Quantifying hillslope erosion and sediment connectivity in the Rogativa catchment, Southeast Spain

Abstract

Soil erosion is a process accelerated by natural and anthropogenic disturbances over time and space, leading to land degradation and causing geomorphological change. It is difficult to investigate the spatial and temporal distribution of soil erosion and sedimentation in data-scare areas, in that case, the use of simplified methods to analyze soil erosion and sediment connectivity variations over time and space can help. Sediment connectivity denotes the transfer of sediment from source to sink areas through channel systems of landscape compartments within a watershed. In this study, we aimed to investigate sediment yield (SY) variation over time and space and understand the link between hillslope soil erosion and sediment connectivity to identify hotspot areas in the Rogativa catchment (∼53 km2) in Southeast Spain. The (specific) sediment yield (S)SY was estimated by combining the Revised Universal Soil Loss Equation (RUSLE) model with the sediment delivery ratio (SDR). The SDR was calculated based on the Index of Connectivity (IC). In the channels, 100% delivery was assumed. In the Rogativa catchment, 58 check dams were constructed in 1976/77. Their trapping efficiency, obtained from field observations of sediment retained behind the checkdams in 2003, was included in the SDR estimation of the checkdams. SY was estimated from accumulated hillslope soil erosion in the local stream network while accounting for sedimentation through the SDR. Soil erosion, IC, SDR, and (S)SY were quantified and compared for the years 1956, 1977, 2001, and 2016, for which different land use maps were available. SY model results for the year 2001 were compared with observed SY (in 2003) behind the check dams. Only for about half of the checkdams, model results were comparable. This is investigated further and could be explained by complex sediment dynamics within the channels and between checkdams (i.e. one check dam retaining part of the sediment, the next downstream checkdam as well, etc) – these dynamics are not included in the RUSLE-SDR model. The RUSLE-generated soil erosion and sediment connectivity signatures (IC, SDR, and (S) SY) showed higher values in the channels and croplands than in hillslopes and decreased over time due to significant changes in land use and construction of check dams in the catchment. Moreover, the combined proportion of erosion-connectivity patterns showed about 7% of the area adjacent to some of the streams was found both highly erodible and highly connected, which indicates an adverse erosion-prone part. It is possible to apply this method to understand SY amount and distribution and identify hotspot locations in drainage systems with limited field data in data-scarce semi-arid areas like the Rogativa catchment. However, more field observations to validate the models to identify hotspot locations and investigate river network systems rather than focusing only on hillslopes, which could help to know where to intervene in the catchment.Keywords: Soil erosion-RUSLE, Sediment connectivity, Sediment delivery ratio, Sediment yield, hotspot location