利用化學計量分析區分農產品之來源—以茶葉與大蒜為例

林振男; 彭宗仁; 王博軒; 林毓雯; 劉滄棽; Jhen-Nan Lin; Tsung-Ren Peng; Po-Hsuan Wang; Yu-Wen Lin; Tsang-Sen Liu

doi:10.6156/JTAR.202306_72(2).0004

利用化學計量分析區分農產品之來源—以茶葉與大蒜為例

Other Title

Apply Chemometrics Approach to Differentiate the Geographical Origin of Agricultural Products: Case Studies of Tea and Garlic

Journal

台灣農業研究

Journal Volume

72

Journal Issue

2

Pages

127-149

Date Issued

2023-06-04

Author(s)

林振男

彭宗仁

王博軒

林毓雯

劉滄棽

Jhen-Nan Lin

Tsung-Ren Peng

Po-Hsuan Wang

Yu-Wen Lin

Tsang-Sen Liu

DOI

10.6156/JTAR.202306_72(2).0004

URI

https://scholars.tari.gov.tw/handle/123456789/19915

Abstract

本試驗利用化學計量法區分茶葉與大蒜的地理來源。此化學計量分析係利用多變量分析方法，將微量元素含量及同位素組成數據進行統計處理，建立各產地之判別式，並據此估算以判別茶葉及大蒜產地來源。茶葉樣品來自台灣、中國、越南、泰國及印尼；大蒜樣品來自台灣、阿根廷、韓國、中國及越南。分析結果顯示，透過化學計量分析區分茶葉地理來源的整體判別率為96%，個別地理區判別率最高可達100%；大蒜試驗的整體判別率為93%，個別地理區判別率最高可達100%。除上述優異之判別成果外，有幾個研究經驗可作為後續相關研究之重要參考依據：(1) 微量元素是區分農作物產區地理來源的最重要因子；如果再加入同位素數據，則可提高識別率。(2) 減少參與化學計量分析之地區數目，可提升盲樣判釋的成功率。(3) 應儘量透過相關檢定，篩選出最有相關的自變數以優化判別式。 This experiment used chemometrics approach to discriminate the geographical origin of tea and garlic. The chemometrics, based on the multivariate analysis, statistically process the trace element concentration and isotopic composition data of agricultural samples studied to establish a discriminant function for each geographical origin. The discriminant function was used to estimate the geographical origin of the tea and garlic studied. The tea samples were taken from Taiwan, China, Vietnam, Thailand, and Indonesia, and the garlic samples were from Taiwan, Argentina, Korea, China, and Vietnam. The results show that the overall discrimination rate of this approach in differentiating the geographical origin of tea was about 96%, and the highest discrimination rate of individual geographical areas was 100%. For comparison, the corresponding discrimination rates of the garlic test were 93% and 100%, respectively. In addition to the excellent discriminant results, some research experiences can be important references for follow-up related research: (1) Trace elements are the most important factor while discriminating the geographical origin of crop production areas; if isotopic data are added, the discrimination rate can be elevated. (2) Minimizing the number of regions involved in the chemometric analysis can elevate the discrimination rate of the blind sample. (3) The most relevant independent variables should be screened out as far as possible to optimize the discriminant function.

Subjects

化學計量分析

產源判別率

茶葉

大蒜

Chemometrics

Discriminant rate of geographical origin

tea

garlic

Publisher

農業試驗所

Type

journal article

File(s)

Name

72-2-4.pdf

Size

2.48 MB

Format

Adobe PDF

Checksum

(MD5):38f9166002ae192ed145d1a97e058a2c