Optimization of Ensemble Modeling Approach for Studying Climatic Niche and Conservation Status Assessment for Endemic Taxa

Abstract

Modeling ecological niche of rare endemic species is an understated issue especially for data-deficient endangered taxa. Syzygium travancoricum Gamble. a critically endangered plant in southern Western Ghats of India was considered to examine the role of pseudo-absence and modeling algorithms in endemic species niche formulation. This aided in reassessing IUCN status emphasizing on geographical distribution.
Seven different modeling algorithms were attempted considering different sets of pseudo-absence points through BIOMOD2 program in R and MaxEnt. The outputs were compared with the extent of occurrence (EOO) and area of occupancy (AOO), as done in conservation status assessment exercise by IUCN.
Modeling algorithm and pseudo-absence points play an important role in model development. Ensemble methods displayed wider variations in outputs despite having good statistical performance thus emphasizing on importance of multiple algorithm trials for data deficient taxa. Both model and topology based methods have shown much higher values for EOO and AOO which contradict it’s critically endangered status; thus, suggesting re-evaluation of the threat status with comprehensive field investigation.