The use of Vis/NIRS and chemometric analysis to predict fruit defects and postharvest behaviour of 'Nules Clementine' mandarin fruit

The use of chemometrics to analyse Vis/NIRS signal collected from intact 'Nules Clementine' mandarin fruit at harvest, to predict the rind physico-chemical profile after eight weeks postharvest was explored. Vis/NIRS signals of 150 fruit were obtained immediately after harvest. Reference data on the rind were obtained after eight-week storage, including colour index (CI), rind dry matter (DM), and concentration of sugars. Partial least squares (PLS) regression was applied to develop models. Principal component analysis (PCA) followed by PLS-discriminant analysis (PLS-DA) were used to classify fruit according to canopy position. Optimal PLS model performances for DM, sucrose, glucose and fructose were obtained using multiple scatter correction pre-processing, showing respective residual predictive deviation (RPD) of 3.39, 1.75, 2.19 and 3.08. Clusters of sample distribution in the PCA and PLS-DA models based on canopy position were obtained. The results demonstrated the potential applications of Vis/NIRS to predict postharvest behaviour of mandarin fruit.