Classification and prediction of early-to-late ripening apricot quality using spectroscopic techniques combined with chemometric tools

Abstract

Abstract A DA-meter, a hand-held instrument developed from Vis/NIR spectroscopy, and a portable chroma meter based on CIELab coordinates were used to classify the maturity stage of six early-to-late apricot varieties at the wholesale market, using the linear discriminant analysis (LDA) and the k Nearest Neighbours (k-NN), and to predict their quality and nutraceutical attributes by a stepwise regression analysis. Results pointed out a significant variation among cultivars analysed, regardless of the ripening season. LDA highlighted the key role of fresh weight and flesh firmness (FF), and to a lesser extent of sugar content (TSS) and dry matter (DM), in discriminating apricot cultivars. Based on LDA results and DA-meter index (IAD), it was possible to distinguish samples into four robust ripening classes (0.00-0.10; 0.11-0.20; 0.21-0.30; >0.30). The colorimetric indices allowed to define three classes (a* 70; a* from 5 to 20; a*>20) and to assess the relative ranges of TSS and DM values. Finally, good predictive multi-cultivar models were carried out for TSS, DM, FF, and total carotenoid content (R2: 0.60–0.65). These techniques are promising tools to assess fruit quality, to identify fruit uniform ripening classes and to predict the quality attributes of early-to-late fresh apricots.