Summary of the results of the statistical tests of hypotheses for the comparison of the predictive models, in each scenario.

Only significant p-values are depicted in the table. Empty cells correspond to statistically non-significant probabilities. “Model 1 > Model 2” means that the alternative hypothesis (which is accepted if the p-value is < 0.05) is that the mean accuracy is greater for the Model 1 than for the Model 2. P-values are classified as follows: 0.05–0.01 = *; 0.01–0.001 = **; <0.001 = ***. In Model B, significant relationships among inputs are not represented (only among outputs) to avoid noise from variables that we do not predict and are expectable (i.e., the relationship among environmental characteristics).
(XLSX)