Abstract
Carcass traits are crucial for the meat duck industry, but their measurement in live animals poses challenges, limiting the effectiveness of traditional breeding methods. Genomic selection (GS) is a powerful approach in animal breeding that enhances selection efficiency and reduces costs. However, research on effective genomic selection strategies for improving carcass traits in poultry, particularly commercial purebred ducks, has been scarce. This study analyzes the phenotypic and genetic parameters of 39 duck carcass traits in 1649 ducks of a commercial breed, focusing on carcass weight and component measurements. Pedigree heritability estimates indicated high values for gizzard weight (0.46) and its percentage (0.48), as well as feet weight (0.43) and wingette tip weight percentage (0.43). Moderate heritability was observed for over half of the traits, including carcass weight (0.27) and eviscerated weight (0.27). Phenotypic and genetic correlations among weight traits were generally positive, except for abdominal fat. We applied pedigree BLUP and genomic BLUP (GBLUP) methods for predictive analysis based on cross-validation. GBLUP demonstrated predictive reliability ranging from 0.07 to 0.44, significantly outperforming traditional BLUP methods. Notably, 33 of 39 traits exhibited improved predictive reliability with GBLUP. We designed three breeding strategies of GS based on parental sibling and sibling information, revealing that inter&intra-generational prediction combining parental sibling and sibling information yielded the highest reliability for most traits. This study marks a significant innovation by establishing effective genomic selection strategies tailored for carcass traits in poultry, demonstrating the potential for enhancing predictive accuracy and informing future breeding strategies. Expanding the reference population in each generation and utilizing it for predictions in the next hatch is recommended to improve breeding outcomes and enhance meat yield and quality in duck production.