Phosphorus transfer from soil to surface waters

The release of phosphate ions (PO4-P) from soils to surface waters and overland flow was examined in this study. Sand and undisturbed soil cores were ponded in beakers and the surface waters were subjected to a gentle swirling action by a paddle agitator. Slabs of soil were also placed in a flume, saturated and subjected to overland flow. A comparison of the mass of PO4-P released to surface waters in the agitator test with that released to surface runoff in the flume tests and in the literature on field losses at the plot and sub-catchment scales was made. Leighton Buzzard silica sand was washed free of all phosphate ions and a constant weight of this sand was placed in 11 beakers; each sand preparation was saturated with a constant volume of water of differing concentration of P as KH2PO4. The sands were then carefully ponded with deionised water, that was gently swirled by the agitator rotating at 20 rpm. Samples of the ponded water were taken at increasing times, up to 12 h. Layers of the sand were then carefully recovered in 5 mm depth increments and the porewater extracted using vacuum filtration and analysed for PO4-P ions to determine the depth of release of P from the porewater. A Fickian diffusion model was used to model the transfer of PO4-P to the surface water. Cores of undisturbed soil in grassland for the agitator test, soil samples from the soil surrounding the cores for analysis for P extracted by Morgan's reagent and slabs of soil from adjoining ground for placement in a flume for surface runoff tests were taken from sites of known low, medium and high Morgan's P-values. Additional cores were taken at a high Morgan's P site to evaluate the release of P from soil at several depths in the agitator. Sub-layers were employed to determine the depth distribution of Morgan's P and if a relationship existed between Morgan's P at the several depths and the concentration of P released to the surface water. The agitator sand tests showed that the depth to which P was released from the sand porewater (the depth of interaction) was about 20 mm below the sand surface in agreement with previous studies. After an initial short period of convective transfer, release of P from the sand porewater to the surface water proceeded in accordance with Fickian diffusion law. Model output and measured data were in good agreement. Morgan's P on 0-20 mm deep grassland soil samples gave good correlation with mass of P released to the surface water but values from 0 to 100 mm deep samples did not. Morgan's P in the grassland soil samples showed an exponential decline with depth. Agitator test P from many soil cores behaved in accordance with Fickian diffusion law but others exhibited fluctuating concentrations of P with time, suggestive of sorption-desorption during the tests. The masses of P released from grassland soils in the agitator tests were comparable with those released in the flume tests and with those measured in surface runoff from small plots and small sub-catchments as reported in the literature, suggesting that, with further development, the agitator test might be a suitable environmental test for P release to surface runoff. The results are discussed in relation to environmental tests for soil P and the potential for control of P loss to surface waters.