Genetic parameters for growth traits of mediterranean buffaloes from Brazil, estimated by Bayesian inference

Abstract

Quantitative analysis of growth genetic parameters is not available for many breeds of buffaloes making selection and breeding decisions an empirical process that lacks robustness. The objective of this study was to estimate heritability for birth weight (BW), weight at 205 days (W205) and 365 days (W365) of age using Bayesian inference. The Brazilian Program for Genetic Improvement of Buffaloes provided the data. For the traits BW, W205 and W365 of Brazilian Mediterranean buffaloes 5169, 3792 and 3883 observations have been employed for the analysis, respectively. In order to obtain the estimates of variance, univariate analyses were conducted using the Gibbs sampler included in the MTGSAM software. The model for BW, W205 and W365 included additive direct and maternal genetic random effects, random maternal permanent environmental effect and contemporary group that was treated as a fixed effect. The convergence diagnosis was performed employing Geweke, a method that uses an algorithm from the Bayesian Output Analysis package that was implemented using R software environment. The average values for weight traits were 37.6 +/- 4.7 kg for BW, 192.7 +/- 40.3 kg for W205 and 298.6 +/- 67.4 kg for W365. The heritability posterior distributions for direct and maternal effects were symmetric and close to those expected in a normal distribution. Direct heritability estimates obtained using the modes were 0.30 (BW), 0.52 (W205) and 0.54 (W365). The maternal heritability coefficient estimates were 0.31, 0.19 and 0.21 for BW, W205 and W365, respectively. Our data suggests that all growth traits and mainly W205 and W365, have clear potential for yield improvement through direct genetic selection.