Fine root lignin content is well predictable with near-infrared spectroscopy

Area of Science:

• Ecology
• Biogeochemistry
• Analytical Chemistry

Background:

• Root lignin influences decomposition and the terrestrial carbon cycle.
• Traditional lignin measurement methods are labor-intensive, limiting ecological study sample sizes.
• High-throughput methods are needed to predict root lignin content.

Purpose of the Study:

• To explore the applicability of near-infrared spectroscopy (NIRS) for predicting fine root lignin content.
• To develop and validate NIRS calibration models for lignin prediction.
• To compare NIRS model performance with existing literature benchmarks.

Main Methods:

• Fine root lignin content was measured using the Acetylbromid (AcBr) method in 73 grassland plots.
• Near-infrared spectroscopy (NIRS) calibration and prediction models were developed using partial least square regression (PLSR).
• PLSR was combined with spectral variable selection to improve model performance and identify key wavelengths.

Main Results:

• Initial NIRS-PLSR models showed moderate prediction accuracy for lignin content (RPD=1.96, R²=0.74).
• Combining PLSR with spectral variable selection significantly improved model performance (RPD=2.67, R²=0.86).
• The improved models identified chemically relevant wavelength regions for lignin prediction.

Conclusions:

• NIRS, particularly when combined with spectral variable selection, provides a rapid and accurate method for analyzing root lignin content in herbaceous plants.
• This approach overcomes the limitations of traditional methods, enabling larger sample sizes in ecological research.
• The findings support the use of NIRS for routine lignin analysis in ecological and global climate change studies.