Grain quality characterization of indigenous rice (Oryza sativa L.) landraces using mixed clustering and selection index approaches

Grain quality remains a key determinant of consumer acceptance and market value in rice. Despite the widespread cultivation of high-yielding varieties, many are rejected by consumers due to inferior cooking and eating quality. This study evaluated 345 rice landraces from diverse Indian ecologies to assess eight physical, three biochemical, and five cooking quality traits. Significant variation was observed across all traits. Kernel length (KL) showed a strong positive correlation with kernel length after cooking (KLAC; r = 0.545) and head rice recovery (HRR; r = 0.192), while KLAC was positively associated with elongation ratio (ER; r = 0.575). KL and ER jointly explained 92% of the variation in KLAC. HRR, a critical milling quality trait, was influenced by grain chalkiness and milling percentage, with several long slender landraces showing unexpectedly high HRR (> 60%). Biochemical traits such as amylose content (AC), gel consistency (GC), and alkali spreading value (ASV) revealed significant variation, with most landraces showing intermediate AC (20–25%) and preferred gelatinization temperature (ASV 4–5). Cooking traits such as high KLAC, strong aroma, and high ER were present in several unique landraces. Seven aromatic landraces and 21 low-chalkiness landraces were identified. A multi-trait Inverse Variance Weightage Selection Index (IVWSI) was used to rank landraces, identifying Bagi, Lallusagar, Kalepali, Chhuhia Gurmatia, and Karaki as top performers. These landraces represent valuable genetic resources for developing rice varieties that meet evolving consumer preferences for cooking quality, texture, aroma, and appearance, especially in diverse and high-demand markets.