Probability distributions, vulnerability and sensitivity in Fagus crenata forests following predicted climate changes in Japan

Question: How much is the probability distribution of Fagus crenata forests predicted to change under a climate change scenario by the 2090s, and what are the potential impacts on these forests? What are the main factors inducing such changes? Location: The major islands of Japan. Methods: A predictive distribution model was developed with four climatic factors (summer precipitation, PRS; winter precipitation, PRW; minimum temperature of the coldest month, TMC; and warmth index, WI) and five non-climatic factors (topography, surface geology, soil, slope aspect and inclination). A climate change scenario was applied to the model. Results: Areas with high probability (>0.5) were predicted to decrease by 91%, retreating from the southwest, shrinking in central regions, and expanding northeastwards beyond their current northern limits. A vulnerability index (the reciprocal of the predicted probability) suggests that Kyushu, Shikoku, the Pacific Ocean side of Honshu and southwest Hokkaido will have high numbers of many vulnerable F. crenata forests. The forests with high negative sensitivity indices (the difference between simulated probabilities of occurrence under current and predicted climates) mainly occur in southwest Hokkaido and the Sea of Japan side of northern Honshu. Conclusion: F. crenata forest distributions may retreat from some islands due to a high WI. The predicted northeastward shift in northern Hokkaido is associated with increased TMC and PRS. High vulnerability and negative sensitivity of the forests in southern Hokkaido are due to increased WI.