Effect of rainfall and tillage direction on the evolution of surface crusts, soil hydraulic properties and runoff generation for a sandy loam soil

The study was aimed at evaluating the effect of rainfall and tillage-induced soil surface characteristics on infiltration and runoff on a 2.8 ha catchment located in the central region of Senegal. This was done by simulating 30 min rain storms applied at a constant rate of about 70 mm h−1, on 10 runoff micro-plots of 1 m2, five being freshly harrowed perpendicularly to the slope and five along the slope (1%) of the catchment. Runoff was automatically recorded at the outlet of each plot. Hydraulic properties such as capillary sorptivity and hydraulic conductivity of the sandy loam soil close to saturation were determined by running 48 infiltration tests with a tension disc infiltrometer. That allowed the calculation of a mean characteristic pore size hydraulically active and a time to ponding. Superficial water storage capacity was estimated using data collected with an electronic relief meter. Because the soil was subject to surface crusting, crust-types as well as their spatial distribution within micro-plots and their evolution with time were identified and monitored by taking photographs at different times after tillage. The results showed that the surface crust-types as well as their tillage dependent dynamics greatly explain the decrease of hydraulic conductivity and sorptivity as the cumulative rainfall since tillage increases. The exponential decaying rates were found to be significantly greater for the soil harrowed along the slope (where the runoff crust-type covers more than 60% of the surface after 140 mm of rain) than across to the slope (where crusts are mainly of structural (60%) and erosion (40%) types). That makes ponding time smaller and runoff more important. Also it was shown that soil hydraulic properties after about 160 mm of rain were close to those of untilled plot not submitted to any rain. That indicates that the effects of tillage are short lived.