Soils potentially vulnerable to phosphorus losses: speciation of inorganic and organic phosphorus and estimation of leaching losses

Eutrophication is an important threat to aquatic ecosystems world-wide, and reliable identification of areas vulnerable to phosphorus (P) losses from diffuse sources is essential for high efficiency of mitigation measures. In this three-step study we investigated (i) relationships between the agronomic (Olsen-P and P-AL) and environmental soil P tests (P-CaCl2) with molecular techniques (P-31 NMR and XANES) followed by (ii) rainfall simulation experiment on topsoil lysimeters and (iii) comparison to long-term field measurements of water quality. Soil samples were collected from seven sites indicated to be vulnerable to nutrient losses due to underlying geology. High P release correlated to standard agronomic P tests (Olsen P, r = 0.67; and P-AL, r = 0.74) and low P sorption capacity (r = - 0.5). High content of iron-bound P compounds indicated more labile P and higher release of dissolved P (r = 0.67). The leaching experiment showed that three out of four soils with high initial soil P status had both higher P leaching concentrations before fertilization (0.83-7.7 mg P l(-1)) compared to soil with low initial soil P status (0.007-0.23 mg P l(-1)), and higher increase in P concentrations after fertilization. Higher soil P sorption capacity reduced P leaching losses. Finally, long-term monitoring data show no significant trends in P losses in a field with low initial P content and moderate P fertilization rates whereas high and over time increasing P losses were recorded in a field with high initial soil P content and repetitively high P fertilization rates.