Random regressions models to describe the genetic variation of milk yield over multiple parities in buffaloes

The objectives of this study were to estimate (co)variance functions for additive genetic and permanent environmental effects, as well as the genetic parameters for milk yield over multiple parities, using random regressions models (RRM). Records of 4,757 complete lactations of Murrah breed buffaloes from 12 herds were analyzed. Ages at calving were between 2 and 11 years. The model included the additive genetic and permanent environmental random effects and the fixed effects of contemporary groups (herd, year and calving season) and milking frequency (1 or 2). A cubic regression on Legendre orthogonal polynomials of ages was used to model the mean trend. The additive genetic and permanent environmental effects were modeled by Legendre orthogonal polynomials. Residual variances were considered homogenous or heterogeneous, modeled through variance functions or step functions with 5, 7 or 10 classes. Results from Akaike's and Schwarz's Bayesian information criterion indicated that a RRM considering a third order polynomial for the additive genetic and permanent environmental effects and a step function with 5 classes for residual variances fitted best. Heritability estimates obtained by this model varied from 0.10 to 0.28. Genetic correlations were high between consecutive ages, but decreased when intervals between ages increased.