Table 3. Validation statistics and jack-knife analysis of variable contributions to the models for all taxa (50^th percentile), Alcyoniina, Antipatharia and Calcaxonia.

	All Taxa				Alcyoniina				Antipatharia				Calcaxonia
Cross-validation cell	1	2	3	4	1	2	3	4	1*	2	3	4	1	2	3	4
Validation statistics
Test AUC	0.82	0.846	0.815	0.943	0.871	0.861	0.881	0.95	0.577	0.835	0.929	0.912	0.787	0.923	0.922	0.896
Test AUC standard deviation	0.016	0.012	0.016	0.005	0.026	0.015	0.013	0.004	0.204	0.033	0.017	0.037	0.036	0.016	0.01	0.011
10th percentile training presence	0.4866	0.5028	0.4545	0.4222	0.515	0.577	0.4962	0.4586	0.434	0.4829	0.2917	0.4321	0.6061	0.5946	0.4228	0.4631
Maximum test sensitivity plus specificity	0.181	0.262	0.412	0.379	0.153	0.162	0.098	0.35	0.57	0.054	0.375	0.26	0.046	0.195	0.117	0.323
Jack-knife of variable importance (jack of regularized training gain)
Depth	0.7343	0.8258	1.0306	0.6101	0.9197	1.0217	1.2143	0.7726	1.0188	1.1635	0.528	1.0441	1.1326	1.1511	1.3354	1.1089
Dissolved Oxygen	0.5784	0.5817	0.6399	0.5601	0.8359	0.8946	0.9726	0.9285	1.2431	1.3786	0.4208	1.4051	1.0852	1.0692	1.1143	1.3166
Calcite Saturation State	0.6823	0.7536	0.9572	0.5682	0.9203	1.0206	1.1697	0.8571	1.2913	1.4405	0.5438	1.3341	1.1359	1.1417	1.2385	1.2377
Particulate Organic Carbon	0.6244	0.7207	0.9181	0.5533	0.7423	0.8689	1.0206	0.6541	0.798	0.7914	0.2422	1.0263	0.5591	0.55	0.9316	0.7141
Salinity	0.815	0.8898	1.0605	0.6733	0.9825	1.1053	1.2755	0.8786	1.2911	1.4415	0.5035	1.3727	1.2273	1.2426	1.3415	1.296
Slope (1 km)	0.9349	0.9133	0.8904	0.5992	1.2309	1.3442	1.3521	1.0203	1.4655	1.5807	1.3852	1.3643	1.5481	1.4591	1.4562	1.3894
Temperature	0.749	0.8451	1.0054	0.6698	0.9414	1.0662	1.1859	0.8904	1.4282	1.627	0.7199	1.4986	1.2905	1.2951	1.3924	1.3747
Test AUC for a single variable
Depth	0.9045	0.7761	0.7048	0.8734	0.9403	0.8161	0.784	0.8639	0.6869	0.6496	0.8357	0.7562	0.8125	0.8368	0.8067	0.8675
Dissolved Oxygen	0.7588	0.8111	0.7253	0.7746	0.8504	0.8013	0.7904	0.8379	0.5127	0.7592	0.7807	0.6544	0.7517	0.853	0.7973	0.7244
Calcite Saturation State	0.8791	0.7738	0.7312	0.8449	0.9165	0.8303	0.8193	0.845	0.5853	0.6969	0.8488	0.8689	0.8115	0.8512	0.8235	0.8435
Particulate Organic Carbon	0.8558	0.7159	0.6523	0.7759	0.7956	0.7	0.673	0.6805	0.611	0.771	0.6574	0.5644	0.693	0.6625	0.6173	0.6079
Salinity	0.8185	0.8036	0.7082	0.8618	0.9113	0.7971	0.7875	0.8921	0.6308	0.8093	0.8468	0.5811	0.7874	0.8524	0.8133	0.8549
Slope (1 km)	0.6041	0.7819	0.8298	0.9205	0.7739	0.8421	0.8948	0.9515	0.7553	0.7945	0.9116	0.9571	0.727	0.9184	0.9324	0.9436
Temperature	0.8882	0.7622	0.6842	0.8656	0.9331	0.7594	0.7889	0.8476	0.6478	0.6091	0.8535	0.7822	0.8095	0.8448	0.8045	0.8603

Higher values for the regularized training gain of the jack-knife test indicates greater contribution to the model for a variable (these values are not directly comparable between the different taxa). Test AUC numbers in parentheses are the standard deviation of the Test AUC scores. The top three variables are highlighted in bold for each taxon, both for the jack-knife variable contribution and test AUC values for Maxent models generated using a single variable. *indicates cross-validation cells that were eliminated due to low Test AUC scores.