Understanding ephemeral gully erosion: Relation to the apparent soil erodibility concept

Ephemeral gully erosion is considered one of the dominant sources of soil loss from the agricultural landscape but has been historically overlooked in traditional soil erosion assessment. Efforts to develop an adequate modeling framework addressing ephemeral gully erosion are hampered by the lack of data and research tools tailored to this erosion pathway. Ephemeral gullies are assumed to form where surface flow concentration exerts hydraulic forces exceeding a given threshold for channel initiation. Consequently, models attempting to predict ephemeral gully location and erosion only consider surface flow as the main controlling factor. Subsurface hydrology is often given a secondary role in ephemeral gully erosion since only its effect on surface flow quantity is considered. Recent developments in soil erosion research support the hypothesis of a first order control of subsurface hydrology on soil erosion. This evidential control of subsurface hydrology on soil erosion is a special case of a more general concept describing soil erodibility as an apparent property composed of a constant intrinsic part and an extrinsic component depending on field conditions. Difficulties of current ephemeral gully erosion prediction models to address gully erosion on fields with limited runoff such as those managed under No-Till farming suggest that the apparent soil erodibility concept might play a significant role in gully initiation and development. This thesis uses the specific case where soil erodibility was modified by subsurface hydrology to show the relationship between the apparent soil erodibility concept and ephemeral gully erosion. To complete this study, research tools specifically designed to address ephemeral gully processes were developed. A 9.75 m × 3.66 m hillslope section with controllable surface and subsurface hydrologic conditions was designed and built to adequately represent natural hillslope scale processes. To monitor soil erosion and deposition, channel network development and hillslope form evolution, a digital photogrammetric technique was developed and tested for soil erosion applications. A series of rainfall-runoff experiments was conducted on the hillslope set under drainage or oversaturation (seepage) condition. Soil loss was monitored by collecting runoff samples and by digitizing the soil surface at regular time intervals using digital photogrammetry. Results suggest that subsurface hydrology had a significant impact on erosion rate and channel network development. On average, erosion rates were 1.85 times higher under seepage than they were under drainage. From the elevation change data, we found that channels developed at a rate 1.5 times higher under seepage condition than they did under drainage condition. The mechanism of channel development was also affected by subsurface hydrology. While channel development occurred at a steady rate under drainage condition, two phases of development with contrasting rates were observed under seepage condition. Finally, soil loss and elevation change data were also used to validate key theories in soil erosion modeling. Measured effects of rainfall-runoff intensity on interrill sediment load, support a previously proposed model for interrill erosion relating sediment load to rainfall intensity and to the square root of runoff rate while elevation change patterns observed during each rainfall-runoff event were found to be more consistent with the simultaneous erosion deposition theory as opposed to the sediment transport capacity concept for sediment transport mechanism. This study demonstrated that subsurface hydrology might be a controlling factor in the location, initiation and rate of development of ephemeral gullies