Ensemble modelling enables prediction of suitable habitat of critically endangered White bellied Heron in Bhutan (Ardea insignis) under climate change

Identifying and mapping potential suitable habitats for effective conservation is important for conservation of globally threatened flora and fauna in an era of climate change. It is the topic of emerging interest in the current times and in this study we integrate the insights gained from ensemble species distribution modelling to identify and map the potential suitable habitats of critically endangered A.insignis for conservation purposes. We used ensemble species distribution modelling to model and map the current potential suitable habitats and also predicted the future potential suitable habitats for the species under future climate change scenarios. The predictive accuracy of the final ensemble species distribution model was pretty good ( i . e. AUC value of 0.908 and a TSS value of 0.702 for committee averaging method and AUC value of 0.921 and a TSS value of 0.701 for ensemble my mean method ) and the Precipitation of the Coldest Quarter was the most influential bioclimatic variables governing the distribution of A.insignis with Variable importance score mean of 0.456, followed by Precipitation Seasonality with importance score mean of 0.267, and Annual Mean Temperature with importance scores mean of 0.220.The central-western and central parts of Bhutan along the Punatsangchhu and Mangdechhu were predicted as a very potential suitable habitat for A.insignis by the final ensemble model. However, a shift in potential suitable habitat was observed from central-western and central parts of Bhutan to south eastern parts of Bhutan along the Manas river basins as predicted under two future climate change scenarios.