Presentation of a model simulating the response of lakes to fertilizations to reduce radiocesium levels in fish

Lake fertilization is a potential remedy for toxic contamination in lakes with low productivity, which are known to be sensitive to, e.g. radionuclides, metals and organic toxins. This study presents a model to plan the duration and predict the outcome of fertilization in lakes. Several methods of lake fertilization have been used in field experiments: (1) Lake and wetland liming using ‘mixed’ lime with added phosphorus; (2) treatment with commercial fertilizers; and (3) using effluents (containing phosphorus) from fish farms. This model is basically a dynamic model using differential equations to handle fluxes, amounts and concentrations. It also includes several empirical relationships, because the model is meant to be used in practice and the driving variables should be few and readily available, like catchment and lake morphometric data. The model is primarily intended to be used as a sub-model predicting realistic changes in phosphorus and potassium concentrations and in lake pH caused by these remedial measures within the framework of a more extensive lake model for radiocesium. It is, however, meant to be based on the fundamental processes regulating phosphorus fluxes in lakes, and could be of interest also in contexts other than radioecology. It is driven by the amount of fertilizer added to the lake and the month of treatment. The consequences for the spread, biological uptake and concentrations of radiocesium are also discussed and simulated. The phosphorus model is, and must be, easy to handle since all the input data should be derived either from maps or from regular monitoring programs. In this paper, the model is presented and calibrations are examined. The results of the calibrations indicate that the model should be useful for managers to optimise lake fertilization, e.g. in contexts of lake radioecology where the benefits and drawbacks of different remedial strategies are evaluated (multi-attribute analysis).