Modelos de afilamento para Pinus taeda L. ajustados segundo a forma do tronco e métodos de estratificação

• Ano de defesa: 2009
• Autor(a) principal: Souza, Carlos Alberto Martinelli de
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; BR; Recursos Florestais e Engenharia Florestal; UFSM; Programa de Pós-Graduação em Engenharia Florestal
• Palavras-chave em Português: Precisão; Modelos segmentados; Modelos não segmentados; Dendrocronologia; Sortimentos; Accuracy; Segmented models; Non-segmented models; Dendrochronology; Assortments; CNPQ::CIENCIAS AGRARIAS::RECURSOS FLORESTAIS E ENGENHARIA FLORESTAL
• Link de acesso: http://repositorio.ufsm.br/handle/1/3730

This study was carried out in order to analyze the performance of segmented and non-segmented taper models in terms of estimates for trees profiles with 1, 2 and 3 shape change points in stems; to test the accuracy of estimates of stem profiles and commercial height based on an adjusting taper model with non-stratified data, stratified in classes of DBH (diameter at breast height) and stratified in classes of shape quotient; and to examine the influence of the results on the formulation of a classification table for wood assortments. The data to this study were measured in Pinus taeda L. stands from Klabin S.A. in Telêmaco Borba-PR. The sample trees grew up in stands managed to sawmill, with two thinnings and coppicing, plantation of 1.600 trees/ha, spacing of 2,50 x 2,50 m. Forty trees were cut down and cubed by the Smalian method in the following positions: 0,10, 0,30, 0,80, 1,30, and in each meter until the total height. In these positions, stem transversal slices were obtained to study the reconstruction of growth using dendrocronology and it generated a database of 649 trees aged between 8 and 30 years, among which the samples to this research were selected. Aiming at the best model to estimate the stem profiles, for all kinds of trees the non-segmented models were superior, the 5th degree polynomial showed a better performance and the Garay model presented the second best result. Considering the best method to illustrate the data, the stratification by classes of shape quotient was more accurate, emphasizing K0,7h and K0,5h classes. The assortments table proved that the higher the trees dimensions, the higher the possibility of obtaining more products and, consequently, the better the performance, with a lower residual value. The validation also showed the accuracy of the models adopted in the assortments table.