Soil erosion is a major environmental concern, with the potential to severely impact soil and water quality. Assessments of soil erosion are normally carried out using model predictions. Cesium-137 can be used to provide estimates of soil erosion at a landscape scale, and it remains the best tool to make such assessments. Two major weaknesses affect the conventional use of this assessment technique. (i) The period of time that assessments span from about 1960, when the peak of atmospheric deposition of 137Cs occurred, to the date of sampling. As such, is not possible to assess current land management practices and their impacts on soil erosion. (ii) The use of reference sites to estimate the level of 137Cs deposition at a sampling site. Uncertainties in reference values can exceed 20%, greatly affecting the reliability of soil erosion estimates.
The objective of this study was to use repeated sampling for 137Cs to provide more accurate estimates of soil erosion for current land management practices on a cultivated field near Cyrus, Minnesota. Samples were collect in 1999 and again in 2007 from two, perpendicular transects. These samples were characterized in the field for pedologic classification and they were analyzed for soil organic carbon and inorganic carbon at the USDA-ARS laboratories in Morris, Minnesota. Samples were analyzed at the University of Manitoba for 137Cs. The repeated sampling for 137Cs provided highly accurate estimated of soil erosion over the 8-year interval. The erosion rates during that time were very high with losses exceeding 100 t ha yr-1 on convex upper slope positions. The strong interactions between water and tillage erosion were also evident. As well, geostatistical analyses of the soil carbon and soil erosion data were examined to assess the relationship between soil erosion and soil carbon.