Estimates of Genetic Covariance Functions for Growth of Nelore Cattle Assuming a Parametric Correlation Structure for Animal Permanent Environmental Effects

INTRODUCTION Random regressions (RR) on Legendre polynomial (LP) of age have been used to model longitudinal data such as milk production and weights at different ages. Usually such models involve high degree polynomials, especially for direct permanent environmental effects and, consequently, a large number of parameters to be estimated. High order polynomials are associated with sampling problems (Kirkpatrick et al., 1994) and high computational requirements. Parametric correlation functions combined with variance functions provide a more parsimonious alternative, and have been suggested to model within animal (co)variances (Foulley et al., 2000). This paper presents estimates of genetic covariance functions for growth of Nelore cattle, fitting a parametric correlation function to describe the covariance structure between animal, permanent environmental effects.