Phylogenetic conflict is often caused by incomplete lineage sorting (ILS) and interspecific gene flow. The most frequent gene tree may not necessarily represent the species tree due to the influence of gene flow. Gene trees that are consistent with species tree are suggested to be enriched in genomic regions less affected by ILS and gene flow, such as regions with low recombination and nucleotide diversity. Here, we investigated the phylogenetic relationships within three clades comprising a total of 13 species in the songbird family Paridae (tits) using one de novo genome assembly and 82 re-sequenced genomes. Based on multiple phylogenetic inference approaches and divergence time estimates of genomic regions reflecting different phylogenetic topologies, we found that the true species tree was not represented by the most frequent gene tree in one of the three clades (Poecile tits), mainly due to the effect of gene flow. For the two other clades, gene trees concordant with the species trees were significantly enriched in autosomal genomic regions with low recombination and low nucleotide diversity, but no such enrichment was observed in the Poecile clade. Interestingly, gene trees consistent with the species tree were over-represented on the Z chromosome in all three clades, possibly due to faster lineage sorting and lower gene flow on the Z chromosome than on autosomes. Our study highlights the confounding effects of recombination and nucleotide diversity on species tree estimation and the potential role of the Z chromosome in phylogenetic reconstruction in the face of ILS and gene flow.
Reconstructing the species tree in the face of incomplete lineage sorting and gene flow in three clades within the avian family Paridae