No abstract

Recently, there have been contradictory results regarding relationships among 3 Australian parrot species known commonly as the Northern (Platycercus venustus), Pale-headed (P. adscitus), and Eastern rosella (P. eximius). An early phylogenetic analysis of the rosella genus Platycercus found that P. adscitus and P. eximius, 2 species putatively forming a hybrid zone, were sister taxa, while the geographically isolated P. venustus was in turn their sister taxon. However, a recent reexamination using multilocus sequence data found that evolutionary relationships differed depending on marker choice. Mitochondrial DNA (mtDNA) supported the earlier finding, but nuclear sequence data reflected an alternative relationship aligning P. adscitus and P. venustus as sister taxa. Two hypotheses were proposed to reconcile the discordance: (1) incomplete lineage sorting, and (2) mtDNA introgression from P. adscitus into populations of P. eximius. To clarify the relationships among these 3 species and to distinguish between the proposed hypotheses, restriction site-associated DNA (RAD) sequencing data from 23 individuals (~4 individuals per subspecies) were examined. Species trees were generated based on datasets either excluding loci with missing data (393 loci) or including them (5 nonoverlapping subsets of 1,000 loci). Findings strongly supported a sister relationship between P. adscitus and P. venustus (posterior probabilities of 1 at all nodes), suggesting that the discordant patterns previously identified may have been the result of mitochondrial introgression and/or capture. This and other potential explanations are discussed.

The study of hybrid zones advances understanding of the speciation process, and approaches incorporating genomic data are increasingly used to draw significant conclusions about the impact of hybridisation. Despite the progress made, the complex interplay of factors that can lead to substantially variable hybridisation outcomes are still not well understood, and many systems and/or groups remain comparatively poorly studied. Our study aims to broaden the literature on avian hybrid zones, investigating a potentially geographically and temporally complex putative hybrid zone between two native Australian non-sister parrot species, the pale-headed and eastern rosellas (Platycercus adscitus and Platycercus eximius, respectively). We analysed six plumage traits and >1400 RADseq loci and detected hybrid individuals and an unexpectedly complex geographic structure. The hybrid zone is larger than previously described due to either observer bias or its movement over recent decades. It comprises different subregions where genetic and plumage signals of admixture vary markedly in their concordance. Evidence of contemporary hybridisation (later generation and backcrossed individuals) both within and beyond the previously defined zone, when coupled with a lack of F1 hybrids and differential patterns of introgression among potentially diagnostic loci, indicates a lack of post-zygotic barriers to gene flow between species. Despite ongoing gene flow, species boundaries are likely maintained largely by strong pre-mating barriers. These findings are discussed in detail and future avenues for research into this system are proposed, which would be of benefit to the speciation and hybrid zone literature.

Recent work has highlighted the need to account for hierarchical patterns of genetic structure when estimating evolutionary and ecological parameters of interest. This caution is particularly relevant to studies of riverine organisms, where hierarchical structure appears to be commonplace. Here, we indirectly estimate dispersal distance in a hierarchically structured freshwater fish, Mogurnda adspersa. Microsatellite and mitochondrial DNA (mtDNA) data were obtained for 443 individuals across 27 sites separated by an average of 1.3 km within creeks of southeastern Queensland, Australia. Significant genetic structure was found among sites (mtDNA Φ(ST) = 0.508; microsatellite F(ST) = 0.225, F'(ST) = 0.340). Various clustering methods produced congruent patterns of hierarchical structure reflecting stream architecture. Partial mantel tests identified contiguous sets of sample sites where isolation by distance (IBD) explained F(ST) variation without significant contribution of hierarchical structure. Analysis of mean natal dispersal distance (σ) within sets of IBD-linked sample sites suggested most dispersal occurs over less than 1 km, and the average effective density (D(e)) was estimated at 11.5 individuals km(-1); indicating sedentary behavior and small effective population size are responsible for the remarkable patterns of genetic structure observed. Our results demonstrate that Rousset's regression-based method is applicable to estimating the scale of dispersal in riverine organisms and that identifying contiguous populations that satisfy the assumptions of this model is achievable with genetic clustering methods and partial correlations.