Depletion, accumulation and availability of soil phosphorus in the Askov long-term field experiment

Soil phosphorus (P) reserves, built up over decades of intensive agriculture, may account for most of the crop P uptake, provided adequate supply of other plant nutrients. Whether crops grown on soils with reduced supply of other nutrients obtain similar use-efficiency of soil P reserves remains unclear. In treatments of the Askov Long-Term Experiment (initiated in 1894 on light sandy loam), we quantified changes in soil total P and in plant-available P (Olsen P, water extractable P and P offtake in wheat grains) when P-depleted soil started receiving P in rock phosphate and when P application to soil with moderate P levels ceased during 1997-2017. Additionally we studied treatments with soil kept unfertilised for >100 years and with soil first being P depleted and then exposed to surplus dressings of P, nitrogen (N) and potassium in cattle manure. For soil kept unfertilised for >100 years, average grain P offtake was 6 kg ha(-1) and Olsen P averaged 4.6 mg kg(-1), representing the lower asymptotic level of plant-available P. Adding igneous rock phosphate to severely P-depleted soil with no N fertilisation had little effect on Olsen P, water extractable P (Pw), grain yields and P offtake. For soils with moderate levels of available P, withholding P application for 20 years reduced contents of Olsen P by 56% (from 16 to 7 mg P kg(-1)) and of Pw by 63% (from 4.5 to 1.7 mg P kg(-1)). However, the level of plant-available P was still above that of unfertilised soil. Application of animal manure to P-depleted soil gradually raised soil P availability, grain yield and P offtake, but it took 20 years to restore levels of plant-available P. Our study suggests symmetry between rates of depletion and accumulation of plant-available P in soil.