Total Glucosinolate Preservation and Near Infrared Prediction in Rapeseed Meal
View/ Open
Abstract
Glucosinolates (GLS) and their hydrolysis products are of great food and feed safety concern because they are responsible for both the beneficial and harmful properties of GLS-containing plants. Understanding GLS storage stability and total GLS concentration in Brassica oil meals is important to ensuring livestock health. The storage stability of GLS and potential of near infrared spectroscopy (NIRS) for screening the total GLS content of various Brassica meals obtained globally and over multiple growing seasons was evaluated. Decreases were observed in meal stored at 4 oC. GLS storage stability within stored Brassica meals was possible for 18 months and possibly longer providing the seed meals are protected from exposure to moisture conditions that promote endogenous myrosinase hydrolysis. NIRS spectra data from 400 to 2500 nm were recorded on various Brassica meal samples (186) at 2 nm intervals. A global calibration using the Brassica database was developed for both ground and unground meal samples with a modified partial least squares regression analysis of conventional laboratory analysis. The optimum NIR calibrations utilized the first derivative and standard normal variate data preprocessing. The ground NIRS calibration for total GLS resulted in a coefficient of determination (R2) and standard error of the calibration (SEC) and relative predictive determinant (RPD) of 0.96, 6.05, and 6.32, respectively, while the unground NIR calibration had a R2, SEC, and RPD of 0.93, 7.65, and 5.88, respectively. Finally, a sample set (20) with known GLS concentration (by HPLC) was split and one subset was analyzed via NIR “as is” and the other subset was analyzed by NIR after drying for 16 hours at 60 oC in a vacuum oven. The dried Brassica meal sample set had a slightly better residual (HPLC - NIR) standard deviation (4.57) and average residual (-0.74), compared to the “as is” moisture sample set standard deviation (5.00) and average residual (-1.26). The use of NIRS as a routine analytical method for total GLS in Brassica meals destined for animal feeds has great potential. In addition, the low cost of the NIR analysis may be attractive for manufacturers of Brassica meals.