A quick and precise online near-infrared spectroscopy assay for high-throughput screening biomass digestibility in large scale sugarcane germplasm

Abstract

The near-infrared spectroscopy (NIRS) has been used for efficient characterization and rapid assay of biomass saccharification in other energy plants, but its application in sugarcane is not reported yet. The current study collected a total of 541 sugarcane accessions to take an online NIRS assay. Among these sugarcane collections, we observed large variations in biomass digestibility, particularly for fermentable hexose and total sugar yield from fresh sugarcane stalks, which were detected ranging from 69.88 to 239.86 kg t-1 and 66.56–228.55 kg t-1 respectively. Using the modified partial least squares method, six reliable NIRS models were obtained with a high coefficient of determination (R2) and the ratio of prediction to deviation (RPD) values during calibration, internal-cross validation and external validation. Notably, the equation for fermentable hexose exhibited the most consistently high R2 (0.98) and RPD (6.62) values, as well as retaining relatively low root mean square error during calibration (3.74 kg t-1) and validation (4.19 kg t-1), indicating excellent predictive capacity. All models demonstrated accurate and stable prediction performance in the two-year large-scale germplasm resources evaluation, and the optima accessions with high or low biomass digestibility can be screened out consistently. Therefore, this study provides a precise and consistent NIRS assay for high throughput scanning of biomass digestibility in sugarcane.