Soil physicists studying solute transport in the laboratory are often accused of ignoring the "real world" problems by using small soil cores and bare soil surfaces. We are now trying to overcome these "real" problems by using larger soil cores and studying solute transport under pasture. The soil is the Manawatu fine sandy loam. In the field we carefully carved a "representative" soil core of 300 mm length and diameter

The "real world" as taken back to the laboratory

We initially tried to model solute transport through the "representative" fence post, but gave up after the idea of a few sleepless nights in the lab, collecting effluent would prove fruitless. The second soil core we collected looked more promising although you could now accuse us of sampling bias by not using the first. For applying water to the pasture we developed a rainfall simulator which consists of a sprinkler-reservoir (S), a pressure head regulator (P), and 2 water reservoirs (R). The cylindrical sprinkler has a plexiglass base, through which 120 holes were drilled. Conical capillary tubes were then inserted into the holes, to which hypodermic needles were attached. The 120 capillary tubes lie on concentric circles around the axis of the sprinkler-reservoir. A constant head of water is maintained in the sprinkler-reservoir by the water reservoir, based on the Mariotte bottle principle.

The sprinkler-reservoir is mounted to a DC motor, sitting on top of a steel frame, which also holds one of the water reservoirs. The motor rotates the entire sprinkler-reservoir to ensure an equal distribution of drops over the area.

The flow rate in the rainfall simulator can be altered by using either different hypodermic needles or different water heads. For this experiment we used needles of 27 gauge and a head of 60 mm, resulting in a flow rate of about 10 mm h^-1. Suction at the bottom of the soil ensures unsaturated flow.

The rainfall simulator in operation

For studying chemical through the soil a pulse of SrC1₂ (600 kg ha^-1) was applied to the pasture and then leached with a weak solution of 0.0025 M Ca(SO₄)₂. the transport of Cl ^- was monitored by collecting effluent at the bottom of the column. A dispersivity of 75 mm was found, similar to previously determined values for soil columns in the lab with bare soil surfaces. A "mobile" water content of 90% was obtained, also consistent with the earlier results. It seems that the effect of the vegetative cover on solute transport is small if plant uptake is negligible. We are currently performing experiments on similar soil cores to assess the effect of the initial soil surface water content on solute transport.

Acknowledgments