Convergent multi-index selection identifies field-validated nitrogen-efficient donors in a 3K-derived rice panel
Abstract
Improving nitrogen use efficiency (NUE) in rice requires selection of donors that can sustain grain yield under limited N conditions by utilizing acquired N efficiently and partitioning it towards grain. Usually, donors have been identified by employing single index or controlled-environment screening rather than testing in fields. Present study utilised 234 rice accessions grown in field (2022, 2023) with recommended (CN, 120 kg N ha^−1) and zero-applied N (LN, low N). Eighteen traits including GY, total N uptake (TNU), N harvest index (NHI), N-utilisation efficiency (NUtE), nitrate reductase (NR) and glutamine synthetase (GS) activities were measured. Mixed models exhibited broad-sense heritability (H^2) and environment-wise best linear unbiased predictions which supported multivariate clustering and exploratory path and structural equation analysis. Under LN, decrease in GY, TNU, NR and GS were observed while NUtE and NHI increased, exhibiting greater reduction in TNU than GY. The H^2 was high for GY, NUtE and NHI. The N-efficient clusters showed combined high GY, N uptake and partitioning. Accessions selected by multi-trait indices (MTSI, MGIDI and Smith–Hazel) resulted in 15 consensus genotypes, of which 12 passed the LN-stress filter (GY under LN and stress tolerance index). These donors showed higher values for GY and NUtE as well as reached heading date about 21 d earlier as compared with overall genotypic mean. Further, they were statistically comparable to IR-64 (check) for GY, NUtE, NHI and TNU. Thus, the selection framework and donors offered a field-tested genetic resource for rice NUE improvement based primarily on GY and N uptake under N limitation.