ABSTRACT
In this work, with the help of multivariate statistical techniques, it was studied the influence of morphological and physical soil factors and litter nutrient content on site classification of a 47 years old Araucaria angustifolia (Bert.) O. Ktze. stand located on Canela National Forest, Rio Grande do Sul State, Brazil. A stratified sampling, covering three natural sites that received four angle-count sampling (ACS) each one, according to the methodology of BITTERLICH (1984) were installed. In the area of the ACS the tree with the average dominant height was felled for the trunk analysis and determination of the height growth, as defined by Pollanchütz, and soil samples at deeps of 0-10 cm and 10-20 cm were collected with am extractive cylinder. These samples were equidistantly distributed under the crown projection of the average dominant tree. Later, each sample was divided into two geometric horizons, A1 e A2, from which the physical characteristics of the soil were determined. Litter samples were obtained and their chemical content was analyzed. The cluster analysis classified the group of observations into three distinct groups, according to their locations in the different natural sites. The discriminant analysis confirmed the efficiency of the groupings and indicated, with a significance level of 0.01, that the variables soil depth, contents of clay, fine sand and loam, soil density and height mean annual increment presented a higher discriminating power of separating the groups. The first one corresponds to the site of higher yearly mean growth in height and it is located on the plains and low hillsides, where soils, are deep, loamy and sandy. The second, correspond to the lower height growth sites, located on intermontane plains which occurs in the form of irregular lowlands where soils are deep, clay, compacted and with rock residuals along its profile. The third, with yearly mean growth in height intermediary when compared with the two first ones, enclose the most steepests areas of the study with based rocks.
Key words:
Araucaria angustifolia; growth; soil; multivariate analysis