The aim of this study was to perform ecological and functional clustering of tree species in southwestern Amazon. Developed from data from 95 permanent plots of 1 ha each, all individuals with diameter at breast height (DBH) a?￥10 cm were measured. The species grouping was performed in three stages: (1) cluster analysis, using the variables: diameter annual periodic increment -(IPA DAP ) considering three competition levels (high, medium and low) and the 95th percentile of the diameters (P 95 ) cumulative frequency distribution (Ward hierarchical method); (2) Discriminant analysis, using the variables P 95 and IPA DAP by Fisher's method and (3) subjective stage, considering the species ecological characteristics. The Ward and Fisher methods used for discriminant and cluster analyses were effective for species grouping resulting on the formation of 10 groups. Variables: IPA DAP and and P 95 were efficient on the formed groups discrimination.. Variations in the growth rates for the overall mean data were reduced wen calculated for each group of species.
展开▼
