Community Research and Development Information Service - CORDIS

Periodic Report Summary 3 - FISHBOOST (Improving European aquaculture by advancing selective breeding to the next level for the six main finfish species)

Project Context and Objectives:
Only about 10% of today’s global aquaculture production use genetically improved stocks. In Europe, some breeding programmes consist of only the basic components of a breeding scheme. Hence, there is large potential to increase efficiency and profit by domestication and genetic improvement of farmed finfish. The main challenge of FISHBOOST is to realise this potential into economic and social acceptable breeding schemes, and advance these for each of the six target species. Acknowledging the different capacities of the species, the aim of FISHBOOST is: ‘To improve the efficiency and profitability of European aquaculture by advancing selective breeding to the next level for each of the six main finfish species through collaborative research with industry’. FISHBOOST considers the main components of breeding programmes for Atlantic salmon, common carp, European seabass, gilthead seabream, rainbow trout and turbot. 14 well-recognised RTD participants in Europe on aquaculture breeding collaborate from February 2014 until January 2019 in this comprehensive research project with 7 SMEs, 4 large industries and 1 NGO throughout Europe. These partners are in the lead of the development of their species’ breeding programmes or are vectors between industry and RTD. A mixture of low- and high-technological advances depending on current capacities of the species is being developed to move each species’ breeding program to the next level.
Project Results:
WP1 studies the genetics behind the resistance of the six finfish species to specific diseases. Large-scale and powerful disease challenge experiments have been performed for rainbow trout, seabass, seabream (x2), carp and turbot. All of these experiments resulted in informative disease resistance measures. This data was used in one of the largest co-ordinated genotyping-by-sequencing efforts (RAD Sequencing) of aquaculture species in the world to date to successfully generate genome-wide genetic marker data for disease challenged fish. Individual fish samples were genotyped using various RAD sequencing approaches (RAD-Seq, ddRAD-Seq, 2bRAD-Seq). In all cases, the genotyping resulted in population-level, genome-wide SNP marker data for the samples. These results have highlighted the utility of RAD techniques for cost-effective genotyping-by sequencing in aquaculture species, and may lead to wider scale implementation of genomics into aquaculture breeding programmes. Data and samples from the disease challenge and RAD-sequencing experiments have been used to estimate genetic parameters for resistance, detect and validate genetic markers affecting resistance, calculate genomic predictions and will, ultimately, apply the results in selective breeding for resistance to disease in each of these species. WP2 is about production efficiency traits. In FISHBOOST a lot of effort is put on getting indicator traits for production efficiency. The recording of traits directly on the selection candidates makes the selection more efficient. Focus has also been on winter growth, which is an important trait for survival and production efficiency. A method has been developed to study genetic effects of simultaneous changes of several ingredients in feed trials. WP3 is about genomic tools, and here the performance and design of genomic selection breeding schemes has been tested, and the use of genomic breeding values was found beneficial in all studied cases. A software pipeline to calculate genomic breeding values has been developed. Models accounting for dominance effects have also been tested. WP6 performs economic assessment of breeding programs. A survey among producers showed how many breeding programs exist per species and revealed the main characteristics of these breeding programs. Market shares of breeding companies per species were: 65-68% for trout, 93-95% for salmon, 43-56% for seabass, 60-66 for seabream, and 100% for turbot, showing that selective breeding plays an important role in European aquaculture. Bioeconomical models for the derivation of economic values for growth, feed intake, mortality and uniformity have been developed to quantify the importance of these traits for the European aquaculture breeding industry, and cost benefit analyses of the traits and sectors are now being conducted. WP5 is about producer perceptions of aquaculture breeding. Interviews and a survey among producers and representative organisations were used to identify positive points, concerns and expected sensitive areas of aquaculture breeding for producers and advantages and challenges for the aquaculture producers regarding implementation of breeding practises, existing and novel breeding methods and in the importance of different traits. Results from all other WPs are now used as input in WP4 to optimise specific parts of the breeding programmes for the six species. Biological and technological specificities and constraints of each species are accounted for in this optimisation. Software to optimise the selection of next generations parents has been developed for the breeding industry.
Potential Impact:
The most important impacts of FISHBOOST arise when commercial fish breeders integrate new breeding technologies and adopt accurate measurements of new traits to attain genetic improvement. Thereby, they take the development of the species’ breeding programs to the next level. Optimised breeding designs, including economic assessment of these, and genomics tools for e.g. improving accuracy of selection for the six species are developed. FISHBOOST addresses two economically and sustainably important groups of traits that are difficult to improve genetically: disease resistance and production efficiency. FISHBOOST investigates the societal views on acceptable systems and methods of aquaculture breeding. Higher productivity, through more efficient fish, may lead to reduced prices of aquaculture products for the consumers. Reduced mortality in aquaculture production through selection for disease resistance reduces impact on wild fish and increases health and welfare of the farmed fish, thus alleviates societal concerns.
Scientific, technological, commercial, social and environmental impact of FISHBOOST:
• Estimates of heritability have been calculated for diseases in common carp (KHV), European seabass (VNN), pasteurellosis and Sparicotyle (gilthead seabream), Philasterides (turbot). Encouragingly, the estimates of heritability were universally significant, and often moderate in size. The DNA samples collected from these successful disease challenge experiment have been sequenced and genotyped using RAD sequencing in large-scale sequencing experiments. These datasets were combined to map individual loci (QTL) affecting resistance to the diseases. Genome-wide significant QTL were detected for every experiment, highlighting the adequate statistical power of the disease challenge experiments and subsequent genotyping by sequencing experiments. The QTL identified explain between 12 and 69 % of the genetic variation in the resistance to the diseases under study. While this is a significant proportion of the variation, it does assume independent and additive effects of the QTL, and it should be noted that QTL effects are typically overestimated due to a phenomenon known as the Beavis effect. In addition to the QTL results, the same datasets have enabled new genomic tools and resources to be created for these species, including SNP marker databases and genetic linkage maps. These results can be taken up by the breeding industry. (WP1).
• A genetic epidemiology study on turbot, using the parasite Philasterides, showed significant genetic variation in different components of the disease resistance phenotypes, including resistance, tolerance, resilience and infectivity. Results have broad implications for understanding genetic epidemiology in aquaculture. (WP1)
• One major goal for sustainable aquaculture operations is to improve fish welfare by developing and implementing non-lethal methods to record economically important fish traits normally requiring slaughtering of the fish. Lipid% was one of the potential indicators of feed efficiency and it is also an important product quality trait. The use of a non-lethal method to record fillet lipid% was validated in gilthead seabream, common carp and rainbow trout. This method was implemented in an industry-scale recording in five out of the six species. This method addresses the 3R issue. (WP2)
• In rainbow trout, feed efficiency can be improved via simultaneous selection for rapid growth while holding the lipid percentage constant. This makes it possible to breed for improved feed efficiency, which is a difficult trait to select for, because it cannot be measured individually in fish. Improved feed efficiency reduces environmental impact of the aquaculture production (WP2)
• The control of lipid deposition seems to improve protein retention in rainbow trout, the protein being one of the most expensive raw materials in fish feeds. (WP2).
• In common carp, the 2nd year winter survival was heritable with a heritability of 0.13 and can be improved by selection. In common carp, winter mortality is a major issue for farming operations, and it will be useful to improve it using selective breeding. (WP2).
• In both gilthead seabream and common carp, growth rate during winter has been recognised to have economic interest for the industry. In both species, growth rate during winter was moderately heritable, and can be genetically improved by selection. In both species, increased growth rate during winter had favourable effect on final market body weight. A protocol to apply selection for winter performance at industry level was suggested. (WP2).
• FISHBOOST results on one-year old rainbow trout show that sires differ in their sensitivity to cope with challenging diets with increasing amounts of plant-based ingredients (soya and pea). Nevertheless, breeding programmes selecting for fish performance on the current on-growing feeds will make rainbow trout better adapted to the future feeds that use even more plant-protein concentrates to replace fish meal. With the availability of genetic variation and the limited amount of adverse genotype-by-diet interactions in rainbow trout, breeding programmes are aided to make fish more adapted to the novel feeds. Such information highlights the power of selective breeding to support sustainable development in other areas such as feed development. This should contribute to positive image of fish breeding programmes and to positive consumer perception. (WP2).
• The implemented mixture diet design to quantify fish responses as a surface along several changing diet ingredient mixes allowed the identification of diet mix combinations that support good fish performance, identified as the area on the response surface at which fish performance is at least 95% of the maximum observed performance. This area identifies any combination of the three ingredient mixes that produces almost the same fish performance. This mimics the logic of commercial feed production and the approach can be applied by the industry, yet the method is currently not commonly used. (WP2).
• Significant genetic variation for fillet% was shown in sea bass, sea bream and rainbow trout, with heritability values in the range of 0.24-0.31. Headless carcass%, which is easier to measure, is highly correlated with fillet% (genetic correlations were 0.82 and 0.95) and has a higher heritability (0.32-0.49). Carp is special in that the heritability of fillet% is higher (0.50) than of headless carcass% (0.46). Morphological predictors of fillet% using 2D, 3D and echography measurements have been developed in all species and permit similar and sometimes higher gain than sib selection on fillet yield. The typical gains that can be obtained with 20% selection pressure are in the range 0.5 to 1% fillet yield units per generation. (WP2). Computer simulations concluded that waste is an efficient measure for fillet yield (WP4).
• A new method to pool DNA from individuals with extreme phenotypes (dead/alive) from a disease challenge test and use this data for genomic prediction of individually genotyped fish can reduce costs for genomic selection in aquaculture species, while maintaining high accuracy of selection. (WP3)
• Computer simulation studies revealed that the multitrait genomic prediction increased the accuracy by up to 71% when half of the reference animals were recorded for trait 1 and the other half for trait 2. The highest improvement in accuracy was observed when only 10 sibs were genotyped per family with a low marker density of 10 SNPs per chromosome. The results also indicated that multitrait genomic predictions perform relatively better than single trait predictions for lowly heritability traits that are genetically correlated to high heritability traits. The improvement due to multi-trait predictions was greater when both traits are not measured on the same group of individuals. (WP3)
• In turbot challenged with the parasite Philasterides dicentrarchi, genomic breeding value estimates were clearly more accurate than traditional estimates. It appeared that the LD based method for breeding value estimation also utilised the within family information well. In this real turbot data, it was also found that reducing marker density to about 100 SNPs per chromosome hardly affected prediction accuracies. Both in the data from the carp/Koi Herpes Virus (KHV) experiment and in the seabass/VNN experiment, it was found that KHV-survival needed to be treated as a binary trait to maximise accuracies of genomic breeding value estimates. Also, the BayesB method yielded better genomic breeding values than alternative methods. In both seabream experiments (challenged with Pasteurellosis and S.Chrysophrii), genomic breeding values were found to be significantly more accurate than traditional pedigree-BLUP based estimates. A software pipeline was developed to perform within and across family genomic selection. (WP3).
• When trying to exploit dominance effects by mating, equal or higher performance was obtained by selecting on genomic breeding values and then applying minimum coancestry matings compared to using a mate selection strategy. (WP3).
• BASEPOP is a software tool to select individual fish or strains for new aquaculture breeding programs in an optimal way. FISHBOOSTSEL is a software tool that maximises genetic gain with restriction on inbreeding and assigns matings to different tanks, which is a common constraint in aquaculture breeding. These software tools are freely available to the aquaculture industry. (WP4).
• Knowledge on which perceptions European aquaculture producers and representative organisations have on aquaculture breeding will make the European aquaculture breeding industry better able to decide on acceptable traits and methods in the future. (WP5).
• For common carp and European seabass, economic values have been calculated for production efficiency traits. For Atlantic salmon, also the economic value for resistance to salmon lice was calculated. Cost-benefit analyses quantified the profitability of selective breeding for gilthead seabream, European seabass and Atlantic salmon. (WP6).
• Brochures, the movie in three languages and the site give information about aquaculture breeding and FISHBOOST to society. (WP7).

List of Websites:

Reported by

Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top