Abstract

Near infrared (NIR) spectroscopy is developing as an advanced and non-invasive tool in the wood, wood products and forestry sectors. It may be applied as a rapid and cost effective technique for assessment of different wood quality parameters of timber species. In the present study, NIR spectra of heartwood samples of Tectona grandis (teak) were collected before measuring fibre morphological parameters (fibre length, fibre diameter and fibre lumen diameter)and main chemical constituents (cellulose, hemicellulose, lignin and extractives) using maceration and wet chemistry methods respectively. Multivariate partial least squares (PLS) regression was applied to develop the calibration models between measured values of wood parameters and NIR spectral data. Pre-processing of NIR spectra demonstrated better predictions based on higher values of correlation coefficient for estimation (R²), validation (R_cv²), ratio of performance to deviation (RPD), and lower values of root mean square errors of estimation (RMSEE), cross-validation (RMSECV) and number of latent variable (rank). Internal cross-validation was used to find the optimum rank. Robust calibrations models with high R² (>0.87), low errors and high RPD values (> 2.93) were observed from PLS analysis for fibre morphological parameters and main chemical constituents of teak. These linear models may be applied for rapid and cost effective estimation of different fibre parameters and chemical constituents in routine testing and evaluation procedures for teak.

© 2021 The Author(s)