Estimating genetic parameters and genotype-by-environment interactions in body traits of turbot in two different rearing environments

Estimates of heritability and genotype-by-environment (G x E) interactions for three common traits, such as harvest weight (HW), body length (BL), and condition factor (K) were estimated for harvest-size turbot in an industrial farming system (IFS) and a similar system maintained at a lower temperature (IFSLT). The only difference between the two environments was the variation in water temperature. The experimental population (69 families) was composed of 16 maternal half-sib family groups and 17 full-sib families generated by artificial mating. After 15-months of rearing, 2125 and 2925 individuals from the IFS and IFSLT environments were evaluated, respectively. The genetic analysis was based on an animal model with mean family weight at tagging as a covariate, test tank effect as a fixed effect, and an additive genetic effect plus an effect common to full-sib families as random effects using the restricted maximum likelihood method. Heritability estimates within the IFS environment were medium for HW and BL (0.34 +/- 0.12 and 0.34 +/- 0.10) but very low for K (0.009 +/- 0.03). Heritability estimates within the IFSLT environment for HW, BL, and K were 0.16 +/- 0.05, 0.17 +/- 0.05, and 0.04 +/- 0.04, respectively. The genetic correlations between HW and BL in both environments were very high (0.99) with small standard errors. However, the genetic correlations between K and other two traits (HW and BL) were both not significant. The genetic coefficients of variations for HW in the IFS and IFSLT environments were 20.16 and 9.62, and those for BL were 6.68 and 3.70. The genetic correlations for HW and BL between environments were 0.97 +/- 0.15 and 0.90 +/- 0.12, respectively. Our results suggest weak re-ranking of genotypes and heterogeneity of additive genetic variation across environments for HW and BL. The genetic correlation (0.78) for K was near the break-even point but with a high standard error (0.77). This is the first report on G x E interactions across environments for turbot growth traits, which will be of great value to optimize a turbot selective breeding program. Statement of relevance: Our paper offers guideline to select breeding strategy. (C) 2015 Elsevier B.V. All rights reserved.