Validation of the Century model to estimate the impact of agriculture on soil organic carbon in Southern Brazil

Abstract

Simulation models are important tools in the long-term assessment of soil organic carbon (SOC) stock changes caused by agriculture, but proper parameter adjustments for the underlying environmental context are essential for an accurate assessment of the models ability to predict SOC. The objectives of this study were to calibrate and validate the Century 4.0 model for the analysis of SOC stocks evolution from 1900 (pre-agriculture native vegetation condition) to 2007 (current scenario) and in 2050 (forecast scenario) for farms of the Santana district of the Ijuí municipality (Rio Grande do Sul state) in Southern Brazil. We developed land use and soil management scenarios from historical sources and current surveys. Then we initialized Century with local soil and climate data and calibrated to match observed SOC data by adjustment of internal parameters. Soil organic C stocks for calibration were obtained in 2007 from sampling (0 to 20 cm depth) farm fields under agricultural land use for more than 80 years and an adjacent undisturbed forest tract (representing native soil conditions). Additionally, we adjusted the Century crop parameters to reflect commonly observed corn, wheat and soybean biomass C additions to soils in the study region. We conducted statistical analysis for measured and simulated stocks from ten farms where agricultural land use started between 1901 and 1987. We concluded that the Century model could adequately estimate SOC stock changes as affected by agriculture in this region.