Laboratory estimates obscure the patterns of GHG emissions from agricultural soils
Résumé
Agricultural management practices aimed at sequestering carbon (C) in soils show synergies with many agroecosystem services, but may come at the cost of increased greenhouse gas (GHG) emissions. We performed a systematic literature synthesis, followed by a meta-analysis and experimental measurements, to analyse whether C sequestration practices generate trade-offs in terms of CO2 and N2O emissions. We performed systematic literature searches in the Web of Science to look for articles published worldwide that experimentally assess the effect of: 1. minimising soil disturbance (reduced or no tillage), 2. diversifying agroecosystems (crop rotations, cover crops,
intercropping), and/or 3. increasing organic matter inputs (crop residue retention, organic amendments) versus standard practices. We only included studies that quantified C sequestration and at least another response variable related to the synergies or trade-offs of C sequestration. We retrieved 771 publications, 537 of which were excluded based on the type of article, a focus on nonsoil habitats, forests or organic soils, or experimental designs not matching our criteria. We included 234 studies that report 572 effects of sustainable practices on 228 sites located in 38 countries.
Experiments averaged 10 years of monitoring and the majority reported effects of increasing organic matter inputs and minimising soil disturbance (88%) in temperate and continental climates (75%).
Sustainable practices effectively sequestered more C than standard practices considering all studies together. In total, we detected 353 effects related to CO2 and N2O emissions, 51% of which were measured as field fluxes and 43% as laboratory fluxes. Taking both types of measurements, C sequestration increased GHG emissions, particularly CO2. However, the response of CO2 and N2O63 emissions (increase, neutral, decrease) of sustainable versus standard agricultural practices varied significantly depending on how emissions were measured (field or laboratory fluxes). This result was confirmed in a subsequent meta-analysis, including 71 articles that report 123 independent
experiments: CO2 and N2O emissions increased significantly when using agricultural practices that sequester C, but the magnitude of this trade-off depended on the use of field or laboratory measurements. This pattern was remarkable for CO2, where lab fluxes (N=48) yielded significantly
larger responses than field fluxes (N=50). Finally, we quantified CO2 emissions, both as field and laboratory fluxes, in a long-term (27 year) experiment in central Spain, comparing standard tillage (mouldboard plough 30 cm), reduced tillage (chisel plough, 15 cm) and no tillage (direct seeding). Field
fluxes measured weekly for a month were significantly lower under no tillage compared to standard tillage, while laboratory fluxes under controlled conditions yielded the opposite result. Independent methodological approaches indicate that the use of laboratory fluxes, particularly to measure CO2, might overestimate the magnitude of the trade-offs of C sequestration in terms of climate regulation services.