Model studies on the dynamics of hydrophobic organic compounds in shallow lake ecosystems

The twentieth century witnesses the widespread eutrophication and intensive organic contaminations in earth surface water systems located in highly populated areas, resulting in severe deterioration of water quality in freshwater ecosystems around the globe. This is particularly the case for many freshwater shallow lakes in China. The interaction between excess nutrient loading and enormous organic contaminants discharge has raised increasing attention from both scientists and lake managers, whereas accurate prediction for both substances cannot be properly predicted based on knowledge from either field alone. However, efforts in the related scientific research, particularly the development of relevant modeling tools, remains scarce. To this end, the aim of this thesis is to develop an integrated ecological and chemical modeling tool, which is composed of contaminant fate module (CF), food web accumulation module (FW) and ecological module (EM), in the hope to fulfill the research gap above. We collected three groups of HOCs, namely hexachlorocyclohexanes (HCHs), polycyclic aromatic hydrocarbons (PAHs) and Per- and polyfluoroalkyl substances (PFASs), in multiple compartments from two Chinese shallow lakes that are currently in distinct ecological states, i.e., Lake Small Baiyangdian (in clear state) and Lake Chaohu (in turbid state). In particular, paleo-records of PAHs residual levels in Lake Chaohu in two sediment cores covering the time span of over 60 years were obtained. We elaborated to explicitly investigate the fate, transport and transformation of these contaminants in these two shallow lakes using the developed modeling tool, with either steady state or dynamic simulations (in time scales of both short-term intra-annual (1-2 years) and long-term inter-annual (60 years)). The following issues were addressed: 1) fate of the chemicals in lake environment and the dominant processes; 2) seasonal patterns of chemicals in lakes and the driving factors; 3) long-term dynamics of chemicals in lakes and the driving factors; and 4) impact of abrupt changes in ecosystems on the distribution of contaminations in shallow lakes. For modeling techniques, we implemented uncertainty analysis on the model using both classic Monde Carlo and more advanced Bayesian Markov Chain Monte Carlo (MCMC) algorithm. We recommend to apply MCMC to contaminant modeling approach to make calibration possible and to remove the overestimated uncertainty in predictions. Furthermore, we compared the advantages and disadvantages of our model to other models with similar objectives, and we further proposed a more comprehensive modeling framework that incorporates hydrodynamic models to address spatial variations of contamination, which embraces the fruitful outcomes in aquatic ecosystem modeling. Finally, we advocate to add modeling approach as the third dimension for the ‘contemporary & paleo-observations’ strategy, which together contribute to the ‘golden triangle’ framework. New insights and discoveries may emerge for the evaluation on the organic contaminants in shallow lake systems, which may contribute to ecological and human health risk assessment. This ‘golden triangle’ may serve as the multidiscipline framework for limnologic research in the future.