Characterizing grain quality and breeding implications in rice landraces of contrasting duration groups

Abstract

Traditional rice landraces carry grain-quality and nutritional variation that has largely disappeared from the small number of high-yielding varieties now grown across most of Asia, yet this variation is seldom examined as a whole; cooking behaviour, sensory quality, and nutrient content are usually studied one at a time, and rarely against a duration-matched modern check. This study evaluated 41 landraces alongside three check varieties (ADT 45, Improved White Ponni, and CR 1009 Sub 1), grouped into short-, medium- and long-duration classes, for ten grain and cooking-quality traits, seven organoleptic attributes, and grain iron, zinc and protein content. The trial was laid out in a randomized complete block design with three replications at Pandit Jawaharlal Nehru College of Agriculture and Research Institute, Karaikal, during 2024. Genotypes differed significantly ( p < 0.001) for every quality trait examined. Coefficients of variation were highest for elongation index and kernel length-to-breadth ratio (32.1% and 32.6%) and lowest for water uptake ratio and kernel dimensions after cooking, and broad-sense heritability exceeded 96% for all ten traits, a level attributed mainly to this being a single season trial rather than to unusually weak environmental influence. Hierarchical clustering placed the 44 genotypes into four phenotypic groups, and, notably, the three check varieties fell into three different clusters rather than one shared group, indicating they do not share a common grain-quality profile. A Smith–Hazel selection index built on six non-redundant traits ranked Anaikomban, Improved White Ponni, and Ilupai poo samba highest within the short-, medium- and long-duration groups, respectively; the check Improved White Ponni outperformed every landrace in its group, while the other two checks ranked near the bottom of theirs, showing that check performance was not consistently better or worse than the landraces but depended on the individual variety. A separate six-genotype nutrient panel showed Rathasalai with the highest zinc and protein among short-duration types (though the check ADT 45 had appreciably higher iron), Thanga samba highest for zinc and protein among medium-duration types, and only marginal differences among the three long-duration entries. Organoleptic scoring of all 44 genotypes identified seven, including the check Improved White Ponni as “excellent” for overall acceptability. Taken together, these results indicate that phenotypic divergence in this panel does not track the landrace-versus-check distinction, and point to specific genotypes rather than to landraces as a class as promising parents for future grain-quality and nutritional breeding.