A large database of individual farmer field data (n = 4,107) for rice production in the Northwestern Indo-
Gangetic Plains of India was used to decompose rice yield gaps and to investigate the scope to reduce nitrogen
(N) inputs without compromising yields. Stochastic frontier analysis was used to disentangle efficiency and
resource yield gaps, whereas data on rice yield potential in the region were retrieved from the Global Yield Gap
Atlas to estimate the technology yield gap. Rice yield gaps were small (ca. 2.7 t ha 1, or 20% of potential yield,
Yp) and mostly attributed to the technology yield gap (ca. 1.8 t ha 1, or ca. 15% of Yp). Efficiency and resource
yield gaps were negligible (less than 5% of Yp in most districts). Small yield gaps were associated with high input
use, particularly irrigation water and N, for which small yield responses were observed. N partial factor productivity
(PFP-N) was 45–50 kg grain kg 1 N for fields with efficient N management and approximately 20%
lower for the fields with inefficient N management. Improving PFP-N appears to be best achieved through better
matching of N rates to the variety types cultivated and by adjusting the amount of urea applied in the 3rd split in
correspondance with the amount of diammonium-phosphate applied earlier in the season. Future studies should
assess the potential to reduce irrigation water without compromising rice yield and to broaden the assessment
presented here to other indicators and at the cropping systems level.
