Search

International audience; Feed conversion ratio (FCR) is an important trait to target in breeding programs in order to improve fish farming sustainability and increase environmental efficiency. Due to the complexity of accurately measuring the individual feed consumption of fish, developing a selective breeding programme to improve FCR is a challenge. Using video-recordings of several consecutive meals in tilapia groups, we selected two divergent lines of Nile tilapia for high (FCR+) and low (FCR-) FCR at juvenile stage (12.2 g) combining BLUP and within family selection. After two generations, we observed a 12% realized difference in FCR between both divergent lines, indicating that the inclusion of FCR in a selective breeding programme can be efficient in practice. This divergence was in line with a realized heritability of 0.19 for FCR. The divergence in estimated breeding values of FCR between the two lines was reduced (3%) but still present. Another important result was that the realized genetic correlation between FCR and Daily Weight Gain (DWG) was highly negative (rg = -0.69 + 0.16), meaning that improving growth by selective breeding would also indirectly improve FCR in juvenile Nile tilapia, although direct selection for FCR would be more efficient than indirect selection through growth to improve FCR.

Accurately measuring the phenotype at the individual level is critical to the success of selective breeding programs. Feed efficiency is a key sustainability trait and is typically approached through feed conversion ratio (FCR). This requires measurements of body weight gain (BWG) and feed intake (FI), the latter being technically challenging in fish. We assessed two of the principal methods for measuring feed intake in fish over consecutive days: (1) group rearing 10 fish per group and video recording the meals and (2) rearing fish individually on a restricted ration. Juvenile Nile tilapia (Oreochromis niloticus) from the Genetically Improved Farmed Tilapia (GIFT) strain and the Cirad strain were entered into the study (128 GIFT and 109 Cirad). The GIFT strain were reared over three consecutive periods of 7 days each under different feeding, recording, and rearing scenarios (i) in groups fed an optimal ration (g100) or (ii) fed a 50% restricted ration (g50) both with video records of all meals and (iii) reared in isolation and fed a 50% restrictive ration. The Cirad strain were tested similarly but only for scenarios (i) and (iii). All fish were fed twice daily with a calculated ration. Correlations showed the same trends for the GIFT and the Cirad strains. For the GIFT strain, correlations were positive and significant for BWG and FI measured in scenarios (i) and (ii), 0.49 and 0.63, respectively, and FI measured in scenarios (i) and (iii) (0.50) but not for BWG measured in scenarios (i) and (iii) (0.29, NS). The phenotypic correlation estimated for FCR between scenarios (i) and (ii) with fish fed an optimal or a 50% restricted ration was low and not significant (0.22). Feed Conversion Ratio for GIFT fish reared in groups or in isolation and fed with a restricted ration [scenarios (ii) and (iii)] were not significantly correlated either. Social interactions between fish, potentially impacting their efficiency, may explain the results. Therefore, selective breeding programs seeking to improve feed efficiency will need to carefully plan the feeding rate and the rearing system used to estimate FCR in order to optimize selection for the targeted production system.