Source input elucidation in aquatic systems by factor and principal component analysis of molecular marker data

Abstract

The usefulness of principal component analysis (PCA) and factor analysis (FA) for source input elucidation in environmental studies using molecular markers for sample description was evaluated. A case study involving the determination of aliphatic and chlorinated hydrocarbons, fatty acids, alcohols, chlorophylls and some detergent indicators in water particulates from a deltaic system was selected as a representative testing dataset. PCA afforded useful results to differentiate between major groups of samples but not between geochemical sources. In contrast, FA provided a direct correspondence between factor loadings and marker groups defining geochemically consistent organic matter contributions. For autochthonous compounds, FA has allowed an even more precise characterization of input sources than that obtained by the common “qualitative” molecular marker approach. These results were obtained despite the unsuitable dimensions of the dataset describing the system [102 compounds (variables) × 40 samples]. No improved results were obtained from the application of FA or PCA to subgroups of the data involving smaller number of compounds. This is consistent with the hypothesis that a limited number of sources is responsible for the occurrence of these lipid markers in the delta waters.