Authentication of green asparagus varieties by near-infrared reflectance spectroscopy

Near-infrared reflectance spectroscopy (NIRS) was used for the authenticati on of 2 green asparagus varieties (Taxara and UC-157), grown in Huetor-Taja r (Granada, Spain) protected by the Quality Specific Appellation "Esparrago s de Huetor-Tajar". To develop the prediction model, the method chosen was modified partial least square (MPLS) regression. Two sample sets (N-1 = 219 and N-2 = 145 samples, respectively) were used to obtain the calibration e quations. The standard error of cross-validation (SECV) and the r(2) value were 0.082 and 0.97, respectively, for the Ist calibration set and 0.077 an d 0.97 for the 2nd calibration set. The 2nd chemometric model obtained was tested with independent validation sample set (N-3 = 74 samples), and the r esulting values for standard error of prediction (SEP) and for r(2) were 0. 07 and 0.96, respectively. These results prove that NIRS is an accurate tec hnology for identification and authentication of asparagus varieties and ea sily implemented in industry.