Bayesian analysis for estimation of genetic parameters of calving ease and stillbirth for Canadian Holsteins

Abstract

A total of 129,765 calving ease and stillbirth records of Canadian Holsteins were analyzed. Bayesian analysis was applied to a mixed linear sire and maternal grandsire model, and Gibbs Sampling was used to obtain posterior densities of variance components. Both traits were recorded in ordered categories. For calving ease, scores were transformed to Snell scores, for which higher numbers indicated easier calving. Live births were scored 1, and stillbirths were scored 2. Low heritability (combining both direct and maternal effects) estimates were obtained for calving ease (0.07) and stillbirth (0.09). The specific heritabilities for direct and maternal calving ease and for direct and maternal stillbirth were 0.05, 0.03, 0.04, and 0.06, respectively. Negative correlations were obtained between direct and maternal components for calving ease (-0.16) and stillbirth (-0.24). The genetic correlation between direct calving ease and direct stillbirth was high (-0.59), but the genetic correlation between maternal calving ease and maternal stillbirth was slightly lower (-0.34). The genetic correlations between direct calving ease and maternal stillbirth (0.06) and between maternal calving ease and direct stillbirth (0.04) were positive but very low. The genetic (maternal plus direct), phenotypic, and residual correlations between calving ease and stillbirth were -0.47, -0.19, and -0.16, respectively.