A Decade of Study on Carbon Dioxide Exchange and Methane Emission at Paddy Field in Central Japan

From a decade of study on carbon dioxide (CO₂) exchange and methane emission at Mase paddy flux site (MSE), Central Japan using micrometeorological technique combined with biometric and miscellaneous measurements, it was found that about a half of the net carbon uptake by the paddy field during the growing season was lost abruptly at harvest, and that the remaining half was released gradually to the atmosphere through decomposition of organic matters in the subsequent fallow season lasting for more than seven months. At the MSE site, a local farmer cultivates the most popular rice variety in Japan, and gets average crop yield in the prefecture (5.2 kg ha^-1 in brown rice yield). The coefficient of variation (CV) of crop yield in nine years from 2002 to 2010 was 7%, while the CVs of the total solar radiation and the accumulated temperature in the growing season were 9% and 4%, respectively. The CVs of the total net ecosystem CO₂ exchange (NEE) and evapotranspiration were both 10%. The CV of gross primary production (GPP) estimated from observed NEE was 5%, but it was reduced to only 2% when we excluded GPP in the 2004 growing season, which had extraordinarily large GPP under conditions of high temperature and plenty of solar radiation in mid-growing season. The small inter-seasonal variation in GPP reflects stable Japanese rice production, which is supported by excellent cultivation practices as well as moderate climate conditions.

The net biome production (NBP) was estimated at -0.4±0.7 t C ha^-1 y^-1 on average of nine years. This indicates the ecosystem carbon budget of the study site was almost balanced, although uncertainties originating from partial burning of harvested crop residue and flux measurement in the fallow period using the open-path eddy covariance method still remain. The observed ecosystem carbon budget of the paddy field was generally consistent with monitoring of carbon content in the top soil and its model simulation. The inter-annual variations of NBP were principally caused by management practices such as organic matter application and disposal of crop residue at harvest as well as the CO₂ exchange mentioned above. The disposal of crop residue at harvest also had an impact on methane emission from paddy field. The methane flux measured by the flux-gradient technique displayed noticeable inter-seasonal variations as well as dominant seasonal variations influenced by water management and crop growth. The DNDC-Rice model well simulated not only the seasonal variations but also the inter-seasonal variations of methane emission. The observed and simulated inter-seasonal variations of methane emission suggest that partial burning of crop residue at harvest reduced the methane emission in the next growing season. The carbon budget and methane emission of paddy field may change in future with increases in air temperature and ambient CO₂ concentration, but more probably with changes in management practices, which are influenced by agronomical and socio-ecological factors such as spread of labor-saving cultivation practices and mitigation/adaptation practices for climate change.