Genetic parameters for test-day fat yield estimated by random regression models in dairy buffaloes using bayesian inference

Abstract

This study modeled variations in test-day fat yield of first lactation of Buffaloes cows by random regression model (RRM) fitted by Legendre orthogonal polynomials (LOP) compared to 3 alternatives models fitting B-splines. A total of 10691 monthly test-day fat yield records of 1388 first lactations from buffaloes of the Murrah breed born between 1985 and 2007 from 12 herds in the state of São Paulo, Brazil, were used. The fixed effects common for all models were the contemporary group, defined as the herd-year-month or herd-year and calving season of the test day, numbers of milkings per day (two levels), the covariable dam age at calving (linear and quadratic effects) and the average trend of fat yield was modeled by quartic LOP or while using b-splines, cubic LOP. Estimates of (co)variance components were obtained by a Bayesian framework, applying an animal model, through Gibbs Sampling. The residual variances were grouped in ten classes. The random additive genetic and permanent environmental effects were modeled by cubic and quadratic Legendre orthogonal polynomials, respectively, or using linear b-spline functions with 3 to 6 knots. The heritability estimates were moderate (0.24±0.04), ranged from 0.17 to 0.4. Heritability estimates increased at begin and the end of lactations. According to the deviance information criteria (DIC), the best overall performance for both the additive genetic and permanent environmental effects for fat production was that with three knots located at 5th, 60th, 305th days of lactation. The model which considered Legendre orthogonal polynomials were the worst model. B-spline functions could be applied as an alternative to Legendre polynomials to model covariance functions to test-day fat production.