Use of Random Regression Models to Estimate the Heritability of Weight Growth in Meat Quails

Abstract

The aim of this study was to estimate the best covariance function structure, by means of Legendre polynomials using random regression models (RRM) to describe the genetic variability of the weight growth trajectory of quails and to estimate the genetic parameters. Body weight records of animals at 7, 14, 21, 28, 35, and 42 days of age were used. The orders of fit of the polynomials were gradually increased in orders from 3 to 6 for comparison between the models the Akaike information criteria were used. The models included fixed effects of the contemporary group, the fixed regression coefficient of weight on the Legendre polynomial, random regression coefficients of the direct additive genetic, and the permanent environment effects of the animal, in addition to the residual random effect. The RRM with orders five for the additive genetic effect and six for the permanent environment effect, under the assumption of homogeneity, was the most suitable to describe the genetic variability of the birds’ ponderal growth. There was greater expression of additive genetic variability after 21 days, resulting in distinct and increasing heritability estimates between two periods in ponderal development. Estimates of additive genetic correlations for weights between the ages evaluated indicated that genetic associations are more correlated between body weights at closer ages when compared to ages more distant along the growth trajectory. Thus, it is concluded that selection in birds is more efficient from 28 days of age due to the higher heritability values.