Chalmers Conferences, 9th European Conference on Mathematical and Theoretical Biology

Bayesian method for assessing environmental pollution effects on organism dynamic energy budget
Baba Issa Camara

Last modified: 2014-03-28

Abstract


The Dynamic Energy Budget in ecotoxicology (DEB-tox) is a widely used approach to predict efficiently ecotoxicity tests. The DEB-tox models describe the organism energy allocation for processes of growth, reproduction and maintenance. These DEB-tox models are formulated as a system of differential equations having a large number of parameters.The commonly used methods to estimate the model parameters lead to bias in the joint and marginal probability distributions. In this work, we present the experimental laboratory data obtained by exposing Daphia magna at different doses of deltamethrin. We develop a Matlab code of Bayesian inference method to estimate simultaneously joint and marginal probability distributions of DEB-tox model parameters. Our statistical analysis aims i) to estimate the tolerated toxicity thresholds, ii) to determine the pollution effects on food quality, growth, reproduction and survival of organisms. The proposed approach provides a simple and robust tool for the computation of toxicity tests using DEB-tox models. More generally, because of its flexibility, this approach can be used for toxicological tests involving an even larger number of parameters as in multiple chemical contamination.

Keywords


DEB-tox model, Metropolis-Hastiong algorithm, risk assessment, pesticide, Time-varying exposure