Macronutrient processing by temperate lakes: A dynamic model for long-term, large-scale application.

We developed a model of the biogeochemical and sedimentation behaviour of carbon (C), nitrogen (N) and phosphorus (P) in lakes, designed to be used in long-term (decades to centuries) and large-scale (104-105km2) macronutrient modelling, with a focus on human-induced changes. The model represents settling of inflow suspended particulate matter, production and settling of phytoplankton, decomposition of organic matter in surface sediment, denitrification, and DOM flocculation and decomposition. The model uses 19 parameters, 13 of which are fixed a priori. The remaining 6 were obtained by fitting data from 109 temperate lakes, together with other information from the literature, which between them characterised the stoichiometric incorporation of N and P into phytoplankton via photosynthesis, whole-lake retention of N and P, N removal by denitrification, and the sediment burial of C, N and P. To run the model over the long periods of time necessary to simulate sediment accumulation and properties, simple assumptions were made about increases in inflow concentrations and loads of dissolved N and P and of catchment-derived particulate matter (CPM) during the 20th century. Agreement between observations and calculations is only approximate, but the model is able to capture wide trends in the lakewater and sediment variables, while also making reasonable predictions of net primary production. Modelled results suggest that allochthonous sources of carbon (CPM and dissolved organic matter) contribute more to sediment carbon than the production and settling of algal biomass, but the relative contribution due to algal biomass has increased over time. Simulations for 8 UK lakes with sediment records suggest that during the 20th century average carbon fixation increased 6-fold and carbon burial in sediments by 70%, while the delivery of suspended sediment from the catchments increased by 40% and sediment burial rates of N and P by 131% and 185% respectively.