Augmenting aquatic species sensitivity distributions with interspecies toxicity estimation models.

Area of Science:

• Environmental Toxicology
• Ecological Risk Assessment
• Computational Toxicology

Background:

• Species sensitivity distributions (SSDs) are crucial for ecological risk assessment.
• Current SSD development is often hindered by insufficient toxicity data for diverse species.
• Interspecies correlation estimation (ICE) models offer a potential solution for data augmentation.

Purpose of the Study:

• To evaluate the effectiveness of augmenting limited aquatic species databases with ICE-derived toxicity values for SSD development.
• To compare hazard concentrations at the 5th centile (HC5) from augmented SSDs with those from reference SSDs using extensive measured data.
• To assess the impact of ICE model augmentation on HC5 uncertainty and the influence of species composition and chemical mode of action.

Main Methods:

• Developed SSDs using limited measured data augmented with ICE toxicity values (augmented SSDs).
• Compared HC5 estimates from augmented SSDs with HC5 estimates from reference SSDs based on large, diverse datasets.
• Analyzed the influence of species composition and chemical mode of action on HC5 discrepancies.

Main Results:

• Augmented SSDs showed a high proportion (0.94) of HC5 estimates within a 5-fold range of reference HC5s.
• Species composition similarity to reference SSDs slightly improved HC5 accuracy (0.76 closer).
• Acetylcholinesterase inhibitors exhibited greater HC5 discrepancies when SSDs lacked species diversity.

Conclusions:

• ICE models can effectively augment limited datasets for SSD development without significantly increasing HC5 uncertainty.
• Uncertainty analysis in risk assessments using SSDs must consider species composition, particularly for chemicals with known toxicological differences across taxa.
• Augmenting standard test species databases with ICE models is a viable strategy for robust ecological risk assessment.