Modelling the biogeochemical cycle of silicon in soils: Application to a temperate forest ecosystem

We investigated the biogeochemical cycling of silicon (Si) in an acidic brown soil covered by a coniferous forest (Douglas fir). Based on published and original data, we constructed a conceptual model and used a modified version of the reactive transport code MIN3P for model testing and quantification purposes. The model was first calibrated and further validated with respect to biomass data and Si-concentrations in capillary solutions, which were collected monthly over several years by means of suction-cup lysimeters placed at different soil depths.Following sensitivity tests, the model was calibrated quite accurately (limited to a 10% concentration error) by the adjustment of kinetic constants, longitudinal dispersion, and apparent activation energy for K-feldspar dissolution. Calibrated parameter values were constrained by ranges reported in the literature, when available. Mass balance calculations indicate that an average of 60% of the biogeochemical cycle of Si was controlled by biological processes (i.e. Si-uptake and dissolution of phytoliths). Sensitivity analyses suggest that no more than 55% of the Si-cycle is controlled by weathering of primary silicates. Such a large contribution of biological turnover to Si-cycling may be explained by the combined effects of a relatively large Si-content in the litter fall (i.e. specifically in the needles) and high biomass productivity of the coniferous species considered. In addition to potential implications for the global Si cycle, this investigation raises several fundamental questions concerning the nature of Si-uptake mechanisms and physiological use of Si by trees in natural systems.