An independent validation reveals the potential to predict Hagberg-Perten falling number using spectrometers

Authors: CHEN, CHUN-PENG JAMES - Virginia Tech; HU, YANG - Washington State University; Li, Xianran; MORRIS, CRAIG - Washington State University; Delwiche, Stephen - Steve; CARTER, ARRON - Washington State University; Steber, Camille; ZHANG, ZHIWU - Washington State University

Submitted to: The Plant Phenome Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/8/2023
Publication Date: 5/28/2023
Citation: Chen, C., Hu, Y., Li, X., Morris, C., Delwiche, S.R., Carter, A.H., Steber, C.M., Zhang, Z. 2023. An independent validation reveals the potential to predict Hagberg-Perten falling number using spectrometers. The Plant Phenome Journal. 6(1). Article e20070. https://doi.org/10.1002/ppj2.20070.
DOI: https://doi.org/10.1002/ppj2.20070

Interpretive Summary: The Hagberg-Perten falling number (FN) method is the international standard used to evaluate the damage to grain quality due to the presence of alpha-amylase enzyme in the grain. Low FN due to high alpha-amylase is associated with increased risk of poor end-product quality including cakes that fall and sticky bread and noodles. Low FN can be caused by preharvest sprouting (PHS) and late maturity alpha-amylase (LMA). The FN test requires specialized equipment and personnel, and laboratory facilities, and is time-consuming. The results are received weeks after the grain has been delivered to the elevator. The use of a hyperspectral camera to estimate FN could allow rapid detection on-site allowing elevators to preserve the value of grain by segregating low and high FN grain at delivery. Very little low FN grain can ruin a large quantity of high FN grain. It could also enable selection earlier in the breeding process. Use of independent calibration and validation samples suggested method feasibility.

Technical Abstract: The Hagberg-Perten falling number (FN) method is the international standard used to evaluate the damage to grain quality due to preharvest sprouting (PHS) and late maturity alpha-amylase (LMA). The FN test requires specialized equipment and personnel, and laboratory facilities, and is time-consuming. Developments of fast alternative methods are in critical need. Spectrometers have been studied as a potential tool for fast FN assessment, but none of the studies have achieved direct support for their calibrations through validation using independent samples. In this study, the calibration set had 462 grain samples consisting of 92 varieties grown at 24 locations in 2019, examined using a near-infrared, hyper-spectrometer scanner. In the validation set, 39 samples of 19 varieties, collected from 10 locations in two years (2018 and 2019) that experienced PHS (7), LMA (13), or no low-FN condition (sound, 19), were scanned with a hyper-spectrometer camera. The association between spectra and FN is modeled by partial least square regression. The overall validation correlation accuracy was r = 0.72 between the observed and predicted FN. Among the validation samples, the prediction accuracies are 0.72 and 0.81 from sound and LMA samples, respectively. Both accuracies were significantly higher than PHS samples (r = 0.39). Furthermore, kernel pixel-wise predictions indicated that when FN was somewhat lower than 300 s, altered pixels were concentrated in areas near the embryo, whereas pixels throughout the kernel were affected when FN was quite low. These results showed the potential of using spectrometers to predict FN, which could provide a faster assessment method. Using this faster method, breeders would have an efficient tool to use in the development of varieties with resistance to PHS and LMA. This study shows the potential that growers could separate the damaged grain from sound grain during harvesting and transportation to preserve the value of the sound grain.