MERCURY IN AND FLUXES TO LAKE MICHIGAN SURFICIAL SEDIMENTS

Ronald Rossmann (United States Environmental Protection Agency, Mid-Continent Ecology Division - Duluth, Community-Based Science Support Staff, Large Lakes Research Station, Grosse Ile, MI 48138 USA, rossmann.ronald@epamail.epa.gov

ABSTRACT

Sediment samples were collected from Lake Michigan between 1994 and 1996. Mercury concentrations in surficial sediment ranged between 2 and 260 ng/g with a mean and median of 78 and 79 ng/g, respectively. The standard deviation of the mean for the 118 stations was 65 ng/g. Mercury concentrations were similar to those reported in the literature for Lake Michigan sediments collected from the late 1960's to mid-1970's. With a standard deviation of 6.9 ng/cm²/y, mercury fluxes averaged 7.2 ng/cm²/y and ranged between 0.85 and 32 ng/cm²/y. Regional atmospheric fluxes accounted for 50% of the total mercury flux to the sediments.

INTRODUCTION

With the identification of a mercury problem in fish from Lake St. Clair in 1969, the analysis of environmental samples for mercury became a priority (Thomas 1974). During the period of 1969 through 1975, surficial sediment surveys of the Great Lakes were made which included mercury as one of the parameters measured. Locations sampled and reported include southern Lake Michigan (Kennedy et al. 1971), western Lake Erie (Walters et al. 1972, Wolery and Walters 1974), Lakes St. Clair, Ontario, Erie, Huron, and Superior (Thomas 1974), and Lake Michigan (Cahill 1981). Since 1975, no comprehensive study of mercury in the surficial sediments of any of the Great Lakes has been made. This lack of information prompted the USEPA to analyze samples of opportunity from Lake Superior and Green Bay and to undertake a study of Lake Michigan sediments (Rossmann 1999). Sediment samples were collected from Lake Michigan during the years 1994-1996 as part of the Lake Michigan Mass Balance Project. The specific objective of the sediment mercury study was to document concentrations of mercury in and fluxes to the surficial 1 cm of sediments.

METHODOLOGY

When possible, sediments were collected with a box core. Subcores from the box cores for mercury analysis were sectioned at 1 cm intervals from top to bottom. At locations where box coring was not possible, hard substrate, PONAR grab samples were taken, and the surficial 1 cm was removed. Samples were stored frozen until they were freeze-dried and homogenized for subsampling. Subsamples for mercury analysis were either extracted into 10% (v/v) nitric acid using microwave digestion or into 50% aqua regia with an addition of potassium permanganate using an automated extraction system (Uscinowicz and Rossmann 1997). Extracts were analyzed by flameless atomic absorption using an automated system (Uscinowicz and Rossmann 1997).

RESULTS AND DISCUSSION

The mean mercury concentration for the 118 stations sampled was 78 ng/g which was nearly identical to the median concentration of 79 ng/g (Table 1). Mercury concentrations were less than 20 ng/g at 32% of the stations and less than 160 ng/g at 96% of the stations. Only two samples exceeded 200 ng/g mercury. With a mean of 7.2 ng/cm²/y and standard deviation of 6.9 ng/cm²/y, mercury fluxes ranged between 0.85 and 32 ng/cm²/y. Mercury concentrations and fluxes were highest in the depositional basins of the lake (Figures 1 and 2). Mercury concentrations were higher on the eastern side of the lake than its western side. Highest mercury fluxes were found on the eastern side of the lake.

The lake’s depositional basins were compared for mercury concentrations and fluxes (Table 1). The mean mercury concentrations in the basins varied between 120 and 160 ng/g, within the observed standard deviations. Two anomalies are noteworthy. First, the minimum concentration in the Southern Basin is substantially lower than those for the other basins (see Figure 2 for basin locations). The reason is unknown. Second, the maximum concentration for the Waukegan Basin is significantly higher than those for the other basins, suggesting a historic or current high mercury concentration source to that basin. Other than these noted differences, the sediments in the lake’s depositional basins are amazingly similar, suggesting either a regional source of mercury to the lake (atmospheric) or a well mixed lake that redistributes inputs extremely well prior to sedimentation to the lake bottom.

Table 1. Mercury concentrations in various depositional basins of Lake Michigan.

Mean Standard Median Minimum Maximum
Mean Deviation Median Minimum Maximum
Basin N (ng/g) (ng/g) (ng/g) (ng/g) (ng/g)

Southern 15 120 31 130 72 180
Waukegan 11 160 65 130 100 320
Grand Haven 6 150 19 150 120 170
Milwaukee 3 130 15 130 110 140
Sarian 1 160
Algoma South 8 140 17 140 120 180
Algoma Central 7 130 15 130 100 150
Algoma North 2 130 110 150
Traverse 1 160

Fluxes for each station were calculated as the product of the measured mercury concentration expressed as ng/g multiplied by the Pb-210 sediment rate expressed as g/cm²/y. Like mercury concentrations, mean mercury fluxes did not significantly vary from basin to basin of the lake (Table 2). All fluxes were within one standard deviation of one another. Of interest are the significantly higher minimum fluxes to the Algoma Basin relative to the other basins. In general, basins that are towards the west side of the lake have lower mean and median fluxes than those on the east side of the lake. Of significant note are the relatively high maximum mercury fluxes to the Southern and Grand Haven Basins. Both of these basins are on the east side of the lake. These high fluxes may be related to the transport of materials from the southwestern and southern shore of the lake to the eastern shore especially in the spring. This transport event occurs annually, and the resulting plume has suspended particulate matter concentrations 4 to 10 times that of the lake (Eadie et al. 1996). A large amount of particulate matter with its associated contaminants is transported along the eastern shore where it settles to the lake floor and accumulates in the Southern and Grand Haven Basins.

The contribution of the regional mercury atmospheric source (mercury from outside the Lake Michigan basin) to the total mercury flux was calculated for each station. To account for the effect of sediment focusing at each station, total mercury fluxes were Cs-137 focusing factor corrected by dividing the total flux by the Cs-137 focusing factor. This corrected flux was divided into its regional atmospheric and local (atmospheric, river, and direct runoff derived from within the basin) components. The regional atmospheric flux is the same as the global total atmospheric mercury flux and is 1.2 ng/cm²/y (Porcella 1996, Mason and Sullivan 1997, Rossmann 1999). The present contribution of the regional atmospheric total mercury flux at each station was obtained by dividing it by the total Cs-137 focusing factor corrected flux. For Lake Michigan, regional atmospheric mercury fluxes accounted for 50% of the total mercury flux. This is higher than that reported for Lake Superior (38%) or calculated for Green Bay (14%) (Rossmann 1999). Thus fluxes of mercury to Lake Michigan are derived equally from regional sources outside the basin and local sources from within the basin.

Table 2. Total mercury fluxes to various depositional basins of Lake Michigan.

Mean Standard Median Minimum Maximum
Mean Deviation Median Minimum Maximum
Basin N (ng/cm²/y) (ng/cm²/y) (ng/cm²/y) (ng/cm²/y) (ng/cm²/y)

Southern 15 10. 8.7 6.5 0.85 32.
Waukegan 11 3.4 1.9 2.9 1.4 8.5
Grand Haven 6 10. 12. 4.0 0.94 31.
Milwaukee 3 3.3 1.9 4.0 1.1 4.8
Sarian 1 14.
Algoma South 8 6.9 4.9 5.1 2.8 16.
Algoma Central 7 5.2 2.5 5.2 2.6 9.5
Algoma North 2 7.6 7.1 8.0
Traverse 1 8.0

REFERENCES

Cahill, R. A. (1981), Geochemistry of Recent Lake Michigan Sediments. Illinois Institute of Natural Resources, State Geological Survey Division, Circular 517.

Eadie, B. J., Schwab, D. J., Assel, R. A., Hawley, N., Lansing, M. B., Miller, G. S., Morehead, N. R., and Robbins, J. A. (1996), EOS, Transaction, American Geophysical Union 77:337-338.

Kennedy, J., Ruch, R. R., and Shimp, N. F. (1971), Distribution of mercury in unconsolidated sediments from southern Lake Michigan. Illinois State Geological Survey, Environmental Geology Notes Number 44, Studies of Lake Michigan Bottom Sediments Number Seven.

Mason, R. P., and Sullivan, K. A. (1997), Environ. Sci. Tech. 31:942-947.

Porcella, D. B. (1996), In: Proceedings of 1995 Canadian Mercury Network Workshop, York University, Toronto, http://www.cciw.ca/eman-temp/reports/publications/mercury95/part5.html

Rossmann, R. (1999), J. Great Lakes Res. 25:683-696.

Thomas, R. L. (1974), In: Proceedings of the International Conference on Transport of Persistent Chemicals in Aquatic Ecosystems, pp. I-1 - I-16.

Uscinowicz, T., and Rossmann, R. (1997), In: USEPA Lake Michigan Mass Balance Study (LMMB) Methods Compendium, Volume 2, Organic and Mercury Sample Analysis Techniques, EPA-905-R-97-012b, Great Lakes National Program Office and Office of Water, pp. 473-503.

Walters, L. J., Jr., Herdendorf, C. E., Charlesworth, L. J., Jr., Anders, H. K., Jackson, W. B., Skoch, E. J., Webb, D. K., Kovacik, T. L., and Sikes, C. S. (1972), In: Proc. 15^th Conf. Great Lakes Res., Internat. Assoc. Great Lakes Res., pp. 306-316.

Wolery, T. J. and Walters, L. J., Jr. (1974), In: Proc. 17^th Conf. Great Lakes Res., Internat. Assoc. Great Lakes Res., pp. 235-249.