Nest-site selection and breeding success of passerines in the world's southernmost forests

Background. Birds can maximize their reproductive success through careful selection of nest-sites. The 'total-foliage' hypothesis predicts that nests concealed in vegetation should have higher survival. We propose an additional hypothesis, the 'predator proximity' hypothesis, which states that nests placed farther from predators would have higher survival. We examined these hypotheses in the world's southernmost forests of Navarino Island, in the Cape Horn Biosphere reserve, Chile (55 degrees S). This island has been free of mammalian ground predators until recently, and forest passerines have been subject to depredation only by diurnal and nocturnal raptors. Methods. During three breeding seasons (2014-2017), we monitored 104 nests for the five most abundant open-cup forest-dwelling passerines (Elaenia albiceps, Zonotrichia capensis, Phrygilus patagonicus, Turdus falcklandii, and Anairetes parulus). We identified nest predators using camera traps and assessed whether habitat characteristics affected nest-site selection and survival. Results. Nest predation was the main cause of nest failure (71% of failed nests). Milvago chimango was the most common predator, depredating 13 (87%) of the 15 nests where we could identify a predator. By contrast, the recently introduced mammal Neovison vison, the only ground predator, depredated one nest (7%). Species selected nest-sites with more understory cover and taller understory, which according to the total-foliage hypothesis would provide more concealment against both avian and mammal predators. However, these variables negatively influenced nest survival. The apparent disconnect between selecting nest-sites to avoid predation and the actual risk of predation could be due to recent changes in the predator assemblage driven by an increased abundance of native M. chimango associated with urban development, and/or the introduction of exotic mammalian ground predators to this island. These predator assemblage changes could have resulted in an ecological trap. Further research will be needed to assess hypotheses that could explain this mismatch between nest-site selection and nest survival.