Effect of variety, vintage and winery on the prediction by visible and near infrared spectroscopy of the concentration of glycosylated compounds (G-G) in white grape juice

The use of visible (Vis) and near infrared (NIR) spectroscopy was explored as a rapid, simple and low cost measurement of the concentration of total glycosylated compounds in white grape juice. The effects of variety (Chardonnay, Riesling and Sauvignon Blanc), winery and vintage (2004 to 2006) on the Vis-NIR spectra were also examined. Juice samples from South Australian wineries were scanned in transmittance mode on a FOSS NIRSystems6500 instrument and subjected to laboratory analyses for the measurement of the concentration of total glycosylated compounds (G-G), total soluble solids (TSS), pH and total phenolics (TP). Partial least squares (PLS) regression method was used to relate the G-G reference data to the Vis-NIR spectra. For all samples, PLS regression resulted in a coefficient of determination in calibration ( R ²_cal) and standard error of cross validation (SECV) of 0.82 and 49.15 μM, respectively. Splitting the sample set by variety, winery or vintage improved the PLS calibrations for the variety sets. The results show that Vis-NIR spectroscopy has potential for use as a rapid, semi-quantitative technique to predict G-G concentration in white grape juices as 'low', 'medium' or 'high'. This method will be valuable when taking decisions at the winery during vintage to allocate juices according to their aroma potential. Further studies are in progress to validate the robustness and accuracy of the calibration models.