Preferential phosphorus leaching from an irrigated grassland soil

Intact lysimeters (50 cm diameter, 70 cm deep) of silt loam soil under permanent grassland were used to investigate preferential transport of phosphorus (P) by leaching immediately after application of dairy effluent. Four treatments that received mineral P fertilizer alone (superphosphate at 45 kg P ha^?1 year^?1) or in combination with effluent (at ～ 40–80 kg P ha^?1 year^?1) over 2 years were monitored. Losses of total P from the combined P fertilizer and effluent treatments were 1.6–2.3 kg ha^?1 (60% of overall loss) during eight drainage events following effluent application. The rest of the P lost (40% of overall loss) occurred during 43 drainage events following a significant rainfall or irrigation compared with 0.30 kg ha^?1 from mineral P fertilizer alone. Reactive forms of P (mainly dissolved reactive P: 38–76%) were the dominant fractions in effluent compared with unreactive P forms (mainly particulate unreactive P: 15–56%). In contrast, in leachate following effluent application, particulate unreactive P was the major fraction (71–79%) compared with dissolved reactive P (1–7%). The results were corroborated by ³¹P nuclear magnetic resonance analysis, which showed that inorganic orthophosphate was the predominant P fraction present in the effluent (86%), while orthophosphate monoesters and diesters together comprised up to 88% of P in leachate. This shows that unreactive P forms were selectively transported through soil because of their greater mobility as monoesters (labile monoester P and inositol hexakisphosphate) and diesters. The short‐term strategies for reducing loss of P after application of dairy effluent application should involve increasing the residence time of applied effluent in the soil profile. This can be achieved by applying effluent frequently in small amounts.