Using Monte Carlo analysis to characterize the uncertainty in final acute values derived from aquatic toxicity data

Area of Science:

• Environmental Toxicology
• Ecotoxicology
• Risk Assessment

Background:

• Ambient water quality criteria protect aquatic life from chemical toxicity.
• Current US EPA 1985 Guidelines use a deterministic procedure for Final Acute Values (FAVs).
• This procedure does not fully account for uncertainty in toxicity test results.

Purpose of the Study:

• To characterize uncertainty in FAVs using Monte Carlo analysis.
• To evaluate the impact of data variability on FAVs.
• To improve the understanding and communication of uncertainty in water quality criteria.

Main Methods:

• Utilized copper (Cu) EC50 values from USEPA's 2007 freshwater Cu criteria guidance.
• Applied Monte Carlo simulations to toxicity data, adjusting for water chemistry.
• Simulated scenarios with reduced species and replicate tests.

Main Results:

• Deterministic FAV for complete dataset: 4.68 µg/L.
• Monte Carlo simulations with complete data yielded mean FAV of 4.66 µg/L (5th-95th percentiles: 4.14–5.20 µg/L).
• Reduced data (8 genera, 3 replicates/species) yielded wider FAV distribution (1.22–6.18 µg/L) vs. deterministic 2.80 µg/L.

Conclusions:

• Monte Carlo analysis effectively characterizes uncertainty in FAVs derived from acute toxicity data.
• This method enhances understanding and communication of uncertainty in water quality criteria.
• Improved uncertainty assessment can benefit the development and application of water quality standards.