Estimations Of Genetic Parameters For Growth Traits Using Microsatellites Markers And G×E Interactions Between Two Difierent Environments In Turbot

The turbot (Scophthalmus maximus L.) is a native European marine fish with high commercial value and is the most widely cultivated commercial flatfish around the world. Due to its fast growth and strong tolerance to cold water temperature, turbot was first introduced into China in 1992 and became one of the most plentiful marine fish species in the North China. Recently, inbreeding depression has emerged and growth rates have decreased because the limited sources of original stock originally and the unknown relationship between introduced fish. Therefore, for the purpose of keeping development of turbot farming industry healthy and continuous, genetic selective breeding to improve important turbot traits should be performed. In addition, the accurate estimation of genetic parameters is considerable to guide selective breeding work. In the breeding process of turbot, pedigree is recorded for every spawn. However, there is an interval of at lowest 2 years for turbot from hatching to sexual maturity. Therefore, due to the quite long generation interval, the estimation of genetic parameters usually has to suffer the loss caused by incomplete pedigree information due to many uncontrolled factors. Additional, the unknown relationship between introduce parents would also affect the estimation. In this study, the microsatellites markers were applied according to two strategies in estimation of genetic parameters for body weight trait of turbot. ① For a data set of 19 families produced at 2013, about one third of progenies in each family were selected at random to be genotyped at 12 microsatellite loci. A sort of molecular relatedness called average molecular relatedness (AMR) was calculated between each two individuals, and then the additive genetic correlation matrix was constructed using AMR. Additionally, the genetic parameters of turbot body weight was estimated on the basis of two different animal models where one is without maternal and common environmental effect (Model 1) and the other one with it (Model 2). ② For 39 families at 2013, the parents consisted of 31 males and 17 females were sampled and genotyped by using 12 microsatellite markers. The coancestry between two parents was calculaed and then used to deduce the relatedness between two progenies, which was named parental molecular relatedness (PMR). A single trait animal model and the additive genetic correlation matrix constructed by PMR were ultilized to estimate the genetic parameters. Within the two strategies, there was another relatedness estimated from complete pedigree, which was called pedigree relatedness (PR).Then PR was also used to construct additive genetic correlation matrix which was applied to estimate genetic parameters along with the same animal model, and then the accuracy of breeding values based on PR was compared with that based on AMR or PR to explore the feasibility of the two strategies. The results demonstrated that ① Pearson correlation between AMR and PR was high (0.914). The best models for AMR and PR were Model 2 and 1 respectively, and the heritability estimates based on best model were 0.19 (±0.056) and 0.66 (±0.17). On the basis of both Model 1 and 2, the accuracy of breeding values from AMR was higher than that from PR. ② The Pearson correlation between PMR and PR was 0.872. The heritability from PMR and PR were close, and 0.52 (±0.13) and 0.55 (±0.22) respectively. AMR and PR showed the same accuracy of breeding values (0.81) through cross-validation. Therefore, in sum, AMR and PR were comparable to PR in terms of estimation of genetic parameters for turbot when only the parents were known.In China, the major farming pattern of turbot is the industrialized culture pattern named "green house+deep well sea water". However, the real water temperature was different in different places. In this study, two farming systems were set, one is the industrial farming system (IFS) with a stable water temperature 18-20℃ and the other is a similar system with lower temperature (IFSLT). The only difference between the two environments was in the water temperature span and variation. The experimental population (69 families) was constructed in 2013. At 100 days post-hatch, from each family, about 100 progenies were randomly selected, weighed, tagged using visible implant elastomer and then divided randomly into two equal subsets (i.e.,50 individuals), which were reared in two respective environments. After rearing for 12 months in different systems,2125 and 2925 individuals from IFS and IFSLT environments were measured for harvest body weight (HW) and body length (BL). The trait condition factor (K) was deduced using weight and length. The heritability, genetic correlations and G x E interactions for the three traits were estimated with animal models using the restricted maximum likelihood (REML) method. Within the IFS environment, heritability estimates were medium for HW and BL (0.34 ±0.12 and 0.34 ±0.10) but very low for K (0.009 ± 0.03). Within the IFSLT environment, the heritability estimates of HW, BL and K were 0.16 ± 0.05,0.17 ± 0.05 and 0.04 ± 0.04, respectively. In the two environments, the genetic correlations between HW and BL were both very high (0.99) with small standard errors. However, the genetic correlations between HW and K, BL and K were both insignificantly different from zero (P> 0.05). For HW and BL, the genetic coefficients of variations in IFS and IFSLT environments were 20.16 and 9.62, and 6.68 and 3.70. Genetic correlations between environments were 0.97 ±0.15 and 0.90 ±0.12 for HW and BL. This study found weak re-ranking of genotypes and heterogeneity of additive genetic variations across environments for traits HW and BL. For trait K, the genetic correlation (0.78) was near the break-even correlation but with a high standard error (0.77). In conclusions, weak G×E interactions were observed for traits HW and BL between IFS and IFSLT environments. This is the first report about genotype-by-environment interactions across environments for growth traits of turbot, which is of great value in optimization of selective breeding program of turbot.