Measurement Of Glucosinolates In Broccoli And Its Rapid Measurement Model By Near Infrared Spectroscopy

In the process of high quality breeding of broccoli,it is necessary to select varieties of broccoli with high quality and excellent quality.An important indicator to measure the quality of broccoli is glucosinolates.In order to establish a rapid determination of 4-methyl sulfonyl butyl-glucosinolate(RAA)and 3-indolyl methyl-glucosinolate(GBC)in the broccoli of the near infrared spectroscopy models,first high performance liquid chromatography(HPLC)was used to determine the contents of glucoside in 90 different varieties of broccoli in this study,the samples were then scanned with an XDS near-infrared spectroscopy instrument,then the glucosinolate content obtained by high performance liquid chromatography was input into the chemometrics software supporting the instrument for spectral analysis,and finally the obtained spectral files and chemical results were analyzed on the basis of PLS(Partial Least Squares),different scattering processing methods(SNV;Detrend;SNV+Detrend)and derivative processing methods(FD;SD)were used to preprocess the spectral data to establish the corresponding scaling equation,then the most suitable calibration equation was selected to establish the model,and finally the near infrared spectroscopy model was verified.In this study,a near infrared spectrum calibration model for rapid detection of RAA and GBC was established,which laid a foundation for rapid detection and utilization of nutrient quality components and excellent germplasm resources of broccoli,and could be applied to the breeding practice of broccoli.The main results showed:1.The mean mmol concentration of RAA was the highest and therange of variation was the largest in the distribution of the mmol concentration of glucosinolates in 90 varieties of broccoli,the mean molality was 5.18 μmol/g,and the range of molality was 0.65～17.00μmol/g,the second high concentration was GBC with the mean molar concentration was 4.43 μmol/g and the variation range was 0.17～9.43μmol/g.2.For the indicators of RAA and GBC in glucosinolates of broccoli,the correlation coefficients of internal cross validation were 0.846 and0.892 after SNV second derivative treatment,the external correlation coefficient of the corrected set of the prediction model was 0.867 and0.912.3.In this study,samples other than RAA and GBC models were verified: the external correlation coefficient of the GBC model was 0.960,and the internal cross-validation correlation coefficient was 0.942,the external correlation coefficient of RAA model was 0.918,and the internal cross-validation correlation coefficient was 0.902.4.By scanning and analysing for 90 different varieties of broccoli with the XDS near infrared spectrum instrument in this study,and through the establishment of calibration model,the rapid and nondestructive prediction of glucosinolates in broccoli could be realized with data accurately and fast,preprocessing the advantages of simple and feasible,for the following quality indicaors of broccoli detection to extract provides a theoretical basis.