Temporal variation of soil sorptivity under conventional and no-till systems determined by a simple laboratory method

Abstract

Soil water sorptivity (S) is an important property that measures the soil capacity to take water rapidly under capillary forces. Usually S is not included in soil laboratory routine experiments because there is not a widely accepted methodology for its determination. The objectives of this work were: i) to propose a modification on the Leeds-Harrison et al. (1994) method (LH) to determine S in undisturbed soil samples; and ii) to determine the temporal variation of S and saturated hydraulic conductivity (K0) in a soil under conventional tillage (CT) and no-tillage (NT) treatments. Additionally, the influence of soil pore size distribution (PoSD) on S was analyzed. Undisturbed soil samples (5 cm height, 5 cm diameter) were collected from the upper 10 depth cm of each plot, from each treatment at four different times during a maize growing season (before seeding (BS), 6 leaf stage (V6), physiological maturity (R5) and after harvest (AH)). PoSD was determined in a sand box apparatus. After that, S was determined in the same samples using a modified Leeds-Harrison approach. For the proposed modification the difference between initial and final water content was actually gravimetrically measured in each sample, rather than considering it equal to the total porosity (TP). The proposed improvement was validated comparing the obtained S values with those calculated using standard one-dimension horizontal infiltration in sieved soil (0.098 vs 0.079 cm s−1/2, respectively) and in calibrated sand (0.041 vs 0.040 cm s−1/2, respectively). These differences were not significant. Both S and K0 were significantly affected by the sampling time in both treatments (mean values ranged between 0.022 and 0.077 cm s−1/2 and 1.57 and 3.75 cm s−1 respectively). We did not find a significant dependence of S with three pore size ranges analyzed. The proposed improvement of the Leeds-Harrison method allowed determining the temporal variation of S in representative undisturbed soil samples.