Assessment of water erosion dynamics by connectivity indices in an agricultural watershed

ARES, MARÍA GUADALUPE; FAJARDO, LUISA; CHAGAS, CELIO; VARNI, MARCELO; MOZO, JOAQUÍN

Congreso; 21st Worl Congress of Soil Science: ?SOIL SCIENCE: Beyond food and fuel?; 2018

Water erosion is a global issue which affects soil quality and nutrients availability, and threatens agricultural productivity. The hydrological and sedimentological connectivity approach to study this process is being developed to understand its spatial heterogeneity and complexity and the dynamics of connections between hillslopes and drainage networks which redistribute the eroded soil. In this context, connectivity indices constitute important tools for erosion modelling. The objective of this work is to assess the sedimentological connectivity during three rainfall-runoff-soil loss events registered in a small basin (560 ha) under no-tillage agriculture by two indices. The first index, the index of sediment connectivity (IC), evaluates the structural connectivity, and is a measure of potential connectivity. It is calculated by geographical information system, using data derived from a digital elevation model (slope, flow direction, flow accumulation) and a map of factorC from Universal Soil Loss equation to consider the resistance to water and sediment fluxes by vegetation during each event. The second index corresponds to the index for physically connected compartments (Cpc), which considers aspects of structural and functional connectivity: the area physically connected, expressed as the area of the drainage network, and the rate of transported sediment during the events. This index complements IC, because it takes into account the variability of each storm and its hydro-sedimentological response. The IC index showed values of 1.35, 1.85 and 1.63 for event 1, 2 and 3, respectively, while Cpc values were 7.9, 16.4 and 175.9 for event 1, 2 and 3, respectively. The indices evidenced the low connectivity induced by the sheet erosion and the good vegetation cover conditions during event 1. Comparatively, during events 2 and 3 the indexes showed discrepancies: high potential connectivity (IC) for fallow period, but medium Cpc in relation with the connected area and the erosivity of the storm event. In contrast, the IC was medium for case 3, with good land cover condition, but Cpc was the highest related to the high erosivity of the rainfall which generated rills in the area. The results indicate that low vegetation cover and rill formation are two key factors which induce connectivity in the area. The adjustment of these indexes and their incorporation in erosion models may improve the understanding of the temporal dynamics of the erosion process.