Introgression of domesticated salmon changes life history and phenology of a wild salmon population

Introgression from local cultivars is a driver of agricultural adaptation in Argentinian weedy rice

Weedy rice, a pervasive and troublesome weed found across the globe, has often evolved through fertilization of rice cultivars with little importance of crop-weed gene flow. In Argentina, weedy rice has been reported as an important constraint since the early 1970s, and, in the last few years, strains with herbicide-resistance are suspected to evolve. Despite their importance, the origin and genetic composition of Argentinian weedy rice as well its adaptation to agricultural environments has not been explored so far. To study this, we conducted genotyping-by-sequencing on samples of Argentinian weedy and cultivated rice and compared them with published data from weedy, cultivated and wild rice accessions distributed worldwide. In addition, we conducted a phenotypic characterization for weedy-related traits, a herbicide resistance screening and genotyped accessions for known mutations in the acetolactate synthase (ALS) gene, which confers herbicide resistance. Our results revealed large phenotypic variability in Argentinian weedy rice. Most strains were resistant to ALS-inhibiting herbicides with a high frequency of the ALS mutation (A122T) present in Argentinian rice cultivars. Argentinian cultivars belonged to the three major genetic groups of rice: japonica, indica and aus while weeds were mostly aus or aus-indica admixed, resembling weedy rice strains from the Southern Cone region. Phylogenetic analysis supports a single origin for aus-like South American weeds, likely as seed contaminants from the United States, and then admixture with local indica cultivars. Our findings demonstrate that crop to weed introgression can facilitate rapid adaptation to agriculture environments.

Genetic analyses verify sexually mature escaped farmed Atlantic cod and farmed cod eggs in the natural environment

Elucidating the effects of domesticated organisms escaping into the natural environment represents a topic of importance in both evolutionary and conservation biology. However, when excluding the abundant data on salmonids, there is a lack of knowledge on this topic for marine fish aquaculture, which continues to expand globally. In order to bridge this empirical gap, we investigated a suspected escape of sexually mature domesticated Atlantic cod from a commercial marine fish farm in northern Norway. This involved genotyping samples of fish from cages on the farm, putatively identified escapees and wild cod captured in the region and samples of recently spawned eggs collected in the sea. Genetic analyses confirmed a farmed ancestry of the suspected escapees, and significantly, 27% of the sampled cod eggs. Furthermore, statistical analyses revealed a strong reduction in genetic variation in all samples of the farmed cod, including low effective population size and high degree of siblingship. These results thus document the escape of sexually mature adult cod and the release of fertilized domesticated cod eggs into the natural environment. Although it is possible that some of the mature escapees spawned post-escape, the fact that only a single egg of potential hybrid farmed × wild origin was identified, together with the high number of mature cod in the farm, points to within cage spawning as the primary source of these eggs. This suggestion is supported by oceanic particle-drift modelling, verifying that transport of eggs between the farm and the egg sampling locations was plausible. This study represents a rare documentation of interaction between domesticated and wild populations for a marine fish, pointing towards potential impacts on the local wild population.

Most of the escaped farmed salmon entering a river during a 5-year period were infected with one or more viruses

Disease interactions between farmed and wild populations have been poorly documented for most aquaculture species, in part due to the complexities to study this. Here, we tested 567 farmed Atlantic salmon escapees, captured in a Norwegian river during 2014-2018, for five viral infections that are prevalent in global salmonid aquaculture. Over 90% of the escapees were infected with one or more viruses. Overall prevalences were: 75.7% for piscine orthoreovirus (PRV-1), 43.6% for salmonid alphavirus (SAV), 31.2% for piscine myocarditis virus (PMCV), 1.2% for infectious pancreatic necrosis virus (IPNV) and 0.4% for salmon anaemia virus (ISAV). A significantly higher prevalence of PMCV infection was observed in immature compared to mature individuals. The prevalence of both SAV and PMCV infections was higher in fish determined by fatty acid profiling to be 'recent' as opposed to 'early' escapees that had been in the wild for a longer period of time. This is the first study to establish a time-series of viral infection status of escapees entering a river with a native salmon population. Our results demonstrate that farmed escapees represent a continuous source of infectious agents which could potentially be transmitted to wild fish populations.

Genomic evidence for domestication selection in three hatchery populations of Chinook salmon, Oncorhynchus tshawytscha

Fish hatcheries are widely used to enhance fisheries and supplement declining wild populations. However, substantial evidence suggests that hatchery fish are subject to differential selection pressures compared to their wild counterparts. Domestication selection, or adaptation to the hatchery environment, poses a risk to wild populations if traits specific to success in the hatchery environment have a genetic component and there is subsequent introgression between hatchery and wild fish. Few studies have investigated domestication selection in hatcheries on a genomic level, and even fewer have done so in parallel across multiple hatchery-wild population pairs. In this study, we used low-coverage whole-genome sequencing to investigate signals of domestication selection in three separate hatchery populations of Chinook salmon, Oncorhynchus tshawytscha, after approximately seven generations of divergence from their corresponding wild progenitor populations. We sequenced 192 individuals from populations across Southeast Alaska and estimated genotype likelihoods at over six million loci. We discovered a total of 14 outlier peaks displaying high genetic differentiation (F ST) between hatchery-wild pairs, although no peaks were shared across the three comparisons. Peaks were small (53 kb on average) and often displayed elevated absolute genetic divergence (D xy) and linkage disequilibrium, suggesting some level of domestication selection has occurred. Our study provides evidence that domestication selection can lead to genetic differences between hatchery and wild populations in only a few generations. Additionally, our data suggest that population-specific adaptation to hatchery environments likely occurs through different genetic pathways, even for populations with similar standing genetic variation. These results highlight the need to collect paired genotype-phenotype data to understand how domestication may be affecting fitness and to identify potential management practices that may mitigate genetic risks despite multiple pathways of domestication.

Overruled by nature: A plastic response to environmental change disconnects a gene and its trait

In Atlantic salmon, age at maturation is a life history trait governed by a sex-specific trade-off between reproductive success and survival. Following environmental changes across large areas of the Northeast Atlantic, many populations currently display smaller size at age and higher age at maturation. However, whether these changes reflect rapid evolution or plasticity is unknown. Approximately 1500 historical and contemporary salmon from the river Etne in Western Norway, genotyped at 50,000 SNPs, revealed three loci associated with age at maturation. These included vgll3 and six6 which collectively explained 36%-50% of the age at maturation variation in the 1983-1984 period. These two loci also displayed sex-specific epistasis, as the effect of six6 was only detected in males bearing two copies of the late maturation allele for vgll3. Strikingly, despite allelic frequencies at vgll3 remaining unchanged, the combined influence of these genes was nearly absent in all samples from 2013 to 2016, and genome-wide heritability strongly declined between the two time-points. The difference in age at maturation between males and females was upheld in the population despite the loss of effect from the candidate loci, which strongly points towards additional causative mechanisms resolving the sexual conflict. Finally, because admixture with farmed escaped salmon was excluded as the origin of the observed disconnection between gene(s) and maturation age, we conclude that the environmental changes observed in the North Atlantic during the past decades have led to bypassing of the influence of vgll3 and six6 on maturation through growth-driven plasticity.

Fine-scale environmentally associated spatial structure of lumpfish (Cyclopterus lumpus) across the Northwest Atlantic

Lumpfish, Cyclopterus lumpus, have historically been harvested throughout Atlantic Canada and are increasingly in demand as a solution to controlling sea lice in Atlantic salmon farms-a process which involves both the domestication and the transfer of lumpfish between geographic regions. At present, little is known regarding population structure and diversity of wild lumpfish in Atlantic Canada, limiting attempts to assess the potential impacts of escaped lumpfish individuals from salmon pens on currently at-risk wild populations. Here, we characterize the spatial population structure and genomic-environmental associations of wild populations of lumpfish throughout the Northwest Atlantic using both 70K SNP array data and whole-genome re-sequencing data (WGS). At broad spatial scales, our results reveal a large environmentally associated genetic break between the southern populations (Gulf of Maine and Bay of Fundy) and northern populations (Newfoundland and the Gulf of St. Lawrence), linked to variation in ocean temperature and ice cover. At finer spatial scales, evidence of population structure was also evident in a distinct coastal group in Newfoundland and significant isolation by distance across the northern region. Both evidence of consistent environmental associations and elevated genome-wide variation in F ST values among these three regional groups supports their biological relevance. This study represents the first extensive description of population structure of lumpfish in Atlantic Canada, revealing evidence of broad and fine geographic scale environmentally associated genomic diversity. Our results will facilitate the commercial use of lumpfish as a cleaner fish in Atlantic salmon aquaculture, the identification of lumpfish escapees, and the delineation of conservation units of this at-risk species throughout Atlantic Canada.

Genomic dynamics of brown trout populations released to a novel environment

Population translocations occur for a variety of reasons, from displacement due to climate change to human‐induced transfers. Such actions have adverse effects on genetic variation and understanding their microevolutionary consequences requires monitoring. Here, we return to an experimental release of brown trout (Salmo trutta) in order to monitor the genomic effects of population translocations. In 1979, fish from each of two genetically (F_ST = 0.16) and ecologically separate populations were simultaneously released, at one point in time, to a lake system previously void of brown trout. Here, whole‐genome sequencing of pooled DNA (Pool‐seq) is used to characterize diversity within and divergence between the introduced populations and fish inhabiting two lakes downstream of the release sites, sampled 30 years later (c. 5 generations). Present results suggest that while extensive hybridization has occurred, the two introduced populations are unequally represented in the lakes downstream of the release sites. One population, which is ecologically resident in its original habitat, mainly contributes to the lake closest to the release site. The other population, migratory in its natal habitat, is genetically more represented in the lake further downstream. Genomic regions putatively under directional selection in the new habitat are identified, where allele frequencies in both established populations are more similar to the introduced population stemming from a resident population than the migratory one. Results suggest that the microevolutionary consequences of population translocations, for example, hybridization and adaptation, can be rapid and that Pool‐seq can be used as an initial tool to monitor genome‐wide effects.