Background/Question/Methods
We constructed a simple spatially explicit geomorphology/biogeochemistry model that includes erosion, soil formation, dust deposition, and phosphorus (P) loss rates, in order to ask two questions of fundamental importance in terrestrial biogeochemistry: 1) how does soil residence time vary across the Earth's surface? and 2) how does this variation influence the availability of phosphorus in terrestrial ecosystems? These questions are of particular importance in the tropics, where the course of anthropogenic changes over the coming century will likely be mediated by soil nutrient status. One widely cited hypothesis is that many tropical soils are old, and therefore depleted in P. While there is little doubt that soils become progressively more P depleted with time, the assumption that old soils are prevalent in the tropics has never been rigorously tested in a spatially explicit framework. Furthermore, the concept of soil age is difficult to apply in dynamic landscapes in which the simplest formulation soil age is the average residence time of a soil particle as it moves from the bedrock-soil interface to the surface and is eroded away.. At steady state, this residence time is the soil thickness divided by the erosion rate. This formulation is complicated by the inputs of dust, which is added only to the upper soil, and thus may have a different residence time than bedrock-derived material.
Results/Conclusions
Our model quantifies these potentially disparate residence times and their importance for soil P availability for 0.5 x 0.5° pixels across the Earth's surface. As such it is the first process-based global prediction of soil "age" and phosphorus status. We find little support for the assertion that tropical soils have a longer soil residence time than unglaciated temperate soils and project that P status will vary more between stable and eroding geomorphic surfaces than between the tropics and extra tropics. Our model highlights several areas where additional data will be needed to arrive at a more refined prediction of soil P status. These include: 1) the rate of P loss over time in different climates, 2) the role of secondary mineral formation in altering P availability, and 3) the degree to which dust is mixed into soil or eroded away . Nevertheless, we believe our simple model has utility as a starting point in the quest to predict soil nutrient status based on first principles in a spatially explicit framework.