PREDICTION OF TOXICITY FROM CONTAMINATED SEDIMENTS USING PLS-DA AND CP- ANN MODELS M. Alvarez-Guerra 1 , D. Ballabio 2 , J.M. Amigo 3 , R. Bro 3 , J.R. Viguri 1 1 Department of Chemical Engineering and Inorganic Chemistry. ETSIIT. University of Cantabria. Avda. Los Castros, s/n. 39005, Santander, Spain. 2 Milano Chemometrics and QSAR Research Group, Department of Environmental Sciences, University of Milano-Bicocca, P.za della Scienza 1, 20126 Milano, Italy 3 Department of Food Science, Quality and Technology, Faculty of Life Sciences, University of Copenhagen, Rolighedsvej 30, 1958 Frederiksberg C, Denmark. Abstract: The development of screening tools to link chemical concentrations of sediment contaminants to toxicity effects is a challenging task of great concern, and this kind of tools can be very useful in the first steps of sediment quality assessment frameworks, with contribution to their sustainable management. This study presents new approaches for predicting acute toxicity from sediment chemistry based on the application of advanced mathematical techniques (Partial Least Squares-Discriminant Analysis (PLS-DA) and Counter-propagation Artificial Neural Networks (CP-ANNs)) to large databases of matching field-collected chemical and biological effects data. The developed PLS-DA and CP-ANN tools achieved better results than existing approaches for predicting sediment toxicity, and moreover, instead of focussing on individual-contaminant estimations, they consider all the chemicals simultaneously in order to include the effects of complex interactions among them. Despite the inherent limitations of predicting toxicity from common chemical analyses of bulk contaminant concentrations, the results obtained in the validation of our models combined relevant values both of sensitivity (ability to correctly classify a toxic sample as toxic) and of specificity (ability to correctly classify a nontoxic sample as nontoxic); this means that these tools allow decision-makers to achieve trade-offs between high environmental protection (i.e. high sensitivity) and optimal allocation of limited resources (i.e. high specificity). Keywords: environment; sediment pollution; quality assessment; management; toxicity; Partial Least Squares-Discriminant Analysis (PLS-DA); Counter-propagation Artificial Neural Networks (CP-ANNs) 1. INTRODUCTION Sediments can act as sinks of multiple chemicals that accumulate over time, making risk assessment and sustainable management difficult and complex. The assessment of sediment quality should therefore be carried out through tiered decision-making frameworks in sequential steps of increasing complexity and cost. Therefore, the development of tools to link chemical concentrations of contaminants to the potential for observing toxicity is a challenge of great interest, since these tools can be very useful in initial assessment steps for an efficient allocation of limited resources; e.g., they can serve to identify uncontaminated sediments with low probability of toxicity, in which further testing would not be necessary, or to identify sediments that are highly likely to be toxic and would require attention in subsequent management steps. According to this interest, different approaches that link chemical concentrations to the potential for observing adverse biological effects have been developed, like the Logistic Regression Models (LRM) (Field et al., 1999; 2002) or the widely used empirical Sediment Quality Guidelines (Wenning et al., 2005; Alvarez-Guerra et al., 2007). Combining the power of advanced mathematical techniques, such as Partial Least Squares-Discriminant