Development of optimal molecular markers of domestication in Atlantic salmon for assessing introgression in wild populations

Controversy surrounds the actual impacts of Atlantic salmon farming on wild salmonid stocks, fed by the lack of direct evidence for or against many potential impacts, with uncertainty an increasing impediment to sustainable industry development and effective management of wild stocks. This applies to the potential impact of the introgression of farm genomes into locally adapted wild populations from breeding of farm escapes. Escapes do occur and are recognized as inevitable, but are a very small fraction of farm stocks and vary in numbers both locally and temporally. The majority of escapees are expected to die without breeding but some do remain in or ascend rivers and spawn. However, a detailed understanding of actual levels of interbreeding and introgression in most rivers is lacking which, along with an understanding of the adaptive differentiation of farm and wild salmon, is required to establish the actual impact of this potential interaction on the productivity and viability of wild populations.

Detection and quantification of interbreeding and introgression requires diagnostic markers for farm and wild genomes. Genetic differentiation of farm and wild genomes can evolve through founder effects, selective breeding and domestication selection and is observed in respect of a variety of molecular markers. However, existing molecular markers are not fully diagnostic and regionally constrained in their usefulness. Unfortunately, marker panels screened for useful variation have been small and arbitrary such that they are unlikely to include the most informative loci and to be context specific, limiting their power and transferability. To properly address the issue of introgression molecular markers are required that are highly diagnostic across all farm and wild populations. These markers will be in genomic regions involved in domestication and controlling the expression of selected economic traits.

What is known of the genomic architecture of domestication and most economic traits indicates their control is polygenic, making the targeting of specific gene regions in the search for markers difficult. In contrast, recent advances in genomics make possible genome scanning and genome-wide association studies (GWAS) which can provide a high resolution assessment of molecular differentiation between different individuals or populations across the genome. Different GWAS strategies can be employed but two are deemed optimal in the current context. Firstly, a GWAS will be carried out using a new Atlantic salmon SNP (single nucleotide polymorphism) containing 930k nuclear SNPs, recently developed in collaboration with the salmon farming industry. This will be carried out on a broad base of representative farm and wild stocks. Secondly, GWAS will be carried out to identify temporally stable epigenetic DNA-methylation base changes induced by rearing fish in culture by comparing groups of single source wild fish reared in the wild and in culture.

The study will deliver the first general understanding of domestication related molecular genetic differentiation between farmed and wild salmon and identify the best markers for identifying farm salmon in the wild and assessing genetic introgression of farm genes into wild populations. The work will deliver a more robust and generally applicable tool for determining the actual levels of escapes and introgression in wild salmon populations. Following field calibration and independent validation, the diagnostic methodology defined in the study is expected to provide the basis for generating the evidence needed to clarify the debate on levels of escapes and introgression and the long term impacts of introgression on population viability. This will help to define more clearly the path forward for the sustainable development of the salmon farming industry in the UK and elsewhere in the North Atlantic region and help to inform management priorities for wild Atlantic salmon stocks.

The proposed research programme will survey direct DNA base variation and epigenetic DNA base methylation changes in farmed Atlantic salmon strains and wild stocks to identify diagnostic heritable differences. Replicated pooled samples (50 individuals in each pool) will be used to screen 12 different farm strains and 12 different wild stocks from Scotland, Ireland, Norway and Canada, spanning the range of Atlantic salmon, will be screened on a 930,000 Affymetrix SNP array. Estimates of allele frequencies for each SNP will be made for pools based on normalised theta values of the cluster positions of the different genotypes and differences will be assessed across all farmed-wild comparisons as well as similarity in the directionality of differences among farmed strains. Coupled with tests for 'outlier' loci, this will be used to identify the top-ranked SNPs for discrimination. The genomic positions of these SNPs will be mapped using existing linkage information to identify regions associated with domestication. The top 300 ranked SNPs will be further screened for individual genotypes on 300 fish from across the study to validate allele frequency estimates derived from pooled samples. Epigenetic analysis will focus on identifying stable DNA methylation changes induced by rearing fish in culture. Eggs from different wild-origin families will be divided across two experimental groups reared under captive or semi-natural conditions. Male parr from the two groups will be screened for maturation stage and sperm collected for both methylation analyses and egg fertilisation according to a fully factorial breeding design. Embryos from each family will be sampled at an identical developmental stage and analysed for methylation status. Reduced Representation Bisulfate Sequencing on an Illumina HiSeq will be used, to analyse both the sperm of the two experimental groups as well as in embryos fertilised by the two types of sperm (wild and cultured).

The findings will have impact on our understanding of our capacity to monitor and manage direct genetic interactions, and will establish a major step change towards the development of the molecular markers and analytical tools available to managers and regulators. It will be of great value to the regulatory and non-academic community in providing a more robuts appreciation of the nature and extent of the genetic differentiation underlying the adaptive differences between wild and domesticated stocks of Atlantic salmon. The stakeholder community in the UK that will make use of this information to inform the management of wild stocks will be the North Atlantic Salmon Conservation Organisation (NASCO), the Atlantic Salmon Trust (AST), the Rivers and Fisheries Trusts Scotland (RAFTS), the Rivers Trust (England); it is these organisation whom will be targeted for delivery of a non-technical overview of project findings and with whom opportunities will be explored to make public presentations of the science carried out and its findings. Each of these organisations is concerned with the welfare and health of wild salmon populations and engages with scientific bodies, academic communities and the aquaculture industry. Furthermore, government agencies such as Marine Scotland, the Department of Fisheries and Oceans (Canada) and other regulatory bodies in other salmon or aquaculture countries will be end-users of the new understanding and the identified tools to assist them in taking management decisions involving the conservation of wild stocks as well as the regulation of aquaculture industry development. Partners in the proposed project cover the scientific branches of government agencies (in addition to academic partners), thereby creating direct links to their relevant regulatory bodies. By furthering our understanding of the occurrence and degree of genetic interactions between farmed and wild salmon stocks, this work will help to progress the dialogue between wild fish interests and those of the aquaculture industry. It is only by understanding the degree to which genetic interactions among farmed and wild salmon may pose a risk that measures can be taken to ensure both the continued conservation of wild stocks, while allowing for increased expansion of the industry, both of which offer great economic resources to many regions. At the same time, the project results can be exploited by the aquaculture industry with regards to using the information to improve the performance in culture of farmed stocks by helping to define the regions where genes have been selected for to improve overall performance. Thus the findings of the project will also be of interest and communicated to the salmon farming industry through their trade organisation the SSPO whom will be included in the Stakeholder Workshop that will be held at the end of the project.