Study On Quality Of Table Grapes In The Process Of Logistics

Grape berry has unique flavor, proper taste and high nutritional value, which is favored by most consumers. The grapes are vulnerable of hardness decay, rotting, shattering, stem drying because of its relatively postharvest activity, thin skin, rich juice, high moisture content and high sugar content which would greatly reduce the value of the goods during transportation, storage and sale. With the development of the quality requirements of table grapes from the market and the improvement of grape logistics and storage technology, the rapid detection and evaluation of grape quality in the process of logistics in postharvest storage, transportation and marketing has become an important research topic in the field of fruit and vegetable industry.In this paper, three kinds of representative table grapes, Muscat grape, Manai grape and Red Globe grape were chosen as test materials. Conventional quality detection combined with high-tech detecting methods such as near infrared spectroscopy technology, electronic nose analysis, HS-SPME/GC-MS, chromatic difference analysis and texture testing analysis were used to measure grape quality during simulate transportation, storage and shelf-life, analyze their correlation, establish universal grape quality prediction models, discuss the applicability of the models, build fast evaluation system of table grapes in complete logistics chain, and provide the basis for practical application. The main results were as follows:1. Using the Muscat grape as the material, with different vibration time simulating different modes of transport, the sensory index, compared and evaluated the sensory index, soluble solids content, total acid, puncture index and TPA index of grapes in different modes. The near infrared spectroscopy model with 1st derivative and standard normalisation was established in the whole wavelength range; the highest accuracy rate of discriminating the maodes of transportation was up to 96.67%. The PCA and LDA analysis of the electronic nose could achieve distinction of grapes in different vibration time. 67 kinds of aroma substances were detected by GC-MS, including geranic acid, citronellol, linalool, a-terpineol, nerol,(E)-2- acrolein which were the main aroma components of grapes. The relative content decreased during the simulation of transport.2.The Muscat grapes were kept at 0°C for 8, 16, 24, 32, 40 days and all three kinds of grapes were kept at 10°C for 3, 6, 9, 12 days, during which the analysis of sensory evaluatio n, soluble solid content, total acid, Vc, surface color, texture quality were detected. During the storage, the soluble solid content and total acid NIR models were better with MPLS, 1stD log(1 / R) and IMSC, of which the cross validation decision coefficient RCV2 were 0.8312, 0.8270, the prediction decision coefficient Rp2 respectively were 0.9205, 0.8312, and the relative verify errors of analysis, RPD were 3.96, 2.02. The optimal method for TPA parameters of hardness, springiness, cohesiveness and recovery model was PLS combined with 1stD log(1 / R) + SNV and detrend. The SECV were 121.18, 0.0374 and 0.0203, the cross validation correlation coefficient RCV were respectively 0.7985 and 0.8769, 0.8497, 0.7850. In the whole spectral range, applying partial least squares(PLS) combined with second order derivative, scattering spectra pretreatment methods, L* value got the best model, the SECV, RCV2 were 0.4591, 0.9476; a*, b* value were suitable for the first derivative combined with the weighted multiple discrete correction processing, the SECV were respectively 0.1239, 0.4496, and the RCV2 were 0.9508, 0.8648. Red Globe Grape vitamin C content prediction model was with the decision coefficient Rp2 of 0.9318, RPD of 3.64. Therefore, visible / near infrared diffuse reflectance spectroscopy rapid detection model of the grape soluble solids, total acid, texture, VC, surface color were stable and high precision, however the texture detection model had to be improved. The electronic nose with PCA and LDA analysis was feasible and effective to distinguish the storage time of the three kinds of different grapes obviously, and PLS binding factor discriminant could effectively identify the grape storage period. The volatile components relative content variation detected and analyzed by GC-MS were consistent with the electronic nose analysis results.3. The grapes were laid up at 18°C ~20°C, 8°C ~10°C, after stored in 0°C for 20 days, simulating sold at normal temperature and sold in cold containers of supermarkets, measuring physicochemical, sensory, nutritional indexes, in 5 days of shelf-life and getting the varying pattern. The electronic nose detection results showed that, using the Principal Component Analysis methods could effectively distinguish samples of different shelf time and Linear Discriminant Analysis reflects the degree of change in odor. The GC-MS analysis results with the peak area normalization method showed that, the content and composition of Muscat grape volatile substances changed during the shelf life, content of main characteristic aroma component, such as(E)-2-Hexenal, Geranic acid, geraniol and nerol decreased and ethanol, hexanol, acetic acid increased and the total peak area declined. Its changing rules were roughly consistent with the physicochemical indexes and electronic nose analysis results. Therefore, the electronic nose combined with GC-MS method was feasible to determine the aroma quality of Muscat grape at shelf life. The qualitative and quantitative detection of volatile substances of 1st day, 3rd day and 5th day in the room temperature and simulated supermarket shelf of Muscat grapes by HS-SPME/GC-MS technology, confirmed identification of the electronic nose and the variation of other indicators.4. The sensory evaluation and physical and chemical indicators, microbial indicators of Muscat grapes in various aspects of the complete logistics chain were analyzed in correlation and factor rotation, to simplify the evaluation variables by extracting the first four principal factors. The cumulative contribution rate reached 84.37%. According to the factor score coefficient calculating the sample’s 4th common factor score Fjn, then calculating the common factor variance contribution rates of the products, the comprehensive factor score Sn could be calculated, which was used to evaluate and classify the grape quality. In order to establish near infrared diffuse reflectance spectra model of the SSC, total acid, TPA hardness, elastic, these four index, MPLS with 1stD log(1 / R) + SNV pre processing methods of modeling turned out to be the best, which could realize the predictions of multiple index with one model. The results of electronic nose PCA and LDA analysis of the whole logistics chain of samples were better than those of the single link, with the first two main components of the total contribution rate of 98.97%, 96.44%. HS-SPME / GC-MS analysis indicated that the relative content of peak area percentage of characteristic aroma compounds, linalool, geraniol,(E)-2-hexenal with maximum relative content of Muscat grapes in the whole logistics chain changed regularly. Correlation analysis on texture parameters of whole logistics chain were carried out. The texture puncture test parameter, peel strength and toughness had significant correlation, and flesh maximum hardness and the average as the same. TPA test index of flesh firmness and elasticity, cohesion, recovery was extremely significant negative correlated, flexibility and cohesiveness, recovery was significantly positive correlated. Using hardness, cohesiveness and elasticity establishing the artificial masticatory model regression equation was Y=0.853X1-63.19X2+21.061X3-60.315. The standardized regression coefficient r was 0.936, and the significant equations corresponding to the significant level Sig<0.05. So the chewiness of Muscat grapes could be expressed with hardness, cohesiveness, elasticity in the equation above, then to predict and evaluate the texture quality of grape effectively.