Rill Cross-sectional Geometry and Soil Erosion as Influenced by Soil and Tillage Part I. Empirical Study

Abstract

Simulated rainfall was applied to several tillage systems on two soil types at slopes ranging from 1.6 to 10%. These experiments were conducted 0 to 10 days after planting corn in a corn-soybean rotation, and for fallow treatments, the experiments were conducted during the second growing season in which the soil was maintained in a fallow and relatively weed-free condition. Soil surface geometry was recorded using a pin-type rillmeter. Mean rill width, wetted perimeter, and hydraulic radius for each tillage system were characterized as functions of rill cross-sectional area using nonlinear regression. In the Catlin soil, rills appeared to widen as erosion progressed. In the Tama soil, rills appeared to become deeper with reduced width per unit of cross-sectional flow area as erosion progressed. Estimated mean rill widths were related to measured runoff rates and compared to rill width equations reported by Gilley et al. (1990) and Elliot (1988). The equation of Elliot appeared to be a better predictor of mean rill widths. For tillage systems with crop residue cover ranging from 15 to 54% in the Tama soil, both equations tended to underestimate rill width by 20 to 36%. In multiple linear regression equations, estimated mean rill width and hydraulic radius accounted for 2 to 20% of the variation in measured sediment discharge. These regression equations suggested that a 25% increase in rill width would lead to a 25 to 50% reduction in sediment discharge.