Phosphate losses through field drains in a heavy cultivated soil

Losses of total and molybdate-reactive phosphate (MRP) were measured for 4 yr in water flowing from interconnecting mole and pipe drains under 0.24 ha plots of a heavy cracking clay soil. The molybdate-unreactive phosphate (MUP) obtained by difference appeared from the results to be mainly phosphate carried on suspended soil material. The losses of all three categories of phosphate were closely related to the cumulative drainflow, with the same relations covering three of the four years, designated the synoptic years. Losses were much greater in the other year, probably because phosphate was applied after, rather than before, plowing and to very wet soil. The ceiling to annual total phosphate loss, assuming 500 mm drainage, was 0.4 kg ha -1 in the synoptic years, with 76% of the loss occurring as MUP. It was about 1 kg ha -1 in the high-loss year, with 88% as MUP, implying that MUP comprised 96% of the extra loss that year. Halving the phosphate application lessened the loss of total phosphate but not conclusively, and the decrease in MRP loss was not significant. Restricting and thereby delaying drainage lessened losses of MUP, probably because suspended material carrying it was allowed to settle. Increasing the spacing between mole channels from 2 m to 4 m increased the losses of MUP but not MRP. This probably happened because to meet a flow pathway connected to a mole channel, rainwater had to travel further horizontally and collected more phosphate-containing soil material.