Watershed spatial heterogeneity of soil saturated hydraulic conductivity as affected by landscape unit in the critical zone

The spatial heterogeneity and related driving factors of the saturated hydraulic conductivity (K_s) in the Earth's critical zones (CZs) have been documented widely in the literature. However, information regarding the effect of distinct landscape units (i.e., Slope-land, Gully-land, and Tableland) on such heterogeneity at the watershed scale in CZs characterized by complex topographic conditions remains scant. In the current study, we collected undisturbed and disturbed soil samples from the surface soil layer (0–5 cm) at 1195 sampling sites across the ShuangChaGou watershed (0.22 km²) in the Chinese Loess Plateau (CLP). According to classical and geostatistical analyses combined with the structural equation model (SEM), the mean K_s followed the sequence Slope-land (0.25 cm·min⁻¹) > Gully-land (0.187 cm·min⁻¹) > All samples (0.186 cm·min⁻¹) > Tableland (0.14 cm·min⁻¹) (p < 0.05), with the K_s values exhibiting moderate variations except for the Gully-land (strong variation). The LogK_s spatial variation characteristics (i.e., the anisotropy ratio, effective range, and C/(C₀ + C)) considering the influence of topography varied among the landscape units and the entire watershed. The SEM identified that soil bulk density, saturated soil water content, silt content, plan curvature, NDVI, and slope gradient were the controlling factors in sequence affecting the K_s variation for All samples (AS). In contrast, factors controlling K_s variations were different under the Tableland, Gully-land, Slope-land, and AS. The magnitudes, spatial variations and controlling factors of the surface K_s were greatly affected by the landscape units. Our study provided a robust assessment of the spatial variation of K_s in CZs involving complicated topographic features, highlighting the utility of dividing distinct landscape units when evaluating the K_s variation. A good understanding of this information is helpful for developing distributed hydrological models and for evaluating water and solute transportation in CZs of the CLP and similar regions around the world.