Prediction of genetic gains in forest improvement

Abstract

Total gains predicted by various selection criteria were evaluated, based on six characteristics of growth in half-sib families of Eucalyptus camaldulensis. The simulated situation consisted of selection of 25% among 44 families studied and of 17% of the plants in these families, totalizing a selection of 4.25% of the constant individuals of the essay. The selection criteria used were: direct and indirect selection; classical index of Smith & Hazel, with four economic weights, and index of Pesek & Baker, in three situations. The criteria used were similar to each other, with a slight superiority for the classical index when the genetic variation coefficient associated with each analysed characteristic was established as the vector of economic weights.