Estimating Wood Specific Gravity of <i>Ravenala madagascariensis</i> Sonn. Using Near-Infrared Spectroscopy

Abstract

Near InfraRed Spectroscopy (NIRS) has emerged as a promising non-destructive method for wood analysis. In this study, the efficacy of NIRS in predicting the wood specific gravity (WSG) of Ravenala madagascariensis, an endemic non-woody species of Madagascar was assessed. The optimal model, employing "SNV (standard normal variate) + DT (detrending)" pre-treatment and utilizing 11 latent variables, exhibited interesting performance metrics, including an RMSEcv of 0.013 g.cm-3, R²cv of 0.73, and RPDcv of 2.76. Additionally, in independent validation, the model achieved an R² of 0.70 and an RPD of 2.17, with 11 numbers of latent variables. The predictive model's application unveiled significant radial variability in WSG within Ravenala madagascariensis. Specifically, the central zone exhibited lower density (average of 0.082 g.cm-³) than the peripheral zone (0.12 g.cm-³), with a highly significant difference (>0.1% threshold). Furthermore, there was a significant interaction effect between radial portion and compartment on WSG, exceeding a threshold of 1%. However, no such significant effects were observed for radial portion×sites interaction at the 5% significance level. This study contributes valuable insights into the wood properties of this endemic species, enhancing the understanding of its ecological and physical significance.