Soil Erosion and Deposition Evidence in a Small Watershed Using Fallout Cesium‐137

Abstract

Soil measurements of 137Cs, a fallout component of nuclear testing, were made within a 3.8‐ha research watershed in the Limestone Valley of northern Alabama. Objectives of the study were to: (i) compare 137Cs activity in soil profiles on eroded and depositional landscape positions, and (ii) relate soil erosion and deposition calculated by 137Cs methods to erosion estimated by the Universal Soil Loss Equation (USLE) and to sediment discharge measured at the watershed outlet. Triplicate soil profiles were analyzed for 137Cs activity to a 46‐cm depth in cultivated sites representing slightly eroded ridgetop, eroded side‐slope, and depositional downslope positions. These data were compared with baseline 137Cs measurements in undisturbed adjacent woodland. Erosion calculated by 137Cs activity losses from baseline values were compared with the USLE estimates on slope transects. Mean total 137Cs activities on baseline, eroded, slightly eroded, and depositional sites were 4860, 1804, 2807, and 5367 Bq m−2, respectively. Peak 137Cs activity and depth to zero activity were much greater in depositional than in eroded and slightly eroded sites. Estimated mean annual (1954–1987) soil‐erosion rates were similar between 137Cs and USLE methods for a slope transect with slight erosion (28 and 26 Mg ha−1 yr−1). On eroded side‐slopes, 137Cs estimates of erosion were 27 to 80% higher than the USLE estimates. This may be due to rill erosion unaccounted for by the USLE. The USLE estimated losses between 9 and 52 Mg ha−1 yr−1 for uplands during 1984 to 1988 cotton production. Sediment losses measured below the downslope depositional area of the small watershed (Gilbert Farm Research Watershed) were only 1 to 4 Mg ha−1 yr−1. This indirect evidence of extensive sediment deposition in a representative small watershed of karst terrain was supported by 137Cs activity in depositional soil profiles.

Soil indicators of agricultural impacts on northern prairie wetlands: Cottonwood lake research area, north Dakota, USA

“…Analytical software used was the MAESTRO iI Emulation Software Model A64-BI Version 1.40 (EG&G ORTEC Oak Ridge, TN). The analog of Soileau et al (1990) was used to estimate annual soil losses or gains for upland soils for each basin: …”

“…Cattle and other animals common in the northern prairie wetlands may distort ~37Cs horizons by disturbing and mixing wetland sediments. For these reasons, the method used by Soileau et al (1990) was chosen to estimate average annual soil erosion rates on upland areas surrounding the wetlands. The J~TCs study approached the problem of sedimentation indirectly by examining upland soil loss from the wetland basin side slopes.…”

Tillage erosion: a review of controlling factors and implications for soil quality

“…This spatial pattern did not agree with the pattern that can be expected to result from water erosion. Furthermore, comparison Of 137CS derived erosion rates and patterns with results of water erosion models often showed poor agreement (Dejong et al , 1986;Soileau et al, 1990;Bernard and Laverdiere, 1992). Other studies supplied additional evidence that soil erosion occurred on unexpected locations on sloping agricul tural land, eg, studies of soil profile truncation (eg, Daniels et al, 1985;Verity and Anderson, 1990); of spatial variation in crop productivity (eg, Miller et al, 1988;Cao et al, 1994) or whereby elevation differences between agricultural land and adjacent non-cultivated land were used to assess soil erosion (Govers et al , 1993).…”

Using 137Cs to investigate net soil erosion at two soil benchmark sites in Quebec

“…therefore, thE potential for using t''Cs measurements to document rates of soil loss. Since this pioneering work, the technique has been further developed by a number of workers throughout the world (Kachanoski 1987;Loughran 1989;Ritchie and McHenry 1990;Soileau et al 1990;Martz and de Jong 1991; Walling and Quine l99l).…”