Environmental Engineering Reference
In-Depth Information
that particulate matter formed within the lake and caught
by sediment traps had been actively recycled and methyl-
ated in the hypolimnion (Chadwick et al., 2006).
presented in the original works. Some studies ignored dry
deposition, so watershed retention is likely to be consid-
erably underestimated. Overall retention averaged 84%
(median, 89%) for THg (n
23) and 58% (median, 79%)
for MeHg (n
9). In addition to all other uncertainties, this
compilation does not account for volatilization, so some of
the “retained THg” may have volatilized.
Watershed retention ranges from about 55% to 95%
for THg and from -28% (net export) to 95% for MeHg
(Table 8.2). The wide ranges refl ect the importance of
watershed features and processes, which tend to trump
deposition as a control on THg and MeHg export (Driscoll
et al., 2007; Evers et al., 2007). Some of the range is also
attributed to uncertainties in or failure to account for dry
deposition, and inaccurate stream Hg fl ux due to lack of
event sampling. Watershed retention of MeHg is com-
monly less than that of THg, refl ecting internal production
of MeHg; MeHg is cyclically produced and degraded in the
landscape, and MeHg inputs probably have little bearing
on MeHg exports, although THg input (which affects the
supply of Hg available to be methylated) may partly control
MeHg export.
Studies that accounted for dry deposition generally
showed the greatest THg inputs (
Mercury Mass Balances
As a framework for interpreting terrestrial processes affect-
ing Hg dynamics, we review the literature on watershed
mass balances. Quantifying the inputs and outputs from
the landscape establishes boundary conditions (Likens and
Bormann, 1995) from which processes can be inferred and
accumulation (or loss) rates can be determined and evalu-
ated relative to catchment stocks. Watershed input-output
budgets for THg and MeHg have been compiled previously
(Allan and Heyes, 1998; Lee et al., 1998). Grigal (2002) per-
formed a thorough synthesis of the available literature and
condensed THg and MeHg inputs and outputs from more
than 100 studies in the form of histograms. Here we build
on these earlier efforts by tabulating and discussing spe-
cifi c Hg mass balance efforts from multiple landscapes,
including several more recent studies. Finally, we demon-
strate the value of applying Hg mass balances to specifi c
components of the landscape (e.g., an individual wetland)
within a given catchment.
Input-output budgets for THg are subject to inaccuracies
because few studies determine dry deposition. To be fair to
the earlier investigators, the importance of dry deposition
has been widely recognized only in the past 10-15 years,
beginning with the discovery of high Hg fl uxes in through-
fall and litterfall (Driscoll et al., 1994; Hultberg et al., 1995).
We now know that, particularly in forested areas, dry depo-
sition of THg generally dominates inputs, but measure-
ments in conjunction with mass balance studies are still
relatively infrequent and have considerable uncertainty.
Most investigators continue the convention of reporting
watershed Hg retention relative to wet deposition, recog-
nizing that wet deposition may underestimate Hg input
by a factor of 2-4. Quantifi cation of stream Hg output is
also hampered by fi xed interval sampling schedules, which
tend to underestimate stream Hg export because of its
episodic nature (Bishop et al., 1995a, 1995b; Hurley et al.,
1998; Shanley et al., 2008; Demers et al., 2010). Depending
on individual stream Hg dynamics and frequency of the
sampling program, actual Hg stream export could be two
or more times the calculated fl ux. But even taking this
underestimation into account does not alter the general
fi nding that Hg export is small relative to inputs.
These caveats notwithstanding, in Table 8.2 we present
a compilation of several Hg mass balance studies repre-
senting a diversity of landscape types. We have also syn-
thesized typical values of inputs and outputs as well as
internal THg and MeHg fl uxes and stores as an aid to inter-
preting catchment Hg budgets (Figure 8.5). In Table 8.2,
we have calculated watershed retention of THg and MeHg
from published fl uxes of the investigators, using measured
or author-estimated dry deposition and stream fl uxes as
20 µg m
-2
a
-1
) and had
near or above the median THg retention. The Lehstenbach
and Steinkreuz catchments in Germany had among the
highest wet and dry THg deposition, yet had 88% and 95%
retention (Schwesig et al., 2000, 2001). At Río Icacos, Puerto
Rico, where THg inputs were high and dry deposition
was estimated as two times wet deposition, retention was
much lower (54%), possibly due to an internal catchment
source (a negligible factor in most watersheds) and/or the
high erosion rate (Shanley et al., 2008). Allequash Creek,
Wisconsin, had the greatest overall THg retention of nearly
99%, refl ecting estimated dry THg input and low THg out-
put in streamfl ow dominated by low-DOC groundwater
discharge from a sandy aquifer, coupled with low erosion
rates (Shanley et al., 2008). Erosion plays a role by physi-
cally removing Hg that would otherwise be bound indefi -
nitely by sorption to or incorporation in organic matter.
Watershed Hg retention appears to be more complex than
the simple retention of contemporary Hg deposition. Hg accu-
mulation rates in lake sediment are consistent with watershed
retention rates discussed earlier (Swain et al., 1992), but the
sediment record suggests less watershed Hg retention during
the more recent period of enhanced anthropogenic Hg depo-
sition (Lorey and Driscoll, 1999; Kamman and Engstrom,
2002). Meili et al. (2003) suggested that the equilibration time
of catchments to Hg deposition can be measured in centuries.
Catchment characteristics play an important role in Hg reten-
tion. For example, Nelson et al. (2007) studied two forested
catchments in Maine, one of which had been burned in the
1940s. Though receiving similar THg inputs as measured in
throughfall, the burned catchment exports only about one
third as much THg as the unburned catchment. One interpre-
tation for the greater THg retention at the burned catchment
