Multivariate Analysis of Phenotypic Traits to Assess Diversity and Breeding Potential of Rice Hybrids in Pakistan

Abstract

Understanding the genetic architecture and phenotypic diversity within elite germplasm is paramount for accelerating rice breeding programs aimed at enhanced productivity and resilience. This study meticulously analyzed 134 rice hybrids evaluated at KSK over one year, across 13 crucial agronomic and quality traits, employing Principal Component Analysis (PCA) and detailed genetic parameter estimation. ANOVA results confirmed highly significant (P ≤ 0.01 or P ≤ 0.05) phenotypic variations among the 134 hybrids for all measured traits, including Days to Flowering (DF), Days to Maturity (DM), Plant Height (PH), Yield, Grains per Panicle (GPP), and Thousand Grain Weight (TGW), validating the substantial genetic diversity. Significant block effects (P ≤ 0.01 or P ≤ 0.05) were also observed for most traits, underscoring the effectiveness of the experimental design. Trait reliability was high, as indicated by low Coefficients of Variation (CVs), with Yield at 0.51% and DM at 1.56%, ensuring vigorous data quality. PCA effectively summarized the multidimensional data, with the Scree plot demonstrating that the first two principal components, Dim1 and Dim2, cumulatively explained 35.6% of the total phenotypic variance (19.1% and 16.5% respectively), thus capturing the most significant patterns of variation. The PCA biplot revealed that Dim1 was primarily driven by developmental traits, with strong positive loadings for DM and DF. Critically, Yield, GPP, average grain length (AGL), PH, and panicle length (PL) also loaded positively on Dim1, yet their overall vectors indicated an inverse relationship with maturity; earlier-maturing hybrids often exhibited higher yields. Dim2 was predominantly shaped by grain morphology, strongly associating with length width ratio (positive loading) and grain width (negative loading), while head rice recovery (HRR) and elongation ratio (ER) showed limited contribution to these initial components. The distribution of rice hybrids in the PCA space further elucidated phenotypic clustering and environmental influences. Hybrids such as 'Badbaan-07', 'Dhanak (FM - HBS-2001)', and 'EG-1907' consistently clustered in the early-maturing, high-yielding quadrant. Genetic parameter estimation provided quantitative support for breeding potential. Broad-sense heritability (h² (BS) %) was exceptionally high (over 80%) for key traits including GPP (98.52%), TGW (98.36%), Yield (99.84%), HRR (98.98%), PH (91.57%), Tillers per Plant (86.36%), Grain Width (80.99%), Length Width Ratio (94.29%), and ER (94.09%), signifying strong genetic control and high selection efficiency. Genetic Advance (GA) was particularly high for Yield (1396.76) and GPP (74.60), predicting substantial gains through selection. Days to Flowering, Days to Maturity, and AGL also displayed moderate heritability (around 64%), offering reasonable prospects for improvement. In conclusion, this research identified significant genetic diversity and high heritability for critical agronomic and quality traits within the 134 rice hybrids. PCA effectively mapped the complex trait interrelationships and identified promising genotypes based on their multi-trait profiles, especially highlighting the potential of early-maturing, high-yielding accessions. The findings offer invaluable guidance for NARC's breeding program at KSK, facilitating targeted selection of superior germplasm and optimizing breeding strategies for developing high-performing, climate-resilient rice varieties.