Morphological differentiation and diagnostic key development for Sri Lankan wild rice (Oryza) species

Accurate identification of wild Oryza species is essential for conservation, genetic resource management, and crop improvement, yet morphological overlap among taxa often complicates species delimitation. This study evaluates the taxonomic utility of seed and vegetative traits in distinguishing five Sri Lankan wild rice species: Oryza nivara, O. rufipogon, O. eichengeri, O. rhizomatis, and O. granulata. A comprehensive set of qualitative and quantitative characters, including seed morphometrics, embryo traits, seed coat parameters, panicle architecture, awn morphology, and vegetative features, was analysed using univariate statistics and multivariate approaches. While ANOVA and Kruskal–Wallis tests revealed limited discriminatory power of individual traits, principal components analysis (PCA) based on both seed and vegetative datasets clearly separated the species into distinct clusters, demonstrating that coordinated multivariate trait variation provides stronger taxonomic resolution than single-character assessments. Qualitative characters such as awn presence and length, rhizome occurrence, panicle type, stigma colour, and lemma–palea pubescence further enhanced species discrimination. Quantitative parameters, including grain dimensions, seed shape index, seed coat ratio, embryo-to-seed length ratio, and plant height, exhibited significant interspecific variation and were incorporated into a dichotomous key for reliable field identification. The results confirm that integrated morphological analysis effectively captures species-level divergence within Sri Lankan wild Oryza. This study provides a robust morphometric framework and practical diagnostic tools that support taxonomic clarification, germplasm conservation, and the strategic utilization of wild rice genetic resources.