Multifractal analysis of Hg pore size distributions in soils with contrasting structural stability

Parameters are needed to recognize and monitor changes in pore size distributions (PSD) caused by factors such as differences in soil management systems or by disturbance of the soil structure. The objectives of this work were to evaluate the potential of multifractal parameters obtained from mercury injection porosimetry (MIP) curves to distinguish between two soils with contrasting structure stability indices and between distinct stages of the surface of these soils. Samples were collected from the uppermost surface layer of two agricultural soils, before and after simulated rainfall. The first soil was loamy textured, with 4.61% organic matter content and a mean weight diameter (MWD) of 2.136 mm. The second soil was a silty loam with 2.17% organic matter content and a MWD of 0.262 mm, highly susceptible to crusting. Crusted soil surfaces were produced by cumulative 260 mm and 140 mm simulated rainfall on the loamy and the silty loam soil, respectively. Ten replicated samples from the initial freshly-tilled and the crusted soil surfaces were analyzed. In the diameter range of 100-0.005 μm, the freshly-tilled surface of the loamy soil had a significantly (p < 0.05) higher pore volume than its rain-disturbed counterpart, whereas the respective pore volume of the silty loam soil slightly increased following simulated rain. The scaling properties of PSDs measured by MIP could be fitted reasonably well with multifractal models. Generalized dimension spectrum, D_q, led to a better definition of multifractal scaling than singularity spectrum, f(α). Multifractal parameters such as Hölder exponent of order zero, α₀, aperture of the left part of the singularity spectrum (α₀ − α_q+), entropy dimension, D₁, correlation dimension, D₂, as well as indexes (D₀-D₁) and (D₀-D₂) were significantly different between the structurally stable loamy soil and the silty loam soil prone to crusting and between initial and rain-disturbed surface stages (p < 0.05). Moreover, D₁ and (D₀-D₁) were also significantly affected by the interaction between soil type and surface stage. Parameter α₀ ranked as: loam initial < loam rain-disturbed < silty loam initial < silty loam rain-disturbed, whereas the opposite rank was true for entropy dimension, D₁. Consequently, low structural stability or stability decay due to disaggregation by rainfall lead to clustering of PSDs measured by Hg intrusion porosimetry. These results show that multifractal analysis of PSDs may be an appropriate tool for characterizing soil structure stability and also a suitable indicator for assessing soil surface evolution stages.