Scholarlink

<p>Climate change threatens global coffee production, necessitating the use of underutilised species such as <i>Coffea liberica</i> W.Bull as alternatives. Despite its potential, <i>C. liberica</i> germplasm in Peninsular Malaysia remains largely uncharacterised, representing a critical knowledge gap as Southeast Asian coffee markets expand. This study presents the first comprehensive multivariate analysis of 33 Liberica coffee accessions from the MARDI germplasm collection at Kluang, Johor. Using a randomised complete block design with three replications, 13 morphological traits encompassing floral, fruit, and seed characteristics were evaluated. Ten samples per plot were collected and averaged to generate independent plot-level means for statistical analysis (n = 94 plot-level observations). The quantitative traits were subjected to multivariate analyses, including MANOVA, principal component analysis (PCA), Pearson correlation assessments, and plot-level repeatability estimations, following appropriate assumption diagnostics. Categorical traits were assessed using Fisher’s Exact Test and were subsequently integrated with continuous variables through factorial analysis of mixed data (FAMD). Significant multivariate divergence was detected among accessions (MANOVA <i>p</i> &lt; 0.001). Fruit and seed length exhibited relatively high within-trial repeatability (H<sup>2</sup> = 0.56–0.64), representing potential indicators for preliminary assessment under the conditions of this trial, although gains in size must be balanced against processing efficiency. PCA showed that the first two components explained 64.0% of total variance, separating floral traits from fruit and seed dimensions. Strong correlations were noted among floral traits (r = 0.92, <i>p</i> &lt; 0.001), while associations between fruit and seed characteristics were moderate (r = 0.47, <i>p</i> &lt; 0.001). The FAMD analysis further illustrated that the categorical descriptors of fruit, such as colour and shape, showed association patterns with quantitative variations, elucidating 34.8% of the total variance. These findings provide a provisional phenotypic baseline, with multi-location trials across Peninsular and East Malaysia required to assess genotype-by-environment interactions and validate selection stability.</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

From flower to seed: a multivariate analysis of phenotypic diversity and trait relationships in Malaysian Liberica Coffee (Coffea liberica) accessions

  • 期刊论文

  • 发表日期：

    2026年7月10日

查看全文
  • Rupini Yesudasan,
  • Noor Syahira Nasarudin,
  • Vanitha Mariappan,
  • Umikalsum Mohamed Bahari,
  • Radhiah Kasumen,
  • Mohamad Azlan Masri,
  • Mohamad Zulfadli Kamarudin

摘要

Climate change threatens global coffee production, necessitating the use of underutilised species such as Coffea liberica W.Bull as alternatives. Despite its potential, C. liberica germplasm in Peninsular Malaysia remains largely uncharacterised, representing a critical knowledge gap as Southeast Asian coffee markets expand. This study presents the first comprehensive multivariate analysis of 33 Liberica coffee accessions from the MARDI germplasm collection at Kluang, Johor. Using a randomised complete block design with three replications, 13 morphological traits encompassing floral, fruit, and seed characteristics were evaluated. Ten samples per plot were collected and averaged to generate independent plot-level means for statistical analysis (n = 94 plot-level observations). The quantitative traits were subjected to multivariate analyses, including MANOVA, principal component analysis (PCA), Pearson correlation assessments, and plot-level repeatability estimations, following appropriate assumption diagnostics. Categorical traits were assessed using Fisher’s Exact Test and were subsequently integrated with continuous variables through factorial analysis of mixed data (FAMD). Significant multivariate divergence was detected among accessions (MANOVA p < 0.001). Fruit and seed length exhibited relatively high within-trial repeatability (H2 = 0.56–0.64), representing potential indicators for preliminary assessment under the conditions of this trial, although gains in size must be balanced against processing efficiency. PCA showed that the first two components explained 64.0% of total variance, separating floral traits from fruit and seed dimensions. Strong correlations were noted among floral traits (r = 0.92, p < 0.001), while associations between fruit and seed characteristics were moderate (r = 0.47, p < 0.001). The FAMD analysis further illustrated that the categorical descriptors of fruit, such as colour and shape, showed association patterns with quantitative variations, elucidating 34.8% of the total variance. These findings provide a provisional phenotypic baseline, with multi-location trials across Peninsular and East Malaysia required to assess genotype-by-environment interactions and validate selection stability.