Cultivar and origin authentication of ‘Fuji’ and ‘gala’ apples from two dominant origins of China based on quality attributes

Apple quality is closely related to its cultivar and origin. However, the apple quality characteristics of different cultivars and origins are unclear. The hypothesis that some quality indicators can effectively distinguish the cultivar and origin of apples. The result indicated that the discriminant accuracy of the models was above 90%, and the multilayer perceptron neural network (MLP–NN) model was superior to the linear discriminant analysis (LDA) model. The identification accuracy of cultivars was higher than origins. The main reason was that the single fruit weight, vitamin C, total soluble solid, soluble sugar, sweetness value, sorbitol, glucose, fructose, sucrose, malic acid, quinic acid and citric acid of ‘Fuji’ apples were significantly higher than ‘Gala’ apples. This study provides a foundation for the quality evaluation and further geographical traceability studies of apples. Further studies related to the regulatory mechanism of environmental conditions on apple quality characteristics should be explored for theoretical confirmation. •Quality attributes of ‘Fuji’ and ‘Gala’ apples from two dominant origins in China were compared.•Different discriminate models were constructed for apple cultivar and origin authentication.•The MLP–NN model obtained higher training and validation accuracies than LDA model.