Enhancing jatropha genotype selection for wood chemical composition and calorific value using near-infrared spectroscopy

Abstract

Importance of the work: Near-infrared (NIR) spectroscopy has great potential as a
high-throughput characterizing tool, especially for improving wood qualities in jatropha
breeding work.
Objectives: To characterize jatropha wood qualities across various genotypes and to
investigate the relationship between chemical composition and calorific value based on
evaluating the chemical composition and calorific value of jatropha wood using NIR
spectroscopy.
Materials and Methods: Analysis was conducted on 365 wood samples from various
jatropha genotypes. Neutral detergent fiber (NDF), acid detergent fiber (ADF), lignin,
cellulose, hemicellulose and calorific value were all predicted using partial least squares
(PLS) regression models.
Results: The correlation coefficient (r) values of these models ranged from 0.65** to
0.81**, where ** indicates significance at p < 0.01, with low standard error of calibration
and of prediction. In addition, there were significant effects of harvest age and reharvesting, along with notable genotypic variations in all quality parameters that the PLS
models anticipated, with significant correlations between calorific value and NDF (r =
0.65**), ADF (r = 0.42**) and lignin (r = 0.91**).
Main finding: The NIR spectroscopy models developed had modest predictive
performance for calorific value and wood chemical components of jatropha, indicating
that NIR has potential as a rapid and non-destructive tool for preliminary screening and
comparative evaluation of genotypes based on wood quality