Biometricians have been studying tree shapes since the 18th century in Europe, initially using shape quotients. With the advent of linear regression techniques, the field evolved toward the development of volumetric models, which incorporate the effects of tree shapes into the coefficients of the fitted equations. This study presents two methodological themes: the volumetric equivalence form factor (VEFF) based on Pollanschütz concept, and the rational form factor (RFF) derived from ratio estimation. Data from 159 trees were used to fit the RFF, revealing a strong correlation with the h/d variable, indicating that the slenderness index effectively expresses tree form. Five form factor models based on shape quotients were tested, producing reliable estimates with standard errors below 10% and no bias in residuals. Among the fitted volume models, Meyer, Meyer (Mod.), Näslund, Schumacher-Hall, and an additional form factor from the Dwight volume equation were selected for their accuracy and reliability. Other volume equations, such as Takata, Stoate, Dwight, and Péllico Netto and Behling, also showed good results. The second alternative solution for the RFF provided highly reliable estimates with minimal variation in the form factor among the sampled trees. The selected volume equations and their form factors, chosen based on statistical estimates and residual evaluation, are applicable to Pinus taeda L. stands in the Midwest of Santa Catarina, Brazil.
Read full abstract