Die Kastanie im Engadin – oder was halten Baumarten von modellierten Potenzialgebieten? Chestnut in the Engadine – what do tree species think about model predictions?

New IPCC climate projections suggest drastic changes in future climate. We discuss two commonly used modeling approaches, statistical distribution models and dynamic forest succession models, as they are suitable for assessing expected effects of climate change on the tree species distribution in Switzerland and for assisting management decisions in forestry. We discuss the basic assumptions and the strengths and weaknesses of the two approaches, without an understanding of which it is impossible to fully judge the outcome of modeling exercises. We give an overview of results from applying these two modeling approaches in Switzerland and in the Alps and discuss their appropriate use. We believe that these models are an important basis for decision making in the face of highly uncertain development of future climate. Nonetheless, models do not represent an exact copy of reality. Plausibility analyses are necessary in order to assess the results' usefulness and precision. Sensitivity analyses and a critical comparison of model results with expert knowledge of current forests, long measurement time series and other data are important. Also, dialog with practitioners and managers is not only important for checking the plausibility of model predictions under current conditions, but may also serve to improve the evaluation of future projections. We propose to apply models to the whole of Switzerland and to many tree species. Such a concerted national analysis may serve the adaptive management of forests and may strengthen dialog between researchers and practitioners.