Four soil phosphorus (P) tests evaluated by plant P uptake and P balancing in the Ultuna long-term field experiment

Soil phosphorus (P) availability was assessed with four different soil P tests on seven soils of the Ultuna long-term field experiment (Sweden). These four soil P tests were (1) P-H2O (water extractable P); (2) P-H2OC10 (water extractable P upon 10 consecutive extractions); (3) P-AL (ammonium lactate extractable P) and (4) P-CDGT (P desorbable using diffusive gradients in thin films). The suitability of these soil P tests to predict P availability was assessed by correlation with plant P uptake (mean of preceding 11 years) and soil P balancing (input vs. output on plot level for a period of 54 years). The ability to predict these parameters was in the order P-H2OC10 > P-CDGT > P-H2O > P-AL. Thus, methods considering the P-resupply from the soil solid phase to soil solution performed clearly better than equilibrium-based extractions. Our findings suggest that the P-AL test, commonly used for P-fertilizer recommendations in Sweden, could not predict plant P uptake and the soil P balance in a satisfying way in the analysed soils.