The Recovery of Soils after Compaction: A laboratory investigation into the effect of wet/dry cycles on bulk density and soil hydraulic functions

Van den Akker et al. (2003) stated that “European subsoils are more threatened than ever in history” in an editorial referring to the results of the European Union concerted action “Experiences with the impact of subsoil compaction on soil crop growth and environment and ways to prevent subsoil compaction”. Compaction of agricultural soils not only reduces yields but can lead to pollution of surface waters and the release of greenhouse gases. Understanding, therefore, how, and to what extent, subsoil compaction may be reversed seems of vital importance.

The literature suggests that soils with a range of textures could recover structural properties such as total and macro-porosity, infiltration rate and stability, although the extent and depth of these changes varies greatly. Factors affecting the extent and depth of changes are the rate of wetting and drying and the matric potentials achieved during drying (influenced by crop cover and irrigation practices as well as climate), organic matter content and tillage history.

There is, however, a limited amount of information regarding changes to the pore-size distribution and unsaturated hydraulic conductivity of soils with wet/dry cycles. Although there are models able to predict the compaction that would occur for a given set of conditions, as well as the compactibility of soils and surface crack generation, there are no predictive models that describe structural changes due to wet/dry cycles.

The aims of this study were to assess the changes in bulk density and hydrological parameters of a range of soils, of varying texture and other physical and chemical properties, with wet/dry cycles and to explore the relationships between any measured changes and the measured soil properties. A series of laboratory experiments were designed to measure changes to the total porosity and pore size distribution and hydraulic functions of soil samples. Changes to bulk density and moisture content of undisturbed soil cores at field capacity with wet/dry cycles were monitored. Changes to the moisture release curve and hydraulic conductivity curve (as well as bulk density) of remoulded clayey subsoil samples, with wet/dry cycles, were estimated using the multi-step outflow method. The surface cracking of the samples during the wetting and drying process was also explored. Changes to these parameters were statistically compared between soils and between initial bulk densities, and related to soil properties such as texture, organic matter content and Atterberg limits with a general linear model for repeated measures.

Changes to the bulk density of the soils after three wet/dry cycles were found to be significantly different (at the 95 % probability level) between soils and between initial bulk densities. It was found that 90 % of the changes in bulk density could be predicted from the initial bulk density and the liquid limit of soils alone.

The porosity gained or lost during wet/dry cycles was not confined to the macro-pore region, but included changes to pores that were undrained at pressures up to and including 100 m of water. Changes in the pore size distribution, as described by the parameters of the water release curve, modelled with the van Genuchten-Mualem equation, were found to be statistically related to soil texture and Atterberg limits.

Changes to the hydraulic conductivity function of the soils were found to be affected by changes to the connectivity of the pore network, not just to changes in the volume of conducting pores. This was particularly true of the saturated hydraulic conductivity which was found to increase the most in soil samples that experienced a concomitant increase in bulk density and decrease in macro-porosity. Changes to the area of surface cracks were found to be significantly related to changes in the saturated hydraulic conductivity and to changes in the parameters of the van Genuchten-Mualem equation. It was also found that cracks did not close up entirely when the samples were saturated and there were many instances where wetting increased the area of surface cracks, relative to the preceding dry event, implying that the wetting phase was as important in structural formation as the drying phase.

There is a strong suggestion that it would be possible to construct a relatively simple model, based on easily measurable parameters, which could be used to predict the changes to soil structure achievable with wet/dry cycles. Further research is needed, therefore, that extends the data set upon which the predictive model designed in this study was constructed. Refinement of the model to include information on the rates of wetting and drying would also contribute greatly to its applicability to the field situation. However, important steps have been made by this study towards a more comprehensive understanding of soil structural dynamics.