Soil properties controlling inorganic phosphorus availability: general results from a national forest network and a global compilation of the literature

Incorporating the phosphorus (P) cycle in climate-carbon cycle models-or calibrating pedotransfer functions to predict available soil P-are important issues. To achieve them we need to improve our understanding of the P cycle by focusing on processes and on the factors which control P dynamics in soils. The aim of the present study was to evaluate the generality of the relationships between physical-chemical soil properties and the availability of inorganic P (i.e. the dynamics of phosphate ions at the solid-to-solution interface), and to test the predictive capacity of these relationships. We used the French permanent network of forest monitoring (102 forests with contrasting soil properties, called network dataset) and a global compilation of published data from different ecosystems (60 studies, mainly in forests, grasslands, or croplands, called compilation dataset). All studies used an isotopic dilution method to quantify the availability of inorganic P. Results showed generality of the dominant role of aluminum and iron oxides and organic carbon in controlling the dynamics of phosphate ions in acidic and non-acidic soils. Inversely, soil texture, pH and CaCO3 generally had no or only little effect. The control of inorganic P availability by oxides and organic carbon was confirmed by the compilation dataset, even in non-forest soils. Relationships obtained with the network dataset correctly predicted available soil inorganic P, suggesting that the dynamics of phosphate ions in soils could be simulated by including the main controlling soil properties in models. Our study provides predictive tools which could be included in diagnostic systems for the long-term management of soil fertility.