Case studies: species delimitation
In Anopheles and in particular in Drosophila , the number of species-level entities recognized was much higher than the currently recognized number of morphospecies. All or most morphospecies were split into multiple species (Fig. 5; Figure S24). SODA exhibited a higher tendency to split individual morphospecies into multiple species than tr2. In Darwin’s finches, over-splitting was also visible, especially in the earlier-diverging species. However, we also found many morphospecies in the genera Geospiza and Camarhynchus to be lumped into a single species. This problem was prevalent when using tr2, but it also occurred with SODA. In Heliconius , over-splitting was less of a problem, but it occurred in some cases. Species-level entities mostly corresponded to established (sub)species in which, however, some entities were lumped (Heliconius melpomene aglaope / H. m. amaryllis ; Heliconius melpomene melpomene [from Panama] /H. m. rosina ). Since over-splitting (or lumping in the case of Darwin’s finches) was the overwhelming problem in almost all analyses, we refrained from performing additional analyses with the full WGS data (such analyses were not performed in the studies that generated the WGS data either).