Relationship between Sect. Sasa distributions and climate was modelled by using a classification tree model and clarified at ca. 1-km2 spatial resolutions in Japan. Occurrence probability was then predicted under both current climate and the future climate change scenario RCM20 (2081-2100) to assess the impact of climate change. Five climatic factors (warmth index: WI, minimum temperature of the coldest month: TMC, summer precipitation: PRS, maximum snow water equivalent: MSW, winter rainfall: WR) were used as predictor variables, and the species distribution data obtained from Phytosociological Relevé Data Base was used as a response variable. Deviance-Weighted Scores revealed that the most influential factor for the species distribution was MSW, followed by WI, PRS, WR and TMC. Predicted potential habitat was divided into “suitable habitat” and “marginal habitat”, based on the Receiver Operating Characteristic analysis. Climatic conditions for suitable habitat were WI ≥ 31.6°C·month and MSW ≥ 131.6mm. The area of suitable habitats was predicted to decrease 46.7% due to decrease in MSW, and 69.9% of marginal habitats were predicted to turn into nonhabitat due to increase in WI under the global warming.
