326 Genetic evaluation of lamb survival in Canadian sheep using a random regression model.

Abstract Improving lamb survival is essential for sustainable sheep production, particularly because high prolificacy can lead to increased mortality. The objective of this study was to implement a random regression model to evaluate lamb survival (LS) and estimate genetic parameters across different survival ages, from 1 to 50 days (weaning age). The dataset analyzed included 1,951,084 animals, progeny of 409,304 ewes and 35,118 rams, raised in 2,208 flocks across Canada between 1986 to 2024. A single-trait Bayesian random regression model (RRM), which included the systematic effects of year-month of lambing and age of the dam-class of sex of lambs born, and the random effects of flock-year-management group, litter of birth dam, direct animal additive genetic, animal permanent environment, maternal genetic, and maternal permanent environment. Third-degree Legendre orthogonal polynomials were used to fit the longitudinal direct animal additive genetic, animal permanent environment, maternal genetic random effects and year-month of lambing systematic effect. A homogeneous residual variance across ages was assumed. The average heritability estimates for LS across ages ranged from 0.15 to 0.20 and were consistently higher than when analyzing survival as a categorical trait with five categories (heritability= 0.05), i.e. death at birth: mummified (1) or stillborn (2), between birth and 10 days (3), between 10 and 50 days (4), and survival until weaning at 50 days (5), which is currently the model used to evaluate LS in the Canadian sheep evaluation. Maternal heritability estimates were high across all ages (ranged from 0.52 to 0.53), showing a significant and consistent genetic maternal influence on survival across all ages. This highlights the importance of considering maternal traits in breeding and management strategies to improve lamb survival. The estimated direct genetic correlations across ages ranged from moderate to high (0.75 to 0.90), while the maternal genetic correlations were slightly higher (0.88 to 0.95). These findings suggest that LS between 1 and 50 days of age has a common genetic background indicated by similar heritabilities and moderate to high genetic correlation across ages. Genetic selection based on RRM breeding values is expected to yield greater selection responses for improving lamb survival in Canadian sheep compared to the current evaluation based on a categorical LS trait.