The role of vegetation on gully erosion stabilization at a severely degraded landscape: A case study from Calhoun Experimental Critical Zone Observatory

Development and validation of process-based gully erosion models is difficult because observations of the formation and progression of gullies are limited. In this study, analytic formulations of the two dominant gullying processes, namely, plunge pool erosion and slab failure, are utilized to simulate the gullying processes in the 4-km2 Holcombe's Branch watershed. In order to calibrate parameters of the gully erosion model, gully features (e.g., depth and area) extracted from a high-resolution LiDAR map are used. After the calibration, the gully model is able to delineate the spatial extent of gullies whose statistics are in close agreement with the gullies extracted from the LiDAR DEM. Several simulations with the calibrated model are explored to evaluate the effectiveness of various gully remedial measures, such as backfilling and revegetation. The results show that in the short-term, the reshaping of the topographical surface by backfilling and compacting gullies is effective in slowing down the growth of gullies (e.g., backfilling decreased the spatial extent of gullies by 21-46% and decreased the average depth of gullies by up to 9%). Revegetation, however, is a more effective approach to stabilizing gullies that would otherwise expand if no gully remedial measures are implemented. Analyses of our simulations show that the gully stabilization effect of revegetation varies over a wide range, i.e., leading to 23-69% reduction of the spatial extent of gullies and up to 45% reduction in the depth of gullies, depending on the selection of plant species and management practices.