Scholarlink

<p>Leaf morphology is a key taxonomic character in angiosperms, yet qualitative descriptions often fail to resolve closely related species exhibiting subtle shape variation. In this study, elliptic Fourier analysis (EFA) was employed to quantitatively assess leaf outline variation among ten species of <i>Ardisia</i> (Primulaceae) collected from wild populations in Assam, India. Leaf outlines were digitized using vector-based tracing, standardized through centering, scaling, alignment and interpolation, and analysed using elliptic Fourier descriptors. Harmonic power analysis indicated that 19 harmonics were sufficient to explain more than 95% of cumulative leaf shape variance, ensuring an optimal balance between descriptive accuracy and statistical efficiency. Principal component analysis of Fourier coefficients revealed clear species-level clustering with minimal overlap among taxa. Permutational multivariate analysis of variance (PERMANOVA) confirmed significant interspecific differences in leaf shape, with species identity explaining a large proportion of shape variation. Linear discriminant analysis (LDA) achieved high classification accuracy, indicating strong discriminatory potential of leaf outline descriptors. The results suggest that quantitative outline-based morphometrics captures biologically meaningful taxonomic signal in <i>Ardisia</i> within the sampled dataset. Overall, this study highlights the utility of elliptic Fourier–based leaf shape analysis as a complementary tool for species delimitation in morphologically challenging plant genera.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Quantitative discrimination of Ardisia species using elliptic Fourier analysis of leaf outlines

  • 期刊论文

  • 发表日期：

    2026年5月13日

查看全文
  • Manash Baruah,
  • Bhaskar Dowarah,
  • Nilakshee Devi,
  • Anurag Kashyap,
  • Bijoy Neog

摘要

Leaf morphology is a key taxonomic character in angiosperms, yet qualitative descriptions often fail to resolve closely related species exhibiting subtle shape variation. In this study, elliptic Fourier analysis (EFA) was employed to quantitatively assess leaf outline variation among ten species of Ardisia (Primulaceae) collected from wild populations in Assam, India. Leaf outlines were digitized using vector-based tracing, standardized through centering, scaling, alignment and interpolation, and analysed using elliptic Fourier descriptors. Harmonic power analysis indicated that 19 harmonics were sufficient to explain more than 95% of cumulative leaf shape variance, ensuring an optimal balance between descriptive accuracy and statistical efficiency. Principal component analysis of Fourier coefficients revealed clear species-level clustering with minimal overlap among taxa. Permutational multivariate analysis of variance (PERMANOVA) confirmed significant interspecific differences in leaf shape, with species identity explaining a large proportion of shape variation. Linear discriminant analysis (LDA) achieved high classification accuracy, indicating strong discriminatory potential of leaf outline descriptors. The results suggest that quantitative outline-based morphometrics captures biologically meaningful taxonomic signal in Ardisia within the sampled dataset. Overall, this study highlights the utility of elliptic Fourier–based leaf shape analysis as a complementary tool for species delimitation in morphologically challenging plant genera.