Modeling biogeochemical processes of phosphorus for global food supply.

Harvests of crops, their trade and consumption, soil erosion, fertilization and recycling of organic waste generate fluxes of phosphorus in and out of the soil that continuously change the worldwide spatial distribution of total phosphorus in arable soils. Furthermore, due to variability in the properties of the virgin soils and the different histories of agricultural practices, on a planetary scale, the distribution of total soil phosphorus is very heterogeneous. There are two key relationships that determine how this distribution and its change over time affect crop yields. One is the relationship between total soil phosphorus and bioavailable soil phosphorus and the second is the relationship between bioavailable soil phosphorus and yields. Both of these depend on environmental variables such as soil properties and climate. We propose a model in which these relationships are described probabilistically and integrated with the dynamic feedbacks of P cycling in the human ecosystem. The model we propose is a first step towards evaluating the large-scale effects of different nutrient management scenarios. One application of particular interest is to evaluate the vulnerability of different regions to an increased scarcity in P mineral fertilizers. Another is to evaluate different regions' deficiency in total soil phosphorus compared with the level at which they could sustain their maximum potential yield without external mineral inputs of phosphorus but solely by recycling organic matter to close the nutrient cycle.