簡化綠豆種原形態結構多樣性之研究

本研究之主要目的為針對一具有全球代表性的綠豆種原，就其形態性狀的多樣性結構進行簡化。本研究共計評估了5,363份種原，所評估之形態性狀包括15個「測度性狀」與6個「等級性狀」，共計21個性狀。為簡化形態性狀多樣性結構所進行之多變數統計分析包括主成分分析、主成分因子分析及最大可能率因子分析。經由主成分分析與主成分因子分析結果，由第1至第6成分（因子）的累加變異解釋量僅達62%，顯示這兩種方法均無法達成只以少數獨立綜合指標簡化此綠豆種原形態性狀多樣性結構的預期目標，因此進行後續種原分群的策略之一為：以包含全部21個形態性狀之原始數據結構直接進行分群分析。另經由最大可能率因子分析求得函數解之方式，可將形態性狀多樣性結構化簡至只以前3個因子即可代表92%之累積變異解釋量，因此進行後續種原分群的策略之二為：以最大可能率因子分析所得之前3個因子或以因子載荷量較大的重要性狀進行分群分析。

關鍵字

綠豆；種原；形態性狀；遺傳多樣性；多變數分析

並列摘要

The objective of this study is to simplify the structure of morphological diversity within the world collection of mungbean. In this study, total of 5,363 accessions were analyzed. Twenty-one morphological traits comprising fifteen numerical traits and six coded traits were evaluated. The multivariate analysis for structure simplification includes principal component analysis, principal component factor analysis and maximum likelihood factor analysis. The results of principal component analysis and principal component factor analysis showed the first six components (factors) only accounted for 62% of total variation. It is evident the structure of morphological diversity could not be reduced to few independent composite index with principal component analysis and principal component factor analysis. As for the result of maximum likelihood factor analysis, the first three factors could account for 92% of total variation. Therefore, the first strategy for successively clustering mungbean collection is to conduct the cluster analysis based on total 21 morphological traits. The second strategy is to conduct the cluster analysis based on 3 retained factors from maximum likelihood factor analysis or important traits with high factor loadings.

並列關鍵字

Mungbean ； Germplasm ； Morphological traits ； Genetic diversity ； Multivariate analysis

國際替代計量

簡化綠豆種原形態結構多樣性之研究

全文下載

主題瀏覽