Plant evapotranspiration is recognised to affect soil suction of slopes and landfill covers. Previous work has focused on evapotranspiration-induced suction by a single plant, with little attention paid to the effects of planting density. The aim of this study is to quantify any changes in tree growth and tree-induced suction during evapotranspiration and rainfall under different planting densities for non-mixed-species plantations. A tree species, Schefflera heptaphylla, which is commonly found in Asia, was planted in silty sand at spacings of 60, 120 and 180 mm, representing three different planting densities. For each case, three replicates were tested to consider tree variability. In total, the responses of suction for 297 seedlings subjected to drying and a rainfall event with a 10-year return period were measured. The test results show that reducing the tree spacing from 180 to 60 mm induced greater tree–tree competition for water, as indicated by a 364% increase in peak suction upon evapotranspiration. Such tree–tree interaction led to: (a) a 19–35% reduction in the leaf area index; (b) a 17–36% decrease in root length; and (c) an obvious decay of roots. Upon the rainfall event, the infiltration rate for vegetated soil with trees planted at a spacing of 60 mm was up to 247% higher than those for soil with a wider tree spacing, where mainly fresh roots were found. Although most suction within the root zone (i.e. top 100 mm) was lost due to increased infiltration at 60 mm spacing, suctions in deeper depths below root zone were largely preserved.

Grass is recognized as being beneficial in reducing rainfall infiltration in some kinds of surface cover systems such as landfill cover, because rainwater discharges as surface runoff due to reduced water permeability caused by evapotranspiration-induced soil suction as well as foliage interception. However, the distributions of grass-induced suction in various compacted soils during rainfall are rarely reported. Moreover, it is not straightforward to determine an optimum soil dry density for minimizing rainfall infiltration and at the same time encouraging plant growth. This is because there are conflicting requirements for vegetated cover systems, i.e., compacted soil should not be too dense as to impede root growth, while on the other hand to minimize infiltration. This study thus aims to investigate, quantify, and compare grass-induced suction distributions in silty sand compacted at different densities when subjected to artificial rainfall in the laboratory. A grass species, Cynodon dactylon, which is common in many parts of Asia, was selected for testing. Compacted soil with and without a growing grass patch was tested at three relative compactions (RCs) of 70%, 80%, and 95%, in six test boxes. Test results reveal that at an RC of 95%, suction (40 kPa) retained in vegetated soil after rainfall is 100% higher than that (20 kPa) in bare soil. Among the vegetated soil compacted at the three RCs, suction retained was the highest at an RC of 95% (40 kPa), whereas suction decreased to 0 kPa at an RC of 70% after rainfall. As the average depth of grass roots decreased by 36% due to an increase in RC from 70% to 95%, the depth of influence of suction for vegetated soil at an RC of 95% reduced to less than half of root depth, which was the shallowest among the three compacted soil specimens.

Direct measurement of a permeability function, k(ψ) (where ψ = suction), of an unsaturated soil is often tedious and time-consuming. Thus, researchers have proposed various semiempirical predictive methods to determine k(ψ) indirectly from a soil-water characteristic curve (SWCC) or a particle-size distribution. Also, stress effects on k(ψ) are generally not considered. To explore any stress effect on k(ψ) , a new 1-m-high stress-controllable soil column is developed to measure stress-dependent SWCC (SDSWCC) and k(ψ) of soils directly. The objectives of this paper are to investigate the effects of (1) two stress-state variables (matric suction and net normal stress) and (2) a drying-wetting cycle on both SDSWCC and k(ψ) of a compacted decomposed silty clay. Each compacted soil column is subjected to an evaporation-ponding cycle under various vertical net normal stresses. The variations of pore-water pressure and volumetric water content (VWC, θw ) profiles with time are measured instantaneously. The k(ψ) ...