Differential response to abiotic stress controls species distributions at biogeographic transition zones

3 March 2017

Sommer, Brigitte; Beger, Maria; Harrison, Peter; Babcock, Russ; Pandolfi, John M.

Understanding range limits is critical to predicting species responses to climate change. Subtropical environments, where many species overlap at their range margins, are cooler, more light-limited and variable than tropical environments. It is thus likely that species respond variably to these multi-stressor regimes and that factors other than mean climatic conditions drive biodiversity patterns. Here, we tested these hypotheses for scleractinian corals at their high-latitude range limits in eastern Australia and investigated the role of mean climatic conditions and of parameters linked to abiotic stress in explaining the distribution and abundance of different groups of species. We found that environmental drivers varied among taxa and were predominantly linked to abiotic stress. The distribution and abundance of tropical species and gradients in species richness (alpha diversity) and turnover (beta diversity) were best explained by light limitation, whereas minimum temperatures and temperature fluctuations best explained gradients in subtropical species, species nestedness and functional diversity. Variation in community structure (considering species composition and abundance) was most closely linked to the combined thermal and light regime. Our study demonstrates the role of abiotic stress in controlling the distribution of species towards their high-latitude range limits and suggests that, at biogeographic transition zones, robust predictions of the impacts of climate change require approaches that account for various aspects of physiological stress and for species abundances and characteristics. These findings support the hypothesis that abiotic stress controls high-latitude range limits and caution that projections solely based on mean temperature could underestimate species’ vulnerabilities to climate change.