484
G.£. Glass et al.
where dated values were available. This indicates that natural sources of mercury
deposition and sediment accumulation were relatively low within the watershed,
in contrast with recent times.
The lowest mercury concentrations, mean values of 17 and 25 ng Hg/g
sediment (dry wt.), occurred at the furthest upstream site on the St Louis River
and in Lake Superior deeper strata sediments, respectively.
The upstream reservoirs of this study (Knife Falls and Potlatch) exhibited
fewer contaminated sediment strata and lower overall mercury concentrations
than those downstream (Scanlon, Thomson, Forbay, and Fond du Lac). This
indicates that significant loadings of mercury had occurred from localized
anthropogenic sources.
The highest mercury concentrations (3500 ng Hg/g sediment, dry wt.)
occurred in a sediment depositional zone of the Scanlon Reservoir at 9095 cm depth into the sediments.
The occurrence of elevated mercury levels in sediment strata appear as
subsurface maxima in concentration profiles as a function of depth into the
sediments. These maxima indicate a beginning and ending of significant
anthropogenic mercury loadings to the river system over a specified time period,
and show dilution by low mercury-containing sediments at both the start and
end of the loading period.
Several sampling sites exhibited high mercury concentrations (300 to
2000 ng Hg/g sediment, d. w.) in surface strata (0-10 cm). These observations
are contrary to the expectation that because point source mercury discharges
ceased during the 1970S, all sediment mercury contamination would be buried by
relatively clean sediments. This confirms the hypothesis that varying sedimentation
rates, high flow episodes, and possible new sources have caused the highest mercury
concentrations to exist in the surface strata for some sediment-sampling sites.
Average mercury concentrations (per reservoir) in Hexagenia are correlated
with average mercury in surface sediment levels, with the highest averages of
both observed in the Thomson Reservoir.
The observations of elevated mercury in surface sediments of certain reservoir
areas, and the significant positive correlation between mercury in sediment and
Hexagenia, indicate that sediment mercury can act as a mercury contamination
source to fish food organisms, and therefore, indirectly influence fish residue
levels within those reservoirs.
The areas exhibiting high mercury levels in surface sediment on the Thomson
Reservoir are the focus of further studies on testing mitigation approaches and
identifying bioaccumulation mechanisms.
Ackllowledgemellts alld Disclaimer We acknowledge and thank the Great Lakes National Program
Office, USEPA ti,r major financial support (Grant No. GL995478) and the Duluth EPA lab for salary
(G.E.G.), space, and equipment support for this project. For their parts in sampling and analysis we
thank Dan Engstrom (UM LRC), James Huber, Randy Rude, Kent Schmidt, Vadim Melnichuk, Ed
Swain (MPCA), and Mark Zacker (POL). We also thank Jane Reed, Charles Lippert, and Kim Nieman
for assistance in data presentation.
The mention of trade names and sources of commercial products is for identification and does not
constitute endorsement or recommendation for use. The views expressed in this chapter are
G.£. Glass et al.
where dated values were available. This indicates that natural sources of mercury
deposition and sediment accumulation were relatively low within the watershed,
in contrast with recent times.
The lowest mercury concentrations, mean values of 17 and 25 ng Hg/g
sediment (dry wt.), occurred at the furthest upstream site on the St Louis River
and in Lake Superior deeper strata sediments, respectively.
The upstream reservoirs of this study (Knife Falls and Potlatch) exhibited
fewer contaminated sediment strata and lower overall mercury concentrations
than those downstream (Scanlon, Thomson, Forbay, and Fond du Lac). This
indicates that significant loadings of mercury had occurred from localized
anthropogenic sources.
The highest mercury concentrations (3500 ng Hg/g sediment, dry wt.)
occurred in a sediment depositional zone of the Scanlon Reservoir at 9095 cm depth into the sediments.
The occurrence of elevated mercury levels in sediment strata appear as
subsurface maxima in concentration profiles as a function of depth into the
sediments. These maxima indicate a beginning and ending of significant
anthropogenic mercury loadings to the river system over a specified time period,
and show dilution by low mercury-containing sediments at both the start and
end of the loading period.
Several sampling sites exhibited high mercury concentrations (300 to
2000 ng Hg/g sediment, d. w.) in surface strata (0-10 cm). These observations
are contrary to the expectation that because point source mercury discharges
ceased during the 1970S, all sediment mercury contamination would be buried by
relatively clean sediments. This confirms the hypothesis that varying sedimentation
rates, high flow episodes, and possible new sources have caused the highest mercury
concentrations to exist in the surface strata for some sediment-sampling sites.
Average mercury concentrations (per reservoir) in Hexagenia are correlated
with average mercury in surface sediment levels, with the highest averages of
both observed in the Thomson Reservoir.
The observations of elevated mercury in surface sediments of certain reservoir
areas, and the significant positive correlation between mercury in sediment and
Hexagenia, indicate that sediment mercury can act as a mercury contamination
source to fish food organisms, and therefore, indirectly influence fish residue
levels within those reservoirs.
The areas exhibiting high mercury levels in surface sediment on the Thomson
Reservoir are the focus of further studies on testing mitigation approaches and
identifying bioaccumulation mechanisms.
Ackllowledgemellts alld Disclaimer We acknowledge and thank the Great Lakes National Program
Office, USEPA ti,r major financial support (Grant No. GL995478) and the Duluth EPA lab for salary
(G.E.G.), space, and equipment support for this project. For their parts in sampling and analysis we
thank Dan Engstrom (UM LRC), James Huber, Randy Rude, Kent Schmidt, Vadim Melnichuk, Ed
Swain (MPCA), and Mark Zacker (POL). We also thank Jane Reed, Charles Lippert, and Kim Nieman
for assistance in data presentation.
The mention of trade names and sources of commercial products is for identification and does not
constitute endorsement or recommendation for use. The views expressed in this chapter are