Abstract
We explored the efficacy of species tree methods at the family level in birds, using the Australo-Papuan Fairy-wrens (Passeriformes: Maluridae) as a model system. Fairy-wrens of the genus Malurus are known for high intensities of sexual selection, resulting in some cases in rapid speciation. This history suggests that incomplete lineage sorting (ILS) of neutrally evolving loci could be substantial, a situation that could compromise traditional methods of combining loci in phylogenetic analysis. Using 18 molecular markers (5 anonymous loci, 7 exons, 5 introns, and 1 mitochondrial DNA locus), we show that gene tree monophyly across species could be rejected for 16 of 18 loci, suggesting substantial ILS at the family level in these birds. Using the software Concaterpillar, we also detect three statistically distinct clusters of gene trees among the 18 loci. Despite substantial variation in gene trees, species trees constructed using four different species tree estimation methods (BEST, BUCKy, and STAR) were generally well supported and similar to each other and to the concatenation tree, with a few mild discordances at nodes that could be explained by rapid and recent speciation events. By contrast, minimizing deep coalescences produced a species tree that was topologically more divergent from those of the other methods as measured by multidimensional scaling of trees. Additionally, gene and species trees were topologically more similar in the BEST analysis, presumably because of the species tree prior employed in BEST which appropriately assumes that gene trees are correlated with each other and with the species tree. Among the 18 loci, we also discovered 102 independent indel markers, which also proved phylogenetically informative, primarily among genera, and displayed a ∼4-fold bias towards deletions. As suggested in earlier work, the grasswrens (Amytornis) are sister to the rest of the family and the emu-wrens (Stipiturus) are sister to fairy-wrens (Malurus, Clytomyias). Our study shows that ILS is common at the family level in birds yet, despite this, species tree methods converge on broadly similar results for this family.
Highlights
In the field of phylogenetics, single-locus approaches using mitochondrial DNA have proved powerful because of mtDNA’s advantages over other genetic markers: relatively small population size, high mutation rate, putative lack of recombination and ease of access to due to its high copy number and the availability of primer sequences and whole genomes (Avise 2000; Zink and Barrowclough 2008)
Using tests of phylogenetic signal for each of our 18 loci, we have demonstrated that incomplete lineage sorting (ILS) is a genuine feature of our data set, as opposed to noise generated from mutational effects of lack of phylogenetic signal (Huang et al 2010)
The four species tree methods we used (BEST, STAR, BUCKy, minimizing deep coalescences (MDC)) generated estimates of phylogeny that were generally congruent and similar to that produced by concatenation, MDC produced species trees that were more divergent from those of other methods in multidimensional space
Summary
In the field of phylogenetics, single-locus approaches using mitochondrial DNA (mtDNA) have proved powerful because of mtDNA’s advantages over other genetic markers: relatively small population size, high mutation rate, putative lack of recombination and ease of access to due to its high copy number and the availability of primer sequences and whole genomes (Avise 2000; Zink and Barrowclough 2008). The high level of plumage divergence among species and populations likely contributes to strong pre-mating isolating mechanisms, in some clades, such as the largely allopatric chestnut-shouldered clade, plumage divergence is less extreme (Rowley and Russell 1997) If this history is accurate, we can regard ILS rather than hybridization as the main source of heterogeneity in gene tree topologies in this group. There have been several prior molecular phylogenetic studies on part or all of this family and they have used allozyme or DNA sequence data (Christidis and Schodde 1997; Christidis 1999; Christidis et al, 2010; Donnellan et al, 2009; Gardner et al 2010) They mainly addressed phylogenetic relationships among species within and between the genera Stipiturus, Amytornis and Malurus as well as biogeographic patterns of strong association between species and habitats. We discuss each of these methods in further detail below
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
