Analysis of prediction factors in yields by maize selection (<em>Zea mays</em> L.)

Abstract

The history and use of the Factor Analysis in agronomic studies, especially for the prediction of selection yields, is presented. The Factor Analysis was applied on data provided by the evaluation of 166 half-sib families of popcorn at the National Research Center in Maize and Sorghum, Sete Lagoas MG Brazil, in 1997-1998. Selection yield prediction based on scores from the Factor Analysis favored yields in the most important traits. Scores were used for yield prediction by direct and indirect selection and by several selection indices. The Factor Analysis is a useful multivariate method for prediction of selection yields and may either substitute the traditional method or supplement selection by selection indices