A reconstruction of historical changes in sediment sources, sediment transfer and sediment yield for a small, semi-arid Karoo catchment, South Africa

Land degradation is widespread in South Africa but few long-term data sets are available to help identify the timing of this degradation and its impact on catchment sediment yields. We report a case study reconstructing the erosion history of a small upland catchment in the Sneeuberg mountains of South Africa based on the dating of reservoir sediments using 210Pb and 137Cs. Sediment sources have been identified using mineral magnetic measurements. Additional information has been derived from the analysis of aerial photographs and satellite images and the measurement of badland erosion using erosion pins. Results suggest that sediment yields increased dramatically from < 30 t km-2 yr-1 in the early part of the 20th Century to ∼1600 t km-2 yr-1 in the 1970s, subsequently falling to ∼850 t km-2 yr-1 in the last decade. Contemporary erosion pin measurements suggest that badlands are still extremely active and help maintain current high yields. Analysis of aerial photographs and satellite images indicate that badlands and extensive gullies existed in the catchment before 1945. An analysis of mineral magnetic signatures suggested that badlands made some contribution to the reservoir sediment in the 1930s, but did not significantly increase sediment yields. Photographic evidence suggests that a major badland became connected to the main channel network (and to the reservoir) from the early 1960s onwards. This interpretation was confirmed by significant and sustained changes in the mineral magnetic signatures of the reservoir sediments. The results indicated that a delayed increase in sediment yield, despite extensive badland development much earlier, appears to have been caused by changes in catchment connectivity rather than being a direct response to contemporary changes in land degradation within the catchment. These results add additional complications to our understanding of the relationship between the response to, and recovery from, periods of overgrazing in the 1930s that probably initiated the badland development