Predictive Framework for Species Sensitivity Distribution Curves of Emerging Contaminants: A Comparative Study in Marine and Freshwater Environments
Yang Huang, Fei Li, Chenyu Wang, Yuefa Teng, Rong Sun, Jianhui Tang, Chenglong Ji, Huifeng Wu, Yitao Pan, Jiayin Dai
Abstract
Emerging pollutants increasingly threaten aquatic ecosystems and human health. However, the risk assessment of these pollutants is constrained by insufficient toxicity data for aquatic species. In this study, a predictive framework has been established to predict the toxicity of potential pollutants across 16 aquatic species, including marine and freshwater species of varying trophic levels and to simulate their species sensitivity distributions (SSDs). Machine learning analysis indicated that per- and polyfluoroalkyl substances (PFASs) and organophosphate esters (OPEs) exhibited higher toxicity in marine species compared to freshwater species. The presence of aromatic rings and chlorine substituents was associated with increased toxicity of the OPEs, and the toxicity of alkyl OPEs showed a positive correlation with side chain length. Key parameters affecting PFAS toxicity included the chain length, substituent type, and number of ether bonds. Additionally, a comprehensive ecological risk assessment framework, integrating persistence, bioaccumulation, toxicity, and concentration, was developed to evaluate 13 PFASs and 6 OPEs in the Bohai Sea. Notably, our assessment framework identified several compounds, such as perfluorododecanoic acid, perfluoro(3,5-dioxahexanoic) acid, and tris(1-chloro-2-propyl) phosphate, previously considered as low or nonrisk, as posing significant ecological risks in this region. Thus, this framework offers a robust tool for advancing the risk assessment of emerging aquatic pollutants.