Recent advances in genotyping-by-sequencing have enabled genome-wide association studies in nonmodel species including those in aquaculture programs. As with other aquaculture species, rainbow trout and steelhead (Oncorhynchus mykiss) are susceptible to disease and outbreaks can lead to significant losses. Fish culturists have therefore been pursuing strategies to prevent losses to common pathogens such as Flavobacterium psychrophilum (the etiological agent for bacterial cold water disease [CWD]) and infectious hematopoietic necrosis virus (IHNV) by adjusting feed formulations, vaccine development, and selective breeding. However, discovery of genetic markers linked to disease resistance offers the potential to use marker-assisted selection to increase resistance and reduce outbreaks. For this study we sampled juvenile fish from 40 families from 2-yr classes that either survived or died after controlled exposure to either CWD or IHNV. Restriction site−associated DNA sequencing produced 4661 polymorphic single-nucleotide polymorphism loci after strict filtering. Genotypes from individual survivors and mortalities were then used to test for association between disease resistance and genotype at each locus using the program TASSEL. After we accounted for kinship and stratification of the samples, tests revealed 12 single-nucleotide polymorphism markers that were highly associated with resistance to CWD and 19 markers associated with resistance to IHNV. These markers are candidates for further investigation and are expected to be useful for marker assisted selection in future broodstock selection for various aquaculture programs.

Background Infectious hematopoietic necrosis (IHN) is a disease of salmonid fish that is caused by the IHN virus (IHNV). Under intensive aquaculture conditions, IHNV can cause significant mortality and economic losses. Currently, there is no proven and cost-effective method for IHNV control. Clear Springs Foods, Inc. has been applying selective breeding to improve genetic resistance to IHNV in their rainbow trout breeding program. The goals of this study were to elucidate the genetic architecture of IHNV resistance in this commercial population by performing genome-wide association studies (GWAS) with multiple regression single-step methods and to assess if genomic selection can improve the accuracy of genetic merit predictions over conventional pedigree-based best linear unbiased prediction (PBLUP) using cross-validation analysis. Results Ten moderate-effect quantitative trait loci (QTL) associated with resistance to IHNV that jointly explained up to 42% of the additive genetic variance were detected in our GWAS. Only three of the 10 QTL were detected by both single-step Bayesian multiple regression (ssBMR) and weighted single-step GBLUP (wssGBLUP) methods. The accuracy of breeding value predictions with wssGBLUP (0.33–0.39) was substantially better than with PBLUP (0.13–0.24). Conclusions Our comprehensive genome-wide scan for QTL revealed that genetic resistance to IHNV is controlled by the oligogenic inheritance of up to 10 moderate-effect QTL and many small-effect loci in this commercial rainbow trout breeding population. Taken together, our results suggest that whole genome-enabled selection models will be more effective than the conventional pedigree-based method for breeding value estimation or the marker-assisted selection approach for improving the genetic resistance of rainbow trout to IHNV in this population.

Rainbow trout from 23 families were evaluated for growth and resistance to the bacterial coldwater disease (BCWD) caused by Flavobacterium psychrophilum and infectious haematopoietic necrosis (IHN) caused by IHN virus. Average family weights were between 161 and 263 g with an average of 225 g at 213 days post-fertilization with specific growth rates ranging from 2.37 to 2.88. Per cent survival of fish challenged with F. psychrophilum was between 18% and 100%, while for those challenged with IHNV, the range was between 12% and 93%. Significant positive correlations were found for end body weight and resistance to IHN (P < 0.05) and for early body weight and resistance to BCWD (P < 0.1). However, no significant correlations were detected between resistance to both pathogens or disease resistance and overall genetic diversity or diversity within the major histocompatibility locus.

Two hatches involving 11,158 pullets from 22 Single Comb White Leghorn strain crosses were used in these experiments: 13 from rapid-feathering (k+/-) and 9 from slow-feathering (K/-) dams, some of which shared common genetic backgrounds; and all from 5 homozygous k+ sire types. Birds were grown intermingled within hatch, and were subsequently caged in two laying houses, one with 30.5 cm x 40.6-cm cages and the other with 40.6 cm x 30.5-cm cages, at densities of three or four birds/cage. Data were collected by cage on hen-day rate of lay from 50% production to 52 wk of age; (HD%), number of eggs produced per hen housed (EHH), percentage mortality (%M) to 52 wk of age; 35-wk BW (BW35); egg weight (EW34), specific gravity (SG34), and percentage bloodspots (BS%34) at 34 wk of age. The effects of hatches or house (H), density (D), maternal feathering type (MFT), genetic strain within MFT (GS/MFT), bank level within house (BL/H), and all possible interactions were included in the analysis model. Hatch, house, and cage type were completely confounded, so no general conclusion can be made concerning the effects of the reverse and standard cages on performance. The H effects were significant for all traits except HD%. Cage D significantly affected all traits except SG34 and BS%34. Pullets housed at three/cage weighed 26 g less, produced 5.5 more eggs that were .3 g/egg lighter, laid at a 1.7% higher hen-day rate, and had 3.6 less %M from 20 to 52 wk of age than those housed at four/cage. The MFT significantly affected all traits except %M, SG34, and BS%34. The k+/- daughters from K/- dams weighed 34 g less, produced at a 2.8% lower HD%, laid 6.9 fewer eggs, which eggs averaged .6 g/egg less than the eggs from the k+/- daughters of the k+/- dams. Interactions were in general small and nonsignificant. However, the H x GS/MFT interactions were significant for HD%, EHH, BW35, %M, and EW34 (P less than .05); their significance is suggestive of the presence of genotype x environment interactions.

No abstract