Rick Stuart-Smith (1), Graham Edgar (1), Amanda Bates (2)
1 Institute for Marine and Antarctic Studies, University of Tasmania, private bag 49, Hobart 7001. rstuarts@utas.edu.au
2 Ocean and Earth Science, National Oceanography Centre Southampton, University of Southampton, Southampton, UK
Using records in the Reef Life Survey (RLS) global dataset and in the Global Biodiversity Information Facility (GBIF) we found that the present-day distributions of species on rocky and coral reefs all around the world tend to be centred in tropical and temperate regions, and that there are relatively few fish and invertebrate species in this environment which can be considered truly subtropical. These distribution patterns result in a phenomenon called ‘thermal bias’, where ecological communities are dominated by species which have, on average, a warmer or cooler thermal affinity than local mean temperatures. Thermal bias can be calculated using the Community Temperature Index and can be used to predict whether the average individual in a community may theoretically respond positively or negatively to warmer conditions (in the absence of changing ecological interactions). We also extended the concept by assessing another metric of species’ thermal distributions – the buffer between the local temperature and the maximum temperature each species has ever been recorded at in the wild. From a global analysis of patterns in this thermal buffer, we found that present-day geographic distributions of marine species are more important than global variation in warming rates in shaping the probability of extirpation of species from local communities (i.e. how range contractions will change communities) over the next century. In some locations it will take much more than a century before the thermal buffer is eroded for many of the species that comprise today’s reef communities, even under rapid warming.