基于主成分分析对不同产地蜂蜜的鉴别研究

Identification of Honey from Different Producing Areas Based on Principal Component Analysis

摘要: 本文利用蜂蜜中各个成分含量的差异，建立鉴别蜂蜜产地的方法。以我国东北地区3个产地的215批次蜂蜜样品为研究对象，对不同产地蜂蜜成分的指标进行测定及分析，并利用R语言软件，对不同产地蜂蜜成分的指标数据进行单因素方差分析和主成分分析。结果显示，不同产地蜂蜜成分差异显著性水平为P < 0.01，蜂蜜成分在不同产地间具有极显著差异。主成分分析的前2个主成分累积贡献率达到82.1%，能表达蜂蜜成分的主要信息，主成分分散点图中，同一产地样品成簇聚集，不同产地样品分布于相对独立的空间，能有效区分不同产地的蜂蜜样品。葡萄糖、果糖、蔗糖、麦芽糖、淀粉酶等蜂蜜成分对第一主成分(PC1)贡献率较大，表明这些蜂蜜成分含量的差异是区分不同产地蜂蜜样品的主要依据，本实验通过对蜂蜜成分的主成分分析，为蜂蜜的质量控制提供试验依据，提供蜂蜜产地鉴别手段。

Abstract: According to the differences of components in honey, the method of identifying honey producing area was established. We took 215 batches of honey samples from 3 areas in northeast China as the research object, the components of honey from different areas were determined and analyzed. The differences of components in honey in different producing areas were conducted based on one-way ANOVA and principal component analysis using R studio. The results showed that the honey components among different producing areas were significantly different and the significant level was P < 0.01. The cumulative contribution rate of the first two principal components in principal compo-nent analysis was 82.1%, which could represent the main information of honey components. The samples from the same place were clustered together in the plot of principal components, while the samples from different places were separated from each other. This method can effectively distinguish the producing areas of honey samples. The honey components such as glucose, fructose, sucrose, maltose and amylase contributed more to the first principal component (PC1), which indicated that these honey components was the main basis to distinguish the producing areas of honey samples. Principal component analysis of honey provides the experimental basis for the quality control of honey, and the identification method of honey producing area.