Modelling potential biotope composition on a regional scale revealed that climate variables are stronger drivers than soil variables.

Aim: Environmental conditions define the suitability of an area for biotopes, and any area can be suitable for several biotopes. However, most previous studies modelled the distribution of single biotopes ignoring the potential co‐occurrence of biotopes in one area, which limits the usefulness of such models for conservation and restoration planning. In this study, we described the potential biotope composition of an area in response to environmental conditions. Location: Bavaria, Federal State of Germany. Methods: Based on the Bavarian biotope mapping data, we modelled the distribution of 29 terrestrial biotopes based on six climate variables and six chemical and four physical soil properties using the species distribution modelling algorithm Maxent. Results: For most biotopes, we found that climate variables were more important than soil variables for the biotope distribution and that the area of the predicted biotope distribution was larger than the observed distribution. The potential biotope composition illustrated that while 8% of the area in Bavaria was not sufficiently suitable for any analysed biotope, 92% of the modelled area in Bavaria was suitable for at least one biotope, 84% for two and 77% for at least three biotopes. The difference in suitability between the most suitable biotopes in composition was minor. Further, over one‐quarter of the modelled area was suitable for 6–8 different biotopes. Main Conclusions: Our study showed that considering a composition of potentially suitable biotopes in a raster cell, instead of only the most suitable biotope, provides valuable information to identify conservation priorities and restoration opportunities. [ABSTRACT FROM AUTHOR]