Table 2. Summary of performance statistics of C. fimbriata MaxEnt models.

Model No.	Variables	MaxEnt settings		Test AUCcv (±SD)	Test OR		Model Ranka
		Features	RM		0%	10%
1	bio1, bio2, bio13, bio14, bio15, bio19, Elevation	LQH	1.5	0.970 ± 0.012	0.017	0.110	1
2	Same as above	LQH	2.0	0.969 ± 0.011	0.037	0.133	6
3	Same as above	LQPTH	1.5	0.971 ± 0.012	0.093	0.150	10
4	Same as above	LQPTH	2.0	0.965 ± 0.014	0.090	0.200	11
5	Same as above	LQP	1.0	0.951 ± 0.017	0.057	0.130	4
6	Same as above	LQP	1.5	0.946 ± 0.017	0.033	0.143	8
7	Same as above	LQP	2.0	0.939 ± 0.021	0.020	0.110	2
8	Same as above	LH	1.5	0.971 ± 0.012	0.020	0.133	5
9	Same as above	LH	2.0	0.967 ± 0.013	0.033	0.150	9
10	Same as above	LQPT	1.5	0.956 ± 0.021	0.057	0.223	13
11	Same as above	LQPT	2.0	0.948 ± 0.020	0.040	0.220	12
12	Same as above	LQPH	1.5	0.974 ± 0.010	0.070	0.137	7
13	Same as above	LQPH	2.0	0.967 ± 0.012	0.053	0.110	3

See Table 1 for variables’ full names. L, Q, P, T and H are linear, quadratic, product, threshold and hinge features, respectively. RM is regularization multiplier, and SD is standard deviation. OR is test omission rate. Test AUC_cv is MaxEnt 10-fold cross-validation Area Under the ROC curve.

^aThe model with the highest performance is highlighted in bold.