Polycyclic Aromatic Hydrocarbons (PAHs), Nitro-PAHs and Petroleum Biomarkers in Lake Michigan.

The overall objective of this research was to understand the distribution, sources and risks of PAHs and NPAHs in Lake Michigan, to characterize their emissions from a major source (diesel exhaust), and to model and predict their environmental fate in the Lake Michigan basin. The petroleum biomarkers hopanes and steranes were characterized along with PAHs and NPAHs to provide more information on hydrocarbon contamination. Several types of samples were collected in Lake Michigan, including bottom sediments and fish. Homogenized composite samples were analyzed. Ecological and human health risks were evaluated by the toxic equivalency approach and sediment quality guidelines. Emissions of target compounds from diesel exhaust were characterized to examine the effects of fuel type, engine load and after-treatment. Finally, fugacity-based multimedia models were used to evaluate the overall behaviors of PAHs and NPAHs in Lake Michigan. In summary, the results suggest the ubiquitous distribution of target compounds in sediments and fish from Lake Michigan. PAHs seem to display "biodilution" through the aquatic biota. Sediment concentrations show a spatial trend with high concentrations near-shore and low concentrations in the middle of the lake. Decrease of PAHs in Lake Michigan sediments from the 1990s is also suggested. Source apportionment identifies vehicle exhaust as the most important PAH source in this region, and bench tests demonstrate the effectiveness of alternative fuels and after-treatment in reducing the emissions. Finally, the modeling results suggest that the environmental concentrations and emission rates of PAHs and NPAHs can be estimated mutually by simple multimedia models. This study provides new information regarding levels and risks of PAHs and NPAHs in the Great Lakes basin, especially the first report of NPAHs in sediment and biota, which can help target these chemicals for pollution prevention and reduction. This work obtains baseline data for trending progress towards elimination of toxic substances in the basin. It can also deepen our understanding to their behaviors in aquatic biota, and provide valuable information for fish advisories. Moreover, the modeling helps to improve understanding of the overall behavior of these compounds, and the estimated emission rates can complement and improve existing emission inventories.