Projection of current and future distribution of adaptive genetic units in an alpine ungulate.

Climate projections predict major changes in alpine environments by the end of the 21st century. To avoid climate-induced maladaptation and extinction, many animal populations will either need to move to more suitable habitats or adapt in situ to novel conditions. Since populations of a species exhibit genetic variation related to local adaptation, it is important to incorporate this variation into predictive models to help assess the ability of the species to survive climate change. Here, we evaluate how the adaptive genetic variation of a mountain ungulate-the Northern chamois (Rupicapra rupicapra)-could be impacted by future global warming. Based on genotype-environment association analyses of 429 chamois using a ddRAD sequencing approach, we identified genetic variation associated with climatic gradients across the European Alps. We then delineated adaptive genetic units and projected the optimal distribution of these adaptive groups in the future. Our results suggest the presence of local adaptation to climate in Northern chamois with similar genetic adaptive responses in geographically distant but climatically similar populations. Furthermore, our results predict that future climatic changes will modify the Northern chamois adaptive landscape considerably, with various degrees of maladaptation risk.

Extra-pair paternity, breeding density, and synchrony in natural cavities versus nestboxes in two passerine birds.

Most of what is known about extra-pair paternity in hole-nesting birds derives from studies using artificial nesting sites, such as nestboxes. However, it has rarely been investigated whether inference drawn from breeding events taking place in nestboxes matches what would be observed under natural conditions, that is, in natural cavities. We here report on a variation in promiscuity in blue tits and great tits nesting in natural cavities and nestboxes in an urban forest in Warsaw, Poland. Specifically, we tested whether local breeding density, local breeding synchrony, and extra-pair paternity (inferred from SNP data generated with a high-throughput genotyping by sequencing method) differed between birds nesting in natural cavities and nestboxes. In both blue tits and great tits, the frequency of extra-pair paternity was similar between the two cavity types. In blue tits, we observed shorter nearest neighbor distance, higher neighbor density, and higher synchronous neighbor density (i.e., density of fertile females) in nestboxes relative to natural cavities. No such pattern was found in great tits. Moreover, we detected a positive relationship between the proportion of extra-pair offspring in the nest and neighbor density around the nest in blue tits. Our results revealed that the provisioning of nestboxes did not change rates of extra-pair paternity, suggesting that conclusions drawn from nestbox studies might adequately represent the natural variation in extra-pair matings in some species or sites. However, the observed differences in spatiotemporal components of breeding dynamics highlight the fact that these parameters should be carefully considered when comparing mating behavior across studies and/or sites.

Population genetics reveals divergent lineages and ongoing hybridization in a declining migratory fish species complex.

Deciphering the effects of historical and recent demographic processes responsible for the spatial patterns of genetic diversity and structure is a key objective in evolutionary and conservation biology. Using population genetic analyses, we investigated the demographic history, the contemporary genetic diversity and structure, and the occurrence of hybridization and introgression of two species of anadromous fish with contrasting life history strategies and which have undergone recent demographic declines, the allis shad (Alosa alosa) and the twaite shad (Alosa fallax). We genotyped 706 individuals from 20 rivers and 5 sites at sea in Southern Europe at thirteen microsatellite markers. Genetic structure between populations was lower for the nearly semelparous species A. alosa, which disperses greater distances compared to the iteroparous species, A. fallax. Individuals caught at sea were assigned at the river level for A. fallax and at the region level for A. alosa. Using an approximate Bayesian computation framework, we inferred that the most likely long term historical divergence scenario between both species and lineages involved historical separation followed by secondary contact accompanied by strong population size decline. Accordingly, we found evidence for contemporary hybridization and bidirectional introgression due to gene flow between both species and lineages. Moreover, our results support the existence of at least one distinct species in the Mediterrannean sea: A. agone in Golfe du Lion area, and another divergent lineage in Corsica. Overall, our results shed light on the interplay between historical and recent demographic processes and life history strategies in shaping population genetic diversity and structure of closely related species. The recent demographic decline of these species' populations and their hybridization should be carefully considered while implementing conservation programs.

Evolving spatial conservation prioritization with intraspecific genetic data.

Spatial conservation prioritization (SCP) is a planning framework used to identify new conservation areas on the basis of the spatial distribution of species, ecosystems, and their services to human societies. The ongoing accumulation of intraspecific genetic data on a variety of species offers a way to gain knowledge of intraspecific genetic diversity and to estimate several population characteristics useful in conservation, such as dispersal and population size. Here, we review how intraspecific genetic data have been integrated into SCP and highlight their potential for identifying conservation area networks that represent intraspecific genetic diversity comprehensively and that ensure the long-term persistence of biodiversity in the face of global change.

Epigenetics and the city: Non-parallel DNA methylation modifications across pairs of urban-forest Great tit populations.

Identifying the molecular mechanisms involved in rapid adaptation to novel environments and determining their predictability are central questions in evolutionary biology and pressing issues due to rapid global changes. Complementary to genetic responses to selection, faster epigenetic variations such as modifications of DNA methylation may play a substantial role in rapid adaptation. In the context of rampant urbanization, joint examinations of genomic and epigenomic mechanisms are still lacking. Here, we investigated genomic (SNP) and epigenomic (CpG methylation) responses to urban life in a passerine bird, the Great tit (Parus major). To test whether urban evolution is predictable (i.e. parallel) or involves mostly nonparallel molecular processes among cities, we analysed both SNP and CpG methylation variations across three distinct pairs of city and forest Great tit populations in Europe. Our analyses reveal a polygenic response to urban life, with both many genes putatively under weak divergent selection and multiple differentially methylated regions (DMRs) between forest and city great tits. DMRs mainly overlapped transcription start sites and promotor regions, suggesting their importance in modulating gene expression. Both genomic and epigenomic outliers were found in genomic regions enriched for genes with biological functions related to the nervous system, immunity, or behavioural, hormonal and stress responses. Interestingly, comparisons across the three pairs of city-forest populations suggested little parallelism in both genetic and epigenetic responses. Our results confirm, at both the genetic and epigenetic levels, hypotheses of polygenic and largely nonparallel mechanisms of rapid adaptation in novel environments such as urbanized areas.

Linking genetic, morphological, and behavioural divergence between inland island and mainland deer mice.

The island syndrome hypothesis (ISH) stipulates that, as a result of local selection pressures and restricted gene flow, individuals from island populations should differ from individuals within mainland populations. Specifically, island populations are predicted to contain individuals that are larger, less aggressive, more sociable, and that invest more in their offspring. To date, tests of the ISH have mainly compared oceanic islands to continental sites, and rarely smaller spatial scales such as inland watersheds. Here, using a novel set of genome-wide SNP markers in wild deer mice (Peromyscus maniculatus) we conducted a genomic assessment of predictions underlying the ISH in an inland riverine island system: analysing island-mainland population structure, and quantifying heritability of phenotypes thought to underlie the ISH. We found clear genomic differentiation between the island and mainland populations and moderate to high marker-based heritability estimates for overall variation in traits previously found to differ in line with the ISH between mainland and island locations. FST outlier analyses highlighted 12 loci associated with differentiation between mainland and island populations. Together these results suggest that the island populations examined are on independent evolutionary trajectories, the traits considered have a genetic basis (rather than phenotypic variation being solely due to phenotypic plasticity). Coupled with the previous results showing significant phenotypic differentiation between the island and mainland groups in this system, this study suggests that the ISH can hold even on a small spatial scale.

Riverscape genetics in brook lamprey: genetic diversity is less influenced by river fragmentation than by gene flow with the anadromous ecotype.

Understanding the effect of human-induced landscape fragmentation on gene flow and evolutionary potential of wild populations has become a major concern. Here, we investigated the effect of riverscape fragmentation on patterns of genetic diversity in the freshwater resident European brook lamprey (Lampetra planeri) that has a low ability to pass obstacles to migration. We tested the hypotheses of (i) asymmetric gene flow following water current and (ii) an effect of gene flow with the closely related anadromous river lamprey (L. fluviatilis) ecotype on L. planeri genetic diversity. We genotyped 2472 individuals, including 225 L. fluviatilis, sampled from 81 sites upstream and downstream barriers to migration, in 29 western European rivers. Linear modelling revealed a strong positive relationship between genetic diversity and the distance from the river source, consistent with expected patterns of decreased gene flow into upstream populations. However, the presence of anthropogenic barriers had a moderate effect on spatial genetic structure. Accordingly, we found evidence for downstream-directed gene flow, supporting the hypothesis that barriers do not limit dispersal mediated by water flow. Downstream L. planeri populations in sympatry with L. fluviatilis displayed consistently higher genetic diversity. We conclude that genetic drift and slight downstream gene flow drive the genetic make-up of upstream L. planeri populations whereas gene flow between ecotypes maintains higher levels of genetic diversity in L. planeri populations sympatric with L. fluviatilis. We discuss the implications of these results for the design of conservation strategies of lamprey, and other freshwater organisms with several ecotypes, in fragmented dendritic river networks.

Demographic history and genomics of local adaptation in blue tit populations.

Understanding the genomic processes underlying local adaptation is a central aim of modern evolutionary biology. This task requires identifying footprints of local selection but also estimating spatio-temporal variations in population demography and variations in recombination rate and in diversity along the genome. Here, we investigated these parameters in blue tit populations inhabiting deciduous versus evergreen forests, and insular versus mainland areas, in the context of a previously described strong phenotypic differentiation. Neighboring population pairs of deciduous and evergreen habitats were weakly genetically differentiated (F ST = 0.003 on average), nevertheless with a statistically significant effect of habitat type on the overall genetic structure. This low differentiation was consistent with the strong and long-lasting gene flow between populations inferred by demographic modeling. In turn, insular and mainland populations were moderately differentiated (F ST = 0.08 on average), in line with the inference of moderate ancestral migration, followed by isolation since the end of the last glaciation. Effective population sizes were large, yet smaller on the island than on the mainland. Weak and nonparallel footprints of divergent selection between deciduous and evergreen populations were consistent with their high connectivity and the probable polygenic nature of local adaptation in these habitats. In turn, stronger footprints of divergent selection were identified between long isolated insular versus mainland birds and were more often found in regions of low recombination, as expected from theory. Lastly, we identified a genomic inversion on the mainland, spanning 2.8 Mb. These results provide insights into the demographic history and genetic architecture of local adaptation in blue tit populations at multiple geographic scales.

Surface temperatures of non-incubated eggs in great tits (Parus major) are strongly associated with ambient temperature.

Temperature is one of the best investigated environmental factors in ecological life-history studies and is increasingly considered in the contexts of climate change and urbanization. In avian ecology, few studies have examined the associations between thermal dynamics in the nest environment and its neighbouring air. Here, we placed avian nests and non-incubated eggs inside nest boxes at various air temperatures that ranged from 0.3 to 33.1 °C, both in the field and in laboratory conditions. We measured how the design of the boxes, their compass orientation and their location in more or less urbanized environments affected the surface temperature of nests and eggs. We also assessed whether covering the eggs with lining material influenced their surface temperature. Overall, across all performed tests, we found that the surface temperature of nests and eggs strongly reflected the air temperature measured outside of the nest boxes. While the design of the nest boxes had little influence on the temperature of nests and eggs, orienting the nest boxes to the north or to the west significantly decreased their surface temperature. The presence of lining material also kept eggs slightly warmer when air temperatures were low. Altogether these results suggest that non-incubated eggs are not well protected against extreme air temperatures prior to the onset of incubation. From an evolutionary point of view, producers of ectotherm eggs need therefore to time egg-laying appropriately in order to avoid unfavourable thermal nest environments.

Assortative Mating in Animals and Its Role for Speciation.

Evolutionary theory predicts that positive assortative mating-the tendency of similar individuals to mate with each other-plays a key role for speciation by generating reproductive isolation between diverging populations. However, comprehensive tests for an effect of assortative mating on species richness at the macroevolutionary scale are lacking. We used a meta-analytic approach to test the hypothesis that the strength of assortative mating within populations is positively related to species richness across a broad range of animal taxa. Specifically, we ran a phylogenetically independent meta-analysis using an extensive database of 1,447 effect sizes for the strength of assortative mating, encompassing 307 species from 130 families and 14 classes. Our results suggest that there is no relationship between the strength of assortative mating and species richness across and within major taxonomic groups and trait categories. Moreover, our analysis confirms an earlier finding that animals typically mate assortatively (global Pearson correlation coefficient: r=0.36; 95% confidence interval: 0.19-0.52) when accounting for phylogenetic nonindependence. We argue that future advances will rely on a better understanding of the evolutionary causes and consequences of the observed intra- and interspecific variation in the strength of assortative mating.

No evidence of inbreeding depression in fast declining herds of migratory caribou.

Identifying inbreeding depression early in small and declining populations is essential for management and conservation decisions. Correlations between heterozygosity and fitness (HFCs) provide a way to identify inbreeding depression without prior knowledge of kinship among individuals. In Northern Quebec and Labrador, the size of two herds of migratory caribou (Rivière-George, RG and Rivière-aux-Feuilles, RAF) has declined by one to two orders of magnitude in the last three decades. This raises the question of a possible increase in inbreeding depression originating from, and possibly contributing to, the demographic decline in those populations. Here, we tested for the association of genomic inbreeding indices (estimated with 22,073 SNPs) with body mass and survival in 400 caribou sampled in RG and RAF herds between 1996 and 2016. We found no association of individual heterozygosity or inbreeding coefficient with body mass or annual survival. Furthermore, those genomic inbreeding indices remained stable over the period monitored. These results suggest that the rapid and intense demographic decline of the herds did not cause inbreeding depression in those populations. Although we found no evidence for HFCs, if demographic decline continues, it is possible that such inbreeding depression would be triggered.

On the importance of time scales when studying adaptive evolution.

Long-term field studies coupled with quantitative genomics offer a powerful means to understand the genetic bases underlying quantitative traits and their evolutionary changes. However, analyzing and interpreting the time scales at which adaptive evolution occurs is challenging. First, while evolution is predictable in the short term, with strikingly rapid phenotypic changes in data series, it remains unpredictable in the long term. Second, while the temporal dynamics of some loci with large effects on phenotypic variation and fitness have been characterized, this task can be complicated in cases of highly polygenic trait architecture implicating numerous small effect size loci, or when statistical tests are sensitive to the heterogeneity of some key characteristics of the genome, like variation in recombination rate along the chromosomes. After introducing these aforementioned challenges, we discuss a recent investigation of the genomic architecture and spatio-temporal variation in great tit bill length, which was related to the recent use of bird feeders. We discuss how this case study illustrates the importance of considering different temporal scales and evolutionary mechanisms both while analyzing trait temporal trends and when searching for and interpreting the signals of putative genomic footprints of selection. More generally this commentary discusses interesting challenges for unraveling the time scale at which adaptive traits evolve and their genomic bases.

RAD-sequencing for estimating genomic relatedness matrix-based heritability in the wild: A case study in roe deer.

Estimating the evolutionary potential of quantitative traits and reliably predicting responses to selection in wild populations are important challenges in evolutionary biology. The genomic revolution has opened up opportunities for measuring relatedness among individuals with precision, enabling pedigree-free estimation of trait heritabilities in wild populations. However, until now, most quantitative genetic studies based on a genomic relatedness matrix (GRM) have focused on long-term monitored populations for which traditional pedigrees were also available, and have often had access to knowledge of genome sequence and variability. Here, we investigated the potential of RAD-sequencing for estimating heritability in a free-ranging roe deer (Capreolous capreolus) population for which no prior genomic resources were available. We propose a step-by-step analytical framework to optimize the quality and quantity of the genomic data and explore the impact of the single nucleotide polymorphism (SNP) calling and filtering processes on the GRM structure and GRM-based heritability estimates. As expected, our results show that sequence coverage strongly affects the number of recovered loci, the genotyping error rate and the amount of missing data. Ultimately, this had little effect on heritability estimates and their standard errors, provided that the GRM was built from a minimum number of loci (above 7,000). Genomic relatedness matrix-based heritability estimates thus appear robust to a moderate level of genotyping errors in the SNP data set. We also showed that quality filters, such as the removal of low-frequency variants, affect the relatedness structure of the GRM, generating lower h2 estimates. Our work illustrates the huge potential of RAD-sequencing for estimating GRM-based heritability in virtually any natural population.

Comparing genomic signatures of domestication in two Atlantic salmon (Salmo salar L.) populations with different geographical origins.

Selective breeding and genetic improvement have left detectable signatures on the genomes of domestic species. The elucidation of such signatures is fundamental for detecting genomic regions of biological relevance to domestication and improving management practices. In aquaculture, domestication was carried out independently in different locations worldwide, which provides opportunities to study the parallel effects of domestication on the genome of individuals that have been selected for similar traits. In this study, we aimed to detect potential genomic signatures of domestication in two independent pairs of wild/domesticated Atlantic salmon populations of Canadian and Scottish origins, respectively. Putative genomic regions under divergent selection were investigated using a 200K SNP array by combining three different statistical methods based either on allele frequencies (LFMM, Bayescan) or haplotype differentiation (Rsb). We identified 337 and 270 SNPs potentially under divergent selection in wild and hatchery populations of Canadian and Scottish origins, respectively. We observed little overlap between results obtained from different statistical methods, highlighting the need to test complementary approaches for detecting a broad range of genomic footprints of selection. The vast majority of the outliers detected were population-specific but we found four candidate genes that were shared between the populations. We propose that these candidate genes may play a role in the parallel process of domestication. Overall, our results suggest that genetic drift may have override the effect of artificial selection and/or point toward a different genetic basis underlying the expression of similar traits in different domesticated strains. Finally, it is likely that domestication may predominantly target polygenic traits (e.g., growth) such that its genomic impact might be more difficult to detect with methods assuming selective sweeps.

Impact of supplementation on deleterious mutation distribution in an exploited salmonid.

Deleterious mutations have important implications for the evolutionary trajectories of populations. While several studies recently investigated the dynamics of deleterious mutations in wild populations, no study has yet explored the fate of deleterious mutations in a context of populations managed by supplementation. Here, based on a dataset of nine wild and 15 supplemented Lake Trout populations genotyped at 4,982 single nucleotide polymorphisms (SNP)s by means of genotype by sequencing (GBS), we explored the effect of supplementation on the frequency of putatively deleterious variants. Three main findings are consequential for the management of fish populations. First, an increase in neutral genetic diversity in stocked populations compared with unstocked ones was observed. Second, putatively deleterious mutations were more likely to be found in unstocked than in stocked populations, suggesting a lower efficiency to purge deleterious mutations in unstocked lakes. Third, a population currently used as a major source for supplementation is characterized by several fixed putatively deleterious alleles. Therefore, other source populations with lower abundance of putatively deleterious mutations should be favored as sources of supplementation. We discuss management implications of our results, especially pertaining to the joint identification of neutral and deleterious mutations that could help refining the choice of source and sink populations for supplementation in order to maximize their evolutionary potential and to limit their mutation load.

Great tits and the city: Distribution of genomic diversity and gene-environment associations along an urbanization gradient.

Urbanization is a growing concern challenging the evolutionary potential of wild populations by reducing genetic diversity and imposing new selection regimes affecting many key fitness traits. However, genomic footprints of urbanization have received little attention so far. Using RAD sequencing, we investigated the genomewide effects of urbanization on neutral and adaptive genomic diversity in 140 adult great tits Parus major collected in locations with contrasted urbanization levels (from a natural forest to highly urbanized areas of a city; Montpellier, France). Heterozygosity was slightly lower in the more urbanized sites compared to the more rural ones. Low but significant effect of urbanization on genetic differentiation was found, at the site level but not at the nest level, indicative of the geographic scale of urbanization impact and of the potential for local adaptation despite gene flow. Gene-environment association tests identified numerous SNPs with small association scores to urbanization, distributed across the genome, from which a subset of 97 SNPs explained up to 81% of the variance in urbanization, overall suggesting a polygenic response to selection in the urban environment. These findings open stimulating perspectives for broader applications of high-resolution genomic tools on other cities and larger sample sizes to investigate the consistency of the effects of urbanization on the spatial distribution of genetic diversity and the polygenic nature of gene-urbanization association.

Gene flow does not prevent personality and morphological differentiation between two blue tit populations.

Understanding the causes and consequences of population phenotypic divergence is a central goal in ecology and evolution. Phenotypic divergence among populations can result from genetic divergence, phenotypic plasticity or a combination of the two. However, few studies have deciphered these mechanisms for populations geographically close and connected by gene flow, especially in the case of personality traits. In this study, we used a common garden experiment to explore the genetic basis of the phenotypic divergence observed between two blue tit (Cyanistes caeruleus) populations inhabiting contrasting habitats separated by 25 km, for two personality traits (exploration speed and handling aggression), one physiological trait (heart rate during restraint) and two morphological traits (tarsus length and body mass). Blue tit nestlings were removed from their population and raised in a common garden for up to 5 years. We then compared adult phenotypes between the two populations, as well as trait-specific Qst and Fst . Our results revealed differences between populations similar to those found in the wild, suggesting a genetic divergence for all traits. Qst -Fst comparisons revealed that the trait divergences likely result from dissimilar selection patterns rather than from genetic drift. Our study is one of the first to report a Qst -Fst comparison for personality traits and adds to the growing body of evidence that population genetic divergence is possible at a small scale for a variety of traits including behavioural traits.

Do genetic drift and accumulation of deleterious mutations preclude adaptation? Empirical investigation using RADseq in a northern lacustrine fish.

Understanding genomic signatures of divergent selection underlying long-term adaptation in populations located in heterogeneous environments is a key goal in evolutionary biology. In this study, we investigated neutral, adaptive and deleterious genetic variation using 7,192 SNPs in 31 Lake Trout (Salvelinus namaycush) populations (n = 673) from Québec, Canada. Average genetic diversity was low, weakly shared among lakes, and positively correlated with lake size, indicating a major role for genetic drift subsequent to lake isolation. Putatively deleterious mutations were on average at lower frequencies than the other SNPs, and their abundance relative to the entire polymorphism in each population was positively correlated with inbreeding, suggesting that the effectiveness of purifying selection was negatively correlated with inbreeding, as predicted from theory. Despite evidence for pronounced genetic drift and inbreeding, several outlier loci were associated with temperature and found in or close to genes with biologically relevant functions notably related to heat stress and immune responses. Outcomes of gene-temperature associations were influenced by the inclusion of the most inbred populations, in which allele frequencies deviated the most from model predictions. This result illustrates challenge in identifying gene-environment associations in cases of high genetic drift and restricted gene flow and suggests limited adaptation in populations experiencing higher inbreeding. We discuss the relevance of these findings for the conservation and management, notably regarding stocking and genetic rescue, of Lake Trout populations and other species inhabiting highly fragmented habitats.

Investigating the extent of parallelism in morphological and genomic divergence among lake trout ecotypes in Lake Superior.

Understanding the emergence of species through the process of ecological speciation is a central question in evolutionary biology which also has implications for conservation and management. Lake trout (Salvelinus namaycush) is renowned for the occurrence of different ecotypes linked to resource and habitat use throughout North America. We aimed to unravel the fine genetic structure of the four lake trout ecotypes in Lake Superior. A total of 486 individuals from four sites were genotyped at 6822 filtered SNPs using RADseq technology. Our results revealed different extent of morphological and genetic differentiation within the different sites. Overall, genetic differentiation was weak but significant and was on average three times higher between sites (mean FST  = 0.016) than between ecotypes within sites (mean FST  = 0.005) indicating higher level of gene flow or a more recent shared ancestor between ecotypes within each site than between populations of the same ecotype. Evidence of divergent selection was also found between ecotypes and/or in association with morphological variation. Outlier loci found in genes related to lipid metabolism and visual acuity were of particular interest in this context of ecotypic divergence. However, we did not find clear indication of parallelism at the genomic level, despite the presence of phenotypic parallelism among some ecotypes from different sampling sites. Overall, the occurrence of different levels of both genomic and phenotypic differentiation between ecotypes within each site with several differentiated loci linked to relevant biological functions supports the presence of a continuum of divergence in lake trout.

Inferring the demographic history underlying parallel genomic divergence among pairs of parasitic and nonparasitic lamprey ecotypes.

Understanding the evolutionary mechanisms generating parallel genomic divergence patterns among replicate ecotype pairs remains an important challenge in speciation research. We investigated the genomic divergence between the anadromous parasitic river lamprey (Lampetra fluviatilis) and the freshwater-resident nonparasitic brook lamprey (Lampetra planeri) in nine population pairs displaying variable levels of geographic connectivity. We genotyped 338 individuals with RAD sequencing and inferred the demographic divergence history of each population pair using a diffusion approximation method. Divergence patterns in geographically connected population pairs were better explained by introgression after secondary contact, whereas disconnected population pairs have retained a signal of ancient migration. In all ecotype pairs, models accounting for differential introgression among loci outperformed homogeneous migration models. Generating neutral predictions from the inferred divergence scenarios to detect highly differentiated markers identified greater proportions of outliers in disconnected population pairs than in connected pairs. However, increased similarity in the most divergent genomic regions was found among connected ecotype pairs, indicating that gene flow was instrumental in generating parallelism at the molecular level. These results suggest that heterogeneous genomic differentiation and parallelism among replicate ecotype pairs have partly emerged through restricted introgression in genomic islands.