How well can we assess impacts of agricultural land management changes on the total greenhouse gas balance (CO2, CH4 and N2O) of tropical rice-cropping systems with a biogeochemical model?

Paddy rice is the main cropping system in Southeast Asia. However, water scarcity arising from competition from other sectors, rainfall variability and climate change increasingly challenges global rice production. One option to adapt to lower water availability is switching from paddy rice to less irrigation intensive upland cropping systems. Such land management change (LMC) is likely to significantly affect ecosystem carbon and nitrogen cycling and its greenhouse gas (GHG) balance. This study evaluates how well the ecosystem model LandscapeDNDC is able to simulate observed emissions of methane (CH4), nitrous oxide (N2O) and carbon dioxide (CO2) from different tropical cropping rotations, i.e., double- and triple-cropped paddy rice, aerobic rice–paddy rice and maize–paddy rice (rice: O. sativa, maize: Zea mays) and how management changes to rice dominated lowland systems will affect the GHG balance on short (a few years) and long (several decades) time scales. LandscapeDNDC predicts seasonal emissions of CH4 and N2O across different cropping rotations (including LMC) with R2 values of 0.85 and 0.78 and average underestimations of 15 and 14%, respectively. In addition to emissions of CH4 and N2O, LandscapeDNDC also captures the long-term development of soil organic carbon (SOC). Soil oxygen status, growth of photosynthetic active aquatic biomass as well as decomposability of harvest residues significantly influence SOC development. Simulation results demonstrate that short-term GHG balances after LMC considerably differ from long-term balances. Simulated total GHG emissions three years after LMC are highest for upland crop − paddy rice rotations due to pronounced decomposition of soil organic carbon. In contrast, the total GHG emissions are highest for double cropping of paddy rice and are clearly dominated by CH4 emissions over a longer period of several decades. Simulation results suggest that approx. 2.8–3.4 t C ha−1 yr−1 residue incorporation after harvest is needed to achieve stable SOC stocks in mixed upland crop–paddy rice systems after LMC from double-cropped paddy rice systems.