Direction-dependent behaviour of hydraulic and mechanical properties in structured soils under conventional and conservation tillage

The development of soil structure units with defined forms and dimensions (e.g. platy by soil compaction or prismatic up to subangular-blocky by swelling–shrinkage processes) can lead to direction-dependent behaviour of mechanical and hydraulic properties. However, little research has investigated direction-dependent behaviour directly. Undisturbed samples were collected at different horizons and orientations (vertical and horizontal) of Stagnic Luvisols derived from glacial till (Weichselian moraine region in Northern Germany). A direct shear test determined the cohesion (c) and the angle of internal friction (φ). The water retention curve (WRC), the saturated hydraulic conductivity (ks) and the air permeability (ka) were also measured. The air-filled porosity (ɛa) was determined and pore continuity indices (N) and blocked porosities (ɛb) were derived from the relationship between ka and ɛa.Although the pore volume as a scalar is isotrop, the saturated hydraulic conductivity and air permeability can be anisotropic. In the seedbed (SB) and plough pan (PP) of conventionally managed soils the effective porosity is non-direction-dependent, however, differences in ks as a function of sampling direction can reach one order of magnitude in PP (ksh > ksv). The shear strength parameters do not present a significant anisotropy, although, a pronounced spatial orientation of soil aggregates (e.g. induced by soil compaction in a plough pan) lead to direction-dependent shear strength (by σn: 10 kPa, σtv: 12 kPa and σth: 19 kPa). This behaviour was especially observed in pore continuity indices (e.g. vertical and horizontal oriented aggregates observed in Bvg and PP presented ɛbv < ɛbh and ɛbv > ɛbh, respectively) showing that the identification of soil structure can be used as the first parameter to estimate if hydraulic properties present a direction-dependent behaviour at the scale of the soil horizon, which is relevant in modelling transport processes.