Open AccessInferring the phylogenetic position of the Drosophila flavopilosa group: Incongruence within and between mitochondrial and nuclear multilocus datasets
Author(s) -
De Ré Francine C.,
Robe Lizandra J.,
Wallau Gabriel L.,
Loreto Elgion L. S.
Publication year - 2017
Publication title -
journal of zoological systematics and evolutionary research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.769
H-Index - 50
eISSN - 1439-0469
pISSN - 0947-5745
DOI - 10.1111/jzs.12170
Subject(s) - biology , phylogenetic tree , drosophila virilis , mitochondrial dna , evolutionary biology , clade , nuclear gene , phylogenetics , drosophila (subgenus) , genetics , gene
Abstract Inconsistencies and incongruences in phylogenetics may result from inaccurate reconstructions or from real differences between gene and species trees. The flavopilosa group of species is assumed to be part of the virilis–repleta radiation of the Drosophila subgenus, but its positioning within this clade is still controversial. To address the possible causes of extensive incongruence among phylogenies regarding this relationship, concordance patterns within and between mitochondrial and nuclear loci were inferred, with the aim of providing a more confident description of the phylogenetic position of Drosophila incompta . For this, Bayesian and Bayesian concordance analyses were performed using mitochondrial (13 protein‐coding genes) and nuclear (25 protein‐coding genes) datasets, in a broader (18 Drosophila species alongside Phortica variegata as an outgroup) and a restricted form (four species of the sister radiations virilis–repleta/ Hawaiian Drosophila with D. simulans as an outgroup), considering the effects of saturation and compositional bias. Our results revealed the presence of significant inconsistencies between Drosophila mitochondrial and nuclear genomes, and the two groups of markers present contrasting signals in regard to the phylogenetic positioning of the D. flavopilosa group. Whereas mitochondrial genomes support the clade formed by D. incompta and Drosophila mojavensis , the nuclear genome indicates D. incompta and Drosophila virilis are sister species. The inconsistency remains even after double control of saturation and sampling differences in nucleotide‐partitioned and amino acid sequences, and even with the use of maximum likelihood approaches followed by the implementation of Partition Addition Bootstrap Alteration principles. Therefore, we argue that it stems mainly from differences in the evolutionary histories of both sets of markers, with the nuclear phylogeny taken as the best depiction of the evolutionary history of the flavopilosa group. Thus, our results highlight the importance of integrative approaches for phylogenetic reconstruction, especially for groups of putatively complex evolutionary history, as in the case of Drosophila .