Modeling the potential habitat of English yew (Taxus baccata L.) in the Hyrcanian forests of Iran

In this study, using four commonly used modeling methods for assessing the potential habitat of yew species (generalized linear and additive models, classification and regression tree and random forest) and maps of the bioclimatic and topographic, the suitability map of the yew species in Hyrcanian forests was prepared. The evaluation of models by using AUC criterion (GLM=0.863, GAM=0.871, CTA=0.781, and RF=0.922) showed that the random forest model has the best performance in which slope percentage and solar radiation index are the most and least important variables, respectively. The present study showed that the presence probability of yew species in relation to slope percentage in the study area increases from 30% and then levels off. In the current conditions, the area of suitable habitats for yew species is about 10.5% of the studied area. The results of this research indicated that in the current climatic conditions, the suitable areas for presence of yew species are in east and center of the study area that will help to manage, protect and restore the habitats of this invaluable species.