Species delimitation is fundamental to deciphering the mechanisms that generate and maintain biodiversity. Alpha taxonomy historically relied on expert knowledge to describe new species using phenotypic and biogeographic evidence, which has the appearance of investigator subjectivity. In contrast, DNA-based methods using the multispecies coalescent model (MSC) promise a more objective approach to describing biodiversity. However, recent criticisms of the MSC suggest that under some conditions, molecular delimitation may over-split lineages and identify species that do not reflect biological reality. Here, we reconcile these two approaches using empirical data for the Greenthroat Darter complex (Etheostoma lepidum), a small freshwater fish species with a disjunct distribution in Texas and New Mexico, USA. We demonstrate that MSC species delimitation recognizes all nine sampled populations as distinct species, sometimes splitting specimens from a single locality into multiple species. However, environmental, phenotypic, and biogeographic evidence do not fully corroborate the nine species supported by the MSC. Instead, the collective evidence indicates that E. lepidum fits a delimitation model comprised of three species that are consistent with the molecular phylogeny; Etheostoma lepidum (Greenthroat Darter) in rivers draining the eastern Edwards Plateau, Etheostoma cf. lepidum (Texas Darter) in the Concho and San Saba River systems, and Etheostoma cf. lepidum (Pecos Darter) in the Pecos River. One newly delimited species, E. cf. lepidum Pecos Darter, is likely highly imperiled due to its localized distribution in the Pecos River and reliance on vanishing spring-fed stream habitats. The impending biodiversity crisis makes integrative and swift species delimitation more necessary than ever. Our study exemplifies how classic taxonomic expertise combined with molecular phylogenetics can produce a more robust description of threatened biodiversity.