PROCEEDINGS OF 1995
CANADIAN MERCURY NETWORK WORKSHOP
[INLINE]
NATURAL SOURCES OF MERCURY: FIELD METHODS FOR CHARACTERIZING MONITORING SITES
P. E. Rasmussen
Geological Survey of Canada, 601 Booth St., Ottawa, Ontario K1A 0E8.
Introduction
An understanding of natural sources and pathways of Hg in the
environment is important for the
* assessment of health risks associated with lithogenic Hg
(Dunnette, 1988),
* evaluation of sites slated for hydroelectric reservoir development
(Rannie and Punter, 1987; Rasmussen, 1993c), and
* measurement of the Hg flux from local geological sources at remote
monitoring sites (Rasmussen, 1994a).
Background Hg concentrations vary widely from place to place
depending on the local geology. To characterize natural sources of
Hg at a given monitoring site, it is necessary to determine what
range of concentrations can be considered "background" in that
setting. It is also necessary to identify areas in the ecosystem
where anomalously high Hg concentrations occur naturally.
Sampling Strategy
In environmental monitoring applications, the strategy known as
"search sampling" is recommended to locate a "hotspot" of elevated
concentration arising from a buried source (Gilbert, 1987). This
strategy, which was adopted from geochemical exploration techniques,
requires
* sampling at regular intervals along lines in a grid pattern,
* sample spacing appropriate for the size of the target, and
* a clear and unambiguous definition of "hotspot" (anomaly).
Examples of the search sampling technique are drawn from a detailed
study of the spatial variation of total Hg concentrations in
vegetation and soil over an area of approximately 150 km2
(Rasmussen, 1993). The study area is located in the southern
Canadian Shield, west of the town of Huntsville, Ontario. Details of
the study area and analytical methodology have been published
previously (Rasmussen et al., 1991; Rasmussen and Gardner, 1992;
Rasmussen, 1994b; 1995).
Survey Design
Locating geological Hg anomalies requires a survey design that
accounts for bedrock geochemistry, bedrock structural features, and
the geochemistry and permeability of the glacial overburden. Deeply
buried geological Hg sources may be reflected by Hg anomalies in the
surface environment, provided that mobility is favoured by the
ambient geochemical conditions and the presence of a permeable zone
that permits migration. The dominant geochemical controls on the
migration of Hg in the environment are adsorption/desorption
reactions and factors which affect Hg volatility (temperature,
humidity, and barometric pressure) (Klusman and Jaacks, 1987;
Klusman and Webster, 1981).
The literature indicates that fault-related Hg anomalies tend to be
restricted in areal extent, with diameters ranging from 10 to 150 m
(Kovalevsky, 1986). Because of their small size, fault-related Hg
anomalies are easily missed if the sample spacing is too wide.
Consequently, the interpretation of lineaments from satellite images
can be a useful aid in designing a cost-effective sampling program.
In the Huntsville study, structural lineaments were located on a
satellite image, and a sample collection grid was designed to
transect the lineaments at approximately right angles (Rasmussen and
Gardner, 1992). Initial surveys were conducted at a reconnaissance
scale (sample spacing 200 to 500 m) and follow-up surveys were
conducted at a detailed scale (sample spacing 10 to 50 m).
Sample Collection
The geochemical exploration literature reports success in defining
buried geological features by measuring total Hg concentrations in
soil and vegetation samples collected at regular intervals along
transects. However, extreme care is required to minimize all sources
of within-site variation that may obscure differences between sites.
Of prime concern is the avoidance of contamination during all stages
of sample collection, handling and processing. The Huntsville survey
included a detailed study of sources of within-site variation,
including natural variance in Hg distribution within the site,
variance introduced by inconsistent sampling and processing methods,
and analytical variance (Rasmussen et al., 1991; Rasmussen, 1994b).
The test for natural within-site variation compared tissue of the
same age from the same organ of different plants of the same species
growing at the same site. Within-site variation averaged 18.8% RSD
(N = 23 sites) using strict sampling and processing controls. Two
sources of laboratory variation, instrumental error (2.2% RSD; N =
480 digests) and analytical error (3.7% RSD; N = 61 samples) were
insignificant by comparison (Rasmussen, 1994b).
Vegetation Surveys
For surveying purposes, the most informative plant species are those
which
* demonstrate a tendency to accumulate Hg,
* occur commonly in the study area, and
* display sensitivity to spatial changes in ambient Hg
concentrations, such that between-site variation is much greater
than within-site variation.
In the Huntsville area, certain lower plants best satisfied these
criteria, namely Pleurozium schreberi, a pleurocarpous moss;
Polytrichum commune, an acrocarpous moss; and Lycopodium
dendroideum, a clubmoss (Rasmussen et al., 1991; Rasmussen, 1994b).
It is very important when sampling vegetation for Hg surveying
purposes to compare tissue of the same age and the same organ from
the same species. If possible, sampling at least two separate plants
of one species is recommended to obtain a representative Hg value
for one site. Vegetation surveys should be completed in as short a
time as possible (preferably less than two weeks) to minimize error
caused by temporal variation (Rasmussen, 1995).
Soil Surveys
For surveying purposes, the B horizon is generally preferred over
the organic surface horizon as the B horizon tends to be a less
heterogeneous sampling medium (Dunn, 1987). Vertical variation in Hg
content of a soil profile is significant and it is therefore
important to sample consistently at the same depth below the upper
contact of the B horizon. A correction for the amount of organic
matter in soil samples is generally recommended, due to the affinity
of Hg for organic matter (Carr et al., 1986). In the Huntsville
study, for example, a linear correlation of R2 = 0.6 between Hg
(ng/g) and organic carbon (%C) was observed for 13 samples collected
from various horizons in 3 soil pits dug at the same location
(Rasmussen, 1993b). In B horizon samples collected from 57
well-drained, upland sites in the Huntsville area, the organic
carbon content varied by an order of magnitude (from 0.8 to 8.3%).
Normalizing Hg against organic carbon is a method used to eliminate
"false Hg anomalies" caused by elevated proportions of organic
matter in the soil sample, rather than by the influence of a
geological source.
Statistical treatment of data
Threshold concentrations are established to distinguish between
background and anomalous concentration populations in each sample
type. Threshold concentrations are variously defined as the "upper
limit of background variation" or as the "minimum anomalous value".
This technique has been used to interpret spatial variations of
lithogenic Hg in soil by Van Kooten (1987), Varekamp and Buseck
(1983), Kodosky (1989), and Williams (1985). Two statistical methods
of determining threshold Hg concentrations were used in the
Huntsville study: the cumulative probability graph technique
developed by Tennant and White (1959) and Sinclair (1974), and the
gap statistic technique developed by Miesch (1981; software
developed for IBM-PC by Koch, 1987). There was excellent agreement
between the two methods, allowing a rigorous definition of Hg
anomalies observed in the Huntsville watershed system. These
anomalies were characterized by Hg concentrations ranging from 2 to
12 times background, and were interpreted to be fault-related Hg
dispersion halos (Rasmussen and Gardner, 1992; Rasmussen, 1993a;
1993b).
Summary
To characterize local natural sources of Hg at a monitoring site,
the study design requires
* appropriate sample spacing for the size of the potential Hg
anomaly,
* consistent sampling to minimize sources of background variation
that will obscure the anomaly, and
* a statistically defined "threshold" concentration to distinguish
"background" and "anomalous" Hg concentration populations in each
sample type.
Acknowledgements. Funding for the Huntsville soil and vegetation
surveys from the Ontario Ministry of the Environment, an NSERC
Strategic Grant (P.I.: Pam Welbourn) and two Ontario Graduate
Scholarships is gratefully acknowledged. The study formed part of
the author's PhD thesis, under the supervision of Jerome Nriagu and
Sherry Schiff.
References
Carr, G.R., Wilmshurst, J.R. and Ryall, W.R., 1986. Evaluation of
mercury pathfinder techniques: base metal and uranium deposits.
Journal of Geochemical Exploration, 26: 1-117.
Dunn, C.E., 1987. Developments in Biogeochemical Exploration. in
Proceedings of Exploration '87: Third Decennial International
Conference on Geophysical and Geochemical Exploration for Minerals
and Groundwater, Toronto. G.D. Garland. (ed.), Spec. Vol. Ont. Geol.
Surv. 1989, pp. 417-438.
Dunnette, D.A., 1988. Assessment of health risk from lithogenic
mercury. Proceedings, 196th American Chemical Society National
Meeting., Los Angeles, U.S.A.
Gilbert, Richard O., 1987. Statistical Methods for Environmental
Pollution Monitoring. Van Nostrand Reinhold Company, Ltd. New York.
Klusman, R.W. and Jaacks, J.A., 1987. Environmental influences upon
mercury, radon and helium concentrations in soil gases at a site
near Denver, Colorado. J. Geochem. Explor., 27: 259-280.
Klusman, R.W. and Webster, J.D., 1981. Meteorological noise in
crustal gas emission and relevance to geochemical exploration. J.
Geochem. Explor., 15: 63-76.
Koch, George S., 1987. Exploration Geochemical Analysis with the
IBM-PC. Van Nostrand Reinhold Company, Ltd. New York.
Kodosky, L.G., 1989. Surface mercury geochemistry as a guide to
volcanic vent structure and zones of high heat flow in the Valley of
Ten Thousand Smokes, Katmai National Park, Alaska. J. Volcan.
Geotherm. Res. 38: 227-242.
Kovalevsky, A.L., 1986. Mercury-biogeochemical exploration for
mineral deposits. Biogeochemistry 2: 211-220.
Miesch, A.T., 1981. Estimation of the geochemical threshold and its
statistical significance. J. Geochem. Explor. 16: 49-76.
Rannie, W.F. and Punter, E., 1987. Survey of existing data on levels
and sources of mercury within the region covered by the Canada -
Manitoba Mercury Agreement. Canada - Manitoba Agreement on the Study
and Monitoring of Mercury in the Churchill River Diversion.
Technical Appendices to the Summary Report, vol. 1., 179 p.
Rasmussen, P.E., Mierle, G. and Nriagu, J.O., 1991. The analysis of
vegetation for total mercury. Water, Air and Soil Pollution. 56:
379-390.
Rasmussen, P.E. and Gardner H.D., 1992. Remote sensing used in a
mercury biogeochemical survey, Huntsville, Ontario; in Proceedings
of 15th Canadian Symposium on Remote Sensing. (J.K. Hornsby, D.J.
King and N.A. Prout eds.). Toronto, Canada, 1-4 June 1992, Ontario
Ministry of Natural Resources, pp. 287-292.
Rasmussen, P.E., 1993a. The environmental significance of geological
sources of mercury: a Precambrian Shield watershed study. PhD
thesis, Earth Sciences Department, University of Waterloo, 379p.
Rasmussen, P.E., 1993b. A soil mercury survey in a Precambrian
Shield watershed. 9th International Conference on Heavy Metals in
the Environment. (R.J. Allan and J.O. Nriagu eds.). Toronto, Canada,
12-17 September 1993. Environment Canada. 2:62-65.
Rasmussen, P.E., 1993c. The significance of crustal sources of
mercury to hydroelectric reservoir development in the Precambrian
Shield. Ontario Hydro Research Division, Report No. 93-64-K, 17p.
Rasmussen, P.E., 1994a. Current methods of estimating atmospheric
mercury fluxes in remote areas. Environmental Science and
Technology, 28(13): 2233-2241.
Rasmussen, P.E. 1994b. Mercury in vegetation of the Precambrian
Shield; in Mercury Pollution: Integration and Synthesis. (C.J.
Watras and J.W. Huckabee eds.) Lewis Publishers/CRC Press, Boca
Raton, FL, USA. Chapter IV-5, pp. 417-425.
Rasmussen, P.E., 1995. Temporal variation of mercury in vegetation.
Water, Air and Soil Pollution, 80: 1039-1042.
Sinclair, A.J., 1974. Selection of threshold values in geochemical
data using probability graphs. J. Geochem. Explor. 3: 129-149.
Tennant, C.B. and White, M.L., 1959. Study of the distribution of
some geochemical data. Econ. Geol. 54: 1281-1290.
Van Kooten, G. K., 1987. Geothermal Exploration using surface
mercury geochemistry. J. Volcan. Geotherm. Res. 31: 269-280.
Varekamp, J.C. and Buseck, P.R., 1983. Hg anomalies in soils: a
geochemical exploration method for geothermal areas. Geothermics.
12: 29-47.
Williams, Stanley N., 1985. Soil radon and elemental mercury
distribution and relation to magmatic resurgence at Long Valley
Caldera. Science. 229: 551-553.
[INLINE]
back to intro next section
BACK TO
*********************************************************************