Background The red soil hilly regions in the South China are highly susceptible to severe soil erosion due to the bare ground and increased surface runoff from heavy rainfall. It has been proved that straw-returning methods may effectively reduce soil erosion by improving water infiltration, enhancing subsurface runoff, and increasing deep infiltration. However, the effects of different straw-returning methods on rainfall-runoff characteristics are not clear. This study aims to clarify these impact mechanisms and provide guidance for the optimization of soil and water conservation measures.

Methods The experiment was conducted in De'an county, Jiangxi province. With bare land (CK) as the control, four levels of straw-returning amount of 0.30, 0.45, 0.6 and 0.75 kg/m² were set for each of the three types of straw-returning methods, namely, crushed and mulch (CV), crushed and tillage (RT), and mulch-crushed and tillage (CR), with a total of thirteen treatments. Through the artificial rainfall test, the runoff volume of different stratification was collected by setting up runoff outlets in the soil tank devices to study the characteristics of subsurface runoff and deep infiltration.

Results 1) Different amounts of straw application significantly increased subsurface runoff by 2.33 to 9.31 times compared to CK. Specifically, the CV demonstrated a rapid increase in subsurface runoff with a greater amount of straw-returning. In contrast, the RT showed no significant correlation. Furthermore, regarding runoff processes, the RT approach showed a distinct single-peak curve with a steep post-peak descent, while both the CR and CV treatments stabilized upon attaining the peak and only commenced their descent after the termination of rainfall. The RT significantly reduced the initial runoff time for subsurface runoff (the reduction was 32.31% of CK), while CR effectively prolonged the formation of runoff peaks and significantly increased the peak runoff intensity. 2) Different amounts of straw application significantly increased deep infiltration by 2.33 to 9.31 times. The CR had a strong positive correlation (R²= 0.9563) between the amount of straw-returning and deep infiltration. The CR significantly reduced the initial runoff time of deep infiltration (the reduction was 44.71% of CK) and increased the peak runoff intensity, while the CV effectively prolonged the peak runoff time. 3) Straw-returning resulted in a significant decrease in surface runoff, dropping from 82% to merely 4% or 1% under the CV and CR, respectively. In contrast, the RT caused a reduction to 49%.

Conclusions Straw-mulching primarily affects runoff by increasing surface roughness and decreasing raindrop impact, whereas straw incorporation through tillage works by improving soil porosity and lowering bulk density. The CR significantly changes runoff fractions and output processes at a lower amount of straw-returning, providing the best soil and water conservation effect. The research results provide valuable insights into the scientific utilization of straw resources in red soil hilly regions, facilitating the development of more effective soil management practices.