Abstract
Analytical methods to assess quality and origin of food are well established, but frequently time consuming, costly, and destructive. Modern spectroscopic techniques such as near infrared spectroscopy (NIRS) have gained interest as fast, nondestructive methods for quality control and traceability in the fruit supply chain. In this work, NIRS combined with chemometrics was successfully used to classify 'Golden Delicious' apples from three different orchard elevation levels (225, 650, and 1000 m above sea level) as well as nine cultivars ('Braeburn', 'Coop39 - Crimson Crisp (R)', 'Fuji', 'Fujion', 'Gala', 'CIV323 - Isaaq (R)', 'Coop43 - Juliet (R)', 'SQ159 - Natyra (R)', 'UEB32642 - Opal (R)'). Principal component analysis (PCA) and quadratic discriminant analysis (QDA) based on PCA scores were used to classify the apples (n = 842) according to their orchard elevation and cultivar. Full cross validation (leave-one-out) was used as validation method in the development of the prediction models. PCA-DA models correctly classified 93.6% and 77.9% of the high- and low-elevation grown 'Golden Delicious', respectively. For the intermediate orchard level, a correct classification rate of 57.1% was achieved. Five ('Braeburn', 'Coop39 - Crimson Crisp (R)', 'Gala', 'CIV323 - Isaaq (R)', 'SQ159 - Natyra (R)') of nine apple cultivars were classified correctly at 100%, whereas 96.2% of 'Fuji', 92.3% of 'UEB32642 - Opal (R)', and 76.9% of 'Fujion' and 'Coop43 - Juliet (R)' were correctly recognized. When the models were validated using independent samples, a correct classification rate of 87.5% for orchard elevation and 86.3% for cultivar was found, respectively. Our results highlight the potential of NIRS combined with PCA-QDA as a non-destructive and fast analytical method to trace the origin of apples in terms of orchard elevation and to classify apple cultivars.