Silvicultural performance of Parkia multijuga Benth. progenies in the Amazonas three years after planting

Progeny tests are strategies for ex situ genetic conservation and genetic improvement, as they allow the selection of genotypes of better quality and high productivity. However, the tree genetic improvement has not been properly applied to native species of the Amazon region. We investigated if Parkia multijuga progenies have different silvicultural performance and if it is possible to cluster them into different performance classes. The experiment was settled in a complete random block design with 14 progenies, a total of 18 seedlings from each progeny planted in six blocks (252 plants). Three years after planting, the following variables were evaluated: diameter at breast height (DBH), total height, stem height, crown diameter, crown length, crown area, crown ratio, slenderness (S), range index (RI), salience index (SI), bifurcation occurrence, percentage of individuals with rectilinear stem and phytosanitary status. Survival, mean annual increment in height and diameter and the integrated response index (IRI = survival and increments) were also calculated. IRI varied about three times between the best and the worst progeny, of which the progenies 7, 6, 11 and 10 had the highest IRI values. The Principal Component Analysis of the most interesting silvicultural characteristics (IRI, DBH, GE, rectilinear stem, bifurcation, and phytosanitary status) allowed to group progenies 6, 7 and 11 among those with the best silvicultural performance. Most progenies showed intermediate and similar performance, being the worst performance for most traits related to the progeny 2. Progenies of Parkia multijuga diverge as to the silvicultural performance during the initial establishment of the plantations. Therefore, the grouping of progenies with better performance is fundamental in the selection of superior genetic material for the composition of productive plantations of this species in the Amazon region.