A Khon Kaen University team screened 336 Thai indigenous upland-rice accessions for seedling leaf rolling in 2022–2023, then advanced 23 to severe water-deficit tests at seedling, active-tillering and reproductive-flowering stages. Performance depended on developmental stage: ULR014 showed broad resilience and recovery, while ULR026 yielded 14.8 g per plant with a 3.27% reduction and ULR130 14.3 g with a reported 0.00% reduction under flowering-stage stress. Clustering suggested parents with complementary strengths, but controlled phenotyping does not make them release-ready cultivars or establish farmer-field yield.
Key findings
- Of 336 accessions screened at seedling stage, 23 advanced to multi-stage testing; ULR014, ULR018 and ULR092 showed notably low seedling leaf-rolling scores.
- After rewatering at tillering, ULR086, ULR014 and ULR397 recorded 31.5–36.9% tiller regeneration, indicating recovery under the experimental conditions.
- Under terminal drought at flowering, ULR026 yielded 14.8 g per plant with 3.27% loss; ULR130 14.3 g with 0.00% loss; ULR129 13.0 g with 8.45% loss; and ULR014 11.4 g with 12.31% loss.
- ULR014, ULR254 and ULR397 showed high proportions of completely filled grain, while cluster 4—ULR014, ULR254, ULR032, ULR346 and IR62266—displayed multi-stage resilience.
- Cluster 2—ULR026, ULR130, ULR129, ULR075, ULR071 and ULR003—had the lowest mean yield reduction at 13.3%, illustrating that reproductive-stage tolerance need not follow seedling performance.
Why this matters globally
Climate change requires breeding programmes to screen drought response at yield-determining stages rather than rely only on rapid seedling assays. This work offers a multi-stage pipeline and tropical germplasm that could combine early recovery with reproductive-stage yield stability. Its global value lies in breeding logic and parental resources, not a guarantee of yield across environments.
Thai researcher contribution
Khon Kaen University brought Thai indigenous upland-rice diversity into a phenotyping pipeline linking seedling response, tillering recovery and flowering-stage yield. It identifies ULR014, ULR026, ULR130 and ULR397 as parents worth considering for different targets, giving local germplasm a role in tropical food-system adaptation.
Limitations to consider
Advancing 23 lines after a seedling screen could exclude accessions that are weak early but tolerant at flowering. Controlled severe water deficit does not reproduce variation in soil, rainfall, pests and farm management. Yield in grams per plant is not multi-location plot yield; ULR130's 0.00% reduction is therefore condition-specific, not absolute resistance. The study does not establish genes or mechanisms, test crosses, or validate stability across seasons and locations before cultivar use.