The contributions of the roots, stems, and leaves of three grass species to water erosion reduction on spoil heaps

[Display omitted] •Soil erosion processes and flow hydraculis on spoil heaps in response to different grass types and components were revealed.•No significant differences of erosion reduction were found between grass with taproot and with fibrous root systems.•The contribution rate to erosion reduction of different grass components depended on rainfall intensity and grass type.•Grass reduced soil erosion on spoil heaps by reducing soil erodibility and increasing critical shear stress. Soil erosion on spoil heaps is one of the global environmental issues, where grass recovery is often implemented. However, information regarding how the erosion processes and control efficiency on spoil heaps differ as functions of grass types and components is lacking. In this study, a series of artificial rainfalls (48, 72, and 108 mm h−1) were conducted on eight plots (each one was replicated) that were established on steep spoil heaps and treated with three different grass types (Artemisia gmelinii (AG), Medicago sativa (MS), and Cynodon dactylon (CD)), with measurements of the runoff generation, sediment yield, and runoff velocity. Bare plot (BS) was set up as a control. Three scenarios regarding erosion efficiencies that by intact grass (IG), a combination of grass stems and roots (NL), and only grass root (OR), were considered. The results showed: 1) grass recovery can significantly reduce soil loss rate (SLR) on spoil heaps up to 99%, and there was no significant difference in erosion reduction efficiency between grasses with a taproot (AG: 93%) and fibrous root (MS: 92%; CD: 89%) systems under the IG treatments (P > 0.05); 2) The effect of the different grass treatments on SLR reduction depended on the rainfall intensity. The contribution rate of roots to SLR reduction was 68%–90% at a rainfall intensity of 48 mm h−1, while the stem and leaf contribution rates were 39%–156% at intensities of 72 and 108 mm h−1, respectively. While incomplete grass even increased the SLR, especially in the OR treatment, where the contribution rate of root to SLR reduction was −2%– −68%. 3) The IG treatment reduced soil erodibility (Kr) by 81.43%–83.80%, while the Kr values of the OR treatment were 1.51–3.53 times greater, as compared to the BS plots. In most cases, the critical shear stress (τc) on the grass plots was 1.09–2.14 times greater than those in the BS plots. These results would guide the selection of grass species for controlling spoil heaps erosion, also highlight the neces