Reply to Yan and Akiyama: Nitrous oxide emissions from rice and their mitigation potential depend on the nature of intermittent flooding

Our fundamental message (1) is that under intense forms of intermittent flooding—a technique used to reduce methane emissions from rice farms (2)—emissions of nitrous oxide (N₂O), a long-term climate forcer (Fig. 1), can be very high.

Fig. 1.

General understanding of climate impacts of rice farms under continuous flooding or alternate wetting and drying (AWD) (10) compared with our highest rice-N₂O from two studies (1, 7) to highlight that N₂O, not methane, is the dominant greenhouse gas emitted under intense forms of intermittent flooding.

In response to Yan and Akiyama (2), we stand by our statement (1) that “none of the major rice-producing countries…officially report rice-N₂O or related emission factors” [(EFs); percentage of added fertilizer-N converted to nitrous oxide (N₂O)]. Although China, India, and Indonesia (3–5) mention that rice farms can produce N₂O and/or refer vaguely to using country-specific EFs, they do not provide EFs for rice or clarify what fraction of soil N₂O, if any, came from rice. If these countries have used an average rice-N₂O EF of 0.3% from the Intergovernmental Panel on Climate Change guidelines (6), it is a hundred times smaller than our highest EF (31%). This just reinforces our assessment that the potential of high rice-N₂O should be a consideration in these inventories.

We are consistent in comparing the highest rice-N₂O flux from among replicates reported by Lagomarsino et al. (7) with the highest flux from among our replicates. Our highest average rice-N₂O flux (22.7 kg N₂O·ha⁻¹·season⁻¹, Farm 3 2012) would still be approximately three times the previous highest (7.7 kg N₂O·ha⁻¹·season⁻¹) (7).

Yan and Akiyama (2) question both the extent of mitigation possible through water management and our assertion that nitrogen management is not central to reducing rice-N₂O. While N availability is necessary for N₂O production, the N use rate was not central to rice-N₂O, which was very low at two farms where the added N rate was very high (Farms 2 and 4). When farms are flooded and soil-oxygen content is low, either denitrification is limited by availability of oxidized N due to reduced nitrification or N₂O converts into N₂ (last denitrification step) (8). We do not claim that 90% mitigation can be achieved everywhere, but rather that this was the maximum mitigation observed in our study. We have already pointed out that more research is needed to minimize climate impacts per unit yield (see also SI appendix, figure S38 in ref. 1). Our global risk and mitigation analysis (9) suggests that up to 60% of the net climate impact of irrigated rice farms could be mitigated through water management without changing inorganic or organic fertilizer rates.

One study can determine neither the average business-as-usual climate impacts of rice cultivation nor the average mitigation potential of any farming technique(s). Our insight is that rice-N₂O is a potentially large problem that needs due attention and can be managed. We invite mapping of flooding regimes at farmer-managed farms (as opposed to research-station plots that have consistent access to water/electricity) and more studies with over 50% sampling intensity at a range of intermittently flooded farms (Fig. 2).

Fig. 2.

Examples of flooding regimes.