Phylogeny of terraranan frogs based on 2, 665 loci and impacts of missing data on phylogenomic analyses. Issue 7 (3rd October 2021)

• Record Type: Journal Article
• Authors: Barrientos, Lucas S.; Streicher, Jeffrey W.; Miller, Elizabeth C.; Pie, Marcio R.; Wiens, John J.; Crawford, Andrew J.
• Is Part Of: Systematics and biodiversity. Volume 19:Issue 7(2021)
• Page Start: 818
• Page End: 833
• Publication Date: 2021-10-03
• Subjects: amphibians -- Anura -- missing data -- phylogenomics -- Terraranae -- ultraconserved elements; Biodiversity -- Periodicals; Biology -- Classification -- Periodicals; Natural history -- Periodicals; Biodiversity; Biology; Classification; Periodicals; 578
• DOI: 10.1080/14772000.2021.1933249
• Languages: English
• ISSNs: 1478-0933

Abstract

Abstract : Terraranae is a large clade of New World direct-developing frogs that includes 3–5 families and >1, 100 described species (∼15% of all named frog species). The relationships among major groups of terraranan frogs have been highly contentious, including conflicts among four recent phylogenomic studies utilizing 95, 220, 389, and 2, 214 nuclear loci, respectively. In this paper, we re-evaluate relationships within Terraranae using a novel genomic dataset for 16 ingroup species representing most terraranan families and subfamilies. The preferred data matrix consisted of 2, 665 nuclear loci from ultraconserved elements (UCEs), with a total of 743, 419 aligned base pairs and 57% missing data. Concatenated likelihood analyses and coalescent-based species-tree analyses recovered strong statistical support for the following relationships among terraranan families: (Brachycephalidae, (Eleutherodactylidae, (Craugastoridae + 'Strabomantidae'))). This position for Brachycephalidae agrees with two previous phylogenomic studies but conflicts with two others. Our results infer Strabomantis (of the Strabomantidae) to be with (or within) Craugastor (Craugastoridae), rather than with other strabomantid genera. This renders Strabomantidae paraphyletic with respect to Craugastoridae. Our results also suggest that Pristimantinae is paraphyletic with respect to Holoadeninae. We also find that using matrices with less missing data (and concomitantly fewer UCE loci) generally resulted in trees with lower mean branch support and problematic phylogenies (e.g. non-monophyly of terraranans). Overall, our results help resolve controversial relationships within one of the largest clades of frogs, with a dataset including ∼7 times more loci than those used in previous studies focused on this clade.