Caracterização da madeira de híbridos de Corymbia via espectroscopia no infravermelho próximo (NIRS)

Abstract

The wood of Corymbia shows promise in the forestry sector due to its resistance to biotic and abiotic factors and its high wood basic density. However, detailed analysis of its quality is necessary for industrial application, which can be facilitated by non destructive evaluation methods. Although near-infrared spectroscopy (NIRS) is effective in this regard, there is a shortage of studies applying this technique to Corymbia wood. Therefore, this study aimed to apply NIRS as a tool for assessing the quality of hybrid Corymbia clone materials. Non-destructive wood chip samples were collected from 1059 three-year-old trees of 34 hybrid Corymbia clones at breast height (DBH). These samples were analyzed using a NIR spectrometer, and based on the spectra, 100 trees were selected to determine reference values for basic density, lignin, extractives, holocellulose, and ash. The models were calibrated using partial least squares regression analysis (PLS-R) and evaluated according to validation parameters. The developed models for basic density (R²cv=0.71; RMSECV=19), total lignin (R²cv=0.63; RMSECV=0.58), total extractives (R²cv=0.72; RMSECV=0.60), holocellulose (R²cv=0.61; RMSECV=1.05), and ash (R²cv=0.42; RMSECV=0.22) demonstrated satisfactory performance. Model predictions highlighted clones 11 and 23 as promising for cellulose production due to their low lignin and extractive content, combined with high wood basic density. Meanwhile, clones 12, 15, 16, and 17 were considered suitable for charcoal production due to their high basic density and high lignin content. The study demonstrated that NIRS is a valuable tool for non-destructive analysis of Corymbia wood quality, especially in large tree populations, providing relevant information for the forestry sector.