Speciation in the presence of gene flow: population genomics of closely related and diverging Eucalyptus species

Abstract

Speciation is a complex process that is fundamental to the origins of biological diversity. While there has been considerable progress in our understanding of speciation, there are still many unanswered questions, especially regarding barriers to gene flow in diverging populations. Eucalyptus is an appropriate system for investigating speciation mechanisms since it comprises species that are rapidly evolving across heterogeneous environments. We examined patterns of genetic variation within and among six closely related Eucalyptus species in subgenus Eucalyptus section Eucalyptus in south-eastern Australia (commonly known as the "green ashes"). We used reduced representation genome sequencing to genotype samples from populations across altitudinal and latitudinal gradients. We found one species, Eucalyptus cunninghamii, to be highly genetically differentiated from the others, and a population of mallees from Mount Banks to be genetically distinct and therefore likely to be a new undescribed species. Only modest levels of differentiation were found between all other species in the study. There was population structure within some species (e.g., E. obstans) corresponding to geographical factors, indicating that vicariance may have played a role in the evolution of the group. Overall, we found that lineages within the green ashes are differentiated to varying extents, from strongly diverged to much earlier stages of the speciation continuum. Furthermore, our results suggest the green ashes represent a group where a range of mechanisms (e.g., reticulate evolution and vicariance) have been operating in concert. These findings not only offer insights into recent speciation mechanisms in Eucalyptus, but also other species complexes.

Fig. 1

Distribution of the study species showing: Eucalyptus cunninghamii (), E. laophila (), E. stricta (), E. langleyi (), E. obstans (), and E. dendromorpha (). Leaf morphology and variations in bud and capsule size of each species are also shown (Klaphake , pp. 45–49). Locations sampled in the present study are shown. Details of species and populations sampled (and location codes) are presented in Table 1. Maps were generated using Australia’s Virtual Herbarium (2015)

Fig. 2

Principle coordnate analysis (PCoA) derived from the DArTseq SNP dataset showing: a all study taxa (based on 11,739 DArTseq markers), b the focal group based on 10,868 DArTseq markers and including all taxa except for Eucalyptus cunninghamii, NSW908486 from Stanwell Tops and E. sp. Mount Banks (axis 1 and 2 shown), and c the focal group (axis 1 and 3 shown). GBMWHA, Greater Blue Mountains World Heritage Area

Fig. 3

Population structure of the green ash eucalypts (based on 4783 DArTseq SNP markers). The optimum value of K was calculated using the method of Evanno et al. (2005) (the ∆K plots from STRUCTURE HARVESTER for each STRUCTURE analysis are presented in Supplementary Fig. S3). Each graph produced from STRUCTURE was also inspected to ensure the optimum value of K for each species was calculated. STRUCTURE barplots with different values of K are shown in Supplementary Fig. S4. Location codes correspond to those presented in Fig. 1

Fig. 4

TreeMix analysis of the green ash eucalypts (based on 11,739 SNPs) showing the a inferred maximum likelihood phylogeny showing four migration events and b residual fit plotted from the maximum likelihood tree in (a). In (a) the directionality of gene flow is indicated by arrows and coloured (yellow to red) according to their weight (0–50%). In (b), the colour bar to the right of the matrix indicates degree of relatedness between populations, with residuals above zero indicating populations that are more closely related to each other in the data than in the best-fit tree (i.e., bluer shades indicate population pairs that are candidates for admixture events). Sample codes correspond to those presented in Table 1. TreeMix analyses under varying numbers of admixture events (0–3 events) are shown in Supplementary Fig. S7