In organisms reproducing sexually, speciation occurs when increasing divergence results in pre- or post-zygotic reproductive isolation between lineages. Studies focusing on reproductive isolation origin in early stages of speciation are common. Many rely on indirect measures of introgression providing limited information on the genomic architecture of reproductive isolation maintenance in the long term. This study focuses on direct measures of introgression between two species in a late stage of speciation. We used ddRADseq genotyping in a natural hybrid zone between Podarcis bocagei and P. carbonelli to examine admixture extent, estimate effective selection, analyse hybrid zone stability, and assess variation in selection against introgression across the genome. Hybridization was confirmed the narrow and bimodal hybrid zone demonstrating the existence of strong mechanisms of reproductive isolation. Simulations suggested that simple premating barriers were not enough to explain the observed distribution of admixture classes, pointing out the role of post-mating isolation. A geographic cline approach confirmed strong reproductive isolation and high effective selection preventing extensive introgression outside of the contact zone; and detected a signal of hybrid zone movement towards P. bocagei distribution. Genomic cline analysis revealed heterogeneous patterns of introgression among loci within the syntopy zone, but most of the loci do not introgress more or less than the genomic average maintaining a strong association with the genomic background of origin. However, genomic clines can be driven by confounding effects resulting in incongruences between both cline approaches. Importantly, overall patterns of introgression seem to result from strong intrinsic barriers across the genome, without clear “islands of differentiation”. Last, an important role of the Z chromosome in reproductive isolation is suggested.

Many long-term genetic monitoring programs began before next-generation sequencing became widely available. Older programs can now transition to new marker systems usually consisting of 1000s of SNP loci, but there are still important questions about comparability, precision, and accuracy of key metrics estimated using SNPs. Ideally, transitioned programs should capitalize on new information without sacrificing continuity of inference across the time series. We combined existing microsatellite-based genetic monitoring information with SNP-based microhaplotypes obtained from archived samples of Rio Grande silvery minnow (Hybognathus amarus) across a 20-year time series to evaluate point estimates and trajectories of key genetic metrics. Demographic and genetic monitoring bracketed multiple collapses of the wild population, and included cases where captive-born repatriates comprised the majority of spawners in the wild. Even with smaller sample sizes, microhaplotypes yielded comparable and in some cases more precise estimates of variance genetic effective population size, multilocus heterozygosity and inbreeding compared to microsatellites because many more microhaplotype loci were available. Microhaplotypes also recorded shifts in allele frequencies associated with population bottlenecks. Trends in microhaplotype-based inbreeding metrics were associated with the fraction of hatchery-reared repatriates to the wild, and should be incorporated into future genomic monitoring. Although differences in accuracy and precision of some metrics were observed between marker types, biological inferences and management recommendations were consistent.