An NIRS-based assay of chemical composition and biomass digestibility for rapid selection of Jerusalem artichoke clones

Abstract

Background: High-throughput evaluation of lignocellulosic biomass feedstock quality is the key to the successful commercialization of bioethanol production. Currently, wet chemical methods for the determination of chemical composition and biomass digestibility are expensive and time-consuming, thus hindering comprehensive feedstock quality assessments based on these biomass specifications. To find the ideal bioethanol feedstock, we perform a near-infrared spectroscopic (NIRS) assay to rapidly and comprehensively analyze the chemical composition and biomass digestibility of 59 Jerusalem artichoke (Helianthus tuberosus L., abbreviated JA) clones collected from 24 provinces in six regions of China. Results: The distinct geographical distribution of JA accessions generated varied chemical composition as well as related biomass digestibility (after soluble sugars extraction and mild alkali pretreatment). Notably, the soluble sugars, cellulose, hemicellulose, lignin, ash, and released hexoses, pentoses, and total carbohydrates were rapidly and perfectly predicted by partial least squares regression coupled with model population analyses (MPA), which exhibited significantly higher predictive performance than controls. Subsequently, grey relational grade analysis was employed to correlate chemical composition and biomass digestibility with feedstock quality score (FQS), resulting in the assignment of tested JA clones to five feedstock quality grades (FQGs). Ultimately, the FQGs of JA clones were successfully classified using partial least squares-discriminant analysis model coupled with MPA, attaining a significantly higher correct rate of 97.8% in the calibration subset and 91.1% in the validation subset. Conclusions: Based on the diversity of JA clones, the present study has not only rapidly and precisely examined the biomass composition and digestibility with MPA-optimized NIRS models but has also selected the ideal JA clones according to FQS. This method provides a new insight into the selection of ideal bioethanol feedstock for high-efficiency bioethanol production.