Soil Water Infiltration Under Bare Soil and Maize Straw: A Field Comparison of a Double-Ring Infiltrometer and a Rainfall Simulator

Soil water infiltration (SWI) is central to hydrology and erosion assessment, but the measurement method and surface condition may affect the rates obtained. We compared a double-ring infiltrometer (DRI) with a rainfall simulator (InfiAsper) on a Dystric Acrisol with a sandy clay loam A horizon, under bare tilled soil and under maize straw. InfiAsper applied five rainfall treatments of equal erosivity, combining intensities of 44.6–75.0 mm h−1 with durations of 106–38 min, on four to five plots per treatment and surface. DRI tests were carried out in the same fields (22 under bare soil, 13 under ground cover) and analysed as an independent group. We (i) determined the final infiltration rate (SIR_f), (ii) compared DRI with InfiAsper, and (iii) evaluated the Kostiakov–Lewis, Horton and Philip models. Under bare soil, InfiAsper means ranged from 21.6 to 37.7 mm h−1 and the DRI mean was 86 mm h−1. A planned contrast between DRI and the mean of the five InfiAsper treatments gave a DRI/InfiAsper ratio of 1.80 (95% CI 1.00–3.23; p = 0.049); with the 20 tests of at least 15 readings, it was 2.13 (1.22–3.71). Under ground cover, InfiAsper means ranged from 15.4 to 26.3 mm h−1 and the ratio was 0.65 (0.36–1.18) with all 13 tests and 1.73 (0.93–3.22) with the five tests that ran for at least 75 min. The ratio tended to be larger on bare soil than under ground cover, depending on which ground-cover tests were used. The intensity–duration combination did not significantly affect InfiAsper SIR_f. Horton fitted best for both methods and Philip worst. DRI thus gave roughly twice the final infiltration rate of rainfall simulation on bare soil, whereas no excess was found under ground cover.