Regional breeding bird assessment of the Antarctic Peninsula

Conservation of seabirds in remote polar regions requires accurate information on the location of breeding sites, which is often limited by logistical constraints of surveying large areas. On the Antarctic Peninsula, many seabird colonies are visited briefly but regularly by commercial cruise vessels, a platform from which we were able to collect presence/absence data on the entire community of seabirds. We used a multistate occupancy model, accounting for limited detection during surveys, to estimate the probability of presence and breeding of all 16 species native to the area. Our results provide a much clearer view than that was previously available of how avian diversity is distributed across the region's network of multi-species colonies and reveals species-specific differences in the effect of sea-ice concentration and site fidelity on breeding probability. Several breeding sites host an unusually large number of breeding species, but these known richness hotspots are scattered throughout the region and we were unable to identify any clear gradients in species richness that might explain why some sites are so species rich. While accounting for detection failure accelerates the pace of reliable inference on species occupancy, we find that as many as ten years of repeated visits are often required to fully catalog the seabird richness at bare rock sites along the Antarctic Peninsula. This work highlights the challenges of identifying high priority sites for special protection or management and the importance for continued surveys, even at nominally well-studied locations.