No abstract

Little information is available concerning the performance of grass strips for erosion control from steep cropland. An experiment was conducted on 5-m-long grass strips with slopes of 3°~15° that were subjected to silt laden runoff and simulated rainfall, to investigate the sediment trapping processes. The grass strips had three treatments including intact grass control (C), no litter (dead grass material covering the soil surface was removed) (NL), and no litter or leaves (only 2~3 cm grass stems and roots were reserved) (NLL).Generally the grass strips had a high effectiveness in trapping sediment from steep cropland runoff. Sediment trapping effi ciency (STE) decreased with increasing slope gradient, and even for a 15° slope, STE was still more than 40%. Most sediment deposited in the backwater region before each grass strips. The removal of grass litter or/and leaves had no signifi cant infl uence on STE. The sediment median size (D 50 ) in infl ow was greater than that in outfl ow, and the difference (ΔD 50 ) decreased with increasing slope. A positive power relationship between STE and ΔD 50 can be obtained. Grass strips were more effective in trapping sediments coarser than 10 or 25 μm, but sediments fi ner than 1 μm were more readily removed from runoff than particles in the range of 2 to approximately 10 μm.Grass litter had less infl uence on fl ow velocity than leaves because the deposited sediment partially covered the litter layer. Mean fl ow velocity and its standard deviation were negatively correlated with STE, and they can help make good estimation of STE. Results from this study should be useful in planting and managing forage grass to effectively conserve soil loss by runoff from steep slopes on the Loess Plateau of China.