Phylogenomic analyses reveal non-monophyly of the antbird genera Herpsilochmus and Sakesphorus (Thamnophilidae), with description of a new genus for Herpsilochmus sellowi

Abstract

The family Thamnophilidae is a species-rich Neotropical radiation of passerine birds. Current classification of its 235 species is mostly based on morphological similarities, but recent studies integrating comprehensive phenotypic and phylogenetic data have redefined taxonomic limits of several taxa. Here, we assess generic relationships of Herpsilochmus, Sakesphorus, Thamnophilus, Biatas, and Dysithamnus using DNA sequences from the mitochondrion, nuclear exons, and ultraconserved elements, with further attention to interspecific relationships within Herpsilochmus. We show that Herpsilochmus and Sakesphorus are not monophyletic. We resolve Herpsilochmus sellowi as a deep-branch sister to the monotypic genus Biatas and Sakesphorus cristatus as sister to a clade comprising Herpsilochmus sensu stricto and Dysithamnus. These results are consistent across loci, obtained via concatenation and coalescent-based analyses, and supported by likelihood-ratio tests of the distribution of our sampled coalescent histories. The phenotypic distinctiveness of both H. sellowi and Biatas argues against merging them into a single genus. Because no generic name is available for H. sellowi, we describe a monotypic genus. The polyphyly of Sakesphorus warrants recognition of the available generic name Sakesphoroides for the distinctive and monotypic S. cristatus. Furthermore, we recover 6 well-supported species groups within Herpsilochmus sensu stricto. Within the context of the family as a whole, the ubiquity of long terminal branches representing monotypic genera points to extinction events among ancestors of these lineages. We suggest that retention of ancestral characters or random genetic drift coupled with extensive extinction could explain the high degree of morphological and ecological similarity across these taxa, but we highlight the potential role of the environment in driving adaptive phenotypic convergence. Finally, our results send a cautionary message against the blind use of phylogenies containing imputed data based on taxonomy due to the increasingly frequent mismatches between traditional taxonomic classification and molecular phylogenies. LAY SUMMARY We conduct phylogenomic analyses to infer generic relationships of Herpsilochmus, Sakesphorus, Thamnophilus, Biatas, and Dysithamnus. Using thousands of ultraconserved elements, exons, and mitochondrial DNA, we consistently show that the genus Herpsilochmus is not monophyletic because H. sellowi is a sister of the monotypic species Biatas nigropectus. Because the phenotypic distinctiveness of H. sellowi and B. nigropectus argues against merging them into a single genus and there is no available generic name for H. sellowi, we describe a monotypic genus for this species. We also found polyphyly of the genus Sakesphorus, warranting recognition of the available generic name Sakesphoroides for Sakesphorus cristatus. Our results provide a robust framework for downstream analyses of biogeographic and phenotypic evolution of Herpsilochmus antwrens and allies. This study adds to the increasing body of literature documenting the mismatch between traditional avian taxonomic classifications based on external morphology and evolutionary histories traced by modern genetic tools.