Insights into biogeochemical cycling from a soil evolution model and long-term chronosequences

Abstract. Despite the importance of soil processes for global biogeochemical cycles, our capability for predicting soil evolution over geological timescales is poorly constrained. We attempt to probe our understanding and predictive capability of this evolutionary process by developing a mechanistic soil evolution model, based on an existing model framework, and comparing the predictions with observations from soil chronosequences in Hawaii. Our soil evolution model includes the major processes of pedogenesis: mineral weathering, percolation of rainfall, leaching of solutes, surface erosion, bioturbation, the effects of vegetation in terms of organic matter input and nutrient cycling and can be applied to various bedrock compositions and climates. The specific properties the model simulates over timescales of tens to hundreds of thousand years are, soil depth, vertical profiles of elemental composition, soil solution pH and organic carbon distribution. We demonstrate with this model the significant role that vegetation plays in accelerating the rate of weathering and hence soil profile development. Comparisons with soils that have developed on Hawaiian basalts reveal a remarkably good agreement with Na, Ca and Mg profiles suggesting that the model captures well the key components of soil formation. Nevertheless, differences between modelled and observed K and P are substantial. The fact that these are important plant nutrients suggests that a process likely missing from our model is the active role of vegetation in selectively acquiring nutrients. This study therefore indirectly indicates the valuable role that vegetation can play in accelerating the weathering and thus release of these globally important nutrients into the biosphere.