No abstract

No abstract

No abstract

Background and Aims The parameters commonly examined upon receival of grapes at wineries are insufficient for comprehensive quality assessment. Grape quality is mainly determined by the ripeness and phytosanitary status of the grapes and is indicated by changes in chemical composition. Thus, we have evaluated the feasibility of an on‐line assessment for real‐time analysis of crushed grapes by near infrared/visible (NIR/VIS) spectroscopy at production scale upon their receival. Methods and Results Spectrometers were successfully integrated into the receival lines of two wineries during three vintages. The spectra of 1160 samples were collected. Calibration models were established for the spectral ranges of 450–850 nm (VIS) and 1050–1650 nm (NIR) by partial least squares regressions for ten relevant parameters. A global calibration model and more specific models relying on grape colour, vintage, winery and grape cultivar were established. The best performing models specific to individual cultivars resulted in a coefficient of determination of the validation (R2val) of 0.767, 0.732 and 0.747 for relative density, fructose and glucose concentration; 0.787, 0.540, 0.877 and 0.862 for pH value, tartaric acid and malic acid concentration and titratable acidity; and 0.698, 0.777 and 0.574 for the grape rot indicators glycerol, gluconic acid and acetic acid, respectively. Conclusions Models suitable for semi‐quantitative prediction of parameters related to composition were established as being valuable for multifactorial, holistic assessment of the composition of grape deliveries. Specialisation of the calibration models bears the potential for further improvement, thus allowing comprehensive assessment of grape composition by NIR/VIS spectroscopy. Significance of the Study On‐line implementation of NIR/VIS spectroscopy was shown to be a promising tool for rapid and objective assessment of the composition of grapes upon receival at wineries, which may enable improved systems for payment and quality management. Further acceptance criteria may be developed and food safety may be improved by detecting and rejecting spoiled grapes potentially containing mycotoxins.

No abstract