Watermelon (Citrullus lanatus) is an important fruit crop where understanding genetic variability and trait dissection in segregating populations can accelerate breeding for yield and quality. This study assessed phenotypic variability, transgressive segregation, trait correlations, and multivariate patterns across F₂ to F₅ generations, integrating molecular markers for genetic validation. In the F₂ generation, fruit weight ranged from 1600 to 3200 g with moderate variability and high Narrow sense heritability (0.71), indicating strong genetic control. Transgressive segregation was pronounced, with > 99% of progeny exceeding parental trait ranges. Across F₃–F₅ generations, fruit weight remained variable (1600–4500 g in F₄; 1900–3400 g in F₅), while ovary length, fruit weight, and TSS (%) continued to exhibit transgressive expression. Size-related traits (fruit diameter, rind thickness) were relatively stable. Correlation analysis showed strong positive associations between fruit weight and size traits in early generations, which weakened over time, reflecting recombination and trait decoupling under selection. PCA identified fruit size (PC1, 49.7%) and shape (PC2, 28.9%) as major axes of variation, while TSS (PC3) remained largely independent, suggesting scope for improving sweetness independently of fruit size. K-means clustering separated genotypes into two groups: large, heavy fruits versus smaller, sweeter fruits. Molecular analysis using 110 SSR markers identified 20 polymorphic loci generating 56 alleles (mean PIC 0.45). Seven SSRs confirmed F₁ hybrid heterozygosity, and segregation analysis in 267 F₂ individuals generally fit Mendelian expectations except at locus WmSSR27. STRUCTURE analysis revealed two genetic clusters corresponding to homozygous and heterozygous groups. The results demonstrate substantial genetic variability, significant transgressive segregation, and independent inheritance of key traits. These findings provide valuable opportunities for breeding watermelons with improved yield, fruit shape, and sweetness through multivariate trait dissection and marker-assisted strategies.
Choudhary et al. (Tue,) studied this question.