Ecological and biological indicators of the accuracy of species distribution models: lessons from European bryophytes

The predictive power of species distribution models (SDMs) varies substantially among depending on their ecological and life-history traits, but which these traits are the most relevant how they influence ‘predictability' remains an area debate. Here, we address questions in bryophytes. SDMs employing macroclimatic, topographic edaphic predictors were calibrated for 411 Europe externally evaluated using independent dataset. Regression implemented to determine whether characteristics, including preference niche breadth, accuracy SDMs. Variation SDM was significantly explained by supporting hypothesis that strength species–environment correlations is affected characteristics themselves. percent variance however, varied between 9 57% evaluation metrics used. lower correlation observed MaxKappa Boyce index than with under curve (AUC) MaxTSS suggests former less suitable latter determining based traits. decreased from restricted pristine habitats thriving eutrophic high levels human disturbance. widespread man-made fact opens door spread now ubiquitous species, even environments would primarily not be them. Such likely occur anywhere, reach very full occupancy rates, thereby decreasing aiming at predicting distributions. AUC higher important a conservation context, as eutrophic, disturbed precisely ones relevance.