Scholarlink

<p>Rice (<i>Oryza sativa</i> L.) is fundamental to global food security, yet yield stagnation under changing climatic conditions necessitates improved utilization of phenotypic diversity in breeding programs. This study evaluated 62 rice genotypes, including traditional landraces from West Bengal and Tripura and commonly grown improved varieties, for 13 quantitative traits across two cropping seasons. The panel captured wide variation in grain type, maturity duration, and plant architecture typical of eastern India. Multivariate analyses revealed that reproductive sink-related traits—panicle weight, slpikelet number, test weight, and grain yield per plant—were the primary contributors to phenotypic variation, while tillering and panicle number defined a secondary axis of diversity. K-means clustering, validated by silhouette analysis, grouped genotypes into four distinct classes based on yield potential, phenology, and tillering behaviour. Nonlinear 3D t-SNE visualization further resolved fine-scale patterns of genotypic similarity beyond linear methods. Outlier detection identified grain breadth, panicle weight, and grain length-to-breadth ratio as key drivers of phenotypic divergence, while correlation analysis demonstrated that yield was predominantly influenced by sink-related traits, with early flowering and maturity showing negative associations with productivity. By integrating linear, nonlinear, and statistical association approaches, this study presents a novel and robust framework for deciphering complex phenotypic variation and provides actionable insights for parental selection, ideotype breeding, and the development of stable, high-yielding rice cultivars.</p> Graphical abstract <p></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

Multivariate and nonlinear analyses to characterize phenotypic variability in landraces and cultivars of rice (Oryza sativa L.)

  • 期刊论文

  • 发表日期：

    2026年1月28日

查看全文
  • Soham Hazra,
  • Debarpita Datta Ray,
  • Abhisek Hazra,
  • Avishek Chatterjee,
  • Ankur Mukhopadhyay,
  • Subhadwip Ghorai,
  • Pabitra Kumar Ghosh,
  • Suvojit Bose,
  • Sourav Roy,
  • Tina Nath,
  • Sabuj Ghosh,
  • Deep Pal,
  • Rahul De,
  • Supriya Dey,
  • Maman Khatun,
  • Poulomi Sen

摘要

Rice (Oryza sativa L.) is fundamental to global food security, yet yield stagnation under changing climatic conditions necessitates improved utilization of phenotypic diversity in breeding programs. This study evaluated 62 rice genotypes, including traditional landraces from West Bengal and Tripura and commonly grown improved varieties, for 13 quantitative traits across two cropping seasons. The panel captured wide variation in grain type, maturity duration, and plant architecture typical of eastern India. Multivariate analyses revealed that reproductive sink-related traits—panicle weight, slpikelet number, test weight, and grain yield per plant—were the primary contributors to phenotypic variation, while tillering and panicle number defined a secondary axis of diversity. K-means clustering, validated by silhouette analysis, grouped genotypes into four distinct classes based on yield potential, phenology, and tillering behaviour. Nonlinear 3D t-SNE visualization further resolved fine-scale patterns of genotypic similarity beyond linear methods. Outlier detection identified grain breadth, panicle weight, and grain length-to-breadth ratio as key drivers of phenotypic divergence, while correlation analysis demonstrated that yield was predominantly influenced by sink-related traits, with early flowering and maturity showing negative associations with productivity. By integrating linear, nonlinear, and statistical association approaches, this study presents a novel and robust framework for deciphering complex phenotypic variation and provides actionable insights for parental selection, ideotype breeding, and the development of stable, high-yielding rice cultivars.

Graphical abstract