Ensemble Species Distribution Modeling

An ensemble of projections of SDM was obtained for the 11,495 selected species. The ensemble included projections with Generalized Additive Models, Boosting Regression Trees, Generalized Linear Models and Random Forests. Models were calibrated for the baseline period using 70% of observations randomly sampled from the initial data and evaluated against the remaining 30% data using the true skill statistic (TSS^{29 (link)}). Presence data were randomly drawn from the gridded range maps. For absences, we considered data in a reasonable buffer around the presence data to avoid having over-optimistic predictive accuracies^{30 (link)}. To be consistent with assumed realistic dispersal distances (see next paragraph) and in line with previous analyses, we selected absence data in 2000, 3000, and 4000 km buffer around amphibian, mammal and bird species ranges, respectively. This analysis was repeated four times, thus providing a four-fold internal cross-validation of the models (biomod package^{15 (link)}). The quality of the models was ‘very high’ to ‘excellent’ with an average TSS of 0.83 (Supplementary Fig. 2). Since the quality of the models strongly affect projection uncertainties, we tested three thresholds below which models were removed from projection ensembles. We used a threshold of TSS = 0.4, which is usually considered a minimum for retaining reliable models^{29 (link)}. We also used thresholds of 0.6 and 0.7, and investigated their effects on uncertainties of species-based sensitivity metrics. For all further analyses, the most drastic threshold (TSS = 0.7) was then used (red lines in Supplementary Fig. 2).
For each species, all calibrated models (4 SDMs × 4 repetitions) were then used to project the potential distribution of each species under both current and projected future climatic conditions (Supplementary Fig. 1).