Trait-Based Clustering Analysis Reveals Phenotypic Diversity and Regional Differentiation in Capsella bursa-pastoris (Shepherd’s Purse)
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Morphological diversity and its association with geographic origin are crucial for the conservation and genetic use of Capsella bursa-pastoris (shepherd’s purse). In this study, 35 accessions from Korea and abroad were evaluated using 13 phenotypic traits of shoots and roots. Factor Analysis of Mixed Data (FAMD) and Partitioning Around Medoids (PAM) clustering classified the accessions into six clusters (C1–C6), with C5 and C6 distinctly separated, indicating unique phenotypic compositions shaped by trait combinations or regional origin. Root size-related traits such as length, area, and volume were most pronounced in C1, whereas C6 exhibited the smallest values. C5 showed thicker roots despite shorter lengths, and C4 had the greatest number of root forks and links. Root link length (RLL) was highest in C6, reflecting diverse root architectures. Correlation analysis revealed generally positive relationships among root traits, while RGB and RLL associations varied by cluster, with C5 showing a strong correlation (r = 0.91). Regional composition differed among clusters, and Shannon’s diversity index for leaf-lobed type (LL) was highest in C6 and lowest in C5. These results elucidate phenotypic diversity and population structure in C. bursa-pastoris, supporting efficient germplasm classification and selection for breeding.
Abstract
Morphological diversity and its association with geographic origin are crucial for the conservation and genetic use of Capsella bursa-pastoris (shepherd’s purse). In this study, 35 accessions from Korea and abroad were evaluated using 13 phenotypic traits of shoots and roots. Factor Analysis of Mixed Data (FAMD) and Partitioning Around Medoids (PAM) clustering classified the accessions into six clusters (C1–C6), with C5 and C6 distinctly separated, indicating unique phenotypic compositions shaped by trait combinations or regional origin. Root size-related traits such as length, area, and volume were most pronounced in C1, whereas C6 exhibited the smallest values. C5 showed thicker roots despite shorter lengths, and C4 had the greatest number of root forks and links. Root link length (RLL) was highest in C6, reflecting diverse root architectures. Correlation analysis revealed generally positive relationships among root traits, while RGB and RLL associations varied by cluster, with C5 showing a strong correlation (r = 0.91). Regional composition differed among clusters, and Shannon’s diversity index for leaf-lobed type (LL) was highest in C6 and lowest in C5. These results elucidate phenotypic diversity and population structure in C. bursa-pastoris, supporting efficient germplasm classification and selection for breeding.