Methane and nitrous oxide emissions from paddy fields in Japan: An assessment of controlling factor using an intensive regional data set

Rice paddy fields, producing a major staple food to support growing world populations, represent a major source of greenhouse gases (GHGs) from agricultural ecosystems. The GHG emissions, mainly as CH4 and N2O from paddy ecosystems, are highly sensitive to both environmental and management factors. Yet the identification of specific factors, a fundamental step for GHG inventory and mitigation, is often limited by data availability. Here, we compiled 572 and 174 data on CH4 and N2O emissions, respectively, from paddy fields across Japan, which arguably represents the most intensive GHG data set from paddy fields per region. We hypothesized that statistical analyses of the intensive data set allow the identification of key factors and possible mechanisms that have not been fully appreciated in the previous studies. Important environmental factors newly identified for CH4 emission were soil type and precipitation pattern. The soil emitted CH4 the most was Histosols (172% higher) and the least was Andosols (32% lower) compared to the other soil types. Our analysis also revealed that the region of severe summer rainfall (southwestern Japan) tended to have higher CH4 emission. The most critical management-related factor was straw incorporation and its timing had significant impact as previously reported. Specifically, CH4 emission was 242% and 59% higher by pre-puddling and post-harvest incorporation, respectively. The CH4 response to straw incorporation had relatively large uncertainty, which partly resulted from the variation in straw mass and soil type (esp. Andosols). In addition, the soils having inherently low CH4 emission due presumably to more oxidized conditions had significantly higher response to straw incorporation. Organic amendment increased CH4 by 35%, while water management effect was unclear. We also found that N2O accounted only for 5.5% of total global warming potential from the paddy fields and was mainly emitted in fallow season (84% of annual emission). The amount of nitrogen fertilizer added, the commonly-used factor to estimate N2O emission (e.g., IPCC guideline) showed no significant relationship with the N2O emission in rice growing season, which may be explained by very low level of fertilizer application in Japanese paddy fields (typically < 100 kg ha(-1) y(-1)) compared to other parts of the world. While some of the findings are unique to specific regions (e.g., Andosols), new findings on the factors and potential mechanisms controlling GHG emissions from rice paddy ecosystems would be useful to develop strategies for regional GHG estimate and for modeling biogeochemical cycle in rice paddy ecosystems.