Evaluation of quality and nutraceutical content of blueberries (Vaccinium corymbosum L.) by near and mid-infrared spectroscopy

Blueberries (family Ericaceae; genus Vaccinium corymbosum L.) are interesting fruit for potential health benefits due to their bioactive compounds (anthocyanins, flavonoids, polyphenols and ascorbic acid). Blueberry quality indices and nutritional properties are generally determined by instrumental techniques (LC-MS; HPLC) that are time and chemical consuming, laborious and expensive. The aim of this work was to examine the potential of MIR and near-infrared (NIR) spectroscopy to evaluate both ripening indices and nutraceutical properties of two blueberry cvs, 'Brigitta' and 'Duke', grown in Valtellina (northern Italy), harvest 2005 and 2006, collected weekly from mature green to full ripe berry stage. Fruit from each ripening class was analysed for total soluble solids (TSS), total phenols, total flavonoids and total anthocyanins, ascorbate and spectroscopic analysis, such as FT-IR and FT-NIR. The spectral datasets were correlated with technological ripening parameters and with nutraceutical compounds by using partial least squares (PLS) regression algorithms both in full cross-validation and test set validation. The PLS analysis for quantitative prediction of TSS gave good calibration models in near (root mean standard error in cross-validation (RMSECV) = 0.50% and root mean standard error in test set validation (RMSEP) = 0.65%) and mid-infrared regions (RMSECV = 0.30% and RMSEP = 0.36%). The calibration models developed in the near infrared region were able to evaluate the content of total phenols (RMSECV = 0.14 mg catechin/g and RMSEP = 0.18 mg catechin/g), total flavonoids (RMSECV = 0.20 mg catechin/g and RMSEP = 0.25 mg catechin/g) and total anthocyanins (RMSECV = 0.25 mg malvidin/g and RMSEP = 0.22 mg catechin/g). The models developed in the mid-infrared region proved to be well correlated with analytically determinated phenolic compounds. The predicted results by the regression models developed for ascorbic acid in near and mid-infrared regions were not as good as those obtained for the phenolic parameters. On the whole, near and mid-infrared spectroscopy gave promising results as they were rapid and non-destructive methods able to evaluate at the same time, technological ripening indices and nutraceutical compounds of blueberries. These techniques could be valid and simple tools to reduce the analytical time and cost of monitoring blueberry quality. (C) 2008 Elsevier B.V. All rights reserved.