Modelling Nutrient Fluxes in Floodplain Lakes

Abstract

Only a few years ago Mitsch and Fennessy (1991) stated that modelling nutrient exchanges in wetlands was “now” developing. However, their statement applies to temperate wetlands rather than tropical floodplain lakes, for which still very few modelling studies exist. Their hydrological and hydrochemical complexity, often ill-defined boundaries, as well as complex and little understood interface processes (e.g. sediment-water exchange of chemicals, free water solutes and flooded terrestrial and aquatic/semiaquatic vegetation) distinguish them from lake ecosystems that have been subject to modelling efforts for several decades. Due to these distinctions, the question remains open whether mathematical approaches derived from “established” lake models are appropriate for wetlands as well (Mitsch and Fennessy 1991). For these “established” models, Reckhow (1994) concluded that “limited observational data and limited scientific knowledge are often incompatible with the highly detailed model structures” of water quality simulation models, and thus shed further doubt on their potential usefulness for floodplain ecosystems. Provided that data and knowledge limitations are overcome, multidimensional hydrodynamic models (Mertes 1994), and approaches including Geographic Information Systems (Costanza and Maxwell 1991) might become appropriate for future modelling of nutrient fluxes in floodplain lakes. Given the present state of knowledge, we consider highly aggregated point models, rather than complex spatial simulation models, as useful tools to identify causalities and improve our understanding of nutrient fluxes in floodplain lakes. However, the validity of describing floodplain lakes with spatially zero-dimensional models needs to be established for every case study, and may limit their general application.KeywordsAquatic MacrophyteNutrient FluxFloodplain LakeNutrient StockLake Water QualityThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.