文献类型：期刊文献

中文题名：基于多元素含量的统计学方法鉴别我国不同产地核桃

英文题名：Identification of Walnut from Different Regions of China by Statistical Methods Based on the Determination of Multi-element Contents

作者：任传义[1] 程军勇[2] 陈振超[1] 倪张林[1] 汤富彬[1]

第一作者：任传义

机构：[1]中国林业科学研究院亚热带林业研究所,国家林业局经济林产品质量检验检测中心(杭州),浙江杭州311400;[2]湖北省林业科学研究院,湖北武汉430079

年份：2017

卷号：30

期号：5

起止页码：779-787

中文期刊名：林业科学研究

外文期刊名：Forest Research

收录：CSTPCD;;Scopus;北大核心:【北大核心2014】；CSCD:【CSCD2017_2018】；

基金：中央级公益性科研院所基本科研业务费专项资金项目(CAFYBB2017SZ002)

语种：中文

中文关键词：核桃;多元素;主成分分析;判别分析;产地鉴别

外文关键词：walnut ; multi-element; principal component analysis ; linear discriminant analysis ; geographical origin discriminant

分类号：S664.1

摘要：[目的]对我国不同产地核桃进行鉴别,为核桃的原产地保护提供一定的基础数据和理论依据。[方法]应用电感耦合等离子体质谱(ICP-MS)测定全国8个核桃主产省128份核桃样品中35种元素含量,结合单因素方差分析、主成分分析(PCA)和线性判别分析(LDA),建立判别模型,对核桃产地进行鉴别。[结果]表明:核桃中主要微量营养元素为Fe、Zn、Cu和Ni等,重金属(Pb、Cd和As)及稀土元素含量较低;单因素方差分析结果表明,不同地区核桃样品中元素组成差异显著(P<0.05);主成分分析表明,不同地区核桃样品中的特征元素为稀土元素以及Co、Fe、Rb、Zn、Tl、Cu、Cd、Ba、Sm、Sc、Mo和Ti等元素;应用线性判别分析建立了不同产地核桃判别模型,8个省份核桃整体判别正确率为99.2%,同时,应用线性判别分析建立了核桃地理标志产品与非地理标志产品判别模型,6种核桃地理标志与非地理标志产品整体判别正确率为95.7%。[结论]通过测定核桃中多种元素含量,结合主成分分析(PCA)和线性判别分析(LDA)等方法,可对不同产地核桃进行鉴别。
[ Objective ] To identify the walnut from major producing provinces in China and provide some basic data and theoretical basis for the protection of geographical indication. [ Method ] The contents of 35 elements in 128 walnut samples from eight major producing provinces of China were determined by inductively coupled plasma mass spectrometry （ICP-MS） , the discriminant model was established by one-way analysis of variance, principal compo- nent analysis （PCA） and linear discriminant analysis （LDA） to identify walnut from different areas. [ Result ] It was found that the contents of Fe, Zn, Cu and Ni were the most abundant nutrient elements in walnut, and the con- tents of heavy metal （ Pb, Cd and As） and rare earth elements were in relatively low level. One-way analysis of va- fiance indicated that there were significant differences in the elemental composition of walnut samples from different regions （P 〈0.05）. The PCA showed that Fe, Ti, Rb, B, Ba, Cu, Zn, Ba, Mo, A1, Pb and rare earth elements were inferred to be the characteristic elements of walnut samples from different regions, and these elements could explain 64.33% of the total variance. LDA was applied to construct the classification model of walnuts according to their geographical origins, and the accuracy was as high as 99.2%. LDA was also applied to construct the model of identifying the walnut with geographical indication from that without geographical indication, the accuracy was 95.7%. [ Conclusion] Through the determination of multi-element contents in walnut combined with principal component analysis （PCA） and linear discriminant analysis （LDA） , the walnut from different regions can be identified successfully.