Untangling the contribution of adaptive versus non-adaptive processes in the evolution of reproductive isolation between Coenonympha butterflies.

Speciation is a key evolutionary process which has been studied in numerous organisms and at multiple scales, from lineage radiation to gene expression. However, the factors explaining the rise of new species are not yet fully understood, and the relative contribution of neutral versus selective evolutionary processes in triggering and maintaining reproductive isolation between lineages is still debated. To explore this question, we study the divergence of two butterfly species, Coenonympha arcania and C. gardetta (Nymphalidae), which diverged relatively recently but show strong ecological differences. Whole genome sequence data reveal high overall differentiation between the two lineages, best explained by a long period of isolation at the early stage of their divergence. Demographically explicit approaches identify that 6.6% of the genome (32.7 Mbp) is impermeable to gene flow between the two species. Lots of these barrier loci are located on the Z chromosome, potentially spanning 75% of its length which would indicate that a large Z effect is at play in this speciation. Moreover, only a small proportion of barriers showed signatures of selection, suggesting that non-adaptive processes largely contributed to the build-up of reproductive isolation. Therefore, although genes involved in stress response and response to hypoxia are interesting candidates under selection, the adaptation of C. gardetta to alpine conditions may not to be the main driver of speciation. Our study brings an original example of intertwined adaptive and non-adaptive processes leading to reproductive isolation in a speciation with secondary contact and improves our understanding of the genomic underpinnings of species divergence.