The phylogenomic forest of bird trees contains a hard polytomy at the root of Neoaves

Alexander Suh

doi:10.1111/zsc.12213

Abstract

Birds have arguably been the most intensely studied animal group for their phylogenetic relationships. However, the recent advent of genome‐scale phylogenomics has made the forest of bird phylogenies even more complex and confusing. Here, in this perspective piece, I show that most parts of the avian Tree of Life are now firmly established as reproducible phylogenetic hypotheses. This is to the exception of the deepest relationships among Neoaves. Using phylogenetic networks and simulations, I argue that the very onset of the super‐rapid neoavian radiation is irresolvable because of eight near‐simultaneous speciation events. Such a hard polytomy of nine taxa translates into 2 027 025 possible rooted bifurcating trees. Accordingly, recent genome‐scale phylogenies show extremely complex conflicts in this (and only this) part of the avian Tree of Life. I predict that the upcoming years of avian phylogenomics will witness many more, highly conflicting tree topologies regarding the early neoavian polytomy. I further caution against bootstrapping in the era of genomics and suggest to instead use reproducibility (e.g. independent methods or data types) as support for phylogenetic hypotheses. The early neoavian polytomy coincides with the Cretaceous–Paleogene (K‐Pg) mass extinction and is, to my knowledge, the first empirical example of a hard polytomy.

Highlights

The last few years have produced so many different phylogenetic trees of birds that it has become almost impossible to see the forest for the trees
I show that genome-scale phylogenomics has resolved most of these disputes, and I argue that the presence of a hard polytomy explains all of the remaining irresolvable branches of the neoavian Tree of Life
As carefully analysed retroposon presence/absence patterns contain negligible amounts of homoplasy (Shedlock et al 2004; Ray et al 2006; Han et al 2011), these discordances have been attributed to an extreme prevalence of incomplete lineage sorting (ILS) in some parts of the neoavian Tree of Life (Suh et al 2015)

Summary

Introduction

The last few years have produced so many different phylogenetic trees of birds that it has become almost impossible to see the forest for the trees. I show that genome-scale phylogenomics has resolved most of these disputes, and I argue that the presence of a hard polytomy explains all of the remaining irresolvable branches of the neoavian Tree of Life. To test for a hard polytomy, I generated phylogenetic networks in Splitstree4 (Huson & Bryant 2006) using publicly available trees/data sets (Jarvis et al 2014, 2015; Prum et al 2015; Suh et al 2015).

Results

Conclusion