The team focused on two widely used pesticide groups: Pyrethroids and Organophosphates. These chemicals are commonly found in agricultural and household pest control products, and often make their way into rivers, lakes, and other water bodies. Once there, they can pose serious risks to aquatic life, from tiny water fleas to fish. Untersteiner and his colleagues aimed to determine how toxic these substances are to different species and to identify which organisms are most sensitive to them.

Using the QSAR Toolbox, the researchers predicted the concentrations at which these pesticides begin to harm aquatic species. Their models relied on the chemical and physical properties of each substance – most notably their “log Kow” value, which is a measure of how readily a chemical dissolves in fat versus water.

For many of the pesticides, the team’s predicted toxicity levels closely matched those observed in experimental studies. For example, the predictions for the pesticide cypermethrin closely matched known data from previous toxicity tests on water fleas.

The team also used the SSD Toolbox to build so-called “species sensitivity distributions”. These models help environmental regulators to understand how different species react to the same chemical.

By analysing data from dozens of aquatic species, the researchers calculated the “HC05 value” for each pesticide, which is the concentration that would be harmful to 5% of species in an ecosystem. This kind of information is essential for setting environmental safety standards.

According to Untersteiner’s team, the HC05 values calculated for organophosphates were mostly in line with current European environmental quality standards. The values they calculated for pyrethroids were slightly less conservative than EU standards – suggesting there may be room to adjust those safety thresholds.

Importantly, the study demonstrates the value of using computer models as part of chemical risk assessment. Methods such as QSAR and SSD not only reduce the need for animal testing, but also offer faster, more cost-effective ways to rapidly screen various chemicals for ecological impact.