Serviços Personalizados
Journal
Artigo
Indicadores
Citado por SciELO
Acessos
Links relacionados
Similares em SciELO
Compartilhar
Agrociencia
versão On-line ISSN 2521-9766versão impressa ISSN 1405-3195
Resumo
RAMIREZ-VALVERDE, Rodolfo et al. Univariate vs multivariate analyses in genetic evaluation of growth traits in two cattle breeds. Agrociencia [online]. 2007, vol.41, n.3, pp.271-282. ISSN 2521-9766.
The prediction of genetic values of animals for variables with sequential selection requires the identification of statistical models that maximize the response to selection. The objective of the present study was to compare the use of univariate and multivariate analyses of growth variables in the genetic evaluation of Angus and Tropicarne cattle. The variables analyzed were the weights at birth (BW), at weaning (WW) and at one year (YW), obtained from the genealogical (n=9933) and productive records of the Asociación Angus Mexicana, and the Asociación Mexicana de Criadores de Ganado Tropicarne (n=5724). The genetic evaluations were carried out using an Animal Model, with univariate and bivariate analyses in both populations, and trivariate analysis in Tropicarne. The comparison criteria were: magnitude of the estimators of genetic parameters, prediction ability of the models, and size of the accuracy of the genetic values predicted for BW, WW, milk and YW; in addition, the effect of the alternatives of analysis in the ranking of animals was evaluated. For some variables, the use of bivariate analyses improved the estimation of the direct and maternal heritabilities, the accuracy of the predictions of the genetic values, and the prediction ability of the models, but no advantages were observed when using trivariate analyses. The use of multivariate analysis with respect to univariate analysis changed the ranking of animals in nearly all of the variables, which suggests the need to evaluate the most appropriate alternative of analysis for population and specific variables. In Angus, individual analysis is recommended for BW, and bivariate analysis for WW and YW; whereas in Tropicarne, for BW, bivariate analysis of BW and WW is recommended, and for WW and YW, the bivariate analysis of both traits.
Palavras-chave : Univariate and multivariate analyses; Angus; beef cattle; genetic parameters; Tropicarne; genetic value.