Abstract
•Short-term, laboratory studies outnumbered long-term, field-realistic studies.•Most studies had low replication of exposure concentrations.•Most studies did not verify their nominal exposure concentrations.•Biomarkers were widely used but unsuitable for ecosystem-level evaluations.•The perceived low environmental risk of these drugs is based on limited evidence.
Pharmaceuticals are an important class of micropollutants within the freshwater environment. There is sufficient evidence of their impacts on exposed biota to warrant further research to better assess their risk. We reviewed ecotoxicological freshwater studies conducted from 2010-2020 on carbamazepine, diclofenac, and ibuprofen, three high-use pharmaceuticals frequently detected globally in surface waters. One hundred and thirteen studies encompassing short-term (mean exposure duration 6.4 days) and long-term exposures (mean 37.8 days) were reviewed. Study designs were examined and critiqued using a qualitative analysis, and a quantitative analysis compared toxicities (lowest observed effect concentration, LOEC) between the three pharmaceuticals in both short- and long-term tests. Short-term tests were predominant (60% of studies), as reported in past related reviews. Examination of experimental designs highlighted important limitations. Most studies had low replication (n=3 per treatment) and provided no effect sizes for their findings, and 55% of studies did not supply measurements of actual exposure concentrations. Thirty-five percent of studies detected clear effects (mostly negative) at environmentally relevant concentrations. In short-term studies, biomarkers were the most sensitive endpoints; the same applied to community-level endpoints in long-term studies. The LOECs of all three pharmaceuticals (0.05-10 μg L−1) were similar in both short- and long-term studies. Future research should prioritise long-term, environmentally relevant exposures, using a combination of biomarker and community-level endpoints. Due to the acute toxicity of pharmaceuticals, future studies should investigate how these contaminants may modify communities and ecosystems, rather than single-species approaches. Finally, a comprehensive meta-analysis would be pertinent when enough long-term, environmentally relevant studies with rigorous designs and effect size information have accumulated.