A Bayesian mixture model for estimating intergeneration chronic toxicity.

Area of Science:

• Environmental toxicology
• Aquatic ecotoxicology
• Population dynamics

Background:

• Assessing multigenerational toxic effects is vital for understanding long-term chemical pollution impacts in aquatic ecosystems.
• Standard statistical methods often struggle with complex, multi-generational experimental data.
• There is a need for advanced statistical approaches to analyze intergenerational toxicity data.

Purpose of the Study:

• To apply a Bayesian mixture model with random-effects to evaluate intergenerational copper (Cu) exposure on the reproductive output of Tigriopus japonicus.
• To analyze reproductive data across three generations to understand the subtle effects of Cu exposure.
• To compare the robustness of Bayesian models against standard statistical methods for ecotoxicological data.

Main Methods:

• Utilized a Bayesian mixture model with random-effects to analyze three generations of Tigriopus japonicus reproductive data.
• Specified nonstandard statistical distributions and accounted for correlations within the experimental data.
• Investigated the impact of Cu exposure on ovisac maturation rate and nauplii production.

Main Results:

• Intergenerational Cu exposure negatively affected current generation reproductive output.
• Both current and parent generation Cu exposures demonstrated detrimental effects on reproduction.
• Evidence suggests potential adaptation to parental Cu exposure, but with a reproductive cost when Cu concentrations differed.

Conclusions:

• Bayesian mixture and random-effects models offer a robust framework for analyzing complex multigenerational ecotoxicological data.
• The developed model provides more reliable inferences than traditional statistical methods.
• This approach enhances the understanding of chemical toxicity and adaptive responses in aquatic populations.