Hydrological conditions influence the prediction of soil phosphorus indices on phosphorus leaching in hydromorphic...

Subsurface leaching is a major pathway of phosphorus (P) loss from agricultural landscapes and soil P indices (SPIs) have been developed to assess P leaching risk. However, most SPIs are developed for well-drained agricultural soils, their ability to predict P leaching risk in hydromorphic soils remains underexplored. Using soil water P concentration data from a 3-year field monitoring (weekly or biweekly) and the SPIs from the same hydromorphic soils in an agricultural catchment in Western France, we tested the ability (Pearson r²) of SPIs to predict soil water P concentrations and assessed the influence of hydrological conditions on these relationships. Eight SPIs were tested including Dyer P, Olsen P, total P, water-extractable P (WEP), equilibrium soil P concentration (ECPo), and three degree of P saturation indices (DPS_Dyer, DPS_Olsen, and DPS_Pöthig). Results showed high temporal variations in the positive correlations between the SPIs and molybdate-reactive P (MRP) and total dissolved P (TDP), with mean r² ranging 0.36-0.64 for MRP and 0.34-0.58 for TDP. The prediction effectiveness of most SPIs showed no significant difference among hydrological years, while those of Olsen P, ECPo, DPS_Olsen, DPS_Pöthig, and TP were significantly lower in winter waterlogging seasons compared to rewetting/drying seasons. This suggests a greater influence of intra-annual hydrological conditions on SPIs' prediction effectiveness, particularly for MRP. This study highlights the need to consider the resilience of SPIs to hydrological variations and proposes that aggressive indicators like Dyer P and DPS_Dyer should be prioritized when developing P leaching risk indicators for hydromorphic agricultural soils.