Newhall Navel Orange Main Producing Areas Of Fruit Quality Characteristics And Origin Recognition

Excellent quality, good shape and bright color make Newhall orange fruits mainly serve as fresh fruits in market, and get vastly cultivated in most main citrus production areas. However, due to environmental alterations, fruit qualities significantly vary in geographical locations, which lead to obvious price difference in market. Yet there were rare researches about the influence of factors on fruit quality about one citrus variety in so many production areas and the effective recognition technology of production areas. In this study, Newhall navel orange fruits were collected from17main production areas to analyze the influences of geographical location on the fruit qualities, to research the collection method of NIR spectroscopies, and to set up the production area recognition model, with an aim to trace their production areas.Experiment results showed that:1, Fruits from different geographical locations were different in weight, shape, the content of total soluble solids (TSS), titratable acidity (TA), Vitamin C (VC), sugar and organic acid contents, mineral elements and amino acid compositions in both pulp and peel. The mentioned meteorological factors had no significant influence on fruit shape and mean weight. While TSS showed significantly positive correlation with mean annual temperature, annual rainfall, and sunshine hours. TA and VC were closely associated with mean annual temperature and sunshine hours, respectively.2, The main sugar contents (fructose, glucose, sucrose) of Newhall orange showed that peel contained less sugar contents than pulp. The most intense fluctuation in sucrose was observed in pulp, followed by glucose and fructose. Whereas the fructose and glucose which had the same content were the most intense fluctuation in peel, the sucrose was the least. The four contents of organic acid also exhibited a remarkable variation with geographical location, and the citric acid was the predominant organic acid in both peel and pulp, the other organic acids were a trace of existence in the fruits. Except for citric acid, other organic acids in peel were significantly lower than those in pulp. The mentioned meteorological factors had no significant influence on sugar contents, while precipitation was the main factor to influence the organic acid contents, however, the level was not significant.3, Newhall navel orange fruits contained abundant total amino acids, peel had the higher content of total animo acids than pulp. Except for Arg, Asp, Gln, Cys and Lys, the other animo acids contained higher content in peel than in pulp. The contents of Thr, Ser, Gln, Gly, Cys, Val, Lie, Leu, Tyr, Phe and His in peel and pulp were relatively stable, while those of the remaining exhibited dramatic variation as the alternation in geographical locations. Precipitation was the most important factor to influence the animo acids content, followed by sunshine hours.4, Potassium (K) was the most abundant element presented in citrus peel and pulp, followed by P, Ca and Mg. Compared to pulp, Ca, Mg, Mn in peel were notably higher. Fruits from’Leibo’were rich in K and those from ’Guilin’,’Beibei’ and ’Fengjie’ were rich in Se, Zn and Fe, respectively. Precipitation and sunshine hours were the main influential factors, but the levels were not significant.5, The spectrum data were further processed through principal component analysis (PCA), the juice NIR spectroscopy which containing abundant characters were chosen as the import information, then established and compared artificial neural network (ANN), support vector machine (SVM) and genetic algorithm (GA) optimized SVM. The results showed that (1) The established three-layer ANN classifier was trained, whose accuracy of classifying navel orange’s origin reached up to81.45%when there were11input neurons and13hidden neurons.(2) The studied one-to-one extended SVM classifier with radial basis function being the core function, exhibited better performance than ANN with an accuracy of86.98%when the number of PC was20.(3) GA-SVM classifier took into account the interaction of individual inputs, where the PCA-processed results was optimized by GA. During the experiments, classification accuracy hit89.72%when the population, generation, mating probability, and mutation probability were200,100,0.7and0.01respectively.(4) Subsequent researches found the highest classification accuracy of GA-SVM was80%when taking the spectrum of fruit equator, and69%for the base, inferior to that of orange juice. As verified, NIR spectra of Newhall Navel Orange can be used for classifying its production origin effectively. Considering the accuracy, GA-SVM classifier was regarded with the most excellence among three investigated classifiers. Spectra of orange juice was accordingly selected as the best data to analyze origins traceability.