The EPIC model was used to simulate soil erosion and soil C content at 100 randomly selected sites in the US corn belt. Four management scenarios were run for 100 years: (1) current mix of tillage practices maintained; (2) current trend of conversion to mulch-till and no-till maintained; (3) trend to increased no-till; (4) trend to increased no-till with addition of winter wheat cover crop. As expected, the three alternative scenarios resulted in substantial decreases in soil erosion compared to the current mix of tillage practices. C content of the top 15 cm of soil increased for the alternative scenarios, while remaining approximately constant for the current tillage mix. However, total soil C to a depth of 1 m from the original surface decreased for all scenarios except for the no-till plus winter wheat cover crop scenario. Extrapolated to the entire US corn belt, the model results suggest that, under the current mix of tillage practices, soils used for corn and/or soybean production will lose 3.2 x 10^6 tons of C per year for the next 100 years. About 21% of this loss will be C transported off-site by soil erosion; an unknown fraction of this C will be released to the atmosphere. For the base trend and increased no-till trend, these soils are projected to lose 2.2 x 10^6 t-C yr-1 and 1.0 x 10^6 t-C yr-1, respectively. Under the increased no-till plus cover crop scenario, these soils become a small sink of 0.1 x 10^6 t-C yr-1. Thus, a shift from current tillage practices to widespread use of no-till plus winter cover could conserve and sequester a total of 3.3 x 10^6 t-C yr-1 in the soil for the next 100 years.
