About the project

The main goals achieved in the project were:

1. A simple stepwise method for risk assessment analysis for the marine environment based on multiple biological responses (biomarker approach) was constructed. The analytical power of Artificial Neural Networks (ANN) for water quality classification based on biomarker data was assessed.
2. Environmental quality information was extracted from biological responses of marine organisms. The link between biomarker responses and water containing PAHs was studied and the output is satisfying. General Additive Models proved to be a useful and robust method in describing relationships between biomarkers and combinations of biomarker responses.
3. Suitable sets of biomarkers for monitoring programs, depending of the type of pollutants present in the environment, were to be defined. Data deficiency was the main challenge for this goal. Sets of biomarkers, focused on PAHs and metals, were established and used in the water quality classification performed with ANN. A field validation study was also performed.
4. A mapping system (Geographic Information Systems) to make the tool accessible for public environmental agencies was applied.
5. This project has shown the possibility to extract valuable information from complex data indicating pollution (biomarkers), using sophisticated statistical and/or modelling tools, displaying the results in an easy and comprehensible way. Water quality classification was attempted with different "classifiers" other than ANN, to explore new methods extrapolating maximum information from the actual incomplete dataset, and to test possible future use with different species and more complete datasets.