The objective of this work was to evaluate the efficiency of the Z index (sum of standardized variables index), in the selection of full-sib progenies between Eucalyptus grandis and Eucalyptus urophylla via least square method (LSM) and best linear unbiased prediction (Blup) procedures. The experiments were carried out in two soil types and environments, in a randomized complete block design, with five replicates, eight plants per plot and 3x2 m spacing. Wood average annual increment, basic density, screened pulp yield, and effective alkali were evaluated for three-year-old plants. The Z index was obtained for each plant in order to compare the efficiency of selection for the best progenies by LSM and Blup procedures. Subsequently, the predicted averages to Z index were obtained and ranked. For comparison, the classic index was used. The coincidence among the best progenies selected by classical and Z index was good. However, the Z index allows a graphic display that enables to check in which characters the progenie has a disability. There is no difference between the best progenies selected by the Z index via LSM and Blup.
Eucalyptus; genetic correlation; forest breeding; selection index
