Study On The Identification Of Pork Kinds And Comparison Of Meat Flavor And Quality Between Three Chinese Indigenous And Lean Pig

The Chinese indigenous pork is favored deeply by consumers for rich flavor, delicious and storability. In order to construct Chinese indigenous pork brand feature, extend the quality superiority, and fight against fake and shoddy, the industry urgently need to master the difference of flavor, quality and storability between Chinese indigenous pork and lean pork, and provide a rapid and accurate method for identifying the high quality indigenous pork.One lean pork (DLY pork) and three Chinese indigenous pork--AQ (Anqingliubai), WP (Wannanhua Pig), DP (Dingyuan Pig) were used as the research objects in this paper. We analyzed the differences of pork quality, antioxidant capacity and flavor components (volatile flavor components, free amino acids and inosinic acid) between indigenous pork and DLY pork, Also we constructed a rapid and accurate method to identify the high quality indigenous pork based on the differences of flavor component. The main results of this study were summarized as follows:1. The difference of meat quality and antioxidant capacity was significant between three Chinese indigenous pig and lean pig.Three Chinese indigenous pork was obviously superior than DLY pork (P<0.05) on intramuscular fat (IMF) content, water holding capacity and color. The unsaturated fatty acid content in Chinese indigenous pork was significantly higher than DLY (P<0.05), while the saturated fatty acids was significantly lower than DLY pork. The total antioxidant capacity and antioxidant enzymes activity of superoxide dismutase (SOD), catalase (CAT) and glutathione peroxidase (GSH-Px) in three Chinese indigenous pork were significantly higher than DLY pork no matter before storage or after 72 h storage, The value of thiobarbituric acid (TBA value) was less than DLY pork. The three Chinese indigenous pork was significantly better than DLY pork on storage time and the ability of keeping flavor under the same condition.2. Constructed a method for rapidly detecting the volatile flavor components, free amino acids and inosinic acid(1) The better condition for producing and collecting volatile flavour:The pork was cooked under 120℃ for 25min, followed by 30 minutes of headspace extraction using a 75μm CAR/PDMS SPME fiber at 60℃ water bath. The variation coefficient of volatile compound concentration were all less than 9%.(2) Constructed a rapid detection method to detect IMP and FAA. The variation coefficient of FAA was less than 5%, and the variation coefficient of IMP was less than 3.2%, which indicated the two methods had good repeatability3. The differences of volatile flavor between three indigenous pork and DLY pork was mainly caused by 6 aldehydes,4 ketones,3 alcohols, and 2 furans, under the specific cooking conditions. And the differences of taste were mainly due to FAA and IMP concentration.After analyzing the differences of flavour using principal component analysis method, we found that the different volatile flavor components between three indigenous pork and DLY pork were significant and the difference was mainly caused by 6 aldehydes,4 ketones,3 alcohols and 2 furans. FAA and IMP had significant effects to taste difference between the three indigenous pork and DLY pork, while no significant impact on flavor of AQ and DP pork4. Three discriminant models were established based on the weighted information of flavor, taste, flavor and taste based by stepwise discriminant analysis(1) Four Fisher discriminant functions were established with the concentration of nonanal, Octanal and 1-Octen-3-ol as variables, The four function as follows:DLY:F1=1.894×x1+0.377×x2+0.34×x3-784.611;AQ:F2=2.666×x1+1.474×x2+1.775×x3-5098.685;WP:F3=4.809×xi+0.772×x2+2.710×x3-6715.748;DP:F4=1.793×x1+0.449×x2+3.000×x3-3928.242。In the function:x1 represented 1-Octen-3-ol concentration, x2 represented nonanal concentration, x3 represented Octanal concentration. The original values of three substances concentrations were substituted into the four functions, then four function values were compared, the function type of the highest value is the sample type. With the external data validation, the accuracy of identifying the DLY pork samples was 100%, while the identification capability for indigenous pork was weak. No indigenous pork samples were misjudged as DLY pork samples.(2) According to the concentrations of the FAA and IMP, four Fisher discriminant functions were established with threonine, valine, and isoleucine as variablesDLY:F,=4.717x×1+5.888×x2+79.722×x3-129.509;AQ:F2=39.914×x1+16.288×x2+46.300×x3-971.908;WP:F3=19.407×x1+21.995×x2+31.435×x3-287.751;DP:F4=36.012×x1+29.606×x2+71.750×x3-943.315。In the formula:x1 represented threonine concentration, x2 represented valine concentration, x3 represented isoleucine concentration. With the external data validation, the accuracy of identifying WP pork samples was 100%, the DLY pork samples was 93.3%, and the AQ and DP pork samples was 60% and 53.3%, respectively. No AQ and DP pork was misjudged as DLY pork samples.(3) After the weighted information stepwise discriminant analysis of volatile flavor compounds aroma value (weighting coefficient 0.02), free amino acid and inosine acid concentration (weighting coefficient 2), we got four Fisher discriminant functionwith hexanal, octanal, and glutamic acid as variables.DLY:F1-67.547×x1+16.112×x2+7.639×x3-260.372;AQ:F2=104.445×x1+98.892×x2+42.401×x3-3619.014;WP:F3=212.779×x1+110.299×x2+22.649×x3-4599.775;DP:F4=91.867×x1+139.652×x2+46.162×x3-5746.747.In the formula:x1 represented hexanal aroma value, x2 represented octanal aroma value, x3 represented glutamic acid concentration. With this modle, the accuracy of identifying the DLY pork samples was 100%, the WP pork samples was 93.3%, the DP pork samples was 86.7%, and the AQ pork samples is 80%. No DP and WP pork samples was misjudged as DLY pork in all the external data validation.(4) In the three identifying models of pork type, the model by taste or flavor has low accuracy, while the detection was quick and easy; the model by weighted identification model has the highest accuracy, but needs more comprehensive detections.