Oil and protein analysis of whole rapeseed kernels by near infrared reflectance spectroscopy

Publisher Website
Google Scholar
Abstract
Oil, protein, chlorophyll and glucosinolate content were analyzed in whole rapeseed kernels by use of a near infrared reflectance technique. Oil and protein content could be estimated with high correlation and good accuracy when predicted results for 89 samples were compared to standard laboratory results. For oil, a multiple correlation coefficient (R) of 0.954 and a standard error of estimate (Sy) value of 0.83 were obtained when reflectance was measured at eight wavelengths. For protein, R = 0.964 and Sy = 0.88 were obtained when reflectance was measured at six wavelengths. Significantly lower correlations were obtained for prediction of chlorophyll (R = 0.506) and glucosinolate (R = 0.707) content, and presently near infrared data cannot be used to measure these two constituents. For the prediction of oil and protein, the levels of accuracy obtained are sufficient for many analytical purposes, and if needed, the accuracy can be improved by repeated measurements. The method is rapid, involves no sample preparation, and leaves intact, viable seed available for other purposes.