Projected climate impacts on the natural pasturelands of the Italian alps

This study simulates the impacts of climate changes on distribution and composition of the main pasture macro-types located over the Italian Alps using spatially informative layers as predictor variables. More specifically, a machine learning approach (Random Forest), firstly calibrated for the present period and then applied to future conditions as simulated by HadCM3 General Circulation Model, was used to forecast potential expansion/reduction and/or altitudinal shifts of the alpine pasture macro-types. Future climate conditions, as depicted by HadCM3 in A2 and B2 SRES scenarios, will have impacts of great concern on pasture composition and location. Except for pastures dominated by xeric species and by Nardus stricta, which showed wide expansions over the study area, a troublesome reduction was predicted for the rarest pastures (i.e. dominated by Festuca rubra) or for the macro-types currently restricted at the highest altitudes (i.e. dominated by Carex curvula, by Carex firma and by Sesleria varia).