Near‐infrared reflectance spectroscopy calibrations for assessment of oil, phenols, glucosinolates and fatty acid content in the intact seeds of oilseed Brassica species

• Published in: Journal of the science of food and agriculture 2018-08, Vol.98 (11), p.4050-4057
• Main Authors: Sen, Rahul, Sharma, Sanjula, Kaur, Gurpreet, Banga, Surinder S
• Format: Article
• Language: English
• Subjects: Brassica; Brassica juncea; Brassica napus; Calibration; calibration model; chemical methods; Correlation coefficients; external validation; Fatty acids; Genotypes; Germplasm; Glucosinolates; Infrared spectroscopy; Linoleic acid; Mathematical models; Mustard; non‐destructive analysis; Oil; Oils & fats; Oilseeds; Oleic acid; Organic chemistry; Phenols; Plant breeding; Quality control; Rapeseed; Reflectance; Seeds; Spectroscopy; Spectrum analysis
• ISSN: 0022-5142; 1097-0010
• DOI: 10.1002/jsfa.8919

BACKGROUND Very few near‐infrared reflectance spectroscopy (NIRS) calibration models are available for non‐destructive estimation of seed quality traits in Brassica juncea. Those that are available also fail to adequately discern variation for oleic acid (C18:1), linolenic (C18:3) fatty acids, meal glucosinolates and phenols. We report the development of a new NIRS calibration equation that is expected to fill the gaps in the existing NIRS equations. RESULTS Calibrations were based on the reference values of important quality traits estimated from a purposely selected germplasm set comprising 240 genotypes of B. juncea and 193 of B. napus. We were able to develop optimal NIRS‐based calibration models for oil, phenols, glucosinolates, oleic acid, linoleic acid and erucic acid for B. juncea and B. napus. Correlation coefficients (RSQ) of the external validations appeared greater than 0.7 for the majority of traits, such as oil (0.766, 0.865), phenols (0.821, 0.915), glucosinolates (0.951, 0.986), oleic acid (0.814. 0.810), linoleic acid (0.974, 0.781) and erucic acid (0.963, 0.943) for B. juncea and B. napus, respectively. CONCLUSION The results demonstrate the robust predictive power of the developed calibration models for rapid estimation of many quality traits in intact rapeseed‐mustard seeds which will assist plant breeders in effective screening and selection of lines in quality improvement breeding programmes. © 2018 Society of Chemical Industry