Efficiency of the Volumetric Replacement Method in Estimating Rill Erosion under Different Slopes and Soil Textures in Semi-Arid Region, Zanjan Province

Rills are usually found on the sloping fields worldwide, especially in semi-arid slopes, where vegetation covers are often poor and soils are weakly aggregated. Rill erosion is recognized as an important process of water erosion on agricultural land in these regions and causes a grate amount of soil loss. Understanding rill erosion rate is important in the prediction of soil erosion and the prevention of soil loss in the lands. Rill erosion is often easy to observe but difficult to measure because of its complexity and stochastic nature. A common method used to determine rill erosion rate is measuring sediment concentration distribution of eroding rill flow under different flow rates. However, it is not only time-consuming but also had to measure. The volume Replacement Method is an easy method to estimate soil loss from rills in the sloped lands. Limited information is available concerning the ability of this method in different soil textures under slope gradients. Therefore, this study was conducted to evaluate the ability of the method to estimate rill erosion of semi-arid soils.

This study was conducted on three different soil textures i.e. loam, clay loam and sandy clay loam under four slope gradients including 5, 10, 15 and 20% using factorial arrangement based on completely randomized block design with three replications in the laboratory. A flume with 0.3 m width and 4 m length was subdivided into strips of 0.1 m width and 4 m length to imitate eroding rills. Soil samples for each soil texture were passed from 8-mm sieve and packed into the flumes at its bulk density in the field. Prior to each experimental run, the soil materials were pre-wetted to reach to water-holding capacity. Tap water was introduced into the rill from the upper end, through a water supply tank and a pump at a constant flow rate of 0.5 L.min-1. After erosion, the flume was lowered to the horizontal position for the measurements of eroded rill volumes. The rill volume was determined using soil samples passed from a 2-mm sieve. Soil loss mass eroded from soil surface was computed using rill volume and original soil bulk density packed into the flume. This value was considered as estimated value using the Volumetric Replacement Method for each soil texture under different slope gradients. The performance of the method was assessed using the measured data for each soil and slope gradient using error measures such as root mean square error (RMSE) and mean absolute error (ME).

Significant differences were found among soil textures and slope gradients as well as their interaction on rill erosion rate. The highest rill erosion rate was observed in clay loam (3.166 g.m-2.s-1), whereas sandy clay loam showed the minimum susceptibility to rill detachment (0.962 g.m-2.s-1). Higher fine particles (clay) and lower aggregation as well as weak aggregate stability are the major reasons for higher susceptibility of clay loam to rill erosion. The rill erosion was more sensitive to slope gradient than soil texture and the strongest dependency of rill erosion on slope gradient was found in clay loam (R2= 0.99). With an increase in slope gradient, rill erosion strongly increased except for loam. The Volumetric Replacement Method overestimated rill erosion in all soils and slope gradients. The highest overestimation was observed in sandy clay loam (RMSE= 2.72 g/m2.sec and ME= 7.02 g/m2.sec), whereas the lowest overestimation value was in loam (RMSE= 0.60 g/m2.sec and ME= 3.86 g/m2.sec). The performance of the Volumetric Replacement Method decreased in higher slope gradients and the highest overestimation was observed under 20% slope gradient (RMSE=1.86 g/m2.sec and ME= 3.84 g/m2.sec).

Rill erosion is strongly affected by soil texture and slope gradient. Particle size distribution, aggregates percentage and their stability can control the soil’s susceptibility to detach by concentrated water flow. The Volumetric Replacement Method showed higher uncertainty as evaluated in the semi-arid soil textures especially under steep slopes. The change of soil physical properties by water flow especially bulk density result in errors in determination of rill volume by using this method. The higher change of physical properties by concentrated flow occurs in fine soil textures and steeper slopes. Additionally, continuous sedimentation along the rills imposes other errors in estimating soil loss mass from the rills.