Abstract
Biodiversity conservation planning requires accurate, current information about species status and threats. Although introduced mammals are the greatest threat to seabirds globally, data on introduced species is lacking for many seabird breeding islands. To inform conservation planning, we used trail cameras to document the presence, relative abundance, and seasonal and diel attendance of introduced and native vertebrates within pink-footed shearwater (Ardenna creatopus) breeding colonies on Isla Mocha (five colonies, 2015-2020) and Isla Robinson Crusoe (Juan Fernández Archipelago), Chile (one colony, 2019-2020). The most commonly detected species were pink-footed shearwaters and introduced rats (Rattus spp.) on Isla Mocha, and European rabbits (Oryctolagus cuniculus) and pink-footed shearwaters on Isla Robinson Crusoe. Introduced mammals observed, in order of greatest catch-per-unit-effort, were rats, cats (Felis catus), dogs (Canis lupus familiaris), and European hares (Lepus europaeus) on Isla Mocha and European rabbits, cats, cattle (Bos taurus), rats, dogs, mice (Mus musculus), and southern coati (Nasua nasua) on Isla Robinson Crusoe. Especially noteworthy results for pink-footed shearwater conservation were the presence of cats during all monitoring months in shearwater colonies on both islands, that catch-per-unit-effort of rabbits was greater than shearwaters on Isla Robinson Crusoe, and that rats were the most observed vertebrates after shearwaters on Isla Mocha. Pink-footed shearwaters were regularly present on the islands from October through May. Presence and relative catch-per-unit-effort of pink-footed shearwaters qualitatively matched the species' known breeding phenology. The regular presence and temporal overlap with shearwaters of cats, rats, rabbits, and cattle within shearwater colonies, coupled with the irregular presence of dogs, coati, hares, and mice, indicated a serious conservation threat for pink-footed shearwaters and other native insular fauna and flora. Finally, our study provides a widely applicable model for analysis of multi-year trail camera data collected with unstandardized settings.