Studies on Adulteration Detection Methods for Taiwan Honey

• Main Authors: Yi-Chung Chen, 陳宜君
• Other Authors: 區少梅
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/91883558749686909068

碩士 === 國立中興大學 === 食品暨應用生物科技學系 === 94 === In recent years, there are many Thailand-imported honey and adulterated honey existing in the Taiwan honey market. It results in adversely affecting the Taiwan beekeeper’s and consumer’s benefits. Therefore, the purpose of this study was to investigate the methods for determining the adulteration of Taiwan honey as a reference for governmental inspection and to search the methods for the consumer to easily discriminating the adulteration of Taiwan honey as well. Totally 118 honey samples which were obtained from Taiwan major production areas were used as experimental materials. After physicochemical analysis, there were 17 samples classified as adulterated honey samples due to their contents of constituents either not pass the Chinese National Standards (CNS) or unusually above the normal readings. These 17 samples as the adulterated samples were used for the discrimination study. In addition, there were another 10 honey samples collected from the market as the unknown samples for checking the feasibility of established discriminating model. The results of linear discrimination analysis (LDA) for 12 physicochemical items of 118 honey samples showed that the adulterated samples could be successfully discriminated. However, for testing the discrimination model, there were 2 out of 10 unknown samples misclassified. From the results of LDA with stepwise method, there were 7 constituents, such as moisture, diastase activity, HMF, fructose, maltose and ash selected to be able to 100% discriminate the adulterated samples. And there was only one unknown sample misclassified by using these 7 constituents. Using Near Infrared Spectroscope (NIRS) spectrum absorption value of 20 wavelengths which were selected by Fisher Weight to discriminate 118 samples, the percentage of successful discrimination was 99.15%. But still one of 10 unknown samples could not be discriminated successfully with this method. However, after adding back this misclassified unknown sample to the discrimination model, the percentage of successful discrimination could be up to 100%. For searching methods for consumers to easily distinguish the adulterated Taiwan honey, two simple methods for consumer use were found feasible in this study. One is to dilute the honey with water 3 times and examine its transparency. The adulterated honey showed a higher clarity than the pure honey. The other one is to dilute the honey with water 40 times and then heavily shacked. The pure honey showed a relatively large amount of bubbles on the top of layer while the adulterated honey showed as a clear liquid without any bubbles.