Relating 137Cs and sediment yield from uncultivated catchments: the role of particle size composition of soil and sediment in calculating soil erosion rates at the catchment scale

Abstract

Abstract Purpose Soil loss is considered one of the most important consequences of land degradation as it affects the production of agricultural and forested areas, and the natural equilibrium of aquatic ecosystems downstream. For these reasons, the availability of tools and techniques able to identify areas at risk of land degradation is essential. Over the last 3–4 decades, theoretical models, based on the use of 137Cs, an anthropogenic radiotracer, proved to be very effective for this purpose. However, these models require specific information on soil and sediment particle size to provide estimates of soil erosion or deposition and this information is summarised by a particle size correction factor ‘P’. Empirical methods of calculation of this factor assume the basic hypothesis that a particle size selectivity takes place in erosion processes and this results in a general enrichment of the fine component in sediments and a corresponding higher radionuclide activity. In this contribution, we demonstrate that this hypothesis is not valid everywhere, and consequently, the P factor cannot be estimated using traditional approaches. Materials and methods A long-term experiment, conducted in Southern Italy and based on two small experimental catchments (approximately 1.5 ha in size), for which measurements of sediment yield are available for the period 1978–2020, is used in this work. More specifically, 137Cs measurements carried out within the catchments and on a reference area provided the basis to obtain long-term estimates of soil erosion rate in these sites. Combined measurements of 137Cs activity and particle size on both soils and sediments, obtained for 46 events, were also carried out to explore possible particle size effects on the final estimates of soil loss. Results and discussion Particle size analyses of soil and sediments showed that there is evidence of a general enrichment of the eroded soil in the finer size fractions. Conversely, radiometric analyses revealed that 137Cs activity in sediments is generally lower than that in surface soil. These results reflect both the decreasing 137Cs activity associated with depth in undisturbed soils and the higher specific surface area of the deeper horizon in these soils. These findings preclude the application of the available empirical models to calculate P, and suggest the opportunity to use, for long-term estimates of soil erosion, a particle size correction factor P = 1. This assumption and an uncertainty analysis associated with the spatial variability of the 137Cs reference value were incorporated into the Diffusion and Migration Model (DMM) to obtain estimates of soil erosion rates for the study catchments. Conclusions The final estimates of soil erosion provided by the DMM showed values very close to the measurements of sediment yield obtained for the two catchments during the study period. The overall results demonstrated that the DMM, if properly calibrated using specific information of particle size and of 137Cs reference value, can be considered a useful tool to individuate areas more prone to risks of land degradation and to identify appropriate strategies able to reduce soil loss in forested sites.