The size of the sample used in species distribution models affects how well the model predicts where species are found. This study looked at how different sample sizes in the testing dataset can impact the accuracy of two important statistics used to evaluate these models. The researchers found that as the sample size increases, the accuracy of the statistics also improves. Both statistics, AUC and Se*, were equally accurate in estimating the model's predictive ability. It was also found that a sample size of at least 20-30 is needed for accurate estimates, especially for rare or poorly known species.