Novel approach to ecotoxicological risk assessment of sediments cores around the shipwreck by the use of self-organizing maps.

Marine and coastal pollution plays an increasingly important role due to recent severe accidents which drew attention to the consequences of oil spills causing widespread devastation of marine ecosystems. All these problems cannot be solved without conducting environmental studies in the area of possible oil spill and performing chemometric evaluation of the data obtained looking for similar patterns among pollutants and optimize environmental monitoring during eventual spills and possible remediation actions - what is the aim of the work presented. Following the chemical and ecotoxicological studies self-organising maps technique has been applied as a competitive learning algorithm based on unsupervised learning process. Summarizing it can be stated that biotests enable assessing the impact of complex chemical mixtures on the organisms inhabiting particular ecosystems. Short and simple application of biotests cannot easily explain the observable toxicity without more complex chemometric evaluation of datasets obtained describing dependence between xenobiotics and toxicological results.