Land use transformations in the Brazilian Savanna: A decade of soil erosion and runoff measurements

Abstract

Changes in land cover and land use (LULC) are one of the main drivers of erosion and runoff. How ever, most research has relied on short-term observations and only focused on one or two land cover types. We investigated the long-term trade-off between common agricultural land covers (sugarcane, pasture, and soybean), runoff, and soil loss rates. We compared these to native forest (wooded Cerrado) and bare soil. The field observations were done in 100 m² experimental plots in Brazil maintained during the past 10 years. The paper provides three main contributions: (1) long-term runoff and soil loss rates of plots under different LULCs, (2) comparison of runoff, soil loss, and pedological characteristics between plots constructed 10 years apart, and (3) analysis of the trade-off between different LULCs. When ranking land covers based on runoff and soil loss rates, there is a shift in ranking positions, making it difficult to determine which one is more environmentally harmful. However, it is evident that whatever agricultural practice is used, there is a significant impact when compared to native forest. For example, the area converted to pasture resulted in almost 20 times higher runoff, while conversion to sugarcane resulted in 5 times higher soil loss. Not only land cover plays a major influencing factor, but also weathering exposing time. Areas under the same land cover and environmental conditions had different rates of soil loss and runoff due to long-term exposure effects such as soil crusting. Our findings have high relevance for the hydrological and agricultural community by demonstrating (i) the magnitude of trade-off in terms of soil loss and runoff due to land cover changes and (ii) that soil loss should not be assumed to be a linear process over time, as it is commonly assumed.