28
R. Ebinghaus et al.
lakes in Northern Wisconsin as part of the MTL (mercury in temperate lakes)
project. All investigated lakes are seepage lakes (meaning they have no
permanent surface water in- or outflow) and are not directly impacted by
anthropogenic Hg sources. Consequently, the general finding was that atmospheric deposition was the major source of Hg for these lakes, which represent
net sinks for atmospheric mercury since the modeled volatilization rates were
always significantly smaller than the total Hg deposition from the atmosphere
(consisting of measured wet deposition through rain and snow and estimated
particulate dry deposition). Even though two out of three parameters in this
mass balance are not determined experimentally, it still seems reasonable to
conclude from these studies that lakes are net mercury sinks on an annual basis.
In contrast, studies in anthropogenically contaminated lakes like the Davis Creek
Reservoir in northern California show the opposite behavior, with the evasion
flux being at least twice as big as the atmospheric Hg deposition (Porcella 1994).
This is reflected in the fact that dissolved Hg( 0) concentrations in a
contaminated lake (Mason et al. 1995a) are five times higher than in remote
lakes (Vandal et al. 1995). Some studies suggest that DMM may be directly
formed from inorganic Hg under suboxic/oxic conditions and could, by
decomposition, be the actual source of MMM observed in freshwaters (Mason
et a1. 1995a).
The ratio between atmospheric deposition and volatilization varied greatly
between the individual MTL lakes, apparently due to differences in the water
chemistry between the lakes (Watras et a1. 1994). Between 10 and 50% of the
mercury deposited from the atmosphere were reemitted, and the revolatilization
was correlated positively with pH, DOC, and residence time in the upper water
column. Calculated flux rates range from 1 to 5 pg m -2 year -1 (0.11 to
0.57 ng m- 2 h- '). All mercury not reemitted was usually removed from the
water column via sedimentation. It was also observed that MeHg was wet
deposited from the atmosphere (about 1% of the total Hg deposition), and all
deposited MeHg was removed by sedimentation. It has to be noted that the
calculated flux rates on an hourly basis are much smaller than those measured at
the Swedish lakes mentioned above. Since both the American and Swedish study
lakes should be fairly comparable and also receive comparable atmospheric Hg
inputs, it cannot be ruled out that the model calculations underestimate Hg
volatilization considerably, which could make them less important sinks or even
- in the worst case - alter the overall role of lakes in regional Hg budgets from
sinks to sources. In the Lake Gardsjon study, it was estimated that the lake could
be a net source of Hg( 0) during warm dry summer months, but is a net sink for
the rest of the year (Lindberg et a1. 1996).
The mechanism of Hg volatilization from lakes remains unclear. In contrast to
oceans, information on DMM in lake waters is not known to the authors. Thus,
Hg(o) is the only volatile Hg species contributing to the quantification of the
emission flux. Since it is known that Hg(H) reduction can occur through both
biotic and abiotic reactions, attempts have been made to further elucidate the
formation of Hg(o). It has been found that mercury reduction is lower in heatsterilized lake water, suggesting that the reduction process might be biologically
mediated (Vandal et al. 1995). This is in agreement with studies in Upper Mystic
R. Ebinghaus et al.
lakes in Northern Wisconsin as part of the MTL (mercury in temperate lakes)
project. All investigated lakes are seepage lakes (meaning they have no
permanent surface water in- or outflow) and are not directly impacted by
anthropogenic Hg sources. Consequently, the general finding was that atmospheric deposition was the major source of Hg for these lakes, which represent
net sinks for atmospheric mercury since the modeled volatilization rates were
always significantly smaller than the total Hg deposition from the atmosphere
(consisting of measured wet deposition through rain and snow and estimated
particulate dry deposition). Even though two out of three parameters in this
mass balance are not determined experimentally, it still seems reasonable to
conclude from these studies that lakes are net mercury sinks on an annual basis.
In contrast, studies in anthropogenically contaminated lakes like the Davis Creek
Reservoir in northern California show the opposite behavior, with the evasion
flux being at least twice as big as the atmospheric Hg deposition (Porcella 1994).
This is reflected in the fact that dissolved Hg( 0) concentrations in a
contaminated lake (Mason et al. 1995a) are five times higher than in remote
lakes (Vandal et al. 1995). Some studies suggest that DMM may be directly
formed from inorganic Hg under suboxic/oxic conditions and could, by
decomposition, be the actual source of MMM observed in freshwaters (Mason
et a1. 1995a).
The ratio between atmospheric deposition and volatilization varied greatly
between the individual MTL lakes, apparently due to differences in the water
chemistry between the lakes (Watras et a1. 1994). Between 10 and 50% of the
mercury deposited from the atmosphere were reemitted, and the revolatilization
was correlated positively with pH, DOC, and residence time in the upper water
column. Calculated flux rates range from 1 to 5 pg m -2 year -1 (0.11 to
0.57 ng m- 2 h- '). All mercury not reemitted was usually removed from the
water column via sedimentation. It was also observed that MeHg was wet
deposited from the atmosphere (about 1% of the total Hg deposition), and all
deposited MeHg was removed by sedimentation. It has to be noted that the
calculated flux rates on an hourly basis are much smaller than those measured at
the Swedish lakes mentioned above. Since both the American and Swedish study
lakes should be fairly comparable and also receive comparable atmospheric Hg
inputs, it cannot be ruled out that the model calculations underestimate Hg
volatilization considerably, which could make them less important sinks or even
- in the worst case - alter the overall role of lakes in regional Hg budgets from
sinks to sources. In the Lake Gardsjon study, it was estimated that the lake could
be a net source of Hg( 0) during warm dry summer months, but is a net sink for
the rest of the year (Lindberg et a1. 1996).
The mechanism of Hg volatilization from lakes remains unclear. In contrast to
oceans, information on DMM in lake waters is not known to the authors. Thus,
Hg(o) is the only volatile Hg species contributing to the quantification of the
emission flux. Since it is known that Hg(H) reduction can occur through both
biotic and abiotic reactions, attempts have been made to further elucidate the
formation of Hg(o). It has been found that mercury reduction is lower in heatsterilized lake water, suggesting that the reduction process might be biologically
mediated (Vandal et al. 1995). This is in agreement with studies in Upper Mystic