Estimating shrinkage, microfibril angle and density of Eucalyptus wood using near infrared spectroscopy

摘要

Developing calibration models from near infrared (NIR) spectra for quickly and reasonably estimating key wood traits has become a prime step in tree breeding programmes. While the optimal procedures for the calibration of NIR-based models for chemical-related wood traits, stiffness and wood density of wood are relatively well established, relatively few studies have dealt with NIR calibrations for wood shrinkage and cellulose microfibril angle (MFA) of cell walls. MFA has major effects on wood stiffness and shrinkage; however, this ultra-structural cell wall feature has been under-utilised by breeding programmes due to the difficulty in measuring it. The aim of this study was to develop NIR models for estimating radial and tangential shrinkage and MFA in Eucalyptus wood. The calibration for MFA provided NIR predicted values suitable for making a selection of the most favourable trees for a given end-use in breeding programmes. The correlation (r_p~2) between NIR predicted and X-ray diffraction derived values of MFA was 0.75 and the root mean standard error of the predictions was 1.3 degrees. These estimates appear to be adequate for selecting potential trees in genetic investigations. Calibrations were also established for predicting the dimensional stability of the wood. However, the statistics for shrinkage indicated that the models were not satisfactory as they provided NIR estimated values of limited applicability. Further studies would be useful for establishing adequate protocols for NIR calibrations of these important (but complicated to measure) wood characteristics.