A major goal of speciation research is identifying loci that underpin barriers to gene flow. Population genomics takes a 'bottom-up' approach, scanning the genome for molecular signatures of processes that drive or maintain divergence. However, interpreting the 'genomic landscape' of speciation is complicated, because genome scans conflate multiple processes, most of which are not informative about gene flow. However, studying replicated population contrasts, including multiple incidences of secondary contact, can strengthen inferences. In this paper, we use linked-read sequencing (haplotagging), FST scans and genealogical methods to characterise the genomic landscape associated with replicate hybrid zone formation. We studied two flower colour varieties of the common snapdragon, Antirrhinum majus subspecies majus, that form secondary hybrid zones in multiple independent valleys in the Pyrenees. Consistent with past work, we found very low differentiation at one well-studied zone (Planoles). However, at a second zone (Avellanet), we found stronger differentiation and greater heterogeneity, which we argue is due to differences in the amount of introgression following secondary contact. Topology weighting of genealogical trees identified loci where haplotype diversity was associated with the two snapdragon varieties. Two of the strongest associations were at previously identified flower colour loci: Flavia, that affects yellow pigmentation, and Rosea/Eluta, two linked loci that affect magenta pigmentation. Preliminary analysis of coalescence times provides additional evidence for selective sweeps at these loci and barriers to gene flow. Our study highlights the impact of demographic history on the differentiation landscape, emphasising the need to distinguish between historical divergence and recent introgression.
Pal et al. (Mon,) studied this question.