Date of Award


Document Type

Open Access Master's Thesis

Degree Name

Master of Science in Environmental Engineering Science (MS)

Administrative Home Department

Department of Civil and Environmental Engineering

Advisor 1

Noel Urban

Committee Member 1

Gordon Paterson

Committee Member 2

Judith Perlinger


Polychlorinated biphenyls (PCBs) are ubiquitous contaminants worldwide and are the most frequent contaminant at US and bi-national Great Lakes areas of concern (AOCs). This study evaluated existing evidence to answer the questions: how does total PCB contamination compare between fish species, are there spatial patterns in PCB contamination in fish species, and have our remediation efforts been effective? Using multi-variate statistics (e.g. analysis of variance (ANOVA), principal component analysis (PCA), and multiple linear regression (MLR)), this research evaluated Michigan Department of Environmental Quality (MDEQ) fish contaminant monitoring data to attempt to answer the questions above.

PCB concentrations between the species were only significantly different when compared to carp (p < 0.05). There were no statistically significant differences between smallmouth bass, largemouth bass, and walleye. A congener analysis showed that there are similar general patterns in contamination but there is some nuance. For example, carp, smallmouth bass, and largemouth bass have significantly different concentrations of low and high molecular weight congeners (p < 0.05) but that is not the case for walleye (p > 0.05). This has implications for how we select suites of congeners used for “total PCB” quantification as well as toxicity.

Sources of PCBs were difficult to establish in Michigan’s rivers. It is clear that PCB concentrations in fish at AOC sites are higher than fish at non-AOC sites (p < 0.001). However, we were unable to determine definitively whether rivers were impacted by local or atmospheric sources of PCBs, but carp, largemouth bass, and walleye in Lake St Clair have similar congener profiles with high concentrations of lighter congeners. ANOVA and MLR showed that site status (AOC or non-AOC) is an important variable to consider when explaining the variance in PCB concentrations and in explaining PCB concentrations. Despite its statistical significance, it has little explanatory power in the ANOVA (partial ή2 = 0.017 – 0.040) and has a low correlation to total PCB concentrations in the MLR (0.122, p < 0.001).

Evaluating the efficacy of remediation was done with limited data. A comparison of the quantification methodologies showed that combined trends often contradicted the more accurate congener-based trends in either direction or magnitude. The PCB half-lives ranged from 3.0 – 13.5 year-1 at AOC sites and 0.64 – 17.5 year-1 at non-AOC sites. These values are not different from the atmospheric half-lives.

Considered together, this work indicates the need for a better monitoring framework to facilitate a complete understanding of sources of PCBs to Michigan’s rivers and a sampling program with the spatial extent and frequency to evaluate remediation efforts. Additionally, PCB researchers at federal and state agencies, generally, should establish a standardized suite of PCB congeners used to quantify total PCB concentrations that will permit comparisons across studies.