Examining historical mercury sources in the Saint Louis River estuary: How legacy contamination influences biological mercury levels in Great Lakes coastal regions

Abstract

Industrial chemical contamination within coastal regions of the Great Lakes can pose serious risks to wetland habitat and offshore fisheries, often resulting in fish consumption advisories that directly affect human and wildlife health. Mercury (Hg) is a contaminant of concern in many of these highly urbanized and industrialized coastal regions, one of which is the Saint Louis River estuary (SLRE), the second largest tributary to Lake Superior. The SLRE has legacy Hg contamination that drives high Hg concentrations within sediments, but it is unclear whether legacy-derived Hg actively cycles within the food web. To understand the relative contributions of legacy versus contemporary Hg sources in coastal zones, Hg, carbon, and nitrogen stable isotope ratios were measured in sediments and food webs of SLRE and the Bad River, an estuarine reference site. Hg stable isotope values revealed that legacy contamination of Hg was widespread and heterogeneously distributed in sediments of SLRE, even in areas lacking industrial Hg sources. Similar isotope values were found in benthic invertebrates, riparian spiders, and prey fish from SLRE, confirming legacy Hg reaches the SLRE food web. Direct comparison of prey fish from SLRE and the Bad River confirmed that Hg isotope differences between the sites were not attributable to fractionation associated with rapid Hg bioaccumulation at estuarine mouths, but due to the presence of industrial Hg within SLRE. The Hg stable isotope values of game fish in both estuaries were dependent on fish migration and diet within the estuaries and extending into Lake Superior. These results indicate that Hg from legacy contamination is actively cycling within the SLRE food web and, through migration, this Hg also extends into Lake Superior via game fish. Understanding sources and the movement of Hg within the estuarine food web better informs restoration strategies for other impaired Great Lakes coastal zones.

Keywords: Bioaccumulation; Estuary; Great Lakes; Mercury; Site assessment; Stable isotopes.

Fig. 1:

Distribution of surficial (0–4 cm) sediment (a) HgT concentrations in SLRE and (b) source contributions in SLRE determined by Hg isotopes. Inset (c) shows the location of SLRE and the Bad River reference site in Lake Superior and inset (d) shows sediment concentrations within the Bad River and Slough regions. Sampling regions are divided by dashed lines. Pie charts outlined in boxes in panel b represent the regional source profile in sediments. Sources were identified as industrial (defined by Hg within Howards Bay), watershed (as defined by the Bad River sites), and precipitation (defined from literature values -Sherman et al. 2015 and Gratz et al. 2010). Region-specific maps for SLRE concentrations and source attributions are found in Fig. S1, S4, and S5. Shape files for regions of remedial and restoration activity were provided by the Minnesota Pollution Control Agency. The base map image is the intellectual property of Ersi and is used herein under license. Copyright (c) Esri and its licensors.

Fig. 2

Mercury stable isotope ratios (δ²⁰²Hg) in SLRE and Bad River surficial sediments (0–4 cm) as a function of log transformed Hg concentration (HgT). Sediments within the study regions show a continuum between regional background Hg, defined by the Bad River, and industrially influenced sediments, such as sediment hotspots in Erie Pier Ponds and Howards Bay. A subset of sediments from Allouez Bay and Bad River Slough deviate from gradient, likely due to higher organic matter content which acts as a sink for Hg within these regions. Error bars represent the 2SD of isotope measurements performed on certified reference material IAEA SL-1.

Fig. 3.

Comparison of δ²⁰²Hg_COR for mixed benthic invertebrates (dashed line), dragonfly nymphs (open box), riparian spiders (lined box), and prey fish (filled box) for SLRE and the Bad River. Sample counts are denoted on the x axis under the corresponding boxplots; mixed benthic invertebrate samples are n = 1 per region. Boxplots describe the 25–75% percentile, whiskers denote 1.5 interquartile range, median values are marked as the inner line of the box, and average is denoted by the square symbol.

Fig. 4.

Isotope biplot of δ²⁰²Hg_cor and δ¹³C for prey fish (shiners and Yellow Perch) in SLRE (blue colors) and the Bad River (orange colors). The solid orange line is the slope of the linear regression for Bad River prey fish (y = 0.019 x – 0.038, r² =0.53, n = 37). The solid blue line is the linear regression for SLRE and Bad River Slough prey fish (y = 0.019 x −0.014, r² = 0.15, n = 74), which had a weaker regression coefficient. These lines are proxies for the habitat related Hg isotope shift that occurs within these estuaries. Error bars represent the 2SD of isotope measurements performed on certified reference material IAEA-407.

Fig. 5.

Mercury isotope biplot for (a) Walleye and (b) Northern Pike in relation to dietary contributions from Lake Superior and total mercury (HgT) concentrations (wet weight). Dietary contributions are represeted as green to blue for SLRE and orange to red for the Bad River. The size of the circle represents the HgT concentration in the sample. Error bars represent the 2SD of isotope measurements performed on certified reference material IAEA-407.