than in an area lacking eagle owls. The abandonment of kite territories increased near owl nests, but close coexistence was maintained by new kite territories being established occasionally in areas of high predation risk and high food abundance, which probably functioned as ecological traps. Compared to a random nest dispersion, colonial nesting was avoided within 1 km of an owl pair and peaked in conditions of medium predation risk. At the population level, kite density and productivity were related to a complex interaction of IGP risk and food abundance, and were probably shaped by a mixture of top‐down and bottom‐up effects. These results may apply to many other vertebrate mesopredators, whose individual decisions and population responses are probably the result of a trade‐off between predator avoidance and food acquisition. At the community level, the density of different diurnal avian predators responded differently to IGP risk, some of them even increasing with this risk. The diversity of the assemblage peaked at medium–high eagle owl density. Future studies of top vertebrate predators need to take more account of IGP effects. Their conservation management needs to be highly case‐specific and framed within a broader ecosystem perspective.