U.S. EPA Office of Research
and Development's Science
To Achieve Results (STAR)
Research in Progress
Vol.4 Issue 1 October2000
MERCURY TRANSPORT AND FATE IN WATERSHEDS
In December 1997, EPA's Mercury Study
Report to Congress identified mercury as a critical
human health and environmental problem needing
additional scientific and technical research. Mercury
poses risks to humans and wildlife, particularly
because it concentrates in the tissues of animals as
it moves up through a food chain. Observed ad-
verse effects in mammals, fish and birds have
included behavioral and neurological abnormalities,
impaired growth and development, reproductive
abnormalities such as fetal deformities, and in some
Sources & Paths of Mercury in the Environment
cases complete reproductive failure. Fish consump-
tion is the dominant exposure pathway for humans and
wildlife. In some cases, mostly in the past, wildlife
were killed by extreme environmental concentrations
of mercury, for example from seed grains treated with
mercury, or from unusally severe instances of waste
release. But the most pervasive wildlife effects involve
reduced breeding success, which now poses severe
consequences for water birds throughout North
America and some endangered species, including
panthers, in the Florida Everglades.
OOON00002
Methyl Mercury
in Fish (CHJHg
-------
More U.S. waters are closed
to fishing because of mercury
contamination than any other toxic
contamination problem. Because
fish consumption can be a signifi-
cant route of human exposure to
toxic chemicals, the U.S. Food and
Drug Administration (FDA) issues
guidelines to states on fish con-
tamination levels that warrant
closures of fishing waters to
commercial or recreational fishing.
The principal sources of fish
contamination, perhaps surpris-
ingly, are air emissions of mercury
from coal burning
power plants, munici-
pal waste incinerators
and other industrial
sources. The emitted
mercury compounds
readily settle, either
directly on water, or on
land from which they
runoff to water bodies.
Additional significant
amounts reach surface
and ground water due to
urban runoff and leaching from
mines and waste disposal sites.
Mercury can also be directly
discharged in inadequately treated
municipal or industrial wastewa-
ters. A number of biological and
chemical processes that are
partially understood occur in soils,
water bodies and sediments that
cause mercury to react with or-
ganic matter to form methyl mer-
cury (MeHg), the most toxic form of
Hg, which is readily taken up by
plants and animals. A principal
focus of the research studies
described in this report is to
achieve a better understanding of
the rates at which formation or
transformation of such compounds
occurs in watersheds and
waterbodies and how these are
affected by releases of various
forms of mercury from natural
sources, from residues left by
historic human activities, and from
current human activities.
Many human sources of
mercury have been eliminated or
much reduced by laws and regula-
tions. Its use in paints and pesti-
cides has been banned, and
industrial and municipal emissions
have been greatly curtailed through
process changes and recovery
technologies. However, some
major human uses continue,
leading to on-going emissions.
Mercury compounds left
by past industrial
activities and mining
also continue to
cycle through
land and water
ecosystems. In
addition, there
significant
atun^l sources.
Mercury's various chernical forms
differ in tdxicity. It is readily
crfanged from one form to another
in passing from air to land to
surface waters, wetlands and
sediments, and then released back
again to the atmosphere. It is a
highly challenging problem for
environmental managers and
scientists to understand the poten-
tial risks posed by mercury to
wildlife and humans, and to deter-
mine what regulatory and non-
regulatory approaches will most
effectively eliminate those risks.
STAR MERCURY
TRANSPORT AND
FATE RESEARCH
PROJECTS
In 1999, EPA's STAR program
funded a set of studies to develop
a better understanding of
mercury's terrestrial and aquatic
fate and transformation processes
that influence ecological and
human exposures. In several
projects, there is a particular focus
on microbial processes, which are
at present poorly understood.
Another area of focus is model
development. Improved models
are needed to predict ecosystem
responses to changes in anthropo-
genic inputs of mercury that could
result from various emission
reduction options and scenarios.
Whole-Ecosystem
Studies
A research team from Syra-
cuse University, Cornell Univer-
sity and Smith College is studying
chemical and biological control of
mercury cycling in upland, wetland
and lake ecosystems in the
Adirondack region of New York.
Many Adirondack lakes have
elevated mercury concentrations in
water and fish tissue, and health
advisories warning against fresh-
water fish consumption have been
mandated throughout the region.
The concentration of mercury
in fish tissue has been found to be
higher in more acidic water bodies.
High acidity can occur naturally,
but is exacerbated by acid deposi-
tion from coal-burning power
-------
plants, mine drainage and other
pollution sources. Conversely,
high levels of dissolved organic
carbon, resulting from the decay of
organic matter such as plants, can
decrease mercury uptake by fish.
Consequently, lakes that receive
substantial drainage from wetlands
are often characterized by rela-
tively low levels of mercury in fish.
Recent lake sediment studies in
the Northeast have shown more
than threefold increases in mercury
deposition since 1850, suggesting
the degree to which industrial
pollution has contributed to re-
gional mercury contamination.
Moreover, for reasons that are not
yet understood, these studies
suggest that watershed retention of
atmospheric mercury has de-
creased markedly over the last 60
years, from 95% retention in the
1930s to 75% retention today.
The Syracuse-led team is gather-
ing data on transport and transfor-
mation patterns of mercury com-
pounds in an Adirondack forest,
wetlands and surface water. They
will focus on mechanisms control-
ling MeHg levels and transport in
wetland pore waters and open
waters. Based on these findings
and other current data, and on
historic patterns of mercury trans-
port and fate, they will develop a
lake/watershed mercury cycling
model. This will be used to predict
impacts on ecosystem mercury
levels that might result from emis-
Minnesota
Illi
Pennsylvania
Great Lakes Watershed
sion reductions in the Adirondack
region.
The opportunity to fully track
and quantify MeHg production,
transport and fate in two large
ecosystems is offered by a study in
the Experimental Lakes Area of
northwestern Ontario, and by
extensive work in the Florida
Everglades. An international
research team from the University
of Maryland's Chesapeake
Biological Laboratory, the Acad-
emy of Natural Sciences Estua-
rine Research Center, and the
Freshwater Institute of the
Canada Department of Fisheries
and Oceans is conducting com-
parative studies on these large
systems. They are assessing
mercury methylation in a catch-
ment (large watershed) that in-
cludes uplands, wetlands and a
lake. This STAR-funded work is
one portion of a larger scale,
whole-ecosystem experiment in
which all mercury fate, transport,
transformation and uptake, from
atmospheric loading through fish
mercury concentrations, is being
traced. The STAR study compo-
nent focuses on how MeHg
changes in response to changes in
mercury loading. Stable isotope
tracing methods will be used to
measure accumulation, production
and degradation rates of MeHg in
upland and wetland soils and lake
sediments, together with key
biogeochemical parameters that
affect mercury methylation and
bioavailability. The tracer method
will allow them to track the fate of
"new" mercury deposits vs. "old"
mercury stored in sediments and
soils, and to track the bioavailability
-------
of new mercury over time across
the components of the watershed.
The Experimental Lakes Area
(ELA) is representative of north
central and northeastern US
ecosystems affected by mercury.
To provide a comparison in a very
different ecosystem and region,
this team will conduct parallel
mercury tracing studies in the
Florida Everglades. Here, mercury
addition experiments will be
conducted within wetland enclo-
sures. Additional field work,
together with transport and fate
data from other Everglades stud-
ies, will support estimates of MeHg
fate in this warm, highly organically
enriched system. Taken together,
these two studies will provide data
that bracket the most likely condi-
tions in most U.S. watersheds
impaired by mercury pollution and
provide information on regional and
landscape variability in methylation
and subsequent mercury accumu-
lation in fish. Results should
provide the ability to predict
changes in levels of various forms
of mercury in the environment
caused by reduced mercury
emissions.
Fish consumption advisories
have been issued for Lake Supe-
rior due to mercury contamination.
Transport, fate and bioavailability
of mercury in the whole Lake
Superior Basin ecosystem are
being studied by a team from the
University of Wisconsin at
Madison, Lake Superior State
University, and the Wisconsin
Department of Natural Re-
sources. Preliminary studies of
several rivers in the area have
found that wetland- and forest-
dominated watersheds attenuate
total mercury transport in the
Basin. And, very importantly, these
watershed types enhance the
production of the MeHg, the most
bioaccumulative, toxic form of
mercury. Preliminary estimates
suggest that while air deposition
provides about 75% of the total
mercury input, inputs from water-
sheds provide more than 80% of
the MeHg. Furthermore, while
atmospheric inputs are distributed
across the vast surface area of the
Lake, watershed influences are
concentrated in certain nearshore
zones. The specific physical and
chemical factors regulating
bioavailability are not known. For
instance, perhaps the MeHg
produced in some components of
watersheds is largely unavailable
for bioconcentration due to its
association with certain high
molecular weight organic materials,
or specific soil types. This study
will investigate sites where it is
hypothesized that enhanced
bioaccumulation may occur,
including sub-watersheds, river-
lake mixing zones and open
waters. Field work will be supple-
mented with lab studies of pro-
cesses influencing fate, transport
and bioavailability of MeHg and
total mercury under similar condi-
tions. Results will be incorporated
into models predicting the influence
of each part of the ecosystem on
the ultimate delivery of the more
toxic and bioaccumulative forms of
mercury to the Lake's waters and
aquatic life.
-------
The Re-emission
Problem
Recent national mercury
emissions inventories estimate that
re-emission of previously deposited
mercury from historic human
sources is similar in magnitude to
industrial emissions. U.S. indus-
trial emission is approximately 240
tons per year, while estimates of
re-emission from soils, sediments
and terrestrial and aquatic vegeta-
tion range from 150 to 600 tons per
year. The role of soils, water
bodies and vegetation as mercury
sources will affect the success of
any program aimed at reducing
mercury loading in and from
watersheds. Processes that
convert mercury to the elemental
form, which is highly volatile, also
diminish its availability for methyla-
tion and uptake by living organ-
isms. The University of Michigan
is examining an important hypoth-
esis: that the processes through
which monovalent and divalent
mercury are converted to the
elemental form are enhanced by
sunlight. If true, it indicates that
when there is strong light penetra-
tion into water bodies, the monova-
lent and divalent forms of mercury
associated with the organic matter
abundant in wetlands and some
lakes may be preferentially con-
verted to readily released elemen-
tal mercury. It would thus be far
less likely to form the bioavailable
MeHg that accumulates in food
chains. Lab and shipboard experi-
ments will assess the following and
other factors in the Saginaw Bay/
Lake Huron system: 1) what
degradation products, containing
what forms of mercury, are created
when locally typical organic matter
decays in different amounts of
light?; 2) what intracellular or
extracellular pathways in organ-
isms can result in methylation of
these compounds?; 3) whether
conversion of monovalent and
divalent mercury to its elemental
form is activated in solid phases in
soils and wetlands; and 4) what
effect does oxidation of elemental
mercury have on its tendency to be
released? Results will help distin-
guish which components of the
watershed are the most likely
souces of mercury re emission.
For example, perhaps clear lakes
could be identified as less serious
sources of mercury re-emission
than wetlands with murky waters,
and thus with less light-induced
conversion of mercury into the less
bioavailable forms. This will help in
modeling likely successes of
decreasing mercury contamination
in fish and other organisms for
given watershed if new mercury
emissions are decreased.
Emission from Water
to Air
The natural chemical pro-
cesses of reduction and oxidation,
called the "redox" cycle, control the
proportions of mercury present in
various forms. This in turn affects
rates of mercury exchange be-
tween water bodies and the atmo-
sphere. Princeton University is
performing a comprehensive
analysis of all components of this
cycle. Some water bodies release
mercury in large amounts to air,
while others do not, resulting in
major differences in the amounts of
mercury in the water, sediments
and organisms. A series of lab and
field experiments will study the
chemical and biological redox
mechanisms that transform mer-
cury between its divalent form,
Hg(ll), and the volatile elemental
form, Hg(0). The forms differ in the
electron charges of the mercury
atoms. The experiments test three
complementary hypotheses: 1)
biological reduction of Hg(ll) to
Hg(0) normally occurs through a
two-electron transfer reaction
mediated by certain enzymes in
two groups of photosynthetic
microorganisms, phytoplankton
and cyanobacteria; 2) chemical
reduction of Hg(ll) occurs in two
distinct one-electron transfer
reactions, first the reduction of
Hg(ll) to Hg(l) which requires a
high energy reducing agent (typi-
cally formed in light) such as the
superoxide anion or an organic
radical (probably a semiquinone),
followed by the reduction of Hg(l)
to Hg(0) by organic matter typically
present in natural waters; and 3)
oxidation of elemental mercury
requires first oxidation of Hg(0) to
Hg(l), likely caused by the same
radicals, superoxide or
semiquinones, and then oxidation
of Hg(l) to Hg(ll) by oxygen, which
is facilitated by chloride complex-
ation of the ionic mercury species.
Starting with simple modeled
systems and building up to com-
plex lab situations more like those
of natural waters, the lab experi-
ments are designed to establish the
-------
mechanisms and rates of these
processes and to provide tech-
niques and probes for field work.
Field experiments will include a
range of values for the key param-
eters, and are designed to establish
the actual occurrence of the mecha-
nisms in nature and to provide data
on process rates. Results should
let us understand how factors
including solar irradiation, dissolved
organic matter concentration, pH
and biological productivity control
rates of mercury loss from water
bodies to the atmosphere, and
hence why water bodies with similar
mercury inputs end up with different
mercury levels in water, sediments
and fish.
Sulfur-Mediated
Microbial Effects
Sulfur, in all its chemical
forms, plays important roles in the
methylation of mercury by naturally
occuring sediment and soil bacteria.
Researchers from the University of
Maryland's Chesapeake Biologi-
cal Laboratory and the Academy
of Natural Sciences Esturaine
Research Center are using lab and
field studies to investigate the
hypothesis that high levels of sulfide
(a reduced form of sulfur) diminish
methylation by lessening the degree
to which mercury is taken up by
methylating bacteria. The scientists
also hypothesize that when there is
little sulfur present as sulfate (an
oxidized form), little methylation
occurs because many methylating
bacteria need sulfate as a nutrient.
Also, they will investigate the role of
some sulfate-using bacteria in the
converse process of demethylation,
together with assessing rates under
various conditions of abiotic
demethylation, a purely chemical
process with no bacteria or other
organisms involved. Results will
provide a comprehensive picture of
the soil and sediment processes
leading to greater or lesser produc-
tion of the more toxic and
bioaccumulative forms of mercury.
Mercury from Mine
Wastes
Mercury mining has left a
legacy of contaminated wastes
throughout the mining belts of the
United States. The problem is
compounded by weathering, which
transforms the mercury into addi-
tional chemical forms, redistributes
it in watersheds and through the air,
and may increase bioavailability to
organisms. Mine wastes can have
significant regional environmental
impacts, since airborne mercury
sometimes reaches uncontami-
nated ecosystems far from point
sources. Stanford University and
the University of Nevada at Reno
are studying the physical and
chemical processes that control
chemical transformation, release
and distribution of mercury from
mine wastes, with the overall
objective of assessing risks to local
and regional ecosystems.
The researchers will determine
the mix of mercury compounds in
environmental samples, and use lab
experiments to examine sorption
processes of mercury on mineral
particles typical of sediments
downstream from mine sites,
transport on colloidal particles, and
the effects of sulfate and chloride
on sorption processes. They will
also monitor atmospheric emissions
from mine waste sites representing
different weathering and climatic
regimes, and compare emission
levels of each type of mercury with
the distribution of types (the "spe-
ciation") in the mine wastes. This
will test the hypothesis that emis-
sions of mercury to the atmosphere
are closely linked to mercury
speciation in mine wastes, and are
influenced by the amount of avail-
able elemental mercury, the other
mineral components of the wastes,
and weather. Based on these
results, they will develop an area
emission model by incorporating the
speciation data into a Geographic
Information System. With this they
will "scale up," to estimate mercury
emissions over entire contaminated
mine sites. If this research in-
creases our understanding of the
contributions of various mercury
sources to local and regional
mercury budgets, results will help
assess whether particular control
options for various point sources
will be effective in reducing local
and regional risks. Results will also
provide information useful in global
mercury modeling efforts.
-------
Mercury in a Coastal
Environment
When polluted freshwater
mixes with saline coastal water
chemical processes typically lead to
precipitation (settling out) of many
toxic pollutants, including mercury
compounds. Estuarine water
bodies are thus major repositories
for riverborne/watershed derived
mercury. The University of Con-
necticut is studying microbiological,
chemical and physical aspects of
mercury cycling in Long Island
Sound. For decades the Sound and
its river tributaries served as the
receiving water for discharges
associated with older industrial
facilities. Many of these facilities
no longer exist, but in some cases a
legacy of contamination remains,
particularly in sediments associated
with freshwater and saltwater
interfaces. Preliminary work has
found large emissions of elemental
mercury from the waters of the
Sound to the atmosphere. At
present it is not known how rates of
historic and new mercury inputs and
deposition compare to rates at
which it is being released to the air.
This research team will test the
following hypotheses: 1) that the
principal influences on mercury
distributions are the the amount that
is present from place to place in
bioavailable forms and the levels of
chemical reducing agents, bacterial
activity and solar radiation from
place to place; (2) that organic
matter production, reflected in the
difference between net primary
production and bacterial consump-
tion, can serve as a surrogate for
the amount of active organic
substances that influence the
biogeochemical fate of mercury; (3)
that estuarine reactions (i.e., mixing
of riverborne mercury with seawater
high in chlorine and positively
charged ions) increase the propor-
tion of bioavailable mercury avail-
able for reduction; 4) that direct
sewage discharge into contained
brackish waters leads to enhanced
localized production of elemental
mercury; (5) given the availability of
reactive mercury species, and the
limited light penetration in coastal
waters, elemental mercury is the
predominant mercury cycling
product of bacterial activity in the
zone of well oxygenated waters;
and 6) net synthesis of MeHg is
most significant in redox "transition
zones", primarily the shallower
sediments and waters in basins that
experience seasonal low oxygen
conditions (i.e., the western and
central Sound). Results of these
studies will be incorporated into a
biogeochemical mercury cycling
and mixing model, aspects of which
will be relevant to other coastal
regions. Such a model can be used
to predict water quality and fish
mercury levels that may result from
alternate pollution abatement
strategies.
Historic Analysis of
Sources and
Environmental
Conditions
Efforts to reduce mercury
emissions tend to have been based
on the assumption that in most
water bodies, particularly lakes,
mercury accumulation in fish would
be more or less proportional to
atmospheric deposition. Given the
strong evidence that mercury
deposition is now about 3 to 4 times
greater than natural rates, a first
estimate would be that fish should
be 3 to 4 times more contaminated
than natural levels. But a team of
Minnesota researchers has deter-
mined, based on comparing modern
fish to fish preserved in the 1930s,
that mercury concentrations in fish
may have increased by a factor of
10 in the low alkalinity lakes that are
common in their region. It is
possible that enhanced
bioaccumulation could be caused
by changes in the biological com-
munity. For example, there could be
-------
higher levels in food chains, and
thus more biomagnification, if there
have been changing prey abun-
dances. However, recent sediment
cores from sixteen Minnesota lakes
demonstrate that, beginning about
1940, MeHg was an increasingly
large percentage of the total. This
suggests a fundamental change in
methylation efficiency, weakening
support for the competing hypoth-
esis of longer food chains.
Researchers from the Univer-
sity of Minnesota, the Minnesota
Pollution Control Agency and the
Science Museum of Minnesota
are investigating the anthropogenic
ecosystem changes that have
become widespread since World
War II to assess which are the most
likely contributors to enhanced
methylation. The changes include
increased sulfate deposition,
increased nutrient loads and
increased loading of mercury to
wetlands because of soil distur-
bance. The study will establish the
relative importance of atmospheric,
in-lake, and various wetland
sources of MeHg to a lake in a
forested watershed. They will
determine mercury retention and
release in different wetland types
and conduct experiments to deter-
mine which methylation-enhancing
processes are occuring. Finally,
they will add a hydrologically-based
wetland geographic information
system module to a widely used
Mercury Cycling Model, so the
findings can be applied to other
lakes. The model will be, among
other applications, relevant to
states' needs to develop the total
maximum daily load estimates
required to support water quality
protection actions.
Get online for more
detailed information
www.epa.gov/ncerqa
G. E. Brown, Jr., T. R. Ireland, D.
Grolimund, C. S. Kim (Stanford University,
CA); M. S. Gustin (University of Nevada-
Reno (NV)); J. J. Rytuba (collaborator,
U.S. Geological Survey, Menlo Park (CA).
Processes Controlling the Chemical/
Isotopic Speciation and Distribution of
Mercury from Contaminated Mine Sites.
C. T. Driscoll (Syracuse University, NY); J.
Yavitt (Cornell University, NY); R. Newton
(Smith College, MA); R. Munson
(TetraTech lnc.,UT). Chemical and
Biological Control of Mercury Cycling in
Upland, Wetland, and Lake Ecosystems in
the Northeastern U.S.
W. F. Fitzgerald, P.T. Visscher (University
of Connecticut, CT). Microbiological and
Physicochemical Aspects of Mercury
Cycling in the Coastal/Estuarine Waters of
Long Island Sound and Its River-Seawater
Mixing Zones
C. C. Gilmour, A. Heyes (The Academy of
Natural Sciences Estuarine Research
Center, MD); R. P. Mason (University of
Maryland, Chesapeake Biological
Laboratory); J. M. Rudd (collaborator,
Canada Department of Fisheries and
Oceans, Freshwater Institute). Re-
sponse of Methylmercury Production and
Accumulation to Changes in Hg Loading:
A Whole-ecosystem Mercury Loading
Study
J. P. Hurley (University of Wisconsin at
Madison, and Wisconsin Department of
Natural Resources, Wl); D.E. Armstrong,
M.M. Shafer (University of Wisconsin at
Madison, Wl); Richard C. Back (Lake
Superior State University, Ml). Water-
shed Influences on Transport, Fate, and
Bioavailability of Mercury in Lake
Superior
R. P. Mason (University of Maryland,
Chesapeake Biological Laboratory) and
C. C. Gilmour (TheAcademy of Natural
Sciences Estuarine Research Center,
MD). Understanding the Role of Sulfur in
the Production and Fate of Methylmer-
cury in Watersheds
F. M. M. Morel (Princeton University, NJ).
The Redox Cycle of Mercury in Natural
Waters
J. O. Nriagu, G. Keeler, J. Lehman
(University of Michigan, Ml); S. Lindberg,
Hong Zhang (collaborators, Oak Ridge
National Laboratory). Photo Induced
Reduction of Mercury in Lakes, Wet-
lands, and Soils
E. B. Swain, J. Jeremiason (Minnesota
Pollution Control Agency, MN); P. L.
Brezonik, E.Nater, J.Cotner (University
of Minnesota, MN); D. Engstrom, J.
Almendinger (Science Museum of
Minnesota, MN), Reed Harris (TetraTech
Inc., CA). Methylmercury Sources to
Lakes in Forested Watersheds: Has
Enhanced Methylation Increased
Mercury in Fish Relative to Atmospheric
Deposition?
------- |