Statistical cautions when estimating DEBtox parameters.

Area of Science:

• Ecotoxicology
• Environmental Science
• Computational Biology

Background:

• Dynamic Energy Budget (DEB) models provide mechanistic insights into organismal responses to contaminants.
• DEBtox models, a specific application of DEB theory, are increasingly recognized for analyzing ecotoxicity data.
• The application of DEBtox within regulatory frameworks for ecotoxicology is emerging.

Purpose of the Study:

• To clarify and detail the model-building process for DEBtox models.
• To statistically evaluate the quality of the parameter estimation process using a least squares approach for DEBtox models.
• To identify and address statistical issues in fitting DEBtox models to standard ecotoxicity test data, specifically the 21-day Daphnia magna reproduction test.

Main Methods:

• Step-by-step rederivation of DEBtox model equations.
• Statistical evaluation of parameter estimation using a least squares approach.
• Analysis of both experimental and simulated data from the 21-day Daphnia magna reproduction test.

Main Results:

• Some DEBtox model equations yielded different results upon rederivation compared to published literature.
• Statistical issues were identified in the fitting process of DEBtox models to OECD-type reproduction data.
• Particular attention is required for parameter estimates and the interpretation of their confidence intervals when fitting DEBtox models.

Conclusions:

• The study highlights critical statistical considerations for applying DEBtox models in ecotoxicology, particularly concerning parameter estimation and confidence intervals.
• Refined understanding of DEBtox model building and statistical evaluation is crucial for their reliable use.
• Further attention to statistical methodologies is needed for the robust implementation of DEBtox models in regulatory ecotoxicology.