Overestimation of N₂O mitigation potential by water management in rice paddy fields

In PNAS, Kritee et al. (1) report extremely high nitrous oxide (N₂O) fluxes from rice paddy fields with intermittent irrigation and alternate wetting and drying (AWD) and conclude that N₂O can be reduced by up to 90%—with nitrogen management not playing a central role. However, we believe that this conclusion could potentially mislead readers by exaggerating the data, causing inaccurate interpretation. In their article, we observe the widespread selective use and misinterpretation of works cited.

Kritee et al. (1) begin by stating that N₂O emissions were not included in previous estimates of greenhouse gas (GHG) emissions from rice cultivation: “None of the major rice-producing countries, including the two leading rice producers, China and India, officially report rice-N₂O or related emission factors in their national GHG inventories submitted to the United Nations.” However, China, India, and Indonesia (2–4), the top three rice-producing countries, have included rice-N₂O emissions in their national communications according to the tier 1/tier 2 method in the Intergovernmental Panel on Climate Change (IPCC) guidelines (5, 6). Notably, rice-N₂O emission is listed in the agricultural soils category in the IPCC guidelines and not in the rice cultivation category. Numerous other studies reported increases in rice-N₂O emissions from AWD (7, 8).

Kritee et al. (1) report extremely high N₂O emissions from an intermittently flooded rice farm in India (33 kg N₂O·ha⁻¹·season⁻¹; three times higher than a previously recorded extreme value of ∼10 kg N₂O·ha⁻¹·season⁻¹). Actually, the previous extreme reported by Lagomarsino et al. (9) was 7.7 kg N₂O·ha⁻¹·season⁻¹. Therefore, the value reported by Kritee et al. was four times higher than that previously reported. This extreme value of 33 kg N₂O·ha⁻¹·season⁻¹ corresponded to one of the three replicates of a single treatment (dataset S1 in ref. 1), and the average of these triplicates was 17.4 kg N₂O·ha⁻¹·season⁻¹. It is unusual and misleading to state the value of a single replicate without providing the treatment mean in the abstract. Although such an extreme value can be observed in a single field given the large variability in rice-N₂O, it is highly risky to extrapolate an extreme value of a single field to a subcontinental/global scale. In the supplemental text, Kritee et al. emphasize the importance of high sampling frequency. Indeed, low sampling frequency will result in higher uncertainty but not necessarily in underestimation. Low sampling frequencies can also overestimate N₂O emissions, even by a factor of 935% (10).

Kritee et al. (1) conclude that rice GHG emissions could be reduced by up to 90%, implying that the reduction can be achieved by water management and not by nitrogen management. However, rice yields using “alternate” practices (APs) were severely reduced compared with those using “baseline” practices (BPs) in their study. The only farm that did not lose rice yields using APs had increased N₂O emission compared with those using BPs. In other farms, the average yield loss was 24%, with a maximum of 44%, which is unacceptable to farmers. The authors argue that yield reduction is not associated with net climate impact reduction. However, in their study, N₂O emission reduction was achieved by decreasing the inorganic nitrogen input by 71%, which is unsustainable.