97. Genetic Parameters for a Longitudinal Longevity Indicator Trait in U.S. Katahdin Ewes.

Abstract

Genetic selection for ewe longevity can enhance flock productivity but is challenging as it is sex-limited and expressed late in life. The Katahdin is a hair breed noted for parasite resistance and maternal performance and was the first in the U.S. to adopt genomic selection. This study estimated genetic parameters for longevity across different age-at-lambing periods in Katahdin sheep. Longevity was defined longitudinally and cumulatively: a value of 1 was assigned each period a ewe lambed, and values were summed across periods. The final value represents the total number of litters over 14 periods, spanning 9–169 months, grouped in ∼8-mo intervals. This interval reflects that accelerated breeding in sheep can result in up to three lambings over two years. After quality control, 14,983 ewes born between 1989 and 2021 from 98 flocks were considered. Ewes were daughters of 1,536 sires and 7,573 dams and produced 72,636 lambs in 41,348 litters between 1992 and 2024. Only ewes with complete lifetime records were included. A single-trait random regression model using third-order Legendre polynomial was fitted. Variance components were estimated using 500,000 iterations, 200,000 burn-in, and a thinning interval of 10. Convergence was checked via Heidelberger Welch and Geweke diagnostics and visual inspection. The statistical model was: y = Xβ + Hα + zμ + wδ + ε, where y is the vector of observations of cumulative longevity for each age, β is the vector of systematic effects, including combined birth and rearing type, categories of age at first lambing, and year-season as a fixed regression curve, α is the vector of random regression coefficients for the contemporary group (flock–year–season), μ is the vector of random regression coefficients for direct additive genetic effects, δ is the vector of random regression coefficients for permanent environmental effects, and 𝛆 is the vector of residuals. The incidence matrices X, H, Z, and W link y to β, α, μ, and δ, respectively. Heritability estimates for different ages were low (Table 1), but increased gradually from 0.04 in the earliest age-at-lambing group (9 to 16 mo) to ≥ 0.11 in later periods (65 to 169 mo). The heritability estimates (± Posterior Standard Deviation, PSD) for ewe longevity traits ranged from 0.044 ± 0.008 to 0.113 ± 0.011 (Table 1). The lower heritability values at earlier ages were associated with proportionally lower additive genetic variance. From a practical standpoint, overlapping highest posterior density HPD intervals indicated that heritability estimates were similar from 33 to 169 mo. Additionally, genetic correlations among age-at-lambing periods beyond 2.75 yr were strong (0.94 to 1.00). Thus, random regression models allow the inclusion of animals still in production, enabling selection for longevity in ewes younger than three years.

For image description, please refer to the figure legend and surrounding text.