Phosphorus loss in drainflow from intensively managed grassland soils

The loss of P in subsurface drainage from lysimeters (0.5 ha), managed as e ither monoculture grass or grass-clover for the last 9 yr have been quantif ied. Both systems received two to three cattle slurry applications annually and were cut two to three times before being grazed by dairy cattle. Miner al fertilizer P was applied only to the grass-clover (about 25 kg P ha(-1) yr(-1)). After 9 yr, NaHCO3-soluble P in the topsoil (0-10 cm) averaged 38 and 47 mg P kg-l for the grass and grass-clover respectively, giving an ave rage increase of 1.0 mg Olsen-P kg(-1) yr(-1) the grass-clover. Drainage-we ighted molybdate-reactive phosphorus (MRP) and total phosphorus (TP) concen trations ranged from 0.16 to 0.38 mg P L-1 and 0.45 to 0.79 mg P L-1, respe ctively during the 2-yr study period. The MRP and TP losses in subsurface-f low from the grass-clover (1.68-2.03 and 3.47-5.03 kg P ha(-1) yr(-1), resp ectively) were significantly larger than those from the grass (1.27-1.34 an d 2.97-3.58 kg P ha(-1) yr(-1), respectively). Averaged across years and pa stures, MRP accounted for 42% of the TP loss, while a non-MRP form accounte d for 41% of the TP loss through field drains. Particulate-associated P rep resented about 17% of the TP loss. The P losses in subsurface runoff measur ed in the present study were much larger than previous estimates. The resul ts also showed that, despite the subsoil having much larger P-sorption capa city than the topsoil, significant amounts of P losses could occur through preferential hydrological pathways.