PSV-17 Estimates of variance components and genetic parameters for weight, carcass, and reproductive traits in U.S. Suffolk sheep including genomic information.

Abstract The Suffolk sheep breed is widely used in the U.S. as a terminal sire, valued for its growth and carcass traits. In the National Sheep Improvement Program (NSIP), breeders aim to enhance these traits through selection. However, the genetic parameters currently used by NSIP in its Suffolk evaluation have not been updated in over two decades, and genomic information has yet to be integrated. Therefore, our aim was to re-estimate genetic parameters for growth, carcass, and reproductive traits in this breed. Additionally, we evaluated the impact of genomic data on genetic parameter estimates and theoretical accuracies of estimated breeding values (EBV). Data were birth weight (BWT), weaning weight (WWT), post-weaning weight (PWWT), post-weaning ultrasound eye muscle (PEMD) and fat (PFD) depth, and number of lambs born (NLB) and weaned (NLW). The number of records, contemporary groups, flocks, and genotypes, and the model components are in Table 1. The pedigree consisted of 70,271 animals. The data were pre-adjusted for several fixed effects, including birth and rearing type, dam age, and recording age for weight traits or weight for ultrasound traits. Contemporary group (flock, year, season, management group, sex, and recording date for all but BWT) was the only fixed effect fitted in the models. No covariance between direct and maternal additive genetic effects was included. Variance components were estimated fitting the A matrix for pedigree-based and H matrix for genomic-based evaluations (1,467 animals genotyped with the 50K SNP chip). After quality control, 41,782 SNPs remained. BLUPF90+ software was used with default parameters, and theoretical accuracies calculated from the prediction error variances. Except for PWWT and PEMD, the re-estimated variance components were like those currently used by NSIP (Table 2). However, we observed a substantial decrease in phenotypic variance and heritability for PWWT (currently: 63.15 kg² and 0.28; re-estimated: 34.60 kg² and 0.10) and an increase in phenotypic variance with a lower heritability for PEMD (currently: 4.37 mm² and 0.39; re-estimated: 9.69 mm² and 0.23). Given the importance of these traits for the selection goal of U.S. terminal sire breeders, these differences suggest that realized genetic gains may differ from those predicted using current NSIP EBV. These changes also suggest that the design of selection indexes may need re-evaluation. Theoretical accuracy improved across all traits with inclusion of genomic data, ranging from 0.01 (5%) for PFD to 0.03 (13%) for NLW, demonstrating enhanced prediction reliability. This increment was even higher for NLW (0.06, 18%) on genotyped sheep. Although relatively few animals were genotyped, integrating genomic data resulted in increased theoretical accuracies, reinforcing its potential to further enhance genetic evaluations as more genomic data are incorporated. In contrast, genomic data had minimal effect on variance component estimates, likely due to limited genotyping.