Combining species sensitivity distribution (SSD) model and thermodynamic index (exergy) for system-level ecological risk assessment of contaminates in aquatic ecosystems.

After reviewing the species- and community-level ecological risk assessments (ERAs) of chemicals in the aquatic environment, the present study attempted to propose a third stage of ERA, i.e., the ecosystem-level ERA. Based on the species sensitivity distribution model (SSD) and thermodynamic theory, the exergy and biomass indicators of communities from various trophic levels (TLs) were introduced to improve ecological connotation of SSDs. The species were classified into three TLs based on algae (TL1), invertebrates (TL2), and vertebrates (TL3), and the weight of each TL was determined based on relative biomass and β value, which indicated a holistic contribution of each species or community to the ecosystem. Then, a system-level ERA protocol was successfully established, and the community- and system-level ecological risks of 10 typical toxic micro-organic pollutants in the western area of Lake Chaohu and its inflowing rivers were evaluated. System-level ERA curves (ExSSD) were mainly affected by the community-level SSD at TL2 for most chemicals in the present study. The uncertain boundary of ExSSD was mostly related to TLs with a wider uncertain boundary, but had little relation to the weight of each TL. The results of system-level ERAs revealed that dibutyl phthalate had the highest eco-risk, whereas γ-hexachlorocyclohexane presented the lowest eco-risk. Results of the system-level ERA were not fully consistent with the those of community-level ERA owing to the lack of a sufficient dataset, SSD model type, and ecosystem structure, as indicated by the weight of each TL. The successful application of ExSSD in Lake Chaohu signifies the start of the third stage of ERA at the system-level, and it also provides a scientific basis for ecosystem-level ERA, aquatic ecosystem protection, and future water safety management. However, there were some limitations, including sufficient data dependence, neglect of ecological interactions, and neglect of environmental parameters such as natural organic matter. We propose to employ toxicogenomics to enrich the toxicity database, to simulate the interaction using the ecological dynamic model, and to introduce the chemical fate model into the system-level ERA.