Estimation of the stability of topsoil aggregates in areas affected by water erosion using selected soil and terrain properties

The stability of soil structure is crucial for providing most of the essential functions of soil. In our previous research in a region of Chernozem soil, multiple linear regression analysis showed that soil aggregate stability within this morphologically diverse study area could be predicted using the soil organic matter content and one of the terrain attributes. Therefore, this present study aims to determine if a similar procedure can be applied to areas of different soil types, namely a Haplic Luvisol, two Haplic Cambisols, and a Calcaric Leptosol. Basic soil properties such as the organic carbon (Cₒₓ) content, Fe content, Mn content, CaCO₃ content, pH, magnetic susceptibility, bulk density (ρz), electrical conductivity, and water-stable aggregate (WSA) index were measured, and terrain attributes were derived using a system for automated geoscientific analyses and a digital elevation model. It was found that it was not possible to propose similar predictive models for all locations. The models included different combinations of the Cₒₓ content, CaCO₃ content, Fe content, ρz, and various terrain attributes. The Cₒₓ content was the main parameter that positively affected the WSA index values in the areas of Luvisol and Leptosol soils. Magnetic susceptibility was the main parameter that positively correlated with the WSA index in the two areas of Cambisols, and in one Cambisol area, the WSA index was also positively correlated with the ratio between amorphous and crystalline Fe forms. Negative impacts were observed for ρz (Luvisol and one Cambisol) and the CaCO₃ content (Leptosol). From the terrain attributes, slope was mostly included in the predictive models. The other attributes were the catchment slope and profile curvature. The best model prediction assessed based on the coefficient of determination was found for the Leptosol area (R² =0.69), followed by the Luvisol are (R² =0.60) and both Cambisol areas (R² =0.52, respectively 0.33).