Assessing the potential effect of different land cover management practices on sediment yield from an abandoned farmland catchment in the Spanish Pyrenees

Abstract

Purpose: Sediment delivery from headwater catchments to reservoirs is a serious threat to reservoir sustainability, and is a critical issue in Mediterranean environments where water resources are scarce. In this study we assessed the consequences of two landscape management scenarios (natural vegetation recovery and scrub clearance) on soil erosion and sediment yield. The results were analyzed in relation to predicted and measured rates of soil erosion and sediment yield, with the aim of promoting better management practices. Materials and methods: The study area was the Arnás River catchment (284 ha), which is located in the central Spanish Pyrenees; the area includes abandoned and poorly managed fields. The combination of the RUSLE and SEDD models of soil erosion and sediment delivery was evaluated in terms of its ability to predict annual rates of sediment yield, using field measurement data for seven water years at the gauging station. The consequences of natural plant succession in other areas of the Spanish Pyrenees, and scrub clearance practices implemented by certain regional governments to increase grazing meadow areas and reduce the incidence of wildfires, were spatially analyzed using GIS. The main sediment source areas were identified and their specific and total sediment yield rates were calculated. Results and discussion: The predicted soil loss under existing conditions was 2.6 Mg ha–1 yr–1, with 5% of the surface area affected by rates greater than 2 tons. The measured sediment yield (SY) range was 69–534 Mg yr–1. The maximum SY detected was associated with an extraordinary debris flow. The predicted rates were strongly correlated to measured rates (Nash–Sutcliffe coefficient = 0.72). The main sources of were alluvial deposits (Specific Sediment Yield = 51 Mg ha–1 yr–1), bare soil (SSY = 12), unpaved trails (SSY = 11), lots (SSY = 4) and pastures (SSY = 1). Under a scenario of vegetation recovery, decreases of 3, 17 and 16% in soil loss, and sediment delivery and yield (respectively) are predicted, whereas increases of 15, 5 and 2% are predicted following scrub clearance practices. Conclusions: Coupling the RUSLE and SEDD models enabled estimation of annual values of soil erosion and delivery in monitored and unmonitored catchments of small and medium size, making this approach a useful tool for risk analysis. Management practices that combine fire-risk control, by the implementation of scrub clearance practices, with the effects of plant succession on the sediment production are suggested as the best management strategy.