Environmental Engineering Reference
In-Depth Information
Even in the mineral soil, Hg is presumed to be nearly
exclusively associated with organic matter; Grigal (2003)
used C content as a surrogate to estimate a national (U.S.)
forest soil Hg inventory. Few measurements of MeHg in
mineral soil exist, but it is generally less than 2% of THg
(Grigal, 2003). Interestingly, while soil C content decreases
sharply with depth from the forest fl oor to mineral soil,
many investigators have found that the Hg:C ratios are
higher in the mineral soil (Aastrup et al., 1991; Grigal et al.,
1994; Schwesig et al., 1999).
The accumulations of Hg near the terrestrial-atmo-
spheric interface, coupled with the known affi nity of Hg
for organic matter, leave little doubt that this surface Hg
is predominantly atmospheric. But as one moves deeper
into the mineral soil and Hg concentrations decrease, back-
ground geogenic Hg probably becomes increasingly impor-
tant. Some of the Hg in mineral soil may have originated
from natural atmospheric deposition of Hg that moved
slowly downward over millennia. In Sweden, the Hg con-
centration in mineral soil at ~50 cm depth is uniform from
south to north, despite a strong gradient in Hg deposition,
which shows up clearly in the surface mor layer (Figure 8.2).
Moreover, the total store of Hg in most soils is too great to
be explained by recent anthropogenic emissions. In assess-
ing the rate of Hg sequestration by soil and anthropogenic
effects, it would be helpful to know how much of the Hg
present is native (geogenic) and how much is atmospheric.
Quantifi cation of Hg in unweathered parent material may
help estimate the geogenic Hg proportion.
concluded that volatilization must be large. Some plot scale
measurements in the Adirondack Mountains, United States,
support volatilization fl uxes approaching the magnitude of
Hg wet deposition fl uxes—for example, 7.0 µg m
-2
a
-1
from
a forest fl oor (Choi and Holsen, 2009b) and 4.6 µg m
-2
a
-1
from a wetland (Selvendiran et al., 2008). Global Hg mass
balance considerations likewise suggest that a sizable frac-
tion of terrestrial Hg deposition is revolatilized to the atmo-
sphere (Mason and Sheu, 2002).
Several watershed studies suggest lower Hg volatiliza-
tion losses. In an unpolluted boreal forest in Canada, St.
Louis et al. (2001) concluded that volatilization amounted
to only about 10% of Hg deposition. In the METAALICUS
study, volatilization was directly measured from isoto-
pically distinct Hg input, and was found to occur only
in the fi rst few months after application (Lindberg
et al., 2003), accounting for about 8% of the applied isotope
(Hintelmann et al., 2002). In an overview of catchment Hg
cycling, Krabbenhoft et al. (2005) concluded that volatil-
ization from forest soils was a comparatively small fl ux. In
montane grasslands, Fritsche et al. (2008) found net uptake
of atmospheric Hg, while Converse et al. (2010) found a
bidirectional fl ux dependent on season. Ultraviolet radia-
tion is thought to be the primary driver of volatilization
both from foliar surfaces (Graydon et al., 2006) and forest
soils (Choi and Holsen, 2009a). Johnson et al. (2003) also
showed that soil moisture and other factors affect Hg vola-
tilization from soils and that it is not driven by diffusion.
For more on Hg volatilization see chapter 7.
Mercury Outputs
Stream Mercury Export
The two main pathways of Hg output from the terrestrial
landscape are volatilization to the atmosphere and export
in streamwater. Volatilization is diffi cult to quantify, and
assessments of its importance in the terrestrial Hg cycle
range widely. Streamwater Hg export is more readily quan-
tifi ed, and is typically only a small fraction of Hg in deposi-
tion, which itself is measured in parts per trillion. Yet this
“small fraction of a trace amount” exported from terrestrial
landscapes has profound ecologic signifi cance. In this sec-
tion, we examine the magnitude and processes controlling
THg and MeHg outputs from catchments.
TOTAL MERCURY
Export by streamwater is the dominant loss pathway for
THg in most catchments. Most streams have detectable
levels of THg (
0.1 ng L
-1
) at all times. At base fl ow, con-
centrations generally range from 0.5 to 2 ng L
-1
(Hurley
et al., 1995; Babiarz et al., 1998; Balogh et al., 1998a, 1998b,
2005; Hurley et al., 1998; Scherbatskoy et al., 1998), and are
sustained by dissolved THg in groundwater. Nearly all
Hg
in streamwater occurs as Hg(II) in dissolved or particulate
form. MeHg also occurs in dissolved or particulate form.
Dissolved Hg is operationally defi ned as the Hg fraction
that passes a fi lter membrane (pore size range, 0.2-0.7 µm),
but Babiarz et al. (2001) showed that much of the
Volatilization
0.2-µm
Hg fraction is colloidal. For this reason the term
fi ltered Hg
is preferable when referring to discrete samples. In addi-
tion to Hg(II), some dissolved gaseous Hg(0) is present but
typically represents
Terrestrial soils are a primary sink for atmospherically
deposited mercury, but some of this Hg is volatilized back
to the atmosphere. Estimates of Hg volatilized from soil
or vegetation surfaces to the atmosphere vary widely and
error bars on the estimates are large. Hg deposition and
volatilization are both too small relative to the soil Hg pool
to calculate volatilization from an annual mass balance
approach. In a long-term mass balance, Grigal (2002) com-
puted that soil Hg accretion since deglaciation accounted
for only about one quarter of the soil Hg pool, and therefore
1% of the aqueous Hg in streams. Dis-
solved gaseous Hg(0) is important in lakes and other water
bodies because it regulates the amount of dry deposition
of Hg to the water surface (O'Driscoll et al., 2004; Eckley
et al., 2005).
Catchments release only a minute amount of their stored
Hg to st rea mwater, but st rea m Hg fl ux tends to be focused in
