GEOLOGIC CONTROL OF NUTRIENT AVAILABILITY IN TERRESTRIAL ECOSYSTEMS

We developed a model to explore the effects of soil production and erosion on phosphorus availability in terrestrial ecosystems. Over millennial timescales, nutrient availability is determined by the mass balance of ecosystem inputs and outputs, and for those elements derived from rock weathering the balance generally tips towards outputs over the course of soil development. In particular, the loss of soil phosphorus (P) over time result in P limitation of plant primary productivity on the oldest, weathered soils along a few, intensively studied chronosequences. These results have been extrapolated to suggest widespread P limitation in the lowland tropics, where climatic conditions promote the rapid depletion of soil nutrients and the absence of recent glaciations suggests that on average soils are relatively old.

We developed a quantitative geomorphologic framework with which to explore this assumption, and the effects of erosion rate and weathering zone thickness on soil P availability. Using data from three chronosequences in Hawai'i and New Zealand to derive field-based estimates of available P inputs and losses, and published uplift rates and weathering zone thicknesses, we identified tectonic settings that are likely to support P-limited ecosystems, at least in areas where geologic uplift and erosion are balanced. Our model predicts the occurrence of transient P limitation in young ecosystems where mineral weathering to form biologically available P is outpaced by rapid physical erosion. In contrast, P limitation appears unlikely in areas of moderate erosion rates, such as much of Central America and Southeast Asia, because erosion drives the advection of P bearing fresh rock into the rooting zone fast enough to compensate for losses. Finally, we expect soils on stable cratons with negligible uplift and erosion, such as the Amazon Basin and Western Africa, to exhibit widespread P depletion in the absence of exogenous nutrient inputs. Because the model results do not include estimates of exogenous P fluxes, they tend to overestimate the extent of P depletion in terrestrial ecosystems. Given these results, we argue for a more nuanced exploration of the spatial distribution of P depletion in the lowland tropics.