Quantifying the breeding distribution and habitat use of the snow petrel (Pagodroma nivea), the world’s most southerly breeding vertebrate

Abstract

Seabirds in the Southern Ocean serve as important ecological indicators of ecosystem status, responding to environmental conditions at both local and regional scales. However, knowledge of the spatial distribution of many polar seabird species is incomplete due to logistical difficulties of accessing remote breeding locations. A prime example is the snow petrel Pagodroma nivea, the most southerly breeding vertebrate in the world, whose breeding distribution has not been assessed in almost three decades. This thesis aims to quantify this species’ breeding distribution, characterise breeding habitat, and test whether remote sensing can detect known breeding sites. To do so, records of breeding locations, including population estimates when available, were collected from previously published work. Local scale environmental conditions at breeding sites (lithology, temperature, precipitation and wind speed), distance to the coast and regional sea-ice conditions accessible within defined foraging ranges were characterised. Two large breeding sites were subsequently selected for remote sensing, with image enhancement and unsupervised classification performed. The results provide the first updated version of the circumpolar breeding distribution, in which 456 breeding sites are now known, 158 more than the previous inventory. Most known breeding sites are biased towards the location of research stations, indicating more remote breeding sites remain undiscovered. As a cavity-nesting species, the distribution is partly controlled by cavity availability, and results suggest preferential use of cavities in intrusive igneous and high-grade metamorphic lithologies, with the majority of the known breeding population located on the latter. Breeding snow petrels face a central-place foraging constraint, needing to repeatedly return to their nests, and it has been hypothesised therefore that the breeding distribution is limited by distance to pack-ice, where they forage. Characterising regional sea-ice conditions in areas accessible from breeding sites (foraging habitat) supports this, with a median distance from breeding sites to the November ice edge of 430 km. However, the most remote sites are > 1000 km from this foraging habitat. The lack of accessible foraging habitat, due to the year-round persistence of high concentration sea ice in the Weddell Sea, likely explains the absence of breeding sites on the eastern Antarctic Peninsula. However, other gaps in the breeding distribution remain unexplained. The results of remote sensing indicate that if we are to detect breeding sites remotely, better spectral and spatial resolution imagery will be needed, as well as ground truthing data recorded at breeding sites. As ~70% of known breeding sites were recorded before 2000, more consistent and detailed data on breeding sites and breeding populations are also needed to better understand the distribution. Similarly, more widespread long-term studies of snow petrel populations are needed in order to predict the response of this species to climate change.