Environmental or geological changes can create new niches that drive ecological species divergence without the immediate cessation of gene flow. However, few such cases have been characterized. On a recently formed volcano, Mt. Etna, Senecio aethnensis and S. chrysanthemifolius inhabit contrasting environments of high and low altitude, respectively. They have very distinct phenotypes, despite hybridizing promiscuously, and thus may represent an important example of ecological speciation “in action,” possibly as a response to the rapid geological changes that Mt. Etna has recently undergone. To elucidate the species’ evolutionary history, and help establish the species as a study system for speciation genomics, we sequenced the transcriptomes of the two Etnean species, and the outgroup, S. vernalis, using Illumina sequencing. Despite the species’ substantial phenotypic divergence, synonymous divergence between the high- and low-altitude species was low (dS = 0.016 ± 0.017 [SD]). A comparison of species divergence models with and without gene flow provided unequivocal support in favor of the former and demonstrated a recent time of species divergence (153,080 ya ± 11,470 [SE]) that coincides with the growth of Mt. Etna to the altitudes that separate the species today. Analysis of dN/dS revealed wide variation in selective constraint between genes, and evidence that highly expressed genes, more “multifunctional” genes, and those with more paralogs were under elevated purifying selection. Taken together, these results are consistent with a model of ecological speciation, potentially as a response to the emergence of a new, high-altitude niche as the volcano grew.

Summary Microbes can have profound effects on their hosts, driving natural selection, promoting speciation and determining species distributions. However, soil‐dwelling microbes are rarely investigated as drivers of evolutionary change in plants.We used metabarcoding and experimental manipulation of soil microbiomes to investigate the impact of soil and root microbes in a well‐known case of sympatric speciation, the Howea palms of Lord Howe Island (Australia). Whereas H. forsteriana can grow on both calcareous and volcanic soils, H. belmoreana is restricted to, but more successful on, volcanic soil, indicating a trade‐off in adaptation to the two soil types.We suggest a novel explanation for this trade‐off. Arbuscular mycorrhizal fungi (AMF) are significantly depleted in H. forsteriana on volcanic soil, relative to both H. belmoreana on volcanic soil and H. forsteriana on calcareous soil. This is mirrored by the results of survival experiments, where the sterilization of natural soil reduces Howea fitness in every soil–species combination except H. forsteriana on volcanic soil. Furthermore, AMF‐associated genes exhibit evidence of divergent selection between Howea species.These results show a mechanism by which divergent adaptation can have knock‐on effects on host–microbe interactions, thereby reducing interspecific competition and promoting the coexistence of plant sister species.

Hybrid zones typically form as a result of species coming into secondary contact, but can also be established in situ as an ecotonal hybrid zone, a situation which has been reported far less frequently. An altitudinal hybrid zone on Mount Etna between two ragwort species (the low elevation Senecio chrysanthemifolius and high elevation S. aethnensis) could potentially represent either of these possibilities. However, a scenario of secondary contact vs. speciation with gene flow has not been explicitly tested. Here, we test these alternatives and demonstrate that the data do not support secondary contact. Furthermore, we report that the previous analyses of speciation history of these species were based on admixed populations, which has led to inflated estimates of ongoing, interspecific gene flow. Our new analyses, based on 'pure' S. aethnensis and S. chrysanthemifolius populations, reveal gene exchange of less than one effective migrant per generation, a level low enough to allow the species to accumulate neutral, genomewide differences. Overall, our results are consistent with a scenario of speciation with gene flow and a divergence time which coincides with the rise of Mt. Etna to altitudes above 2000 m (~150 KY). Further work to quantify the role of adaptation to contrasting environments of high and low altitudes will be needed to support the scenario of recent ecological speciation in this system.

The hybrid zone on Mount Etna (Sicily) between Senecio aethnensis and Senecio chrysanthemifolius (two morphologically and physiologically distinct species) is a classic example of an altitudinal cline. Hybridization at intermediate altitudes and gradients in phenotypic and life-history traits occur along altitudinal transects of the volcano. The cline is considered to be a good example of ecological selection with species differences arising by divergent selection opposing gene flow. However, the possibility that the cline formed from recent secondary contact following an allopatric phase is difficult to exclude. We demonstrate a recent split between S. aethnensis and S. chrysanthemifolius (as recent as ∼32,000 years ago) and sufficient gene flow (2Nm > 1) to have prevented divergence (implicating a role for diversifying selection in the maintenance of the cline). Differentially expressed genes between S. aethnensis and S. chrysanthemifolius exhibit significantly higher genetic divergence relative to "expression invariant" controls, suggesting that species differences may in part be mediated by divergent selection on differentially expressed genes involved with altitude-related adaptation. The recent split time and the absence of fixed differences between these two ecologically distinct species suggest the rapid evolution to an altitudinal cline involving selection on both sequence and expression variation.

Phenotypic plasticity in ancestral populations is hypothesized to facilitate adaptation, but evidence is piecemeal and often contradictory. Further, whether ancestral plasticity increases the probability of parallel adaptive changes has not been explored. The most general finding is that ancestral responses to a new environment are reversed following adaptation (known as reversion). We investigated the contribution of ancestral plasticity to adaptive evolution of gene expression in two independently evolved lineages of zinc-tolerant Silene uniflora. We found that the general pattern of reversion is driven by the absence of a widespread stress response in zinc-adapted plants compared with zinc-sensitive plants. We show that ancestral plasticity that moves expression closer to the optimum value in the new environment influences the evolution of gene expression among genes that are likely to be involved in adaptation and increases the chance that genes are recruited repeatedly during adaptation. However, despite convergence in gene expression levels between independently adapted lineages, ancestral plasticity does not influence how similar expression values of adaptive genes become. Surprisingly, we also observed that ancestral plasticity that increases fitness often becomes genetically determined and fixed, that is, genetically assimilated. These results emphasize the important role of ancestral plasticity in parallel adaptation.