Storage or loss of soil active carbon in cropland soils: The effect of agricultural practices and hydrology

Few studies have simultaneously addressed the issue of the short- and long-term hydrological control of organic carbon (OC) export from soils and the role of the leaching process in the long-term dynamics of the soil OC pool. We combined short- and long-term approaches by investigating dissolved organic carbon (DOC) at the outlet of a small drainage catchment and establishing a relationship between DOC concentrations (3.5 +/- 1.8 mgC L-1 on average) and subsurface runoff (175 mm yr(-1) on average). We then calculated the annual DOC export as a function of average annual water runoff for a 9-year period. We assumed that the annual flux of leaching is proportional to the active soil OC stock, which we compared with data from the literature. We added a leaching function to the AMG two-compartment model of soil carbon dynamics. The innovative use of the Rock-Eval method for agricultural soils made it possible to determine the stable and active carbon fractions (OCp and OCA, respectively), necessary to characterize the system in the model, for 52 plots in organic and conventional agricultural farms in the Seine Basin. No significant difference was found in OC for the two agricultural systems (11.4 +/- 2.5 gC kg(-1) vs. 12.3 +/- 4.2 gC kg(-1), respectively, for the 0 to 30 cm layer). Using the AMG model with its leaching function, we calculated the equilibrium value of OCA, representing the size of the OCA pool that would be reached in the long term under constant farming practices and hydrological conditions in a given plot. Deviation from this equilibrium indicates whether carbon storage or loss occurs. Overall, for the plots sampled in the Seine Basin, an annual carbon loss of similar to-0.24 % yr(-1) of the total OC pool was found. This may increase by 15% (i.e., to similar to-0.28% yr(-1)) under higher subsurface runoff, which is plausible under ongoing climate change (e.g., 600 mm yr(-1) vs. 175 mm yr(-1) currently observed).