IMPACTS OF CONTEMPORARY SEDIMENT YIELD ON THE USE OF IN SITU 10BE IN RIVER SEDIMENT AS A TOOL FOR ESTIMATING BACKGROUND EROSION RATES

In situ ¹⁰Be concentrations in river sediment are thought to be well buffered against changes in contemporary erosion rates because most sediment originates from well-mixed upper soil horizons. However, the buffering capacity of ¹⁰Be as a long-term erosion rate monitor can be undermined in catchments where long- and short-term, sub-basin erosion rates differ in time and space due to varying land-use, slope, and lithology.

We measured in situ ¹⁰Be in 54 samples of river sediment collected in Yunnan, SW China, and used the data to calculate apparent long-term erosion rates which vary between 12 and 195 m/My. Samples were collected from three tributary basins of the Mekong River between 21° and 27° N that are 200, 1000, and 2000 km² in area. Total relief for each catchment ranged from 1180 to 2000 m, and mean annual precipitation ranged from 870 to 1300 mm/yr. Mean slope was the best predictor of erosion rate (R²=0.37, p<0.001).

At 13 stream junctions, we collected samples from each contributing stream and below the confluence. For each set of samples collected upstream of a junction, we estimated the downstream concentration of ¹⁰Be by convolving the long-term sediment flux, derived from the ¹⁰Be erosion rate, the sub-basin contributing area, and the measured sub-basin ¹⁰Be concentration (e.g. Granger et al., 1996). When we compared the measured and calculated downstream concentration of ¹⁰Be, we found they did not agree well. Measured concentrations ranged from -48% to +79% of the calculated concentration; however, the mean difference for the 13 junctions is within 0.5% of the calculated value indicating that on average, the downstream ¹⁰Be concentrations are meaningful.

Results suggest that the apparent long-term sub-basin sediment flux estimated from ¹⁰Be erosion rates and used to calculate isotope concentration downstream of junctions does not equate to the modern sediment flux in all instances. This discordance occurs because altered contemporary sediment yield in a sub-basin relative to long-term sediment yield influences the ¹⁰Be concentration in sediment downstream of junctions. Intensive land-use such as agriculture, clear cutting, or dams, therefore, can change mixing ratios between sub-basins in a drainage network away from the long-term average, introducing bias in in situ ¹⁰Be samples collected downstream.