A fast and efficient phenotyping method to estimate sugarcane stalk bending properties using near-infrared spectroscopy

Abstract

Lodging is a critical factor that impedes sugarcane growth and restricts the production of sucrose and refined ethanol. Considering bending properties of the stalk are highly associated with lodging resistance, there is currently no method for assessing them in a rapid and accurate manner, which limits the development of lodging-resistant varieties. In this study, a high-throughput phenotyping assay for sugarcane stalk bending strength (SBS) and flexural rigidity (EI) characterization was developed by combining partial least squares (PLS) and artificial neural networks (ANN) approaches via near-infrared spectrum calibration. Both strategies demonstrated high coefficient of determination (R²) and ratio performance deviation (RPD) values during calibration, external validation, and internal cross-validation. Compared with the PSL, the ANN algorithm exhibited better performance for both two types of bending properties calibration, with the R²cv and RPD values as high as 0.94 and 4.18, respectively. Most importantly, these models exhibited consistently excellent predictive performance in successive multi-year phenotypic analyses, where the extreme genotypes could be successfully screened out from a large-scale sugarcane population. In conclusion, this study refines a high-throughput phenotyping approach for the determination of mechanical strength in sugarcane stalks that could be applied to the breeding of lodging resistant sugarcane varieties.

Code and data availability

公開状態または取得可能な本文経路を確認できませんでした。

No evidence-backed public reproduction asset is currently recorded.