Genetic improvement of farmed tilapias: Genetic parameters for body weight at harvest in Nile tilapia (Oreochromis niloticus) during five generations of testing in multiple environments

Phenotypic and genetic parameters for harvest body weight were estimated within and across multiple test environments and five consecutive generations of the synthetic GIFT population. Of a total of about 62,000 individually tagged fish, 43,066 individuals from 461 sires and 815 dams across generations had body weights recorded in two to eight different test environments per generation from 1991 to 1996. The test environments included earthen ponds (fertilized with inorganic fertilizer, organic manure, or on-farm agricultural residues), cage culture, rice-fish, and ponds at test stations located in different agro-climatic regions. Heritability estimates for harvest body weight within test environments and generation were on average h2 = 0.31 (range 0.06–0.68) and the estimates for effects common to full sibs other than additive genetic effects c2 = 0.09 (range 0.04–0.16). The estimates tended to be higher in cage test environments. After adjusting for heterogeneous variances of harvest body weight across generations, test environments and sexes, the estimates across test environments within generations were on average h2 = 0.23 and c2 = 0.03 over the five generations, while the estimate across all test environments and generations was h2 = 0.16 ± 0.02 and c2 = 0.10 ± 0.01. The genetic correlations (rg) between harvest body weights of sibs in different test environments were in general high (0.53–0.99, mean 0.89) for all environments except for intensive cage culture, implying minor genotype by environment (G × E) interactions. The genetic correlations involving the intensive cage culture environment were lower (0.08–0.43) and not significantly different from zero, suggesting that G × E interactions may occur if test environments differ widely. It is proposed that this G × E interaction may involve effects of sexual maturation on growth in pond environments. The overall heritability for body weight at harvest of females (0.20 ± 0.02) was not significantly different from that of males (0.16 ± 0.02). However, these estimates were lower in ponds than in cages, in particular for male body weight. Estimates of genetic correlations between harvest body weight of sib males and females also suggested a moderate sex by genotype interaction in ponds (rg = 0.78 ± 0.04) that did not occur in cages (rg = 0.97 ± 0.04). These results are consistent with the realization that the GIFT population responds well to selection for increased body weight at harvest across a wide range of pond farm environments (including rice–fish culture) without the need to develop environment specific selection lines. The benefits of specialized selection for intensive cage farming systems should be investigated further