Genotype by environment interaction for growth of sole (Solea solea) reared in an intensive aquaculture system and in a semi-natural environment

The objective of the current study was to assess the extent of genotype by environment interaction for growth of sole (Solea solea) grown in an intensive recirculation aquaculture system (RAS) and in a semi-natural outdoor pond (POND). The RAS environment consisted of four indoor shallow raceways without sand, where fish were stocked at a density of 40 fish/m(2) and fed with dry pelleted feed. The POND environment consisted of one 100 m(2) outdoor pond, where fish were stocked at densities of 1.4 fish/m(2), and fish could prey on live ragworms present in the sediment. S. solea (n = 2800) offspring, produced by natural mating of wild broodstock, and with initial body weight of 64 +/- 20 g, were randomly divided over the two environments. After a growing period of 165 days, all fish were harvested and harvest weight (HW, g), specific growth rate (SGR, %BW/d) and sex of all fish was assessed; molecular relatedness between animals was estimated using 9 microsatellite markers. In POND 980 fish and in RAS 774 fish were successfully genotyped and used in the analysis. SGR was higher for sole reared in POND compared to RAS (0.60 +/- 0.01 vs. 0.57 +/- 0.01). Pearson correlation of initial body weight with SGR was negative, and more so in POND compared to RAS (-0.30 vs. -0.16, respectively). Genetic variance and heritability estimates for SGR were higher in POND (h(2) = 0.20 +/- 0.05) than in RAS (h(2) = 0.04 +/- 0.02). Genetic correlations for HW and SGR of sole reared in RAS and in POND were 0.56 +/- 0.34 and 0.27 +/- 0.3, respectively. The differences in heritable variation and the low genetic correlations of growth of sole between environments suggest strong genotype by environment interaction. These results are important in developing breeding programs for sole because the accuracy of selection and genetic gain for growth of sole may differ between environments. Low genetic correlations for growth between environments imply that the best genotypes in an intensive aquaculture environment are not to be necessarily the best genotypes in more natural environments such as ponds. (C) 2013 Elsevier B. V. All rights reserved.