Taxon-toxicity study of fish to typical transition metals: Most sensitive species are edible fish

Water Quality Criteria and Ecological Risk Assessment of Typical Transition Metals in South Asia

“…Different species show differences in sensitivities to the same chemical constituents, and this variation can be defined as SSD. In this study, NPKDE-SSDs were constructed based on our previous study [ 3 ], in which NPKED-SSDs have been proven could better reflect the overall trend of changes in species toxicity data than parametric models. The goodness-of-fit of SSDs were evaluated with the Kolmogorov–Smirnov (K-S) test, root mean square error (RMSE), and error sum of squares (SSE).…”

“…If the proportion of aquatic species protected is set at 95%, the acceptable concentration of the contaminant is hazardous for the remaining 5% and is therefore defined as HC5 [ 2 , 3 , 20 ], which is the concentration value corresponding to 5% cumulative probability on the SSD curve. Acute and chronic HC5 values are the basis for deriving acute and chronic WQC [ 20 ], which were calculated using an assessment factor of 2 to avoid uncertainties (Equation (1)) [ 21 ]: WQC = HC5/2 …”

“…Transition metals are widespread in aquatic environments and can adversely affect aquatic organisms and human health when their concentrations exceed threshold values [ 1 , 2 , 3 ]. Water pollution caused by transition metals, which generally originate from industrial discharge and mining activities, has recently attracted considerable global attention [ 4 ].…”

“…It is estimated from a sample of toxicity data for various species that are used as surrogates and assumed to represent the community structure of a specific ecosystem and visualized as a cumulative distribution function (CDF) to describe the SSD curve [ 15 ]. SSDs were widely applied to derive WQC for several transition metals, such as Cu, Ni, Zn, and cadmium (Cd), to protect freshwater aquatic organisms, but rarely used to derive WQC for multiple pollutants [ 2 , 3 , 12 ]. The predicted no-effect concentration (PNEC) can also be calculated via the SSD and the entropy method for ecological risk assessment [ 4 ].…”

“…In our previous study, we proved that a non-parametric kernel density estimation (NPKDE)-SSD better reflects the overall trend of changes in species toxicity data than parametric models and is more adequate in determining the internal structural characteristics of toxicity data [ 16 ]. For sensitive species, an NPKDE-SSD also yields better fits, more robustness, and better predictions than parametric models, making the calculated WQC more reliable [ 3 , 4 ].…”