Predator protection or similar habitat selection in red-breasted goose nesting associations: Extremes along a continuum

We tested the predator protection and similar habitat hypotheses in relation to red-breasted goose, Branta ruficollis, nesting associations. Geese began laying 1-3 weeks after all associated species. In almost all cases they nested on the mainland only if raptors were also present and always followed raptors when they changed eyries between years. They selected peregrines, Falco peregrinus, and snowy owls, Nyctea scandiaca, as associates in preference to rough-legged buzzards, Buteo lagopus, even though the latter were several times more abundant along river corridors. Nest defence experiments with a surrogate Arctic fox, Alopex lagopus, showed that this could be explained by differences in nest defence intensity, rather than habitat types selected. Similar experiments also suggested that gulls were much less aggressive than owls or falcons. Foxes rarely approached, and were easily repelled from goose colonies associated with owls and peregrines, but gulls were apparently incapable of repelling foxes. Breeding success was much higher for geese nesting with raptors than for those on islands and geese apparently preferred to nest with owls in peak lemming years than to remain on islands. These findings support the hypothesis that red-breasted geese actively choose to nest near raptors; however, they probably associate with gulls primarily because both species select fox-free islands. Compared with other studies, red-breasted geese nesting with raptors and on islands apparently represent two extremes in a continuum of nesting associations generally seen in birds. We discuss why the behaviour might have evolved and argue that this may be the only known bird species whose evolution has been facilitated primarily by the exploitation of the nest defence behaviour of aggressive raptorial hosts.