Author Correction: Hybridization is a recurrent evolutionary stimulus in wild yeast speciation.

The original version of the Supplementary Information associated with this Article contained errors in Supplementary Figures 2, 12, 20 and 22. The HTML has been updated to include a corrected version of the Supplementary Information; the original incorrect versions of these Figures can be found as Supplementary Information associated with this Correction.

Hybridization is a recurrent evolutionary stimulus in wild yeast speciation.

Hybridization can result in reproductively isolated and phenotypically distinct lineages that evolve as independent hybrid species. How frequently hybridization leads to speciation remains largely unknown. Here we examine the potential recurrence of hybrid speciation in the wild yeast Saccharomyces paradoxus in North America, which comprises two endemic lineages SpB and SpC, and an incipient hybrid species, SpC*. Using whole-genome sequences from more than 300 strains, we uncover the hybrid origin of another group, SpD, that emerged from hybridization between SpC* and one of its parental species, the widespread SpB. We show that SpD has the potential to evolve as a novel hybrid species, because it displays phenotypic novelties that include an intermediate transcriptome profile, and partial reproductive isolation with its most abundant sympatric parental species, SpB. Our findings show that repetitive cycles of divergence and hybridization quickly generate diversity and reproductive isolation, providing the raw material for speciation by hybridization.

Persistence of Resident and Transplanted Genotypes of the Undomesticated Yeast Saccharomyces paradoxus in Forest Soil.

One might expect yeasts in soil to be highly dispersed via water or insects, forming ephemeral, genetically heterogeneous populations subject to competition and environmental stochasticity. Here, we report persistence of genotypes of the yeast Saccharomyces paradoxus in space and time. Within 1 km2 in a mixed hardwood forest on scales from centimeters to tens of meters, we detected persistence over 3 years of native genotypes, identified by single nucleotide polymorphisms (SNPs) genome-wide, of the wild yeast Saccharomyces paradoxus growing around Quercus rubra and Quercus alba Yeasts were recovered by enrichment in ethanol-containing medium, which measures only presence or absence, not abundance. Additional transplantation experiments employed strains marked with spontaneous defects in the URA3 gene, which also confer resistance to 5-fluoroorotic acid (5FOA). Plating soil suspensions from transplant sites on 5FOA-containing medium permitted one-step quantification of yeast CFU, with no interference from other unmarked yeasts or microorganisms. After an initial steep decrease in abundance, the yeast densities fluctuated over time, increasing in association with rainfall and decreasing in association with drought. After 18 months, the transplanted yeasts remained in place on the nine sites. In vitro transplantation experiments into nonsterile soil in petri dishes showed similar patterns of persistence and response to moisture and drought. To determine whether Saccharomyces cerevisiae, not previously recovered from soils regionally, can persist in our cold climate sites, we transplanted marked S. cerevisiae alone and in mixture with S. paradoxus in the fall of 2017. Five months later, S. cerevisiae persisted to the same extent as S. paradoxusIMPORTANCESaccharomyces yeasts are intensively studied in biological research and in their domesticated roles in brewing and baking, and yet, remarkably little is known about their mode of life in forest soils. We report here that resident genotypes of the yeast S. paradoxus are persistent on a time scale of years in their microhabitats in forest soils. We also show that resident genotypes can be replaced by transplanted yeast genotypes. The high inoculum levels in experimental transplantations rapidly decreased over time, but the transplanted genotypes persisted at low abundance. We conclude that, in forest soils, Saccharomyces yeasts exist at very low abundance and that dispersal events are rare.

Richard Paul Korf (1925-2016).

"With poetry, the tune is in the words themselves-and once you begin to hear it, it will stay with you." Richard P. Korf, notes to his narration of John Brown's Body.

Fungus Causing White-Nose Syndrome in Bats Accumulates Genetic Variability in North America with No Sign of Recombination.

Emerging fungal diseases of wildlife are on the rise worldwide, and the white-nose syndrome (WNS) epidemic in North American bats is a catastrophic example. The causal agent of WNS is a single clone of the fungus Pseudogymnoascus destructans. Early evolutionary change in this clonal population has major implications for disease ecology and conservation. Accumulation of variation in the fungus through mutation, and shuffling of variation through recombination, could affect the virulence and transmissibility of the fungus and the durability of what appears to be resistance arising in some bat populations. Our genome-wide analysis shows that the clonal population of P. destructans has expanded in size from a single genotype, has begun to accumulate variation through mutation, and presents no evidence as yet of genetic exchange among individuals. IMPORTANCE Since its discovery in 2006, the emerging infectious disease known as white-nose syndrome has killed millions of bats in North America, making it one of the most devastating wildlife epidemics in recorded history. We demonstrate that there has been as yet only spontaneous mutation across the North American population of P. destructans, and we find no indication of recombination. Thus, selective forces, which might otherwise impact pathogenic virulence, have so far had essentially no genetic variation on which to act. Our study confirmed the time of origin for the first and, thus far, only introduction of P. destructans to North America. This system provides an unprecedented opportunity to follow the evolution of a host-pathogen interaction unfolding in real time.

Population genomics reveals structure at the individual, host-tree scale and persistence of genotypic variants of the undomesticated yeast Saccharomyces paradoxus in a natural woodland.

Genetic diversity in experimental, domesticated and wild populations of the related yeasts, Saccharomyces cerevisiae and Saccharomyces paradoxus, has been well described at the global scale. We investigated the population genomics of a local population on a small spatial scale to address two main questions. First, is there genomic variation in a S. paradoxus population at a spatial scale spanning centimetres (microsites) to tens of metres? Second, does the distribution of genomic variants persist over time? Our sample consisted of 42 S. paradoxus strains from 2014 and 43 strains from 2015 collected from the same 72 microsites around four host trees (Quercus rubra and Quercus alba) within 1 km2 in a mixed hardwood forest in southern Ontario. Six additional S. paradoxus strains recovered from adjacent maple and beech trees in 2015 are also included in the sample. Whole-genome sequencing and genomic SNP analysis revealed five differentiated groups (clades) within the sampled area. The signal of persistence of genotypes in their microsites from 2014 to 2015 was highly significant. Isolates from the same tree tended to be more related than strains from different trees, with limited evidence of dispersal between trees. In growth assays, one genotype had a significantly longer lag phase than the other strains. Our results indicate that different clades coexist at fine spatial scale and that population structure persists over at least a one-year interval in these wild yeasts, suggesting the efficacy of yearly sampling to follow longer term genetic dynamics in future studies.

Genome Sequences of Three Phytopathogenic Species of the Magnaporthaceae Family of Fungi.

Magnaporthaceae is a family of ascomycetes that includes three fungi of great economic importance: Magnaporthe oryzae, Gaeumannomyces graminis var. tritici, and Magnaporthe poae. These three fungi cause widespread disease and loss in cereal and grass crops, including rice blast disease (M. oryzae), take-all disease in wheat and other grasses (G. graminis), and summer patch disease in turf grasses (M. poae). Here, we present the finished genome sequence for M. oryzae and draft sequences for M. poae and G. graminis var. tritici. We used multiple technologies to sequence and annotate the genomes of M. oryzae, M. poae, and G. graminis var. tritici. The M. oryzae genome is now finished to seven chromosomes whereas M. poae and G. graminis var. tritici are sequenced to 40.0× and 25.0× coverage respectively. Gene models were developed by the use of multiple computational techniques and further supported by RNAseq data. In addition, we performed preliminary analysis of genome architecture and repetitive element DNA.

The underlying structure of adaptation under strong selection in 12 experimental yeast populations.

The aims of this study were to determine (i) whether adaptation under strong selection occurred through mutations in a narrow target of one or a few nucleotide sites or a broad target of numerous sites and (ii) whether the programs of adaptation previously observed from three experimental populations were unique or shared among populations that underwent parallel evolution. We used archived population samples from a previous study, representing 500 generations of experimental evolution in 12 populations under strong selection, 6 populations in a high-salt environment and 6 populations in a low-glucose environment. Each set of six populations included four with sexual reproduction and two with exclusively asexual reproduction. Populations were sampled as resequenced genomes of 115 individuals and as bulk samples from which frequencies of mutant alleles were estimated. In a high-salt environment, a broad target of 11 mutations within the proton exporter, PMA1, was observed among the six populations, in addition to expansions of the ENA gene cluster. This pattern was shared among populations that underwent parallel evolution. In a low-glucose environment, two programs of adaptation were observed. The originally observed pattern of mutation in MDS3/MKT1 in population M8 was a narrow target of a single nucleotide, unique to this population. Among the other five populations, the three mutations were shared in a broad target, sensing/signaling genes RAS1 and RAS2. RAS1/RAS2 mutations were not observed in the high-salt populations; PMA1 mutations were observed only in a high-salt environment.

PRP8 inteins in species of the genus Botrytis and other ascomycetes.

The mobile elements termed inteins have a sporadic distribution in microorganisms. It is unclear how these elements are maintained. Inteins are intervening protein sequences that autocatalytically excise themselves from a precursor. Excision is a post-translational process referred to as 'protein splicing' in which the sequences flanking the intein are ligated, reforming the mature host protein. Some inteins contain a homing endonuclease domain (HEG) that is proposed to facilitate propagation of the intein element within a gene pool. We have previously demonstrated that the HEG of the PRP8 intein is highly active during meiosis in Botrytis cinerea. Here we analysed the Prp8 gene status in 21 additional Botrytis species to obtain insight into the mode of intein inheritance within the Botrytis lineage. Of the 21 species, 15 contained a PRP8 intein whereas six did not. The analysis was extended to closely related (Sclerotiniaceae) and distantly related (Ascomycota) taxa, focussing on evolutionary diversification of the PRP8 intein, including their possible acquisition by horizontal transfer and loss by deletion. Evidence was obtained for the occurrence of genetic footprints of previous intein occupation. There is no compelling evidence of horizontal transfer among species. Three distinct states of the Prp8 allele were identified, distributed over different orders within the Ascomycota: an occupied allele; an empty allele that was never occupied; an empty allele that was presumably previously occupied, from which the intein was precisely deleted. The presence of the genetic footprint identifies 20 species (including Neurospora crassa, Magnaporthe oryzae and Fusarium oxysporum) that previously contained the intein but have lost it entirely, while only 18 species (including Podospora anserina and Fusarium graminearum) appear never to have contained a PRP8 intein. The analysis indicates that inteins may be maintained in an equilibrium state.

Evidence for a common toolbox based on necrotrophy in a fungal lineage spanning necrotrophs, biotrophs, endophytes, host generalists and specialists.

The Sclerotiniaceae (Ascomycotina, Leotiomycetes) is a relatively recently evolved lineage of necrotrophic host generalists, and necrotrophic or biotrophic host specialists, some latent or symptomless. We hypothesized that they inherited a basic toolbox of genes for plant symbiosis from their common ancestor. Maintenance and evolutionary diversification of symbiosis could require selection on toolbox genes or on timing and magnitude of gene expression. The genes studied were chosen because their products have been previously investigated as pathogenicity factors in the Sclerotiniaceae. They encode proteins associated with cell wall degradation: acid protease 1 (acp1), aspartyl protease (asps), and polygalacturonases (pg1, pg3, pg5, pg6), and the oxalic acid (OA) pathway: a zinc finger transcription factor (pac1), and oxaloacetate acetylhydrolase (oah), catalyst in OA production, essential for full symptom production in Sclerotinia sclerotiorum. Site-specific likelihood analyses provided evidence for purifying selection in all 8 pathogenicity-related genes. Consistent with an evolutionary arms race model, positive selection was detected in 5 of 8 genes. Only generalists produced large, proliferating disease lesions on excised Arabidopsis thaliana leaves and oxalic acid by 72 hours in vitro. In planta expression of oah was 10-300 times greater among the necrotrophic host generalists than necrotrophic and biotrophic host specialists; pac1 was not differentially expressed. Ability to amplify 6/8 pathogenicity related genes and produce oxalic acid in all genera are consistent with the common toolbox hypothesis for this gene sample. That our data did not distinguish biotrophs from necrotrophs is consistent with 1) a common toolbox based on necrotrophy and 2) the most conservative interpretation of the 3-locus housekeeping gene phylogeny--a baseline of necrotrophy from which forms of biotrophy emerged at least twice. Early oah overexpression likely expands the host range of necrotrophic generalists in the Sclerotiniaceae, while specialists and biotrophs deploy oah, or other as-yet-unknown toolbox genes, differently.

Cellular effects and epistasis among three determinants of adaptation in experimental populations of Saccharomyces cerevisiae.

Epistatic interactions in which the phenotypic effect of an allele is conditional on its genetic background have been shown to play a central part in various evolutionary processes. In a previous study (J. B. Anderson et al., Curr. Biol. 20:1383-1388, 2010; J. R. Dettman, C. Sirjusingh, L. M. Kohn, and J. B. Anderson, Nature 447:585-588, 2007), beginning with a common ancestor, we identified three determinants of fitness as mutant alleles (each designated with the letter "e") that arose in replicate Saccharomyces cerevisiae populations propagated in two different environments, a low-glucose and a high-salt environment. In a low-glucose environment, MDS3e and MKT1e interacted positively to confer a fitness advantage. Also, PMA1e from a high-salt environment interacted negatively with MKT1e in a low-glucose environment, an example of a Dobzhansky-Muller incompatibility that confers reproductive isolation. Here we showed that the negative interaction between PMA1e and MKT1e is mediated by alterations in intracellular pH, while the positive interaction between MDS3e and MKT1e is mediated by changes in gene expression affecting glucose transporter genes. We specifically addressed the evolutionary significance of the positive interaction by showing that the presence of the MDS3 mutation is a necessary condition for the spread and fixation of the new mutations at the identical site in MKT1. The expected mutations in MKT1 rose to high frequencies in two of three experimental populations carrying MDS3e but not in any of three populations carrying the ancestral allele. These data show how positive and negative epistasis can contribute to adaptation and reproductive isolation.

Determinants of divergent adaptation and Dobzhansky-Muller interaction in experimental yeast populations.

Divergent adaptation can be associated with reproductive isolation in speciation [1]. We recently demonstrated the link between divergent adaptation and the onset of reproductive isolation in experimental populations of the yeast Saccharomyces cerevisiae evolved from a single progenitor in either a high-salt or a low-glucose environment [2]. Here, whole-genome resequencing and comparative genome hybridization of representatives of three populations revealed 17 mutations, six of which explained the adaptive increases in mitotic fitness. In two populations evolved in high salt, two different mutations occurred in the proton efflux pump gene PMA1 and the global transcriptional repressor gene CYC8; the ENA genes encoding sodium efflux pumps were overexpressed once through expansion of this gene cluster and once because of mutation in the regulator CYC8. In the population from low glucose, one mutation occurred in MDS3, which modulates growth at high pH, and one in MKT1, a global regulator of mRNAs encoding mitochondrial proteins, the latter recapitulating a naturally occurring variant. A Dobzhansky-Muller (DM) incompatibility between the evolved alleles of PMA1 and MKT1 strongly depressed fitness in the low-glucose environment. This DM interaction is the first reported between experimentally evolved alleles of known genes and shows how reproductive isolation can arise rapidly when divergent selection is strong.

Genome-wide investigation of reproductive isolation in experimental lineages and natural species of Neurospora: identifying candidate regions by microarray-based genotyping and mapping.

Inherent incompatibilities between genetic components from genomes of different species may cause intrinsic reproductive isolation. In evolution experiments designed to instigate speciation in laboratory populations of the filamentous fungus Neurospora, we previously discovered a pair of incompatibility loci (dfe and dma) that interact negatively to cause severe defects in sexual reproduction. Here we show that the dfe-dma incompatibility also is a significant cause of genetic isolation between two naturally occurring species of Neurospora (N. crassa and N. intermedia). The strong incompatibility interaction has a simple genetic basis (two biallelic loci) and antagonistic epistasis occurs between heterospecific alleles only, consistent with the Dobzhansky-Muller model of genic incompatibility. We developed microarray-based, restriction-site associated DNA (RAD) markers that identified approximately 1500 polymorphisms between the genomes of the two species, and constructed the first interspecific physical map of Neurospora. With this new mapping resource, the approximate genomic locations of the incompatibility loci were determined using three different approaches: genome scanning, bulk-segregant analyses, and introgression. These population, quantitative, and classical genetics methods concordantly identified two candidate regions, narrowing the search for each incompatibility locus to only approximately 2% of the nuclear genome. This study demonstrates how advances in high-throughput, genome-wide genotyping can be applied to mapping reproductive isolation genes and speciation research.

Exploring the evolutionary ecology of fungal endophytes in agricultural systems: using functional traits to reveal mechanisms in community processes.

All plants, including crop species, harbor a community of fungal endophyte species, yet we know little about the biotic factors that are important in endophyte community assembly. We suggest that the most direct route to understanding the mechanisms underlying community assembly is through the study of functional trait variation in the host and its fungal consortium. We review studies on crop endophytes that investigate plant and fungal traits likely to be important in endophyte community processes. We focus on approaches that could speed detection of general trends in endophyte community assembly: (i) use of the 'assembly rules' concept to identify specific mechanisms that influence endophyte community dynamics, (ii) measurement of functional trait variation in plants and fungi to better understand endophyte community processes and plant-fungal interactions, and (iii) investigation of microbe-microbe interactions, and fungal traits that mediate them. This approach is well suited for research in agricultural systems, where pair-wise host-fungus interactions and mechanisms of fungal-fungal competition have frequently been described. Areas for consideration include the possibility that human manipulation of crop phenotype and deployment of fungal biocontrol species can significantly influence endophyte community assembly. Evaluation of endophyte assembly rules may help to fine-tune crop management strategies.

Single nucleotide polymorphism-based diagnostic system for crop-associated Sclerotinia species.

A molecular diagnostic system using single nucleotide polymorphisms (SNPs) was developed to identify four Sclerotinia species: S. sclerotiorum (Lib.) de Bary, S. minor Jagger, S. trifoliorum Erikss., and the undescribed species Sclerotinia species 1. DNAs of samples are hybridized with each of five 15-bp oligonucleotide probes containing an SNP site midsequence unique to each species. For additional verification, hybridizations were performed using diagnostic single nucleotide substitutions at a 17-bp sequence of the calmodulin locus. The accuracy of these procedures was compared to that of a restriction fragment length polymorphism (RFLP) method based on Southern hybridizations of EcoRI-digested genomic DNA probed with the ribosomal DNA-containing plasmid probe pMF2, previously shown to differentiate S. sclerotiorum, S. minor, and S. trifoliorum. The efficiency of the SNP-based assay as a diagnostic test was evaluated in a blind screening of 48 Sclerotinia isolates from agricultural and wild hosts. One isolate of Botrytis cinerea was used as a negative control. The SNP-based assay accurately identified 96% of Sclerotinia isolates and could be performed faster than RFLP profiling using pMF2. This method shows promise for accurate, high-throughput species identification.

Divergent adaptation promotes reproductive isolation among experimental populations of the filamentous fungus Neurospora.

BACKGROUND: An open, focal issue in evolutionary biology is how reproductive isolation and speciation are initiated; elucidation of mechanisms with empirical evidence has lagged behind theory. Under ecological speciation, reproductive isolation between populations is predicted to evolve incidentally as a by-product of adaptation to divergent environments. The increased genetic diversity associated with interspecific hybridization has also been theorized to promote the development of reproductive isolation among independent populations. Using the fungal model Neurospora, we founded experimental lineages from both intra- and interspecific crosses, and evolved them in one of two sub-optimal, selective environments. We then measured the influence that initial genetic diversity and the direction of selection (parallel versus divergent) had on the evolution of reproductive isolation. RESULTS: When assayed in the selective environment in which they were evolved, lineages typically had greater asexual fitness than the progenitors and the lineages that were evolved in the alternate, selective environment. Assays for reproductive isolation showed that matings between lineages that were adapted to the same environment had greater sexual reproductive success than matings between lineages that were adapted to different environments. Evidence of this differential reproductive success was observed at two stages of the sexual cycle. For one of the two observed incompatibility phenotypes, results from genetic analyses were consistent with a two-locus, two-allele model with asymmetric (gender-specific), antagonistic epistasis. The effects of divergent adaptation on reproductive isolation were more pronounced for populations with greater initial genetic variation. CONCLUSION: Divergent selection resulted in divergent adaptation and environmental specialization, consistent with fixation of different alleles in different environments. When brought together by mating, these alleles interacted negatively and had detrimental effects on sexual reproductive success, in agreement with the Dobzhansky-Muller model of genetic incompatibilities. As predicted by ecological speciation, greater reproductive isolation was observed among divergent-adapted lineages than among parallel-adapted lineages. These results support that, given adequate standing genetic variation, divergent adaptation can indirectly cause the evolution of reproductive isolation, and eventually lead to speciation.

Marker stability throughout 400 days of in vitro hyphal growth in the filamentous ascomycete, Sclerotinia sclerotiorum.

The stability of routinely used, population genetic markers through approximately 1 year of continuous laboratory growth was investigated in the common, plant pathogentic ascomycete Sclerotinia sclerotiorum. Given reports of accelerated mutation rates at higher temperatures, both a permissive temperature, 22 degrees C, and a temperature at the high end of tolerance, 30 degrees C, were employed. Because mycelial growth rate was tracked among mitotic lineages established for each strain, a subsidiary objective was addressed, testing the stability of a 30 degrees C-competent phenotype. Twelve laboratory strains of S. sclerotiorum, including the genome sequence isolate, 1980, were propagated serially for up to 400 days at 22 degrees C. Five of these strains were also propagated at 30 degrees C. No mutations were observed in mycelial compatibility groupings (MCGs), DNA fingerprints, alleles at 7 microsatellite loci, or alleles at 56 AFLP loci. All of these markers show variation in field populations, which are likely much larger and influenced by different and more stochastic environmental processes. In S. sclerotiorum, population genetic markers were stable over time through serial transfer and growth of laboratory strains at both 22 degrees C and 30 degrees C. The strain isolated after extended drought and capable of infecting plants at 28 degrees C demonstrated the stability of its high temperature-competent phenotype, in addition to its stable growth rate at 22 degrees C. This observation has implications for modeling pathogen tolerance or adaptation under conditions of environmental stochasticity, including climate warming.

Host-synthesized secondary compounds influence the in vitro interactions between fungal endophytes of maize.

Maize produces a suite of allelopathic secondary metabolites, the benzoxazinoids. 2,4-Dihydroxy-7-methoxy-2H-1,4-benzoxazin-3-one and 2,4-dihydroxy-2H-1,4-benzoxazin-3-one reside as glucosides in plant tissue and spontaneously degrade to 6-methoxy-2-benzoxazolinone (MBOA) and 2-benzoxazolinone (BOA) upon plant cell disruption. Several maize-associated fungi in the genus Fusarium can metabolize MBOA and BOA. BOA tolerance levels in 10 species of Fusarium and in the maize endophytes Nigrospora oryzae, Acremonium zeae, and Periconia macrospinosa were characterized. BOA tolerance ranged from 0.25 to 1.10 mg/ml among species. The influence of substrate alteration by one species on the subsequent growth of another species was assessed in the presence and absence of BOA. The colony area of the secondary colonizer in heterospecific interactions was compared to that in autospecific interactions (one isolate follows itself). In the presence of BOA, four of six secondary colonizers had greater growth (facilitation) when primary colonizers had higher BOA tolerance than the secondary colonizer. When the primary colonizer had lower tolerance than the secondary, three of six secondary colonizers were inhibited (competition) and three not significantly affected. In BOA-free medium, the number of isolates that were facilitated or inhibited was the same regardless of the tolerance level of the primary colonizer. Two of six secondary colonizers were facilitated, two inhibited, and two not significantly affected. This study provides some support for facilitation in stressful conditions under the Menge-Sutherland model. The results are not consistent with the corresponding prediction of competition in the absence of stress. The hypothesis drawn from these data is that in the presence of a toxin, fungal species that detoxify their substrate can enhance the colonization rate of less tolerant fungi.

Incipient speciation by divergent adaptation and antagonistic epistasis in yeast.

Establishing the conditions that promote the evolution of reproductive isolation and speciation has long been a goal in evolutionary biology. In ecological speciation, reproductive isolation between populations evolves as a by-product of divergent selection and the resulting environment-specific adaptations. The leading genetic model of reproductive isolation predicts that hybrid inferiority is caused by antagonistic epistasis between incompatible alleles at interacting loci. The fundamental link between divergent adaptation and reproductive isolation through genetic incompatibilities has been predicted, but has not been directly demonstrated experimentally. Here we empirically tested key predictions of speciation theory by evolving the initial stages of speciation in experimental populations of the yeast Saccharomyces cerevisiae. After replicate populations adapted to two divergent environments, we consistently observed the evolution of two forms of postzygotic isolation in hybrids: reduced rate of mitotic reproduction and reduced efficiency of meiotic reproduction. This divergent selection resulted in greater reproductive isolation than parallel selection, as predicted by the ecological speciation theory. Our experimental system allowed controlled comparison of the relative importance of ecological and genetic isolation, and we demonstrated that hybrid inferiority can be ecological and/or genetic in basis. Overall, our results show that adaptation to divergent environments promotes the evolution of reproductive isolation through antagonistic epistasis, providing evidence of a plausible common avenue to speciation and adaptive radiation in nature.