Selective breeding for juvenile survival in Chinese tongue sole (Cynoglossus semilaevis): Heritability and selection response

Chinese tongue sole (Cynoglossus semilaevis) is an economically important marine flatfish which is now severely threatened by various bacterial pathogens (especially at juvenile stage) in China. As we all known, it is of great importance to increase the natural disease resistance of farmed fish. So, the aim of this study was to verify the exitance of genetic variance of natural disease resistance and to detect the selection response by using juvenile natural survival data (involving four year-classes and three generations with 221 full-sib families, 195,589 individuals). Survival was defined as binary trait (dead/alive) fitted in two cross-sectional models (i.e. crosssectional linear sire-dam model (CLM) and cross-sectional threshold (logit) sire-dam model (CTM)). Heritabilities of survival were estimated with each generation dataset and with complete dataset. Heritability estimates varied among generations regardless of model used, i.e., 0.01-0.17 and 0.03-0.25 for CLM and CTM respectively. On the observed (CLM) and underlying (CTM) scale with complete dataset, the heritabilities were 0.09 +/- 0.04 and 0.13 +/- 0.06 respectively. Both models performed nearly identical and very high selection accuracy (> 0.99), the accuracy of selection obtained from CLM (0.993) was slightly higher than CTM (0.991). By cross-validation, the prediction accuracy of CLM is 21% higher than CTM, which was 0.885 and 0.730 for CLM and CTM respectively. The average of predicted genetic gain for each generation was 14.89%, and the average of realized genetic gain was 8.10% per generation for juvenile survival. These results confirmed the existence of genetic variation for juvenile natural survival and highlighted the enormous potential for improving natural survival by selective breeding in tongue sole.