Genetic parameters and genotype-by-environment (GxE) interactions for growth have been estimated in many studies in aquaculture fish, but are scarcer for quality traits, and almost non-existent for GxE interaction estimates on those traits. In the present work, 253 full-sib families from 33 males and 23 females of European seabass were produced in a partly factorial mating design. All fish were reared in common garden from incubation, then dispatched at 14 months of age to three fish farms representing a wide variety of rearing systems (pond, cage, raceway) and environmental conditions (Portugal, Italy, Israel). Around 400 g mean weight, 512 to 639 fish in each site were dissected to evaluate the share of different body parts.,The heritability of most yields (evaluated as the residual of the regression of a body part weight on body weight) across all environments ranged from 0.35 0.10 to 0.63 0.10, except for digestive tract yield which showed very low heritability (0.02 ± 0.05). Heritability within site was generally higher, as most traits showed some level of GxE interactions, with genetic correlations between sites ranging from 0.66 to 0.92. We also validated a simple predictor of viscera yield, measurable on live individuals with ultrasound tomography, which had a very high genetic correlation with measured viscera yield (rG= 0.89 ± 0.04). These results will help design selective breeding programs for quality traits in European seabass.

Coinfection by multiple pathogens is common in aquaculture. Since 2012, Pacific oysters in Europe have been affected by two main pathogens, with additive or cooperative pathogenicity: the ostreid herpes virus type 1 (OsHV-1) and bacterium Vibrio aestuarianus. In oysters, genetic improvement by selective breeding is effective in mitigating diseases caused by single pathogens, but little is known about resistance to coinfections. This study aimed to investigate the genetic parameters and genomic architecture of disease resistance in C. gigas by comparing experimental infections made with each of the two pathogens, as well as coinfection with both. One hundred families were produced using parents from three origins: two wild populations experiencing different pathogen pressures and oysters experimentally selected for their higher resistance to OsHV-1 and V. aestuarianus infections. The experimental infections were carried out on 7-month old juveniles, leading to survival rates of 52%, 55% and 43%, for the bacterial, the viral and dual infection, respectively. Individuals were genotyped using a 57 k SNP array. Survival varied widely among families and parental origins. Heritability estimates for survival ranged from 0.20 to 0.50 for the three infections, and genetic correlations were low between V. aestuarianus and the two other experimental infections, but very high between OsHV-1 and coinfection. Genome-wide association study revealed a polygenic architecture for all traits. However, five quantitative trait loci (QTLs) were detected in the OsHV-1 treatment as well as for coinfection with one specific genomic region on linkage group 6 being related to higher survival. Prediction accuracy was higher using a genomic model than a pedigree-based model, particularly for OsHV-1 and coinfection, for which larger numbers of individuals were genotyped. Our results suggest that (1) geographic origin of oysters should be considered when establishing a breeding program for improved survival; (2) use of genomic selection (GS) and QTL mapping may lead to more efficient selection and faster genetic gain; and (3) the coinfection challenge used in this study, which is likely to be closer to field conditions than the other treatments, is practical and may be suitable for breeding programs. Our findings represent a significant step towards using genomic information to improve disease resistance in selective breeding programs for the Pacific oyster, and possibly other aquaculture species.

To grow and develop, shrimps must shed their exoskeletons. In shrimp an inter-individual variability in molting time is spread over about a week. The genetic basis of this variability and its implications for selective breeding programs to improve growth or quality traits remains largely unexplored. We estimated the effect of molting stage on genetic parameters of growth and color traits in the Pacific blue shrimp Litopenaeus stylirostris. As the variability of molting time may also be partly genetically determined, we estimated the heritabilities of molt stages and molting time under the hypothesis that some families may have the additive genetic potential to molt earlier (or later) than others. To address these questions, 1200 shrimp progenies (mean body weight 14.5 g) from 49 full-and half-sib families, bred according to a partly factorial design with double-sire insemination, were phenotyped over 3 successive days, 3 to 5 days before the full moon. Body weight, body length, head length and molt stage were recorded on uncooked shrimp and external color traits on uncooked or both uncooked and cooked shrimp. Molt stage had a significant phenotypic effect on all traits except for body length and head-to-length ratio. The heritability estimates for both growth and color traits were, however, similar in models with or without molting, and the estimated breeding values were highly correlated between these models. The heritability of molt stage with a four-class threshold model was low (0.07 [0.02-0.28]). When the stage classes were combined into just two to simplify the description of the molting process, the heritability of early or late molting was intermediate (from 0.16 [0.10-0.28] to 0.22 [0.13-0.40]), indicating that this trait is partly under additive genetic determinism rather than solely under environmental control. The genetic correlations of molting traits with body weight were not different from zero. These results are discussed for selective breeding practices and shrimp molting biology to consider in future study molting as a factor to improve selection practices for disease resistance, feed efficiency and robustness.

Selective breeding plays a key role in the sustainable development of aquaculture by improving productivity and efficiency. While breeding programs have initially focused on growth traits, their potential to mitigate the environmental footprint of fish farming remains poorly explored. Here, we present a case-study on the environmental consequences of selective breeding during five generations (15 years) for growth and fillet yield in gilthead sea bream Sparus aurata by integrating genetic gain projections into a full supply chain Life Cycle Assessment (LCA), extending from hatchery to final consumer. The potential impacts were calculated in terms of 1 tonne of fish produced at the farm level and were also expressed per tonne of flesh consumed. Performance data and genetic parameters from families in a seabream breeding program were used to estimate breeding values under different selection scenarios. Selection for growth significantly shortened the production cycle, while selection for fillet yield increased the proportion of edible product. Although environmental gains at the farm level were modest (<1 %), substantial reductions (up to 7 % in climate change impacts and cumulative energy demand) were observed when impacts were expressed per tonne of fillet consumed. Feed production and consumption remained the dominant source of environmental burden across all scenarios. By explicitly accounting for post-farm stages - processing, distribution, consumption, and waste management - our study demonstrates that genetic improvement, particularly for fillet yield, can meaningfully reduce the environmental footprint of aquaculture products delivered to the consumer. This case study underlines the importance of integrating full supply chain assessments into genetic selection strategies to drive sustainable breeding in aquaculture.

Genotype-by-environment (GxE) interaction in aquaculture is usually estimated for continuous traits and based on data from a limited number of 2-3 rearing sites. Here we report the results of a GxE study for resistance to Pacific oyster mortality syndrome (POMS), a multi-factorial disease that severely impacts Pacific oyster production worldwide. The syndrome is largely associated with Ostreid Herpes Virus 1 (OsHV-1). Resistance to OsHV-1 in Crassostrea gigas has been shown to be heritable, meaning that selective breeding is a suitable strategy for reducing mortalities. However, limited information was available about GxE interaction or the possible need to consider it in selective breeding. Survival of two cohorts (C1 and C2), consisting of a total of 104 full-sib families, was evaluated during the summer of 2013 in 7 sites along the French Atlantic, Channel and Mediterranean coasts. Mean survivals in autumn 2013 were 12.6% +/- 10.9 and 4.6% +/- 6.4 for C1 and C2, respectively. Genetic parameters were computed by MCMC, which is suitable for binary data like survival. Heritability estimates ranged from 0.16 to 0.42 depending on site and cohort, with a mean of 0.24 [0.20; 0.27] when including all data. GxE interactions were estimated by the genetic correlations between pairs of sites. Genetic correlations were high (rho > 0.80) for C1 between most tidal Atlantic and Channel sites, and intermediate between tidal sites and a Mediterranean lagoon site, while they were lower and more variable for C2 (0.21-0.77). Expected genetic gains were maximal when production site was the same as selection site. They were closed to this expected maximum when the selection site was different from the production site along the Atlantic or Channel coast. Limited GxE interaction along the French Atlantic coast is favorable to wide dissemination of genetically improved oysters along this coast. Limited potential improvement was shown in the Mediterranean site if selection was carried out elsewhere, confirming the specificity of this environment. Consequently, a specific strategy such as dedicated breeding should be used to achieve genetic progress for this site.

Some cultured shrimp populations have been domesticated from a very small number of founders, which likely limits their genetic variability and heightens inbreeding. The population of Pacific blue shrimp Litopenaeus stylirostris in New Caledonia is one such case. It has undergone domestication since 1980 (36 generations at the time of this study) but was based on only four initial founders. Polymorphic SNPs for pedigree assignment and assessment of genetic variability were identified by partial RAD sequencing. The low genetic variability of this domesticated population compared with a wild Mexican population of reference was confirmed with 820 SNP markers. To estimate genetic parameters, full-sib and half-sib families (n = 148) were bred from 74 dams and 92 sires using a double-sire insemination of each dam with spermatophores from two sires, the first time that this method has been used in shrimp genetics. At 6 months of age (14.5 g), 1200 progenies were then phenotyped for body weight and external color by digital vision, before and after cooking, using the CIE L*a*b* color space international reference system. This allowed 99.4 % of the progenies were to be successfully assigned to parents with 171 SNP markers. Heritability was high for body weight and color traits before and after cooking (h2 = 0.41-0.59). Correlations between the blue color (b*) of uncooked shrimps and uncooked body weight was negative but did not differ from zero (rg =-0.25 +/- 0.24). The blue (b*) value of uncooked shrimps was positively correlated with the red (a*) value of cooked shrimps (rg = 0.48 +/- 0.14). All results are favorable to pursuing a selective breeding program for blue shrimp in New Caledonia using this population despite its limited number of initial founders and more than 36 generations of domestication. Genetic factors potentially contributing to the results are discussed.

This study focuses on genetic resistance to infectious pancreatic necrosis (IPN), a highly contagious disease caused by an aquatic birnavirus (IPNV) which especially affects salmonids worldwide. The objectives were to estimate the heritability of IPN resistance and to fine map quantitative trait loci (QTL) using a Bayesian Sparse Linear Mixed Model to identify candidate genes possibly linked to IPN resistance in two successive generations from a French commercial strain of rainbow trout. For each generation, 2000 fish were experimentally exposed by bath to IPNV and mortalities were monitored daily during 5 weeks. All fish were genotyped using a medium-density 57 K single nucleotide polymorphism (SNP) chip and imputed to high-density genotypes (665 K SNPs). The mean survival rate was 70

In aquaculture, sterile triploids are commonly used for production as sterility gives them potential gains in growth, yields, and quality. However, they cannot be reproduced, and DNA parentage assignment to their diploid or tetraploid parents is required to estimate breeding values for triploid phenotypes. No publicly available software has the ability to assign triploids to their parents. Here, we updated the R package APIS to support triploids induced from diploid parents. First, we created new exclusion and likelihood tables that account for the double allelic contribution of the dam and the recombination that can occur during female meiosis. As the effective recombination rate of each marker with the centromere is usually unknown, we set it at 0.5 and found that this value maximizes the assignment rate even for markers with high or low recombination rates. The number of markers needed for a high true assignment rate did not strongly depend on the proportion of missing parental genotypes. The assignment power was however affected by the quality of the markers (minor allele frequency, call rate). Altogether, 96-192 SNPs were required to have a high parentage assignment rate in a real rainbow trout dataset of 1,232 triploid progenies from 288 parents. The likelihood approach was more efficient than exclusion when the power of the marker set was limiting. When more markers were used, exclusion was more advantageous, with sensitivity reaching unity, very low false discovery rate (<0.01), and excellent specificity (0.96-0.99). Thus, APIS provides an efficient solution to assign triploids to their diploid parents.

Triploidy is very useful in both aquaculture and some cultivated plants as the induced sterility helps to enhance growth and product quality, as well as acting as a barrier against the contamination of wild populations by escapees. To use genetic information from triploids for academic or breeding purposes, an efficient and robust method to genotype triploids is needed. We developed such a method for genotype calling from SNP arrays, and we implemented it in the R package named GenoTriplo. Our method requires no prior information on cluster positions and remains unaffected by shifted luminescence signals. The method relies on starting the clustering algorithm with an initial higher number of groups than expected from the ploidy level of the samples, followed by merging groups that are too close to each other to be considered as distinct genotypes. Accurate classification of SNPs is achieved through multiple thresholds of quality controls. We compared the performance of GenoTriplo with that of fitPoly, the only published method for triploid SNP genotyping with a free software access. This was assessed by comparing the genotypes generated by both methods for a dataset of 1232 triploid rainbow trout genotyped for 38,033 SNPs. The two methods were consistent for 89% of the genotypes, but for 26% of the SNPs, they exhibited a discrepancy in the number of different genotypes identified. For these SNPs, GenoTriplo had >95% concordance with fitPoly when fitPoly genotyped better. On the contrary, when GenoTriplo genotyped better, fitPoly had less than 50% concordance with GenoTriplo. GenoTriplo was more robust with less genotyping errors. It is also efficient at identifying low-frequency genotypes in the sample set. Finally, we assessed parentage assignment based on GenoTriplo genotyping and observed significant differences in mismatch rates between the best and second-best couples, indicating high confidence in the results. GenoTriplo could also be used to genotype diploids as well as individuals with higher ploidy level by adjusting a few input parameters.

The Percidae family comprises many fish species of major importance for aquaculture and fisheries. Based on three new chromosome-scale assemblies in Perca fluviatilis, Perca schrenkii and Sander vitreus along with additional percid fish reference genomes, we provide an evolutionary and comparative genomic analysis of their sex-determination systems. We explored the fate of a duplicated anti-Mullerian hormone receptor type-2 gene (amhr2bY), previously suggested to be the master sex determining (MSD) gene in P. flavescens. Phylogenetically related and structurally similar amhr2 duplications (amhr2b) were found in P. schrenkii and Sander lucioperca, potentially dating this duplication event to their last common ancestor around 19-27 Mya. In P. fluviatilis and S. vitreus, this amhr2b duplicate has been lost while it was subject to amplification in S. lucioperca. Analyses of the amhr2b locus in P. schrenkii suggest that this duplication could be also male-specific as it is in P. flavescens. In P. fluviatilis, a relatively small (100 kb) non-recombinant sex-determining region (SDR) was characterized on chromosome-18 using population-genomics approaches. This SDR is characterized by many male-specific single-nucleotide variants (SNVs) and no large duplication/insertion event, suggesting that P. fluviatilis has a male heterogametic sex determination system (XX/XY), generated by allelic diversification. This SDR contains six annotated genes, including three (c18h1orf198, hsdl1, tbc1d32) with higher expression in testis than ovary. Together, our results provide a new example of the highly dynamic sex chromosome turnover in teleosts and provide new genomic resources for Percidae, including sex-genotyping tools for all three known Perca species.

The development of sustainable aquaculture relies on replacing marine raw materials like fish meal (FM) and fish oil (FO). Emerging alternatives, such as single-cell proteins and alternative lipids, offer promise. This study explored the effects of partially substituting FM with 10% bacterial protein (Methylococcus capsulatus) and completely replacing FO with a blend of poultry oil (PO) and DHA-rich microalgae oil in European sea bass (Dicentrarchus labrax) of unselected (WT) and selected (HG) genotypes. The results indicated that bacterial protein had no adverse impact on fish growth. The HG group demonstrated better growth and feed conversion due to genetic selection. This study also analysed the dietary and genotype effects on body lipid composition and fatty acid profiles. Notably, the HG fish had lower levels of major fatty acids (EPA, DHA, n-3 FAs, and n-3 LC-PUFAs) in their fillets, but not in their whole-body composition. These differences influenced sensory and qualitative aspects. Electronic sensory analyses (the first e-sensory profiling conducted for genetic purposes in fish) showed more significant differences due to diet in the WT group, with a less variable pattern in the e-tongue score in for the HG group. The volatile profiles showed no significant differences. In summary, combining selected fish genotypes with innovative feeds is a step forward in aquaculture. It maximizes nutrient utilization, enhances fish growth, and improves product quality. This approach becomes increasingly important in scenarios with limited FM/FO availability, promoting sustainability in aquaculture.

Functional ingredients have profiled as suitable candidates for reinforcing the fish antioxidant response and stress tolerance. In addition, selective breeding strategies have also demonstrated a correlation between fish growth performance and susceptibility to stressful culture conditions as a key component in species domestication processes. The aim of the present study is to evaluate the ability of a selected high-growth genotype of 300 days post-hatch European sea bass (Dicentrarchus labrax) juveniles to use different functional additives as endogenous antioxidant capacity and stress resistance boosters when supplemented in low fish meal (FM) and fish oil (FO) diets. Three isoenergetic and isonitrogenous diets (10% FM/6% FO) were supplemented with 200 ppm of a blend of garlic and Labiatae plant oils (PHYTO0.02), 1000 ppm of a mixture of citrus flavonoids and Asteraceae and Labiatae plant essential oils (PHYTO0.1) or 5000 ppm of galactomannan-oligosaccharides (GMOS0.5). A reference diet was void of supplementation. The fish were fed the experimental diets for 72 days and subjected to a H2O2 exposure oxidative stress challenge. The fish stress response was evaluated through measuring the circulating plasma cortisol levels and the fish gill antioxidant response by the relative gene expression analysis of nfΚβ2, il-1b, hif-1a, nd5, cyb, cox, sod, cat, gpx, tnf-1α and caspase 9. After the oxidative stress challenge, the genotype origin determined the capacity of the recovery of basal cortisol levels after an acute stress response, presenting GS fish with a better pattern of recovery. All functional diets induced a significant upregulation of cat gill gene expression levels compared to fish fed the control diet, regardless of the fish genotype. Altogether, suggesting an increased capacity of the growth selected European sea bass genotype to cope with the potential negative side-effects associated to an H2O2 bath exposure.

One of the main objectives of the present study was an effective replacement of dietary fishmeal/fish oil (FM/FO) by new raw materials without negatively affecting European sea bass (Dicentrarchus labrax) performance and health status within a selection breeding context. The genomic selection of this species is still in its infancy if compared to salmonids and in particular, the role of D.labrax genome in shaping the gut microbiome, has been scarcely investigated. Wildtype (WT) and selected (GS) sea bass were fed with two diets: a control (20% FM/ 5-9% FO), and a "future" (F) diet, in which FM was decreased to 10% being replaced by poultry meal, whereas FO was completely replaced by a blend of rapeseed, poultry, and algae oils. The morphological evaluation of the intestine revealed a well-organized folding pattern and a conserved gut epithelial barrier for all fish groups. Despite a basal level of inflammation in the proximal intestine of WT fish, no differences were observed neither in the morphometric characteristics of goblet cells nor in the expression of GALT-related genes in response to fish genotype or diet. At distal intestine, WT fish showed a higher inflammatory status and larger goblet cells than GS fish and within the same genotype, fish fed the F diet had in general larger goblet cells. In distal intestine, a significant effect was found on the expression of 3 out of 7 target GALT genes. In particular, the expression of cytokines il-1 beta, tnf-alpha, and il-10 was different, showing an interaction effect diet x genotype. Diet had a lower influence upon gut bacterial composition than genotype. Indeed, regardless of the diet, WT fish showed higher species richness than GS genotype and this could be a direct consequence of selective breeding for multiple traits selection including growth, external morphology for lower abdominal fat deposition and adaptation to multiple and successive feed sources and composition across generations of selection. Furthermore, the gut microbiota of GS fish shared a reduced individual variability, indicating an enhanced capacity to cope with changes in diet composition. The less changes of GS sea bass at the level of gut bacterial composition in cumulating data collected with the feeds, demonstrate a capacity to reshape their microbiota thus better adapting to the diet, but with no negative impact on their growth performances, and even a better growth. A significant genotype effect was found for specific bacterial taxa, such as Paracoccus genus and other genera belonging to Moraxellaceae family, which were enriched in WT fish, regardless of the diet. Interestingly, the relative abundance of Paracoccus genus was positively correlated with higher proinflammatory cytokine il-1 beta expression found in distal intestine of wildtype sea bass.

The sustainable development of the aquaculture industry relies on the use of alternative conventional and emergent raw materials that contributes to a circular economy and to reduce the dependency on fish meals and fish oils coming from oceanic fish populations. Additionally, the genetic selection of farmed fish that can display higher growth and feed utilization when fed alternative feeds, is pointed out to be a complementary valuable tool to facilitate the implementation of circular economy approaches. The main purpose of the present study was to determine the effectiveness of genetic selection for growth in European sea bass, in response to a challenge with an alternative diet that aimed to partially replaced fishmeal (FM) by poultry meal (PM) and totally replace fish oil (FO) by a blend of poultry oil (PO) with a novel microalgae oil. The two families of fish juveniles were obtained by in vitro fertilization of selected for a multi-trait including high growth (genetically selected, GS) or non-selected (wild type, WT) broodstocks and then were nutritionally challenged with a control diet that mirrored a standard commercial diet with fishmeal (20%) and fish oil (7%), or a Future diet that partially replaced the FM by PM and totally replaced the FO by a blend of rapeseed oil, PO, and a novel DHA rich-algal oil. From the second month of feeding until the end of the trial, European sea bass that was selected since for 7 generations performed better in terms of growth than the wild-type genotype, possibly related with an apparent favored feed and nutrient utilization. Furthermore, selection decreased the perivisceral fat and increased the nutritional value of flesh by increasing DHA (in g/ 100 g flesh) and ARA contents. In contrast, the dietary treatment showed little effect on fish growth performance, denoting the successful partial replacement of FM by PM and the total replacement of FO by a blend of poultry oil and an emergent microalgal oil. However, Future diet tended to reduce the ADCs of some amino acids, as well as showed an additive effect to genotype in increasing the n-3 PUFA of flesh. Altogether, our data demonstrate that multi-trait genetic selection of European sea bass improve fish plasticity to cope with the variations of ingredients in alternative feeds with low FM/FO.

Additional file 2: Table S1. Number of SNPs removed after several quality filters. Details of the number of SNPs removed with quality controls applied on whole-genome sequences and 57K SNP chip data.

Abstract Background Selective breeding is a promising solution to reduce the vulnerability of fish farms to heat waves, which are predicted to increase in intensity and frequency. However, limited information about the genetic architecture of acute hyperthermia resistance in fish is available. Two batches of sibs from a rainbow trout commercial line were produced: the first (N = 1382) was phenotyped for acute hyperthermia resistance at nine months of age and the second (N = 1506) was phenotyped for main production traits (growth, body length, muscle fat content and carcass yield) at 20 months of age. Fish were genotyped on a 57 K single nucleotide polymorphism (SNP) array and their genotypes were imputed to high-density based on the parent’s genotypes from a 665 K SNP array. Results The heritability estimate of resistance to acute hyperthermia was 0.29 ± 0.05, confirming the potential of selective breeding for this trait. Since genetic correlations of acute hyperthermia resistance with the main production traits near harvest age were all close to zero, selecting for acute hyperthermia resistance should not impact the main production traits, and vice-versa. A genome-wide association study revealed that resistance to acute hyperthermia is a highly polygenic trait, with six quantitative trait loci (QTL) detected, but explaining less than 5% of the genetic variance. Two of these QTL, including the most significant one, may explain differences in acute hyperthermia resistance across INRAE isogenic lines of rainbow trout. Differences in mean acute hyperthermia resistance phenotypes between homozygotes at the most significant SNP was 69% of the phenotypic standard deviation, showing promising potential for marker-assisted selection. We identified 89 candidate genes within the QTL regions, among which the most convincing functional candidates are dnajc7, hsp70b, nkiras2, cdk12, phb, fkbp10, ddx5, cygb1, enpp7, pdhx and acly. Conclusions This study provides valuable insight into the genetic architecture of acute hyperthermia resistance in juvenile rainbow trout. We show that the selection potential for this trait is substantial and selection for this trait should not be too detrimental to improvement of other traits of interest. Identified functional candidate genes provide new knowledge on the physiological mechanisms involved in acute hyperthermia resistance, such as protein chaperoning, oxidative stress response, homeostasis maintenance and cell survival.