The net greenhouse gas balance (NGHGB), estimated by combining direct and indirect greenhouse gas (GHG) emissions, can reveal whether an agricultural system is a sink or source of GHGs. Currently, two types of methods, referred to here as crop-based and soil-based approaches, are widely used to estimate the NGHGB of agricultural systems on annual and seasonal crop timescales. However, the two approaches may produce contradictory results, and few studies have tested which approach is more reliable. In this study, we examined the two approaches using experimental data from an intercropping trial with straw removal and a tillage trial with straw return. The results of the two approaches provided different views of the two trials. In the intercropping trial, NGHGB estimated by the crop-based approach indicated that monocultured maize (M) was a source of GHGs (-1315 kg CO2-eq ha(-1)), whereas maize-soybean intercropping (MS) was a sink (107 kg CO2-eq ha(-1)). When estimated by the soil-based approach, both cropping systems were sources (-3410 for M and -2638 kg CO2-eg ha(-1) for MS). In the tillage trial, mouldboard ploughing (MP) and rotary tillage (RT) mitigated GHG emissions by 22,451 and 21,500 kg CO2-eq ha(-1), respectively, as estimated by the crop-based approach. However, by the soil-based approach, both tillage methods were sources of GHGs: -3533 for MP and -2241 kg CO2-eq ha(-1) for RT. The crop-based approach calculates a GHG sink on the basis of the returned crop biomass (and other organic matter input) and estimates considerably more GHG mitigation potential than that calculated from the variations in soil organic carbon storage by the soil-based approach. These results indicate that the crop-based approach estimates higher GHG mitigation benefits compared to the soil-based approach and may overestimate the potential of GHG mitigation in agricultural systems.
