Soil derived phosphorus in surface runoff from grazed grassland lysimeters

Seven 1 ha grazed lysimeter plots, managed as intensive grassland for the last 12 years, were monitored for total phosphorus (TP) and molybdate reactive phosphorus (MRP) in surface runoff plus interflow to 30 cm depth, for up to ten events during 1994. The mean MRP and TP concentrations determined were 40 and 122 mu g l(-1), respectively, but the data were heavily skewed by low frequency high intensity events. Thus concentrations of MRP and TP of over 1200 and 1700 pg l(-1) respectively were determined in extreme events, causing a TP export of up to 18 g ha(-1) per h and, during one event of 30 h duration, over 0.5 kg TP ha(-1) was estimated to have been removed, representing a significant proportion of the triple super phosphate fertiliser added 6 days earlier. One storm was monitored at 3 h intervals and the patterns for TP concentration and TP load were closely related to discharge. MRP concentration did not follow the hydrographic pattern. Excluding the low frequency high intensity events, an empirical model TP (mu g l(-1))=58+[42 x discharge] (1 s(-1)) was postulated, which although significant (P = 0.0053), only accounted for 14% of the relationship. It was thus concluded that understanding of release mechanisms in the majority of high frequency, low intensity events is inadequate. Conversely, during one period of high phosphorus (P) export, the mechanisms responsible were suggested to be a combination of the presence of cattle (excretal returns and poaching), timing of inorganic P fertiliser additions, and intensity of rainfall. Grassland soils are a significant source of diffuse P inputs to surface and estuarine waters and may cause eutrophication. Copyright (C) 1996 Elsevier Science Ltd