Vegetation barrier and tillage effects on runoff and sediment in an alley crop system on a Luvisol in Burkina Faso

The effects of vegetation barriers and tillage on runoff and soil loss were evaluated in an alley crop system at a research station in central Burkina Faso. On a 2% slope of a sandy loam various local species (grasses, woody species and a succulent) were planted as conservation barriers in order to examine their influence on sediment transport. After each erosive storm, runoff and sediment yield was determined. The dense effective barriers (Andropogon gayanus and dense natural vegetation) slow down flow velocity, build up backwater and promote sedimentation uphill. The through flow in the less effective barriers with woody species and succulents (Ziziphus mauritania and Agave sisalana) was slightly hampered and flow velocity was not reduced enough, resulting in a higher soil transport. Under degraded conditions soil loss diminished 50% with less effective and 70–90% with effective barriers. During the initial cropping phase (light tillage; sowing) erosion was reduced 40–60% with effective barriers and showed an increase of 45% with less effective barriers. In the full tillage (weeding) period erosion decreased by 80–90% for effective and 70% for less effective barriers, aided by the development of the barrier and the crop on the alley. Barriers of natural vegetation and A. gayanus are preferred for diminishing soil loss. nt yield could best be predicted by the erosivity index (AIm), second best by runoff amount (mm), closely followed by maximum peak intensity. All these parameters are related to the volume of overland flow needed to transport soil particles. Correlation of soil loss with small rain showers was poor and correlation with big showers was good. Sediment transport with no barrier had the highest correlation, closely followed by less effective barriers. Due to the heterogeneity in development of the effective barrier, correlations were much lower. The bulk of soil loss was only dependent on a few extreme events during the observation period.