Quantifying Relative Rates of Upland and Bank Erosion Using Radionuclide Tracers in AN Agricultural Watershed

Watershed evolution is governed by interactions between different earth surface processes that are mediated by water through precipitation and surface-subsurface flows. One key process affecting the evolution of watersheds is the movement of sediment. To evaluate the movement of sediment at the watershed scale, a common practice is to quantify sediment loads at the outlet of a system, i.e., develop a sediment budget. However, most sediment budgets at watershed outlets are net balances of different source areas. Different processes delivering sediment to streams and waterways include erosion of upland soils, collapse of channel banks, and resuspension of bed sediments. Distinguishing material from these different sources requires that each source sediment has a unique set of characteristics (or signature). Only by differentiating the source materials to the suspended sediment load of a stream can accurate measures of upland sediment delivery, bank erosion, and bed resuspension be determined. A simplified method to coarsely differentiate eroded surface soils and channel sediments to the suspended sediment load of a stream was used during single runoff events in a representative agricultural watershed of the US Midwest. The method utilized the activities of two naturally occurring radioisotopes, 7Be and 210Pbxs, as tracers and a simple two end-member mixing model. This model was applied to sediment loads calculated from different methods including discrete grab samples and continuous measurements. The proportion of eroded upland soils was high during the beginning stages of the events, while the channel became the dominant contributor to the suspended load near the peaks and on the falling limbs of the event hydrographs. The primary reason for this change in the dominant sediment source was that the upland source quickly became exhausted, as shown with clockwise hysteresis. Combining the quantitative suspended sediment measurements with the load partitioning analysis revealed that the majority of sediment (~60%) transported during the events was derived from the uplands, while the channel source contributed the remainder of sediment especially during the falling limbs of the hydrographs. These results compared favorably with estimates of bank erosion rates measured using traditional erosion pins and channel surveys, as well as automated photoelectric erosion pins.