Predicting metal toxicity to aquatic life:a first step towards integrating a QICAR approach with the BLM framework

The biotic ligand model (BLM) is a tool that quantifies the bioavailability of metals, accounting for both metal speciation and hydrogeochemical properties. The BLM assumes that metal ions bind to biologically active sites, the so-called biotic ligands (BLs), and that metal toxicity is caused by the binding of metal ions to BLs. The binding affinity of a given metal ion for the BLs is represented by a conditional binding constant (K). This parameter is typically obtained via metal toxicity testing. However, metal toxicity testing is time consuming, labor intensive, and expensive. Thus, there is an urgent need for a system to predict BLMs for data-poor metals. Herein, relationships between the log K values of known metals (derived from the literature) and metal ionic characters were investigated to construct quantitative ion character-activity relationship (QICAR) equations. The results showed that log K was significantly correlated with the softness index (σp) in rainbow trout (adjusted R² = 0.901; F = 65.0; p < 0.01) and Daphnia magna (adjusted R² = 0.871; F = 41.5; p < 0.01). The σp-based QICAR equations were coupled with the BLM to construct the novel QICAR-BLM, and this model was used to predict the median effective concentration (EC₅₀) of metals with unknown toxicity in various aqueous solutions. The log K and EC₅₀ values predicted by the QICAR-BLM were reasonably consistent with previously published measured values; differences between predicted and measured values were generally less than an order of magnitude. In summary, the novel QICAR-BLM developed herein is a promising tool with which to predict the toxicity of metals of unknown toxicity. This model can serve as a framework for the establishment of water quality standards and ecological risk assessments.