Application of ensemble modelling method in predicting the effects of climate change on the distribution of Fritillaria imperialis L.

One of the major impacts of climate change on plants is the change in their geographical range. Predicting the potential distribution of endangered plants in response to climate change is essential for their conservation and policy management. This study aimed at predicting the effect of climate change on the geographical distribution of Fritillaria imperialis in Chaharmahal & Bakhtiari province in the central Zagros region. We used RCP4.5 and RCP8.5 (two climate change scenario) and general circulation model HadGEM2-CC in order to predict the future geographic distribution of the species. In this study, we used 19 bioclimatic variables related to rainfall and temperature and three physiographical variables to create the models. In this regard, we used 5 modelling approaches, Generalized Linear Model (GLM), Classification Tree Analysis (CTA), Artificial Neural Network (ANN), Generalized Boosting Method (GBM) and Random Forest (RF) under the ensemble using biomod2 package in the free statistical software environment R 3.3.2. The results of the model showed that annual precipitation and temperature seasonality had the most important to provide habitat suitability of this species. All the models were accurate enough to predict the species distribution, among them, RF presented the most reliable model for the prediction. Under RCP4.5 and RCP8.5 climate scenario, the habitat of the species will shrink to 19.7% and 61, respectively in 2050.