Avaliação da influencia dos fatores sitio, idade, densidade e posição sociologica na relação hipsometrica para Pinus taeda nas regiões central e sudoeste do Estado do Parana
| Autor: | Cardoso, Denise Jeton |
|---|---|
| Přispěvatelé: | Universidade Federal do Paraná. Setor de Ciencias Agrárias. Programa de Pós-Graduaçao em Engenharia Florestal, Machado, Sebastião do Amaral, 1939 |
| Jazyk: | portugalština |
| Rok vydání: | 2013 |
| Předmět: | |
| Zdroj: | Repositório Institucional da UFPR Universidade Federal do Paraná (UFPR) instacron:UFPR |
| Popis: | The main objective of this research was to analyze the effects of age, site, density and sociologic position on the height-diameter relationship The data carne from the measurement of height and diameter of trees in 137 plots of Pinus taeda, with area of 600 m2, scatered in the State of Parana, Brazil, with ages ranging from 4 to 27 years old. The comparison among regression lines was done through the analysis of covariance. The comparison of means of values h/d was also done through statistic tests. Generic equations including age, site, density and sociologic position as independent variables were also tested a secondary proposed objective was to test the ideal number of trees (pairs of values h/d) for the fitting of the hipsometric relationship. The criteria for the selection of regression models in all phases of this research were the standart error of estimate (syx, syx%), and the residuals distributions. In a preliminary analysis they were selected four regression models (Stoffells, Curtis, Henricksen and Prodan), which were also used for the analysis of the other factors. To analyze the effect of age it was opted for a modal which included the reason height over dominant height (h/hdom =b0 + b1 In d), reducing considerably the standard error of estimate, In a general way, the covariance analysis detected a greater influence of age and site on the intercept (b0) of regression lines. Refering to the inclination coefficient (b1), the influences of density seemed to be stronger, producing steeper regression lines for the greater densities for all analyzed age classes. Refering the sociologic position for the three considered diameter classes, as dominated, mean and dominant, it was detected significant differences among them. The relationship h/d presented decreasing values from the dominated to the dominant diameter class. It was verified that 31 to 40 pairs of values h/d were better to fit the height-diameter equation. The use of a generic equation for the used set of data, seemed quite liable. From the four tested models the one that deserved more attention was: ln h = ln hdom + b0 + b1 ((1/d) - (1/dmax)), with standard error of estimate in percent equal to 6,97%, and presenting a good residual distribution. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|