Data Types and the Phylogeny of Neoaves
Department of Biology, University of Florida, Gainesville, FL 32611, USA
Authors to whom correspondence should be addressed.
Received: 15 November 2020 / Revised: 14 December 2020 / Accepted: 15 December 2020 / Published: 5 January 2021
Some of the earliest studies using molecular data to resolve evolutionary history separated birds into three main groups: Paleognathae (ostriches and allies), Galloanseres (ducks and chickens), and Neoaves (the remaining ~95% of avian species). The early evolution of Neoaves, however, has remained challenging to understand, even as data from whole genomes have become available. We have recently proposed that some of the conflicts among recent studies may be due to the type of genomic data that is analyzed (regions that code for proteins versus regions that do not). However, a rigorous examination of this hypothesis using coding and non-coding data from the same genomic regions sequenced from a relatively large number of species has not yet been conducted. Here we perform such an analysis and show that data type does influence the methods used to infer evolutionary relationships from molecular sequences. We also show that conducting analyses using models of sequence evolution that were chosen to minimize reconstruction errors result in coding and non-coding trees that are much more similar, and we add to the evidence that non-coding data provide better information regarding neoavian relationships. While a few relationships remain problematic, we are approaching a good understanding of the evolutionary history for major avian groups.