Figure 4.

Cumulative error rates per pairwise distance threshold value. Hypothetical threshold values were evaluated in intervals of 0.1% pairwise distance values to assess how well they performed in separating intraspecific from interspecific divergence values, assuming that the morphospecies are true representations of species boundaries. Observed genetic distances were calculated using the post-reassessment dataset. At a given test threshold value, “erroneous lumpings” refer to cases of distinct morphospecies that are grouped together into one provisional species, due to having nearest neighbour interspecific genetic distances that fall below the threshold. “Erroneous splittings” refer to single morphospecies split into two or more provisional species at that threshold, due to having a maximum intraspecific divergence above the threshold.