应用光谱分析方法测定牛肝菌的产地和不同部位矿物质含量

Traceability of Boletus Edulis Origin by Multispectral Analysis Combined With Mineral Elements From Different Parts

我国是全球最大的牛肝菌出口国,云南则是国内牛肝菌最大产区。美味牛肝菌香郁爽滑、营养丰富,备受消费者青睐,但因不同的地理气候和环境差异,导致品质参差不齐。开展云南省不同产区美味牛肝菌产地鉴别,有利于提高商品质量控制。本研究采集云南省13个产地美味牛肝菌样品124份,使用傅里叶变换中红外光谱(FTIR-MIR)、傅里叶变换近红外光谱(FTIR-NIR)、紫外可见光谱(UV-Vis)和电感耦合等离子体原子发射光谱法测定光谱信息与不同部位矿质元素含量并进行分析,对原始光谱进行平滑(Savitzky-Golay, SG)、二阶导数(SD)、标准正态变换(SNV)等预处理,数据使用Kennard-Stone分类法分为训练集与预测集,通过偏最小二乘辨别分析(PLS-DA)和支持向量机(SVM)建立分类模型后进行对比分析,寻找最佳的产地鉴别方法。结果显示:(1)元素测定方法中茶叶标准物质回收率在91.00%~106.00%之间,方法准确可靠;(2)美味牛肝菌富含元素K, P, Mg, Na和Ca且不同部位相同产地,不同产地相同部位元素之间存在差异性,可能与美味牛肝菌不同部位的富集能力和不同产地地理环境差异相关;(3)中级融合通过主成分分析(PCA)提取重要信息,其中FTIR-MIR和UV-Vis光谱数据的累计贡献率达到83.50%和66.70%,代表重要信息变量;(4)在PLS-DA与SVM模型中,数据融合后的产地鉴别效果基本高于单一数据鉴别,说明数据融合策略在美味牛肝产地鉴别中效果显著;(5)采用Hottelling T2检测法对数据融合进行异常值检验,结果表明模型未超过置信区间,具有准确性与可信性;(6)PLS-DA模型的初级融合和中级融合结果都高于SVM,说明PLS-DA模型中级融合可以作为产地鉴别的最佳方法。多种光谱结合不同部位矿质元素可准确鉴别不同产地美味牛肝菌,为云南美味牛肝菌地域品质差异评价提供有效的分析方法。

China is the world’s largest exporter of Boletus,and Yunnan Province is the largest producer of Boletus eduils in China.The delicious Boletus eduils is fragrant and nutritious,and it is popular among consumers.However,due to different geographical climates and environmental differences,the quality is uneven.The production of Boletus edulis in different production areas in Yunnan Province was identified and the quality control was improved.In this study,124 samples of delicious Boletus edulis from 13 producing areas around Yunnan were collected,using Fourier to transformed mid-infrared spectroscopy(FTIR-MIR),Fourier transform near-infrared spectroscopy(FTIR-NIR),and UV-visible spectroscopy(UV-Vis).Inductively coupled plasma atomic emission spectrometry was used to determine the spectral information and mineral content of different parts and analyze them.The original spectrum is smoothed(Savitzky-Golay SG),second derivative SD,standard normal variate(SNV)and other pre-processing.The data is divided into a training set and prediction set by Kennard-Stone classification.The classification model is established by partial least square discriminant analysis(PLS-DA)and support vector machine(SVM),and then the comparative analysis is carried out to find the best method of origin identification.The results show that:(1)The recovery rate of standard tea materials in the element determination method is between 91.00%and 106.00%,and the method is accurate and reliable.(2)Boletus edulis is rich in elements K,P,Mg,Na,Ca and the same place of origin in different parts.There are differences between the same parts in different places,which may be different from the enrichment ability of different parts of Boletus eduils.The geographical environment of the place of origin is related.(3)Intermediate fusion extracts important information through Principal component analysis(PCA).The cumulative contribution rate of FTIR-MIR and UV-Vis spectral data reaches 83.50%and 66.70%,which can represent important information variables.(4)In the PLS-DA and SVM models,the identification effect of the data after fusion is higher than that of the single data identification,indicating that the data fusion strategy is effective in the identification of delicious bovine liver.(5)Using the Hoteling T2 to perform the outlier tests on data fusion.The results show that the model establishment does not exceed the confidence interval,and the model has accuracy and credibility.(6)The primary fusion and intermediate fusion results of the PLS-DA model are higher than the SVM,indicating that the PLS-DA models intermediate fusion can be used as the best method for identification.Multi-spectral combined with mineral elements in different parts can accurately identify the delicious Boletus eduils from different habitats,and provide an effective analysis method for the regional quality difference evaluation of Yunnan Boletus eduils.