The performances of the ecological niche models were surprisingly good (AUC ranged from 0.85-0.97). The best predictive models were found for Polyalthia hypoleuca (AUC = 0.97). The levels of accuracy for plants derived from the testing data varied relatively behind the training data (ranging from 0.81 to 0.92). The disagreement may have occurred because there were fewer points for plant species. Nevertheless, the ecological niche models were considered to be excellent in discriminating between predicted presence and predicted absence (Hosmer & Lewshow, 2000).