The objective of this work was to evaluate the possibility of using selection indexes in factorial complexes, in the simultaneous selection of characters of Coffea canephora (Conilon var.), when predicting the selection gains. Fourteen characters and 40 genotypes were analyzed in two experiments of the genetic improvement program of 'Conilon' coffee, developed by Incaper, carried out in Marilândia and Sooretama, ES, Brazil. The experiments were conducted on a randomized block design, with four replicates and two useful plants in each plot. Factorial complexes eliminated the multicolinearity problem, allowing the appropriate use of the selection index theory for simultaneous improvements of characters. The selection indexes by Smith and Hazel and Pesek & Baker were used, and the first one was the best alternative to obtain adequate simultaneous gains. The combined use of multivariate techniques of the factor analysis, for previous structural simplification in the number of characters, and selection indexes to predict simultaneous gains are efficient alternatives in the coffee cropping genetic improvement.
coffee; plant breeding; selection gains; multivariate analysis
