Articles
Phylogenomics of Orchidaceae: concatenated and coalescent analyses of the plastomes provide different insights into orchid phylogeny
Article number
1414_8
Pages
89 – 104
Language
English
Abstract
In the light of recent evidence, coalescent methods can accommodate inconsistencies of gene trees and perform well in estimating species trees for a large number of genes.
Using 76 protein-coding genes of plastomes from 73 orchids, we evaluated the impact of concatenated analyses (including Bayesian inference, maximum likelihood and Most Parsimony) and coalescent analyses on the reconstructed phylogeny and estimated the timing of diversification for Orchidaceae. Our results indicated that the phylogenetic positions of several tribes and subtribes, such as Epipogiinae, Gastrodieae, were unresolved or weakly supported in concatenated analyses.
Instead, the phylogenetic relationships of Orchidaceae were well estimated in coalescent analyses using genes with high phylogenetic information (GHPI) or GHPI in combination with ndh genes. Orchidaceae are subdivided into six clades instead of five in concatenated analyses.
Seven clades were resolved as successive sisters to higher Epidendroideae. There are likely two periods of stasis during the early diversification of Orchidaceae. The stem age of subfamily Vanilloideae was 10 Ma younger than previous results.
Our results indicate that coalescent analyses might provide new insights into the phylogeny and temporal evolution of Orchidaceae and represent a good example that coalescent analyses are more robust in resolving long-branch artifacts than concatenated analyses.
Using 76 protein-coding genes of plastomes from 73 orchids, we evaluated the impact of concatenated analyses (including Bayesian inference, maximum likelihood and Most Parsimony) and coalescent analyses on the reconstructed phylogeny and estimated the timing of diversification for Orchidaceae. Our results indicated that the phylogenetic positions of several tribes and subtribes, such as Epipogiinae, Gastrodieae, were unresolved or weakly supported in concatenated analyses.
Instead, the phylogenetic relationships of Orchidaceae were well estimated in coalescent analyses using genes with high phylogenetic information (GHPI) or GHPI in combination with ndh genes. Orchidaceae are subdivided into six clades instead of five in concatenated analyses.
Seven clades were resolved as successive sisters to higher Epidendroideae. There are likely two periods of stasis during the early diversification of Orchidaceae. The stem age of subfamily Vanilloideae was 10 Ma younger than previous results.
Our results indicate that coalescent analyses might provide new insights into the phylogeny and temporal evolution of Orchidaceae and represent a good example that coalescent analyses are more robust in resolving long-branch artifacts than concatenated analyses.
This paper comes with a set of 6 supplementary tables and 33 supplementary figures.
Click here to download the .zip file containing the supplemental data.
Publication
Authors
X. Ma, Z.H. Li, M.W. Chase, Y. Wen, Y. Han, J.W. Li, X.H. Jin
Keywords
coalescent analyses, concatenated analyses, Orchidaceae, phylogenomics, plastome, Vanilloideae
Online Articles (11)