Revista Brasileira de Zootecnia
On-line version ISSN 1806-9290
PIRES, Aldrin Vieira et al. Evaluation of models to estimate (co)variance components in reproductive traits of swine. Rev. Bras. Zootec. [online]. 2000, vol.29, n.6, pp. 1684-1688. ISSN 1806-9290. http://dx.doi.org/10.1590/S1516-35982000000600013.
Reproductive traits on three swine breeds, Duroc, Landrace and Large White, were used to evaluate the importance of inclusion of maternal genetic and the permanent environmental effects in animal models were analyzed. The traits for litter size at birth, litter size at weaning, litter weight at birth, litter weigh at 21 days, and loss rate were evaluated by the restricted maximum likelihood (REML) method, and one likelihood value for each trait and breed were obtained. The likelihood ratio test was used to indicate the best model in the animal genetic evaluation. The models were: model 1 contained the direct genetic effect, model 2 contained the direct and maternal genetic effects, model 3 contained the direct and common litter effects, and model 4 included direct, maternal and common litter effects. Different models were indicated for different traits and breeds, but the model 4 was more adapted for most of the traits and breeds. These results evidenced the importance of the inclusion of maternal genetic additive and common litter effects in the evaluation of swine litter traits, to estimate, with more precision, the genetic parameters of these herds.
Keywords : animal model; likelihood ratio test; maternal effect; REML.