Available soil phosphorus, phosphorus buffering and soil cover determine most variation in phosphorus concentration in runoff from pastoral sites

Phosphorus (P) runoff from agricultural land contributes to elevated P concentrations in surface water. Although P concentrations in runoff are often higher when soil P is increased, attempts to quantify this relationship in field studies have been inconclusive, either because of confounding within experiments or methodological differences between experiments. The present study attempted to address these issues using simulated rainfall at 15 pastoral locations with varied lithology, land-use intensity, plant cover and soil-P properties in SE Australia, at a total of 38 on-farm sites, each with paired plots. At 27 sites, one of each pair was covered with 70 % shade-cloth to reduce rainfall energy. Although these were uncultivated pastoral sites and surface soil erosion rates were all relatively low, eroded surface soil had sufficiently high concentrations of P to make a major contribution to total P in runoff from some sites, even when plant-available soil P was low. We conclude that greater vegetative cover is required to manage P in runoff than to manage soil erosion. When soil was adequately protected against erosion by either high vegetative cover or shade-cloth, greater than 90 % of the variation in both dissolved reactive P and total P was explained by two commonly measured soil variables: soil-test P (bicarbonate extract) and P-buffering index (where P was in the range where plants may respond to fertiliser). It appears that available soil P can be raised with fertiliser to improve agronomic production without fear of significantly increasing P losses in runoff provided soil test P remains below the agronomic critical concentration defined by the soil P-buffering index. These findings have broad relevance for environmental risk assessment and management where the data required for more complex procedures are unavailable.