Genetic Evaluation and Ranking of Different Animal Models Using Mallows C_p Statistic for Post Weaning Growth in Tuli Cattle of Zimbabwe

Abstract

An animal model utilizes all relationships available in a given data set. Estimates for variance components for additive direct, additive maternal, maternal environmental and direct environmental effects, and their covariances between direct and maternal genetic effects for post weaning growth traits have been obtained with different models and models were ranked for suitability using Mallows Cp statistic. Variance components for additive direct, additive maternal, permanent environmental maternal effects, the covariance between additive direct and maternal effects were estimated by restricted maximum likelihood, fitting five animal models from 3112 yearling records of indigenous Tuli cattle of Zimbabwe. All investigated models included a random direct genetic effect, but different combinations of random maternal genetic and permanent environmental effects as well as for direct-maternal genetic covariance. The direct heritability (h²_a) ranged from 0.00 to 0.18, when the maternal genetic effects were included in the model, while (h²_a) estimates were 0.48 to 0.75, when maternal effects were excluded. Ignoring maternal effects resulted in high estimates of (h²_a). The maternal heritability (h²_m) was 0.30 when only maternal genetic effects were included in the model and were 0.10 and 0.18 when the permanent environmental effect of the dam was fitted. The permanent environmental effects of the dam were around 5% of the total phenotypic variance and a small and positive covariance between direct and maternal genetic effects (σ²_am) was observed. The difference between direct and maternal heritability for yearling weight were not obvious, indicating that it may be worthwhile to utilize the maternal genetic variation for accurate ranking of animals.