Multi-Model Analysis to Predict the Potential Suitable Habitat for Quercus oblongata D. Don (Fagaceae) in the Western Himalayan Region

Abstract

Quercus oblongata D. Don is an economically and ecologically important tree species. It is distributed in the western Himalayan region of India as well as Nepal, Pakistan, Thailand, and Vietnam. The present study explored the potential distribution and habitat suitability of Q. oblongata using ensemble modeling. Eleven environmental variables (eight bioclimatic and three topographic), and 121 occurrence records were used in the analysis. Four algorithms: Generalized Additive Model (GAM), Generalized Linear Model (GLM), Random Forest (RF), and Maximum Entropy (Max Ent) were used to build the ensemble model for potential suitable habitat of Q. oblongata. To evaluate the model performance, AUC and TSS metrics were used, which showed high AUC (>0.94) and TSS (>0.86) metrics for all the models. The environmental variables that displayed high contribution in the prediction were temperature annual range (bio7, 48.5%), precipitation of wettest period (bio13 ; 41%), and elevation (elev ; 40%). The total suitable area was 22634 km2, including the least (12546 km2), moderate (7935 km2), and highly (2153 km2) suitable areas. Habitat suitability of Q. oblongata is predicted in most of the regions of Nainital, Almora, central Tehri Garhwal, eastern Mussoorie and Chakrata region of Dehradun District of Uttarakhand. A sizeable wide patch was found in Southern Chamba with Northern Kangra District, including the northern region of Mandi District of Himachal Pradesh. The predicted suitable habitat can be used for future exploration for the study of genetic diversity and conservation purposes.