Use of Different Models for Estimation of Genetic Parameters and Genetic Trends of Performance Test Traits of Gilts

Abstract

The aims of this research were to compare estimates of variance components using different animal models and to determine the most suitable mixed model for estimating genetic parameters and genetic trends for traits in performance test of gilts of different breeds using REML. A total of 73129 gilts of four genotypes in the period of 2009 to 2013 were included in the analyses. Four mixed models were constructed. Information criterion of Akaike (AIC) and Bayesian information criterion (BIC) were used to suggest which model is an adequate model for evaluation of genetics parameters. With the introduction of certain factors in the models, reduction in components of variance and heritability in all studied traits was observed. Heritability traits in four genotypes and models were at medium to high degree of heritability. The resulting genetic trends were different between the models and the coefficients of determination (R2) were relatively high. Average gain and meat percentage were established positive (favourable) or negative (unfavourable) genetic trends in all models, while back fat thickness and lateral back fat thickness in all models were established to have positive (unfavourable) genetic trends. Based on the obtained results in this study, it is concluded that it is necessary to include mixed models in the estimation of breeding values in order to eliminate their influence, which significantly affects the variation of important traits for selection. In addition, with the inclusion of a greater number of parameters in mixed models, the models become more accurate and provide more accurate assessment of genetic and breeding value.