Using Bayesian Belief Networks and Fuzzy Logic to Evaluate Aquatic Ecological Risk

are three issues: 1) ecological effects quotient is a low-level toxicity risk estimate for chemical pressure because it uses a simple number to represent a complex biological problem; 2) besides biological toxicity, species future status is also relevant to physical and biological pressures which are difficult to express by ecological effects quotient; 3) we need a higher level methodology to consider all ecological pressures. To address the issues, this study has five objectives: 1) to design the ecological risk procedure for Bayesian belief networks (BBN) approach; 2) to represent the relationships between physical, chemical and biological pressures and effects with BBN; 3) to construct conditional probabilities in BBN by fuzzy logic; and 4) study case.