A MULTIVARIATE APPROACH TO ANALYSE NATIVE FOREST TREE SPECIE SEEDS

• Autor(a) principal: Lúcio, Alessandro Dal Col
• Data de Publicação: 2015
• Outros Autores: Fortes, Fabiano de Oliveira, Storck, Lindolfo, Filho, Alberto Cargnelutti
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Cerne (Online)
• Texto Completo: https://cerne.ufla.br/site/index.php/CERNE/article/view/404
• topic: Cluster analysis; principal component analysis; storage

This work grouped, by species, the most similar seed tree, using the variables observed in exotic forest species of the Brazilian flora of seeds collected in the Forest Research and Soil Conservation Center of Santa Maria, Rio Grande do Sul, analyzed from January, 1997, to march, 2003. For the cluster analysis, all the species that possessed four or more analyses per lot were analyzed by the hierarchical Clustering method, of the standardized Euclidian medium distance, being also a principal component analysis technique for reducing the number of variables. The species Callistemon speciosus, Cassia fistula, Eucalyptus grandis, Eucalyptus robusta, Eucalyptus saligna, Eucalyptus tereticornis, Delonix regia, Jacaranda mimosaefolia e Pinus elliottii presented more than four analyses per lot, in which the third and fourth main components explained 80% of the total variation. The cluster analysis was efficient in the separation of the groups of all tested species, as well as the method of the main components.