Evaluating climatic threats to habitat types based on co-occurrence patterns of characteristic species

Species distribution models (SDMs) are used to project how suitable ranges of species shift under a warming climate. Conservation management, however, commonly targets habitat types rather than individual species. Such habitat types are often defined by the co-occurrence of a set of characteristic species. Here, we develop a co-occurrence-based index (CRI); which measures how the representation of habitat types in a particular area may change in a future climate. The index is based on stacking projections of distribution models of characteristic species and accounts for changes both in potential range size of each species individually and in spatial range overlap among characteristic species, i.e. co-occurrence patterns. We illustrate the approach by modelling the changing representation of 68 habitat types in Austria under two different climate scenarios. We base index calculations on SDM projections under either the assumption of unrestricted mobility (‘full-dispersal’) or of complete immobility (‘no-dispersal’) of individual species. Moreover, we compare results to those achieved with a simpler occurrence-based index (OI); which only accounts for change in species’ range sizes. All three alternative index calculations suggest that most modelled habitat types will lose area (in particular mires, wetlands and siliceous alpine grasslands) and only a minority will profit from a warming climate (in particular forests of dry and warm sites). ‘Full-dispersal’ CRI and OI are closely, but not perfectly correlated. Importantly, for more than half of the habitat types, accounting for changing co-occurrence patterns amplifies projected losses. The ‘no-dispersal’ CRI (CRInd) delivers the most alarming projections, indicating considerable spatial turn-over of sites suitable to the habitat types. Taken together, our results suggest that modelling re-distribution of habitat types which are defined by species combinations needs to account for modifications of co-occurrence patterns. Moreover, conservation should acknowledge that novel combinations of species will likely emerge under a warming climate.