基于矿物元素指纹技术的西北地区枸杞产地判别分析  

Discriminant analysis of wolfberry origin in northwest China based on mineral element fingerprint technology

作　　者：王彩艳 开建荣 李彩虹 王晓菁 杨春霞 王晓静 WANG Caiyan;KAI Jianrong;LI Caihong;WANG Xiaojing;YANG Chunxia;WANG Xiaojing(Ningxia Research Institute of Quality Standards and Testing Technology of Agricultural Products,Yinchuan 750002,Ningxia,China)

机构地区：[1]宁夏农产品质量标准与检测技术研究所,宁夏银川750002

出　　处：《经济林研究》2023年第3期187-196,共10页Non-wood Forest Research

基　　金：宁夏回族自治区自然科学基金项目(2021AAC03283);“十四五”农业高质量发展和生态保护科技创新示范项目(NGSB-2021-5);2022年第二批宁夏回族自治区人才专项青年拔尖人才项目。

摘　　要：[目的]探讨矿物元素指纹技术在枸杞原产地判别中的可行性,为枸杞地理标志产品保护提供参考。[方法]采用电感耦合等离子体质谱仪(ICP-MS)测定宁夏、青海、甘肃和新疆产枸杞果实中56种矿物元素含量,对矿物元素含量进行方差分析、主成分分析及判别分析,基于产地间含量有显著差异的矿物元素构建枸杞产地判别模型。[结果]各产地枸杞果实中矿物元素含量有其各自的指纹特征,56种矿物元素中有47种元素的含量在不同产区间存在显著差异(P <0.05),青海产枸杞果实中大部分元素含量高于其他地区。通过主成分分析提取出13个主成分,累计方差贡献率为79.656%,并筛选出Nd、Ce、Hf、Bi、Ho、Mn、Mg、Au、Ir、Ba、Sn、Ru、Sr、Cr、Sc、Rb、Cs、Zn、Cu、Cd、Li、Ta共22种枸杞的特征指标。以地域间含量具有极显著差异的44种矿物元素指标进行Fisher判别分析,确定了Al、As、B、Co、Cu、Dy、Mo、Nb、Ni、Sb、Se、Tl、U、Y、Fe、Zn、P、Ca、Ir、Hf共20种枸杞的有效溯源指标,建立了基于Fisher线性判别分析和OPLS-DA判别分析的宁夏、青海、甘肃和新疆4个产地枸杞的产地判别模型,整体正确判别率分别为97.4%和95.8%,使用模型对不同产地枸杞的区分判别较为准确,因此,建立的判别模型可被用于枸杞的原产地判别。[结论]可以基于矿物元素指纹的差异有效鉴别不同产地的枸杞。【Objective】This study aims to explore the feasibility of mineral element fingerprint technology in the identification of origin of wolfberry,and provide technical support for the protection of geographical indications of wolfberry products.【Method】The contents of 56 mineral elements in wolfberry from Ningxia,Qinghai,Gansu and Xinjiang were determined by inductively coupled plasma mass spectrometry(ICP-MS),and the variance analysis,principal component analysis and discriminant analysis were carried out for the mineral element content.The discriminant model of wolfberry was established based on the mineral elements with significant difference between producing areas.【Result】By analyzing the differences of mineral elements in wolfberry from Ningxia,Gansu,Qinghai and Xinjiang,the results showed that each producing area of wolfberry had its own fingerprint characteristics,there were significant differences in the contents of 47 elements out of 56 mineral elements in different production areas,most elements in Qinghai wolfberry were higher than those in other areas.13 principal components were extracted by principal component analysis,and the cumulative variance contribution rate was 79.656%,the characteristic indexes of 22 kinds of wolfberry(Nd,Ce,Hf,Bi,Ho,Mn,Mg,Au,Ir,Ba,Sn,Ru,Sr,Cr,Sc,Rb,Cs,Zn,Cu,Cd,Li,Ta)were screened out.The characteristic indexes of wolfberry were screened out.Geographical indicators of 44 sorts of elements which had a very significant differences were analyzed by Fisher discriminant method to determine 22 effective traceability index including Al,As,B,Co,Cu,Dy,Mo,Nb,Ni,Sb,Se,Tl,U,Y,Fe,Zn,P,Ca,Ir and Hf.Based on Fisher linear discriminant analysis and OPLS-DA discriminant analysis,the discriminant models of four wolfberry producing areas in Ningxia,Qinghai,Gansu and Xinjiang were established.The overall correct discriminant rates were 97.4%and 95.8%,respectively,the model could be used to distinguish the origin of wolfberry from different habitats,so the mineral elements can be used to

关 键 词：枸杞 矿物元素 FISHER线性判别分析 偏最小二乘法判别分析 溯源 

分 类 号：S663.9[农业科学—果树学]