Aggregate breakdown in clay soils under simulated rain and effects on infiltration

Wet sieving was used to separate aggregate-size distributions of four clay soils after pre-treatments of simulated rainfall, tension wetting and immersion wetting. Infiltration rates of columns of the soils were also measured under simulated rainfall. During rainfall, samples for wet sieving and the infiltrating columns were either bare or covered with a cloth mesh designed to absorb raindrop impact without affecting rates of wetting. Two swelling clay soils, one non-swelling clay soil and one clay soil dominated by sesquioxides, were used. For the first three soils, rate of wetting was the major factor governing aggregate breakdown. Significant changes in Mean Weight Diameter (MWD) occurred during the first few minutes of rainfall whether the samples were covered or not. As the rain continued, further breakdown was detected only in the uncovered samples. MWD of the sesquioxide soil decreased slightly during immersion, but most aggregate breakdown resulted from the impact of raindrops. Infiltration into the soil columns was virtually unrestricted if the soils were covered. Slaking without drop impact did not interfere with water entry. On bare soils, positive correlations were found between cumulative rainfall and the per cent of soil particles <0.12 mm diameter. An inverse relationship was found between this particle size range and infiltration rates. It is suggested that 25% of the <0.12 mm fraction must be present before infiltration rates decline.