The Science of the Total Environment 259 2000.

6171

Assessing the contribution of natural sources to

regional atmospheric mercury budgets
U

Mae Sexauer Gustina, , Steven E. Lindbergb, Kenneth Austina ,

Mark Coolbaugha, Alan Vettea , Hong Zhangb
a

Department of En ironmental and Resource Sciences MS 370, Uni ersity of Ne ada-Reno, Reno NV 89557, USA
b
En ironmental Sciences Di ision, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6038, USA
Received 27 September 1999; accepted 8 March 2000

Abstract
Naturally mercury-enriched substrate is a long-lived source of mercury to the global atmospheric mercury cycle.
Field flux chambers, laboratory gas exchange chambers and micrometeorological methods may be applied to estimate
emissions from these sources. However, field chamber experimental design may affect the magnitude of the fluxes
measured, and the laboratory chamber only provides a minimum estimate of flux. Many factors, such as mercury
concentration and speciation in substrate, light, precipitation, and temperature, influence the emission of mercury
from the substrate. Mercury concentration in the substrate is a dominant factor controlling emissions and may be
used to predict emissions from regions of mercury enrichment. Mercury fluxes measured from three areas of natural
enrichment and three areas with low levels of mercury enrichment are 15 orders of magnitude greater than the
value applied to global belts of natural enrichment. Preliminary scaling of emissions from one of these areas and for
western North America indicates that mercury enriched areas may be significant sources of mercury to the
atmosphere, and that their contribution to regional and global atmospheric budgets needs to be reassessed. 2000
Elsevier Science B.V. All rights reserved.
Keywords: Mercury; Atmosphere; Emmisions

Corresponding author. Tel.: q1-775-784-4203; fax: q1-775-784-4789.

E-mail address: msg@scs.unr.edu M.S. Gustin..
0048-9697r00r$ - see front matter 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 8 - 9 6 9 7 0 0 . 0 0 5 5 6 - 8

62

M.S. Gustin et al. r The Science of the Total En ironment 259 (2000) 6171

1. Introduction
Areas which are geologically or naturally enriched in mercury Hg. are concentrated globally
in association with plate tectonic boundaries, recent volcanism, precious and base metal mineralization, and high crustal heat flow. Although this
manuscript focuses on natural source emissions in
North America, those geologic settings where Hg
is enriched in substrate are distributed globally in
three broad belts. One belt extends from the
southern tip of South America along the western
edges of South, Central and North America to
Alaska. This belt then swings westward to northeast Asia and bifurcates north of Japan. One
splay follows the island arcs through Japan, the
Philippines and Indonesia to New Zealand
and eastern Australia. The second follows the
Himalaya and Alpine Belts to terminate in Spain
Pennington, 1959.. Recent work has documented
the deposition of mercury in sea-floor hydrothermal settings Stoffers et al., 1999. indicating that
mid-ocean ridge systems may be important sources
of mercury to the oceans. Areas containing important commercial concentrations of Hg are
found in China, Italy, Mexico, the Philippines,
Peru, Spain, the Soviet Union, the USA and
Yugoslavia. The worlds most important deposits
of mercury are found in the Mediterranean region, i.e. Almaden, Spain, Monte Amiata, Italy
and Idria, Slovenia Bailey et al., 1973.. In addition, some sedimentary rocks, especially shales
and those containing organic materials, may be
enriched in Hg Turekian and Wedepohl, 1961;
Marowsky and Wedepohl, 1971; Rasmussen et al.,
1998.. We consider the substrate to be enriched
in Hg when it contains ) 0.1 g gy1 of Hg. This
concentration was chosen because ; 0.06 g gy1
represents the average upper crustal abundance
of Hg Wedepohl, 1995. and ; 0.08 g gy1 , the
average concentration of Hg in igneous rocks
Turekian and Wedepohl, 1961; Marowsky and
Wedepohl, 1971.. It has long been recognized
that Hg-enriched substrate emits Hg vapor to the
air cf. Kothny, 1973.. Because of this, and the
association of Hg with precious and base metal
deposits, elevated concentrations of Hg vapor in
air have been used as an exploration tool cf.

Hawkes and Williston, 1962; Williston, 1964; McCarthy, 1972; Varenkamp, 1983; Klussman, 1993..
Mercury is considered a global pollutant because of its potential for long-range transport via
the atmospheric pathway and deposition in
ecosystems remote from anthropogenic Hg
sources. In contrast to anthropogenic point
sources of atmospheric Hg, which are short lived
and emit a high percentage of particulate and
reactive gaseous Hg, natural sources are long
lived ) 10 4 years. and emit primarily elemental
Hg. The latter form of Hg is considered to enter
the global atmospheric Hg pool, whereas particulate Hg and reactive gaseous Hg are deposited
locally and regionally. Estimates of the contribution of Hg from natural sources to the atmosphere have varied widely cf. Lantzy and
Mackenzie, 1979; Watson, 1979; Varenkamp and
Buseck, 1986; Nriagu, 1989; Lindqvist et al., 1991..
In developing recent Hg emission inventories,
most researchers have considered the natural
source term to have contributed a constant
amount of Hg to the atmosphere overtime. A rate
equivalent to ; 1.5 ng my2 hy1 is now widely
used to represent large scale average emissions
from global belts of natural Hg enrichment cf.
Lindqvist et al., 1991; Mason et al., 1994.. This
value has often been derived from the difference
between global anthropogenic emissions and deposition Nriagu and Pacyna, 1988; Nriagu, 1989;
Lindqvist et al., 1991; Mason et al., 1994., neither
of which are well constrained cf. Porcella et al.,
1995; Pai et al., 1998.. Recent measurements of
Hg emissions from naturally enriched areas e.g.
Ferrara et al., 1998; Rasmussen et al., 1998;
Gustin et al., 1999; this paper. suggest that the
contribution of Hg from natural sources to the
regional Hg budgets may be underestimated. In
addition, there are no reliable data to demonstrate that natural source emissions have remained constant over time. Given that the disturbance of naturally enriched areas through
processes such as mining and land development
has resulted in an increase of the exposed surface
area of naturally enriched substrate over time, it
is unlikely that emissions have remained constant.
Emissions from natural sources need to be
better constrained in order to assess the effec-

M.S. Gustin et al. r The Science of the Total En ironment 259 (2000) 6171

tiveness of regulations and controls for Hg emissions imposed on point sources, and to develop a
complete understanding of the biogeochemical
cycling of Hg in the environment. In order to
develop regional estimates of Hg emissions from
naturally enriched landscapes we must: 1. determine whether methods applied to constrain
emissions from naturally enriched terrains obtain
representative data; 2. determine those environmental factors most important in controlling natural emissions; 3. develop a database of emissions from representative landscapes; and 4. develop a means of scaling up Hg fluxes measured
in situ to large areas. We describe here our
assessment of the current state of these research
needs.

2. Methods assessment
The methods that have been used to measure
Hg emissions in situ include field flux chambers
cf. Xiao et al., 1991; Kim and Lindberg, 1995;
Carpi and Lindberg, 1997; Gustin et al., 1999.
and micrometeorological methods Lindberg et
al., 1995; Poissant and Casimir, 1998; Gustin et
al., 1999.. In addition, laboratory gas exchange
chambers may be used to provide a minimum
estimate of in situ flux from substrate Gustin et
al., 1999a.. These methods were recently compared as part of the Nevada SToRMS mercury
flux methods intercomparison held in Reno,
Nevada cf. Gustin, 1998; Gustin et al., 1999;
Lindberg et al., 1999; Poissant et al., 1999;
Wallschlaeger et al., 1999.. This study demonstrated that micrometeorological methods provide
a useful areal average of Hg fluxes from naturally
enriched areas. However, because the terrain associated with naturally enriched areas is often
rugged and complex, micrometeorological methods may not generally be applicable, and more
portable field chambers must be utilized. Intercomparison of Hg fluxes obtained using these two
methods during Nevada SToRMS demonstrated
that the mean Hg emissions obtained with field
chambers underestimated the flux derived from
micrometeorological data by approximately a factor of 3 Gustin et al., 1999.. Recent research has

63

demonstrated that chamber design and air

turnover rates can significantly influence emissions Gustin et al., 1999; Wallschlaeger et al.,
1999.. These findings have led to the development
of an optimized field chamber that measures
emissions that are more representative of those
occurring from an unconstrained surface, and also
to a rough correction factor that may be applied
to data obtained with chambers which utilized a
lower turnover rate.

3. Assessment of factors controlling emissions

In order to be able to scale up emissions from
naturally enriched substrate we must understand
those factors controlling emissions. Factors that
have been demonstrated to influence Hg emissions from substrate include temperature Lindberg et al., 1979, 1995; Gustin et al., 1997,
1998a,b., Hg speciation and concentration Gustin
et al., 1998a; Rasmussen et al. 1998; Zhang and
Lindberg, 1999., incident radiation Carpi and
Lindberg, 1997; Gustin et al., 1998a., turbulence
Poissant et al., 1999., soil gas concentrations
Ebinghaus et al., 2000. and precipitation Lindberg et al., 1999.. The relative importance of
these various factors is not clearly understood.
However, data collected from a wide variety of
substrates indicate that substrate Hg concentration is a dominant parameter controlling flux
from naturally enriched areas. Fig. 1 shows Hg
fluxes collected from 30 locations in western USA
and Canada using a variety of field chamber designs. If the fluxes are adjusted to account for
differences in chamber turnover time, the log
transformed regression coefficient improves from
0.67 to 0.87, while the slope and intercept remain
essentially unchanged. This plot suggests that for
a wide range of substrate Hg concentrations, this
may be a practical parameter to use to begin to
scale up in situ derived Hg- fluxes to make regional emission estimates. Rasmussen et al. 1998.
reported a similar correlation between the average flux and total Hg concentration in the substrate. These data, along with some additional data
from Canadian natural sources, have been included in Fig. 1 and fall nicely along the line

64

M.S. Gustin et al. r The Science of the Total En ironment 259 (2000) 6171

Fig. 1. Scatter plot showing the logarithm of Hg flux vs. the logarithm of substrate mercury concentration. Mercury fluxes were
measured using a variety of field chamber designs.

developed from our data. In addition to our field

data, the controlled environment of a laboratory
gas exchange chamber was used to demonstrate
that flux in the dark was strongly correlated with
substrate Hg concentrations r2 s 0.997. whereas
in the light, the correlation was far weaker r2 s
0.41. Gustin et al., 1998a.. This demonstrates
that with the addition of other environmental
parameters in a controlled setting the degree to
which soil Hg concentrations control emissions
may vary.

4. Measurement of emissions from representative

areas
Areas of natural enrichment of Hg can be
subdivided into two basic types, those with Hg
concentrated in sufficient quantity for mining or
prospecting maximum ) 50 g gy1 ., and those
with lower concentrations 0.1 to ; 50 g gy1 .
which are distributed over broad areas. The latter
includes areas of enrichment associated with base
and precious metal mineralization, volcanism, and
hydrothermal or geothermal processes associated
with high crustal heat flow. To develop an estimate of emissions from these areas, an assessment of rock types, geologic structures, and gen-

eral distribution of Hg in association with these

features needs to be made.
In the western USA there are two major geologic settings in which mercury mineralization,
especially ore grade, is found: one contains silicacarbonate alteration of mafic volcanic rocks
that is attributed to regional metamorphism; and
the other is associated with epithermal hydrothermal systems in which Hg is deposited in veins and
hot spring systems. We have begun to characterize emissions from three areas representing these
types of deposits; the New Idria Mining District,
CA, which was the second largest Hg producing
district in the USA a silicacarbonate type deposit.; the Ivanhoe District, NV, an epithermal
hot spring deposit formed ; 15 million years ago;
and the Steamboat Springs Geothermal Area,
NV, an active hot spring system Fig. 2.. Fig. 3
summarizes Hg emissions measured from these
three locations using optimized field chambers
developed as a result of the Nevada SToRMS
project. The New Idria data were collected at
three sites remote from where any mining had
occurred in the district. The Ivanhoe data were
collected from the Silver Cloud ore zone, which is
dense siliceous sinter containing cinnabar, from
mine tailings or waste rock, and from desert soils

M.S. Gustin et al. r The Science of the Total En ironment 259 (2000) 6171

65

Fig. 2. Map showing the field sites discussed in this manuscript. The Canadian locations are the research sites of Rasmussen et al.
1998.. This figure is from Gustin et al. 1999a..

Fig. 3. Mercury emissions measured from three areas where a range of concentration of mercury enrichment in substrate is
present. Numbers at the end of the bars represent the mercury concentration in substrate at each site in micrograms per gram.

66

M.S. Gustin et al. r The Science of the Total En ironment 259 (2000) 6171

Fig. 4. Mercury emissions measured from areas with low levels of mercury enrichment associated with precious metal mineralization Carlin and Comstock. and with organic rich shales Rove Shale.. The estimated flux from global mercuriferous belts is from
Lindqvist et al. 1991. and the data from the Rove Shale is from Rasmussen et al. 1998..

surrounding the area of mining and ore processing all within an ; 1-km2 area.. The Steamboat
Springs data includes an average emission from
the Nevada SToRMS field site, part of an undisturbed area of low Hg-enrichment surrounding an
area where Hg was mined. Mercury flux was also
measured at the Steamboat Mine site, which has
been anthropogenically disturbed and altered.
Fig. 4 shows emissions measured from areas
with low levels of natural enrichment, which cover
far more surface area than those zones of high
Hg enrichment. The average flux estimated for
mercuriferous belts in global models is shown for
comparison. The Carlin Trend and Comstock ore
zones of Nevada have Hg enrichment associated
with gold and silver mineralization. The Proterozoic Rove Shale, that outcrops in the USA and
Canada, covers 300 km2 . Rasmussen et al. 1998.
measured a flux of 35 ng my1 hy1 from this unit.

5. Scaling up
In order to derive estimates of annual Hg flux
for a given district or region, individual flux measurements must be integrated over space and
time. This process is complex because of the large

number of field variables affecting surface-to-air

Hg flux, including substrate and vegetation type,
the Hg concentration and speciation, the presence of anthropogenic disturbance, and environmental conditions such as temperature, light, and
precipitation. Geographic Information Systems
GIS. is a tool in which all external variables
controlling flux along with their spatial and temporal relationships can be modeled to obtain an
average flux for an area.
To begin to scale up emissions from naturally
enriched areas, the New Idria mining district was
chosen because of our established database for
both Hg concentrations of substrate and measured in situ Hg fluxes. The geologic map of
Eckel and Myers 1946. was used as the working
map base and the areal extent was 190 km2 . The
New Idria district is underlain by Jurassic and
Cretaceous sedimentary rocks of the Franciscan
Group and Panoce Formation that were intruded
by a serpentinite mass or dome Figs. 5 and 6..
Hydrothermal fluids transformed serpentinite
along faults into a silicacarbonate rock and deposited Hg in the silica carbonate altered rocks,
as well as near the contacts of the Panoce Formation and Franciscan Group rocks with the serpentinite. A copy of the geologic map was scanned,

M.S. Gustin et al. r The Science of the Total En ironment 259 (2000) 6171

67

Fig. 5. Geologic map of New Idria mining district modified from Eckel and Myers 1946..

then digitized into the GIS program ArcView.

Polygons defining areas of Hg enrichment were
digitized for each rock type, and average Hg soil
concentrations and Hg fluxes were assigned to
each polygon. At New Idria, Hg flux measurements were taken from areas with no vegetation,
in order to unravel the cause and effect of relationships between Hg-bearing rocks, soils, environmental conditions and Hg emissions. For the
purposes of this study the Hg flux data, most of
which was collected during late summerrearly
fall, are assumed to represent average climatic
and environmental conditions. To account for reduced fluxes at night and during time periods of
cold temperatures in winter, fluxes were calculated for only a 12-h day. To characterize Hg
fluxes, measurements were taken from ; 20 representative locations using a newly developed field

flux chamber with higher turnover rates. These

fluxes were applied to similar geologic units
throughout the district to develop an emission
estimate. An independent emission estimate was
developed using soil Hg concentrations obtained
from the flux measurement locations, data from
other field sites sampled during two field reconnaissance surveys, and unpublished data from
mining company databases, along with the linear
relationship established between soil mercury
concentrations and fluxes Fig. 1.. Using the Hg
flux data, a total annual atmospheric flux of 26 kg
yeary1 or an emission rate of 135 g my2 yeary1
was estimated. Using the soil concentration data,
the annual flux was estimated to be 16 kg yeary1
or 83 g my2 yeary1 . Considering the reconnaissance nature of this initial survey, these two numbers are not considered significantly different.

68

M.S. Gustin et al. r The Science of the Total En ironment 259 (2000) 6171

Fig. 6. Map showing the mercury flux values assigned to the New Idria Mining District based upon in situ flux measurements, rock
types and rock alteration.

Fluxes from this small area would vastly exceed

that expected to be deposited to the area via dry
deposition and the area average of 2030 ng m2 h
is significantly higher than the value applied to
naturally enriched areas in global models.

6. Estimating emissions from western North

America
One of the worlds larger Hg metallogenic
provinces extends from south-western Canada into
Mexico Bailey et al., 1973.. In order to obtain an
estimate of Hg flux from this belt, emissions from
both concentrated sources of Hg and diffuse areas of low levels of enrichment need to be quantified. To begin this process, the contribution from
reported Hg occurrences in western North America was crudely approximated by: 1. compiling

reported Hg occurrences into an ArcInfo GIS

database; 2. ranking them into categories based
on mercury production Table 1.; 3. assigning an
annual Hg flux and surface area to each category;
and 4. calculating the predicted total annual flux
by multiplying the number of deposits by the flux
assigned to each category. Documentation on Hg
districts, mines, prospects, and occurrences was
obtained from a variety of sources, including regional metallogenic maps, state mineral records
and commodity maps, and the USGS Mineral
Resource Data System cf. Jenkins, 1939; Bailey
and Phoenix, 1944; Lee et al., 1987; Mason and
Arndt, 1996.. Geographical areas included in the
study were Nevada, California, Washington, Oregon, Idaho, Utah, Arizona, Alaska, and Mexico.
Annual fluxes were assigned based on area averages being obtained in ongoing scaling exercises.
Prospects and occurrences are in general associ-

M.S. Gustin et al. r The Science of the Total En ironment 259 (2000) 6171

69

Table 1
Parameters used for scaling up mercury emissions from identified areas of mining or prospecting in the western USA
Production
flasks of Hg.

Area applied
km2 .

Emission rate
applied
ng my2 hy1 .

) 1000
1001000
- 100
Prospect or
occurrence
Geothermal
areas

25
20
15
100

30
30
30
10

7
5
4
9

39
11
200
911

100

1259

ated with broad areas of mineralization which

cover larger areas of low levels of mercury enrichment when compared with discrete mercury deposits, thus they were assigned a somewhat larger
area than those areas where Hg was mined.
Geothermal fluids tend to carry high concentrations of mercury in the vapor phase so although
assigned a small surface area of expression their
high concentration of mercury contributes to a
high emission. The total annual flux calculated
for those documented Hg deposits in western
USA and Mexico is estimated to be ; 10 Mg
yeary1 . Estimates on a state-by-state basis ranged
from a high of 2.1 Mg yeary1 for Nevada to a low
of 0.08 Mg yeary1 for Arizona. This scaling exercise is greatly limited by the adequacy of the data
base of located mercury deposits and occurrences, and does not include an assessment of the
contribution from the relatively broad areas of
low levels of Hg enrichment. For example, Arizona does not have a lot of recorded mercury
deposits and occurrences, however, it is a state
with abundant mineralization with which low levels of mercury enrichment are associated. There
are many geothermal areas, precious and base
metal deposits and recent volcanic deposits associated with this part of the world that have low
levels of mercury enrichment and we hypothesize
that emissions from these areas could be even
greater than those estimated for the smaller, but
more enriched sites. More work is needed to
constrain those factors controlling emissions and
emissions from diffuse sources. With the addition
of this information we may better assess the contribution of mercury to the atmosphere from nat-

Annual emission
kg yeary1 .

Number of
reported areas

ural sources. In preliminary scaling of emissions,

natural sources appear to be significant contributors of mercury to the atmosphere. In fact, we
may need to reassess the atmospheric residence
time of mercury and the biogeochemical cycling
of mercury to account for the emissions being
measured.

Acknowledgements
This work was funded by EPRI project manager, Mary Ann Allan. and an EPA-STAR
Grant R825249 project manager, William
Stelz.. The authors thank the participants in the
Nevada SToRMS Project and Cambior Exploration USA, Inc. who provided the database
for Mexican mercury deposits.
References
Bailey E, Clark A, Smith R. United States mineral resources:
mercury. USGS Prof Paper 1973;820:404414.
Bailey E, Phoenix D. Quicksilver deposits in Nevada. Univ
Nevada Bull 1944;38-5:206. Geology and Mining Series no.
41..
Carpi A, Lindberg S. Sunlight-mediated emission of elemental
mercury from soil amended with municipal sewage sludge.
Environ Sci Technol 1997;31:20852091.
Ebinghaus R, Kock H, Hempel M. Determination of mercury
in ambient air using highly time resolved online methods.
Gefahrstoffe-Reinhaltung der Luft 2000, in press.
Eckel E, Myers W. Quicksilver deposits of the new Idria
district San Benito and Fresno counties, California. Calif J
Mines Geol 1946;42:81124.
Ferrara R, Maserti B, Anderson M et al. Atmospheric mer-

70

M.S. Gustin et al. r The Science of the Total En ironment 259 (2000) 6171

cury concentrations and fluxes in the Almaden Mining

District Spain.. Atmos Environ 1998;32:38973904.
Gustin M. Mercury flux measurements: an intercomparison
and assessment. EPRI Palo Alto CA, Report TR 111346,
1998.
Gustin M, Sexauer Lindberg S, Marsik F et al. The Nevada
SToRMS project: measurement of mercury emissions from
naturally enriched landscapes. J Geophys Res Atmos
1999;104D17.:21 83121 844.
Gustin M, Rasmussen P, Edwards G, Schroeder W, Kemp J.
Application of a laboratory gas exchange chamber for
assessment of in situ mercury emissions. J Geophys Res
Atmos 1999a;104D17.:21 87321 878.
Gustin M, Sexauer Maxey RA, Rasmussen P, Beister H.
Mechanisms influencing the volatile loss of mercury from
soil. Symposium Volume Air and Waste Management Association meeting September 1998. Measurement of Air
Toxics and Related Air Pollutants 1998a:224235.
Gustin MS, Maxey RA, Rasmussen P, Biester H. Mechanisms
influencing the volatile loss of mercury from soil. Symposium volume of the Air and Waste Management Association on the Measurement of Toxic and Related Air Pollutants, September 1998b Cary; NC, USA: 224235.
Gustin M, Taylor G, Maxey R. Effect of temperature and air
movement on the flux of elemental mercury from substrate
to the atmosphere. J Geophys Res 1997;102:38913898.
Hawkes H, Williston S. Mercury vapor as a guide to
leadzincsilver deposits. Min Cong J 1962;48:3032.
Jenkins O. Outline geologic map of California showing quicksilver properties. California Division of Mines and Geology,
Economic Mineral Map of California; 391, State of California Dept. of Natural Resources Division of Mines, 1939.
Kim K, Lindberg S. Design and initial test of a dynamic
enclosure chamber for measurements of vapor-phase mercury fluxes over soils. Water Air Soil Pollut 1995;80:
10591068.
Klussman R. Soil gas and related methods for natural resource exploration. Chichester, England: John Wiley and
Sons, 1993:483.
Kothny E. The three-phase equilibrium of mercury in nature.
Trace elements in the environment, Symposium Sponsored
by the Division of Water Air and Waste Chemistry at the
162nd meeting of the ACS, 15 September 1971, Washington, DC, 1973, pp. 4980.
Lantzy R, Mackenzie F. Atmospheric trace metals: global
cycles and relevance to geochemical exploration. Geochim
Cosmochim Acta 1979;43:511525.
Lee M, Guild P, Schruben P. Preliminary metallogenic map of
North America: a listing of deposits by commodity. USGS
Circ 1987;858-C:138.
Lindberg S, Zhang H, Gustin M et al. Increases in mercury
emissions from desert soils in response to rainfall and
irrigation. J Geophys Res Atmos 1999;104 D17 .:
21 87921 888.
Lindberg S, Kim Y, Meyers T, Owens J. A micrometeorological gradient approach for quantifying airrsurface exchange

of mercury vapor: tests over contaminated soils. Environ

Sci Technol 1995;29:126135.
Lindberg S, Jackson D, Huckabee J, Janzen S, Levin S, Lund
J. Atmospheric emission and plant uptake of mercury from
agricultural soils near the Almaden mercury mine. J Environ Qual 1979;8:572578.
Lindqvist O, Johansson K, Aastrup M et al. Mercury in the
Swedish environment: recent research on causes consequences and corrective methods. Water Air and Soil Pollut
1991;55:2332.
Marowsky G, Wedepohl K. General trends in the behavior of
Cd Hg Tl and Bi in some major rock forming processes.
Geochim Cosmochim Acta 1971;35:12551267.
Mason G, Arndt R. Mineral resources data system. USGS
Digital Data Series DDS-0020 1996 available on CD..
Mason R, Fitzgerald W, Morel F. The biogeochemical cycling
of elemental mercury: anthropogenic influences. Geochim
Cosmochim Acta 1994;58:31913198.
McCarthy J. Mercury vapor and other volatile components in
the air as guides to ore deposits. J. Geochem Explor
1972;1:143162.
Nriagu J. A global assessment of natural sources of atmospheric trace metals. Nature 1989;338:4749.
Nriagu J, Pacyna J. Quantitative assessment of worldwide
contamination of air water and soil by trace metals. Nature
1988;333:134139.
Pai P, Heisler S, Joshi A. An emissions inventory for regional
atmospheric modeling of mercury. Water Air Soil Pollut
1998;101:298308.
Pennington JW. Mercury: a materials survey. US Bureau
Mines Circ 1959;7941:92.
Porcella D, Chu M, Allan MA. Inventory of North American
mercury emissions to the atmosphere. In: Baeyens W,
Ebinghaus R, Vasiliev O, editors. Global and regional
mercury cycles: sources fluxes and mass balances NATOASI-Series, 21. Bordrecht, Netherlands: Kluwer Academic
Publishers, 1995:359380.
Poissant L, Pilote M, Casimir A. Mercury flux measurements
in a naturally enriched area: correlation with environmental conditions. J Geophys Res Atmos 1999;104D17.:21
84521 857.
Poissant L, Casimir A. Waterair and soilair exchange rate
of total gaseous mercury measured at background sites.
Atmos Environ 1998;32:883893.
Rasmussen P, Edwards G, Kemp J, Fitzgerald-Hubble C,
Schroeder W. Towards an improved natural sources inventory for mercury. Met Environ: Int Symp: Metallur Soc Can
Inst Min Metall Pet CIM. 1998; 18.
Stoffers P, Hannington M, Wright I, Herzig P, de Ronde C.
Elemental mercury at submarine hydrothermal vents in the
Bay of Plenty, Taupo volcanic zone, New Zealand. Geology
1999;10:931934.
Turekian K, Wedepohl K. Distribution of the elements in
some major units of the earths crust. Bull Geol Soc Am
1961;72:175192.

M.S. Gustin et al. r The Science of the Total En ironment 259 (2000) 6171
Varenkainp J. Hg anomalies in soils: a geochemical exploration method for geothermal areas. Geothermics 1983;
12:2947.
Varenkamp J, Buseck P. Global mercury flux from volcanic
and geothermal processes. Appl Geochem 1986;1:6573.
Watson W. In: Nriagu JO, editor. Economic considerations in
controlling mercury pollution. Biogeochemistry of mercury
in the environment. Amsterdam: Elsevier, 1979:4177.
Wedepohl K. The composition of continental crust. Geochim
Cosmochim Acta 1995;59:12171232.
Williston S. The mercury halo method of exploration. Eng
Min J 1964;165:98101.

71

Wallschlaeger D, Turner R, London J et al. Factors influencing the measurement of mercury emissions from soils measured with field flux chambers. J Geophys Res Atmos
1999;104D17.:21 85921 871.
Xiao Z, Munthe J, Schroeder W, Lindqvist O. Vertical fluxes
of volatile mercury over forest soil and lake surfaces in
Sweden. Tellus 1991;43B:267279.
Zhang H, Lindberg S. Processes influencing the emission of
mercury from soils: a conceptual model. J Geophys Res
Atmos 1999; in press.