PSX-7 - Heritability estimates for longevity-related traits in U.S. Katahdin ewes

Abstract: Genetic selection for increased ewe longevity can directly contribute to improved flock profitability and welfare. Therefore, this study aimed to estimate heritabilities for eight longevity-related traits in 12,149 Katahdin ewes. The ewes were daughters of 1,157 sires and 6,147 dams and were born between 1989 and 2020 in 58 flocks located across the U.S. The National Sheep Improvement Program provided the data. Age at last lambing (ALL); productive life (PL; number of days between the first and last lambing); total number of litters (TLI), lambs born (TNB), and lambs weaned (TNW) over ewe’s lifetime; total litter weight at lambing (TLB) and weaning (TLW) over ewe’s lifetime; and TLW divided by the ewe’s weight adjusted for 120 d (TLWadj) were evaluated (Table 1). Body weights at lambing and weaning (WW) were adjusted for sex effect, while WW was also adjusted for 60 d old. Ewes with lambing interval and age at the first lambing greater than 720 d and 1,095 d, respectively, were excluded. Moreover, only flocks that reported lambing records across at least seven consecutive years and an average of 20 or more lambing records over the years were evaluated. Fixed effects of contemporary group (flock, year, and season of ewe’s birth), the combined effect of the birth type and rearing type of the ewe, and age class at first lambing, as well as the animal random effect, were included in the animal model fitted. Contemporary groups with less than three ewes were excluded. Moreover, we excluded records deviating ±3.0 SD from the mean. Variance components were estimated by restricted maximum likelihood based on pedigree (by fitting the A matrix) and genomic information (by fitting an H matrix including pedigree and genomic relationships from 10,032 animals genotyped with a 50K single nucleotide polymorphisms (SNP) chip). The criteria used to exclude SNP were MAF < 0.05, call rate < 0.95 (for both SNP and animals), and a maximum difference of 0.15 between the observed and expected heterozygosity (Hardy-Weinberg equilibrium). After quality control, 31,182 SNP were used for the analyses. The analyses were performed using the BLUPF90+ programs. The pedigree-based heritability estimates were: ALL (0.10±0.017), PL (0.11±0.023), TLI (0.10±0.017), TNA (0.11±0.017), TNW (0.11±0.018), TLB (0.12±0.020), TLW (0.10±0.024), and TLWadj (0.07±0.031). The genomic-based heritability estimates were: ALL (0.13±0.018), PL (0.14±0.024), TLI (0.13±0.018), TNB (0.13±0.018), TNW (0.14±0.018), TLB (0.15±0.021), TLW (0.11±0.024), and TLWadj (0.07±0.032). The relative standard error (RSE) of the estimates obtained with the A matrix ranged from 15.5% (TNB) to 44.3% (TLWadj) and from 12.9% (TNW) to 45.7% (TLWadj) when fitting the H matrix. In conclusion, longevity-related traits in U.S. Katahdin ewes are heritable, and estimates are slightly greater (with lower RSE) when incorporating genomic information to estimate variance components.