This study evaluates the applicability of three bacterial short-term genotoxicity test systems to aquatic suspended particulate matter of the Elbe river. This material was sampled in sedimentation vessels after deposition periods of one month. It was extracted in a Soxhlet apparatus with toluene and methanol. Aqueous elutriates were prepared additionally. A solid phase method was developed that enables to incubate bacteria in contact with the particulate material. The test battery consists of two mutagenicity assays (the Ames-test and the Ara-test) and an SOS induction assay (the umu-test). Both mutagenicity assays came to nearly the same assessment of the samples of particulate matter of the Elbe. The quantitative response, however, was higher in the Ara-test. The particulate river material generally induced lower genotoxic potencies in the umu-test than in the mutagenicity assays. This lead to a completely different outcome of the umu-test; 29 out of 35 mutagenic samples were not SOS inducing. No quantitative or rank correlations between the concentrations of anthropogenic contaminants (PAHs, chlorinated hydrocarbons and metals) and the observed effects could be established on a 90% confidence limit. However, there is obvious correspondence between more contaminated regions of the river system and mutagenic effects (Ames- and Ara-test) in the samples from this region, as well as correspondence between low contaminated regions and the absence of mutagenicity. For this reason, the mutagenicity assays appear more favourable to describe the anthropogenic contamination with genotoxins in complex mixtures than the umu-test. The authors recommend the Ara-test for a first genotoxicity screening of complex environmental mixtures. This forward mutagenicity assay is advantageous due to higher effects and lower costs compared to the Ames-test. The development and use of a solid phase version of Ames- and Ara-test revealed the occurrence of a major part of particle-bound mutagens. The hydrophobic nature of these mutagens was also confirmed by the gradually decreasing effects with decreasing lipophilicity of the solvents. The results suggest that the solid phase test and the use of extracts complement each other in detecting mutagens of different lipophilicity. Both versions should be used in order to include a broad variety of compounds.
