An extinct hummingbird species that never was: a cautionary tale about sampling issues in molecular phylogenetics.

Abstract

The selection of species and individuals for molecular analyses critically affects inferences in various fields of systematic biology including phylogenetics, phylogeography, and species delimitation. Especially in areas like the Neotropical region where molecular analyses have recovered substantial within-species divergence and unexpected affinities of populations (Turchetto-Zolet et al. 2013), biases resulting from incomplete taxonomic or geographic sampling may compromise the understanding of phylogenetic relationships (Avendaño et al. 2017). Here we describe a case in which assessments of the validity of a potentially extinct species of Neotropical bird were likely compromised because within-species variation was not accounted for in phylogenetic analyses evaluating the alternative hypothesis that the only known specimen may represent a hybrid.