Wildlife species are challenged and threatened by various infectious diseases that act as important selective forces and demographic drivers of populations. Yet, studies about host genetic factors and disease susceptibility are very limited. Cetacean morbillivirus (CeMV) has emerged as a major viral threat to cetacean populations worldwide, contributing to the death of tens of thousands of individuals of multiple dolphin and whale species. To understand the genomic basis of immune responses to CeMV, we generated and analysed whole genomes of 53 Indo-Pacific bottlenose dolphins (Tursiops aduncus) exposed to Australia’s largest CeMV-related mortality event known to date. The genomic dataset consisted of 7,720,686 SNPs anchored onto 23 chromosome-length scaffolds and 77 short scaffolds. Allele frequency estimates between survivors and non-survivors of the outbreak revealed 11,009 candidate SNPs, of which 498 were annotated to 220 protein coding genes. These included 36 genes with functions related to innate and adaptive immune responses, and cytokine signalling pathways. The list also included genes known to be involved in immune responses to other morbilliviruses, such as measles in humans and the phocine distemper virus in pinnipeds. Our study characterised genomic regions and pathways that likely contribute to CeMV susceptibility and resistance in dolphins, representing a stride towards clarifying the complex interactions of the cetacean immune system. It also emphasises the relevance of whole genome datasets to study the genetics of wildlife diseases.

Biological invasions are accelerating, and invasive species can have large economic impacts as well as severe consequences for biodiversity. During invasions, species can interact, potentially resulting in hybridization. Here, we examined two Cakile species, C. edentula and C. maritima (Brassicaceae), that co-occur and may hybridize during range expansion in separate regions of the globe. Cakile edentula invaded each location first, while C. maritima established later, apparently replacing the former. We assessed the evidence for hybridization in western North America and Australia, where both species have been introduced, and identified source populations with 4561 SNPs using Genotype-by-Sequencing. Our results indicate that the C. edentula in Australia originated from one region of eastern North America while in western North America it is likely from multiple sources. The C. maritima in Australia were derived from at least two different parts of Europe while the introduction in western North America is from one. Although morphological evidence of hybridization is generally limited to mixed species populations in Australia and virtually absent elsewhere, our genetic analysis revealed relatively high levels of hybridization in Australia (58% hybrids) and supported the presence of hybrids in western North America (16%) and New Zealand. Hybrids might be commonly overlooked in invaders, as identification based solely on morphological traits may represent only the tip of the iceberg. Our study reveals a repeated pattern of invasion, hybridization and apparent replacement of one species by another, which offers an opportunity to investigate the role of hybridization and introgression during invasion.

Skin is the largest mammalian organ and the first defensive barrier against the external environment. The skin and fur of mammals can host a wide variety of ectoparasites, many of which are phylogenetically diverse, specialized, and specifically adapted to their hosts. Among hematophagous dipteran parasites, volatile organic compounds (VOCs) are known to serve as important attractants, leading parasites to compatible sources of blood meals. VOCs have been hypothesized to be mediated by host-associated bacteria, which may thereby indirectly influence parasitism. Host-associated bacteria may also influence parasitism directly, as has been observed in interactions between animal gut microbiota and malarial parasites. Hypotheses relating bacterial symbionts and eukaryotic parasitism have rarely been tested among humans and domestic animals, and have to our knowledge never been tested in wild vertebrates. In this study, we use Afrotropical bats, hematophagous ectoparasitic bat flies, and haemosporidian (malarial) parasites vectored by bat flies as a model to test the hypothesis that the vertebrate host microbiome is linked to parasitism in a wild system. We identify significant correlations between bacterial community composition of the skin and dipteran ectoparasite prevalence across four major bat lineages, as well as striking differences in skin microbial network characteristics between ectoparasitized and non-ectoparasitized bats. We also identify links between the oral microbiome and presence of malarial parasites among miniopterid bats. Our results support the hypothesis that microbial symbionts may serve as indirect mediators of parasitism among eukaryotic hosts and parasites.

Genomic-based epidemiology can provide insight into the origins and spread of herbicide resistance mechanisms in weeds. We used kochia (Bassia scoparia) populations resistant to the herbicide glyphosate from across western North America to test the alternative hypotheses that 1) a single EPSPS gene duplication event occurred initially in the Central Great Plains and then subsequently spread to all other geographical areas now exhibiting glyphosate-resistant kochia populations or that 2) gene duplication occurred multiple times in independent events in a case of parallel evolution. We used qPCR markers previously developed for measuring the various units of the EPSPS tandem duplication to investigate whether all glyphosate-resistant plants had the same EPSPS repeat structure. We also investigated population structure using simple sequence repeat (SSR) markers to determine the relatedness of kochia populations from across the Central Great Plains, Northern Plains, and the Pacific Northwest. We identified three distinct EPSPS-duplication haplotypes that had geographic associations with the Central Great Plains, Northern Plains, and Pacific Northwest. Population structure revealed a group of populations around the first reported occurrence of glyphosate resistance in the Central Great Plains, a separate group of Pacific Northwest populations, and some relatedness of populations from geographically isolated areas. The results support at least three independent origins of glyphosate resistance in kochia, followed by substantial and mostly geographically localized gene flow to spread the resistance alleles into diverse genetic backgrounds.

Is a handful of genes responsible for the common starling invasion success?Invasive species have the ability to colonize new habitats across distinct areas of the globe, rapidly adjusting to new biotic and abiotic conditions, and often experiencing little impact from the decrease in effective population size and genetic diversity. Still, as each invading population represents a subsample of the original native distribution, it is frequent to see variability in terms of the genetic makeup of invading populations and consequently differences in invasion success rates across their non-native range (Blackburn et al. 2017). Invasion success results largely from a combination of the genetic diversity in the native populations, the number of founders and founder events that are effectively get established at the introduction sites, the presence of particularly important genes or new interactions between genes, and genotype x environment associations (GEAs) (Lee 2002; Blackburn et al. 2017). From the genetic side of the process, invasive populations undergo dramatic changes compared to their native range due to a combination of genetic bottlenecks from the source population with genetic drift at both the introduction site and at the expansion front as the species invades new habitat (Dlugosch & Parker 2008). Most differences are observed at the neutral level, as quick range expansions lead to an increase in frequency of random neutral variants at the range expansion front, termed allele surfing (Excoffieret al. 2009). But allele surfing might also lead to an increase in frequency of maladapted alleles, which might limit the expansion success of invading populations (Peischl & Excoffier 2017).The common starling (Sturnus vulgaris ) is an example of a worldwide invasive species that has been introduced most successfully from its Palearctic range into three other continents including North America, Africa and Australia. Still, the success of the introductions seems to vary between introduction sites. The biggest difference in introduction success might be that found between North America and Australia. Starling invasion of North America started with an introduction of ~160 birds in the east coast in 1890s that rapidly expanded across all of North America, establishing populations in the west coast within 50 years (Bodt et al. 2020). In Australia, while starlings were introduced at five locations about sixteen times since the 19th century, only two resulted in established populations (Bodt et al. 2020), suggesting the existence of stronger barriers to expansion than in North America. In a From the Cover article in this issue of Molecular Ecology, Stuart and Cardilini et al. (2020) used Genotyping-by-Sequencing to explore how landscape and environmental heterogeneity shaped the genetic population structure and adaptation of the common starling multiple invasions of Australia, and compare it to the patterns observed in North America, examined in Hofmeister et al. (2019). Stuart and Cardilini et al. (2020) determined that starlings in Australia are distributed across three distinct environmental groups characterized by temperature and precipitation in ‘arid’, ‘semi-arid’ and ‘non-arid’. However, neutral and adaptive population structure was incongruent with the defined environmental subdivision, rather showing a pattern of isolation-by-distance (IBD). This pattern was further supported by the levels of genetic diversity and divergence across populations. Genetic diversity decreased with distance from the introduction sites in SE Australia, and genetic differentiation between populations increased with geographic distance (Figure 1A and 1B). These patterns contrast with those observed at other regions of the world where starlings have been introduced (Hofmeister et al. 2019; Bodt et al. 2020). The starling invasion of North America showed a genetic pattern not consistent with IBD, but rather with isolation-by-environment. In accordance to the rapid continent wide expansion, Hofmeister et al. (2019) also found very low differentiation between sampling localities, indicating not only high dispersal ability but also a panmictic population across the entire North America. Studies on invasion genetics clearly indicate that multiple introductions lead to increased genetic diversity, which in turn provide increased chances to expand into more diverse areas (more genomic targets of selection) (Dlugosch & Parker 2008; Blackburnet al. 2017). However, despite multiple successful introductions in Australia vs. a single introduction in North America, the extent of the invasion is smaller in Australia and the levels of population differentiation were found to be higher. The authors propose that despite the higher genetic diversity, there could also be greater constraints for gene flow, potentially attributed to environmental (dis)similarities of the different introduction sites in Australia with the native range. The North American climate is more similar to that of the native range of the starling in the Palearctic, and thus it would be expected that population structure would be more evident in Australia where individuals would have to adapt to more extreme conditions than they were adapted in the native range (Colautti & Lau 2017). To understand the role of environmental variability on expansion, both studies performed outlier-scans and GEAs to identify loci under selection. Contrarily to initial expectations, outlier scans failed at providing strong evidence of selection in association with population structure or environmental variability. In the case of North American starling, this could result from a low genetic diversity across populations, approaching what would be expected in panmixia. In the case in Australian starlings, even in the presence of population structure, outlier detection might simply reflect stochastic patterns as a result of allele surfing. Using redundancy analyses to detect GEAs, both studies found evidence of adaptation to temperature and precipitation. Stuart and Cardilini et al. (2020) found associations of the populations found in ‘arid’ regions with environmental variables that relate to higher and less variable temperatures, and with increased variability in precipitation (Figure 1C and 1D). Associations with the other environmental regimes (‘semi-arid’ and ‘non-arid’) were not as clear, although subdivision of populations in the GEA analysis space appeared to match neutral population structure for these populations. Similarly, Hofmeister et al. (2019) also detected associations between population structure and environmental variables, particularly mean annual temperature.One of the most interesting results that emerges from the combined analysis of both studies is the identification of similar loci under selection in both Australia and North America. Two genes were reported as candidates for selection in both studies, and both are related to cell structure regulation, suggesting a potential association with cell viability in variable and extreme environments. As noted by many other studies, successful invasions might only need to encompass a few genes that then become important for adaptation to novel conditions (Lee 2002; Nadeau & Urban 2019). Selection in just a few key loci can be easily spread in species with high mobility, also benefiting from allele surfing, leading to a rapid increase in frequency at the range expansion front and potentially across the entire invasive range. This might explain how such distinct invasion histories as those found across the starling populations around the world led to contradicting expectations in terms of invasiveness (Bodt et al. 2020). Still, the role of other mechanisms such as plasticity, genomic rearrangements, epigenetic variation, among others, still needs to be further examined to fully explain the differences in starling invasion dynamics (Bock et al. 2017). The results provided by Stuart and Cardilini et al.(2020) evidenced that particularly advantageous loci can be positively selected multiple times, across multiple introduction events and through very distinct geographic contexts, despite low effective population sizes and various demographic scenarios. This manuscript suggests that the common starling invasion success around the world has been greatly impacted by a hand full of genes that allow adaptation to extreme environmental conditions. But additional research in other less successful introduction sites should provide further evidence into the mechanisms involved in the invasion success and allow to build a more comprehensive picture of the mechanisms involved.

Endemics co-occur because they evolved in situ and persist regionally or because they evolved ex situ and later dispersed to shared habitats, generating evolutionary or ecological endemicity centres, respectively. We investigate whether different endemicity centres can intertwine in the region ranging from Alps to Sicily, by studying their butterfly fauna. We gathered an extensive occurrence dataset for butterflies of the study area (27,123 records, 269 species, in cells of 0.5x0.5 degrees of latitude-longitude). We applied molecular-based delimitation methods (GMYC model) to 26,557 COI sequences of Western Palearctic butterflies. We identified entities based on molecular delimitations and the most recent checklist of European butterflies and objectively attributed occurrences to their most probable entity. We obtained a zoogeographic regionalisation based on the 69 endemics of the area. Using phylogenetic ANOVA we tested if endemics from different centres differ from each other and from non-endemics for key ecological traits and divergence time. Endemicity showed high incidence in the Alps and Southern Italy. The regionalisation separated the Alps from the Italian Peninsula and Sicily. The endemics of different centres showed a high turnover and differed in phenology and distribution traits. Endemics are on average younger than non-endemics and the Peninsula-Sicily endemics also have lower variance in divergence than those from the Alps. The observed variation identifies Alpine endemics as paleoendemics, now occupying an ecological centre, and the Peninsula-Sicily ones as neoendemics, that diverged in the region since the Pleistocene. The results challenge the common view of the Alpine-Apennine area as a single “Italian refugium”.

Physical disruption of gene flow among mountains is commonly viewed as an important process for the generation of hyperdiverse tropical mountain biotas. However, the role of in situ diversification within mountains has been seldom explored. Here we evaluate spatially fine-scale patterns of arthropod community assembly within a single mountain to understand the role of dispersal limitation and landscape features as drivers of tropical mountain diversity. We focus on a single tropical sky-island of the Transmexican Volcanic Belt, where we sampled whole-communities of arthropods for eight orders with a comparable design at a spatial scale ranging from 50 m to 20 km, using 840 pitfall traps and whole community metabarcoding. We explored multiple hierarchical levels, from individual haplotypes to lineages at 0.5, 1.5, 3, 5, 7.5% similarity thresholds, to evaluate patterns of richness, turnover and distance decay of similarity with isolation-by-distance and isolation-by-resistance approaches. Our results showed that distance and altitude influence distance decay of similarity at all hierarchical levels. This holds for arthropod groups of contrasting dispersal abilities, but with different strength depending on the spatial scale. Our results suggest long-term persistence of lineages within sky islands, combined with local-scale differentiation, may be an important driver of high arthropod biodiversity in tropical mountains.

Parental age can affect offspring telomere length through heritable and epigenetic-like effects, but at what stage during development these effects are established is not well known. To address this, we conducted a cross-fostering experiment in common gulls (Larus canus) that enabled us distinguish between pre- and post-natal parental age effects on offspring telomere length. Whole clutches were exchanged after clutch completion within and between parental age classes (young and old) and blood samples were collected from chicks at hatching and during the fastest growth phase (11 days later) to measure telomeres. Neither the ages of the natal nor the foster parents’ predicted the telomere length or the change in telomere lengths of their chicks. Telomere length was repeatable within chicks, but increased across development (repeatability = 0.55, Intraclass Correlation Coefficient within sampling events 0.934). Telomere length and the change in telomere length were not predicted by post-natal growth rate. Taken together, these findings suggest that in common gulls, telomere length during early in life is not influenced by parental age or growth rate, which may indicate that protective mechanisms buffer telomeres from external conditions during development in this relatively long-lived species.

Maternal effects have been well documented for offspring morphology and life history traits in plants and terrestrial animals, yet little is known about maternal effects in corals. Further, few studies have explored maternal effects in gene expression. In a previous study, F1 interspecific hybrid and purebred larvae of the coral species Acropora tenuis and A. loripes were settled and exposed to ambient or elevated temperature and pCO2 conditions for seven months. At this stage, the hybrid coral recruits from both ocean conditions exhibited strong maternal effects in several fitness traits. We conducted RNA-sequencing on samples from the same experiment and showed that gene expression of the hybrid Acropora also showed clear maternal effects. Only 40 genes were differentially expressed between hybrids and their maternal progenitor. In contrast, ~2000 differentially expressed genes were observed between hybrids and their paternal progenitors, and between the reciprocal F1 hybrids. These results indicate that maternal effects in coral gene expression can be long-lasting. Unlike findings from most short-term stress experiments in corals, no genes were differentially expressed in the hybrid nor purebred offspring after seven months of exposure to elevated temperature and pCO2 conditions.

Invasive pathogens can be a threat when they affect human health, food production or ecosystem services, by displacing resident species, and we need to understand the cause of their establishment. We studied the patterns and causes of the establishment of the pathogen Dickeya solani that recently invaded potato agrosystems in Europe by assessing its invasion dynamics and its competitive ability against the closely-related resident D. dianthicola species. Epidemiological records over one decade in France revealed the establishment of D. solani and the maintenance of the resident D. dianthicola in potato fields exhibiting blackleg symptoms. Using experimentations, we showed that D. dianthicola caused a higher symptom incidence on aerial parts of potato plants than D. solani, while D. solani was more aggressive on tubers (i.e. with more severe symptoms). In co-infection assays, D. dianthicola outcompeted D. solani in aerial parts, while the two species co-existed in tubers. A comparison of 76 D. solani genomes (56 of which having been sequenced here) revealed balanced frequencies of two previously uncharacterized alleles, VfmBPro and VfmBSer, at the vfmB virulence gene. Experimental inoculations showed that the VfmBSer population was more aggressive on tubers while the VfmBPro population outcompeted the VfmBSer population in stem lesions, suggesting an important role of the vfmB virulence gene in the ecology of the pathogens. This study thus brings novel insights allowing a better understanding of the pattern and causes of the D.solani invasion into potato production agrosystems, and the reasons why the endemic D.dianthicola nevertheless persisted.

The microcrustacean Daphnia is arguably one of the most studied zooplankton species, having a well understood ecology, life history, and a relatively well studied evolutionary history. Despite this wealth of knowledge, species boundaries within this genus often remain elusive and the major evolutionary forces driving the diversity of daphniids remain controversial. This genus contains more than 300 species with multiple cryptic species complexes, with many closely related species able to hybridize. Here we review speciation research in Daphnia within the framework of current speciation theory. We evaluate the role of geography, ecology, and biology in restricting gene flow and promoting diversification. Of the 253 speciation studies on Daphnia, the majority of studies examine geographic barriers (55%). While evidence shows that geographic barriers play a role in species divergence, ecological barriers are also likely prominent in Daphnia speciation. We assess the contribution of ecological and non-ecological reproductive isolating barriers between closely related species of Daphnia and found that none of the reproductive isolating barriers are capable of restricting gene flow completely. Research on reproductive isolating barriers has been disproportionally focused on two species complexes, Daphnia pulex and Daphnia longispina. Lastly, we identify areas of research that remain relatively unexplored and discuss future research directions that build our understanding of speciation in daphniids.

While the identification of microbial eukaryotes using molecular tools is now widespread, additional information are needed to confirm the molecular observation and make the difference between species and population variants, and therefore to better understand the biogeography of microbial eukaryotes. In this issue of Molecular Ecology, Postel et al (2020) not only used three molecular approaches to identify subgroups of Fragilariopsis kerguelensis but also morphology and physiology to better understand the relationship between the three genotypes. They revealed that (1) the three genotypes of the diatom F. kerguelensis have a negligible gene flux; and (2) two of the genotypes are geographically isolated with different physiology but still able to crossbreed; and (3) the last one is omnipresent but reproductively isolated.

Environmental DNA (eDNA) metabarcoding can rapidly characterize the composition and diversity of benthic communities. As such, it has high potential utility for routine environmental assessments of benthic impacts of marine finfish farming. In this study, 126 sediment grab samples from 42 stations were collected along an organic enrichment gradient at six salmon farms in British Columbia, Canada, and benthic biotic community changes were assessed by both eDNA metabarcoding of metazoans and macrofaunal polychaete surveys. The latter was done by analyzing 11,466 individuals using a combination of morpho-taxonomy and DNA barcoding. Study objectives were to: (1) compare biotic signals associated with benthic impacts of salmon farming in the two data types; and (2) identify potential eDNA indicators to facilitate eDNA-based monitoring in Canada. Across both data types, alpha diversity parameters were reduced in sediments near fish cage edge and were negatively correlated with pore-water sulphide concentration. Presence/absence of known indicator taxon Capitella generally agreed well between the two methods despite that they differed in both the volume of sediment sampled and the molecular marker used. In eDNA data, there was a strong negative correlation between Nematoda OTU richness and pore-water sulphide concentration, and multiple approaches were used to identify OTUs related to organic enrichment statuses. We demonstrate that eDNA metabarcoding generates biotic signals that could be leveraged for environmental assessment of benthic impacts of fish farms in multiple ways: both alpha diversity and Nematoda OTU richness could be used to assess the spatial extent of impact, and OTUs related to organic enrichment could be used to develop a local biotic index.

The evolutionary histories of alpine species are often directly associated with responses to glaciation. Deep divergence among populations and complex patterns of genetic variation have been inferred as consequences of persistence within glacier boundaries (i.e. on nunataks), while shallow divergence and limited genetic variation is assumed to result from expansion from large refugia at the edge of ice shields (i.e. massifs de refuge). However, for some species, dependence on specific microhabitats could profoundly influence their spatial and demographic response to glaciation, and such a simple dichotomy may obscure the localization of actual refugia. In this study, we use the Nebria ingens complex (Coleoptera: Carabidae), a water-affiliated ground beetle lineage, to test how drainage basins are linked to their observed population structure. By analyzing mitochondrial COI gene sequences and genome-wide single nucleotide polymorphisms, we find that the major drainage systems of the Sierra Nevada Mountains in California best explain the population structure of the N. ingens complex. In addition, we find that an intermediate morphotype within the N. ingens complex is the product of historical hybridization of N. riversi and N. ingens in the San Joaquin basin during glaciation. This study highlights the importance of considering ecological preferences in how species respond to climate fluctuations and provides an explanation for discordances that are often observed in comparative phylogeographic studies.

Clock genes exhibit substantial control over gene expression and ultimately life-histories using external cues such as photoperiod, and are thus likely to be critical for adaptation to shifting seasonal conditions and novel environments as species redistribute their ranges under climate change. Coding trinucleotide repeats (cTNRs) are found within several clock genes, and may be interesting targets of selection due to their containment within exonic regions and elevated mutation rates. Here, we conduct inter-specific characterization of the NR1D1 cTNR between Canada lynx and bobcat, and intra-specific spatial and environmental association analyses of neutral microsatellites and our functional cTNR marker, to investigate the role of selection on this locus in Canada lynx. We report signatures of divergent selection between lynx and bobcat, with the potential for hybrid-mediated gene flow in the area of range overlap. We also provide evidence that this locus is under selection across Canada lynx in eastern Canada, with both spatial and environmental variables significantly contributing to the explained variation, after controlling for neutral population structure. These results suggest that cTNRs may play an important role in the generation of functional diversity within some mammal species, and allow for contemporary rates of adaptation in wild populations in response to environmental change. We encourage continued investment into the study of cTNR markers to better understand their broader relevance to the evolution and adaptation of mammals.

Seashore paspalum (Paspalum vaginatum Swartz) is a halophytic turfgrass and emerging genomic model system for the study of salt tolerance in cereals and other grasses. Despite recent interest and an increase in available tools, little is known about the diversity present in wild populations of P. vaginatum and its close relative P. distichum. Variation in ploidy, clonal propagation, hybridization, and subgenome composition appear to occur in the wild and may interact to influence geographic patterns of adaptation, particularly in response to environmental salinity levels. Using 218 accessions representing >170 wild collections from throughout the coastal southern United States plus existing USDA germplasm, we employed genotyping-by-sequencing, cpDNA sequencing and flow cytometry to identify genetic differentiation and ploidy variation. Within P. vaginatum, there are two morphologically distinct ecotypes: the fine-textured ecotype is diploid and appears to reproduce in the wild both sexually and by clonal propagation; in contrast, the coarse-textured ecotype consists largely of clonally-propagating triploid and diploid genotypes. The coarse-textured ecotype appears to be derived from hybridization between fine-textured P. vaginatum and an unidentified Paspalum species. These clonally propagating hybrid genotypes are more broadly distributed than clonal fine-textured genotypes and may represent a transition to a more generalist adaptive strategy. The triploid genotypes vary in whether they carry one or two copies of the P. vaginatum subgenome, indicating multiple evolutionary origins. This variation in subgenome composition shows associations with local ocean salinity levels across the sampled populations and may play a role in local adaptation.

Populations of invasive species that colonize and spread in novel environments may differentiate both through demographic processes and local selection. European starlings (Sturnus vulgaris) were introduced to New York in 1890 and subsequently spread throughout North America, becoming one of the most widespread and numerous bird species on the continent. Genome-wide comparisons across starling individuals and populations can identify demographic and/or selective factors that facilitated this rapid and successful expansion. We investigated patterns of genomic diversity and differentiation using reduced-representation genome sequencing (ddRADseq) of 17 winter-season starling populations. Consistent with this species’ high dispersal rate and rapid expansion history, we found low geographic differentiation and few FST outliers even at a continental scale. Despite starting from a founding population of approximately 180 individuals, North American starlings show only a moderate genetic bottleneck, and models suggest a dramatic increase in effective population size since introduction. In genotype-environment associations we found that ~200 single-nucleotide polymorphisms are correlated with temperature and/or precipitation against a background of negligible genome- and range-wide divergence. Local adaptation in North American starlings may have evolved rapidly even in this wide-ranging and evolutionarily young population. This survey of genomic signatures of expansion in North American starlings is the most comprehensive to date and complements ongoing studies of world-wide local adaptation in these highly dispersive and invasive birds.

The repeated occurrence of similar phenotypes in independent lineages (i.e., parallel evolution) in response to similar ecological conditions can provide compelling insights into the process of adaptive evolution. An intriguing question is to what extent repeated phenotypic changes are underlain by repeated changes at the genomic level and whether patterns of genomic divergence differ with the geographic context in which populations evolve. Here, we combine genomic, morphological and ecological datasets to investigate the genomic signatures of divergence across populations of threespine stickleback (Gasterosteus aculeatus) that adapted to contrasting trophic niches (benthic or limnetic) in either sympatry or allopatry. We found that genome-wide differentiation (FST) was an order of magnitude higher and substantially more repeatable for sympatric benthic and limnetic specialists compared to allopatric populations with similar levels of trophic divergence. We identified 55 genomic regions consistently differentiated between sympatric ecotypes that were also associated with benthic vs. limnetic niche across allopatric populations. These candidate regions were enriched on three chromosomes known to be involved in the benthic-limnetic divergence of threespine stickleback. Some candidate regions overlapped with QTL for body shape and trophic traits such as number of gill rakers, traits that strongly differ between benthic and limnetic ecotypes. In sum, our study shows that magnitude and repeatability of genomic signatures of trophic divergence in threespine stickleback highly depend on the geographical context. The identified candidate regions provide starting points to identify functionally important genes for the adaptation to benthic and limnetic trophic niches.