“…An ensemble of species distribution models (Araújo & Guisan, 2006 ) was used to model potential suitable habitat for D. involucrata using the biomod2 package in the R platform (v. 4.0.4; http://cran.r‐project.org ). We chose the ensemble modeling approach because of its ability to create a consensus of the predictions of multiple algorithms and reduce the predictive uncertainty of single algorithm (Kanagaraj et al., 2019 ; Thuiller et al., 2014 ; Zhang, Dong, et al., 2020 ; Zhang, Mammola, et al., 2020 ). Ten algorithms were considered in the ensemble model: artificial neural network (ANN; Ripley, 1996 ), classification tree analysis (CTA; Breiman et al., 1984 ), flexible discriminant analysis (FDA; Hastie et al., 1994 ), generalized additive model (GAM; Hastie & Tibshirani, 1990 ), generalized boosting model (GBM; Ridgeway, 1999 ), generalized linear model (GLM; McCullagh & Nelder, 1989 ), multiple adaptive regression splines (MARS; Friedman, 1991 ), maximum entropy (MAXENT; Phillips et al., 2006 ), random forest (RF; Breiman, 2001 ), and surface range envelope (SRE; Busby, 1991 ).…”