Statistical strategies design based on competition classes of Eucalyptus clones

Abstract

A detailed knowledge of different types of trial designs is essential to establish adequate Eucalyptus breeding strategies. This study compared clone’s development in single-tree-plot (STP) and square-plot (SP) experiments to analyze differences in estimated genetic parameters, ranking, predicted genotypic value and competition ability. Experiments were carried out in the CMPC Celulose Riograndense Company, in the state of Rio Grande do Sul, Brazil. A total of 239 clones were used, 214 of which were common to the STP and SP trials. Results showed that STP is a very promising trial design for accurately ranking the genotypes. The STP and SP tests showed high coincidence in ranking selection, but the best clone yields were overestimated in the STP trial and the yields of the worst clones were underestimated in the STP trial. Therefore, an STP design should be used in initial and intermediary clonal tests, and an SP design should be used in the final stages of a breeding program, in order to estimate the clonal yield at a semi-operational scale. The estimated yield decreasing of clones planted in SP compared to STP is 26%. This value should be used in future STP experiments in the same area and with the same germplasm evaluated in this work. In addition, this research identified classes of aggressive clones, sensitive clones, and clones that were homeostatic to the competitiveness effect. Based on this classification, the optimized multiclonal plantation were suggested aiming to maximize yield by using aggressive and homeostatic clones.