Further Development and Refinement of Interspecies Correlation Estimation Models for Current-Use Dispersants

Area of Science:

• Environmental Toxicology
• Chemical Risk Assessment
• Ecotoxicology

Background:

• Limited species diversity in aquatic toxicity data for current-use dispersants creates uncertainties in hazard assessments.
• Existing Interspecies Correlation Estimation (ICE) models can be re-evaluated and expanded with available toxicity data.

Purpose of the Study:

• To refine existing and develop new dispersant-specific ICE models to increase species diversity in toxicity predictions.
• To assess the accuracy of these ICE models in predicting aquatic toxicity and their utility in developing species sensitivity distributions.

Main Methods:

• Re-evaluated 4 and developed 25 new dispersant-ICE models, incorporating a higher diversity of species (19 species).
• Predicted toxicity for untested species using known toxicity data from surrogate species.
• Compared ICE-predicted toxicity data with empirical data to evaluate model accuracy and impact on species sensitivity distributions.

Main Results:

• New models include species not previously covered, significantly increasing species diversity compared to original models.
• Dispersant-ICE models demonstrated reasonable predictive accuracy, with predictions within 3-fold of observed values (70% for new models, 88% for refined models).
• Species sensitivity distributions developed using ICE-predicted data were largely consistent with those from empirical data (83% agreement).

Conclusions:

• Dispersant-ICE models effectively fill gaps in species diversity for aquatic toxicity data.
• These models aid in addressing concerns regarding species sensitivities and improve the reliability of dispersant hazard assessments.
• The practical application of ICE models, including laboratory-to-field comparisons, supports their use in operational risk management.