Predicting metal toxicity from bioaccumulated concentrations in Gammarus fossarum – Part 2: Estimation of internal lethal thresholds using a toxicodynamic modelling approach

Assessing ecological risks of trace metal elements (TME) in freshwater ecosystems remains challenging, particularly when relying on bioaccumulation data that lack associated toxicity thresholds. In France, large-scale biomonitoring using caged Gammarus fossarum provides extensive datasets of internal TME concentrations, yet their interpretation in terms of biological effects is still limited. In this study, we developed a toxicodynamic (TD) model based on the General Unified Threshold model of Survival (GUTS) to link bioavailable internal concentrations of eight TMEs (As, Cd, Cr, Co, Cu, Ni, Ag and Zn) to lethal effects. Toxicokinetic (TK) parameters were first estimated in males to predict bioavailable internal concentrations under different exposure scenarios. These predictions were then coupled with survival data from a companion study to calibrate the TD model and derive internal toxicity thresholds (i.e. NECi and internal killing rates) for females gammarids and embryos. Results showed that NECi and internal killing rate parameters revealed distinct sensitivity patterns between life stages, with embryos generally more vulnerable. At the national scale, these thresholds were applied to national biomonitoring data; exceedances of NECi thresholds were rare for most TMEs, except for zinc in embryos (8.7% of measurements) and localized cadmium contamination at a historically impacted site. Overall, this study demonstrates that integrating TD modelling with biomonitoring data enables a mechanistic interpretation of bioaccumulation in terms of ecological risk. This approach provides a promising framework to support more realistic and protective environmental risk assessments, particularly when extended to mixture toxicity in freshwater ecosystems.