Fractal analysis of soil water hysteresis as influenced by sewage sludge application

The impact of three types of surface applied sewage sludge from the same lot (fresh, composted, and thermally dried) on the water retention properties of a loam soil (Udic Calciustept) and a loamy sand soil (Typic Haplustalf) from central Catalonia (NE Spain) was investigated using fractal analysis. First, we proposed a composite fractal model that covers both the low and high suction regimes. This model has four fitting parameters: D1 (the pore-solid fractal dimension), D2 (the surface fractal dimension), A1 (a compound parameter that includes D1, the density of water, bulk density, particle density, and the air/water displacement suction), and A2 (a compound parameter that includes D2, and the critical suction and water content separating the low and high suction regimes). This model was fitted to the main wetting and drying branches of soil water retention curves obtained from small-disturbed samples using the chilled mirror dew point method. The equation fitted the data extremely well with adjusted R2 values ≥ 0.99 (p < 0.0001). Analysis of variance (ANOVA) was performed on the resulting parameter estimates. Few significant effects were observed for the loamy sand soil. In contrast, all of the model parameters, except D1, were significantly affected by hysteresis and/or the sludge treatments for the loam soil. Values of A1 and A2 from the main drying branch were significantly higher than corresponding estimates from the wetting branch. This trend was reversed for D2, indicating that pore surfaces are smoother after wetting, as compared to initially dry surfaces. The fresh, composted and thermally dried sludge treatments all significantly increased the A1 parameter relative to the untreated loam soil, possibly by decreasing bulk density. The fresh and thermally dried sludge treatments also significantly increased the A2 parameter. All three sludge types increased D2 relative to the control when this parameter was determined from the main wetting branch of the water retention curve. In contrast, D2 estimated from the main drying branch was not influenced by any of the sludge treatments. These analyses indicate that the effects of sewage sludge on hysteresis of the soil water retention curve were still present 2 years after surface application in the case of the loam soil.