A computational weight of evidence platform to understand critical fish specific biology mediating toxicologically relevant responses to stress

Regulatory approval and marketing of chemicals require environmental risk assessment. It traditionally includes aquatic ecotoxicity studies of species from three trophic levels: algae, crustaceans and fish. The large number of resources, costs, and animals used for toxicity tests, lack of societal acceptance, and phasing out of animal testing for regulatory purposes have resulted in the increasing uptake of the 3Rs (Reduce, Replace, and Refine) principles in safety science and the use of new approach methodologies (NAMs). One such method is identifying chemicals for which fish are the most sensitive aquatic taxa and understanding the biological mechanisms resulting in toxicity.
We performed a large-scale analysis of the ECOTOXicology Knowledgebase (ECOTOX) for comparative sensitivity assessment of different aquatic species to chemicals. We hypothesised that this would help identify group chemicals for which risk assessment based on algae/invertebrate data would be protective of fish and amphibians, too. We developed a Python pipeline to extract, refine, harmonise, and analyse the median lethal concentrations (expressed as LC50s) and No Observed Effect Concentrations (NOECs) for all available laboratory-based aquatic toxicity studies.
Our study shows that most chemicals in ECOTOX have comparable median NOECs and LC50s across different aquatic trophic levels, and fish display higher sensitivity than other species for a small proportion of chemicals. While fewer data points were available for amphibians, a similar trend was observed, with amphibians showing a higher sensitivity than algae/invertebrates to less than 30% of chemicals. Understanding the pathways underlying the apparent higher sensitivity of fish and amphibians to these chemicals may provide a strong rationale for predicting species sensitivity to untested chemicals. It would also bypass the need for chemical testing in fish where a more sensitive lower trophic level is available as an alternative.