Testing mechanisms underlying the Hedley sequential phosphorus extraction of soils

The Hedley fractionation has become the preeminent measure for estimating the bioavailability of phosphorus (P) in soils. However, mechanisms underlying P extractability have never been tested. We hypothesize that P sequentially extracted by individual steps can either be referred to a specific mineral source (Hypothesis 1) or to its binding strength to minerals (Hypothesis 2). We prepared mineral‐P associations in the laboratory using various secondary mineral phases and P forms (orthophosphate, phytic acid, ribonucleic acid), which were then subject to the Hedley sequential extraction scheme (anion exchange resin in form, 0.5 M NaHCO₃, 0.1 M NaOH, 1 M HCl, and concentrated HCl at 80°C). Extracts were analyzed for P as well as for the main mineral‐borne elements by inductively coupled plasma–optical emission spectroscopy (ICP–OES). In order to test if the observed mineral dissolution patterns match those of natural soils, we applied the Hedley fractionation to forest soils comprising various P stocks and measured in addition to extracted P also iron, aluminum, and calcium by ICP – OES. Phosphorus extractability from mineral‐P associations differed between P forms and mineral phases. Adsorbed P always contributed to several or all extracts, Hypothesis 1 was thus not tenable. Aluminum hydroxide, allophane, ferrihydrite, and goethite completely dissolved during Hedley fractionation from the third extraction step onwards. Successive mineral dissolution also occurred for the soil samples. Thus, extracted P represents partly desorbed P from various soil constituents and partly P co‐released upon dissolution of various minerals. Consequently, also Hypothesis 2 could not be confirmed, i.e., the sequential extraction is not suitable to assess different binding strengths between P forms and minerals. We conclude that the method hardly provides information for studies aiming at the mechanistic understanding of P bioavailability in soil.