Model-Averaging Species Sensitivity Distribution for Phthalate Esters and Ecological Risk Assessment in Typical Freshwater Basins of China
The construction of species sensitivity distribution (SSD) models using a single function requires optimization to reduce subjectivity. To minimize model selection uncertainty and align with Chinese freshwater organism effect criteria, this study integrated native freshwater species toxicity data, including experimental and predicted values from interspecies correlation estimation (ICE) and acute-chronic ratio (ACR) methods, and applied a model-averaging approach to construct SSD models for seven representative phthalate esters (PAEs): dimethyl phthalate (DMP), diethyl phthalate (DEP), dibutyl phthalate (DnBP), butyl benzyl phthalate (BBP), bis(2-ethylhexyl) phthalate (DEHP), diisodecyl phthalate (DIDP), and dihexyl phthalate (DnHP). The derived short-term predicted no-effect concentrations (PNECacute) for DMP, DEP, DnBP, BBP, DEHP, DIDP, and DnHP were 16.796, 4.984, 9.064×10⁻², 2.490×10⁻¹, 1.898×10⁻², 1.386×10⁻¹, and 7.428×10⁻² μg·L⁻¹, respectively. Long-term PNECs (PNECchronic) were 3.245×10², 36.500, 1.149, 4.018, 8.949×10⁻², 1.637, and 4.073×10⁻¹ μg·L⁻¹, respectively. These PNECs, based on native species toxicity data and more stringent than existing standards, are recommended as potential references for water quality criteria based on Chinese freshwater organism effects. Ecological risk assessment using the hazard quotient (HQ) method on exposure concentrations from typical Chinese freshwater basins revealed that DEHP and DnBP posed high short-term risks, BBP mainly medium risk, while DMP, DEP, and DnHP showed low or no risk. Long-term risks indicated DEHP at medium to high risk, DnBP mainly medium to low, BBP low or no risk, and DMP, DEP, and DnHP no risk. The overall ecological risk ranking was DEHP > DnBP > BBP > DEP > DMP ≈ DnHP.