Efeito da inclusão da covariância genética direta-materna, na análise para estimar parâmetros genéticos e predizer valores genéticos para ganho de peso em bovinos de corte.

Abstract

The present study has as objective to evaluate the effect of the inclusion in the model of analysis, the covariance between the additive genetic direct and maternal effects, on the estimation of genetic parameters for average daily gain from birth to weaning (GMDND) and from weaning to 550 days of age (GMDDS), and on the prediction of genetic values and rank of the animals, for Angus and Brangus breeds. The data were furnished by Gensys Consultores Associados S/C Ltda. and Natura Genética Sul Americana, and contained records from animals created in many farms, in different regions of Brazil and they were collected from 1968 and 2002. They were estimated genetic parameters and predicted genetic values for average daily gain from birth to weaning (GMDND) and from weaning to 550 days of age (GMDDS), for the two populations. In paper 1, they were studied 11,202 and 4,665 records of Angus breed animals, respectively for GMDND and for GMDDS. The genetic direct and maternal heritability coefficients and genetic values were obtained using the (co)variance components obtained by Restricted Maximum Likelihod Method by Boldman et al. (1995). It was adopted an animal model, considering GMDND as a function of the random effects additive genetic direct, maternal and residual and of the fixed effects of contemporaneous group at weaning (GC205), and the covariables, age at weaning (ID) and age of the cow at parturition (IV), linear and quadratic effects; for GMDDS, the model was the same, only substituting the fixed effects GC205 by contemporaneous group at 550 days of age (GC550) and ID by age at 550 days of age (IS). Each model was used for two analyzes, one do not including the genetic-maternal covariance, equal to zero and the other one, including the covariance previously estimated. I the two analysis, it was included in the model the permanent environment effect of the cow. The animals were ranked according to their genetic values in the two analysis (with and without the inclusion of the covariance between the additive genetic direct and maternal effects) with the objective of to evaluate the correspondence among the ranks getting from the different models. The genetic direct heritabilities were .21 and .55, and the maternal ones were .00 and .22, respectively, including and do not the covariance in the model. The correlations between genetic direct and maternal effects were negatives for GMDND and for GMDDS. It is recommended the inclusion of the covariance in the model, for GMDND because it can change the value of the estimative. The correlation between the ranks of the genetic values (.88 P<0,0001) suggest to occur an alteration in the ranks of the animals for GMDND. In paper 2, for Brangus breed, it were studied 28,949 and 11,884 records, respectively for GMDND and GMDDS. The components of (co)variance used to estimate additive direct and maternal heritabilities and to predict the genetic values were obtained by Restricted Maximum Likelihood Method and the computational program MTDFREML by Boldman et al. (1995). The animal model adopted for GMDND, considered as random, the additive genetic direct and maternal and residual effects, and as fixed, the effects of contemporaneous group at weaning (GC205) and the interaction of the percent of Nellore blood of the bull and cow (INTV), and the covariables age of the cow at parturition (ID) and age at weaning (ID). It was used as random, the permanent environmental effect of the cow. Each model was used for two analysis, one considering the genetic direct and maternal covariance effects, previously estimated and the other considering it equal to zero. The animals were ranked according to their genetic values to evaluate the correspondence of the ranks generated by different models. The genetic direct heritabilities were .14 ± .03 and .21 ± .01 and the maternal ones were .00 ± .01 and .15 ± .02, respectively including and do not, the covariance. The correlations between additive genetic direct and maternal effects were negatives for GMDND (-.25 ± .12) and for GMDDS (-.77 ± .19). It is recommended the inclusion of the covariance in the model because that can change the values of the estimative. The correlation, for GMDND (0,89 P<0,0001), suggest that occur light changes in the ranks of the animals.