“…For the remaining 30%, we used an evaluation dataset (a combination of the systematic datasets MHB, BDM, and MHBBDM) with high precision geography identification (Table 1) from the similar study area to evaluate the models (Edwards Jr et al, 2006;Graham et al, 2008;Hallman and Robinson, 2020). For model evaluations, we used several indices such as the area under the receiver operating characteristic curve (AUC; Jiménez-Valverde, 2012;Fernandes et al, 2019), the true skills statistic (TSS; Allouche et al, 2006;Fernandes et al, 2019), and Cohen's Kappa Statistic (KAPPA; Cohen, 1960;Fernandes et al, 2019) (Li et al, 2020;Smeraldo et al, 2021).…”