基于金属元素组成的新疆石榴产地识别分析

采用电感耦合等离子体原子发射光谱法，对新疆6 个主要产地（库车、吐鲁番、叶城、疏附、喀什、和田）的36 个石榴样品的可食部分（果肉）和籽中12 种金属元素的含量进行测定，采用主成分分析（principalcomponent analysis，PCA）和线性判别分析（linear discrimination analysis，LDA）对石榴可食部分和籽中金属元素进行综合评价。结果表明：PCA得出2 个三因子模型，分别解释了石榴可食部分和籽中金属元素数据的84.29%和60.33%；通过对石榴可食部分中金属元素组成进行PCA，PCA更好地将36 个石榴样品划分为6 类，与实际产地吻合。LDA得出新疆不同产地石榴可食部分和籽的总体验证判别率分别为100%和100%，交互验证判别率分别为100%和94.44%。说明提出的方法具有很好的产地识别作用，可作为石榴产地的一种鉴别方法。

Identification of Pomegranate from Different Producing Areas of Xinjiang Based on Mineral Profile

The concentrations of 12 mineral elements in the edible part and seeds of 36 pomegranate samples collected
from 6 different producing areas of Xinjiang were determined by inductively coupled plasma-atomic emission spectrometry
(ICP-AES) and analyzed by principal component analysis (PCA) and linear discriminate analysis (LDA). Two three-factor
models accounting for 84.29% and 60.33% of data variability for the metal element composition of the edible part and seeds,
respectively were established by PCA. Thirty-six pomegranate samples were classified into six groups by PCA, agreeing
with the producing areas. A satisfactory classification of the edible part and seeds of pomegranate was obtained with
overall correct classification rates of both 100%, and cross-validation rates of 100% and 94.44%, respectively. Thus, this
discrimination method can be applied in the geographic origin discrimination of pomegranate.