Linking vegetation cover patterns to hydrological responses using two process-based pattern indices at the plot scale

Vegetation cover pattern is one of the factors controlling hydrological processes. Spatially distributed models are the primary tools previously applied to document the effect of vegetation cover patterns on runoff and soil erosion. Models provide precise estimations of runoff and sediment yields for a given vegetation cover pattern. However, difficulties in parameterization and the problematic explanation of the causes of runoff and sedimentation rates variation weaken prediction capability of these models. Landscape pattern analysis employing pattern indices based on runoff and soil erosion mechanism provides new tools for finding a solution. In this study, the vegetation cover pattern was linked with runoff and soil erosion by two previously developed pattern indices, which were modified in this study, the Directional Leakiness Index (DLI) and Flowlength. Although they use different formats, both indices involve connectivity of sources areas (interpatch bare areas). The indices were revised by bringing in the functional heterogeneity of the plant cover types and the landscape position. Using both artificial and field verified vegetation cover maps, observed runoff and sediment production on experiment plots, we tested the indices' efficiency and compared the indices with their antecedents. The results illustrate that the modified indices are more effective in indicating runoff at the plot/hillslope scale than their antecedents. However, sediment export levels are not provided by the modified indices. This can be attributed to multi-factor interaction on the hydrological process, the feedback mechanism between the hydrological function of cover patterns and threshold phenomena in hydrological processes.