Beyond climate envelopes: bio-climate modelling accords with observed 25-year changes in seabird populations of the British Isles

Aim: Climate envelope models (CEMs) are used to assess species' vulnerability to predicted changes in climate, based on their distributions. Extinction risk, however, also depends on demographic parameters. Accordingly, we use CEMs for 18 seabird species to test three hypotheses: (i) population sizes are larger in areas where CEMs fitted using distribution data predict more suitable climate; (ii) the presence of this relationship (Hypothesis i) is related to a species' foraging ecology; and (iii) species whose distributions and population sizes conformed most closely to indices of climatic suitability in the mid-1980s experienced the largest population changes following climatic change between 1986 and 2010. Location: Europe. Methods: Climate envelope models fitted at a 50-km resolution using European climatic and distribution data were applied using local climatic data to calculate local climatic suitability indices (CSIs) for 18 species within the British Isles. We then investigated the relationship between CSI and population size at a 10-km resolution and related both the presence of this relationship and goodness-of-fit metrics from the European models to changes in population size (1986-2010). Results: Local population sizes were significantly positively related to local CSI in 50% of species, providing support for Hypothesis (i), and these 50% of species were independently considered to be most vulnerable to changes in food availability at sea in support of Hypothesis (ii). Those species whose distributions and populations most closely conformed to indices of climatic suitability showed the least favourable subsequent changes in population size, over a period in which mean climatic suitability decreased for all species, in support of Hypothesis (iii). Main conclusions: Climate influences the population sizes of multiple seabird species in the British Isles. We highlight the potential for outputs of CEMs fitted with coarse resolution occupancy data to provide information on both local abundance and sensitivity to future climate changes.