Phylogenetic analysis of the Australian rosella parrots (Platycercus) reveals discordance among molecules and plumage

Full Issue

“…However, a recent multilocus analysis of the genus incorporating both mitochondrial (ND2) and nuclear (9 loci) DNA sequence data found that relationships varied depending on marker choice. The single mitochondrial locus found P. adscitus and nominotypical subspecies P. e. eximius to be sister taxa, which supported the original finding, but analyses incorporating nuclear data found P. adscitus to be more closely related to P. venustus than P. eximius (Shipham et al 2015). Additionally, when all data were included in analyses, the 3 rd currently recognized subspecies of P. eximius (P. e. diemenensis), which is confined to the island of Tasmania, aligned with P. e. eximius.…”

“…Additionally, when all data were included in analyses, the 3 rd currently recognized subspecies of P. eximius (P. e. diemenensis), which is confined to the island of Tasmania, aligned with P. e. eximius. However, when examining mitochondrial DNA in isolation (Shipham et al 2015), individuals of P. e. diemenensis were sufficiently divergent for this subspecies to be considered a sister taxon to the rest of the rosellas in the subgenus Violania. Instances of mitonuclear discordance are becoming increasingly prevalent in the literature, and a number of potential explanations have been proposed, such as adaptive introgression, demographic differences between groups of interest, or sex-biased asymmetries in mate selection or offspring survival (see Toews and Brelsford 2012).…”

“…Instances of mitonuclear discordance are becoming increasingly prevalent in the literature, and a number of potential explanations have been proposed, such as adaptive introgression, demographic differences between groups of interest, or sex-biased asymmetries in mate selection or offspring survival (see Toews and Brelsford 2012). For rosellas in the subgenus Violania, 2 hypotheses have been proposed to explain the discrepancy between purely mitochondrial datasets and those incorporating nuclear data (Shipham et al 2015). Firstly, the relationships identified when analyzing nuclear data may be the result of incomplete lineage sorting (ILS), a phenomenon by which ancestral polymorphisms are retained after speciation events (Degnan and Rosenberg 2009, Steiner et al 2012, Corl and Ellegren 2013.…”

“…This has the potential to impede accurate phylogenetic inference and is a common problem when examining nuclear data (McLean et al 2012, Leache et al 2014, particularly for taxa that have arisen relatively recently (Degnan and Rosenberg 2009, Steiner et al 2012, Corl and Ellegren 2013. This hypothesis was supported for Violania by the low posterior probabilities and topological inconsistencies that arose when only nuclear loci were included in a species tree (Shipham et al 2015). Alternatively, hybridization between the 2 parapatrically distributed species may have facilitated mitochondrial introgression or capture, potentially from P. adscitus into the 2 currently recognized subspecies of P. eximius that are found on mainland Australia, the isolated Tasmanian subspecies still showing the 'true' mtDNA of P. eximius.…”

“…Alternatively, hybridization between the 2 parapatrically distributed species may have facilitated mitochondrial introgression or capture, potentially from P. adscitus into the 2 currently recognized subspecies of P. eximius that are found on mainland Australia, the isolated Tasmanian subspecies still showing the 'true' mtDNA of P. eximius. Given that support for species tree topologies found by Shipham et al (2015) was not consistently high, even when all loci were included in analyses (i.e. posterior probabilities for key nodes were as low as 0.61), further study is warranted.…”