Chemicals of Concern in the Great Lakes Basin

Science Center Objects

The Great Lakes are an important freshwater source of drinking water, fisheries, and habitat. Chemicals of concern are introduced to the environment by human activities, but resulting ecological consequences are little understood. With federal and University partners, we are characterizing the presence of contaminants and potential effects to fish in Great Lakes tributaries.

Duluth St. Louis Bay Sampling

USGS hydrologist processes a sediment sample on the St. Louis River near Duluth, Minnesota. (Credit: Jeffrey Ziegeweid, USGS)

Study Objectives: This study aims to identify the occurrence and distribution of selected contaminants of concern in surface water and bottom sediment, and characterize potential effects to fish and wildlife across the U.S. portion of the Great Lakes Basin. This project is a collaboration with U.S. Fish and Wildlife as part of their ‘Early warning program to detect and identify emerging contaminants and their effects to fish and wildlife’ project funded through the Great Lakes Restoration Initiative.

Study locations include 24 tributaries to the Great Lakes located in Minnesota, Wisconsin, Illinois, Michigan, Ohio, and New York. Water and sediment samples were collected during 2010 to 2014 and analyzed for approximately 200 contaminants including: pharmaceuticals, hormones, fragrances, and other chemicals indicative of wastewater sources. The first two years of the study targeted sites with known contaminant sources (e.g. sewer outfalls) and were mostly focused near major metropolitan areas. Subsequent years targeted additional sites to capture other land uses and were designed to achieve a better understanding of the occurrence and distribution of contaminants throughout a range of watersheds.

Locations of sites sampled along US tributaries to the Great Lakes

Locations of sites sampled along US tributaries to the Great Lakes (Credit: USGS)



Data indicate that many of the studied contaminants are ubiquitous across the sampled tributaries and reflect land uses within the watershed. Concentrations of several contaminants exceeded sediment and/or water-quality targets, indicating those contaminants may pose a risk to aquatic biota. Surface waters within watersheds dominated by agricultural land use tended to have more pesticides, whereas waters near urban areas tended to have more pharmaceuticals and wastewater indicators. Additionally, contaminants were detected in complex mixtures highlighting the complicated nature of understanding sources, occurrence, and ecological effects of the contaminants.


To aid in understanding what influences contaminant presence in the environment, a vulnerability tool was developed using statistical models to predict the occurrence of contaminants in surface water and sediment. We used boosted regression tree methods to predict the occurrence of specific classes of contaminants using watershed characteristics such as land use, number of point sources, proximity to nearest point source, and road density. Overall, developed land use and proximity to point sources were important factors in determining the presence of contaminants, but results varied by contaminant class and media (surface water vs. sediment). This vulnerability tool can be used by to inform susceptibility of fisheries and other aquatic management activities to contaminant exposure. Currently, we are collecting additional data to test and validate the statistical models developed for the vulnerability tool.