Model-based refinement of metal water quality criteria: Advances in biotic ligand model, multiple linear regression, and toxicokinetic-toxicodynamic approaches

Abstract

Water quality criteria (WQC) for metals rely heavily on species sensitivity distribution (SSD) models to quantify the variability of biological responses and define concentrations that safeguard aquatic communities. However, traditional SSD approaches often neglect two critical dimensions: site-specific modulation of metal bioavailability and the time-dependent nature of toxic effects. In response, several model-based refinements, including Biotic Ligand Model (BLM), Multiple Linear Regression (MLR), and Toxicokinetic-Toxicodynamic (TK-TD) frameworks, have been developed to enhance the ecological relevance and predictive reliability of WQC. In this review, we highlight the conceptual foundations of each model, their integration into SSD frameworks, and their current status and limitations in supporting WQC development. For TK-TD models, in particular, we propose solutions to address key bottlenecks, including the need for standardized time-series toxicity testing and the development of hybrid modeling frameworks (e.g., kinetic BLM). Together, these advances represent a paradigm shift from static, single-endpoint risk estimates toward dynamic, site-specific, and biologically meaningful WQC.