Lumber yield of Sucupira-Vermelha in central Amazon sawmill

Abstract

The sustainable use of forestry resources has an intrinsic relationship with log yield in sawmills, because the knowledge of this variable can optimize the processed wood production. Thus, the objective of this work was to determine the percentage yield in the sawing of Sucupira-Vermelha (Andira parviflora Ducke.) logs in different diameter classes and to adjust mathematical models to estimate processed wood. It was sampled 56 logs with diameters between 31 and 69 cm grouped into 4 diameter classes. Afterwards the sawing was performed obtaining various pieces of wood which were cubed to determine the yielding percentage. The increase in diameter classes provided an increase in sawn wood yield, which showed a significant difference only between the diameter class 1 and the others studied. Twelve mathematical models, 6 linear and 6 nonlinear, were evaluated. The models quality adjustment and selection was based on the highest adjusted determination coefficient (R²adj), lowest estimation standard error (Syx) and variation coefficient (VC%) and, most homogeneous residues graphic distribution. The nonlinear models of single entry (diameter) and double entry (diameter and length) were the most accurate to estimate the sawn wood yield of Sucupira-Vermelha species.