Journal of Agriculture and Food Research (Dec 2021)
Development of NIRS re-calibration model for ethiopian barley (Hordeum vulgare) lines traits to determine their brewing potential
Abstract
Barley is one of the top five grain crops in Ethiopia, covering approximately 1 million hectares, and is grown for food, animal's feed and industrial purposes. Ethiopian Agricultural Research Institute (EIAR) is also conducting a barley research and breeding program with the aim of improving grain yield and resilience through screening lines. Sample materials used for the experiment were taken from 2018, 2019 and 2020 seasons to capture the variation in the growing year. As a result, the spectral data of the samples and the wet chemical data were generated using official international methods. Subsequently, a recalibration based on 280 samples of breeding lines of different genotypes, growing years and locations was developed in order to predict the protein, extract and friability content characteristics of important barley malt parameters. Wet chemical data were collected from processed/malted barley grains. A calibration model was created on the basis of the calibration software OPUS LAB 7.5 from Bruker Optics GmbH. The protein calibration (R2c = 0.90; RPD = 3.8) can be viewed as broadly applicable; the extract content (R2c = 0.88 and RPD = 2.8) and friability (R2c = 0.85 and RPD = 2.7) could be useful as well as a good predictive model for line screening. The calibration model was missing to predict long-range sample sources before the model range was expanded by recalibration process including different sample sources. Therefore, these NIRS models enable the selection of suitable food and malt barley genotypes at wide range. Since NIRS is fast and inexpensive, the barley improvement program can increase selection intensity, especially in the early testing phase. So the best precision of the NIRS recalibration discovered in this study enables to identify top candidate lines with a minimum of errors.
