Parâmetros genéticos e curva de crescimento para características de desempenho de equinos do Exército Brasileiro

Abstract

This study aimed in the Article 01, to compare the magnitude of the coefficients of heritability and genetic correlations estimated by two different methodologies, and to estimate the genetic and phenotypic trends for the characteristics of height at the withers (AC24) and weight at 24 months of age (P24). In Article 02 the objective was to identify among the various mathematical functions mentioned in the literature, which of them best describes the growth curve for the Brasileiro de Hipismo equine breed Brazilian Equestrian (BH) and for animals without definite breed (SRD) for a population of horses belonging to the Brazilian Army. The estimation of variance components to estimate genetic parameters and predict breeding values, was performed by the Derivative-Free Restricted Maximum Likelihood Method (REML) and by Bayesian Inference Method (BIM). The average heritability estimates obtained by BIM for AC24 (0.54) and P24 (0.41) presented higher values than those obtained by REML (0.37 and 0.52, respectively). The regression of average additive genetic merit on the year of birth, showed that the average breeding values for AC24 presented a slight downward trend over the years, with virtually no genetic tendency, the average breeding values for P24 animals showed oscillatory behavior, with negative genetic trend of -0.38 kg. The genetic correlation between traits, high and positive, suggest that selection for AC24 will increase P24 at the same age. Estimated genetic trends indicate that the selection, which aims to reduce P24 and increase AC24 in this population, has been relatively successful in reducing the P24, however, it is not being efficient to increase AC24. In Article 02 the values obtained for Absolute Mean Deviation were low, ranging from 0.01 to 0.18 for the BH breed animals and from 0.02 to 0.17 for SRD horses, revealing the occurrence of small variations, being lower for the Brody (1924) and Bianchini Sobrinho and higher for the Papajcsik & Bodero (1988) model. The values obtained for the Residual Mean Square were low, ranging from 0.002 to 0.089 for BH breed animals and from 0.002 to 0.042 for SRD animals. Nonlinear models of Brody (1924) and Bianchini Sobrinho were the ones that provided the best fit for the wither height characteristic for BH breed and for SRD animals in a Brazilian Army population.