The land use and land use and management changes are among the main determinants of the stocks of soil organic carbon (SOC). The spatially explicit dynamic modeling is a technique that has been successfully used in evaluations of regional SOC stocks changes. Thus, this study aimed to use the Century 4.0 model associated with geoprocessing and remote sensing techniques to evaluate the effects of agricultural use and soil management changes on dynamics of soil organic carbon (SOC) on farms in Santana (Ijuí, RS, Brazil). A geospatial database of the thematic plans: soil classes, land cover, landforms and beginning of agricultural use, were created in ARCGIS 9.x. After land use and soil management systems had been reconstituted from historical and current surveys, the model was initialized with local soil and climate data and calibrated in two steps. In the first calibration step, internal model parameters based on maize, wheat and soybean biomass C additions to the soil were adjusted, as well as SOC measured in 2007 (0-20 cm layer) of native forest (original condition) and an adjacent farm, where agriculture had been initiated between 1901 and 1930 (older plantation, area 1), sampling an Oxisol, on hilltops. Afterwards, the Century model was statistically validated based on data from 10 fields, initiated at different times, with the same soil type and landscape, sampled in 2007 (0-20 cm layer) divided into four homogeneous areas (Areas 1, 2, 3 and 4) within the district of Santana. In the calibration for generalization, four "subtropical forest " options were created to represent the native forest of each soil type in (Oxisol, Mollisol and two Entisol types) based on the tropical forest option generated in the primary calibration step; the adjustments of this step were maintained and only the parameter related to maximum gross monthly production of forest biomass was changed to stabilize the output variable SOMTC (simulated SOC), after running the model for 3000 years (equilibrium run), matching the measured SOC stocks of each soil type under native vegetation. After calibration, all possible combinations of soils, landforms, land use cover and land use change for the agricultural use of Santana ((1901 to 2007; 54 simulation units) were simulated by the i-Century interface. The model estimated the C dynamics adequately, reproducing the evolution of SOC stocks in the area. According to the estimates, the adoption of conservation management practices ensures the reduction of C losses caused by erosion and microbial decomposition, resulting in increases in SOC stocks.
soil organic matter; carbon modeling; spatially explicit dynamic modeling; management systems
