Assessing the effectiveness of foraging radius models for seabird distributions using biotelemetry and survey data

The design and funding for aerial surveys were provided by Ireland's Dept of Communication, Climate Action and Environment and the Dept of Culture, Heritage and the Gaeltacht, under the ObSERVE Programme established in 2014. We are grateful for the support and assistance of both Depts in undertaking this work. Funding for development of projected distributions was provided by the Petroleum Infrastructure Program (IS013/08), and funding for telemetry work was provided by the Zoological Society of London (Good gifts programme), Petroleum Infrastructure Programme (IS013/08) and the Irish Research Council (GOIPD/2015/81) Ireland's Dept of Agriculture, Fisheries and the Marine. Relatively simple foraging radius models have the potential to generate predictive distributions for a large number of species rapidly, thus providing a cost‐effective alternative to large‐scale surveys or complex modelling approaches. Their effectiveness, however, remains largely untested. Here we compare foraging radius distribution models for all breeding seabirds in Ireland, to distributions of empirical data collected from tracking studies and aerial surveys. At the local/colony level, we compared foraging radius distributions to GPS tracking data from seabirds with short (Atlantic puffin Fratercula arctica, and razorbill Alca torda) and long (Manx shearwater Puffinus puffinus, and European storm‐petrel Hydrobates pelagicus) foraging ranges. At the regional/national level, we compared foraging radius distributions to extensive aerial surveys conducted over a two‐year period. Foraging radius distributions were significantly positively correlated with tracking data for all species except Manx shearwater. Correlations between foraging radius distributions and aerial survey data were also significant, but generally weaker than those for tracking data. Correlations between foraging radius distributions and aerial survey data were benchmarked against generalised additive models (GAMs) of the aerial survey data that included a range of environmental covariates. While GAM distributions had slightly higher correlations with aerial survey data, the results highlight that the foraging radius approach can be a useful and pragmatic approach for assessing breeding distributions for many seabird species. The approach is likely to have acceptable utility in complex, temporally variable ecosystems and when logistic and financial resources are limited. Publisher PDF