摘要：Soil water is a link between precipitation and the functioning of ecological systems. It is therefore critical to understand exactly how soil water regimes are affected by changes in precipitation. This is especially true for the variable water regimes of savanna ecosystems. Therefore, understanding the effects of precipitation on soil water was the central goal of this article. The hydropedological behaviour of a catena in the Stevenson Hamilton Research Supersite of the Kruger National Park was configured as a conceptual model of catchment modelling framework, a toolbox of various model structures and processes. The model was parameterised using measured hydraulic properties of the soils, and calibrated and validated using measured soil matric potentials and derived actual evapotranspiration (aET) data. The model was then used to simulate hydrological response under five different rainfall scenarios, ranging from 30% drier than the normal rainfall to 30% wetter than the normal rainfall. The scenarios also included rainfall years with fewer but larger rain events, that is, more intense rainfall events. In general, the model performed well with Pearson's correlation coefficient (R) values ranging between 0.66 and 0.87 and between 0.58 and 0.69 for correlations with daily soil matric potential and daily aET, respectively. Scenario analysis indicates non-linearity in the response of hydrological processes to changes in precipitation. This is especially evident in a seven-fold increase in the duration of saturation at the seepage surface associated with a 30% increase in rainfall. In general, the impact of drying conditions (30% below average rain) has a greater influence on soil water contents, overland flow and percolation from the riparian zone to bedrock than a 30% increase in rainfall would have on the same process.
