United States
Environmental Protection
Agency
EPA-452/R-96-001ei/'
June 1996
Air
Mercury Study
Report to Congress
Volume V:
An Ecological Assessment of
Anthropogenic Mercury
Emissions in the United States
SAB REVIEW DRAFT
xvEPA
Office of Air Quality Planning £r Standards
and
Office of Research and Development
C66011-2-5
-------
MERCURY STUDY REPORT TO CONGRESS
VOLUME V:
AN ECOLOGICAL ASSESSMENT OF ANTHROPOGENIC
MERCURY EMISSIONS IN THE UNITED STATES
SAB REVIEW DRAFT
June 1996
0 s P.ot««on
Chicago,
60604-3590
Office of Air Quality Planning and Standards
and
Office of Research and Development
U.S. Environmental Protection Agency
-------
age
TABLE OF CONTENTS
',r\
U.S. EPA AUTHORS iv
SCIENTIFIC PEER REVIEWERS v
LIST OF TABLES '. 'vii
LIST OF FIGURES viii
LIST OF SYMBOLS, UNITS AND ACRONYMS ix
EXECUTIVE SUMMARY ES-1
1. INTRODUCTION 1-1
2. PROBLEM FORMULATION %. 2-1
2.1 Stressor Characteristics: Mercury Speciation and Cycling 2-1
2.1.1 Mercury in Air 2-3
2.1.2 Mercury in Surface Water 2-4
2.1.3 Mercury in Soil 2-5
2.2 Potential Exposure Pathways 2-5
2.2.1 Exposure Pathways in Aquatic Systems 2-5
2.2.2 Exposure Pathways in Terrestrial Systems 2-8
2.2.3 Summary of Aquatic and Terrestrial Exposure Pathways 2-10
2.3 Ecological Effects 2-11
2.3.1 Bioaccumulation of Mercury 2-12
2.3.2 Individual Effects 2-22
2.3.3 Population Effects 2-26
2.3.4 Communities and Ecosystems 2-31
2.3.5 Conclusions 2-32
2.4 Ecosystems Potentially at Risk 2-33
2.4.1 Highly Exposed Areas 2-33
2.4.2 Lakes and Streams Impacted by Acid Deposition 2-33
2.4.3 Factors in Addition to Low pH that Contribute to Increased
Bioaccumulation of Mercury in Aquatic Biota 2-34
2.4.4 Sensitive Species 2-34
2.5 Endpoint Selection 2-34
2.6 Conceptual Model for Mercury Fate and Effects in the Environment 2-36
3. EXPOSURE OF PISCIVOROUS AVIAN AND MAMMALIAN WILDLIFE TO
AIRBORNE MERCURY 3-1
3.1 Objectives and Approach 3-1
3.2 Computer Models 3-2
3.3 Current Exposure of Piscivorous Wildlife to Mercury 3-3
3.4 Exposure Estimates Based on Measured Deposition and an Indirect Exposure
Methodology 3-4
3.5 Regional-scale Exposure Estimates 3-6
3.5.1 Analysis of RELMAP Results 3-7
3.5.2 Locations of Socially Valued Environmental Resources 3-7
3.5.3 Airborne Deposition Overlay with Threatened and Endangered Plants . . 3-10
June 1996 i SAB REVIEW DRAFT
-------
TABLE OF CONTENTS (continued)
Paee
3.5.5 Regions of Concern Overlay with Wildlife Species Distribution Maps . . 3-10
3.6 Local-scale Exposure Estimates 3-22
3.6.1 Approach and Assumptions 3-22
3.6.2 Results of Local-scale Exposure Analysis 3-23
4. EFFECT OF AIRBORNE MERCURY ON PISCIVOROUS AVIAN AND
MAMMALIAN WILDLIFE .% 4-1
4.1 Bioaccumulation Factors (BAFs) for Magnification of Methylmercury in the
Aquatic Food Chain 4-1
4.1.1 Definition of BAFs and Overview 4-1
4.1.2 BAF Estimation Methods 4-2
4.1.3 Results of BAF Simulations and Recommended Values 4-3
4.1.4 Sensitivity Analysis 4-4
4.1.5 Uncertainty and Variability .. 4-5
4.1.6 Conclusions 4-6
4.2 Calculation of a Criterion Value for Protection of Piscivorous Wildlife 4-7
4.2.1 Procedure Used to Develop Criterion Values for Wildlife in the Water
Quality Guidance for the Great Lakes System 4-7
4.2.2 Exposure Parameters 4-9
4.2.3 Health Endpoints for Avian Wildlife 4-9
4.2.4 Health Endpoints for Mammalian Wildlife 4-10
4.2.5 Summary of Health Endpoints for Avian and Mammalian Wildlife .... 4-11
4.2.6 Calculation of Wildlife Criterion Values 4-12
4.2.7 Calculation of Mercury Residues in Fish Corresponding to the Wildlife
Criterion Value 4-14
4.2.8 Calculation of a Wildlife Criterion for the Florida Panther 4-14
4.2.9 Comparison of GLWQI Criteria with WC Derived in this Report 4-14
4.3 Uncertainty Analysis 4-16
4.4 Sensitivity Analysis 4-17
4.5 Uncertainties Associated with the GLWQI Methodology 4-17
4.5.1 Limitations of the Toxicity Database 4-18
4.5.2 LOAEL-to-NOAEL Uncertainty Factor UFL 4-19
4.5.3 Validity of BCF/BAF Paradigm 4-23
4.5.4 Selection of Species of Concern 4-24
4.5.5 Trophic Levels at Which Wildlife Feed 4-25
4.5.6 Variability in BAFs at each Trophic Level 4-25
4.5.7 Natural History Considerations 4-26
4.5.8 Individuals Versus Populations 4-26
4.5.9 Species Versus Taxa 4-26
4.5.10 Discussion of Uncertainties Associated with the GLWQI Methodology . 4-27
5. CONCLUSIONS 5-1
June 1996 ii SAB REVIEW DRAFT
-------
TABLE OF CONTENTS (continued)
Page
6. RESEARCH NEEDS 6-1
6.1 Process-based Research 6-1
6.2 Wildlife Toxicity Data 6-1
6.3 Improved Analytical Methods 6-1
6.4 Complexity of Aquatic Food Webs 6-2
6.5 Accumulation in Trophic Levels 1 and 2 6-2
6.6 Field Residue Data 6-2
6.7 Natural History Data "..... 6-3
7. REFERENCES 7-1
APPENDIX A: ESTIMATION OF BIO ACCUMULATION FACTORS FOR
MERCURY IN FISH A-l
June 1996 iii SAB REVIEW DRAFT
-------
U.S. EPA AUTHORS
Principal Author:
John W. Nichols, Ph.D.
Mid-Continent Ecology Division
Office of Research and Development
Duluth,MN
Contributing Authors:
Robert B. Ambrose, Jr., P.E.
Ecosystems Research Division
National Exposure Research Laboratory
Athens, GA
Chris Cubbison, Ph.D
National Center for Environmental Assessment-
Cincinnati
Office of Research and Development
Cincinnati, OH
Anne Fairbrother, Ph.D., D.V.M.
Environmental Research Laboratory-Corvallis
Corvallis, OR
currently with:
Ecological Planning and Toxicology, Inc.
5010 S.W. Hout St
Corvallis, OR 97333
Martha H. Keating
Office of Air Quality Planning and Standards
Research Triangle Park, NC
Kathryn R. Mahaffey, Ph.D.
National Center for Environmental Assessment-
Cincinnati
Office of Research and Development
Cincinnati, OH
Debdas Mukerjee, Ph.D.
National Center for Environmental Assessment-
Cincinnati
Office of Research and Development
Cincinnati, OH
Glenn E. Rice
National Center for Environmental Assessment-
Cincinnati
Office of Research and Development
Cincinnati, OH
David J. Reisman
National Center for Environmental Assessment-
Cincinnati
Office of Research and Development
Cincinnati, OH
Rita Schoeny, Ph.D.
National Center for Environmental Assessment-
Cincinnati
Office of Research and Development
Cincinnati, OH
Jeff Swartout
National Center for Environmental Assessment-
Cincinnati
Office of Research and Development
Cincinnati, OH
Michael Troyer
Office of Science, Planning and Regulatory
Evaluation
Cincinnati, OH
June 1996
IV
SAB REVIEW DRAFT
-------
SCIENTIFIC PEER REVIEWERS
Brian J. Alice, Ph.D.
Harza Northwest, Incorporated
Thomas D. Atkeson, Ph.D.
Florida Department of Environmental
Protection
Steven M. Bartell, PhJX
SENES Oak Ridge, Inc.
Mike Bolger, Ph.D.
U.S. Food and Drug Administration
James P. Butler, Ph.D.
University of Chicago
Argonne National Laboratory
Rick Canady, Ph.D.
Agency for Toxic Substances and Disease
Registry
Rufus Chaney, Ph.D.
U.S. Department of Agriculture
TimEder
Great Lakes Natural Resource Center
National Wildlife Federation for the
States of Michigan and Ohio
William F. Fitzgerald, Ph.D.
University of Connecticut
Avery Point
Robert Goyer, Ph.D.
National Institute of Environmental Health
Sciences
George Gray, Ph.D.
Harvard School of Public Health
Terry Haines, Ph.D.
National Biological Service
Joann L. Held
New Jersey Department of Environmental
Protection & Energy
Gerald J. Keeler, Ph.D.
University of Michigan
Ann Arbor
Leonard Levin, Ph.D.
Electric Power Research Institute
Malcom Meaburn, Ph.D.
National Oceanic and Atmospheric
Administration
U.S. Department of Commerce
Paul Mushak, Ph.D.
PB Associates
Jozef M. Pacyna, Ph.D.
Norwegian Institute for Air Research
Ruth Patterson, Ph.D.
Cancer Prevention Research Program
Fred Gutchinson Cancer Research Center
Donald Porcella, Ph.D.
Electric Power Research Institute
Charles Schmidt
U.S. Department of Energy
Pamela Shubat, Ph.D.
Minnesota Department of Health
Alan H. Stem, Dr.P.H.
New Jersey Department of Environmental
Protection & Energy
Edward B. Swain, Ph.D.
Minnesota Pollution Control Agency
M. Anthony Verity, M.D.
University of California
Los Angeles
June 1996
SAB REVIEW DRAFT
-------
WORK GROUP AND U.S. EPA/ORD REVIEWERS
Core Work Group Reviewers:
Dan Axelrad, U.S. EPA
Office of Policy, Planning and Evaluation
Angela Bandemehr, U.S. EPA
Region 5
Jim Darr, U.S. EPA
Office of Pollution Prevention and Toxic
Substances
Thomas Gentile, State of New York
Department of Environmental Conservation
Arnie Kuzmack, U.S. EPA
Office of Water
David Layland, U.S. EPA
Office of Solid Waste and Emergency
Response
Karen Levy, U.S. EPA
Office of Policy Analysis and Review
Steve Levy, U.S. EPA
Office of Solid Waste and Emergency
Response
Lorraine Randecker, U.S. EPA
Office of Pollution Prevention and Toxic
Substances
Joy Taylor, State of Michigan
Department of Natural Resources
U.S. EPA/ORD Reviewers:
Robert Beliles, Ph.D.,-D.A.B.T.
National Center for Environmental Assessment
Washington, DC
Eletha Brady-Roberts
National Center for Environmental Assessment
Cincinnati, OH
Annie M. Jarabek
National Center for Environmental Assessment
Research Triangle Park, NC
Matthew Lorber
National Center for Environmental Assessment
Washington, DC
Susan Braen Norton
National Center for Environmental Assessment
Washington, DC
Terry Harvey, D.V.M.
National Center for Environmental Assessment
Cincinnati, OH
June 1996
VI
SAB REVIEW DRAFT
-------
LIST OF TABLES
ES-l Percentiles of Predicted Methylmercury Concentrations in Fish (ug/g) Based
on a Total Mercury Dissolved Water Concentration of 0.7 ng/L ES-2
ES-2 Percent of Species Range Overlapping with Regions of Concern ES-3
ES-3 Wildlife Criteria for Mercury ES-9
2-1 . Examples of Effects of Contaminants on Ecosystem Components 2-11
2-2 Nationwide Average of Mercury Residues in Fish 2-16
2-3 Mercury Residues in Tissues of Piscivorous Birds 2-17
2-4 Mercury Residues in Tissues of Piscivorous Mammals 2-20
2-5 Toxicity Values for Aquatic Plants . A. . 2-22
2-6 Mercury Toxicity Increases With Temperature . . . -, 2-23
2-7 Toxicity Values for Fish and Aquatic Invertebrates 2-24
2-8 Examples of Assessment and Measurement Endpoints 2-35
3-1 Models Used to Predict Mercury Air Concentrations, Deposition Fluxes and
Environmental Concentrations 3-2
3-2 Exposure Parameters for Mink, Otter, Kingfisher, Osprey, and Eagle 3-4
3-3 Summary of Sample Calculations of Wildlife Species Methylmercury Exposure from
Fish Digestion, based on Average Fish Residue Values 3-4
3-4 Mercury Concentrations in Water and Sediment Predicted Using a Mercury Air
Concentration of 1 ng/m3, Deposition Rate of 10 ug/m2/yr, and Soil Concentration of
50 ng/g 3-5
3-5 Methylmercury Concentrations in Fish (ug/g) Predicted Using a Mercury Air
Concentration of 1 ng/m3, Deposition Rate of 10 ug/m2/yr, and Soil Concentration of
50 ng/g 3-6
3-6 Percentiles of Predicted Methylmercury Concentrations in Rsh (ug/g) Based on a
Total Mercury Dissolved Water Concentration of 0.7 ng/L 3-6
3-7 Process Parameters for the Model Plants Considered ia the Local Impact Analysis .... 3-22
3-8 Predicted Intakes for Wildlife Receptors for the Eastern Site 3-24
3-9 Predicted Intakes for Wildlife Receptors for the Western Site 3-25
4-1 Summary of Bioaccumulation Factors for Trophic Levels 3 and 4 4-4
4-2 Exposure Parameters for Mink, Otter, Kingfisher, Osprey, and Eagle 4-9
4-3 Species-specific Wildlife Criteria Calculated in the Great Lakes Water Quality
Initiative (GLWQI)a and in the Mercury Study Report to Congress 4-15
4-4 Analysis of LOAEL-to-NOAEL Uncertainty Factor 4-20
June 1996 vii SAB REVIEW DRAFT
-------
LIST OF FIGURES
ES-l Surface Water with pH < 5.5 and Anthropogenic Mercury Deposition ES-4
ES-2 Common Loon Range, Surface Water with pH < 5.5, and Anthropogenic
Mercury Deposition ES-6
ES-3 Florida Panther Range, Surface Water with pH < 5.5, and Anthropogenic
Mercury Deposition (Detail: Eastern U.S.) ES-7
2-1 Cycling of Mercury in Freshwater Lakes 2-2
2-2 Possible Routes of Exposure to Mercury 2-6
2-3 Distribution of Mercury in a Water Body 2-7
2-4 Example Aquatic Food Web 2-8
2-5 Example Terrestrial Foo(l Web . 2-9
3-1 Total Anthropogenic Mercury Deposition 3-8
3-2 Major Rivers and Lakes ." 3-9
3-3 National Resource Lands 3-11
3-4 Threatened and Endangered Plan Species and anthropogenic Mercury Deposition 3-12
3-5 Surface Water and pH< 5.5 and Anthropogenic Mercury Deposition 3-13
3-6 Kingfisher Range, Surface Water with pH < 5.5, and Anthropogenic Mercury
Deposition 3-14
3-7 Bald Eagle Range, Surface Water with pH < 5.5, and Anthropogenic Mercury
Deposition 3-15
3-8 Osprey Range, Surface Water with pH < 5.5, and Anthropogenic Mercury Deposition . . 3-17
3-9 Common Loon Range, Surface Water with pH < 5.5, and Anthropogenic Mercury
Deposition ........... „ 1 3-18
3-10 Florida Panther Range, Surface Water with pH < 5.5, and Anthropogenic Mercury
Deposition „ 3-19
3-11 Mink Range, Surface Water with pH <5.5, and Anthropogenic Mercury Deposition .... 3-20
3-12 River Otter Range, Surface Water with pH <5 J, and Anthropogenic Mercury
Deposition • 3-21
4-1 LOAEL-to-NOAEL Ratio Distribution 4-22
June 1996 viii SAB REVIEW DRAFT
-------
LIST OF SYMBOLS, UNITS AND ACRONYMS
ATP
AWQC
BAF
BAF3
BAF4
BCF
BCF
fflg
BCFMHg
BSAF
BMP
bw
CAA
CAP
CMWI
COMPDEP
d
DDE
DDT
FA
FCM
FD3
FD4
FDER
GLWQI
GM
GSD
ha
Hg°
Hg
2+
HgBAF3
HgBAF4
IEM2
IJC
IMWI
kg
L
LC50
LCUB
LMWC
LOAEL
MCUB
MCM
Adenosine triphosphate
Ambient Water Quality Criteria
Bioaccumulation factor
Aquatic life bioaccumulation factor for trophic level 3
Aquatic life bioaccumulation factor for trophic level 4 .
Bioconcentration factor
Total mercury concentration in fish divided by that in water following a waterborne
exposure to inorganic mercury
Total mercury concentration in fish divided by that in water following a waterborne
exposure to methylmercury
Biota-sediment accumulation factor
Biomagnification factor
Body weight
Clean Air Act as Amended hi 1990
Chlor-alkali plant
Continuous medical waste incinerator
Short range air dispersion model for mercury
Day
p,p-Dichlorodiphenyldichloroethylene
4,4-Dichlorodiphenyltrichloroethane
Average daily amount of food consumed
Food chain multiplier
Fraction of the diet derived from trophic level 3
Fraction of the diet derived from trophic level 4
Florida Department of Environmental Regulation
Great Lakes Water Quality Initiative
Geometric mean
Geometric standard deviation
Hectare
Elemental mercury
Mercurous ion
Mercury II
Total mercury in forage fish (trophic level 3) divided by that in water accumulated
by all possible routes of exposure
Total mercury in piscivorous fish (trophic level 4) divided by that in water
accumulated by all possible routes of exposure
Indirect exposure methodology
International Joint Commission
Intermittent medical waste incinerator
Kilogram
Liter
Lethal concentration (for fifty percent of population)
Large coal-fired utility boiler
Large municipal waste combustor
Lowest-observed-adverse-effect level
Medium coal-fired utility boiler
Mercury cycling model
June 1996
IX
SAB REVIEW DRAFT
-------
LIST OF SYMBOLS, UNITS AND ACRONYMS (continued)
MeHgT
MeHgw
MDNR
m
m3
mg
MOUB
ng
nM
NOAEL
PCBs
PCS
pcti
Pg
pH
PLS
PPF
PPF4
RELMAP
SAB
SCUB
SMWC
sp.
TL3
TL3
UFA
UFS
UFL
U.S. EPA
Hg
uM
WA
we
wcf
wq
wcs
WHO
WtA
Percent of total mercury in fish tissues existing as the methylated form
Percent of total mercury in water existing as the methylated form
Michigan Department of Natural Resources
Meter
Cubic meter
Milligram
Medium oil-fired utility boiler
Nanogram
Nanomole
No-observed-adverse-effect level
Polychlorinated biphenyls
Primary copper smelter
Percentile
Picogram
Logarithm of the reciprocal of the hydrogen ion concentration. A measure of
acidity
Primary lead smelter
Predator-prey factor
The observed ratio of total mercury concentration at trophic level 4, divided by
methylmercury concentrations at trophic level 3
Regional Lagrangian Model of Air Pollution
Science Advisory Board
Small coal-fired utility boiler
Small municipal waste combustor
Species
Trophic level 3
Trophic level 4
Uncertainty factor for species extrapolation
Uncertainty factor for use of less than lifetime study
Uncertainty factor for use of a lowest adverse effect level
U.S. Environmental Protection Agency
Microgram
Micromole
Average daily volume of water consumed
Wildlife criterion level
Final wildlife criterion level
Intermediate wildlife criterion level
Species-specific wildlife criterion level
World Health Organization
Average species weight
June 1996
SAB REVIEW DRAFT
-------
EXECUTIVE SUMMARY
Section 112(n)(l)(B) of the Clean Air Act (CAA), as amended in 1990, directs the U.S.
Environmental Protection Agency (U.S. EPA) to submit to Congress a comprehensive study on
emissions of mercury to the air. Volume V, which addresses the ecological exposure and effects
assessment for mercury and mercury compounds, is part of a seven-volume report developed by U.S.
EPA in response to this directive.
Volume V comprises an ecological assessment for anthropogenic mercury emissions. It follows
the format of the U.S. EPA Framework for Ecological Risk Assessment (U.S. EPA 1992a). The first
step in the Framework is the problem formulation phase, wherein the potential ecological impacts of
mercury are reviewed. This is followed by the presentation of a conceptual model describing how
airborne mercury accumulates in aquatic biota, biomagnifies in aquatic food chains and is consumed
by wildlife that eat contaminated fish. Subsequent steps in the assessment include an exposure
assessment and finally the calculation of criteria for the protection of piscivorous avian and
mammalian wildlife.
Scope of the Assessment
The scope of this assessment was limited solely to anthropogenic mercury that is emitted
directly to the atmosphere. The origins and extent of these emissions are reviewed in Volume II of
this Report. This analysis did not address mercury originating from direct wastewater discharge to
water bodies, mining waste or the application of mercurial pesticides. In a number of instances, these
and other "point" sources have been related to unacceptably high mercury levels in fish, triggering
site-specific fish consumption advisories. Clearly, where such point sources exist, there is a need to
address the combined impacts of mercury originating from all sources, including air emissions.
Mercury in the Environment
Wet deposition is thought to be the primary mechanism by which mercury emitted to the
atmosphere is transported to surface waters and land, although dry deposition may also contribute
substantially. Once deposited, mercury enters aquatic and terrestrial food chains. Mercury
concentrations increase at successively higher trophic levels, as a result of bioconcentration (in prey
organisms), bioaccumulation and biomagnification. Of the various forms of mercury in the
environment, methylmercury has the highest potential for bioaccumulation and biomagnification.
Predators at the top of these food chains are potentially at risk from consumption of mercury in
contaminated prey. Based on the review of available information, it was concluded that piscivorous
(fish-eating) birds and mammals are particularly at risk from airborne mercury emissions. This risk is
likely to be greatest in areas that receive high levels of mercury deposition or, because of a strong
negative correlation between surface water pH and mercury residues in fish, in regions that contain
poorly buffered surface waters.
The assessment endpoint for this ecological assessment is identified as the maintenance of self-
sustaining wildlife populations. Measurement endpoints include the growth and survival of individual
animals, and reproductive success.
June 1996 ES-1 SAB REVIEW DRAFT
-------
Exposure of Piscivorous Wildlife to Mercury
Exposure was characterized in a progressive manner, with varying reliance on computer
models for mercury deposition and fate. The objective of this analysis was to characterize the extent
to which piscivorous wildlife are exposed to mercury originating from airborne emissions. Details on
exposure assessment inputs, methods and results can be found in Volume HI of this Report to
Congress. Four general approaches were used, which are described as follows.
1. Estimation of current average exposure to piscivorous wildlife on a nationwide basis.
The first analysis was conducted without computer models. Estimates of current mercury
exposure to selected piscivorous wildlife species were calculated as the product of the fish
consumption rate and measured mercury concentrations in fish. This analysis was not intended to be a
site-specific analysis, but rather to provide national exposure estimates for piscivorous wildlife. This
analysis used mean total mercury measurements from a nationwide study of fish residues and
published fish consumption data for the selected wildlife species. The relative ranking of exposure in
Hg/kg bw/d of selected wildlife species was as follows: kingfisher > river otter > osprey = mink >
bald eagle.
2. Estimation of mercury levels in fish using measured deposition values and an indirect exposure
methodology.
In the second analysis, measured mercury deposition rates were used as inputs to an indirect
exposure methodology (IEM2) to estimate mercury concentrations in water, soil and fish. Additional
inputs to the IEM2 model include the characteristics of a hypothetical lake and its associated
watershed. The analysis was conducted for two such hypothetical lakes, one located in the Western
U.S., the other located in the Eastern U.S. Residue values were calculated as the product of predicted
mercury concentrations in water and estimated bioaccumulation factor (BAF) values for fish in trophic
levels 3 and 4. (An explanation of trophic levels and the assumptions used in this analysis are
described in Chapter 4 and Appendix of this volume.) Mercury levels in fish estimated in this
manner, summarized in Table ES-1 below, were consistent with measured values from field studies.
Table ES-1
Percentiles of Predicted Methylmercury Concentrations in Fish (jig/g) Based on a
Total Mercury Dissolved Water Concentration of 0.7 ng/L
Parameter
Trophic 3 BAF
Predicted Fish Concentration (ng/g)
Trophic 4 BAF
Predicted Fish Concentration (ng/g)
Geometric
Mean
67,000
0.05
335,000
0.23
Percentile of Distribution
5th
6,400
0.00
22,700
0.02
25th
25,400
0.02
111,000
0.08
50th
66,200
0.05
336,000
0.24
75th
172,400
0.12
1,000,000
0.70
95th
684,000
0.48
4,700,000
3.30
June 1996
ES-2
SAB REVIEW DRAFT
-------
3. Estimation of mercury deposition on a regional scale (40 km grid), and a comparison of these
deposition data with species distribution information.
The third type of analysis was carried out on a national scale. A long-range atmospheric
transport model (RELMAP) was used in conjunction with the mercury emissions inventory to generate
predictions of mercury deposition across the continental U.S. Ecosystems subject to high levels of
mercury deposition (e.g., near sources of mercury emissions, in areas with high deposition rates) will
be more exposed to mercury than ecosystems with lower levels of mercury deposition. The pattern of
mercury deposition nationwide, therefore, will influence which ecoregions and ecosystems might be
exposed to hazardous levels of mercury. Thus, predictions of mercury deposition were compared with
the locations of major lakes and rivers, national resource lands, threatened and endangered plant
species and the distributions of selected piscivorous wildlife species. Additionally, mercury deposition
data were superimposed onto a map of surface waters impacted by acid deposition, because it has been
shown that low pH values are positively correlated with high levels of mercury in fish. Areas
receiving high levels of deposition that also contain large numbers of poorly buffered lakes were
designated "regions of concern" (see Figure ES-1). The extent of overlap of selected species
distributions with these "regions of concern" was characterized.
In the case of plant life, analysis was limited to plotting the location of federally threatened or
endangered species and indicating where threatened populations coincide with estimated high mercury
deposition. Large concentrations of endangered plant populations exposed to high levels of deposition
occur in central and southern Florida, along the northeastern coastal region and scattered throughout
the Midwest. Mercury has been demonstrated to have adverse impacts on a number of plant species.
Avian wildlife considered in this
analysis included piscivorous species with
habitats that are widely distributed (kingfishers)
and narrowly distributed (bald eagles), as well as
birds whose range fell within areas of high
mercury deposition (ospreys and common
loons). All the birds selected were piscivores
that feed at or near the top of aquatic food
chains and are therefore at risk from
biomagnified mercury. Two of the mammals
selected for this analysis (mink and river otters)
are piscivorous and widely distributed. The
other mammal selected, the Florida panther, is
not widely distributed but is listed as an
endangered species. The Florida panther lives in
an environment known to be contaminated with
mercury and preys upon small mammals (such
as raccoons) which may contain high tissue
burdens of mercury. Results for each avian and
mammalian species are summarized in Table
ES-2.
Table ES-2
Percent of Species Range
Overlapping with Regions of Concern
Species
Kingfisher
Bald Eagle
Osprey
Common Loon
Florida Panther
Mink
River Otter
Percent of Range
Impacted
8%
17%
13%
23%
<1%
9%
14%
June 1996
ES-3
SAB REVIEW DRAFT
-------
Figure ES-1
Surface Water and pH < 5.5 and Anthropogenic Mercury Deposition
NORTHEAST
WEST
5 20% pH < = 5.6. Depoiition 1-6 ug/m2
6-20% pH < - 5.6. DopotHJon 6-10 ug/m2
6-20% pH < - 5.6, DopoiMon > 10 ug/m2
> 20% pH < » 6.6, Dopo«ttJon 1-6 ua/n»2
> 20% pH < - 6.6, Deposition 6-10 ug/m2
> 20% pH < = 6.5. Deposition > 10 ug/m2
MID-ATLANTIC
COASTAL PLAIN
FLORIDA
June 1996
ES-4
SAB REVIEW DRAFT
-------
Approximately 8% of the kingfisher's range occurs within regions of concern. Given this
small degree of overlap, mercury does not appear to be a threat to the species nationwide.
Although a recovery in the population of bald eagles in some areas has resulted in a status
upgrade from "endangered" to "threatened" in five states (Michigan, Minnesota, Oregon, Washington
and Wisconsin) bald eagle populations are still depleted throughout much of their range. Bald eagles
can be found seasonally in large numbers hi several geographic locations, but most of these individuals
are transient, and the overall population is still small. Historically, eagle populations hi the U.S. have
been adversely impacted by the effects of bioaccumulative contaminants (primarily DDT and perhaps
PCBs). Approximately 17% of the bald eagle's range overlaps mercury regions of concern. The risk
to eagles posed by mercury appears to be greatest in the Great Lakes region, the northeastern Atlantic
states and south Florida.
Nationwide, approximately 13% of the osprey's total range overlaps regions of concern;
however, a much larger fraction of the osprey's eastern population occurs within these regions.* The
osprey diet consists almost exclusively of fish, and they are known to take dead fish from the water
surface if the fish are fresh. Their position at the top of the aquatic food chain places ospreys at risk
from toxins that bioaccumulate. Osprey populations underwent severe declines during the 1950s
through the 1970s; these declines have been linked to exposure to DDT.
Nearly 23% of the loon's range is located in regions of concern (see Figure ES-2). Moreover,
nearly all of the loon's range occurs in regions where mercury deposition is predicted. Limited data
from the study of mercury point sources showed that loon reproductive success was negatively
correlated with exposure to mercury hi a significant dose-response relationship. Residue data,
combined with field observations, suggest that loon populations hi areas of Minnesota and Wisconsin
may be adversely impacted by mercury originating from airborne deposition.
The Florida panther, an endangered species, lives in an environment known to be
contaminated with mercury and preys upon small mammals (such as raccoons) that may contain high
tissue burdens of mercury. Although the panther's range falls outside of identified regions of concern
(<1%), the species habitat is contiguous with this region (see Figure ES-3). Measured mercury levels
found hi Florida panther tissue have approached levels that are frankly toxic in other feline species.
Approximately 9% of the range of mink habitat coincides with regions of concern nationwide.
Mink occupy a large geographic area and are common throughout the U.S. In general, mink prey on
small mammals for most of the year; however, some populations prey primarily on fish and aquatic
birds. Mink that prey on aquatic animals are most at risk from mercury contamination. In addition
small predators may be at greater risk than large predators due to higher food consumption rate per
unit of body weight.
River otter habitats overlap regions of concern for about 14% of the range for this species
nationwide. River otters occupy large areas of the U.S., but then- population numbers are thought to
be declining in the Midwestern states. The river otter's diet is almost exclusively of aquatic origins
and includes fish (primarily), crayfish, amphibians and aquatic insects. The species of fish taken
depends on the fish's ability to escape capture. The consumption of large, piscivorous fish puts the
river otter at risk from bioaccumulative contaminants such as mercury.
June 1996 ES-5 SAB REVIEW DRAFT
-------
Figure ES-2
Common Loon Range, Surface Water with pH < 5.5, and Anthropogenic Mercury Deposition
NORTHEAST
WEST
Range lor tfti* •peole*
6-20% pH < = 6.6, Deposition 1-6 ug/m2
6-20% pH < - 6.6, Dapoiltion 6-10 ug/m2
6-20% pH < - 6.6. DopoiWon > 10 ug/m2
> 2O% pH < - 6.6, Deposition 1-6 UQ/m2
> 2O% pH < - 6.6, Deposition 6-10 ug/m2
> 20% pH < = 6.6, Deposition > 10 ug/m2
MID-ATLANTIC
COASTAL PLAIN
RORIDA
June 1996
ES-6
SAB REVIEW DRAFT
-------
Figure ES-3
Florida Panther Range, Surface Water with pH < 5.5,
and Anthropogenic Mercury Deposition
(Detail: Eastern ILS.)
Ring* for this spsets*
9-2O% pH < - 5.5, Deposition 1-8 ug/m2
5-20% pH < « 5.5, Deposition 5-10 ug/rn2
5-20% pH < - 5.5, Deposition > 10 ug/m2
> 20% pH < - 5.5, Dvpmhlon 1-5 ug/m2
> 20% pH < - 5.5, Deposition 5-10 ug/m2
> 2O% pH < - 5.5, Deposition > 10 ugim2
June 1996
ES-7
SAB REVIEW DRAFT
-------
Otter population declines do not overlap to a large extent with regions of concern; however, the area
of decline does coincide with RELMAP predictions of high mercury deposition rate.
4. Estimation of mercury deposition on a local scale in areas near emissions point sources.
A final analysis was conducted using a local-scale air dispersion model (COMPDEP), in
addition to the long-range transport data and the indirect exposure methodology, to predict mercury
concentrations in water and fish under a variety of hypothetical emissions scenarios. COMPDEP
simulated mercury deposition originating from model plants representing a range of mercury emissions
source classes. The six source categories were selected based on their estimated annual mercury
emissions or their potential to be localized point sources of concern. The categories selected were
these: municipal waste combustors (MWCs), medical waste incinerators (MWIs), utility boilers, chlor-
alkali plants, primary copper smelters and primary lead smelters.
The analysis was conducted for two hypothetical lakes located in the Western and Eastern U.S.
The proximity of these lakes to the source was varied to examine the effect of this parameter on model
predictions. To account for the long range transport of emitted mercury, the 50th percentile RELMAP
atmospheric concentrations and deposition rates were included hi the estimates from the local air
dispersion model.
These data were used to rank the relative contributions of different emission source types and
the exposure of selected wildlife species. It was concluded from this analysis that local emissions
sources have the potential to increase significantly the exposure of piscivorous birds and mammals to
mercury. The extent of this local contribution depends in turn upon watershed characteristics, facility
type, local meteorology, and terrain. The exposure of a given wildlife species is also highly dependent
upon the fish bioaccumulation factor, the trophic level(s) at which it feeds and the amount of fish
consumed per day.
Effects Assessment for Mercury
Due to the broad range and extent of mercury emissions throughout the United States, many
potential ecological effects could have been considered. Neither the available data nor existing
methodology supported evaluation of all possible risks.
The ecosystem effects of mercury are incompletely understood. No applicable studies of the
effects of mercury on intact ecosystems were found. Characterization of risk for non-human species,
thus, did not attempt to quantify effects of mercury on ecosystems, on plant and animal communities
or on species diversity. Direct effects of methylmercury on fish and other aquatic biota were not
estimated, although there is evidence of adverse impacts on these organisms following point source
releases of mercury and in aquatic environments affected by urban runoff.
Data on methylmercury effects suitable for dose-response assessment are limited to what are
termed "individual effects" hi the U.S. EPA Framework for Ecological Risk Assessment. In general,
selections of wildlife species for dose-response assessment were based on the following factors:
(1) exposure to bioaccumulative contaminants; (2) relevance to establishing species of concern on a
national basis; (3) availability of information with which to calculate criterion values; and (4) evidence
for bioaccumulation and/or adverse effects. The species selected were piscivorous birds and mammals.
Avian species were the bald eagle (Haliaeetus leucocephalus), the osprey (Pandion haliaetus) and the
belted kingfisher (Ceryle alcyon). Mammalian species were the mink (Mustela visori) and the river
otter (Lutra canadensis).
June 1996 ES-8 SAB REVIEW DRAFT
-------
The goal of this assessment was to calculate a water-based wildlife criterion (WC) value that,
if not exceeded, would be protective of piscivorous avian and mammalian wildlife. Because this
assessment (and the local-scale exposure assessment) depends to a large extent on the assignment of
BAFs for mercury in fish at trophic levels 3 and 4, an effort was made to review published field data
from which these BAFs could be estimated. A Monte Carlo analysis was then performed to
characterize the variability around these estimates.
A WC for mercury was estimated as the ratio of a no-adverse-effect dosing level (NOAEL; in
ug/d) to an estimated exposure level (L/d), referenced to a water concentration using a BAF.
Individual wildlife criteria are provided in Table ES-3. This approach is similar to that used in non-
cancer human health risk assessment and was employed previously to estimate a WC for mercury in
the Water Quality Guidance for the Great Lakes System. Species-specific WC values for mercury
were estimated for selected avian and mammalian wildlife. A final mammalian WC was then
calculated as the lowest mean of WC values for each of the two taxonomic classes (birds and
mammals). The final WC for mercury was based on individual WC values calculated for the river
otter and mink, and was estimated to be 346 picograms (ng) total mercury/L water. The WC was
calculated based on a NOAEL for nervous system damage in mink fed mercury in fish or chow. The
avian WC was based on the lowest adverse effect level for behavioral and reproductive effects in three
generations of mallard ducks fed mercury in grain. Existing data were not sufficient to calculate a
WC for the Florida panther; however, a NOAEL based on laboratory data from domestic cats was
identified as 20 ug methylmercury/kg bw/d.
The evaluation of data and calculation of WC in this Report was done in accordance with the
methods and assessments published in the Final Water Quality Guidance for the Great Lakes System:
Final Rule (U.S. EPA 1995). Availability of additional data led to differences in calculated values of
the WC in this Report and those published in the final rule. Differences were the result of three
factors. The Report uses more recent data to derive BAF. Second, the final rule appropriately used
some region-specific assumptions that were not used in the nationwide assessment hi the Report, for
example, consumption of herring gulls by eagles. Finally different endpoints were used for the
evaluation of mammals because the purposes of the assessments in the Report and final rule were
different. In the final rule, a risk-management decision was made to base the wildlife criterion on
endpoints likely to influence whole populations (mortality, growth). In this Report a more sensitive
endpoint was selected with the goal of assessing the full range of effects of mercury. The difference
in the results reflects the amount of discretion allowed under Agency Risk Assessment Guidelines.
Table ES-3
Wildlife Criteria for Mercury
Organism
Mink
River otter
Kingfisher
Osprey
Bald eagle
Wildlife Criterion
(pg/L)
415
278
193
483
538
June 1996
ES-9
SAB REVIEW DRAFT
-------
Of the pathways by which ecosystems and components of ecosystems might be exposed to
atmospheric mercury, exposure of high trophic level wildlife to mercury in food is particularly
important The trophic level and feeding habits of an animal influence the degree to which that
species is exposed to mercury. Mercury biomagnifies in aquatic food chains, with the result that
mercury concentrations in tissue increase as trophic levels increase. Predatory animals primarily
associated with aquatic food chains accumulate more mercury than those associated with terrestrial
food chains. Thus, piscivores and their predators generally have the highest exposure to mercury.
Species with high tissue levels of mercury include otter and mink, which are top mammalian predators
of aquatic food chains. Top avian predators of aquatic-based food chains include raptors such as the
osprey and bald eagle.
Although clear causal links have not been established, mercury originating from airborne
deposition may be a contributing factor to population effects on bald eagles, river otters and mink.
Stronger evidence is available to support the possibility of toxic effects on the common loon and the
Florida panther. Effects of mercury originating from point sources on restricted wildlife populations
have been conclusively demonstrated and provide a tissue residue basis for evaluation of risk to other
populations.
Information presented in Volume V of this Report suggests that the ecosystems most at risk
from mercury releases to air exhibit one or more of the following characteristics:
• They are located in areas where exposure to mercury (e.g., atmospheric deposition of
mercury) is high;
• They include surface waters already impacted by acid deposition;
• They possess characteristics other than low pH that result in high levels of
bioaccumulation; and/or
• They include sensitive species.
Conclusions
The following conclusions are presented in approximate order of degree of certainty in the
conclusion, based on the quality of the underlying database. The conclusions progress from
those with greater certainty to those with lesser certainty.
• Inorganic mercury emitted to the atmosphere deposits on watersheds and is
translocated to waterbodies. A variable proportion of this mercury is transformed by
abiotic and biotic chemical reactions to organic derivatives, including methylmercury.
Methylmercury bioaccumulates in individual organisms, biomagnifies in aquatic food
chains and is also the most toxic form of mercury to which wildlife are exposed.
• The proportion of total mercury in biota that exists as methylmercury tends to increase
with trophic level. Greater than 90% of the mercury contained in freshwater fish
exists as methylmercury. Methylmercury accumulates in fish throughout their lifetime,
although changes in concentration as a function of time may be complicated by growth
dilution and changing dietary habits.
June 1996 ES-10 SAB REVIEW DRAFT
-------
Piscivorous avian and mammalian wildlife are exposed to mercury primarily through
consumption of contaminated fish and accumulate mercury to levels above those in
prey items.
Toxic effects on piscivorous avian and mammalian wildlife due to consumption of
contaminated fish have been observed in association with point source releases of
mercury to the environment
Concentrations of mercury in the tissues of wildlife species have been reported at
levels associated with adverse health effects in laboratory studies in the same species.
Piscivorous birds and mammals receive a greater exposure to mercury than any other
known component of aquatic ecosystems.
\
Vield data are highly suggestive of adverseJoxicological effects in common loons due
to accumulation of mercury originating from airborne emissions. Field data are also
suggestive of adverse toxicological effects in the Florida panther due to mercury;
however, this mercury may have originated from both airborne and non-airborne
sources. Field data suggest that bald eagles have not suffered adverse toxic effects due
to airborne mercury emissions. Field data are insufficient to conclude whether the
mink, river otter, or kingfisher have suffered adverse toxic effects due to airborne
mercury emissions.
BAFs for mercury in fish vary widely; however, field data are sufficient to calculate
representative means for different trophic levels. The recommended estimates hi this
Report for BAFs for topic levels 3 and 4 are 66,200 and 335,000, respectively. In
general, BAFs for fish sampled from poorly buffered surface waters are higher than
those for fish obtained from well buffered surface waters.
Based upon knowledge of mercury bioaccumulation in fish, and of feeding rates and
the identity of prey items consumed by piscivorous wildlife, it is possible to rank, the
relative exposure of different piscivorous wildlife species. Of the five wildlife species
selected for detailed analysis, the relative ranking of exposure to mercury is this:
kingfisher > river otter > osprey = mink > bald eagle. Existing data are insufficient to
estimate the exposure of the Florida panther relative to that of the selected species.
Local emissions sources (<50 km from receptors) have the potential to increase the
exposure of piscivorous wildlife well above that due to remote sources (background).
Based upon knowledge of mercury exposure to wildlife and its toxicity in long-term
feeding studies, criterion values can be calculated for the protection of piscivorous
avian and mammalian wildlife. A wildlife criterion value is defined as the
concentration of total mercury in water that, if not exceeded, protects avian and
mammalian wildlife populations from adverse effects resulting from ingestion of
surface waters and from ingestion of aquatic life taken from these surface waters.
The criterion value protective of piscivorous avian wildlife is 405 pg/L.
The criterion value protective of piscivorous mammalian wildlife is 346 pg/L.
June 1996 ES-11 SAB REVIEW DRAFT
-------
• Modeled estimates of mercury concentration in fish around hypothetical mercury
emissions sources predict exposures at the wildlife WC. The wildlife WC, like the
human RfD, is predicted to be a safe dose over a lifetime. It should be noted,
however, that the wildlife effects used as the basis for the WC are gross clinical
manifestations or death. Expression of subtle adverse effects at these doses cannot be
excluded.
There are many uncertainties associated with this analysis, due to an incomplete understanding
of the toxicitv of mercury and mercury compounds. The sources of uncertainty include the
following:
• Variability in the calculated BAFs is a source of uncertainty. BAFs given in this
Report relate total mercury in fish (most of which exists as methylmercury) to total
mercury in the Vvater column. Methylmercury is the bioaccumulating species of
\ mercury, but existing data are insufficient to estimate BAFs on a methylmercury basis.
.Methods for the speciation of mercury hi environmental samples are rapidly improving
but remain difficult to perform. Questions also remain concerning the bioavailability
of methylmercury associated with paniculate and dissolved organic material. Local
biogeochemical factors that determine net methylation rates are not fully understood
and are not amenable at this time to generalized modeling.
• The representativeness of field data used in establishing the BAFs is a source of
uncertainty. The degree to which the analysis is skewed by the existing data set is
unknown. A disproportionate amount of data is from north-central and northeastern
lakes. The applicability of these data to a national assessment is not known.
• Limitations of the toxirity database present a source of uncertainty. Few controlled
studies of quantifiable effects of mercury exposure in wildlife are available. These are
limited to few species, necessitating the use of uncertainty factors in extrapolating to
species of interest The toxic endpoints reported in existing studies can be considered
severe, raising questions as to the degree of protection against subtle effects offered by
reference doses and water criteria calculated on those endpoints. tlse of less than
lifetime studies for prediction of effects from lifetime exposure is a source of
uncertainty.
• Concern has been raised regarding the possibility of toxic effects in species other than
those piscivorous birds and mammals evaluated in this Report. In particular there is
considerable uncertainty about mercury effects hi biota at trophic levels 1 and 2 in
aquatic ecosystems and about effects in terrestrial systems.
• Lack of knowledge of wildlife feeding habits is a source of uncertainty. Existing
information frequently is anecdotal or confined to evaluations of a particular locality;
the extent to which this information is generalizable is open to question. In some
instances wherein feeding habits are relatively well characterized (e.g., Florida
panther), the extent of mercury contamination of prey is poorly known (e.g., in
raccoons).
• While the methods used to develop wildlife criteria are based on effects in individual
organisms, the stated goal of the assessment is to characterize the potential for adverse
effects in wildlife populations. Factors that contribute to uncertainty in population-
June 1996 ES-12 • SAB REVIEW DRAFT
-------
based assessments include these: variability in the relationship between individuals
and populations; variability in fecundity; lack of data on carrying capacity; and
relationships of one population, of the same or different species, to another population.
• A focus on populations may not always be appropriate. This could be true for
endangered species, which may be highly dependent for the survival of the species on
the health of a few individuals. This may also be true for some regional or local
populations of widespread species; the local population may be "endangered" and thus
dependent on the survival of individuals.
To improve the ecological risk assessment for mercury and mercury compounds. U.S. EPA
would need the following:
• Mechanistic research is needed for better understanding of variability of mercury
effects. This would include studies on the following: factors determining rates of
methylation and demethylation; dietary absorption efficiencies from natural food
sources; effects of dietary choice; and bioavailability of methylmercury in the presence
of dissolved organic material and other material that could bind mercury.
• Data are needed for better definition of adverse effects on the species that were
evaluated in this Report Information is also lacking on species at trophic levels 1 and
2.
• Efforts to develop and standardize methods for analysis of total mercury and
methylmercury in environmental samples (including animal and plant tissue) remain
important.
• The current wildlife criteria are based on linear, four-tiered food chains. Research on
the appropriateness of this design and information that will improve the model are
important.
• Research is needed to reduce uncertainty regarding the accumulation of mercury at
lower trophic levels.
• High quality field data will be useful to support the process-based research described
above, as well as to determine residue concentrations in fish and other aquatic biota
consumed by wildlife.
• There is a need to collect additional natural history data for macroinvetebrates and
amphibians. Seasonal and spatial effects on predation should be explored and
described.
Based on the extant data and knowledge of developing studies the U.S. EPA predicts the
following:
• "Regions of concern" are defined as those geographic areas in the contiguous U.S. that
are thought to receive high levels of mercury deposition and that contain relatively
large numbers (>5% below pH 5.5) of poorly buffered surface waters. The designation
of an area as a region of concern implies an increased risk of mercury toxicity to
June 1996 ES-13 SAB REVIEW DRAFT
-------
wildlife. This designation could be used to define critical habitat, identify wildlife
populations potentially at risk and provide a focus for future research.
• Increased deposition will lead to increased levels in fish.
• Increased levels in fish will lead to toxicity in piscivorous birds and mammals.
• These impacts are most likely to occur in areas that receive high levels of deposition
and that also contain poorly buffered surface waters.
June 1996 ES-14 SAB REVIEW DRAFT
-------
1. INTRODUCTION
Section 112(n)(l)(B) of the Clean Air Act (CAA), as amended in 1990, requires the U.S.
Environmental Protection Agency (U.S. EPA) to submit a study on atmospheric mercury emissions to
Congress. The sources of emissions that must be studied include electric utility steam generating
units, municipal waste combustion units and other sources, including area sources. Congress directed
that the Mercury Study evaluate the rate and mass of mercury emissions, health and environmental
effects, technologies to control such emissions and the costs of such controls.
•
In response to this mandate, U.S. EPA has prepared a seven-volume Mercury Study Report to
Congress. The seven volumes are as follows:
I. Executive Summary
II. An Inventory of Anthropogenic Mercury Emissions in the United States
IE. An Assessment of Exposure from Anthropogenic Mercury Emissions in the United
States
IV. Health Effects of Mercury and Mercury Compounds
V. An Ecological Assessment for Anthropogenic Mercury Emissions in the United States
VI. Characterization of Human Health and Wildlife Risks from Anthropogenic Mercury
Emissions in the United States
VII. An Evaluation of Mercury Control Technologies and Costs
This volume (Volume V) comprises an ecological assessment for airborne mercury emissions.
It provides an overview of the ecological effects of mercury, uses modeling predictions from Volume
m (Exposure Assessment) on airborne mercury concentrations and deposition rates to assess potential
ecological exposures, and reviews available toxicity and bioaccumulation data for the purpose of
developing a criterion for the protection of sensitive wildlife species.
Volume V is composed of three main sections, organized by a format provided by U.S. EPA's
Framework for Ecological Risk Assessment (U.S. EPA, 1992a). Chapter 2 corresponds to the problem
formulation phase of the assessment and reviews the potential ecological impacts of mercury. Based
upon this information it is concluded that piscivorous avian and mammalian wildlife are potentially at
risk due to airborne mercury emissions. A conceptual model is presented to describe how airborne
mercury becomes concentrated in aquatic biota which serve in turn as the primary food source for
piscivorous wildlife. An exposure analysis is conducted in Chapter 3. Effects are analyzed in Chapter
4, culminating in calculation of a criterion value for protection of piscivorous wildlife. Chapter 5
discusses further research needs. References are provided at the end of the volume. An ecological
risk characterization for mercury is developed separately in Volume VI of this Report.
The scope of this assessment is limited to consideration of only that mercury which is emitted
directly to the atmosphere. The origins and extent of these emissions are reviewed in Volume II of
this Report This analysis does not address mercury originating from mine leachate, the manufacturing
and disposal of batteries, dental amalgam (in municipal wastewater), or the application of mercurial
pesticides. In a number of instances, these and other "point" sources have been related to
unacceptably high mercury levels in fish, triggering site-specific fish consumption advisories. Clearly,
where such point sources exist, there is a need to address the combined impacts of mercury originating
from all sources, including air emissions.
June 1996 1-1 SAB REVIEW DRAFT
-------
The exposure analysis for piscivorous wildlife was designed to address the following
questions.
• What is the current degree of exposure of piscivorous avian and mammalian wildlife?
• In what broad geographical areas of the continental United States is there a high
probability for co-occurrence of high mercury deposition rates and wildlife species of
concern?
•
• What is the relative increase in exposure that can be anticipated for wildlife species
that live in proximity to mercury emissions sources?
• What is the relative ranking of source categories with regard to their contribution to
mercury concentrations in fish consumed by piscivorous wildlife?
The first of these questions was addressed by defining in detail what piscivorous wildlife eat
and then characterizing the mercury content of these food items. The second question was addressed
by superimposing the results of a long-range transport analysis onto wildlife distribution information.
The last two questions were addressed by using the results of a local-scale air dispersion model,
combined with an indirect exposure methodology, to generate hypothetical exposure scenarios for
wildlife. This short-range analysis'is analogous to that used in the human health exposure assessment
(Volume m). Descriptions of the long- and short-range air dispersion models and the indirect
exposure methodology are provided in Appendix D to Volume m.
The goal of the effects analysis was to calculate a water-based wildlife criterion value for
mercury which, if not exceeded, would be protective of piscivorous avian and mammalian wildlife.
An effort was then made to calculate fish residue concentrations corresponding to this criterion value.
These residue values are compared in Volume VI with measured values obtained from environmental
sampling efforts.
Owing to its importance for both the ecological and human health assessments, an effort was
made to re-evaluate and calculate bioaccumulation factors (BAFs) for mercury in fish and to
characterize the uncertainties associated with this estimate. The data and methods used to derive these
BAFs are presented in a separate Appendix (Appendix A). A summary of this material is provided in
Section 4.1 as an input to the wildlife criterion development effort.
June 1996 1-2 SAB REVIEW DRAFT
-------
2. PROBLEM FORMULATION
U.S. EPA defines ecological risk assessment as "a process that evaluates the likelihood that
adverse ecological effects may occur or are occurring as a result of exposure to one or more stressors"
(U.S. EPA, 1992a). A "stressor" is defined as any chemical, biological, or physical entity that causes
adverse effects on ecological components, i.e., individuals, populations, communities, or ecosystems.
Although ecological risk assessment follows the same basic risk paradigm as the human health risk
assessment, there are three key differences between the two types.
• Ecological risk assessment can consider effects on populations, communities and
ecosystems in addition to effects on individuals of a single species.
• No single set of ecological values to be protected is applicable in all cases; instead,
they must be selected for each assessment based on both scientific and societal merit
• Nonchemical stressors (e.g., physical disturbances) often need to be evaluated as well
as chemical stressors.
The problem formulation phase of an environmental risk assessment consists of three main
components: (1) characterizing the stressors, potential exposure pathways, ecosystems potentially at
risk, and ecological effects; (2) selecting endpoints (the ecological values to be protected); and (3)
developing a conceptual model of the problem (U.S. EPA, 1992a).
Section 2.1 reviews the characteristics of mercury in the environment, including its various
chemical forms (speciation), chemical transformations and movement within and between the air,
surface water, and soil compartments of the environment (cycling). Section 2.2 identifies the exposure
pathways by which plants and animals can be exposed to mercury in both aquatic and terrestrial
ecosystems. Section 2.3 provides an overview of what is known about the effects of mercury on
organisms, populations, communities and ecosystems. Section 2.4 identifies ecosystems and ecosystem
components that are thought to be most at risk from mercury in the environment. Section 2.5
describes the selection of assessment and measurement endpoints for the ecological risk assessment. A
conceptual model of mercury fate and effects in the environment is presented in Section 2.6, setting
the framework for the risk assessment that follows.
It should be noted that this review of mercury fate and effects is limited to consideration of
only terrestrial and freshwater aquatic ecosystems. It is recognized that mercury that deposits in
coastal areas can be translocated to estuarine environments, and that biota which inhabit these and
nearby marine systems have the potential to be adversely impacted. Presently, however, uncertainties
regarding mercury deposition, cycling, and effects in such environments are so great as to preclude
even a qualitative risk assessment.
2.1 Stressor Characteristics: Mercury Speciation and Cycling
Mercury in the environment can occur in various physical and chemical forms. Physically,
mercury may exist as a gas or liquid, or it may be associated with solid particulates. Chemically,
mercury can exist in three oxidation states:
(1) Hg - elemental mercury, also called metallic mercury;
June 1996 2-1 SAB REVIEW DRAFT
-------
(2) Hg22"1' - mercurous ion (monovalent mercury, mercury I); or
(3) Hg2+ - mercury II (mercuric ion, divalent mercury).
Mercury also reacts with other chemicals to form inorganic compounds (e.g., HgCl2 - mercuric
chloride) and organic compounds (e.g., CH3Hg+ - monomethylmercury, (CH3)2Hg - dimethylmercury,
C6H5HgCl - phenyl mercuric chloride). Figure 2-1 illustrates the major transformations between these
different forms in the environment Dimethylmercury is highly volatile and dissociates to
monomethylmercury at neutral or acid pH (pH < 8) (Huckabee et al., 1979). In contrast,
monomethylmercury is stable and tends to accumulate in living organisms as described below (Bloom,
1992). Throughout this volume, monomethylmercury is referred to simply as methylmercury.
Figure 2-1
Cycling of Mercury in Freshwater Lakes (adapted from Winfrey and Rudd, 1990)
CH,HgCH,
XXXXXXXXXXXXXXXXXX
xxxxxxxxxxxxxxxxxx
xxxxxxxxxxx
X X
X
xxxxxxxxxx
xxxxxxxxxxxxxx x^>«*x
xxxxxxxxxxxx X-^*v xxx
XXXXXXX''/'
,,,,,,,,. . „ . '.'XXX X™' ' XXXXXXXXXXXXXXXXX
x x x x x x .Organic & Inorganic. , x x ^HgS.
Complexes
x x x x x x
xxxxxxxxxxxxxxxxxxxxxxx
As discussed in the box below, methylation is an important step in the mercury cycle that
strongly influences the ecological fate and effects of mercury. Methylmercury is readily accumulated
by fish due, in part, to efficient uptake from dietary sources and to low rates of elimination (Bloom,
1992). It is also the most toxic form of mercury to wildlife (Eisler, 1987).
Mercury cycling and partitioning in the environment are complex phenomena which depend on
numerous environmental parameters. The following sections provide a brief overview of mercury
speciation and partitioning in the atmosphere, surface water and soil, including information from
specific case studies. For a detailed review, see Volume III of this Report to Congress.
June 1996
2-2
SAB REVIEW DRAFT
-------
FOCUS ON METHYLMERCURY
Methylmercury is the form of mercury of particular concern in ecosystems for three reasons.
(1) All forms of mercury can be methylated by natural processes in the environment
(2) Methylmercury bioaccumulates and biomagnifies in aquatic food webs at higher rates and to a greater
extent than any other form of mercury.
(3) Methylmercury is the most toxic form of mercury.
All forms of mercury discharged into surface waters can be converted to methylmercury by natural
processes. In the 1960s, researchers found methylmercury in fish in Swedish lakes, although no discharge of
methylmercury had occurred in those lakes (Bakir et al., 1973). Later research determined that the methylation of
mercury in sediments by anaerobic sulfur-reducing bacteria was a major source of methylmercury in many aquatic
environments (Gilmour and Henry, 1991; Zillioux et aL, 1993). Aerobic bacteria and fungi, including yeasts that
grow best in acid conditions, also can methylate mercury (Eisler, 19S7; Yannai et al., 1991). In addition, fulvic
and humic material may abiotically methylate mercury-(Nagase et al., 1984; Lee et al., 1985;. Weber, 1993). The
major site of methylation in aquatic systems is the sediment, but methylation also occurs in the water column
(Wright and Hamilton, 1982; Xun et aL, 1987; Parks et al., 1989; Bloom and Effler, 1990; Winfrey and Rudd,
1990; Bloom et al., 1991; Gilmour and Henry, 1991; Miskimmin et al., 1992). The rate of mercury methylation
varies with microbial activity, mercury loadings, suspended sediment load, nutrient content, pH and redox
conditions, temperature, and other variables. The net rate of mercury methylation in an ecosystem is determined
by competing rates of methylation and demethylation.
Methylmercury bioaccumulates and biomagnifies in aquatic food webs at higher rates and to a greater
extent than any other form of mercury (Watras and Bloom, 1992). "Bioaccumulation" refers to the net uptake of
a contaminant from the environment into biological tissue via all pathways. It includes the accumulation that may
occur by direct contact of skin or gills with mercury-contaminated water as well as ingestion of mercury-
contaminated food. "Biomagnification" refers to the increase in chemical concentration in organisms at
successively higher trophic levels in a food chain as a result of the ingestion of contaminated organisms at lower
trophic levels. Methylmercury can comprise from 10 percent to over 90 percent of the total mercury in
phytoplankton and zooplankton (trophic levels 1 and 2)(May et al., 1987; Watras and Bloom, 1992), but generally
comprises over 90 percent of the total mercury in fish (trophic levels 3 and 4XHuckabee et al., 1979; Grieb et al.,
1990; Bloom, 1992; Watras and Bloom, 1992). Fish absorb methylmercury efficiently from dietary sources and
store this material in organs and tissues. The biological half-life of methylmercury in fish is difficult to determine
but is generally thought to range from months to years.
Methylmercury is the most hazardous form of mercury to birds, mammals, and aquatic organisms due to
its high stability and strong affinity for sulfur-containing organic compounds (e.g., proteins). Biological
membranes, including the blood-brain barrier and the placenta, that tend to discriminate against other forms of
mercury allow relatively easy passage of methylmercury and dissolved mercury vapor (Eisler, 1987).
Methylmercury can cause death, neurological disorders, organ damage, impaired immune response, impaired
growth and development and reduced reproductive success (Klaassen, 1986). In mammals, fetuses are particularly
sensitive to mercury, experiencing deleterious developmental effects when the mothers appear to be unaffected
(Clarkson, 1990).
2.1.1 Mercury in Air
In the atmosphere, most mercury (95 to over 99 percent) exists as gaseous Hg°; the remainder
generally is comprised of gaseous methylmercury (-0 to 5 percent) and mercury associated with
particulates (Lindqvist, 1991; MDNR, 1993). Gaseous Hg2"1" also may exist in air, especially near
mercury emissions sources. Mercury associated with particulates in air includes Hg2"1", thought to
occur primarily as mercuric chloride (MDNR, 1993).
June 1996 2-3 SAB REVIEW DRAFT
-------
The form of mercury in air affects both the rate and mechanism by which it deposits to earth.
Kg2"*" and methylmercury are more likely to be deposited than Hg° because they are more soluble and
are scavenged by precipitation more easily. Methylmercury is also thought to be dry-deposited more
effectively than Hg°. As a result, although methylmercury and Hg2+ generally comprise less than five
percent of mercury in the atmosphere, they are thought to comprise a higher proportion of deposited
mercury (Lindqvist, 1991).
Wet deposition apparently is the primary mechanism for transporting mercury from the
atmosphere to surface waters and land (Lindqvist, 1991). In the Great Lakes area, for example, wet
deposition is believed to account for 60 to 70 percent of total mercury deposition. Hg2"1" is the
predominant form in precipitation (MDNR, 1993).
2.1.2 Mercury in Surface Water
Mercury can enter surface water as Hg°, Hg2*, or methylmercury. Once in aquatic systems,
mercury can exist in dissolved or paniculate forms, and can undergo the following transformations
(See Figure 2-1) (Lindqvist et al., 1991; Winfrey and Rudd, 1990).
• Hg2"1" in surface waters can be reduced to Hg°.
• Volatile Hg° in surface waters can be released to the atmosphere.
• Atmospheric Hg° may be oxidized to form Hg2"1", and may be deposited/ redeposited to
surface waters.
• Hg2"1" can be methylated in sediments and the water column to form methylmercury.
Each of these reactions can also occur in the reverse directioa The net rate of production of each
mercury species is determined by the balance between forward and reverse reactions.
Estimates of the percent of total mercury in surface waters that exists as methylmercury vary.
Generally, methylmercury makes up less than 20 percent of the total mercury in the water column of
lakes (Kudo et al., 1982; Parks et al., 1989; Bloom and Effler, 1990). In lakes without point
discharges, methylmercury frequently comprises less than ten percent of the total mercury in the water
column (Lee and Hultberg, 1990; Lindqvist, 1991; Porcella et al., 1991; Watras and Bloom, 1992).
Contaminated sediments can serve as an important mercury reservoir, with sediment-bound
mercury recycling back into the aquatic ecosystem for decades or longer. Biological processes affect
this recycling process. For example, bacterial activity affects the transformation of one mercury form
to another (e.g., sulfate-reducing bacteria mediate mercury methylation) (Gilmour and Henry, 1991).
Benthic invertebrates may take up mercury from sediments, making it available to other aquatic
animals through the food chain, and to vertebrates that consume emergent aquatic insects (Hildebrand
et al., 1980; Wren and Stephenson, 1991; Dukerschein et al., 1992; Saouter et al., 1993; Suchanek et
al., 1993). Physical factors, such as reduced pH, stimulate methylmercury production at the
sediment/water interface and thus may accelerate the rate of mercury methylation resulting in increased
accumulation by aquatic organisms (Figure 2-1) (Winfrey and Rudd, 1990).
June 1996 2-4 SAB REVIEW DRAFT
-------
2.1.3 Mercury in Soil
Mercury deposited from the air forms stable complexes with soil particles of high organic or
sulfur content and with humic and fulvic acids (Andersson, 1979; WHO, 1989; Johansson et al., 1991).
These chemical bonds limit mercury's mobility in soils and its availability for uptake by living
organisms. In general, the distribution of mercury in soil is likely to follow the distribution of organic
matter. Mercury has a long retention time in soils. As a result, mercury that has accumulated in soils
may continue to be released to surface waters for long periods of time, possibly hundreds of years
(Johansson et al., 1991)
Hg2"1" in soils can be transformed to other mercury species. Bacteria and organic substances
can reduce Hg2"1" to Hg°, releasing volatile inorganic mercury to the atmosphere. Alternatively,
bacteria and organic substances can methylate mercury, and subsequently demethylate it, depending on
environmental conditions (Allard and Arsenic, 1991; Gilmour and Henry, 1991).
2.2 Potential Exposure Pathways
Plants and animals can be exposed to mercury by direct contact with contaminated
environmental media or ingestion of mercury-contaminated water and food (Figure 2-2). Mercury
deposited in soil can be a source of direct exposure from physical contact (e.g., earthworms, terrestrial
plants). Animals also can ingest mercury in soil, either purposefully (e.g., earthworms) or incidentally
(e.g., grazers). Mercury in the air can be taken up directly by terrestrial or aquatic emergent plants, or
inhaled by terrestrial animals. Mercury in water can be a source of direct exposure to aquatic plants
(e.g., algae, seagrasses) and aquatic animals (e.g., zooplankton, fish) and can be ingested by terrestrial
animals in drinking water (e.g., moose). Finally, both aquatic and terrestrial animals can be exposed
to mercury in contaminated food sources (e.g., fish, piscivorous mammals and birds).
Not all of these potential exposure pathways are equally important, however. The remainder
of this section evaluates the likely importance of different routes of exposure consequent to mercury
release to air. Section 2.2.1 discusses the fate and bioavailability of mercury in aquatic systems and
the pathways by which aquatic plants and animals can be exposed to mercury directly in contaminated
water or indirectly through aquatic food webs. Section 2.22 provides information on the fate and
bioavailability of mercury in terrestrial ecosystems and the pathways by which terrestrial plants and
animals can be exposed. Bioaccumulation of mercury in aquatic and terrestrial organisms is discussed
further in Section 2.3.1.
2.2.1 Exposure Pathways in Aquatic Systems
Figure 2-3 illustrates the potential distribution of mercury in a water body. As shown,
mercury can be present in surface waters in various forms: (1) dissolved in the water; (2) concentrated
in the surface microlayer (the uppermost layer of a surface water); (3) attached to seston1; (4) in the
bottom sediments; and (5) in biota (e.g., fish, macroinvertebrates2).
Seston is suspended paniculate matter, including detritus (dead organic matter) and plankton (i.e., living
plants and animals that passively float or weakly swim in the water column such as algae, water fleas, and
copepods).
2
Macroinvertebrates are invertebrates (i.e., animals without backbones) that are visible to the naked eye,
such as worms, clams, snails, insects and insect larvae, and crayfish.
June 1996 2-5 SAB REVIEW DRAFT
-------
Figure 2-2
Possible Routes of Exposure to Mercury
Mercury Transported
o.o 1,1,1,1,1
Atmospheric Deposition of Mercury
Mercury Transferred
Up Aquatic Food Wrts
Uptake by
Uptake by upi«» oy
BanthfcAnimate Aquatic Plants
C54054-1-1
The form and location of mercury inea water body determines its bioavailability. For example,
dissolved mercury is available for direct uptake by aquatic plants, fish and invertebrates. Mercury that
concentrates in the surface microlayer is available to organisms that live, reproduce, or feed on the
surface of water bodies (i.e., neuston). Mercury attached to seston can be ingested by aquatic animals
that feed on plankton. Additionally, mercury that has deposited on the accumulated in the sediments
is available to benthic (i.e., bottom-dwelling) plants and animals.
Aquatic plants may take up mercury from air, water or sediments (Crowder, 1991). Planktonic
plants (i.e., phytoplankton such as algae) are not rooted; therefore, their only route of exposure is
uptake from water. Both submerged aquatic vegetation and wetland emergent plants are rooted,
however, and can be exposed to mercury in sediments. In locations with mercury-contaminated
sediments, mercury levels in aquatic macrophytes3 have been measured at 0.01 ug/g, indicating that
these plants do not strongly accumulate mercury from sediments (Wells et al., 1980; Crowder et al.,
1988). The ratio between inorganic and organic mercury varies in plants (Crowder, 1991).
Macrophytes are aquatic plants that are large enough to be visible to the naked eye.
June 1996 2-6 SAB REVIEW DRAFT
-------
Figure 2-3
Distribution of Mercury in a Water Body
Dissolves (available to
plants and animals)
Accumulates in fish and
invertebrates (available in
the food web to high*
trophic levels)
Concentrates in
surface microlayer
(concentrated levels
available to plants
and animals)
Methylated by bacteria
in sediments (methylmercury -
highly toxic to animals - becomes
available to biota)
Attaches to seeton (plankton and
suspended detritus); some settles
to bottom sediments (available
to bottom-dwellers), some eaten
(available in the food web to
higher trophic levels)
C54054-1-2
For aquatic animals, the primary exposure routes of concern are direct contact with mercury-
contaminated water and sediments, and ingestion of mercury-contaminated food. Fish can absorb
mercury through the gills, skin and gastrointestinal tract (U.S. EPA, 1985). These fish then become a
source of mercury for piscivorous birds and mammals. Emergent aquatic insects represent another
potential source of mercury for insectivorous birds and mammals (Dukerschein et al., 1992; Saouter et
al., 1993).
As discussed in more detail in Section 2.3, mercury in aquatic biota tends to occur at higher
concentrations in higher trophic levels. An example aquatic food web is shown in Figure 2-4. At the
top trophic levels are piscivores, such as humans, bald eagles, cormorants, herring gulls, loons,
kingfishers, mergansers, herons, egrets, ospreys, bald eagles, river otters, mink, alligators, snapping
turtles and water snakes. The largest of these species (e.g., bald eagle, otter) can prey on fish that
occupy high trophic levels, such as trout and salmon, which in turn feed on smaller "forage" fish, such
as smelt, alewife, minnow, chub, and sculpin. Smaller piscivorous wildlife (e.g., kingfishers, ospreys,
terns) tend to feed on the smaller forage fish, which in turn feed on zooplankton or benthic
invertebrates. Zooplankton (e.g., copepods, water fleas) feed on phytoplankton (i.e., microscopic
algae), and the smaller benthic invertebrates tend to feed on algae and detritus. Thus mercury can be
June 1996
2-7
SAB REVIEW DRAFT
-------
Figure 2-4
Example Aquatic Food Web
Bottom
Feeders
Bacteria
and Fungi
Submerged ^ ^s^,
Aquatic Vegetation ^v4
Emergent
Vegetation
Sunlight
Mineral Nutrients.
Dead Plants
and Animals
transferred and accumulated through three or four trophic levels to reach the prey of piscivorous
wildlife species. In some large lakes, food chains can be even longer.
2.2.2 Exposure Pathways in Terrestrial Systems
Several exposure pathways are possible for both plants and animals in terrestrial systems. The
two main pathways by which terrestrial plants can be exposed to mercury are uptake from soils into
June 1996
2-8
SAB REVIEW DRAFT
-------
the roots and intake from the air via stomata4 or passive absorption. Potential exposure routes for
terrestrial animals include the following: (1) ingestion of mercury-contaminated food; (2) direct
contact with a contaminated medium; (3) ingestion of mercury-contaminated drinking water; and (4)
inhalation. Food ingestion is of primary concern for vertebrate carnivores (including humans) because
mercury accumulates in prey species. Once mercury enters a terrestrial food web, like that shown in
Figure 2-5, it can be transferred in increasing concentrations to higher trophic levels (Talmage and
Walton, 1993).
Figure 2-5
Example Terrestrial Food Web
2.2.2.1 Terrestrial Plants
Uptake by plants plays a major role in the entry of metals to terrestrial food webs. Mercury
uptake by terrestrial vascular plants5 can occur through the roots and/or through the leaves, by way of
stomata and by passive absorption of gaseous mercury (Mosbaek et al., 1988; Crowder, 1991; Maserti
and Ferrara, 1991). A vascular plant's uptake of mercury from the soil depends on soil type,
Stomata (plural of stoma) are the minute openings in the epidermis of leaves, stems, or other plant organs
that allow gas to diffuse in and out.
Plants with roots, stems, and leaves, such as ferns and seed plants.
June 1996 2-9 SAB REVIEW DRAFT
-------
decreasing as organic matter, which binds mercury, increases (WHO, 1989). Uptake of mercury
through leaves is considered to be a negligible source of mercury for beech and spruce (Schmidt,
1987), but is an important route for pine and herbaceous plants (Mosbaek et al., 1988; Maserti and
Ferrara, 1991). Bryophytes and lichens, on the other hand, have no roots and take up metals only
from air or water (WHO, 1989; Crowder, 1991). Some species of bryophytes and lichens can
bioconcentrate mercury to relatively high levels (e.g., up to 1200 ug/g hi Sphagnum sp.) (Siegal et al.,
1985). Some woody plants (e.g., Pinus sp.) also bioconcentrate mercury (Siegal et al., 1987).
2.2.2.2 Terrestrial Animals
Dietary exposure is the primary route of mercury exposure for vertebrate members of
terrestrial food webs. Figure 2-5 illustrates a terrestrial food web. Plants are eaten by a wide diversity
of herbivorous animals (e.g., grasshoppers, caterpillars, mice, voles, rabbits, deer). Insects, earthworms
and other soil macroinvertebrates can accumulate mercury to levels well above those of the soil in
which they reside (Siegel et al., 1975; Helmke et al., 1979; Beyer et al., 1985); and are themselves
consumed by many species of birds, shrews, snaKes, and amphibians. Small mammals, birds, reptiles
and amphibians are consumed by larger predators, such as owls, hawks, eagles, mink, wolves, and big
cats. Thus mercury can be transferred and accumulated through two or three trophic levels to reach
the prey of top carnivores in terrestrial systems.
For these terrestrial animals, exposure to mercury depends largely on the animal's feeding
strategy. For example, generalist herbivores (plant-eaters) may be less exposed to mercury than
species that are specialized in or restricted to feeding on highly exposed plant species (e.g., reindeer
foraging mostly on lichens and bryophytes).
2.2.3 Summary of Aquatic and Terrestrial Exposure Pathways
Food chain transfers of mercury are thought to be the most important exposure pathways in
both aquatic and terrestrial ecosystems. Mercury, however, tends to bioaccumulate and biomagnify
more strongly in aquatic than in terrestrial ecosystems. Several possible explanations exist to explain
this observation. First, the transfer of metals to higher trophic levels depends to some extent on where
the metals are stored within prey organisms. Birds and mammals accumulate mercury in their feathers
and fur, which are not eaten or are poorly digested. In contrast, most of the mercury in fish is
contained in muscle tissue, which is consumed and digested by piscivorous wildlife. In addition,
mercury in terrestrial food webs frequently exists in an inorganic form, rather than as methylmercury.
Inorganic mercury accumulates to only a limited extent in plants and soil organisms and does not
biomagnify in the organisms that feed on them. Finally, aquatic food chains often include more
trophic levels than terrestrial food chains. A typical food chain in aquatic systems would consist of
phytoplankton/algae/detritus -» zooplankton/benthic invertebrates -» small forage fish -» larger
piscivorous fish. Piscivorous birds and mammals would represent the fifth step in the chain. In some
cases a sixth step exists, as when a bald eagle consumes a piscivorous herring gull or when a Florida
panther consumes a raccoon. A typical food chain hi terrestrial systems might be plants -» small
herbivorous mammals -» predatory birds and mammals. Another typical terrestrial food chain would
be plants -> herbivorous insects -» small birds —» birds of prey. In these examples, the top predators
represent the third and fourth step in the chain (although additional steps are possible), instead of a
fifth or sixth level as can be the case for aquatic systems.
June 1996 2-10 SAB REVIEW DRAFT.
-------
23 Ecological Effects
This section provides an overview of potential effects of mercury on ecosystems and
components of ecosystems. Contaminants such as mercury can affect individual organisms,
populations, communities, or ecosystems (Table 2-1). Effects on individuals can be lethal or sublethal,
including behavioral, reproductive and developmental effects. Additionally, effects can be immediate,
due to acute (short-term) exposures, or may be manifested only after chronic Gong-term) exposures.
Table 2-1
Examples of Effects of Contaminants on Ecosystem Components
Component
Examples of Effects
Individual
Change in respiration
Change in behavior (e.g., migration, predator-prey interactions)
Inhibition or induction of enzymes
Increased susceptibility to pathogens
Decreased growth
Decreased reproduction
Death
Population
Decreased genotypic and phenotypic diversity
Decreased biomass
Increased mortality rate
Decreased fecundity rate
Decreased recruitment of juveniles
Increased frequency of disease
Decreased yield
Change in age/size class structure
Extinction
Community
Decreased species diversity
Change in species composition
Decreased food web diversity
Decreased productivity.
Increased algal blooms
Ecosystem
Decreased diversity of communities
Altered nutrient cycling
Decreased resilience
In animals, toxic effects caused by mercury exposure vary depending on a number of factors,
including but not limited to these:
• delivered dose (i.e., amount and duration of exposure);
• the form of mercury to which an organism is exposed;
• physical and chemical parameters of the environment (e.g., pH, temperature);
• the extent to which an organism is exposed to other chemicals or non-chemical
stressors; and
June 1996
2-11
SAB REVIEW DRAFT
-------
• the life stage, age, sex, species, and physiological condition of the exposed organisms.
The remainder of this section provides an overview of potential adverse ecological effects of mercury.
Section 2.3.1 discusses the bioaccumulation and biomagnification of mercury in food chains, Section
2.3.2 reviews individual-level effects, Section 2.3.3 reviews population-level effects, and Section 2.3.4
reviews effects on communities and ecosystems.
2.3.1 Bioaccumulation of Mercury
As discussed previously, plants and animals may absorb mercury from direct exposure to
contaminated media (e.g., water, soil, air). In addition, animals can acquire mercury through ingestion
of mercury-contaminated food. These pathways determine how much mercury an organism is exposed
to from outside sources. An additional factor that determines the effect of mercury on ecological
systems is how much mercury is accumulated by organisms. Mercury accumulation can result in
concentrations' in exposed plants and animals that are higher than those in surrounding media or in
ingested food. This section outlines the basic processes by which mercury accumulates and introduces
the different ways that chemical accumulation in biological systems is measured and expressed.
Three separate terms are commonly used to describe the mechanism by which a contaminant
accumulates in living tissues. The term "bioconcentration" is used to refer to the uptake of a chemical
directly from an organism's surrounding medium (e.g., uptake by a fish from water through the gills
and epithelial tissue, or uptake by earthworms from soil through the skin) and does not include the
ingestion of contaminated food. The term "bioaccumulation" refers to the net uptake of a contaminant
from all possible pathways. It includes the accumulation that may occur by direct exposure to
contaminated media as well as exposure from food. The term "biomagnification" refers to the increase
in chemical concentration in organisms at successively higher trophic levels as a result of the ingestion
of contaminated organisms at lower trophic levels. Mercury is known to bioconcentrate,
bioaccumulate and biomagnify. In fact, mercury is one of the few metals that is known to biomagnify
in aquatic food webs.
Different numerical factors are used to estimate the extent to which a contaminant
bioconcentrates, bioaccumulates and biomagnifies.
• The bioconcentration factor (BCF) is the ratio of a substance's concentration in tissues
(generally expressed on a whole-body basis) to its concentration in the surrounding
medium (e.g., water, soil) in situations where an organism is exposed through the
medium only.
• The bioaccumulation factor (BAF) is the ratio of a substance's concentration in tissue
to its concentration in the surrounding medium (e.g., water, soil), in situations where
the organism is exposed both directly and through the food web.
• The biota-sediment accumulation factor (BSAF) is a specialized form of the BAF
which refers to the chemical concentration in an aquatic organism divided by that in
surficial bottom sediments. To date it has been applied only to bioaccumulative
organic compounds, but in principal it could be applied to mercury also. When
applied to organic compounds, chemical concentrations in tissues and sediment are
generally normalized for lipid content and organic carbon content, respectively.
June 1996 2-12 SAB REVIEW DRAFT
-------
• The predator-prey factor (PPF, also known as the biomagnification factor, or BMP) is
the factor by which a substance's concentration in the organisms at one trophic level
exceeds the concentration in the next lower trophic level. For example, the PPF for
mercury at trophic level 4 equals the observed mercury concentration in trophic level 4
organisms divided by mercury concentration in trophic level 3 organisms.
• The food chain multiplier (FCM) is the factor by which the BAF of a substance at
trophic level 2 or higher exceeds the BCF at trophic level 1. Implied by this definition
is the assumption that organisms at trophic level 1 are at or near chemical equilibrium
with their environment
BAF, BSAF, PPF and FCM values are trophic-level specific. Depending on environmental
levels of mercury, sufficient mercury may accumulate in organisms at one or more trophic levels to
produce adverse effects at the individual, population, community or ecosystem level.
Mercury accumulates in an organism when the rate of uptake exceeds the rate of elimination.
Although all forms of mercury can accumulate to some degree, methylmercury generally accumulates
to a greater extent than other forms of mercury. Methylmercury is absorbed into tissues quickly and
becomes sequestered due to covalent reactions with sulfhydryl groups in proteins and other
macromolecules. Inorganic mercury can also be absorbed but is generally taken up at a slower rate
and with lower efficiency than methylmercury (Eisler, 1987). Elimination of methylmercury takes.
place very slowly resulting in tissue half-lives (the time required for half of the mercury in the tissue
to be eliminated) ranging from months to years (Westermark et al., 1975). Elimination of
methylmercury from fish is so slow that long-term reductions of mercury concentrations in fish are
often due mainly to growth of the fish. In comparison, other mercury compounds are eliminated
relatively quickly resulting hi reduced levels of accumulation (Eisler, 1987).
Methylmercury and total mercury concentrations both tend to increase in aquatic organisms as
trophic level increases in aquatic food webs. In addition, methylmercury generally comprises a
relatively greater percentage of the total mercury content at higher trophic levels (May et al., 1987;
Watras and Bloom, 1992). Accordingly, mercury exposure and accumulation is of particular concern
for animals at the highest trophic levels hi aquatic food webs and for animals that feed on these
organisms.
2.3.1.1 Field-derived BAFs, BSAFs, and PPFs
In this section, BAFs for organisms that occupy the base of aquatic food chains are reviewed,
along with BSAFs for fish and PPFs for avian and mammalian piscivores. BAFs for earthworms and
emergent aquatic insects are also presented because both represent possible vectors for mobilization of
sediment-associated mercury and subsequent translocation to wildlife. BAFs derived for fish are
discussed in detail in Appendix A. Derivation is summarized in Section 4.1.
The only known data from which to estimate BAFs for phytoplankton were reported by
Suchanek et al. (1993) for Clear Lake, California. Averaged across seven sampling sites, the BAF for
total mercury was 100,880, which was about 14% greater than the corresponding value for
zooplankton from the same system (see below). In contrast, the BAF for methylmercury in
phytoplankton was 477,300, which was approximately one-third that for zooplankton (1,273,300).
June 1996 2-13 SAB REVIEW DRAFT
-------
BAFs published for zooplankton, expressed as ratios of total mercury, range from
approximately 35,600 (Sorenson et al., 1990) to 285,200 (Lindqvist, 1991). Intermediate values
include those reported by Watras and Bloom (1992), approximately 56,200 for the reference basin, and
Suchanek et al. (1993), average of 87,260 across sampling sites. This broad range of values typifies
BAF data for aquatic systems generally, and is thought to be due largely to natural variation in the
percentage of total mercury that exists as methylmercury, the principal bioaccumulating species.
Fewer data exist with which to estimate BAFs for zooplankton on a methylmercury basis. Watras and
Bloom (1992) reported a BAF for methylmercury of approximately 1,000,000, which is similar to the
value (1,273,300) obtained by Suchanek et al. (1993).
To date, BSAFs for mercury have not been estimated; however, limited data exist that allow
such calculations to be made. Hildebrand et al. (1980) observed a linear relationship between mercury
in sediment and that in benthic invertebrates. A BSAF of approximately 0.4 is obtained from the
slope of this relationship (after expressing benthos data on a dry weight basis). The relationship
between mercury in fish (rock bass and hog suckers) and that in sediments was reported by Hildebrand
et al. (1980) to be logarithmic. Taking as an average a fish tissue value of 4.0 ug/g (dry weight;
converted from 1.0 ug/g wet weight) and solving for the sediment concentration yields a value of 2.78
ug/g. The BSAF is equal to the ratio of fish and sediment values, or approximately 1.4. Data
presented by Sorenson et al. (1990) yield BSAFs (dry weight basis) of approximately 2.0 and 10.1 for
zooplankton and northern pike, respectively. Data presented by Wren and MacCrimmon (1986) allow
BSAFs to be calculated for two Ontario lakes that differ considerably with respect to total mercury
residues in biota. In both lakes BSAFs (dry weight basis) appeared to be very similar, ranging from
approximately 5.1 (clams) to 24.0 (northern pike) in the less contaminated of the two lakes, and 3.4
(clams) to 27.1 (pike) in the other system. Using the mid-range of values reported by Lindqvist
(1991), BSAFs (dry weight) of approximately 2.2,17.2, 17.7, and 45.7 are obtained for zooplankton,
macroinvetebrates, yellow perch (small and large), and northern pike (large and small), respectively.
Boyer (1982) reported mercury concentrations in fish and sediments from several locations on the
upper Mississippi River. Expressed on a dry weight basis, these data yield BSAFs ranging from 2.5 to
greater than 50.
In summary, BSAFs calculated for mercury in aquatic biota ranged from 0.4 to about 50, and
appeared to increase with trophic level. In both magnitude and in increase with trophic level, BSAFs
for mercury are similar to BSAFs reported for hydrophobic organic compounds (lipid/carbon
normalized). It could be hypothesized, therefore, that similar processes are at work. This is unlikely,
since bioaccumulation of organic compounds is largely a partitioning process, while for mercury the
chemical interactions tend to more specific, often involving the formation of covalent bonds. Because
mercury does not partition into lipid, normalization for lipid content makes little sense. The existence
of strong relationships between mercury and organic carbon content (see for example Wiener et al.,
1982; Lindqvist, 1991) suggests, however, that some type of carbon normalization may be appropriate.
Limited data are available that allow calculation of BAFs for emergent aquatic insects and
earthworms. Saouter et al. (1993) exposed mayflies for 10 days to methylmercury in sediment and
obtained a BAF (wet weight basis) of 4.0. This value would increase somewhat if it was expressed on
a dry weight basis. The concentration of mercury in earthworms collected from an uncontaminated
field site was 27.1 times that of soil and 6.9 times that of decaying vegetation (dry weight basis;
Siegel et al., 1975). In a 12 wk laboratory exposure, earthworms accumulated an average of 21.3
times the mercury concentration of the soil to which they were exposed (including control and
treatment groups; Beyer et al., 1985).
June 1996 2-14 SAB REVIEW DRAFT
-------
PPFs for piscivorous birds and mammals are difficult to determine accurately because residue
data cannot be attributed with any specificity to residues in a particular prey item; feeding observations
for the species in question are rarely reported in these studies. PPFs were, therefore, estimated by
constructing rough averages of residue values in prey items occupying similar trophic levels. For this
analysis, mink, mergansers, and loons were assumed to feed exclusively at trophic level 3. River
otters were assumed to feed at trophic levels 3 (80%) and 4 (20%). All PPFs were calculated as the
ratio of total mercury concentration in muscle tissue to that in the prey items.
PPFs calculated for piscivorous birds ranged from 1.73 (hooded merganser; Vermeer et al.,
1973) to 7.7 (herring gull; Wren et al., 1983). Intermediate values include those estimated for the
common merganser (2.51; Vermeer et al., 1973) and loon (6.8; Wren et al., 1983). PPFs for mammals
ranged from 1.72 (otter; Wren et al., 1986) to 7.7 (mink; Wren et al., 1983). Kucera (1983) reported
that the ratio of mercury concentrations hi mink and otter to that in predatory fish in the same region
was about 10. Thus, it can be shown that mercury biomagnifies hi piscivorous wildlife, although the
extent of this biomagnification is less than that typically reported for persistent organic compounds.
For example, data reported by Braune and Norstrom (1989) suggest that the PPF for PCBs in
piscivorous birds can approach 100. These observations have led to the suggestion that mercury is
eliminated by piscivorous wildlife in feathers and fur, and perhaps also via a demethylation pathway
(Wren et al., 1986); however, extensive elimination would be expected to result in PPFs of 1 or less.
2.3.1.2 Mercury Residues in Fish
Estimation of an average mercury concentration in freshwater fish at any given trophic level
requires the collection of a large number of samples from randomly selected waterbodies. What, in
fact, generally exists in the literature is a collection of individual studies which characterize mercury
concentrations in a relatively small number of fish from restricted geographical regions. Many of
these studies were initiated because of real or suspected problems with mercury in the region of
interest. Thus, a sample developed by a compilation of these data may be biased toward the high-end
of the distribution of mercury concentrations hi fish nationwide.
The most appropriate source for estimating average concentrations of mercury in freshwater
fish appears to be a study conducted by U.S. EPA (U.S. EPA, 1992b, Bahnick et al., 1994). This is
the only identified nationwide fish collection effort that used consistent sampling and mercury
measurement techniques. Samples were obtained from 279 sites across the U.S., based on proximity
to suspected point and non-point pollution sources. Additionally, fish were collected from 35 "remote"
sites that were thought to provide background pollutant concentrations in fish. Whole-body mercury
levels were determined for bottom feeders, and mercury levels in fillets were analyzed for game fish.
The maximum mercury level detected was 1.8 ug/g wet weight, and the mean across all fish and all
sites was 0.26 jag/g (Table 2-2). The highest values were detected in piscivorous game fish (trophic
level 4), including walleye, bass and northern pike. Lower levels were found in herbivores (carp,
sucker), omnivores (catfish), and species that prey extensively on insects (trout and crappie). In
general, this sampling effort did not address fish that occupy trophic level 3 (forage fish). A
"national" average for trophic level 3 fish must, therefore, be estimated by dividing mercury
concentrations in piscivorous fish by an appropriate predator-prey factor (PPF). A PPF for trophic
level 4 (PPF^ can be estimated from existing field data. This was done in Appendix A to Volume V,
resulting in a mean PPF4 of about 5. Dividing this value into the mean residue for trophic level 4 fish
(0.40 ug/g) yields a value for trophic level 3 fish of 0.08 ug/g.
June 1996 2-15 SAB REVIEW DRAFT
-------
Table 2-2
Nationwide Average of Mercury Residues in Fish3
Fish Species
Carp
Sucker (White, Redhorse and Spotted)
Catfish (Channel and Flathead)
Bass (Largmouth, Smallmouth and White)
Walleye pike
Northern pike
Crappie-
Brown Trout
Mean of All Fish Sampled
Mercury Concentration Averaged Across
Sampling Sites (ug/g wet weight)
0.11
0.167
0.16
0.38
0.52
0.31
0.22
0.14
0.26
a Data from Bahnick et al., 1994.
The highest mercury concentrations hi fish generally occur in the blood, spleen, kidney and
liver, and may exceed those in muscle by a factor of 2 -10 (McKim et al., 1976; Olson et al., 1978,
Ribeyre and Boudou, 1984; Boudou and Ribeyre, 1985; Harrison et al., 1990; Niimi and Kissoon,
1994). Owing to the size of these organs relative to that of other tissues, however, most of the
mercury contained in a fish at any given time is associated with muscle tissue.
2.3.1.3 Mercury Residues in Avian and Mammalian Wildlife
A large volume of mercury residue data exists for both avian and mammalian wildlife which
cannot be directly related to mercury concentrations in water or sediment. Nevertheless, these data are
of considerable value because they indicate the range of mercury concentrations that can be expected
in animals inhabiting both contaminated and uncontaminated environments. A comparison of these
residues with those obtained from laboratory dose-response studies provides additional information,
including the extent of difference between "natural background" residues and those which are
associated with toxic effects. Emphasis is placed on piscivorous birds and mammals; however, data is
also provided for the tree swallow because of its link to aquatic sediments through consumption of
emergent insects.
Mercury residues in tissues from birds are given in Table 2-3. The birds represented in this
table include animals taken from polluted environments and individuals collected from environments
for which there were no known point sources. This table is not intended to be an exhaustive
compilation of measured residues, but instead illustrates the range of values encountered in
environmental sampling efforts. Residues which, in the opinion of the author of the Report, were
associated with toxic effects are noted.
June 1996
2-16
SAB REVIEW DRAFT
-------
Table 2-3
Mercury Residues in Tissues of Piscivorous Birds
Species
Bald eagle
Bald eagle
Common loon
Common loon
Common loon
Common loon
Wood stork
Bald eagle
Common loon
Common loon
Common tern
Herring gull
Wood stork
Tree swallow
Common loon
Mercury (ug/g fresh
weight)
13.0 - 21.0
3.7 - 20.0
8.7
2.7
11.0- 18.0
2.0 - 5.0
1.9
0.07 - 0.41
0.40 - 1.10
2.0 - 3.0
3.6
2.3 - 15.8
0.7
0.04 - 0.08
1.6-47.7
Tissue
feathers
feathers
feathers
feathers
feathers
feathers
feathers
eggs
eggs
eggs
eggs
eggs
eggs
•
eggs
liver
Sampling Location
Great Lakes region
Great Lakes region
Minnesota lakes
Minnesota lakes
Wisconsin lakes
Wisconsin lakes
South Florida
15 States (USA)
Wisconsin lakes
Northwestern Ontario
Northwestern Ontario
Clay Lake, Ontario
South Florida
Lower Great Lakes
Northwestern Ontario
^
Comments
adults
nestlings
adults
juveniles
adults
juveniles
juveniles
polluted by point
source; LOAEL -
reproduction
polluted by point
source; LOAEL -
reproduction
polluted by point
source, no adverse
effects
consume emergent
aquatic insects
LOAEL -
reproduction
Reference
1
1
2
2
3
3
4
5
3
6
7
8
9
10
6
June 1996
2-17
SAB REVIEW DRAFT
-------
Table 2-3 (continued)
Mercury Residues in Tissue of Piscivorous Birds
Species
Common loon
Common loon
Common loon
Common goldeneye
Common merganser
Hooded merganser
Herring gull
Common loon
Mercury (ng/g fresh
weight)
9.5 - 90.0
5.6
0.2 - 6.9
0.9 - 19.4
4.4 - 13.1
3.9 - 17.6
0.7 - 4.0
1.5
Tissue
liver
liver
breast muscle
breast muscle
breast muscle
breast muscle
breast muscle
breast muscle
Sampling Location
Wisconsin lakes
Minnesota lakes
Northwestern Ontario
Clay Lake, Ontario
Clay Lake, Ontario
Clay Lake, Ontario
Tadenac Lake, Ontario
Tadenac Lake, Ontario
Comments
adults found dead
adults found injured
polluted by point
source
polluted by point
source
polluted by point
source
polluted by point
source
Reference
3
2
6
8
8
8
11
11
References:
1. Bowerman et al., 1994; range of means across sampling locations.
2. Ensor et al., 1992; mean of birds caught by nightlighting.
3. Belant and Anderson, 1990; range of individual values. Means for feathers (adult and juvenile), eggs and liver were 14.8, 4.0, 0.64 and 40.9, respectively.
4. Burger et al., 1993; mean value. *
5. Wiemeyer et al., 1993; range of means across sampling locations (collected after failure to hatch).
6. Barr, 1986; range of individual values. Means for liver and muscle were 13.0 and 2.3, respectively.
7. Fimreite, 1974.
8. Vermeer et at., 1973; range of individual values. Means for goldeneye, common merganser and hooded merganser were 7.8, 6.8 and 12.3, respectively.
9. Fleming et al., 1984; mean value.
10. Bishop et al., 1995; range of individual values, mean = 0.07.
11. Wren et al., 1983; gull data are reported as the range of individual values, mean = 1.7.
June 1996
2-18
SAB REVIEW DRAFT
-------
Factors contributing to the accumulation of mercury in wild birds are reviewed by
Scheuhammer (1987, 1991). Interpretation of residue data is complicated by the likelihood that
mercury distribution in tissues varies among species, and perhaps also among individuals of a single
species, depending upon age, sex, diet, and other factors. Nevertheless, several generalizations can be
attempted. Mercury levels in feathers of birds experimentally dosed with methylmercury generally
exceed levels in muscle, liver and kidney by a factor of 4 or more (Heinz, 1976a; Stickel et al., 1977;
Finley and Stendell, 1978), and it has been suggested that in free-living birds greater than 50% of the
total body burden of mercury may be present in the plumage (Braune and Gaskin, 1987). Natural
background levels of mercury in feathers of non-piscivorous raptorial birds are thought to range from
1-5 ug/g (dry weight).
Tissue levels of mercury associated with toxic effects in birds appear to exceed those in birds
inhabiting relatively uncontaminated environments by a factor of ten or less. This observation is
consistent with data for other environmental media (water, sediments, fish), which evidence similar
differences between natural "background" levels of mercury and those which cause significant
environmental damage. Owing to their ease of collection, the analysis of bird feathers and eggs has
been suggested as a means of identifying species that are at risk due to mercury. This suggestion has
particular merit in view of the natural variation in mercury levels in the fish upon which these animals
prey. Mercury residues in tissues also tend to integrate variations in mercury loading and elimination
due to changes in dietary habits, migration, egg production, etc.
The nature and abundance of mercury residue information for mammals reflects to a large
degree the availability of specimens as a byproduct of commercial trapping. Thus, abundant residue
data are available for wild muskrat, beaver, fox, weasel, bobcat, marten, fisher, wolf, raccoon,
opossum, mink and river otter. Data are also available for a number of game species, including
squirrels, rabbits, caribou, moose, deer, elk, mountain goat and bear. An extensive compilation of
these data is provided by Wren (1986), along with a review of tissue levels in both wild and
laboratory animals that have been associated with toxic effects. Some of the data from this
compilation are presented in Table 2-4, as well as more recent information. Emphasis was placed on
piscivorous species because of the exposure to these animals from consumption of contaminated fish.
Data from beaver and muskrat have also been included, both to provide a general comparison of
aquatic-based species, and because in several studies data were available for piscivores and herbivores
from the same waterbody. Emphasis was also placed on residues in fur and liver. This was done for
two reasons: (1) the highest residues are generally found in the liver and kidney; however, there are
more reported values for liver. (2) Fur, like feathers, has been suggested as a way of non-invasively
determining the residue status of wild animals and of identifying areas where animals may be at risk
due to mercury intoxication. Finally, data from raccoons trapped in South Florida are included in
Table 2-4 because they are suspected of contributing to mercury exposure in the Florida panther.
In general, the rank order of mercury residues in tissues from wild mink and otter is liver =
kidney > muscle > brain. Levels in fur relative to those in other tissues are variable but in most cases
are higher than those in liver. Comparisons between residues in wild animals with those in animals
experimentally dosed with mercury appear to be complicated by kinetics-based differences in
disposition. Thus, Wobeser et al. (1976b) reported that mercury levels in the fur of experimentally
dosed mink were low (1.5 ug/g) relative to concentrations in liver (24.3), kidney (23.1), muscle (16.0)
or brain (11.9). A similar pattern of distribution was reported for mink by Aulerich et al. (1974). In
contrast, mercury levels in the fur of an individual mink found dying of mercury poisoning were
higher than concentrations in any other tissue (Wobeser and Swift, 1976; see Table 2-4). Apparently,
the length of time over which a dose is obtained dictates its distribution, with redistribution from well-
June 1996 2-19 SAB REVIEW DRAFT
-------
Table 2-4
Mercury Residues in Tissues of Piscivorous Mammals
Species
Otter
Otter
Otter
Mink
Mink
Mink
Raccoon
Muskrat
Beaver
Otter
Otter
Otter
Otter
Otter
Mercury (ug/g fresh
weight)
15.2 - 25.6
6.5 (max. = 63.2)
47.0
10.7 (max. = 17.3)
7.6 (max. = 41.2)
34.9
4.4
0.06
0.03
1.7 - 3.4
2.4 - 4.5
0.3 - 3.0
0.9 - 3.5
96.0
Tissue
fur
fur
fur
fur
fur
fur
fur
fur
fur
liver
liver
liver
liver
liver
. Sampling Location
Georgia
Wisconsin
Clay Lake, Ontario
Georgia
Wisconsin
Saskatchewan
S. Florida
Wisconsin
Wisconsin
Manitoba
Winnipeg R.
Louisiana
Ontario
Clay Lake, Ontario
Comments
polluted, by point
source; death due to
poisoning
.
polluted by point
source; death due to
poisoning
males and females
males and females;
polluted by point
source
residues correlated
with acidity
polluted by point
source; death due to
poisoning
Reference
1
2
3
4
2
5
6
2
2
7
7
8
9
5
June 1996
2-20
SAB REVIEW DRAFT
-------
Table 2-4 (continued)
Mercury Residues in Tissues of Piscivorous Mammals
Species
Otter
Mink
Mink
Mink
Mink
Raccoon
Muskrat
Beaver
Mercury ((Jg/g fresh
weight)
3.3 (max. = 23.6)
0.4- 1.7
2. Umax. = 17.4)
0.1 - 2.6
58.2
1.5-24.0
<0.02
0.04
Tissue
liver
liver
liver
liver
liver
liver
liver
liver
Sampling Location
i
Wisconsin
Manitoba
Wisconsin •
Ontario
Saskatchewan
S. Florida
Wisconsin
Wisconsin
Comments
residues correlated
with acidity
polluted by point
source; death due to
poisoning
Reference
2
7
2
9
5
10
2
2
References:
1. Halbrook, R.S. 1978. Masters thesis. Environmental Pollutants in the River Otter of Georgia. University of Georgia at Athens; range of means across sampling
locations.
2. Sheffy and St. Amant, 1982; mean value.
3. Wren, 1985; one individual.
4. Cumbie, 1975; mean value.
5. Wobeser and Swift, 1976; one individual.
6. Bigler et al., 1975; mean value.
7. Kucera, 1983; Manitoba data are reported as the range of means across sampling locations. Data from the Winnipeg river are reported as a mean value.
8. Beck, 1977; range of means across sampling locations.
9. Wren et al., 1986; range of means across sampling locations.
10. Roelkc et al., 1991; range of means across sampling locations.
June 1996
2-21
SAB REVIEW DRAFT
-------
perfused organs (liver, kidney) to storage tissues (fur and muscle) slowly taking place during lifetime
exposures. These observations suggest that comparisons between mercury residues in wild and
experimental animals should be limited to consideration of well-perfused tissues. Wobeser et al.
(1976b) reported that mercury residues in the liver and kidney of mink that died during a 93-day
feeding study were 24.3 and 23.1 ug/g, respectively.
Somewhat higher values were reported for mink by Aulerich et al. (1974; 55.6 and 37.7 ug/g),
and for otter by O'Connor and Nielson (1980; 39.0 and 33.0 ug/g). Even if comparisons between
wild and experimental animals are limited to well-perfused tissues, questions still remain, however. In
general, mercury residues in all tissues from wild animals that are suspected to have died from
mercury poisoning are about twice those of animals that died from experimental intoxication (Wren,
1985, 1991).
Perhaps the most valid comparison that can be made at this time is that between apparently
unaffected wild animals and wild animals that have died from mercury poisoning. An examination of
Table 2-4 suggests that mercury residues hi tissues from mink and otters from Wisconsin (Sheffy and
jSt Amant, 1982) approached, and in several cases even exceeded, those of the "naturally" poisoned
animals. High mercury residues in fur were also reported for river otters trapped in several locations
across Georgia (Halbrook, 1978). The livers of raccoons captured in South Florida are also notably
high in mercury content (Roelke et al., 1991).
2.3.2 Individual Effects
Exposure to mercury can cause adverse effects hi a wide variety of organisms, including
plants, fish and aquatic invertebrates, birds and mammals. In this section, we review information on
exposure levels that can cause adverse effects in these groups.
2.3.2.1 Individual Effects on Plants
Effects of mercury on aquatic plants include death and sublethal effects. Sublethal effects
include plant senescence, growth inhibition, decreased chlorophyll content, decreased protein and RNA
content, inhibited catalase and protease activities, inhibited and abnormal mitotic activity, increased
free amino acid content, discoloration of floating leaves, and leaf and root necrosis (Boney, 1971;
Stanley, 1974; Muramoto and Oki, 1984; Mhatre and Chaphekar, 1985; Sarkar and Jana, 1986). The
level of mercury that results in toxic effects varies greatly among aquatic plants, as illustrated in Table
2-5.
Table 2-5
Toxicity Values for Aquatic Plants
Water Type
Fresh Water
Salt Water
Hg2* (HgCl or HgN03)
(Hg/L)
Low End
53 (alga)
10 (alga)
High End
3,400 (submerged
aquatic vegetation)
160 (seaweed)
Methylmercury
(Mg/L)
Low End
0.8 (alga)
Not available
High End
6.0 (alga)
Not available
Source: U.S. EPA, 1985.
June 1996
2-22
SAB REVIEW DRAFT
-------
Mercury can also cause death and sublethal effects in terrestrial plants. Sublethal effects on
terrestrial plants include decreased growth, leaf injury, root damage, inhibited root growth and
function, hampered nutrient uptake, chlorophyll decline and reduced photosynthesis (Schlegel et al.,
1987; Lindqvist, 1991; Godbold, 1991).
Methylmercury is more toxic to terrestrial plants than Kg2"1". One to ten nM (nanomolar)
mercuric chloride or methyl mercuric chloride (provided in a nutrient solution) can inhibit root
elongation in spruce seedlings. However, methyl mercuric chloride has a greater effect than mercuric
chloride at the same concentration (Godbold, 1991). Sublethal effects, including decreased
transpiration, decreased chlorophyll concentration, partial stomatal closure, and reduced photosynthesis,
occurred at nutrient solution concentrations of 10 nM methyl mercuric chloride (Schlegel et al., 1987).
2.3.2.2 Individual Effects on Fish and Aquatic Invertebrates
• i
The toxicity of mercury to fish has been reviewed by Eisler (1987) and more recently by
Wiener and Spry (1995). Toxicity varies, depending on the fish's characteristics (e.g., species, lifer
stage, age, size), environmental factors (e.g., temperature, salinity, dissolved oxygen content, hardness,
and the presence of other chemicals) and the form of mercury available. In particular, early life stages
(especially of salmonids) exhibit greater sensitivity to elevated metal concentrations than later life
stages. The toxicity of Kg2"1" compounds to salmonids and catfish tends to increase with temperature
(see Table 2-6). Organomercury compounds, such as methylmercury, generally are much more acutely
toxic than Kg2"1" to aquatic organisms.
Effects of mercury on fish include death, reduced reproduction, impaired growth and
development, behavioral abnormalities, altered blood chemistry, impaired osmoregulation, reduced
feeding rates and predatory success, and effects on oxygen exchange. LC50 values for fish range from
30 ug/L for guppies to 1,000 ug/L for the Mozambique tilapia (U.S. EPA, 1985). Symptoms of acute
mercury poisoning in fish include increased secretion of mucous, flaring of gill opercula, increased
respiration rate, loss of equilibrium and sluggishness. Signs of chronic poisoning include emaciation
(due to reduced food intake), brain lesions, cataracts, inability to capture food, abnormal motor
coordination and various erratic behaviors (e.g., altered feeding behavior) (Weis and Weis, 1989,
1995).
Table 2-6
Mercury Toxicity Increases With Temperature
Temperature (°C)
LC50 (ugfl)
Rainbow Trout with HgCl
5
10
15
Temperature (°C)
400
280
220
LC50 Oig/1)
Juvenile Catfish with Phenylmercuric Acetate
10
16.5
24
1,960
1,360
233
June 1996
2-23
SAB REVIEW DRAFT
-------
It is generally thought that toxic effects on fish in the environment are unlikely, except in the
case of point source pollution discharges. An accumulating body of evidence, however, suggests that
histological changes and effects on behavior, reproduction, and development can occur at water
concentrations as low as 0.1 ug/L (Wiener and Spry, 1995), or about two orders of magnitude higher
than those in a relatively pristine environment
Levels of mercury that induce toxic effects in aquatic invertebrate species vary. For Hg2"1",
acute (LC50) values for invertebrates range from 2.2 ug/L for the cladoceran Daphnia pulex to 2,000
ug/L for the larval forms of three insects (U.S. EPA, 1985). Examples of some specific toxicity
values for fish and aquatic invertebrates are provided in Table 2-7.
Table 2-7
Toxicity Values for Fish and Aquatic Invertebrates
Organism
; o „ ^ c <% ^
Fresh water
invertebrates
Fresh water fish
Rainbow trout
Fresh water AWQC*
Salt water
invertebrates
Salt water fish
Salt water AWQC*
s •...
Fresh water
invertebrates
Fresh water fish
Fresh water AWQC*
Salt water
invertebrates
Salt water AWQC*
Hg2+ (HgCl or HgNO,) (ug/L)
^\t^%^*&e^rBi^^
2.2 (cladoceran) to 2,000 (insect larvae)
30 (guppy) to 1,000 (tilapia)
155 to 420
Methylmercury (ug/L)
^ ,^\ ^ ^;,«, " , ;
Not available
Not available
24 to 84
2.4 (total mercury)
3 .5 (mysid) to 400 (soft clam)b
36 (juvenile spot) to 1,678 (flounderf
Not available
51.1 (mummichog)
2.1 (total mercury)
Y vJrV^ c *C»*ONI* "- s \*
0.96 (cladoceran) to 1.287 (cladoceran)
< 0.23 (minnow) to < 0.26 (minnow)
< 0.04 (cladoceran)
0.29 (brook trout) to 0.93 (brook trout)
0.012 (total mercury)
1.131 (mysid)
Not available
0.025 (total mercury)
* AWQCs are designed to be protective of the aquatic community as a whole.
b As cited in U.S. EPA, 1985, LC50s of 10,000 and 8,700 ng/L for Atlantic clams (Rangia cuneata) were reported by Olson
and Harrell (1973), but Dillon (1977) reported LC50 values of 58 and 122 |ug/L for the same clam species.
c As cited in U.S. EPA, 1985, an LC50 of.2,000 ug/L for mummichogs was reported by Klaunig et al. (1975), but Dorfman
(1977) and Eisler and Hennekey (1977) reported LC50 values of 800 ug/L or less for the same fish species.
Source: U.S. EPA, 1985 except where otherwise noted.
June 1996
2-24
SAB REVIEW DRAFT
-------
2.3.2.3 Individual Effects on Birds
Methylmercury has been shown to be more toxic to birds than inorganic mercury. Mercury
poisoning in birds is characterized by muscular incoordination, falling, slowness, fluffed feathers,
calmness, withdrawal, hyperactivity, hypoactivity and eyelid drooping (reviewed by Eisler, 1987;
Fimreite, 1979; Scheuhammer, 1987, 1991). Acute oral toxicity studies using methylmercury yielded
LD50 values ranging from 2.2 to 23.5 ug/g for mallards (Anas platyrhynchos), 11.0 to 27.0 ug/g for
quail (Coturnix) and 28.3 ug/g for northern bobwhite (Colinus virginianus). Some bird kills have
occurred, generally due to ingestion of mercury-based fungicides applied to grain. Whole-body
residues of mercury in acutely poisoned birds usually exceed 20 ug/g fresh weight and have been
found up to 126 ug/g. Mercury levels observed in such cases are generally highest in the brain,
followed by the liver, kidney, muscle and carcass.
Sublethal effects of mercury on birds include liver damage, kidney damage, neurobehavioral
effects, reduced food consumption, weight loss, spinal cord danfage, effects on enzyme systems,
reduced cardiovascular function, impaired immune response, reduced muscular coordination, unpaired
growth and development, altered blood and serum chemistry, and reproductive effects (Eisler, 1987;
Scheuhammer, 1987, 1991; MDNR, 1993). Reproductive effects, however, are the primary concern
for avian mercury poisoning and can occur at dietary concentrations well below those that cause overt
toxicity.
Scheuhammer (1991) concluded that on the basis of laboratory dose-response studies (Heinz,
1976a; Finley and Stendell, 1978) piscivorous birds consuming diets containing >1 ug/g (dry weight)
methylmercury in their diet (approximately 0.25 ug/g wet weight) will accumulate >20 ug/g (dry
weight) in their feathers. Similar levels in both spiked diets (Heinz, 1974, 1976a,b, 1979) and natural
prey sources (Barr, 1986) have been shown to be toxic to birds. Thus, it appears that mercury levels
in feathers exceeding 20 ug/g should be interpreted as evidence for possible toxic effects. Eisler
(1987) recommended that 5.0 ug/g fresh weight in feathers be used as a criterion for the protection of
birds.
Tissue mercury concentrations that are associated with toxicity in birds are remarkably similar
despite differences in species, dietary exposure level and length of time necessary to produce the effect
(Scheuhammer, 1991). Frank neurological signs are generally associated with 6rain mercury
concentrations of 15 ug/g (wet weight) or higher, and 30 ug/g or more in liver and kidney. Liver
mercury concentrations of 2-12 ug/g (wet weight) were associated with reproductive impairment in
adult pheasants and mallard ducks (Fimreite, 1971; Heinz, 1976a,b). Mortality was observed in newly
hatched ducklings with brain mercury concentrations of 3-7 ug/g (wet weight), while levels four times
these values are required to cause mortality in adults (Stoewsand et al., 1974; Finley et al., 1979;
Scheuhammer, 1988).
Reproductive impairment has been observed in laboratory studies when mercury concentrations
in eggs exceed 0.5 ug/g (Fimreite, 1971; Heinz, 1974, 1976a,b, 1979). Field studies tend to confirm
these results. Reproductive impairment in the loon was associated with mercury levels in eggs ranging
from 2-3 ug/g (Barr, 1986). Adverse effects on hatching and fledging were observed when mercury
concentrations in the eggs of common terns exceeded 3.6 ug/g (Fimreite, 1974). Mercury appeared to
be a contributing factor to reduced reproductive success in raptors at some locations (Odsjo, 1982;
Evans, 1986). In one study, however, hatching in herring gulls appeared to be unaffected, despite the
fact that eggs contained upwards of 10 ug/g of mercury (Vermeer et al., 1973). Lowest-observed-
adverse-effect level (LOAEL) and no-observed-adverse-effect level (NOAEL) values for effects of
June 1996 2-25 SAB REVIEW DRAFT
-------
methylmercury on avian wildlife are derived in Section 4.2.2 of this Report Possible effects on
populations of selected avian species are discussed in Section 2.3.3.
2.3.2.4 Individual Effects on Mammals
Extensive research on the toxicity of mercury to mammals indicates that effects vary
depending on the form of mercury ingested or inhaled. Inorganic mercury is corrosive, and acute
exposure to humans and other mammals may cause burning, irritation, salivation, vomiting, bloody
diarrhea, upper gastrointestinal tract edema, abdominal pain, and hemorrhaging (Klaassen et al., 1986).
Ingestion of mercurial salts in large doses may cause kidney damage. The main toxic effects due to
ingestion of organic mercurials are neurological effects such as paresthesia, visual disturbances, mental
disturbances, hallucinations, ataxia, hearing defects, stupor, coma and death (Klaassen et al., 1986).
The differences between the toxicity of different forms of mercury was exemplified in a study
by Aulerich et al.(1974) using mink (Mustela vffcri) given either 5 ppm methylmercury (in food) ortiO
ppm mercuric chloride. Mink treated with methylmercury died within 30 days, while those treated
with mercuric chloride suffered no ill effects. Methylmercury attacks the central nervous systems in
mammalian wildlife as well as in humans. The nervous system of the developing fetus may be
particularly vulnerable (Bakir et al., 1973), and concern for these effects tends to drive human health
risk assessments for mercury (Clarkson, 1990; reviewed in Volume IV of this Report). Methylmercury
ingestion can also cause reduced food intake, weight loss, muscular atrophy and damage to the
animal's heart, lungs, liver, kidneys or stomach (Klaassen et al., 1986; MDNR, 1993).
Levels of exposure that induce mercury poisoning in mammals vary among species. For
sensitive mammals, effects can occur at a dose of 0.25 ug/g bw/d, or 1.1 ug/g in the diet (Eisler,
1987). Death occurs in sensitive mammal species at 0.1-0.5 ug/g bw/d, or 1.0-5.0 ug/g in the diet
Smaller animals (e.g., minks, monkeys) are generally more susceptible to mercury poisoning than are
larger animals (e.g., mule deer, harp seals), perhaps because of differences in elimination rates. Also,
smaller mammals eat more per unit body weight than larger mammals and, thus, may be exposed to
larger amounts of mercury on a body weight basis. LOAEL and NOAEL values for effects of
methylmercury on mammalian wildlife are derived in Section 4.2.2 of this Report
2.3.3 Population Effects
Mercury contamination has been documented in endangered species such as the Florida
panther and the wood stork, as well as in populations of loons, eagles and furbearers such as mink and
otters. These species are at high risk of mercury exposure and effects because they either are
piscivores or eat piscivores.
2.3.3.1 Loon Populations
It has been suggested by several researchers that loons are at risk from mercury contamination
in aquatic food chains. Loons are primarily piscivorous but also consume benthic macroinvertebrates,
such as crayfish, when water is too turbid for catching fish (Barr, 1973). Mercury levels in crayfish
approach and may even exceed those of prey fish from the same lakes (Barr, 1986). Concern for the
loon is due also to the fact that much of its summer range coincides with areas that are known to be
susceptible to acid deposition. As noted previously, a negative correlation often exists between the pH
of surface water and mercury concentrations in fish.
June 1996 2-26 SAB REVIEW DRAFT
-------
The most comprehensive study of mercury toxicity in wild loons was conducted by Barr
(1986). Loons were collected from three habitats within the Wabigoon River watershed (Ontario,
Canada) both above and below a chlor-alkali plant which discharged mercury into the river system.
The first habitat (designated Cl) consisted of the lakes and river directly downstream of the plant.
Habitat C2 did not receive mercury discharges but was accessible to mercury-contaminated fishes
which originated in Cl. Habitat C3 was upstream from the chlor-alkali plant and received no
appreciable mercury from other sources. Contaminated fish were prevented from entering C3 by a
waterfall. A nearby habitat (C4), not connected to the other three habitats, received no mercury
contamination and served as a control. Human disturbances in all habitats were determined to be
minimal, and concentrations of organochlorine contaminants were low (less than 0.02 ppm total for all
pesticides, including all DDT metabolites, and 0.04 ppm for PCBs).
Barr (1986) found a strong negative correlation between mercury levels in water and
reproductive success in loons as far as 160 km downstream from the mercury source. Mercury in prey
fish and invertebrates declined with increasing distance from the mercury source but contaminated fish
were able to migrate into the uncontaminated C2 habitat Mercury levels in loon tissues (eggs, liver,
muscle and brain in both adults and chicks) were highest in the Cl habitat but were also elevated in
the C2 habitat, presumably because loons were feeding on contaminated prey which migrated from Cl.
Mercury levels in loons from habitat C3 (upstream from mercury source and inaccessible to
contaminated fish) were comparable with levels from the uncontaminated control habitat, C4. With
the exception of the liver, most of the mercury in loon tissues was in the form of methylmercury.
Mercury in the liver appeared to be inorganic, suggesting the existence of a demethylation pathway.
Dose-response relationships appeared to exist between mercury in both prey and various tissues, and
reproductive success. For example, reductions in egg laying and in nest site and territorial fidelity
were associated with prey containing mean mercury concentrations in the range of 0.3-0.4 ng/kg.
Reproductive success was also reduced hi loons with brain or egg levels of 2-3 fig/kg, and in loons
with liver residues above 13 ug/g. No loons reproduced successfully where prey species contained
mercury at levels greater than 0.4 ug/kg.
Ensor et al. (1992) captured 93 loons and collected 128 dead or dying loons from 18 northern
and central counties in Minnesota. Feathers were collected from live loons. Feathers and liver tissue
were collected from the dead loons. In 22 percent of the liver samples, mercury concentrations
exceeded the level (13 ug/g) associated by Barr (1986) with impaired reproduction. Adult loons
contained greater concentrations of mercury (8.7 vs. 2.7 jig/g in feathers and 6.6 vs. 1.1 (ag/g wet
weight liver) than juvenile loons, as expected for a contaminant which bioaccumulates. The mercury
in the juvenile loons was considered to be representative of local mercury contamination since all of
their food would have been obtained from lakes within Minnesota. Mercury in adult loons was
thought to represent contributions from both the summering (Minnesota) and wintering (Gulf of
Mexico) grounds.
Ensor et al. (1992) concluded that juvenile loons which died of disease had significantly higher
mercury levels in feathers than juvenile loons which died from injury or which were caught alive.
Emaciated loons also had significant elevations of mercury in both feathers and liver (significance
level not reported). It was not clear whether elevated mercury in emaciated loons resulted from
concentration of existing mercury stores while body fat and protein were catabolized or whether
mercury contributed to the emaciation. The authors concluded that evidence of adverse impacts on the
Minnesota loon population was sufficient to recommend monitoring of loon populations and mercury
levels.
June 1996 2-27 SAB REVIEW DRAFT
-------
Working in north central Wisconsin, Belant and Anderson (1990) collected both live and dead
loons and addled eggs from abandoned nests. Mercury residues and 14 organochlorine pesticides were
measured in feathers (live and dead loons), brain, muscle, and liver (dead loons). The conclusions
reported in this study were similar to those reached by Ensor et al. (1992). Pesticide concentrations in
dead loons were relatively low. In contrast, mercury levels in liver (mean concentration of 40.9 ng/kg
wet weight) exceeded those reported by Barr (1986) to be associated with reproductive dysfunction.
Scheuhammer and Blancher (1994) reported mercury levels hi fish sampled from lakes
throughout Ontario, Canada hi areas without known point sources of mercury. Up to 30 percent of the"
lakes contained fish with mercury levels which exceeded 0.3 ug/kg (wet weight), which Barr (1986)
had shown previously to be associated with reproductive impairment in loons. The lack of any
identified point source of mercury contamination was considered by the authors to be indirect evidence
of airborne deposition of mercury over large portions of Ontario.
The viability of loon populations within their traditional habitats in the United States is
unclear. None of the studies reviewed was able to demonstrate clear population declines on a regional
or national basis. Several well conducted studies have found that substantial numbers of loons
sampled contained mercury at or above levels demonstrated to reduce reproductive success. It has also
been suggested (but not clearly demonstrated) that sublethal effects of mercury exposure may produce
greater susceptibility to environmental stresses including other contaminants. Mercury also may make
loons more susceptible to secondary infections, especially during stressful activities such as molting
and migration. Investigations in response to a die-off of over 2,500 loons hi the Gulf of Mexico in
1983 found that elevated levels of mercury were associated with abnormally high infestations of
parasites (Barr, 1986).
2.3.3.2 Eagle Populations
Bald eagles are distributed throughout the United States. Many migrate into the lower forty-
eight states only during the winter months; others are resident throughout the year. Bald eagles, like
several other avian species, are thought to have been adversely impacted by DDT and its metabolites.
Because of their status as a federally listed "threatened" species, the potential for mercury exposure to
pose a threat to eagle survival and recovery is a concern.
Researchers have measured mercury residues hi bald eagle feathers (U.S. FWS, 1993; Welch,
1994; Bowerman, 1994), eggs (Grier, 1974; Wiemeyer et al., 1984, 1993;.Grubb et al., 1990; Anthony
et al., 1993) and blood (Anthony et al., 1993; U.S. FWS, 1993; Welch, 1994). Several of these studies
have also included analyses of eagle tissue for other contaminants which might threaten eagle
reproduction.
Wiemeyer et al. (1984) sampled bald eagle eggs that had failed to hatch from nests located in
14 states between 1969 and 1979; eggs were tested for organochlorine residues and mercury. The
highest levels of mercury were detected hi eggs from Maine. Eight organic contaminants were
significantly, negatively correlated with eggshell thinning, a trait often linked with reproductive failure
in birds. When mercury levels were compared with the mean 5-year production rate for eagle nests, a
weak negative correlation was suggestive of an adverse effect of mercury. The effect was confounded,
however, by the co-occurrence of DDE hi many of the eggs with the highest mercury levels. The
authors concluded that mercury contamination appears to have the potential for adverse effects on
eagle production in only a few of the breeding areas sampled, primarily in Maine.
June 1996 2-28 SAB REVIEW DRAFT
-------
Continuing the work begun earlier, Wiemeyer et al. (1993) collected eggs that had failed to
hatch from IS states between 1980 and 1984 and analyzed them for organochlorine pesticides,
polychlorinated biphenyls (PCBs) and mercury. These data were then combined with the data
collected previously (Wiemeyer et al., 1984). As before, DDE was the contaminant most significantly
(negatively) correlated with eggshell thinning, with DDD, DDT and PCBs significantly but less
strongly correlated. The highest levels of DDE, PCBs and mercury occurred in eggs collected in
Maine. Mercury levels in eagle eggs, at or above 0.28 ug/g (wet weight), were significantly correlated
with a reduction hi mean 5-year production rate for eagle nests. This value compares with a value of
0.5 ug/g derived earlier (Wiemeyer et al., 1984). The authors noted, however, that only three egg
samples (all from Maine) contained mercury levels greater than 0.5 ug/g and that these eggs also
contained levels of DDE (>6 ug/g) known to reduce eagle productivity. Wiemeyer et al. (1993)
concluded that recent data provide even less evidence than previously indicated (Wiemeyer et al.,
1984) that contaminants other that DDE are adversely impacting bald eagle productivity. Grubb et al.
(1990), Grier (1974) and Anthony et al. (1993) reached similar conclusions on the lack of evidence for
an association of mercury levels with reproductive failure in bald eagles.
Bowerman and co-workers (Bowerman, 1993; Bowerman et al. 1994) examined the
productivity of bald eagles in six geographic regions, including Lakes Superior, Michigan, Huron, and
Erie, and the states of Michigan and Minnesota. Significant negative correlations existed between
PCS and p,p'-DDE in plasma and reproductive success. Mercury levels in feathers ranged from 9.0 to
23.4 ug/g; there was no correlation with reproductive success.
Welch (1994) sampled eggs, blood and feathers from Maine bald eagles and analyzed for
organochlorine pesticides, PCBs, TCDD equivalents (TCDD-eq) and mercury. Mercury levels in
inland eagles were higher than concentrations in eagles inhabiting the coastline. In general, these
elevated mercury levels appeared to be related to mercury residues hi fish from the two areas.
Productivity was also lower for inland eagle nests; however, the correlation of mercury levels in blood
and feathers with mean productivity (5 and 15 yr) was not significant.
One of the difficulties in evaluating the effect of mercury on the bald eagle is the co-
occurrence of organochlorine compounds such as PCBs, DDE and TCDDs at levels which are known
to affect reproduction adversely. Bowerman (1993) pointed out that the effect of the organochlorine
contaminants may be masking the effect of mercury. The U.S. Fish and Wildlife Service (1993) also
suggested that while mercury was not found in Florida bald eagles at lethal levels, undetected sublethal
effects may be adversely affecting eagle reproduction.
2.3.3.3 Wood Stork Populations
Mercury has been detected hi feathers of the endangered wood stork, although the levels
detected apparently have not caused toxic effects. Young wood storks in Florida had mercury levels
of 1.87 ug/g dry weight; higher mercury levels would be expected for adults from the same area
(Burger et al., 1993). Fleming et al. (1984) reported mercury levels of 0.66 ug/g wet weight in wood
stork eggs, which is somewhat less than Eisler's (1987) recommended criterion of <0.90-2.0 ug/g wet
weight in eggs.
2.3.3.4 Furbearer Populations
In one Ontario incident, an eagle was found scavenging on a mercury-poisoned dead otter at
Clay Lake (Wren, 1985). Mercury levels in the otter (liver - 96 ug/g; kidneys - 58 ug/g; brain - 30
June 1996 , 2-29 SAB REVIEW DRAFT
-------
ug/g) were well above those known to be toxic to otters in laboratory exposures. The primary source
of the mercury was a chlor-alkali plant that discharged mercury directly into the river. Although
population data were not collected, a Native American had discontinued trapping in the area because
furbearers such as otters and mink had disappeared, providing anecdotal evidence of population
declines. Because other stressors could also be present (such as hunting pressure and habitat loss or
degradation), the role of mercury in contributing to any furbearer population declines is uncertain.
In a separate incident, a mink exhibiting unnatural behavior was collected near the mercury- .
contaminated Saskatchewan River (Wobeser and Swift, 1976). Subsequent determination of mercury
levels in the liver (58 ug/g), kidney (32.9 ug/g), muscle (15.2 ug/g), brain (13.4 ug/g) and fur (34.9
ug/g), combined with clinical and pathologic findings, were deemed sufficient by the authors to
conclude that the animal had been poisoned by mercury. Residue levels found hi this animal exceeded
those determined in laboratory studies to be associated with toxicity. The origins of mercury in this
case could not be determined; however, it was suggested that fish from the Saskatchewan contain
mercury at concentrations higher than those known to cause toxicity to mink.
In a study of furbearers obtained from trappers (1972-1975) in the Wisconsin River watershed,
otters contained the highest tissue mercury levels, followed by minks, raccoons, foxes, muskrats and
beavers (Sheffy and St. Amant, 1982).
2.3.3.5 Florida Panther Populations
Mercury is suspected of contributing to population declines of the endangered Florida panther
(Felis concolor coryi). The Florida Panther Interagency Committee (FPIC, 1989) reported that
approximately 100 ppm of mercury was detected in the liver and 130 ppm in the hair of a 4-year-old
female panther. The panther, No. 27, had been radio-instrumented since 1988 and was found dead in
the eastern part of the Florida Everglades National Park (FPIC, 1989). Relatively high levels of
mercury (0.005-20.0 ppm) were detected in archived liver samples from six dead panthers, and levels
ranging from 0.02-130.0 ppm have been measured hi the hair samples from ten live individuals.
Cause of death of the six archived animals was not mentioned in this Report. The FPIC concluded,
however, that panther No. 27 died of mercury poisoning.
The most probable source of mercury contamination in Florida panthers is via the food chain.
The diet of the Florida panther varies greatly depending on prey availability; however, mercury
contamination in raccoons has been documented to occur in a distributional pattern similar to that of
Florida panthers (Roelke et al., 1991). The accumulation of mercury in raccoons is thought to be due
in turn to consumption of contaminated aquatic life, including invertebrates, fish and amphibians. The
panthers most at risk, therefore, appear to be those that consume mercury-contaminated raccoons.
Panthers which prey on deer are less exposed to mercury because deer are herbivores and accumulate
less mercury. Based upon the findings of the FPIC, habitat modifications have been implemented in
the Florida Everglades to increase forage for deer.
In addition to increasing mortality, mercury contamination could decrease reproductive success
in the Florida panther. Methylmercury ingested by a pregnant mammal passes through the placenta to
the developing fetus, potentially causing abortions, stillbirths, congenital defects and behavioral
modifications that result in early neonatal death. Roelke et al. (1991) found a significant inverse
correlation between mercury concentrations in mother panthers and survivorship of the young.
Because so few Florida panthers remain (only 30 to 50 in the wild) (Jordan, 1990), the possibility
June 1996 2-30 SAB REVIEW DRAFT
-------
exists that mercury contamination could contribute to the extinction of this endangered species (Roelke
et al., 1991).
2.3.4 Communities and Ecosystems
2.3.4.1 Aquatic Communities and Ecosystems
Effects of contaminants on aquatic communities have been investigated by examining
functional and structural responses of natural assemblages in laboratory settings to'toxic substances
added singly or in combination. The species diversity of freshwater and brackish-water microbial
communities was reduced by exposure to 40 ug/L of mercuric chloride (Singleton and Guthrie, 1977).
Carbon fixation was reduced by 50 percent in freshwater phytoplankton communities exposed to 0.4
ugL of mercuric chloride, but this effect was mitigated by the presence of humus or sediment (Hongve
et al., 19^80). Mercuric chloride (0.5 ug/L) administered to a marine aquatic community inhibited
phytoplankton growth,-killed or retarded development in copepods and increased the number of viable
bacteria (Kuiper, 1981). The species composition of the phytoplankton also changed, possibly due to
selective reduction of predation by the copepods. Bacterial populations may have increased due to
increased food supply in the form of dead phytoplankton (Kuiper, 1981).
In general, mercury concentrations (as Hg"1"2) required to elicit toxic effects on natural aquatic
communities exceed those commonly measured in surface waters by two or more orders of magnitude
(low ng/L hi waters not impacted by point source discharge; Spry and Wiener, 1991; Wiener and Spry,
1995). Studies of the effects of methylmercury on aquatic assemblages were not found, however, and
it can be reasonably anticipated that the toxicity of methylmercury to these communities would exceed
that of mercuric chloride. Effects of mercury or any other substance at this level of biological
organization could potentially have far-reaching impacts on the entire food chain by changing both
nutrient and energy fluxes.
Field studies of mercury occurrence and effects at the community level are not available.
Moreover, interpreting field studies can be difficult because more than one stressor is often present
For example, high concentrations of toxic substances including mercury have been found in various
locations in the Great Lakes, and several species of piscivorous wildlife have suffered or continue to
suffer reduced reproductive success and population declines in these areas (e.g., Caspian terns, herring
gulls, double-crested cormorants, mink) (Peakall, 1988; Colbom, 1991; Environment Canada, 1991;
Gilbertson et al., 1991). Other bioaccumulative contaminants present in high concentrations, such as
PCBs, dioxins and DDT/DDE have been implicated as the most likely cause in most reported cases
(Colborn, 1991; Gilbertson et al., 1991).
2.3.4.2 Terrestrial Communities and Ecosystems
As noted previously, atmospherically deposited heavy metals such as mercury tend to
accumulate hi top soils. This results in particularly high exposures hi decomposer communities, which
play a crucial role within the natural nutrient cycles of terrestrial ecosystems. Mercury forms stable
complexes with organic substances of high molecular weight (humic acids) and, thus, exhibits limited
mobility within soils, posing a cumulative risk to soil biological activity. Processes that may be
affected by heavy metals in the top soil include soil type, litter decomposition, carbon mineralization,
nitrogen transformation and enzyme activity. Mercury effects on soil microorganisms vary depending
on soil type (Zelles et al., 1986). Mercury generally inhibits ATP, heat production, respiration and
iron reduction by soil microorganisms in sandy soils and, to a lesser extent, in clay. At some
June 1996 2-31 ' SAB REVIEW DRAFT
-------
intermediate concentrations, however, mercury may stimulate microbial activity in peat (Zelles et al.,
1986).
It is difficult to estimate specific toxic levels for microbial-mediated processes and decomposer
communities, due to widely differing soil properties and methodological discrepancies in the literature.
In a report on mercury hi the Swedish environment, Lindqvist (1991) cites a study in which soil
microbial activity was significantly reduced at mercury concentrations ranging from 0.06-0.08 ug/g
dry weight of humus. The concentration of mercury in forest soils in Sweden is in the range 0.01-0.09
ug/g. In a second cited study, however, the mercury concentration in soil required to reduce soil
microbial activity was 50 ug/g. A common effect of metal contamination in soil animal groups is a
decrease in species diversity. In some species, susceptibility to metals is thought to be a secondary
effect due to differences hi food availability rather than metal toxicity per se.
2.3.5 Conclusions
Of the pathways by which ecosystems and components of ecosystems might be exposed to
atmospheric mercury, exposure of high trophic level wildlife to mercury in food is particularly
important The trophic level and feeding habits of an animal influence the degree to which that
species is exposed to mercury. Mercury biomagnifies hi aquatic food chains with the result that tissue
concentrations of mercury increase as trophic levels increase. Predatory animals primarily associated
with aquatic food chains accumulate more mercury than those associated with terrestrial food chains.
Thus, piscivores and other carnivores that prey on piscivores generally have the highest exposure to
mercury. In a study of forbearing mammals in Wisconsin, the species with the highest tissue levels of
mercury were otter and mink, which are top mammalian predators on aquatic food chains (Sheffy and
St Amant, 1982). Top avian predators of aquatic-based food chains include raptors such as the osprey
and bald eagle. Smaller birds feeding at lower levels in aquatic food chains also may be exposed to
substantial amounts of mercury because of their high food consumption rate (consumption/d/g of body
weight) relative to larger birds.
Although clear causal links have not been established, mercury originating from airborne
deposition may be a contributing factor to population effects on bald eagles, river otters and mink.
Stronger evidence is available to support the possibility of toxic effects on the common loon and the
Florida panther. Effects of mercury originating from point sources on restricted wildlife populations
have been conclusively demonstrated and provide a tissue residue basis for evaluation of risk to other
populations. Based upon reviews of both laboratory and field data, mercury levels that exceed the
following values (in ug/g fresh weight) have been suggested as evidence for possible adverse impacts
on avian populations: feathers - 20 ug/g (Scheuhammer, 1991); eggs - 2.0 ug/g (Scheuhammer, 1991;
after conversion from dry weight); liver - 5 ug/g (Zillioux et al., 1993). Such criteria must be used
with caution, however, as residue thresholds both above and below these values have been reported.
Field data for mammals are not as extensive as those for birds; they appear to be in as good agreement
with laboratory data. Mercury residues in mink and otter that were thought to have been poisoned by
mercury originating from a point source exceeded those seen hi dead laboratory animals by a factor of
two or more (Wren, 1991; see Section 2.3.2.4). The reason for this variation is not presently known.
Additional information is needed before tissue-residue-based criteria for piscivorous mammals can be
developed. Criterion values for fish and water that are designed to be protective of piscivorous
wildlife are calculated in Section 4.2 of this Volume.
June 1996 2-32 ' SAB REVIEW DRAFT
-------
2.4 Ecosystems Potentially at Risk
The information presented in Sections 2.2 and 2.3 suggests that the ecosystems most at risk
from mercury releases to ah- exhibit one or more of the following characteristics:
• They are located in areas where exposure to mercury (e.g., atmospheric deposition of
mercury) is high;
• They include surface waters already impacted by acid deposition;
• They possess characteristics other than low pH that result in high levels of
bioaccumulation; and/or
• They include sensitive species.
2.4.1 Highly Exposed Areas
Ecosystems subject to high levels of mercury deposition (e.g., near sources of mercury
emissions, hi areas with high deposition rates) will be more exposed to mercury than ecosystems with
lower levels of mercury deposition. The pattern of mercury deposition nationwide, therefore, will
influence which ecoregions and ecosystems might be exposed to hazardous levels of mercury.
2.4.2 Lakes and Streams Impacted by Acid Deposition
In many aquatic systems the tendency for mercury to bioaccumulate hi fish appears to be
inversely correlated with pH and alkalinity (or acid neutralizing capacity) (reviewed by Spry and
Wiener, 1991). Thus, fish in acidic lakes (pH 6.0 to 6.5 or less) often have higher body or tissue
burdens of mercury than fish hi nearby lakes with higher pH. This relationship has been found for a
variety of fish species and water bodies, including the following:
• various fish species in 14 lakes and 31 streams in Florida (FDER, 1990);
• yellow perch from lakes hi the Upper Michigan peninsula (Grieb et al., 1990);
• yellow perch from seepage lakes in Northern Wisconsin (Cope et al., 1990);
• yellow perch from an experimentally acidified lake in Northern Wisconsin
(Wiener et al., 1990);
• yellow perch from Southern Ontario lakes (Suns and Hitchin, 1990);
• walleyes from Wisconsin lakes (Lathrop et al., 1991);
• largemouth bass from 53 lakes in Florida (Lange et al., 1993);
• northern pike from 80 Minnesota lakes (Sorensen et al., 1990); and
• smallmouth bass from Ontario lakes (McMurtry et al., 1989).
June 1996 2-33 SAB REVIEW DRAFT
-------
The increased accumulation of mercury in low pH lakes appears to be due largely to increased
microbial production of methylmercury (Xun et al, 1987; Bloom et al., 1991; Miskimmin et al., 1992),
although biogeochemical processes that release mercury from sediments have also been implicated
(Rada et al., 1993). The bioavailability of methylmercury is probably also enhanced by decreased
levels of calcium, as is typical of such lakes. Exceptions to the general relationship between pH and
bioaccumulation of mercury exist, however (Fjeld and Rognerud, 1993).
2.4.3 Factors in Addition to Low pH that Contribute to Increased Bioaccumulation of Mercury in
Aquatic Biota
Numerous factors hi addition to low pH can influence the bioaccumulation of mercury in
aquatic biota. These include length of the aquatic food chain, temperature (Bodaly et al., 1993), and
other water chemistry parameters (e.g. dissolved organic material; McMurtry et al., 1989; Nilsson and
Hakanson, 1992; Fjeld and Rognerud, 1993). Physical and chemical characteristics of a watershed
affect the amount of mercury that is translocated from soils to water bodies (McMurtry et al., 1989, -
Johnston et al., 1991; Joslin, 1994). Interrelationships between these factors are poorly understood,
however, and there is no single factor (including pH) that has been correlated with mercury
bioaccumulation hi all cases examined.
2.4.4 Sensitive Species
For the purpose of this discussion, sensitive species are defined as those species that are more
likely than others to experience adverse effects due to mercury contamination. Such species may or
may not be inherently more sensitive on an absorbed dose basis. Sensitive species also may be at risk
because they receive high methylmercury exposures due to their position in the food chain, or because
their populations are already stressed. In the first category are top-level predators of aquatic-based
food webs exposed to high concentrations of mercury in their prey. Examples include raptors (e.g.,
bald eagles, ospreys) and piscivorous waterbirds (e.g., herons, gulls, terns, cormorants). The second
category includes threatened and endangered species, which are species that have already experienced
severe population declines and are at risk of further population declines or extinction.
2.5 Endpoint Selection
A goal of the problem formulation phase
in an ecological risk assessment is to select
ecological endpoints that are relevant to
decisions to be made. An endpoint is defined as
a characteristic of an ecological component (e.g.,
increased mortality in fish) that may be affected
by exposure to a stressor (U.S. EPA, 1992a).
Ecological endpoints can be chosen at any level
of biological organization, from biochemical and
cellular levels through individuals, populations,
communities, and ecosystems (U.S. EPA,
1992a).
U.S. EPA distinguishes two types of endpoints for ecological risk assessment purposes:
assessment endpoints and measurement endpoints (see text box). Assessment endpoints are explicit
expressions of the actual environmental value that is to be protected. Usually the assessment endpoint
Endpoints for Ecological Risk Assessment
Assessment endpoint - an explicit expression of the
environmental value that is to be protected (U.S. EPA,
1992a).
Measurement endpoint - a measurable ecological
characteristic that is related to the valued characteristic
chosen as the assessment endpoint Measurement
endpoints are often expressed as the statistical or
arithmetic summaries of the observations that comprise
the measurement (U.S. EPA, 1992a).
June 1996
2-34
SAB REVIEW DRAFT
-------
cannot be measured directly, so a risk assessor selects a measurement endpoint (or a group of
measurement endpoints) that can be related, either quantitatively or qualitatively, to the assessment
endpoint (U.S. EPA, 1992a).
Table 2-8 provides examples of ecological assessment and measurement endpoints at various
levels of biological organization. In current practice, the most tractable endpoints are at the individual
or population level and include mortality, growth and development and reproduction.
Table 2-8
Examples of Assessment and Measurement Endpoints
Level of Organization
Ecoregion*
Ecosystem
Community
Population
Individual
Abiotic
Assessment Endpoints
Biodiversity
Regional production
Landscape aesthetics
Productive capability
Nutrient balance
Soil balance
Recreational quality
Change to less useful/desired type
Market/sport value
Extinction
Abundance
Yield/production
Frequent gross morbidity
Massive mortality
Range
Survival
Growth and development
Reproduction
Good physical condition
Habitat quality
Measurement Endpoints
Habitat area
Regional production
Other landscape descriptors
Habitat area
Biomass
Productivity
Nutrient export
Species number
Species evenness
Species diversity
Market/sport value
Saprobic index
Occurrence
Numbers/density
Age structure
Fecundity
Yield/production
Frequency of gross morbidity
Mortality rate
Longevity
Growth and development
Fecundity
Overt symptomology
Biomarkers
Temperature
Water flow
Soil characteristics
Sediment characteristics
Source: Adapted from U.S. EPA, 1989.
a An ecoregion is an area (region) of relative homogeneity in ecological systems (based on elevation, soils, latitude,
precipitation).
June 1996
2-35
SAB REVIEW DRAFT
-------
Based on the information provided in Sections 2.1 through 2.4, the ecological components that
appear to toe most at risk from atmospheric mercury are piscivorous mammals and birds that feed at or
near the top of aquatic food chains. Tliis is particularly true of threatened or endangered species that
already have suffered severe population declines due to one or more causes. An appropriate
assessment endpoint, therefore, would be maintenance of self-sustaining populations of these species.
Appropriate measurement endpoints for exposed wildlife species would include growth and survival of
adults or other life-stages, and reproductive success. Alternatively, when such data are difficult to
collect (or are of questionable utility due to the mode of toxic action), it may be necessary to infer
adverse effects on wildlife from laboratory toxicity studies.
2.6 Conceptual Model for Mercury Fate and Effects in the Environment
The final goal of the problem formulation phase in risk assessment is to develop a conceptual
model of how the stressor may affect ecological components of the natural environment (U.S. EPA,
1992a). The conceptual model identifies the ecosystem(s) potentially at risk, exposure pathways
between sources and receptors and the relationsnip(s) between (the) measurement and assessment
endpoints. A preliminary analysis .of the ecosystem, stressor characteristics and ecological effects
helps to define possible exposure scenarios; that is, qualitative descriptions of how the stressors co-
occur with or contact the various ecological components.
A conceptual model of the ecological effects of airborne mercury emissions can be visualized
in Figures 2-1 through 2-5. Mercury is emitted to the atmosphere primarily as the elemental form or
as an inorganic ion. Inorganic mercury returns to earth in wet deposition due to its relatively high
solubility in water and because it adsorbs to airborne particulates. Elemental mercury has a long half-
life in the atmosphere and tends to stay aloft but may react with other chemicals to form inorganic
mercury species. Wet deposition containing mercury falls on watersheds or directly on water bodies.
Mercury deposited to watersheds is rapidly bound to organic matter and tends to accumulate over time.
A portion of this mercury is continually released, however, and is transported in runoff and
groundwater to receiving waters such as lakes, streams and wetlands. Biotic and abiotic chemical
reactions transform mercury in water and associated sediments to organic derivatives (primarily
methylmercury). Organomercurial compounds then accumulate in aquatic food chains, due both to
their tendency to become sequestered in tissues and to the efficiency with which they are transferred
from one trophic level to another. Eventually, mercury in fish is consumed by piscivorous wildlife,
with the resulting potential for adverse lexicological effects. Uptake pathways other than consumption
of contaminated prey (e.g.," inhalation and drinking of contaminated water) are considered to be of
little consequence for piscivorous birds and mammals.
June 1996 2-36 SAB REVIEW DRAFT
-------
3. EXPOSURE OF PISCIVOROUS AVIAN AND MAMMALIAN
WILDLIFE TO AIRBORNE MERCURY
3.1 Objectives and Approach
The objective of this analysis was to characterize the extent to which piscivorous wildlife are
exposed to mercury originating from airborne emissions. Four general approaches were used, which
may be described as follows.
1. Estimation of current average exposure to piscivorous wildlife on a nationwide basis (Section
3.3).
Estimates of current mercury exposure to selected piscivorous wildlife species were calculated
as the product of the fish consumption rate and measured mercury concentrations in fish. This was
not intended to b'e a site-specific analysis, but was instead intended to provide national exposure
estimates for piscivorous wildlife based on typical mercury concentrations in fish. This analysis
utilized mean total mercury measurements from a nationwide study of fish residues and published fish
consumption data for the selected wildlife species.
2. Estimation of mercury levels in fish using measured deposition values and an indirect exposure
methodology (Section 3.4).
Mercury levels in fish were estimated by using measured mercury deposition values as inputs
to an indirect exposure model (IEM2). Additional inputs to the IEM2 model include the
characteristics of a hypothetical lake and its associated watershed. The analysis was conducted for two
such lakes, one located hi the Western U.S., the other located hi the Eastern U.S. Residue values were
calculated as the product of predicted mercury concentrations hi water and estimated BAF values for
fish hi trophic levels 3 and 4. •
3. Estimation of mercury deposition on a regional scale (40 km grid), and a comparison of these
data with species distribution information (Section 3.5).
A long-range atmospheric transport model (RELMAP) was used in conjunction with a mercury
emissions inventory to generate predictions of mercury deposition across the continental U.S. This
information was then compared with wildlife species distributions to characterize the potential for co-
occurrence of high mercury deposition rates and the presence of wildlife species of concern.
Additionally, mercury deposition data were superimposed onto a map of surface waters impacted by
acid deposition, since it has been shown that low pH values are positively correlated with high levels
of mercury in fish.
4. Estimation of mercury deposition on a local scale in areas near emissions point sources
(Section 3.6).
A local-scale atmospheric transport model (COMPDEP) was used to simulate mercury
deposition originating from six different mercury emissions source classes. The analysis was
conducted for two hypothetical lakes located in the Western and Eastern U.S. The proximity of these
lakes to the source was varied to examine the effect of this parameter on model predictions. To
account for the long-range transport of emitted mercury, the 50th percentile RELMAP atmospheric
concentrations and deposition rates were included in the estimates from the local air dispersion model.
June 1996 3-1 SAB REVIEW DRAFT
-------
3.2 Computer Models
The models used for the wildlife exposure assessment are identical to those used for the
human health assessment and are described in detail in Volume in of this Report. Atmospheric
transport models were used to simulate the deposition of mercury at two different geographical scales
(Table 3-1). A regional-scale analysis was conducted using the Regional Lagrangian Model of Air
Pollution (RELMAP). RELMAP calculates annual mean air concentrations and annual mean
deposition rates for each cell in a 40 km grid. This analysis covered the 48 contiguous states and was
based upon a recent inventory of mercury emissions sources (presented in Volume II of this Report).
A local-scale exposure analysis was conducted by using both RELMAP and a local air transport
model, COMPDEP, to generate hypothetical exposure scenarios for six source classes. COMPDEP is
designed to estimate deposition originating from local point sources (<40 km from the receptor). For
this analysis, COMPDEP results were summed with those from the RELMAP analysis, which in this
case can be thought of as providing the regional "background" to which local source material was
added. Three exposure scenarios were evaluated corresponding to lakes located 2.5, 10, and 25 km
from the emissions source. In each case, deposition information was used to estimate mercury
concentrations in water, averaged over the entire lake.
Table 3-1
Models Used to Predict Mercury Air Concentrations,
Deposition Fluxes and Environmental Concentrations
Model
RELMAP
COMPDEP
ffiM2
Description
Predicts average annual atmospheric mercury concentration and wet and dry
deposition flux for each 40 km2 grid in the U.S. due to all antbropocentric sources
of mercury in the U.S.
Predicts average concentration and deposition fluxes
source.
within 50 km of emission
Predicts environmental concentrations based on air concentrations and deposition
rates to watershed and water body.
The IEM2 model was used to translate both regional and local-scale mercury deposition
estimates into mercury levels in soil, water and biota. Mercury levels in fish were calculated from
average water concentrations using estimated BAFs for fish occupying trophic levels 3 and 4. It was
assumed throughout the wildlife exposure analysis that 100% of mercury contained in fish exists as
methylmercury.
The approach used to characterize mercury exposure to piscivorous wildlife is the same as that
used to characterize human exposure to mercury from consumption of contaminated fish (Volume III).
The same methodology was used to facilitate comparisons between exposure levels to human and
wildlife receptors.
June 1996 3-2 SAB REVIEW DRAFT
-------
3.3 Current Exposure of Piscivorous Wildlife to Mercury
Three avian species (eagle, kingfisher and osprey) and two mammalian species (otter and
mink) were assumed to be exposed to methylmercury through the ingestion of contaminated fish. Fish
consumption is thought to be the dominant mercury exposure pathway for piscivores (see Section 2).
Consequently, an analysis of these ecological receptors' methylmercury contact rate based on the daily
ingestion rate of fish is reasonable and appropriate.
The piscivorous bird's or mammal's mercury contact rate from fish consumption can be
estimated as the product of mercury levels in the fish and the daily amount of fish eaten. The trophic
level at which piscivores feed significantly impacts their exposure to mercury. Those piscivores
consuming a diet primarily consisting of trophic level 3 fish are expected to ingest approximately 5
times less methylmercury per gram of fish eaten than those eating trophic level 4 fish from the same
site. This assumed that the predator prey factor for trophic level 3 to 4 fish was approximately 5.
Animals consuming a mixture of trophic level 3 and 4 fish would experience (on a per gram of fish
basis) an intermediate level exposure. Finally, many top level predators are known to consume a
mixture of both aquatic and terrestrially-derived prey. In general, mercury levels in the tissues of
terrestrial animals are much lower than those of fish. The simplifying assumption can, therefore, be
made that most terrestrially-derived prey contain no mercury. A special case exists, however, when a
terrestrial animal (e.g., a raccoon) feeds on aquatic biota and is itself preyed upon by a larger
terrestrial animal (e.g., the Florida panther). A similar situation exists when a piscivorous bird (e.g.,
the herring gull) is consumed by a larger bird (e.g., the bald eagle). In these situations, the potential
exists for the top predator to obtain a higher mercury dose than it would otherwise receive from a
strictly fish-based diet The extent of this increase would depend hi turn upon the proportion of the
diet composed of these mammalian and avian prey items, and the extent to which the prey items
themselves accumulate mercury hi excess of levels found at trophic levels 3 and 4-.
Exposure factors for the present analysis were obtained from two recent compilations of
wildlife dietary habits (U.S. EPA, 1993a, 1995a) and are shown in Table 3-2. Bald eagles were
assumed to eat fish derived from trophic levels 3 and 4, as well as prey derived from other sources.
Expressed as percentages, these prey items were assumed to contribute 74, 18 and 8% of the daily
dietary intake. For this Report, dietary items other than fish were assumed to contain no mercury.
Eagles are, therefore, expected to experience a greater methylmercury exposure per gram of fish
consumed than ospreys and kingfishers, which were assumed to consume only trophic level 3 fish.
Part of this increase, however, is offset by the contribution of uncontaminated prey consumed by
eagles. Similarly, among the mammals, otters, which were assumed to consume an 80/20 mix of
trophic level 3 and 4 fish are expected to have a greater methylmercury exposure per gram of fish
consumed than mink, which were assumed to eat only trophic level 3 fish. In addition, 10% of the
mink diet was assumed to consist of uncontaminated prey items.
The ratio of grams fish consumed per day to piscivore body weight is also significant in
estimating mercury exposure on a g/kg bw/d basis. The greater this ratio, the higher the resulting
mercury exposure, assuming that methylmercury concentrations in fish remain constant. For example,
osprey and kingfishers each consume trophic level 3 fish only. Kingfishers consume an amount of
fish equivalent to about 50% of their body weight each day, while osprey consume roughly 20% of
their body weights in fish per day. The resulting average daily intake of methylmercury in ug/g body
weight will, therefore, be higher hi kingfishers. The source of the measured fish residue data was a
study entitled "A National Study of Chemical Residues hi Fish" conducted by U.S. EPA (1992b) and
also reported in Bahnick et al. (1994). This is the only identified nationwide fish collection effort that
June 1996 3-3 SAB REVIEW DRAFT
-------
Table 3-2
Exposure Parameters for Mink, Otter, Kingfisher, Osprey, and Eagle
Species
Mink
Otter
Kingfisher
Osprey
Eagle
BodyWt
(WtA)
kg
0.80
7.40
0.15
130
4.60
Ingestion
Rate
(FA>
kg/d
0.178
1.220
0.075
0300
0300
Drinking Rate
-------
value. An increasing amount of this type of information is being reported, however, and it is possible
to use these data to bound an analysis of mercury fate and effects in a hypothetical watershed. For the
purposes of this analysis, mercury air concentration, deposition rate and soil concentration (initial
condition) were set equal to 1 ng/m3, 10 ug/m2/yr, and 50 ng/g, respectively (as total mercury). These
values were then used as inputs to an indirect exposure model (IEM2) to predict mercury residues in
fish. The scientific basis for these input values is described in Volume in.
IEM2 uses atmospheric chemical loadings to perform mass balances on a watershed soil
element and a surface water element The mass balances are performed for total mercury, which is
assumed to speciate into three components: Hg°, Hg2"1" and methylmercury. The fraction of mercury
in each of these components is specified for the soil and the surface water elements. Loadings and
chemical properties are given for the individual mercury components, and the overall mercury
transport and loss rates are calculated.
IEM2 first performs a terrestrial mass balance to obtain mercury concentrations in watershed
soils. Soil concentrations are used along with vapor concentrations and deposition rates to calculate
concentrations in various food plants. These are used, in turn, to calculate concentrations in animals.
IEM2 next performs an aquatic mass balance driven by direct atmospheric deposition along with
runoff and erosion loads from watershed soils. This analysis was conducted for two hypothetical
waterbodies located in the Western U.S. and the Eastern U.S. Predicted mercury concentrations in
water, soil and benthic sediments are presented in Table 3-4.
Table 3-4
Mercury Concentrations in Water and Sediment Predicted Using a Mercury Air Concentration
of 1 ng/m3, Deposition Rate of 10 jig/m2/yr, and Soil Concentration of 50 ng/g
Parameter
Total Mercury Water Concentration (ng/L)
Percent of Mercury Dissolved (%)
Predicted Suspended Sediment Concentration (mg/L)
Total Mercury Benthic Sediment Concentration (ng/g)
Eastern Setting
1.02
71
3.17
110
Western Setting
1.00
77
2.15
118
Mercury residues in fish were estimated by making the simplifying assumption that aquatic
food chains can be adequately represented using four trophic levels. Respectively, these trophic levels
are the following: level 1 - phytoplankton (algal producers); level 2 - zooplankton (primary
herbivorous consumers); level 3 - small forage fish (secondary consumers); and level 4 - larger,
piscivorous fish (tertiary consumers). This type of food chain typifies the pelagic assemblages found
in large freshwater lakes, and has been used extensively to model bioaccumulation of hydrophobia
organic compounds (see for example Thomann, 1989; Clark, 1990; Gobas, 1993). It is recognized,
however, that food chain structure can vary considerably among aquatic systems resulting in large
differences in bioaccumulation in a given species of fish (Putter, 1994; Cabana et al., 1994). In
addition, this simplified structure ignores several important groupings of organisms, including benthic
detritivores, macroinvertebrates, generally, and herbivorous fishes.
June 1996
3-5
SAB REVIEW DRAFT
-------
Methylmercury concentrations in fish were derived from total mercury water concentrations by
using BAFs for trophic levels 3 and 4 (Table 3-5). Respectively, these BAFs are 66,200 and 335,000.
The BAFs selected for these calculations were estimated from existing field data. A detailed
description of their derivation is presented in Appendix A.
Table 3-5
Methylmercury Concentrations in Fish (ug/g) Predicted Using a Mercury Air Concentration of
1 ng/m3, Deposition Rate of 10 ug/m2/yr, and Soil Concentration of 50 ng/g
Trophic Level
Trophic Level 3 Fish
Trophic Level 4 Fish
Predicted Fish Concentration (pg/g)
Eastern Site
0.049
0.241
Western Site
0.052
0.259
An effort was also made to simulate the variability around these fish residue values (Table
3-6). This was accomplished by using percentile information for the BAF estimates developed in
Appendix A (Tables A-9 and A-10). The water concentration used to drive the variability analysis
was 0.7 ng/L dissolved total mercury. The results of this analysis demonstrate the large variability in
fish residues that may occur at a given water concentration. This variability is due hi turn to the large
variability hi field-derived BAF values.
Table 3-6
Percentiles of Predicted Methylmercury Concentrations in Fish (ug/g) Based on a
Total Mercury Dissolved Water Concentration of 0.7 ng/L
Parameter
Trophic 3 BAF
Predicted Fish Concentration (ug/g)
Trophic 4 BAF
Predicted Fish Concentration (ng/g)
Geometric
Mean
67,000
0.05
335,000
0.23
Percentile of Distribution
5th
6,400
0.00
22,700
0.02
25th
25,400
0.02
111,000
0.08
50th
66,200
0.05
336,000
0.24
75th
172,400
0.12
1,000,000
0.70
95th
684,000
0.48
4,700,000
330
3.5 Regional-scale Exposure Estimates
There are many stationary, anthropogenic mercury sources in the U.S., and the impact of these
emissions may not be entirely limited to the local area around the facility. To account for impacts of
mercury emitted from these non-local sources, the long-range transport of mercury was simulated
using the RELMAP model. The RELMAP model was used to predict the average annual atmospheric
June 1996
3-6
SAB REVIEW DRAFT
-------
mercury concentration and the wet and dry deposition flux for each 40 km2 grid in the continental
U.S. The emission, transport and fate of airborne mercury over the continental U.S. were modeled
using meteorologic data for the year of 1989. This year was assumed to be a typical year from an
atmospheric dispersion perspective. Inputs to the RELMAP model were obtained from the mercury
emissions inventory presented as Volume n of this Report In all, over 10,000 mercury emitting cells
within the U.S. were addressed. A detailed description of the RELMAP model is provided hi
Appendix D to Volume in.
3.5.1 Analysis of RELMAP Results
In the first stage of analysis, estimated total mercury deposition data were used with
ARC/INFO cartography software to generate U.S. map overlays. The overlays can be applied to
similar scale maps of natural resources and species distributions or combined with additional data such
as acid deposition and alkalinity or pH of surface waters. Figure 3-1 shows RELMAP projections for
total (including wet and dry) anthropogenic mercury deposition. Nearly all the land area east of the
Mississippi River is projected to receive mercury deposition greater than 5 ug/m2. Projections for the
highly industrialized northeast and south Florida are projected to receive more than 20 ug/m2.
RELMAP results are projections that may differ quantitatively from actual sampling data for a given
locale. It is anticipated, however, that additional sampling data will confirm the prediction that
mercury is deposited in significant quantities over large geographic areas.
Limitations on data precluded a quantitative, nation-wide analysis of the exposure of
piscivorous wildlife to mercury. Existing data are sufficient, however, to permit a qualitative analysis.
In the case of plant life, analysis was limited to plotting the location of federally threatened or
endangered species and indicating where threatened populations coincide with estimated high mercury
deposition.
Avian wildlife selected for this analysis included species that are widely distributed
(kingfishers) and narrowly distributed (bald eagles), as weU as birds whose range fell within areas of
high mercury deposition (ospreys and common loons). All the birds selected were piscivores that feed
at or near the top of aquatic food chains and are therefore at risk from biomagnified mercury.
Two of the mammals selected for this analysis (mink and river otters) are piscivorous and
widely distributed. The other mammal selected, the Florida panther, is not widely distributed but is
listed as an endangered species. The Florida panther lives hi an environment known to be
contaminated with mercury and preys upon small mammals (such as raccoons) which may contain
high tissue burdens of mercury.
The maps and map overlays that follow were used to examine in a qualitative fashion the
potential for anthropogenic mercury to impact representative piscivorous species in a variety of
ecosystems. Animal distribution information was obtained from the Nature Conservancy (1994)
3.5.2 Locations of Socially Valued Environmental Resources
Major freshwater lakes and river systems potentially affected by atmospheric mercury
deposition are illustrated in Figure 3-2. Most of the freshwater located in the lower 48 states occurs in
areas where mercury deposition is predicted to be high. Because mercury accumulates in sediments, it
is anticipated that significant mercury inputs to surface waters will continue for a long period of time
June 1996 3-7 SAB REVIEW DRAFT
-------
Figure 3-1
Total Anthropogenic Mercury Deposition
Base Case
Deposition 1-5 ug/m2
Deposition 5-10 ug/m2
Deposition > 10 ug/m2
June 1996
3-8
SAB REVIEW DRAFT
-------
Figure 3-2
Major Rivers and Lakes
Deposition 1-5 ugAn2
Deposition 5-10 ugftnZ
Deposition > 10 ug/m2
June 1996
3-9
SAB REVIEW DRAFT
-------
even if atmospheric deposition is substantially reduced. The Great Lakes are particularly vulnerable
because of the length of time necessary to replenish contaminated water with clean fresh water.
Figure 3-3 shows the location of national resource lands, which include national parks and
monuments, national forests, wildlife refuges and Native American reservation lands. The area of
national resource lands that are predicted to have high mercury deposition is relatively small when
compared with the total area of national resource lands, most of which are located in the western
states. The small size of eastern resources makes them especially vulnerable to the effects of mercury
because depleted wildlife populations cannot easily be repopulated from less-impacted adjoining
regions. Increasingly, natural areas may become "islands" surrounded by development The loss of
biodiversity is an important problem that could be exacerbated by the added stress of mercury toxicity.
3.5.3 Airborne Deposition Overlay with Threatened and Endangered Plants
Figure 3-4 shows the geographic locations of populations of threatened and endangered plant
species overlaid with RELMAP's predicted mercury deposition. Large concentrations of endangered
plant populations exposed to high levels of deposition occur in central and southern Florida, along the
northeastern coastal region and scattered throughout the midwest Mercury has been demonstrated to
have adverse impacts on a number of plant species (see Section 2).
3.5.4 Regions of Concern Defined by High Mercury Deposition Coincident with Acidic Surface
Water
Figure 3-5 overlays airborne deposition predictions and areas where surface water pH is 5.5 or
lower (NAPAP, 1990). All areas where >5% of surface waters are at or below pH 5.5 are
subsequently referred to as "regions of concern". This distinction is based on the observation that
mercury concentrations in fish flesh have been positively correlated with low pH. Designation of a
particular area as a region of concern implies an increased risk of mercury toxicity to wildlife.
Regions of concern could be used to define critical habitat predict potentially endangered species, and
may be useful to identify future research needs.
3.5.5 Regions of Concern Overlay with Wildlife Species Distribution Maps
Figure 3-6 shows the range of kingfisher habitat and areas where this habitat overlaps with
regions of concern. Kingfishers are piscivorous, consuming fish primarily from trophic level 3.
Approximately 8% of the kingfisher's range occurs within regions of concern and mercury does not
appear to be a threat to the species nationwide.
Figure 3-7 overlays the range of bald eagle habitat onto regions of concern. Although a
recovery in the population of bald eagles hi the lower 48 states has resulted hi a status upgrade from
"endangered" to "threatened", bald eagle populations are still depleted throughout this range. Bald
eagles can be found seasonally hi large numbers hi several geographic locations, but most of these
individuals are transient and the overall population is still small. Historically, eagle populations in the
lower 48 states have been adversely impacted by the effects of bioaccumulative contaminants
(primarily DDT and perhaps also PCBs). Approximately 17% of the bald eagle's range overlaps
regions of concern. The risk to eagles posed by mercury appears to be greatest in the Great Lakes
region, the northeastern Atlantic states and south Florida.
June 1996 3-10 SAB REVIEW DRAFT
-------
Figure 3-3
National Resource Lands
National Park, Monument, Lakeshore, Parkway, Battlefield, Recreation area
National forest or grassland
National wildlife refuge, game preserve, fish hatchery
National scenic waterways or wilderness area
Native American reservation lands
Military reservation
June 1996
3-11
SAB REVIEW DRAFT
-------
Figure 3-4
Threatened and Endangered Plan Species and anthropogenic Mercury Deposition
Threatened or Endangered Plant
Deposition 1-5 ug/m2
Deposition 5-10 ug/m2
Deposition > 10 ug/m2
June 1996
3-12
SAB REVIEW DRAFT
-------
Figure 3-5
Surface Water and pH< 5.5 and Anthropogenic Mercury Deposition
NORTHEAST
WEST
IN ERIOR SOUTHEAST
6 20% pH < = 6.6, Deposition 1-6 ug/m2
6-20% pH < - 5.6. Deposition 6-10 ug/m2
6-20% pH < - 5.6. Deposition > 10 ug/m2
> 20% pH < - 6.6, Deposition 1-6 ug/m2
> 20% pH < - 6.6. Deposition 6-10 ug/m2
> 20% pH < = 5.5. Deposition > 10 ug/m2
MID-ATLANTIC
COASTAL PLAIN
aORIDA
June 1996
3-13
SAB REVIEW DRAFT
-------
Figure 3-6
Kingfisher Range, Surface Water with pH < 5.5, and Anthropogenic Mercury Deposition
NORTHEAST
WEST
I j fUna« (or thl* «paole«
B-20% pH < « 6,6, D«sK.iltlon 1-S ufl/m2
6-20% pH < " 5.6, DvpotMon 5-10 ug/m2
B-20% pH < - 6.6, DapotWon > 10 ug/m2
> 20% pH < - 6,6, DapocHlon 1-6 ug/ra2
> 20% pH < - 6.6, Deposition 6-10 Ufl/m2
> 20% pH < = 6.6, Deposltlim > 10 ua/m2
MID-ATLANTIC
COASTAL PLAIN
FLORIDA
-------
Figure 3-7
Bald Eagle Range, Surface Water with pH < 5.5, and Anthropogenic Mercury Deposition
NORTHEAST
WEST
Range for tills •paolec
6-20% pH < = 5.5, Deposition 1-6 u0/m2
6-20% pH < - 5.6, Deposition 6-10 ug/m2
6-20% pH < - 5.6, Dapoiltlan > 10 ug/m2
> 20% pH < - 6.6, Deposition 1-6 Ufl/m2
> 2O% pH < - 6.6, Deposition 6-10 ug/m2
> 20% pH < = 6.6, Deposition > 10 ug/n>2
MID-ATLANTIC
COASTAL PLAIN
aORIDA
June 1996
3-15
SAB REVIEW DRAFT
-------
Figure 3-8 indicates where the range of osprey coincides with regions of concern. Nationwide,
approximately 13% of the osprey's range overlap.s these regions; however, a much larger fraction of
the osprey's eastern population occurs within these regions. The osprey diet consists almost
exclusively of fish, and they are known to take dead fish from the water surface if they are fresh.
Their position at the top of the aquatic food chain places ospreys at risk from toxins that
bioaccumulate. Osprey populations underwent a severe declines during the 1950s through the 1970s
which has been linked to exposure to DDT.
Figure 3-9 depicts areas where the range of the common loon coincides with regions of
concern. Nearly 23% of the loon's range is located in regions of concern. Moreover, nearly all of the
loon's range occurs in regions where mercury deposition is predicted to be high (Figure 3-1). Limited
data from the study of a mercury point source showed that loon reproductive success was negatively
correlated with exposure to mercury in a significant dose-response relationship (Section 2.3.3).
Residue data, combined with field observations, suggest that loon populations in areas of Minnesota
and Wisconsin may be adversely impacted by mercury originating from airborne deposition.
Figure 3-10 shows the Florida panther's range. Although the panther's range falls outside of
identified regions of concern (<1%), the species habitat is contiguous with this region. Section 2.3.3
describes data which establish a link between mercury exposure and adverse impacts on the Florida
panther. Mercury levels found in Florida panther tissue approach levels that are frankly toxic in other
feline species. The State of Florida has taken measures to reduce the risk to panthers posed by
mercury. Existing plans include modification of surface vegetation to increase the number of deer
available as prey in order to reduce the reliance of panthers on raccoons. As indicated previously,
raccoons frequently feed at or near the top of aquatic food chains and can accumulate substantial tissue
burdens of mercury.
Figure 3-11 gives the range of the mink where this habitat coincides with regions of concern
(approximately 9% of the range, nationwide). Mink occupy a large geographic area and are common
throughout this range, although rarely observed because of their nocturnal habits. In general, mink
prey on small mammals for most of the year; however, some populations prey primarily on fish and
aquatic birds. Mink that prey on aquatic animals are most at risk from mercury contamination. In
additional small mammalian and avian predators may be a greater risk than large predators due to
higher food consumption rate per unit of body weight (Section 2.3.2).
Figure 3-12 shows the range of the river otter where this habitat coincides with regions of
concern (approximately 14% of the range, nationwide). River otters occupy large areas of the United
States, but their population numbers are thought to be declining in the midwestern states. The river
otter's diet is almost exclusively of aquatic origins and includes fish (primarily), crayfish, amphibians
and aquatic insects. The species of fish taken depends on the fish's ability to escape capture. The
consumption of large, piscivorous fish puts the river otter at risk from bioaccumulative contaminants
such as mercury. Otter population declines do not overlap to a large extent with regions of concern;
however, the area of decline does coincide with RELMAP predictions of high mercury deposition rate
(Figure 3-1).
June 1996 3-16 SAB REVIEW DRAFT
-------
Figure 3-8
Osprey Range, Surface Water with pH < 5J5,
and Anthropogenic Mercury Deposition
(Detail: Eastern
ftong*tarttti*»fwci**
5-20% pH < - 5.5, Ocpoiitten 1-5 ugAn2
5-20% pH < » 5.5, D»po»Won 5-10 ug/m2
5-20% pH < - S.S, D«po»«oo > 10 agfrnZ
-------
Figure 3-9
Common Loon Range, Surface Water with pH < 5.5, and Anthropogenic Mercury Deposition
NORTHEAST
WEST
Range for tills spade*
6-20% pH < = 5.6, Deposition 1-6 ug/m2
6-20% pH < - 6.6, Deposition 6-10 U9lm2
6-20% pH < - 6.6, DapotHion > 10 ug/m2
> 20% pH < - 6.6, Deposition 1-6 ue/m2
> 20% pH < - 6.6, Deposition 6-10 ug/m2
> 20% pH < = 6.6. Deposition > 10 ug/m2
MID-ATLANTIC
COASTAL PLAIN
FLORIDA
June 1996
3-18
SAB REVIEW DRAFT
-------
Figure 3-10
Florida Panther Range, Surface Water with pH < 5.5,
and Anthropogenic Mercury Deposition
(Detail: Eastern
5-20% pH < - 5.5, DvpoiHion 1-5 ug/m2
5-20% pH < - 5.5. Deposition 5-10 ug/hi2
5-20% pH < - 5.5. Depoitton > 10 ug/m2
> 2O% pH < - 5.5. D«po»hion1-5 ufl/m2
> 20% pH < » 5.5. Dcpnitfon 5-10 uo/m2
> 2O% pH < « 5.5. DvpoaHlon > 10 ug/m2
June 1996
3-19
SAB REVIEW DRAFT
-------
Figure 3-11
Mink Range, Surface Water with pH <5.5, and Anthropogenic Mercury Deposition
NORTHEAST
WEST
Range (or thi» epeciee
6-20% pH < = 5.6, Deposition 1-6 ug/m2
6-20% pH < - 6.6. Deposition 6-10 ug/m2
6-20% pH < - 6.6, Deposition > 10 ug/m2
> 20% pH < - 6.5, DapMl«on1-6 ug/m2
> 20% pH < - 6.6, Deposition 6-10 ug/m2
> 20% pH < = 6.6, Deposition > 10 ug/m2
MID-ATLANTIC
COASTAL PLAIN
FLORIDA
June 1996
3-20
SAB REVIEW DRAFT
-------
Figure 3-12
River Otter Range, Surface Water with pH <5.5, and Anthropogenic Mercury Deposition
NORTHEAST
WEST
Range for this (peoles
6 20% pH < = 6.6. Deposition 1-6 ug/m2
B-20% pH < m 5.6, Deposition 6-10 ug/ro2
6-20% pH < - 6.6, Depotition > 10 ug/m2
> 20% pH < - 6.6, Deposition 1-6 ug/m2
> 20% pH < - 6.5, Deposition 6-10 ug/m2
> 20% pH < = 6.5. Deposition > 10 ug/m2
MID-ATLANTIC
COASTAL PLAIN
FLORIDA
June 1996
3-21
SAB REVIEW DRAFT
-------
3.6 Local-scale Exposure Estimates
3.6.1 Approach and Assumptions
The goal of the local scale analysis was to evaluate the extent to which local mercury
emissions sources have the potential to create locally elevated mercury exposures for piscivorous
wildlife receptors. Air concentrations and deposition rates due to a single local source were predicted
using the COMPDEP model, specifically modified to model the atmospheric transport of mercury. For
the purposes of this study, hypothetical sources were assumed to contribute mercury in addition to that
simulated by RELMAP. Details, of the local-scale modeling exercise are presented in Volume HI of
this Report
Model plants (hypothetical mercury emitters) representing six source classes were developed to
represent a range of mercury emissions sources. The source categories were selected for the indirect
exposur^ analysis based on their estimated annual mercury emissions or their potential to be localized
point sources of concern. The categories selected were these: municipal waste combustors (MWCs),
medical waste incinerators (MWIs), utility boilers, chlor-alkali plants, primary copper smelters, and
primary lead smelters. Table 3-7 shows the process parameters assumed for each of these facilities.
Table 3-7
Process Parameters for the Model Plants Considered in the Local Impact Analysis
MODEL PLANT
Large MWC
Small MWC
Continuous MWI
Intermittent MWI
Large Coal-fired
Utility Boiler
Medium Coal-fired
Utility Boiler
Small Coal-fired
Utility Boiler
Medium Oil-fired
Utility Boiler
Chlor-alkali plant
Primary Copper
Smelter
Primary Lead Smelter
Plant Size
2250tons/d
200 tons/d
15001b/hr
capacity
(1000 Ib/hr
actual)
200 Ib/hr capacity
(133 Ib/hr actual)
975 Megawatts
375 Megawatts
100 Megawatts
285 Megawatts
300 tons
chlorine/d
180 tons Cu/d
304 tons lead/d
Stack
Ht.(tt)
230
140
40
40
732
465
266
290
10
505
350
Stack
Diam
(ft)
95
5
2.7
1.2
27
18
12
14
0.5
15
20
Baseline
Hg
Emission
Rate
(kgfrr)
1330
170
80
2.4
230
90
10
2
380
5360
2680
Spedation %
(Hg^/Hg2*/
methylmercury)
20/60/20
20/60/20
20/60/20
20/60/20
50/30/20
50/30/20
50/30/20
50/30/20
70/30/0
85/10/5
85/10/5
Exit
VeL
(m/sec)
21.9
21.9
7.3
7.3
31.1
26.7
6.6
20.7
0.1
6
2.8
Exit
Temperature
-------
The model plants were placed in hypothetical sites assumed to lie hi the eastern and western
U.S. The hypothetical sites were assumed to have flat terrain. Emitted mercury is thought to
influence both local and regional atmospheric concentrations and deposition. To account for the long
range transport of emitted mercury, the 50th percentile RELMAP atmospheric concentrations and
deposition rates were included in the estimates from the local air dispersion model. Model simulations
were generated for the same hypothetical watersheds and waterbodies described previously. These
waterbodies were assumed to be located 2.5,10 and 25 Km from the sources.
COMPDEP uses hourly meteorological data to estimate hourly air concentrations and
deposition fluxes within 50 km of a point source. For each hour, general plume characteristics are
estimated based on the source parameters (gas exit velocity, temperature, stack diameter, stack height,
wind speed at stack top, atmospheric stability conditions) for that hour. COMPDEP was run using
five years of actual meteorological data. The average values for air concentration and deposition rates
were then used as inputs for to IEM2 model. These values were assumed to be representative for 30
years, the assumed typical lifetime of a facility. During this 30-year period, the mercury concentration
in soil was allowed to build up, taking into account loss processes such as leaching, runoff and
erosion. Simulated values at the end of the 30-year period were then used as input to the water
portion of the IEM2 model to calculate steady-state water concentrations. Finally, the estimated water
concentrations were used to predict methylmercury concentrations in fish that occupy trophic levels 3
and 4. This was accomplished by multiplying the predicted total mercury dissolved water
concentration by the BAF at each trophic level. Wildlife receptors were assumed to ingest the fish at
rates given previously (Table 3-2).
3.6.2 Results of Local-scale Exposure Analysis
High rates of mercury deposition were associated with proximity to industrial sources emitting
substantial levels of divalent mercury (Tables 3-8 and 3-9). Additional factors that contributed to high
local deposition rates include low stack height and slow stack exit gas velocities. In general, total
mercury concentrations hi lake waters located 2.5 km from the source were much higher than levels
predicted at 10 or 25 km. This was due primarily to the dilution of the mercury emissions in the
atmosphere. Mercury concentrations in fish (hence the mercury exposure to piscivores) were
proportional to dissolved mercury levels in the local waters. At 10 and 25 km the water
concentrations and, thus, the predicted levels hi fish, are elevated relative to levels predicted when
only remote sites are modeled (i.e., RELMAP predictions). When the two hypothetical locations were
compared (western and eastern), higher mercury concentrations were predicted to occur in the
environmental media at the eastern location. This was due primarily to higher levels of precipitation
at the eastern site, which tends to remove mercury from the atmosphere. In this modeling effort,
mercury deposition patterns dominate soil, water and fish mercury concentrations. Although (as
described in Volume HI) watershed characteristics influence model simulations, the facility type, local
meteorology and terrain are generally more important for predicting local dissolved mercury
concentrations in water.
June 1996 3-23 SAB REVIEW DRAFT
-------
Table 3-8
Predicted Intakes for Wildlife Receptors for the Eastern Site
Eastern Site
Resource Plus Local Source
Deposition
2-5 km
Large MWC
Small MWC
Continuous MWI
Intermittent MWI
Large Coal-fired Utility Boiler
Medium Coal-fired Utility Boiler
Small Coal-fired Utility Boiler
Medium Oil-fired Utility Boiler
Chlor-alkali plant
Primary Copper Smelter
Primary Lead Smelter
10km
Large MWC
Small MWC
Continuous MWI
Intermittent MWI
Large Coal-fired Utility Boiler
Medium Coal-fired Utility Boiler
Small Coal-fired Utility Boiler
Medium Oil-fired Utility Boiler
Chlor-alkali plant
Primary Copper Smelter
Primary Lead Smelter
25km
Large MWC
Small MWC
Continuous MWI
Intermittent MWI
Surface Water
Concentration
(ng/L)
' 1.022
23.539
3.692
2.291
0.080
0.572
0.451
0.081
0.014
8.990
1307
7.195
3.015
0.527
0362
0.012
0.107
0.074
0.013 '
0.002
0.968
0.210
1.178
0.666
0.134
0.090
0.003
Large Coal-fired Utility Boiler 0.021
Medium Coal-fired Utility Boiler
0.016
Small Coal-fired Utility Boiler 0.003
Medium Oil-fired Utility Boiler 0.000
Chlor-alkali plant 0.209
Primary Copper Smelter 0.051
Primary Lead Smelter 0.295
Fish
Methylmercury
Concentrations
(Pg/g)
Tier 3
0.049
1.122
0.176
0.109
0.004
0.027
0.022
0.004
0.001
0.428
0.062
0343
0.144
0.025
0.017
0.001
0.005
0.004
0:001
0.000
0.046
0.010
0.056
0.032
0.006
0.004
0.000
0.001
0.001
0.000
0.000
0.010
0.002
0.014
Tier 4
0.242
5.575
0.874
0.543
0.019
0.136
0.107
0.019
0.003
2.129
0.309
1.704
0.714
0.125
0.086
0.003
0.025
0.017
0.003
0.000
0.229
0.050
0.279
0.158
0.032
0.021
0.001
0.005
0.004
0.001
0.000
0.050
0.012
0.070
Predicted Methylmercury Intake (ng/g bw/d)
Bald
Eagle
0.009
0.199
0.031
0.019
0.001
0.005
0.004
0.001
0.000
0.076
0.011
0.061
0.026
0.004
0.003
0.000
0.001
0.001
0.000
0.000
0.008
0.002
0.010
0.006
0.001
0.001
0.000
0.000
0.000
0.000
0.000
0.002
0.000
0.002
Osprey
0.010
0.224
0.035
0.022
0.001
0.005
0.004
0.001
0.000
0.086
0.012
0.069
0.029
0.005
0.003
0.000
0.001
0.001
0.000
0.000
0.009
0.002
0.011
0.006
0.001
0.001
0.000
0.000
0.000
0.000
0.000
0.002
0.000
0.003
Kingfisher
0.024
0.561
0.088
0.055
0.002
-0.014
0.011
0.002
0.000
0.214
0.031
0.171
0.072
0.013
0.009
0.000
0.003
0.002
0.000
0.000
0.023
0.005
0.028
0.016
0.003
0.002
0.000
0.001
0.000
0.000
0.000
0.005
0.001
0.007
River
Otter
0.014
0.332
0.052
0.032
0.001
0.008
0.006
0.001
0.000
0.127
0.018
0.101
0.042
0.007
0.005
0.000
0.002
0.001
0.000
0.000
0.014
0.003
0.017
0.009
0.002
0.001
0.000
0.000
0.000
0.000
0.000
0.003
0.001
0.004
Mink
0.010
0.225
0.035
0.022
0.001
0.005
0.004
0.001
0.000
0.086
0.012
0.069
0.029
0.005
0.003
0.000
0.001
0.001
0.000
0.000
0.009
0.002
0.011
0.006
0.001
0.001
0.000
0.000
0.000
0.000
0.000
0.002
0.000
0.003
June 1996
3-24
SAB REVIEW DRAFT
-------
Table 3-9
Predicted Intakes for Wildlife Receptors for the Western Site
Western Site
Regional Plus Local Source
Deposition
2.5km
Large MWC
Small MWC
Continuous MWI
Intermittent MWI
Large Coal-fired Utility Boiler
Medium Coal-fired Utility Boiler
Small Coal-fired Utility Boiler
Medium Oil-fired Utility Boiler
Chlor-alkali plant
Primary Copper Smelter
Primary Lead Smelter
10km
Large MWC
Small MWC
Continuous MWI
Intermittent MWI
Large Coal-fired Utility Boiler
Medium Coal-fired Utility Boiler
Small Coal-fired Utility Boiler
Medium Oil-fired Utility Boiler
Chlor-alkali plant
Primary Copper Smelter
Primary Lead Smelter
25 km
Large MWC
Small MWC
Continuous MWI
Intermittent MWI
Large Coal-fired Utility Boiler
Medium Coal-fired Utility Boiler
Small Coal-fired Utility Boiler
Medium Oil-fired Utility Boiler
Chlor-alkali plant
Primary Copper Smelter
Primary Lead Smelter
Surface Water
Concentration
(ng/L)
1.002
8.818
1.672
1.558
0.060
0.223
0.157
0.032
0.005
8.185
0.438
2.627
2.095
0.417
0323
0.010
0.066
0.039
0.008
0.001
1.007
0.109
0.651
0.776
0.149
0.095
0.003
0.021
0.012
0.003
0.000
0.263
0.039
0.235
Fish Methyunercury
Concentrations
(Mg/g)
Tier 3
0.052
0.458
0.087
0.081
0.003
0.012
0.008
0.002
0.000
0.426
0.023
0.137
0.109
0.022
0.017
0.001
0.003
0.002
0.000
0.000
0.052
0.006
0.034
0.040
0.008
0.005
0.000
0.001
0.001
0.000
0.000
0.014
0.002
0.012
Tier 4
0.259
2.279
0.432
0.403
0.016
0.058
0.041
0.008
0.001
2.115
0.113
0.679
0.541
0.108
0.084
0.003
0.017
0.010
0.002
0.000
0.260
0.028
0.168
0.201
0.038
0.025
0.001
0.005
0.003
0.001
0.000
0.068
0.010
0.061
Predicted Methyunercury Intake (ng/g
Bald
Eagle
0.009
0.081
0.015
0.014
0.001
0.002
0.001
0.000
0.000
0.076
0.004
0.024
0.019
0.004
0.003
0.000
0.001
0.000
0.000
0.000
0.009
0.001
0.006
0.007
0.001
0.001
0.000
0.000
0.000
0.000
0.000
0.002
0.000
0.002
Osprey
0.010
0.092
0.017
0.016
0.001
0.002
0.002
0.000
0.000
0.085
0.005
0.027
0.022
0.004
0.003
0.000
0.001
0.000
0.000
0.000
0.010
0.001
0.007
0.008
0.002
0.001
0.000
0.000
0:000
0.000
0.000
0.003
0.000
0.002
Kingfisher
0.026
0.229
0.043
^0.040
0.002
0.006
0.004
0.001
0.000
0.213
0.011
0.068
0.054
0.011
0.008
0.000
0.002
0.001
0.000
0.000
0.026
0.003
0.017
0.020
0.004
0.002
0.000
0.001
0.000
0.000
0.000
0.007
0.001
0.006
River
Otter
0.015
0.136
0.026
0.024
°-00l
0.003
0.002
0.000
0.000
0.126
0.007
0.040
0.032
0.006
0.005
0.000
0.001
0.001
0.000
0.000
0.015
0.002
0.010
0.012
0.002
0.001
0.000
0.000
0.000
0.000
0.000
0.004
0.001
0.004
bw/d)
Mink
0.010
0.092
0.017
0.016
0.001
0.002
0.002
0.000
0.000
0.085
0.005
0.027
0.022
0.004
0.003
0.000
0.001
0.000
0.000
0.000
0.010
0.001
0.007
0.008
0.002
0.001
0.000
0.000
0.000
0.000
0.000
0.003
0.000
0.002
June 1996
3-25
SAB REVIEW DRAFT
-------
4. EFFECT OF AIRBORNE MERCURY ON PISCIVOROUS AVIAN
AND MAMMALIAN WILDLIFE
As described in Section 2 of this volume, mercury bioconcentrates, bioaccumulates and
biomagnifies in aquatic food chains. These processes result in mercury residues in fish that are much
higher than concentrations in the water in which they live, thereby providing an enriched contaminant
source for piscivorous avian and mammalian wildlife. Section 4.2 presents derivation of a wildlife
criterion level (WC) for mercury that is protective of piscivorous wildlife. This WC is defined as the
concentration of mercury in water that, if not exceeded, protects avian and mammalian wildlife
populations from adverse effects resulting from ingestion of surface waters and from ingestion of
aquatic life taken from these surface waters. The health of wildlife populations may, therefore, be
considered the assessment endpoint of concern.
Calculation of a WC for mercury is based upon the use of a wildlife reference dose approach,
combined with knowledge of the extent to which mercury becomes concentrated in aquatic food
chains. The methods used to calculate this criterion value are based on those described in the
Proposed Great Lakes Water Quality Guidance for the Great Lakes Water Quality Initiative (U.S. EPA,
1993c) and implemented in the final Water Quality Guidance for the Great Lakes System (U.S. EPA,
1995b), henceforth referred to as the "Proposed Guidance" and "Final Guidance," respectively. This
approach yields a single measurement endpoint, which is the total mercury concentration in water that
is believed to be protective of piscivorous wildlife. It should be noted, however, that this endpoint can
be related to total mercury residues in fish through the use of bioaccumulation factors (BAFs).
BAFs for trophic levels 3 and 4 (forage fish and larger, piscivorous fish, respectively) are
estimated in Appendix A. It is recognized that there is considerable natural variability with respect to
the accumulation of mercury in aquatic food chains, which contributes in turn to variability in trophic
relationships and BAFs. In addition, there is a lack of understanding of fundamental processes that
contribute to methylation of mercury and subsequent bioaccumulation in aquatic organisms.
Additional uncertainty derives from ongoing improvements in sampling technique and analytical
methodology.
Tempering these uncertainties is a large and growing volume of both laboratory and field data
for mercury. From the perspective of WC development, field data are of particular interest. The
Proposed Guidance (or GLWQI) provides that when sufficient field data are available, field-derived
BAFs should take precedence over values estimated from laboratory studies or by employing empirical
relationships (e.g., correlation with chemical hydrophobicity). It was thought that when sufficient field
data exist, field-derived BAFs represent the best obtainable estimate of mercury bioaccumulation by
fish. An effort was made, therefore, to collect both field and laboratory data and to characterize the
variation arising from the aforementioned sources and incorporate it into the analysis by using a Monte
Carlo simulation approach. The results of this effort are summarized in Section 4.1.
4.1 Bioaccumulation Factors (BAFs) for Magnification of Methylmercury in the Aquatic
Food Chain
4.1.1 Definition of BAFs and Overview
The bioaccumulation factor (BAF) for any given trophic level is defined as the ratio of the
total mercury concentration in fish flesh divided by the concentration of total dissolved mercury in the
water column. The BAF represents the accumulation of mercury in fish of a specific trophic level
June 1996 4-1 SAB REVIEW DRAFT
-------
from both water intake and predation on contaminated organisms. The BAF is a principal input
variable in the Indirect Exposure Model (see Volume in) and is used to link estimates of mercury
deposition to exposure levels for fish-consuming species.
In this Report BAFs are estimated for trophic level 3 (foraging fish) and trophic level 4
(piscivorous fish) designated as BAF3 and BAF^ respectively. BAF4 is estimated by three different
methods and BAF3 by two. The result, or output, of each estimation is a distribution of BAF values,
each associated with some degree of likelihood. The three methods by which BAF4 is estimated are
the GLWQI method, the BAF x PPF method and the field-derived method from measured BAFs at
trophic level 4. BAF3 is estimated by the GLWQI method and directly from measured BAFs at
trophic level 3. Each of these methods is described in detail in Appendix A and summarized in
Section 4.1.3. BAF4 is intended to be representative of the random selection of a trophic level 4 fish
from a random lake in a random geographical location. It is meant to be used to estimate the
concentration of methylmercury in such a randomly-selected fish when multiplied by the total
dissolved mercury (inorganic and organic combined) concentration in the water column. BAF3 -
performs the same function for trophic level 3 fish.
The general approach used in this analysis is estimation of BAFs using probabilistic Monte
Carlo simulation methods as described hi Appendix A. This approach was taken to allow quantitative
expression of the overall variability surrounding the various estimates of the BAFs and to determine
the relative sensitivity of the estimates to specific individual variables.
4.1.2 BAF Estimation Methods
GLWQI Method
The GLWQI method is essentially the same as that hi the Proposed Guidance (U.S. EPA,
1993c). The formula is given hi equation 1.
BAFn=BCFHgx FCM,, (1)
where
n is the trophic level for which the BAF is estimated,
BCFH is the weighted-average bioconcentration factor (BCF) for total mercury at
trophic level 1 and
is the food-chain multiplier representing the cumulative biomagnification of
mercury from trophic level 2 to trophic level n, n=[3,4].
The formula for BCFHg is given hi equation 2.
BCFHg = (BCFmHg x MeHgw) + (BCF^g x (1 - MeHgw)) (2)
where
BCFmH is the bioconcentration factor for methylmercury at trophic level 1,
BCF is the bioconcentration factor for inorganic mercury at trophic level 1
inorganic mercury at trophic level 1 and
June 1996 4-2 SAB REVIEW DRAFT
-------
MeHgw is the fraction of total mercury in the water column that is in the methyl form.
The formulas for FCM3 and FCM4 are given in equations 3 and 4, respectively.
FCM3 = PPF2 x PPF3 (3)
FCM4= PPF2x PPF3x PPF4 (4)
where
PPF2 is the predator-prey factor at trophic level 2 representing the biomagnification
of mercury in zooplankton as a result of feeding on contaminated
phytoplankton,
PPF3 is the same for trophic level 3 fish feeding on contaminated organisms and
PPF4 is the same for trophic level 4 fish feeding on trophic level 3 fish.
Distributions were assigned to each of the variables in equations 1-4 based on data available in
the published literature. The basis and description of the distribution for each variable are described in
Appendix A. The nominal values for some of the variables are not the same as presented in the
Proposed Guidance (U.S. EPA, 1993c) because of differing assumptions and different approaches to
data analysis. The differences in the input variables do not result in significant differences in the
BAFs estimated by the Monte Carlo simulations and the BAFs calculated in the Proposed Guidance.
BAF x PPF Method
The formula for the calculation of BAF4 by this method is given in equation 5.
BAF4= BAF3xPPF4 (5)
where
BAF3 is the field-measurement-derived distribution for the BAF at trophic level 3 and
PPF4 is the same as for the GLWQI method.
Field-derived Method
This method estimates BAF3 and BAF4 directly from measurements of BAFs in field studies. The
derivation of the BAF distributions is described in Appendix A.
4.1.3 Results of BAF Simulations and Recommended Values
Results of the Monte Carlo simulations for each of the methods are given in Table 4-1, which
shows representative statistics for each BAF output distribution. All of the statistics are given as the
geometric equivalents (antilogs) of the actual values generated by the simulations. There is a large
variance in the distributions, which cannot be separated into variability in BAFs and uncertainty in
their estimation. In the absence of appropriate local data, it is recommended to use the geometric mean
values in Table 4-1. BAFs derived from data collected at the site of concern are preferred to the
estimated values in Table 4-1.
June 1996 4-3 SAB REVIEW DRAFT
-------
Table 4-1
Summary of Bioaccumulation Factors for Trophic Levels 3 and 4
(mean, 5%, and 95% values)
Recommended
Method
Geometric Mean
S^pctl
95th pctl
GSDa
BAF3
66,200
Field-derived
66,200
6,400
684,000
4.14
GLWQI
25,200
2,310
308,000
4.41
BAF4
335,000
BAF3x •
PPF4
335,000
22,700
4,700,000
5.05
Field-derived
400,000
23,600
6,780,000
5.59
GLWQI
136,000
8,760
2,070,000
5.25
Geometric Standard Deviation
The selection of the BAF3 x PPF4 as the recommended approach is based on several
considerations. Although the mean values of all three BAF4 simulations agree within a factor of 3, the
GLWQI results stand somewhat apart The mean value of 136,000 for the GLWQI method falls at the
30th and 35th percentile of the BAF4 distributions for the field-derived and BAF x PPF methods,
respectively. The GLWQI method is also more complex with more variables and assumptions than the
other two approaches. The B AF x PPF and field-derived methods represent a consolidation of earlier
stages of the GLWQI method and should give more accurate results than the GLWQI method provided
that the data defining the distributions are at least as good as the data defining the GLWQI variables.
Five studies are available for defining BAF3; however, three of the critical variables in the GLWQI
method are based on only one or two studies. In addition the field measurements for BAF3 and PPF4
apply directly to variables, while the BCFs in the GLWQI approach do not. That is, the measurements
are taken directly from fish at the appropriate trophic levels for BAF3 and PPF4; BCFmH and BCF^g
apply to phytoplankton (trophic level 1) but are estimated from measurements in trophic level 3 fish.
The BAF3 x PPF4 approach is also less variable than either of the other two methods, as indicated by
the geometric standard deviation (Table 4-1).
4.1.4 Sensitivity Analysis
Sensitivity analyses were conducted to examine the effect of changes in assumptions. Three
factors have been studied in this analysis: sensitivity of output to individual input variables, PPF4
disaggregation and correlation of input variables.
The relative contribution of each input variable to the variance of the output distribution was
determined for the BAF x PPF method. BAF3 and PPF4 contribute 64% and 36% to the variance of
BAF4, respectively. Acquisition of additional data for the determination of BAF3 is expected to
decrease the uncertainty of BAF4. Refinements to the PPF4 distribution, however, will have a
significant impact on the mean value of BAF4 for application in specific scenarios.
June 1996
4-4
SAB REVIEW DRAFT
-------
Different species of piscivores are likely to feed on a restricted fish size range; alternate PPF4
distributions were, thus, constructed from restricted ranges of the data to represent specific fish size
(age) distributions. The distributions represent three ranges of fish sizes, described in Appendix A.
Because of the limited amount of data, these distributions are hypothetical, in part. The alternate PPF4
distributions do have potential significance in real-world scenarios, however. The mid and high PPF4
distributions, for example, represent estimates of PPF4 for standardized (two- to four-year-old fish) and
older tier 4 fish (eg.,large pike), respectively. These estimates could apply to situations where the size
of the consumed fish is known to be at one end of the distribution. The mean estimates of BAF4 vary
from 150,000 to 900,000 using the alternate PPF4 distributions.
The BAF4 simulation based on the BAF x PPF method assumes that the input variables are
independent Both BAF3 and PPF4, however, depend on the concentration of methylmercury in
trophic level 3 fish (BAF3 is directly dependently, and PPF4 is inversely dependent). The assumption
of independence, hi this case, increases variability hi the output distribution. Simulations were run
with varying assumptions of correlation between the two variables; the details are presented* in
Appendix A. The results of the simulations show that a moderately strong correlation (50%) between
B AF3 and PPF4 would have a significant effect on the spread of the output, reducing it by a factor of
3.5. If the correlation were weaker (10%), the spread of the distribution would be reduced by
only 22%.
4.1.5 Uncertainty and Variability
Generally, in the representation of the input and output distributions, there are no distinctions as
to size or species of fish, location or type of lake (eutrophic or oligotrophic), water column pH,
absolute mercury concentrations (in fish or water) or relative methylmercury concen-trations in the
water column. The available data are insufficient to make these distinctions. Field data are heavily
biased towards northern (oligotrophic) lakes and somewhat towards smaller (younger) fish.
There is no distinction between variability and uncertainty in the BAF4 distributions. That is,
the variability in the output distributions reflects both^ naturally variable processes and the uncertainty
around those processes. For example, the BAF4 distributions include variability in the BAF associated
with variations in fish size combined with the variability associated with methylmercury-generating
processes in the water column and the measurement uncertainties associated with both factors.
The large amount of variability evidenced by the data and reflected in the BAF distributions
arises from several identifiable but, as yet, unquantified sources. A primary source of variability in
both B AF3 and PPF4 is the dependence of methylmercury bioaccumulation on the age of the fish.
Although the age of fish is probably a major contributor to the variance of PPF4, the influence of age
on BAF3 is probably much less. Because the value BAF4 is more dependent on the magnitude of
B AF3 than on PPF4 the total reduction in variability of the B AF4 by accounting for fish age may not
be large. A second source of variability is seasonal variation of total dissolved mercury in the water
column. While the concentration of methylmercury in fish flesh is presumably a function of the
varying water concentration, specific values for BAF3 are generally calculated from single
representative values.
Perhaps the greatest source of variability is that of model uncertainty; that is, uncertainty
introduced by failure of the model to account for significant real-world processes. The simple linear
BAF model relating methylmercury in fish to total mercury in water masks a number of nonlinear
processes leading to the formation of bioavailable methylmercury in the water column. Much of the
June 1996 ' 4-5 SAB REVIEW DRAFT
-------
variability in field data applicable to the estimation of mercury BAFs can be attributed to differences
between individual organisms and between aquatic systems. For example, in lake surveys conducted
within a relatively restricted geographic region, large differences can exist between lakes with respect
to mercury concentrations in a given species of fish (see for example Cope et al., 1990; Grieb et al.,
1990; Sorenson et al., 1990; Jackson, 1991; Lange et al., 1993). These observations have led to the
suggestion that much of the variability in fish mercury levels is due to differences in local bio-
geochemistry processes that determine the percentage of total mercury that exists as the methylated
form. In addition, it has been repeatedly shown that mercury in fish accumulates throughout the
lifetime of the individual (Scott and Armstrong, 1972; MacCrimmon et al., 1983; Wren et al., 1983;
Mathers and Johansen, 1985; Skurdal et al., 1985, Wren and MacCrimmon, 1986; Sorenson et al.,
1990; Jackson, 1991; Gutenmann et al., 1992; Glass et al., 1993, Suchanek et al., 1993; Lange et al.,
1993). Reported BAF values for a given species may therefore vary as a function of the ages of the
animals examined. As a result, some researchers have suggested that comparisons between lakes
should be made using "standardized" fish values (that is, a value for a hypothetical 1 kg northern
pike), typically derived by linear regression of residue data collected from individuals of varying size
and/or age (Wren and MacCrimmon, 1986; Sorenson et al., 1990; Meili et al., 1991).
4.1.6 Conclusions
BAFs derived from adequate data collected at the site of concern should be used in lieu of the
estimated values presented in this Report The criteria for defining the adequacy of data are discussed
hi the Data Quality Objectives section of Appendix A. When such values are not available, the use of
the geometric mean values from the BAF3 and BAF4 output distributions generated from the field-
derived BAF3 and PPF4 distributions is the recommended approach. Use of the geometric mean,
rather than the arithmetic mean, is a consequence of the assumption that BAFs are distributed in nature
as the logarithm of the observed value. The recommended approach is more direct and less variable
than the GLWQI method and involves fewer assumptions. Direct application of the field-derived
BAF4 distribution is not recommended because of the current data limitations. The recommendation
as to the use of the (geometric) mean value of these distributions is based on the inability to
distinguish among various sources of uncertainty and variability in the output distributions, with
consequent problems of interpretation of specific percentiles. Because the exposure concern is for
repeated ingestion of contaminated fish, the mean, rather than the median, is the appropriate value.
The median is only useful if the concern was the random selection of a single fish. In this case, the
mean and median values are virtually identical for both distributions because of the choice of the form
of the input distributions.
Reducing the uncertainty in the BAFs generated by these methods will require the collection of
more data representative of the critical factors underlying the observed variability, and the inclusion of
additional terms to explicitly model those factors. For example, the inclusion of an age/size regression
term should account for a substantial portion of the variability in both BAF3 and PPF4. Of longer-
term significance is the modeling of factors representing the methylmercury-generating processes in the
water column, itself, such as those in the Mercury Cycling Model (Hudson et al., 1994).
June 1996 4-6 SAB REVIEW DRAFT
-------
4.2 Calculation of a Criterion Value for Protection of Piscivorous Wildlife
4.2.1 Procedure Used to Develop Criterion Values for Wildlife in the Water Quality Guidance for the
Great Lakes System
The WC for mercury is defined as the concentration of total mercury in surface water that, if not
exceeded, protects both avian and mammalian wildlife that use the water as a drinking or foraging
source. Thus, the WC is the highest aqueous concentration of mercury that causes no significant
reduction in growth, reproduction, viability or usefulness (in a commercial or recreational sense) of a
population of animals exposed over multiple generations. For the purpose of this analysis, the term
"aqueous concentration" refers to the total concentration of all mercury species in filtered water,
including both freely dissolved forms and mercury that is associated with dissolved organic material.
The equation used in this analysis to calculate a WC for mercury is identical to that described in
the Proposed Guidance (U.S. EPA, 1993c) and implemented in the final Water Quality Guidance for
the Great Lakes System (U.S. EPA, 1995b):
(TD X fl/OPD x Wt
A
WA -i- l(FDJ(FA x BAFJ + (FDJ(FA x BAFJ]
where,
WC = wildlife criterion value (pg/L; after converting from ug/L)
WtA = average species weight (g)
WA = average daily volume of water consumed (L/d)
FA = average daily amount of food consumed (g/d)
FD3 = fraction of the diet derived from trophic level 3
FD4 = fraction of the diet derived from trophic level 4
BAF3 = aquatic life bioaccumulation factor for trophic level 3 (L/g; methylmercury
concentration in fish/total mercury in water)
BAF4 = aquatic life bioaccumulation factor for trophic level 4 (L/g; methylmercury
concentration in fish/total mercury in water)
TD = tested dose (ug/g bw/d)
UF = uncertainty factor
In the equation used in this Report the term F (defined in the GLWQI as the food ingestion rate
of prey for a tropic level) is broken into the terms FA and FD above. The UF considers uncertainty in
three areas described below.
A similar equation was first used by the State of Wisconsin to set Wild and Domestic Animal
Criteria (State of Wisconsin, 1989). The entire approach, including both the equation and data
requirements for its parameterization, was later modified by U.S. EPA for incorporation into the
Proposed Guidance (U.S. EPA, 1993c) and Final Guidance (U.S. EPA, 1995b). The method, in its
current form, was reviewed in 1992 at a workshop entitled the National Wildlife Criteria
June 1996 4-7 SAB REVIEW DRAFT
-------
Methodologies Meeting, sponsored by U.S. EPA (U.S. EPA, 1994). Subsequently, it was used to
develop interim Tier IWC for four compounds (PCBs, DDT, dieldrin, and mercury) in the Great
Lakes Basin (U.S. EPA, 1993b). These criteria have received public comment The method has been
reviewed by the Science Advisory Board (SAB) on two occasions, most recently hi June of 1994.
Detailed descriptions of the method, including comparisons with other proposed methods for setting
wildlife criterion values, are presented in U.S. EPA (1993c, 1994).
An examination of this equation reveals both a hazard and an exposure component. The
GLWQI equation includes a term TD for "tested dose". In this Report, data were reviewed to
ascertain an appropriate NOAEL, which was used for the TD. In the absence of a NOAEL, a LOAEL
was used with the addition of an appropriate factor (UF^ to indicate uncertainty around the toxic
threshold. An uncertainty factor (UFA) may be used to provide a margin of safety when applying data
from a species other than the species of concern. A third uncertainty factor (UFS) may be used to
extrapolate from subchronic to chronic exposures. Additional adjustments may be warranted by
toxicokinetic or toxicodynamic considerations. Collectively, the application of the UF to the TD
results hi the estimation of a "reference dose" for subsequent calculation of WC.
The WC for mercury is expressed as the total mercury concentration in filtered water. It is
recognized that methylmercuty is the form of mercury that bioaccumulates in fish. Few laboratories,
however, currently possess the analytical capability to speciate mercury in water from natural sources
(presently, the different forms of methylmercury tend to be operationally defined; e.g., freely
dissolved, weakly bound and strongly bound). In the future, as analytical capabilities improve and a
basic understanding of methylmercury formation and behavior is acquired, it may be advisable to
calculate WC values based on methylmercury concentrations.
A WC for mercury was calculated hi the Proposed Guidance using fixed values for all
parameters hi the equation. Species-specific WC values (WCj) were calculated for each of the wildlife
species of concern (eagle, herring gull, kingfisher, mink, otter). Intermediate WC values (WCj) were
then obtained for avian and mammalian wildlife by calculating the geometric mean of values for
contributing species. The final WC (WCf) was set equal to the lowest of the two resulting
intermediate values and, for mercury, was driven by the calculations for avian species.
The WCf for mercury derived in the Proposed Guidance is 1300 pg/L. A comparison of the
GLWQI criteria for birds and mammals with those derived in this Report is found in Section 4.2.9.
For the present analysis, it was decided to consider some of the same wildlife species considered
hi the Proposed Guidance. Herring gulls, which are indigenous to the Great Lakes region, were not
evaluated hi this Report The avian wildlife for which WC values were calculated are the bald eagle
(Haliaeetus leucocephalus), osprey (Pandion haliaetus) and belted kingfisher (Ceryle alcyori). The
mammalian wildlife for which WC are calculated are the mink (Mustela visori) and river otter (Lutra
canadensis). Each of these species was originally selected after consideration of the following:
(1) then* exposure to bioaccumulative contaminants; (2) relevance to Great Lakes ecosystems;
(3) availability of information with which to calculate criterion values; and (4) evidence for
accumulation and/or adverse effects.
It was recognized that several other wildlife species would satisfy most or all of the selection
criteria presented in the GLWQI. Notable examples include the herring gull (Larus argentatus),
Forster's tern (Sterna forsteri), double-breasted cormorant (Phalacrocorax auritus), wood stork
(Mycteria americana), raccoon (Procyon lotor), snapping turtle (Chelydra serpentina), and American
June 1996 4-8 SAB REVIEW DRAFT
-------
alligator (Alligator mississippiensis). Exposure factors for a large number of wildlife species are
available in a recently published handbook (U.S. EPA, 1993a). A critical evaluation of these data as
they pertain to the development of WC is also available (U.S. EPA, 1995a). AUometric equations may
also be used to calculate both feeding and drinking requirements (see for example, Calder and Braun,
1983; Nagy, 1987). In time, the inclusion of other species, including both amphibians and reptiles,
may be appropriate, particularly if an effort is made to calculate WCf on a regional basis, or if the
species used in the present analysis are not representative of the ecosystem of concern. The present
analysis is intended, however, to be national in scope. Each of the species selected in the Proposed
Guidance is distributed over large portions of the country (see species distributions, Section 3.2), and
in these locations each species is closely tied to water resources via aquatic food chains.
4.2.2 Exposure Parameters
Exposure parameters for the present analysis are shown in Table 4-2. The scientific basis for
these parameters is reviewed elsewhere (U.S. EPA 1993a, 1995a). For this analysis, it was assumed
that prey not attributed to trophic levels 3 and 4 were derived from non-aquatic origins and do not
contain mercury. Were these prey to contain mercury, WC values calculated for the relevant species
would decrease. BAFs for trophic levels 3 and 4 were assigned the values recommended in Section
4.1.6.
Table 4-2
Exposure Parameters for Mink, Otter, Kingfisher, Osprey, and Eagle
Species
Mink
Otter
Kingfisher
Osprey
Eagle
BodyWt
(WtA)
kg
0.80
7.40
0.15
1.50
4.60
Ingestion
Rate
(FA>
kg/d
0.178
1.220
0.075
0.300
0.500
Drinking Rate
L/d
0.081
0.600
0.017
0.077
0.160
Trophic Level of
Wildlife Food
Source
3
3,4
3
3
3,4
% Diet at
Each
Trophic
Level
90
80,20
100
100
74,18
4.2.3 Health Endpoints for Avian Wildlife
Most studies of chronic exposure to birds involve feeding grain contaminated with a mercurial
compound applied to the feed grain. Exposure of birds to mercury commonly occurs when they feed
on seed grain that has been treated with mercurials as a preservative (Eisler, 1987). Fimreite (1970)
identified a LOAEL of 1.1 ug/g/d for growth inhibition in leghorn cockerel chicks (Callus) based
upon 6 ug/g methylmercury dicyandiamide in the feed. Fimreite (1971) also identified a LOAEL for
reproductive effects (reduced survival, reduced egg production, defective shells) of 0.18 ng/g/d in ring-
necked pheasant (Phasianus colchicus) fed seed treated with methylmercury dicyandiamide. Scott
(1977) identified a LOAEL for reproductive effects (reduced fertility, reduced egg number, reduced
survival, defective shells) of 4.9 ng/g/d in domestic chickens.
June 1996
4-9
SAB REVIEW DRAFT
-------
The most comprehensive studies of the effect of mercury on birds were conducted by Heinz and
co-workers. Heinz (1974, 1975, 1976a,b, 1979) assessed the effects of dietary methylmercury
dicyandiamide (0, 0.5 and 3.0 ppm as elemental mercury) over three generations of mallard ducks.
Treatment groups were housed in separate wire pens and fed dry duck mash treated with
methylmercury dicyandiamide. In the first generation, treatment began in adult ducks. Subsequent
generations received treatment beginning at nine days of age.
Initially, Heinz (1974) identified a NOAEL of 0.5 ppm based upon reproductive effects in a 21
week study. In a later study, reproduction in first and second generation ducks was evaluated (Heinz,
1976a,b), and the NOAEL for the first generation was again determined to be 0.5 ppm. The second
generation, however, suffered adverse reproductive effects including eggs laid outside the nest box
(p<0.05), reduced number of ducklings surviving to one week of age (p<0.05), and reduced growth of
ducklings (rxO.05) at the 0.5 ppm dose. Consequently, the LOAEL for reproductive effects for the
second generation was 0.5 ppm with no NOAEL identified. A third generation of mallards also
demonstrated adverse reproductive effects at 0.5 ppm mercury in the diet. Effects observed included
reduced number of sound eggs laid per day (p<0.01) and thinner egg shells (p<0.05).
Heinz (1975, 1979) also examined behavioral effects of mercury exposure in the approach
response of chicks to maternal calls and avoidance of frightening stimuli. In third generation
ducklings there was a reduction in response rate and speed of response to maternal calls (p<0.01).
When data were pooled from all studies and subject to analysis of variance (ANOVA) with multiple
comparisons, alterations of behavior were observed in the lowest dose groups in all generations. These
alterations included reduction in the number of ducklings which approached maternal calls (p<0.01)
and an increase in the distance traveled to avoid a threatening stimulus (p<0.05). In summary, no
NOAEL could be determined for behavioral effects, and the NOAEL for reproductive effects could
only be demonstrated for the first generation.
For the determination of an appropriate LOAEL, it was concluded that effects observed in
second and third generation ducks at 0.5 ppm should not be discounted. It seems likely that the
effects observed in the second and third generations were a result of the earlier onset of dosing (adult
onset versus onset as ducklings). For this reason, 0.5 ppm was selected as a LOAEL for mallard
ducks. Assuming a feeding rate of 128 mg/g/d for adult mallards, the LOAEL for reproduction and
behavior is 0.064 ug Hg/g/d.
4.2.4 Health Endpoints for Mammalian Wildlife
River otters (Lutra canadensis) fed 2 ppm methylmercury (0.09 ug/g) for six months suffered
from anorexia and ataxia (O'Connor and Nielson, 1981). In mink, 27 ppm of dietary phenylmercuric
chloride caused lethality in 40% of the males and 31% of the females within six weeks of exposure
(Borst and Lieshout, 1977).
Wobeser et al. (1976a,b) studied the effects of dietary consumption of methylmercury on ranch
mink. There were two parts to this study, which together formed the basis of Wobeser's dissertation
research (Wobeser, 1973). In the first part (Wobeser et al., 1976a), 25 adult female mink and their
litters were divided into three groups: Group I contained five females and 19 kits (control); Group II
contained 10 females and 34 kits (50% fish diet); and Group III contained 10 females and 29 kits
(75% fish diet). The ration was prepared using mercury-contaminated freshwater drum from Lake
Winnipeg, Manitoba; mercury in fish tissue was assumed for the purposes of the present analysis to
consist primarily of methylmercury. The fish was supplied in a ground, frozen form and was then
June 1996 4-10 SAB REVIEW DRAFT
-------
mixed with cereal and uncontaminated chow to a desired composition of 50 or 75 kg fish/100 kg of
food. All mink were fed once daily in slight excess of consumption. The three exposure groups were
observed for 145 days. Assuming a food consumption rate of 0.16 g/g/d (appropriate to captive
animals; Bleavins and Aulerich, 1981) and an average weight of 0.8 kg for the mink, these treatments
corresponded to dosing levels of approximately 35 and 55 ug/kg bw/d. One female and 3-6 kits were
euthanized every 15 (treatment) or 30 (control) days. Complete necropsies were then performed. No
clinical signs of disease were observed in any of the mink within the experimental period, and no
mortality or growth impairment occurred which could be attributed to the feeding of mercury-
contaminated fish.
In a second experiment (Wobeser et al., 1976b), 30 adult female mink were assigned to one of
six groups of five animals each. The animals were fed chow spiked with methylmercuric chloride at
0.0 (control), 1.1, 1.8, 4.8, 8.3, or 15.0 ug/g (by analysis), corresponding to dosing levels of 180, 290,
770, 1330, and 2400 ug/kg bw/d. Two mink from each group were allowed to die of intoxication or
were euthanized after 93 days (the end of the experiment). Animals were necropsied and the tissues
analyzed for mercury content. All animals in the control group remained clinically normal, and the
only clinical sign in the 1.1 ug/g dose group was a slight tendency for two of the animals to move
more slowly than the others during the last few days of the experiment. Anorexia, posterior ataxia and
lateral recumbency were observed in the other four dose groups. Death occurred within 26-36 days at
4.8 ug/g, and within 19-26 days at 8.3 ug/g. Histopathological abnormalities were seen at 1.1 ug/g,
including pale, yellow livers, lesions hi the central nervous system, and axonal degeneration.
Based upon a review of the Wobeser studies (Wobeser, 1973; Wobeser et al., 1976a,b) it can be
concluded that the LOAEL for subchronic exposure of mink to methylmercury is 180 ug/kg bw/d (1.1
ug/g dose group), using nerve tissue lesions as an effects endpoint. The NOAEL derived from these
studies is 55 ug/kg bw/d. Importantly, it was Wobeser's opinion that had the studies been carried out
for a longer duration, nervous tissue damage observed hi the 1.1 ug/g dose group would have become
manifested as unpaired motor function.
Charbonneau et al. (1974) fed random-bred domestic cats (Felis domesticus) 3, 8.4, 20, 46, 74 or
176 ug/kg/d of mercury, either as methylmercuric chloride in food or as memylmercury-contajninated
fish, 7 d/week for 2 years. Clinical examinations of the animals were conducted periodically.
Neurological examinations, using a modification of the method of McGrath (1960) were conducted
prior to the test, monthly throughout the test and more frequently as clinical signs of methylmercury
toxicosis became apparent. Neurological impairment, including hindrance of the hopping reaction and
hypalgesia, was observed in animals exposed to 46, 74, or 176 ug/kg/d, regardless of whether casts
were fed contaminated fish or spiked food. No treatment-related effects were observed in three lower
dosage groups. Overt signs of toxicity, including ataxia, loss of balance and motor incoordination,
were observed in animals fed 74 or 176 ug/kg/d These findings suggest that 20 ug/kg/d is the
NOAEL and 46 ug/kg/d is the LOAEL for chronic dietary exposure to methylmercury in domestic
cats. Charbonneau et al. (1974) also concluded that there was no difference in the toxicity or
bioavailability between naturally contaminated fish and fish spiked with methylmercuric chloride.
4.2.5 Summary of Health Endooints for Avian and Mammalian Wildlife
The avian chronic TD value was derived from studies by Heinz (1975, 1976a,b, 1979) in which
three generations of mallard ducks (Anas platyrhynchos) were dosed with methylmercury
dicyandiamide (0, 0.5 and 3.0 ppm). The lowest dose, 0.5 ppm (64 ug/kg bw/d), resulted in adverse
June 1996 4-11 SAB REVIEW DRAFT
-------
effects on reproduction and behavior and was designated as a chronic LOAEL. As no NOAEL was
reported, a UFL of 3 was used according to methodology described in U.S. EPA (1995b).
The mammalian chronic NOAEL was derived from studies of subchronic exposure by Wobeser
(1973, 1976a,b) in which mink were dosed with mercury in the form of mercury-contaminated fish
(0.22 and 0.33 ppm, naturally incorporated into fish; 1.1, 1.8, 4.8, 8.3 and 15.0 ppm spiked into the
diet). Effects observed included histopamologic lesions in nerve tissue at 1.1 ppm and higher doses.
Anorexia, ataxia and death occurred at 1.8 ppm and higher doses. The dose of 0.33 ppm (55 jag/kg
bw/d) was selected as the NOAEL for subchronic exposure. As this was a less than lifetime study, a
UFS of 10 was applied to the TD or NOAEL. The subchronic NOAEL/UFS is 5.5 ug/kg bw/d, which
is approximately one-fourth the chronic NOAEL (20 ug/kg/d) estimated from long-term feeding
studies with domestic cats (Charbonneau et al., 1974).
Based on the information above, the TDs used for calculation of a WC for mercury were these:
For avian wildlife - A LOAEL of 64 ug/kg bw/d, with a UFL of 3; and
For mammalian wildlife - A NOAEL of 55 ug/kg bw/d, with a UFS of 10.
4.2.6 Calculation of Wildlife Criterion Values
WC values were calculated for each of the wildlife species of concern using exposure parameters
values recommended in previous sections. UFAs were employed as recommended in the GLWQI to
extrapolate from test species to the species of interest. Because the mammalian TD (NOAEL) was
derived from studies with mink, the UFA for species extrapolation of the mink WC was set equal to
1.0. Otter were considered sufficiently similar to mink so that a UFA of 1 was also considered
appropriate. A UFA of 3 was used for extrapolation of mallard data to the kingfisher, eagle, and
osprey. Calculations of WC values for each of the selected species follow.
For the mink:
(TD x [ll(UFA x UFS x UFj)}) x WtA
wc _ (0.055 mglkg/d x [lf(l xlOx 1)] x 0.8 kg
s 0.081 Ud + [(0.9) (0.178 kgfd x 66,200)]
WCS
June 1996 4-12 SAB REVIEW DRAFT
-------
For the otter:
WC,
we.
(TD x [1/(UFA x UFS x UFJ\ x WtA
WA + [(0.8) (FA x BAFj + (0.2) (FA x BAFJ]
(0.055 mglkgjd x [1/(1 x 10 x 1)] x 7 A kg
0.60 Ljd + [(0.8) (\32kgld x 66,200) + (0.2) (122ks/d x 335,000)]
WCS - 278
For the kingfisher:
(TD x [lj(UFA x UFS x UFJ] x WtA
WCS
WA + [(1.0)
WC = (0.064 mg/kgfd x [l/(3 x I x 3)] x 0.15 kg
5 0.017 + [(1.0) (0.075 x 66,200)]
WCS - 193 pgjL
For the osprey:
WC -
S~
x UF*>] x
[(1.0) (FA x BAFJ]
WC - (0.064 mg/kg/d x [l/(3 x I x 3)] jc 1.5 kg
5 0.077 Lfd + [ (1.0) (0.3 kgld x 66,200) ]
WCS = 483 pg/L
June 1996
4-13
SAB REVIEW DRAFT
-------
For the bald eagle:
WC,
(TD x [ll(UFA x UFS x UFJ] x WtA
[(0.74) (FA x BAFJ + (0.18) (FA x BAF4]
(0.064 mgfkgld x [l/(3 x 1 x 3)] x 4.6 kg
0.16 Lfd + [(0.74) (OSkgfd x 66,200) + (0.18) (OSkgld x 335,000)]
WCS = 538 pgIL
The mean of the two WCS values calculated for mammals is 346 pg/L. The mean of the three
avian values is 405 pg/L. The lowest of these is the WCj calculated for avian species.
Therefore, the WCf for mercury is 346 pg/L.
4.2.7 Calculation of Mercury Residues in Fish Corresponding to the Wildlife Criterion Value
The WC for mercury can be used to calculate corresponding mercury residues in fish through
the use of appropriate BAFs. Using the BAFs presented in Section 4.1, a WC of 346 pg/L
corresponds to methylmercury concentrations hi fish of 0.023 ug/g and 0.116 ug/g for trophic levels 3
and 4, respectively.
4.2.8 Calculation of a Wildlife Criterion for the Florida Panther
Estimates of the NOAEL and LOAEL in domestic cats were not intended for use in the
derivation of a WC for Florida panthers, but were presented instead to provide a comparison with
other mammals. The chronic NOAEL for cats (20 ug/kg/d) is close to that for mammals generally
(5.5 ug/kg/d; that is, the subchronic NOAEL of 55 ug/kg/d divided by a UFS of 10). Cats do not,
therefore, appear to be uniquely sensitive or insensitive to the toxic effects of mercury.
Derivation of a WC to protect the panther is complicated by possibility that prey items (e.g.,
the raccoon) accumulate mercury to an even greater extent than the fish represented by trophic level 4.
Other prey (e.g., deer) probably contain relatively lower levels of mercury. Calculation of a WC
protective of the panther, therefore, requires collection of additional information on the diet of this
species and mercury residues contained therein. These residues would then have to be related back to
corresponding levels hi water through the use of PPFs (e.g., raccoon/fish or other aquatic biota) and
BAFs (aquatic biota/water). Existing data are insufficient to support such an analysis but could be
collected and developed for this purpose.
4.2.9 Comparison of GLWQI Criteria with WC Derived in this Report
The evaluation of data and calculation of WC in this Report was done in accordance with the
methods and assessments published in the draft GLWQI (U.S. EPA 1993a). Availability of additional
June 1996
4-14
SAB REVIEW DRAFT
-------
data and differences in interpretation of those data led to differences in the calculated values of the
WC in this Report and those published in the final GLWQI (U.S. EPA 1995b). Both evaluations
employed the same methodology as described in Section 4.2.1. Both used the same studies as the
basis for WC calculation: for birds, the three generation reproduction study in mallards (Heinz, 1974,
1975, 1976a,b, 1979); and for mammals the subchronic dietary studies in mink (Wobeser et al.,
1976a,b). In addition to these studies, the authors of this Report were able to obtain Wobeser's
dissertation (Wobeser, 1973); this provided some additional information, which was augmented by
discussions with the author.
•
Table 4-3 presents a comparison between the WC calculated in the GLWQI (U.S.EPA, 1995b)
and this Report. All of the WC calculated in this Report are lower (more conservative) than those
published in the GLWQI. All species-specific WC, however, differ less than an order of magnitude
from one another. Range in differences is from nearly four-fold lower for the WC.
Table 4-3
Species-specific Wildlife Criteria Calculated in the Great Lakes Water Quality Initiative
(GLWQI)* and in the Mercury Study Report to Congress
Species
Mink
Otter
Kingfisher
Osprey
Eagle
Wildlife Criterion
(pg/L)
GLWQI
2880
1930
1040
Not done
1920
Mercury Study Report to Congress
415
278
193
483
538
a U.S.EPA, 1995b
In the evaluation of effects in birds, both the GLWQI and this Report identified a LOAEL for
reproductive effects in the second generation of mallards exposed to 0.5 ppm mercury in diet (Heinz
1976b, 1979). In the GLWQI this LOAEL was adjusted to 0.078 mg/kg bw/d by applying an average
food ingestion rate for treated mallards of 0.156 kg/kg-d. This Report converted the LOAEL to 0.064
mg/kg bw/d by application of an assumed feeding rate for adult mallards of 0.128 kg/kg-d. In
calculating the wildlife reference dose, the GLWQI used a UFA of 3 and a UFL of 2. This Report
used a UFA of 3 and a UFL of 3 (see Section 4.5.2 for a discussion of UFL).
In the effects assessment for piscivorous mammals both the GLWQI and this Report used data
on mink administered mercury in the diet. The GLWQI identified a NOAEL of 1.1 ppm. At this
dietary exposure there were changes in the liver, lesions in the central nervous system and axonal
degeneration; moreover, two of the animals in this treatment group were observed at the end of
treatment to move slowly by comparison to other mink. The study authors reported their opinion that
mink treated at 1.1 ppm in the diet for longer than the study would be expected to show clinical signs
of nervous system damage. Animals treated at the next dose, 1.8 ppm, were observed with anorexia,
June 1996
4-15
SAB REVIEW DRAFT
-------
ataxia and increased mortality. Based on these considerations, this Report considered 1.1 ppm to be
LOAEL, and as described in Section 4.2.4, used data from the first part of the study to identify a
NOAEL of 0.33 ppm. This Report used data from Wobeser (1973) to establish the weights of female
mink and kits used in this part of the study; this resulted in slight differences in conversion of dose in
ppm diet to ug/kg bw/d.
In assessment of exposure to birds through consumption of prey, the GLWQI made
assumptions that were appropriate the Great Lakes region. In particular the GLWQI assumed that
mercury contaminated herring' gulls constitutes 6% of the diet of bald eagles. As this Report is a
nationwide assessment, use of this region-specific assumption was not considered appropriate; eagles
were assumed to consume non-fish prey, with no mercury contamination, as 9% of the total diet. The
largest numerical difference hi exposure assessment between the GLWQI and this Report is in the use
of BAR The GLWQI used a BAF of 27,000 for trophic level 3 and a BAF of 140,000 for trophic
level 4. Derivation of the BAF3 of 66,200 and BAF4 of 335,000 used in this Report is described in
Section 4.1.
Thus, the differences between the wildlife criterion in the Guidance and in this Report are a
result of three factors. First, and with the greatest numerical impact, this Report uses more recent data
to derive BAFs. The Supplementary Information Document to the final Water Quality Guidance for
the Great Lakes System noted .that the preliminary report containing these data was available but was
not used because it had not been completed at the time the final guidance was published (U.S. EPA
1995b, p. 144). Second, the Guidance appropriately used some region-specific assumptions that were
not used in this nationwide assessment (e.g., consumption of herring gulls by eagles). Finally,
different endpoints were used because the purposes of the assessments were different In the Guidance
wildlife methodology, a risk-management decision was made to base the wildlife criterion on
endpoints likely to influence wildlife populations (e.g., reproductive/developmental, mortality, growth).
In this Report, a more sensitive endpoint was selected with the goal of assessing the full range of
effects of mercury. The difference in the results reflects the amount of discretion allowed under the
Agency Risk Assessment Guidelines,
43 Uncertainty Analysis
A formal analysis of uncertainty around the WC estimate was not attempted. Such an analysis
would require specification of numeric distributions for each of the parameters in the equation. While
theoretically possible, this approach is of questionable value'since the overall analysis is intended to be
protective of that subset of each species which feeds extensively at the top of aquatic food chains.
Thus, incorporation of data reflecting the range of dietary items upon which the bald eagle feeds
would tend to generate an extremely broad range of WC values for this species. In addition, data for
several of the parameters in the equation, in particular the NOAEL and UF estimates, are presently
sufficient only to generate point estimates.
A restricted uncertainty analysis involving only incorporation of numeric distributions for each
of the BAF estimates could be accomplished using existing data, but would probably not be useful.
As noted previously, BAF distributions generated by Monte Carlo analysis of field data are thought to
reflect real, naturally-derived variation in mercury bioaccumulation and biomagnification. Despite the
relative abundance of such data, BAFs expressed on a total mercury basis remain difficult to interpret.
Because methylmercury is the form of mercury accumulating in fish, WC distributions based on the
distribution of methylmercury BAFs are more likely to yield information of value to risk assessors.
June 1996 4-16 SAB REVIEW DRAFT
-------
4.4 Sensitivity Analysis
In a sensitivity analysis, an attempt is made to characterize the extent to which a calculated
value (e.g., a WC value) changes with changes in the parameters upon which its calculation depends.
Examination of the equation for calculation of WC values suggests that a proportional relationship
exists between the WC and the NOAEL, UF or WtA. The relationships between the WC and
parameters that appear in the denominator are not as apparent and must be explored by varying these
parameters one-by-one in systematic fashion. The analysis is also complicated by the variable
relationship that exists between FD3 and FD4. In the otter and eagle, FD3 and FD4 tend to be
reciprocal (although in the eagle these values do not add up to 1.0). In the mink, however, FD3 is
assigned a value of less than 1.0 and the remainder of the diet is assumed to consist of prey that are
not aquatic in origin and are not contaminated with mercury.
Nevertheless, general conclusions can be reached regarding the sensitivity of WC estimates to
changes in these parameters. These can be described as follows:
• A decrease in any parameter that appears in the denominator will have a larger effect
on WC than an equivalent percentage-wise increase.
• When BAF3 appears alone in the denominator, a percentage-wise increase in BAF3 or
FD3 will cause a less than proportional decrease in the WC; conversely a decrease in
BAF3 or FD3 will cause a greater than proportion increase the WC.
• When both BAF3 and BAF4 appear in the denominator, an equivalent percentage-wise
change in BAF4 (and by extension PPF^ has a greater impact on the WC than a
change in BAF3, but in either case the effect is less than proportional.
• If BAF3 and BAF4 are both allowed to change (holding PPF4 constant), a percentage-
wise increase in BAF3 (and by extension BAF^) will have a less than proportional
effect on WC, while a decrease in BAF3 will have a greater than proportional impact.
• Under all circumstances, a percentage-wise increase in FA will cause a less than
proportional decrease in WC, while a decrease in FA will cause a greater than
proportional increase in WC.
• Owing to its small contribution to the analysis as a whole, large changes in WA have a
very small impact on WC.
With the exception of FA, it is not possible to conclude that for all species the WC is most
sensitive to one or the other of the parameters in the denominator of the equation. For species that
feed at one trophic level, all parameters other than FA have the potential to change WC in a
proportional or greater than proportional manner. For species that feed at two trophic levels, the BAF
at the lower trophic level becomes relatively less important, but it may still have a large impact on
WC if the percentage of the diet represented by this lower trophic level is large (e.g., in the mink).
4.5 Uncertainties Associated with the GLWQI Methodology
Efforts to develop WC for the protection of piscivorous wildlife are relatively recent in origin,
and the methods employed for this-purpose continue to undergo modification and refinement. Owing
June 1996 4-17 SAB REVIEW DRAFT
-------
to the complexity of natural systems, uncertainties associated with the development of WC are to be
expected. Additional uncertainties derive from the relative scarcity of wildlife toxicity information and
the necessity of extrapolating individual-based effects to higher levels of biological organization (e.g.,
populations).
Uncertainties associated with the GLWQI methodology have been reviewed elsewhere (U.S.
EPA, 1994). It is not our intent to repeat this information, but instead to focus on those areas which
are especially pertinent to the development of a WC for mercury/ These are listed below in no
particular order.
4.5.1 Limitations of the Toxicitv Database
Substantial uncertainties underlie most of the toxicity data for mercury in wildlife.
Comparison of NOAELs and LOAELs between species requires adoption of unproven assump-tions
about the uptake, distribution, elimination, and toxic effects of mercury. Conclusions based upon
extrapolation from one species to another are, therefore, tenuous and result in the use of uncertainty
factors for species extrapolation. Additional uncertainties derive from the necessity of extrapolating
from LOAELs to NOAELs, and from subchronic endpoints to chronic endpoints. In some instances
there may also be a need to account for the possibility that test results do not adequately protect the
most sensitive individuals. This may be particularly germane to the case of the Florida panther, when
there is concern for individual animals).
Existing epidemiological data are complicated by the possibility that "naturally incorporated"
mercury is accompanied by other contaminants which are exerting some or all of the observed effect.
Ideally, it is desirable to compare the effects of mercury that has been incorporated naturally with
effects that are due to mercury that has been spiked into a prepared diet By spiking mercury into the
diet, the researcher can better control the dose to the animal. The bioavailability of mercury in such a
formulation may be very different from that which exists naturally. Charbonneau et al. (1976) has
demonstrated that the bioavailability and toxicity of methylmercury to cats is equivalent whether given
hi contaminated fish or spiked in the diet.
Despite a lack of toxicity information and problems concerning its interpretation, estimated
NOAELs for piscivorous birds and mammals are very similar (55 (ig/kg/d for mammals versus 21
ug/kg/d for birds). Moreover, the existence of toxicity information for the mink eliminates the need to
incorporate additional uncertainty factors into the analysis. Unfortunately, similar data for piscivorous
birds do not exist.
EPA cannot test all wildlife species of interest The use of uncertainty factors for species
extrapolation is likely, therefore, to continue. Existing information can be used, however, to suggest
which species should be singled out for testing. Information of this type is reviewed in this document
in several locations and includes species distribution, natural history considerations and exposure
factors. Properly applied, species sensitivity factors probably represent a relatively small source of
error in the calculation of WC values.
Finally, comparisons between wildlife and human NOAELs are complicated by differences in
the ability of a given study to reveal an adverse effect when it occurs. For wildlife, most of the
endpoints selected can be considered severely adverse or frank effects. Very few studies to date have
been designed to study subtle adverse effects or precursors to adverse effects in wildlife.
Developmental neurotoxicity endpoints are of particular interest because of their demonstrated
June 1996 4-18 SAB REVIEW DRAFT
-------
sensitivity in humans. The question arises, therefore: what would the LOAEL or NOAEL for a given
wildlife species have been had the researcher been looking for (or was able to detect) these more
subtle effects? One possible approach to this question is to examine the results of studies in which
both frank and more subtle effects were observed and determine the corresponding difference between
dosage levels.
Clearly, more research of this type is needed. The available data do, however, suggest that
uncertainty factors presently employed to derive chronic reference doses for birds and mammals are
unlikely to be greatly underprotective or overprotective, and are, therefore, reasonable.
4.5.2 LOAEL-to-NOAEL Uncertainty Factor UFL
In determining the WC for mercury exposure in wildlife, a NOAEL is preferred as the value to
be used as the term TD. The WC can be considered a wildlife Reference Dose (RfD) with an
adjustment for exposure assumptions. As is the case for human health RfDs, the wildlife RfD is an
attempt to estimate a threshold dose for adverse effects and then to determine a level below that
threshold dose. It is assumed that daily consumption of an amount of material below the threshold for
adverse effects should be without ill effect.
In cases where studies do not identify a NOAEL, the data are examined to identify a lowest-
observed-adverse-effect level (LOAEL) to be used in estimating the RfD. A UFL of 3 or 10 (based on
EPA Reference Dose methodology) is typically applied when a LOAEL is used in the absence of a
NOAEL.
In determining the RfD for human exposure to methylmercury, a large number of laboratory
animal studies on methylmercury toxicity were summarized as supporting data. Results from many of
those studies permitted estimation of both a LOAEL and a NOAEL. Those studies were examined in
an effort to determine the most appropriate UFL for wildlife exposure to mercury.
The studies examined are summarized in Volume IV of this Report. Nineteen studies were
selected as being the most relevant and appropriate for determining a UFL. Selection criteria included
the following:
• . methylmercury toxicity to nonhuman mammals;
• oral exposure (with preference given to dosing in food or drinking water); and
• chronic or subchronic exposure durations (with exceptions for reproductive and
developmental toxicity where such distinctions are less relevant).
Cancer and genotoxic endpoints were not included because tumors are not often reported in wildlife
toxicity studies. Endpoints included in the analysis (Table 4-4) included lethality, neurotoxicity, renal
toxicity, gastrointestinal toxicity, immunotoxicity, developmental toxicity and reproductive toxicity.
Data abstracted from the studies include the species and sex of the test subjects, toxicologic endpoint,
LOAEL, NOAEL and the ratio between them. The LOAEL:NOAEL ratios were not segregated by
endpoint because there were an insufficient number of studies at most endpoints to determine statistical
significance.
June 1996 4-19 SAB REVIEW DRAFT
-------
Table 4-4
Analysis of LOAEL-to-NOAEL Uncertainty Factor
Endpoinl
Species and Sex
LOAEL
(mg/kg/day)
NOAEL
(mg/kg/day)
RATIO
LOAEL:NOAEL
Study
Lethality
B6C3F1 Mouse M
0.69
Neurotoxicity
Rat (Wislar) M & F
Cat sex NS
Monkey (Macaco fasicularii) M & F
Monkey (Macaco artoides and M. nemestrina) M & F
Renal Toxicity
Mouse (ICR) M
F
Mouse (B6C3F1) M
F
0.25
0.046
0.03
0.5
0.72
0.62
0.14
0.6
0.60
1.15
0.05
0.020
0.02
0.4
0.15
0.11
0.03
0.13
5.0
2.3
1.5
1.25
4.8
5.6
4.7
4.6 »
Mitsumori et al., 1990
Munro et al., 1980
Charbonneau el al., 1976
Sato and Ikuta, 1975
Evans et al., 1977
Hirano el al., 1986
Mitsumori el al., 1990
Gastrointestinal Toxicity
Mouse (B6C3F1) M
Immunotoxicity
Rabbit (New Zealand White) M & F
Developmental Toxicity
Rat (Charles River) F
Rat (Wistar) F
Rat (Charles River) F
Rat (Wistar) offspring of both sexes
0.69
0.4
4.0
0.25
1.4
0.6
0.14
0.04
0.2
0.05
0.7
0.2
4.9
10.0
20.0
5.0
2.0
3.0
Mtaumori et al., 1990
Koller et al., 1977
Nolen et al., 1972
Khera and Tabacova, 1973
Fowler and Woods, 1977
Schreiner et al., 1986
June 1996
4-20
SAB REVIEW DRAFT
-------
Table 4-4 (continued)
Analysis of LOAEL to NOAEL Uncertainty Factor
Endpoint
Species and Sex
Reproductive Toxicity
Rat (Wistar) M
Mouse (ICR) M
Mouse (B6C3F1) M
Monkey (Macaco facicularis) M
Monkey (M. facicularis) F
LOAEL
(mg/kg/day)
0.5
0.72
0.68
0.065
0.06
NOAEL
(mg/kg/day)
0.1
0.15
0.14
0.047
0.04
RATIO
LOAEL-.NOAEL
5.0
4.8
4.9
1.4
1.5
Study
Khera, 1973
Hirano et »!., 1986
Milsumori et al., 1990
Mohamed et al.. 1987
Burbacher et a]., 1988
NS - Not stated.
June 1996
4-21
SAB REVIEW DRAFT
-------
Figure 4-1
LOAEL-to-NOAEL Ratio Distribution
r
10
8
7
5
4
3
2
6
1
•
9
2
s
1 1 1
123456 9 10 19 20
Ratio
June 1996
4-22
SAB REVIEW DRAFT
-------
The ratios of LOAEL-to-NOAELs for laboratory animal studies are plotted versus frequency in
Figure 4-1. These ratios can be thought of as the reduction in the LOAEL necessary to estimate the
corresponding NOAEL. Figure 4-1 illustrates that the majority of ratios lie between one and two
(n=6) and between four and five (n=9). Only one ratio of the 19 plotted was greater than 10. A ratio
of five indicates that the NOAEL observed following exposure to methylmercury is 5-fold less than
the corresponding LOAEL. These data imply that most ratios between LOAELs and their
corresponding NOAELs will be less than 10.
A similar analysis of animal toxicity data (Weil and McCollister, 1963) was provided by
Dourson and Stara (1983). None of the LOAEL-to-NOAEL ratios from studies of 52 chemical
substances exceeded 10. Only two of the 52 ratios exceeded five. The Dourson and Stara (1983)
analysis has been cited in support of the use of a variable UFL of as much as 10 in deriving reference
doses for humans. Dourson and Stara (1983) recommended the application of a relatively large UFL
when estimating a NOAEL from a LOAEL for a severe or frank toxicological effect. Conversely, a
low UF could be applied when the toxicological effect was considered to be relatively mild.
The analysis by Dourson and Stara (1983) and the analysis reported here support the UFL of
two selected for derivation of the avian wildlife criterion in the Great Lakes Water Quality Initiative
(U.S. EPA 1993b). The UFL of three was selected by the authors of this Report for use with the
avian LOAEL from the same data (Heinz, 1975, 1976a,b, 1979) as a reasonable compromise between
UF ratios of two and five. Given the substantial uncertainties in all the values used to calculate the
wildlife criteria for mercury exposure, neither two nor three can be considered to be the only correct
value.
The distribution of LOAEL:NOAEL ratios around two and five primarily reflect the dose
spacing selected for the study designs. Two-fold, 5-fold and 10-fold spacing are common in
experiments of this type. The most appropriate interpretation of the ratios reported here and by
Dourson and Stara (1983) is that the threshold for the toxicologic effects, defined by each study, lies
within the bounds of the experimentally derived LOAEL divided by an UF and that most of the effects
thresholds will be encompassed by using an UFL of 10 or less. It is also likely that the most
appropriate UFL will vary with the toxicological endpoint selected. For studies that identify only a
LOAEL, the principal assumption is that the next lower dose, had it been tested, would be a NOAEL.
This assumption is best applied to studies that identify a LOAEL for mild effects. LOAELs for severe
or frank effects (which are generally no used for human health risk assessment) require a high degree
of professional judgment in applying an UFL.
4.5.3 Validity of BCF/BAF Paradigm
A significant shortcoming of the WC for mercury calculated in the GLWQI is its reliance upon
BCF values determined in the laboratory. This methodology is based on a bioaccumulation paradigm
(steady-state BCF x FCM) that was developed for neutral hydrophobic organic compounds and that
may be inappropriate for application to mercury.
Field studies indicate that many, if not most fish, accumulate mercury throughout their lives,
often in a nearly linear fashion with age (see for example: Scott and Armstrong, 1972; MacCrimmon
et al., 1983; Wren et al., 1983; Mathers and Johansen, 1985; Skurdal et al., 1985; Wren and
MacCrimmon, 1986; Sorenson et al., 1990; Jackson, 1991; Gutenmann et al., 1992; Glass et al., 1993;
Suchanek, 1993; Lange et al., 1993). Moreover, most of the mercury accumulated by fish at trophic
levels 3 and 4 is thought to be taken up from dietary sources. Thus, particularly for long-lived
June 1996 4-23 SAB REVIEW DRAFT
-------
piscivorous fish, a relatively short (one year or less) waterborne exposure cannot duplicate the extent
of accumulation that takes place in nature. In addition, the relationship between a concentration of an •
applied mercury species in the laboratory and the concentrations of multiple species present in the
environment (some of which may not be bioavailable) is completely unknown.
The apparent progress to "steady-state" observed in several chronic laboratory studies (see
McKim et al., 1976) should not be misinterpreted as an actual steady-state condition, but instead
probably reflects growth dilution with rapidly growing fish. Such growth dilution will tend to depress
BCF values during the period of rapid growth, but as the growth rate slows continued accumulation of
mercury would result in an increase in whole-body concentration with age.
In the present analysis, BAF values for trophic levels 3 and 4 were estimated using field
residue data. Uncertainties remain, however, with respect to the naturally-derived variability that
exists around these estimates. A growing body of information suggests that much of this variability
can be attributed to site-specific factors that control the net rate of mercury methylation, but numerous
additional factors may influence the extent to which mercury accumulates and biomagnifies in aquatic
food webs.
4.5.4 Selection of Species of Concern
The species identified for the present analysis were selected because they were considered
likely to be exposed not because of their inherent sensitivity to mercury. Lacking toxicity information,
little guidance is available concerning which wildlife species are most sensitive to mercury. In
addition, there are problems associated with any comparison of laboratory and field data. For
example, laboratory data suggest that mercury residues in eggs exceeding 0.5 fig/g are associated with
impaired reproduction in mallard ducks (Heintz, 1974, 1976a,b, 1979) and ring-necked pheasant
(Funreite, 1971). In contrast, reproduction in herring gulls appears to be unaffected even when egg
residues exceed 10 ug/g (Vermeer et al., 1973). Taken alone, these data suggest that mallards and
pheasant are more sensitive to the toxic effects of mercury than are gulls. This may in fact be true;
however, such comparisons are complicated by the presence/absence of additional stressors such as
confinement, handling, and weather, differences between natural and prepared diets, and the interplay
between "inherited" (egg) residues and that which the chick consumes. Toxicity can be difficult to
observe in a field study, even when it is occurring due to any number of factors. Nest predation is
one such example: in 18 of 38 nests under study by Vermeer et al. (1973) hatching success could not
be evaluated for one reason or another. Moreover, it is possible that in gulls the most sensitive
endpoint for mercury toxicity is not reproduction but some other effect such as neurological
impairment.
Clearly, exposure and sensitivity are related. If, for example, a species was, on a delivered
dose basis, 10 times more sensitive than the eagle, but because of its dietary habits received less than
10% of the dose, it would not be expected to show adverse effects at water concentrations protective
of the eagle. Pharmacokinetic considerations may also be important. Thus, it has been suggested that
birds eliminate a substantial amount of mercury through incorporation into plumage. The frequency
and extent to which birds moult may, therefore, impact their apparent sensitivity in an environmental
setting. It has also been suggested that some birds and mammals demethylate, or otherwise eliminate
mercury by some route other than in hair or plumage (see Wren et al., 1986 for a discussion of these
data). Enhanced elimination would be particularly important if it represented an adaptive strategy for
piscivorous species. The need for toxicity information has already been noted. As such information
becomes available it may be necessary to revise the WC for mercury.
June 1996 4-24 SAB REVIEW DRAFT
-------
There is a need also to consider animals other than birds and mammals. In particular, there is
a need to characterize the exposure of carnivorous reptiles such as the alligator which is known to
consume considerable quantities of fish and which also feeds on animals (e.g., raccoon) which
themselves feed on aquatic biota and are known to accumulate mercury (Roelke et al., 1991).
4.5.5 Trophic Levels at Which Wildlife Feed
The dietary preferences of the wildlife species identified for this analysis are shown in Table
4-2. Justification for these assignments can be found in two recent U.S. EPA publications that were
developed for the purpose of supporting WC calculations (U.S. EPA 1993a, 1995a). It can be
expected, however, that representatives of the same species will be exposed to different levels of
mercury due to different feeding habits and/or differences in the availability of specific prey items.
For example, bald eagles living on the shores of the Great Lakes may consume significant numbers of
herring gulls (Kozie and Anderson, 1991). Since the gulls themselves are piscivores, feeding primarily
at trophic level 3, it has been argued that wheA an eagle consumes a gull it is feeding at trophic level
4 or higher; the gull/forage fish PPF is thought to be about 10, while the PPF for fish at trophic level
4 is believed to be approximately 5 (U.S. EPA, 1995a). Eagles living in other parts of the country, or
migrating into an area during a particular time of year, may consume relatively few fish, feeding
instead on carrion, including rabbits, squirrels, and dead domestic livestock such as pigs and chickens
(Harper et al., 1988). Other populations, however, are critically dependent upon the seasonal
availability of fish, particularly spawning salmonids.
The feeding habits of bald eagles are reviewed extensively elsewhere (U.S. EPA, 1993a,
1995a). The intent of this discussion is not to characterize the food preferences of the eagle, but
instead to demonstrate how difficult it is to characterize wildlife feeding habits on a nationwide, year-
around basis. For some species, such as the kingfisher and river otter, it can be reasonably assumed
that fish always comprise a high percentage of the diet For others, such as the eagle and mink,
considerable variations in diet are likely to exist Still others, such as the Florida panther, consume
prey (e.g., the raccoon) which, as a species, consume variable amounts of aquatic biota, but which in
south Florida are thought to represent a close link to the aquatic food chain.
Since mercury bioaccumulation is largely a problem associated with aquatic ecosystems, it is
reasonable to focus attention on populations of selected wildlife species whose feeding habits are tied
to these systems. Existing data permit a general treatment of mercury exposure and effects on such
populations. A more accurate characterization of the risk posed by mercury to a specific group of
animals occupying a given location will depend upon the collection of necessary supporting
information such as food habits, migratory behavior, breeding biology, and mercury residues in
preferred prey items.
4.5.6 Variability in BAFs at each Trophic Level
A concern related to the issue of feeding preference is the possibility that trophic levels
presently assigned to the wildlife species in this analysis overestimate the actual extent to which they
are exposed to mercury. This is because BAFs are developed to represent the average value for a
trophic level, when in fact piscivorous birds and mammals are more likely to target prey at the lower
end of the size (age) distribution. Thus, eagles are more likely to consume a 1 kg northern pike than a
10 kg individual, yet both are represented in the BAF for trophic level 4. Similarly, kingfishers are
probably limited to smaller representatives of trophic level 3 than would be true of an osprey. The
reason that these differences are important is that mercury tends to accumulate throughout the life of
June 1996 4-25 SAB REVIEW DRAFT
-------
an individual fish with the result that concentrations in an older individual at a given trophic level may
far exceed those in a younger individual.
The need in this study to apply BAF estimates on a nationwide basis precludes further
refinement. It may, however, be possible to explore this issue by using the Monte Carlo method to
analyze individual datasets. Specifically, it would be of interest to determine whether percentile
information from the Monte Carlo distributions can be related to fish of. known size. Eventually, it
may be possible to use this or another approach to refine BAF estimates for mercury.
4.5.7 Natural History Considerations
Natural exposures are likely to vary in both spatial and temporal domains. This is particularly
true of species that migrate, including the bald eagle, osprey and belted kingfisher. The necessity of
incorporating this type of information and the means by which this can be accomplished are open
questions. t
4.5.8 Individuals Versus Populations
The methods used to develop a WC for mercury are based on effects data from individual
organisms. The stated assessment endpoint for this analysis, however, is the health of wildlife
.populations. The relationship between individuals and populations is likely to vary with the species
and a large number of environmental factors (e.g., availability of food in a given breeding season).
For a given population, the loss of a significant number of individuals may have little effect,
particularly if environmental factors (like carrying capacity) limit population size. For other
populations, in particular those with low fecundity, loss of a relatively few individuals could have a
large impact. Clearly, there is a need to be able to extrapolate toxic effects on individuals to effects
on populations. Unfortunately, this type of analysis is complicated by numerous factors (such as
relationship of one population to other populations) and is essentially impossible to apply on a national
scale.
Finally, a focus on populations may not always be appropriate, particularly when endangered
species are involved. The same may also be true when various factors contribute to the possibility of
regional effects. For example, 95% of eagles nationwide might be protected by a WC for avian
species, but in a given region mortality could approach 100% if low pH of surface waters contribute to
higher than average accumulation of mercury in the aquatic food chain.
4.5.9 Species Versus Taxa
The WC developed for mercury in birds was calculated as the geometric mean of values for
three species. Similarly, the geometric mean of values for two species was used to represent all
mammals. This approach is reasonable if the WC calculated for each species within a taxa are similar,
but would fail to protect species for which the WC value is much lower than the others with which it
was averaged. In the latter case, averaging would effectively lead to protection to <100% of all
species.
In the present analysis, WC values calculated for eagles, osprey and kingfisher were within a
factor of three of one another. WC values for mink and otter agreed to within a factor of about two.
As additional data are gathered, there is a need to identify species which, by virtue of sensitivity
June 1996 4-26 SAB REVIEW DRAFT
-------
and/or exposure, are particularly vulnerable to mercury. Decisions could then be made concerning the
advisability of special measures to insure their protection.
4.5.10 Discussion of Uncertainties Associated with the GLWQI Methodology
The existing limited data suggest that BAF values represent the most important source of
uncertainty in present efforts to calculate water-based WC values, although a lack of toxicity
information and incomplete knowledge of what wildlife eat contribute substantially. Considerable
progress has been made in understanding and predicting how lakewater characteristics (e.g., pH,
temperature,dissolved organic carbon) affect methylation rates, and in time it may be possible to adjust
BAF predictions as needed to represent surface waters of concern. The prospect for continuing
uncertainty surrounding these estimates-argues, however, for adoption of a residue-based approach; that
is, the use of measured mercury residues in fish and wildlife to identify populations at risk.
It is important to-recognize that BAF values are calculated as the ratio of a tissue
concentration and a water concentration. Emphasis has been placed on problems associated with
obtaining the numerator in this equation. Considerable uncertainty, however, also exists with respect
to the denominator. In several instances it has been shown that with improved analytical methods,
mercury levels in a given water body tend to come "down", resulting in an increase in the apparent
BAF. This "decline" is usually not thought to be real but instead reflects improvements in sampling
technique and analytical methods.
It is also unclear which of the mercury species are bioaccumlative and should, therefore,
appear in the denominator. Presently, the denominator in most studies consists of total amount of
mercury in filtered water. It is more likely that there may be multiple "pools" of mercury, each of
which is bioavailable to varying degrees. In this regard it is important to realize that even in highly
polluted systems >99% of all methylmercury is complexed, either in biomass, or with dissolved
organic material, paniculate material and sediments.
An effort was made to treat the uncertainty in BAF estimates by using a Monte Carlo
simulation approach. The advantage of this approach is that it explicitly treats known variation in
these parameters thereby providing for the statistical possibility of a high or low end result. In
addition, the distributions themselves follow from the processes at work. As more information about
mercury is obtained, the distributions themselves can be improved. One example of this relationship
has already been discussed; namely, the fact that a skewed BAF distribution for trophic level 4 would
tend to follow from random sampling of a fish population due to the relative scarcity of the oldest
individuals. An additional example is the highly skewed distribution of methylmercury values as a
percent of total, which can be adequately represented as a beta distribution. With respect to the
definition of these distributions, it is important to recall the possibility of regional bias introduced
previously. Thus, it could be argued that FCMs based on regression of data for a large number of
lakes should be given greater weight (perhaps equal to the number of lakes) than data from a single
location. This, however, would only serve to increase the degree of regional bias that is already
present.
June 1996 4-27 SAB REVIEW DRAFT
-------
5. CONCLUSIONS
The following conclusions are presented in approximate order of degree of certainty in the
conclusion, based on the quality of the underlying database. The conclusions progress from
those with greater certainty to those with lesser certainty.
• Mercury emitted to the atmosphere deposits on watersheds and is translocated to waterbodies.
A variable proportion of this mercury is transformed by abiotic and biotic chemical reactions
to organic derivatives, including methylmercury. Methylmercury bioaccumulates in individual
organisms, biomagnifies in aquatic food chains and is also the most toxic form of mercury to
which wildlife are exposed.
• The proportion of total mercury in biota that exists as methylmercury tends to increase with
trophic level. Greater than 90% of the meYcury contained in freshwater fish exists as V
methylmercury. Methylmercury accumulates in fish throughout their lifetime, although
changes hi concentration as a function of tune may be complicated by growth dilution and
changing dietary habits.
• Piscivorous avian and mammalian wildlife are exposed to mercury primarily through
consumption of contaminated fish and accumulate mercury to levels above those in prey items.
• Toxic effects on piscivorous avian and mammalian wildlife due to consumption of
contaminated fish have been observed in association with point source releases of mercury to
the environment.
• Concentrations of mercury in the tissues of wildlife species have been reported at levels
associated with adverse health effects in laboratory studies in the same species.
• Piscivorous birds and mammals receive a greater exposure to mercury than any other known
component of aquatic ecosystems.
• Field data are highly suggestive of adverse lexicological effects in common loons due to
accumulation of mercury originating from airborne emissions. Field data are also suggestive
of adverse lexicological effects in the Florida panther due to mercury; however, this mercury
may have originated from both airborne and non-airborne sources. Field data suggest that bald
eagles have not suffered adverse toxic effects due to airborne mercury emissions. Field data
are insufficient to conclude whether the mink, otter, or kingfisher have suffered adverse toxic
effects due to airborne mercury emissions.
• BAFs for mercury in fish vary widely; however, field data are sufficient to calculate
representative means for different trophic levels. The recommended estimates in this Report
for BAFs for topic levels 3 and 4 are 66,200 and 335,000, respectively. In general, BAFs for
fish sampled from poorly buffered surface waters are higher than those for fish obtained from
well buffered surface waters.
• Based upon knowledge of mercury bioaccumulation in fish, and of feeding rates and the
identity of prey items consumed by piscivorous wildlife, it is possible to rank the relative
exposure of different piscivorous wildlife species. Of the five wildlife species selected for
detailed analysis, the relative ranking of exposure to mercury is this: kingfisher > otter >
June 1996 5-1 SAB REVIEW DRAFT
-------
osprey = mink >. bald eagle. Existing data are insufficient to estimate the exposure of the
Florida panther relative to that of the selected species.
• Local emissions sources (<50 km from receptors) have the potential to increase the exposure
of piscivorous wildlife well above that due to sources located more than 50 km from the
receptors (i.e., "remote" sources).
• Based upon knowledge of mercury exposure to wildlife and its toxicity in long-term feeding
studies, criterion values can be calculated for the protection of piscivorous avian and
mammalian wildlife. A wildlife criterion value is defined as the concentration of total mercury
in water which, if not exceeded, protects avian and mammalian wildlife populations from
adverse effects resulting from ingestion of surface waters and from ingestion of aquatic life
taken from these surface waters.
\
• The criterion value protective of piscivorous avian wildlife is 405 pg/L.
• The criterion value protective of piscivorous mammalian wildlife is 346 pg/L.
• Modeled estimates of mercury concentration hi fish around hypothetical mercury emissions
sources predict exposures at the wildlife WC. The wildlife WC, like the human RfD, is
predicted to be a safe dose over a lifetime. It should be noted, however, that the wildlife
effects used as the basis for the WC are gross clinical manifestations or death. Expression of
subtle adverse effects at these doses cannot be excluded.
There are many uncertainties associated with this analysis, due to an incomplete understanding
of the toxicity of mercury and mercury compounds. The sources of uncertainty include the
following:
• Variability hi the calculated B AFs is a source of uncertainty. BAFs given hi this Report relate
total mercury hi fish (most of which exists as methylmercury) to total mercury in the water
column. Methylmercury is the bioaccumulating species, but existing data are insufficient to
estimate BAFs on a methylmercury basis. Methods for the speciation of mercury in
environmental samples are rapidly improving but remain difficult to perform. Questions also
remain concerning the bioavailability of methylmercury associated with paniculate and
dissolved organic material. Local biogeochemical factors that determine net methylation rates
are not fully understood and are not amenable at this time to generalized modeling.
• The representativeness of field data used in establishing the BAFs is a source of uncertainty.
The degree to which the analysis is skewed by the existing data set is unknown. A
disproportionate amount of data is from north-central and northeastern lakes. The applicability
of these data to a national assessment is not known.
• Limitations of the toxicity database present a source of uncertainty. Few controlled studies of
quantifiable effects of mercury exposure in wildlife are available. These are limited to few
species, necessitating the use of uncertainty factors in extrapolating to species of interest. The
toxic endpoints reported in existing studies can be considered severe, raising questions as to
the degree of protection against subtle effects offered by reference doses and water criteria
calculated on those endpoints. Use of less than lifetime studies for prediction of effects from
lifetime exposure is a source of uncertainty.
June 1996 5-2 SAB REVIEW DRAFT
-------
• Concern has been raised regarding the possibility of toxic effects in species other than those
piscivorous birds and mammals evaluated hi this Report. In particular there is considerable
uncertainty about mercury effects in biota at trophic levels 1 and 2 in aquatic ecosystems and
about effects in terrestrial systems.
• Lack of knowledge of wildlife feeding habits is a source of uncertainty. Existing information
frequently is anecdotal or confined to evaluations of a particular locality; the extent to which
this information is generalizable is open to question. In some instances wherein feeding habits
are relatively well characterized (e.g., Florida panther), the extent of mercury contamination of
prey is poorly known (e.g., in raccoons).
• While the methods used to develop wildlife criteria are based on effects in individual
organisms, the stated goal of the assessment is to characterize the potential for adverse effects
in wildlife populations. Factors which contribute to uncertainty in population-based
assessments include these: variability in the relationship between individuals and populations;
variability in fecundity; lack of data on carrying capacity; and relationships of one population,
of the same or different species, to another population.
• A focus on populations may not always be appropriate. This could be true for endangered
species, which may be highly dependent for the survival of the species on the health of a few
individuals. This may also be true for some regional or local populations of widespread
species; the local population may be "endangered" and, thus, dependent on the survival of
individuals.
To improve the ecological risk assessment for mercury and mercury compounds, ILS. EPA
would need the following:
• Mechanistic research is needed for better understanding of variability of mercury effects. This
would include studies on the following: factors determining rates of methylation and
demethylation; dietary absorption efficiencies from natural food sources; effects of dietary
choice; and bioavailability of methylmercury in the presence of dissolved organic material and
oti^er material which could bind mercury.
• Data are needed for better definition of adverse effects on the species which were evaluated in
this Report. Information is also lacking on species at trophic levels 1 and 2.
• Efforts to develop and standardize methods for analysis of total mercury and methylmercury in
environmental samples (including animal and plant tissue) remain important.
• The current wildlife criteria are based on linear, four-tiered food chains. Research on the
appropriateness of this design and information which will improve the model are important.
• Research is needed to reduce uncertainty regarding the accumulation of mercury at lower
trophic levels.
• High quality field data will be useful to support the process-based research described above, as
well as to determine residue concentrations in fish and other aquatic biota consumed by
wildlife.
June 1996 5-3 SAB REVIEW DRAFT
-------
Based on the extant data and knowledge of developing studies, the U.S. EPA predicts the
following:
• "Regions of concern" are defined as those geographic areas hi the contiguous U.S. that are
thought to receive high levels of mercury deposition and that contain relatively large numbers
(>5% below pH 5.5) of poorly buffered surface waters. The designation of an area as a region
of concern implies an increased risk of mercury toxicity to wildlife. This designation could be
used to define critical habitat, identify wildlife populations potentially at risk, and provide a
focus for future research.
• Increased deposition will lead to increased levels hi fish.
• Increased levels hi fish will lead to toxicity hi piscivorous birds and mammals.
• These impacts are most likely to occur hi areas that receive high levels of deposition and that
also contain poorly buffered surface waters.
June 1996 5-4 SAB REVIEW DRAFT
-------
6. RESEARCH NEEDS
Mercury is unusual among environmental contaminants in that levels that are likely to cause
significant environmental damage exceed those thought to be present "naturally" by less than two (and
perhaps closer to one) order(s) of magnitude. Conservative use of uncertainty factors can, therefore,
lead to calculation of WC or other similar criterion values that are lower than mercury residues
actually measured in the environment With this in mind, there are two general areas within which
research progress must be made if environmental assessments are to be improved. The first area
pertains to basic information on the fate and effects of mercury in the environment, which would result
in reduced use of uncertainty factors and ensure that WC, BAFs, and other estimates, are based on a
mechanistic understanding of the relevant processes. The second area is an improvement in the ability
to detect ecological damage when it is in fact occurring. The present assessment of the "ecological
impacts" of anthropogenic mercury emissions is largely limited to consideration of toxic effects on
individuals. Models that would permit extrapolation of these results to populations (the simplest
extrapolation of individual-based information) do not exist for most species. Further extrapolation to
communities and ecosystems is presently out of the question.
Throughout this assessment, uncertainties, discussed above and elsewhere in the text, have
limited the scope of possible conclusions. Although lack of sufficient data is a limiting factor in all
phases of this assessment, a number of research needs have emerged as being especially important.
These are described below and are presented in no particular order.
6.1 Process-based Research
Mechanistic information is needed to understand the variability that presently typifies the
mercury literature. This research includes laboratory and field studies to identify the determinants of
mercury accumulation in aquatic food chains and kinetic information that would allow researchers to
describe the dynamics of these systems. Areas of uncertainty include these: (1) factors that determine
net rates of methylation and demethylation; (2) dietary absorption efficiency from natural food sources;
(3) effect of dietary choice; and (4) bioavailability of methylmercury in the presence of dissolved
organic material and other potential ligands.
In time it is anticipated that this information can be used to develop process-based models for
mercury bioaccumulation in fish and other aquatic biota. Significant progress in this direction is
represented by the Mercury Cycling Model (MCM), presently being developed and evaluated by the
Electric Power Research Institute (Hudson et al., 1994).
6.2 Wildlife Toxicity Data
There is a need to reduce the present reliance on a relatively few toxicity studies for WC
development. Additional data are needed for wildlife that constitute the most exposed organisms in
various parts of the country (e.g., the Florida panther). There is also a critical requirement for toxicity
data that can be related to effects on populations (see Table 2-1), including effects on organisms that
comprise the lower trophic levels.
6.3 Improved Analytical Methods
Efforts to develop and standardize methods for analysis of total mercury and methylmercury in
environmental samples should be continued. Such methods must recognize the importance of
contamination, both during the collection of such samples and during their analysis. It is particularly
June 1996 6-1 SAB REVIEW DRAFT
-------
important that mercury measurements, which are at present operationally defined (e.g., "soluble",
"adsorbed to organic material"), be made in such a way that mercury residues in fish can be correlated
with the bioavailable mercury pool.
As validated methods become available, it is important to analyze for both total and
methylmercury whenever possible so that differences between aquatic systems can be definitively
linked to differences in methylmercury levels. Analyzing the two mercury species together will
contribute to an understanding of existing data, most of which is reported as total mercury. It is also
anticipated that developing BAFs in terms of methylmercury will reduce the variability that currently
exists around BAF estimates based on total mercury.
6.4 Complexity of Aquatic Food Webs
Present efforts to develop WC values for mercury are based on linear, four-tiered food chain
models. Research is needed to determine the appropriateness of this simple paradigm and to develop
alternatives if field data suggest otherwise. Of particular interest is whether zooplankton and
phytoplankton should be modeled as two different trophic levels. Current information for detritivores
and benthic invertebrates is extremely limited, even though their importance hi mobilizing hydrophobic
organic contaminants has been demonstrated.
6.5 Accumulation in Trophic Levels 1 and 2
Ongoing efforts to understand mercury bioaccumulation in aquatic systems continue to be
focused on trophic levels 3 and 4, despite the fact that uncertainties in PPFs are relatively small.
Additional emphasis should be placed on research at the lower trophic levels. In particular, there is a
need to understand the determinants of mercury accumulation hi phytoplankton and zooplankton, and
how rapid changes hi plankton biomass impact these values.
6.6 Field Residue Data
High-quality field data are needed to support process-based research efforts and to determine
residue concentrations in the fish and other aquatic biota that wildlife eat Whenever possible, it is
desirable to collect residue data at all trophic levels and to analyze mercury levels in the abiotic
compartments of a system (eg., water and sediments). It is particularly important that such
measurements be made hi a broader array of aquatic ecosystem types (including both lakes and rivers)
so that a better understanding of mercury cycling and accumulation can be obtained.
Residue data from wildlife are also needed to identify populations that are being adversely
impacted or are potentially at risk. Feathers and fur hold considerable promise in this regard because
of the potential for "non-invasive" determination of mercury residues. Laboratory research is required,
however, to allow interpretation of these data. Factors such as age, sex, and time to last moult are
likely to result hi variability among individuals of a single population, and need to be understood.
Sampling efforts should be targeted on areas receiving high levels of mercury deposition and/or
regions containing large numbers of poorly buffered surface waters, as discussed throughout this
Report.
June 1996 6-2 SAB REVIEW DRAFT
-------
6.7 Natural History Data
The development of WC requires knowledge of what wildlife eat. Fish sampling efforts are
frequently focused on species that are relevant to human consumers but that may be of little
significance to wildlife. There is an additional need to collect information for macroinvertebrates and
amphibians. Seasonal and spatial effects on predation should be explored and methods developed to
describe this information adequately. Additional life history data is needed to characterize fully the
nature and extent of exposure to mercury. Complicating factors must be considered, including
migratory behaviors and sex-specific differences in distribution and resource allocation. It is
particularly important that information be collected to support the development of predictive
population models for sensitive species.
June 1996 6-3 SAB REVIEW DRAFT
-------
7. REFERENCES
AUard, B. and I. Arsenic (1991). Abiotic reduction of mercury by humic substances in aquatic system
- An important process for the mercury cycle. Water Air Soil Pollut. 56:457-464.
Andersson, A. (1979). Mercury in soils. In: The Biogeochemistry of Mercury in the Environment.
J.O. Nriagu (Ed.), Elsevier/North Holland Biomedical Press, Amsterdam, The Netherlands, pp. 79-106.
Anthony, R.G., M.G. Garrett and C.A. Schuler (1993). Environmental contaminants in bald eagles in
the Columbia River estuary. J. WildL Manage. 57:10-19.
Aulerich, R.J., R.K. Ringer and S. Iwamoto (1974). Effects of dietary mercury in mink. Arch.
Environ. Contain. Toxicol. 2(1):43-51.
Bahnick, D., C. Sauer, B. Butterworth and D. Kuehl (1994). A national study of mercury
contamination in fish. Chemosphere 29:537-546.
Bakir, R, S.F. Damluji, L. Amin-Zaki, M. Murtadha, A. Khalidi, N.Y. Al-Rawi, S. Tikriti, and H.E.
Dhahir (1973). Methylmercury poisoning in Iraq. Science 181: 230-240.
Barr, J.F. (1973). Feeding biology of the common loon (Gavia immer) in oligotrophic lakes of the
Canadian Shield. Ph.D. dissertation, University of Guelph, Canada.
Barr, J.F. (1986). Population dynamics of the common loon (Gavia immer) associated with
mercury-contaminated waters in northwestern Ontario. Occasional Paper No. 56, Canadian Wildlife
Service.
Beck, D.L. (1977). Pesticide and heavy metal residues in Louisiana river otter. M.S. thesis,
University of Texas, Houston, TX.
Belant, J.L. and R.K Anderson (1990). Environmental contaminants in common loons from northern
Wisconsin. Pass. Pigeon 52:306-310.
Beyer, W.N., E. Cromartie and G.B. Moment (1985). Accumulation of methylmercury in the
earthworm, Eisenia foetida, and its effect on regeneration. Bull Environ. Contam. Toxicol. 35:157-
162.
Bigler, W.J., R.H. Jenkins, R.M. Cumbie, G.L. Hoff and E. Prather (1975). Wildlife and
environmental health. Raccoons as indicators of zoonoses and pollutants in southeastern U.S.A. J.
Amer. Med. Assoc. 167:592-597.
Bishop, C.A., M.D. Koster, A.A. Chek, D.J.T. Hussell and K. Jock (1995). Chlorinated hydrocarbons
and mercury in sediments, red-winged blackbirds (Agelaius phoeniceus) and tree swallows
(Tachycineta biocolor) from wetlands in the Great Lakes-SL Lawrence River basin. Environ. Toxicol.
Chem. 14:491-501.
Bleavins, M.R. and R.J. Aulerich (1981). Feed consumption and food passage in mink (Mustela
vison) and European ferrets (Mustela putorius furo). Lab. Animal Sci. 31:268-269.
June 1996 7-1 SAB REVIEW DRAFT
-------
Bloom, N.S (1992). On the chemical form of mercury in edible fish and marine invertebrate tissue.
Can. J. Fish. Aquat. Sci. 49:1010-1017.
Bloom, N.S. and S.W. Effler (1990). Seasonal variability in the mercury speciation of Onondaga Lake
(New York). Water Air Soil Pollut. 53:251-265.
Bloom, N.S., C.J. Watras and J.P. Hurley (1991). Impact of acidification on the methylmercury cycle
of remote seepage lakes. Water Air Soil Pollut. 56:477-491.
Bodaly, R.A., J.W.M Rudd, R.J.P. Fudge and C.A. Kelly (1993). Mercury concentrations in fish
related to size of remote Canadian shield lakes. Can. J. Fish. Aquat. Sci. 50:980-987.
Boney, AD. (1971). Sub-lethal effects of mercury on marine algae. Mar. Pollut. Bull. 2:69-71.
Borst, H.A. and C.G. Lieshout (1977). Phenylmercuric acetate intoxication in mink. Tijdschr.
Diergeneesk 102:495-503.
Boudou, A. and F. Ribeyre (1985). Experimental study of trophic contamination of Salmo gairdneri
by two mercury compounds - HgC^ and CH3HgCl - analysis at the organism and organ levels. Water
Air Soil Pollut. 26:137-148.
Bowerman, W.W. IV (1993). Regulation of bald eagle (Haliaeetus leucocephalus) productivity in the
Great Lakes basin: An ecological and lexicological approach. Ph.D. dissertation, Michigan State
University, East Lansing, MI.
Bowerman, W.W. IV., E.D. Evans, J.P. Giesy and S. Postupalsky (1994). Using feathers to assess risk
of mercury and selenium to bald eagle reproduction in the Great Lakes regioa Arch. Environ.
Contam. Toxicol. 27:294-298.
Boyer, H.A. (1982). Trace elements in the water, sediments, and fish of the upper Mississippi River,
twin cities metropolitan area. In: Contaminants in the Upper Mississippi River. J.G. Wiener, R.V.
Anderson and D.R. McConville (Eds.). Butterworth Publishers, Boston, MA, pp. 195-230.
Braune, B.M. and D.E. Gaskin (1987). Mercury levels in Bonaparte's gulls (Larus Philadelphia)
during autumn molt in the Quoddy region, New Brunswick, Canada. Arch. Environ. Contam. Toxicol.
16:539-549.
Braune, B.M. and R.J. Norstrom (1989). Dynamics of organochlorine compounds in herring gulls: III.
Tissue distribution and bioaccumulation in Lake Ontario gulls. Environ. Toxicol. Chem. 8:957-968.
Burbacher, T.M., M.K. Mohamed and N.K. Mottett. 1988. Methylmercury effects on reproduction
and offspring size at birth. Reprod. Toxicol. l(4):267-278.
Burger, J., J.A. Rodgers, Jr. and M. Gochfeld (1993). Heavy metal and selenium levels in endangered
wood storks Mycteria americana from nesting colonies in Florida and Costa Rica. Arch. Environ.
Contam. Toxicol. 24:417-420.
June 1996 7-2 SAB REVIEW DRAFT
-------
Cabana, G., A. Tremblay, J. Kalff and J.B. Rasmussen (1994). Pelagic food chain structure in Ontario
lakes: a determinant of mercury levels in lake trout (Salvelinus namaycush). Can. J. Fish. Aquat. Sci.
51:381-389.
Calder, W.A. and E.J. Braun (1983). Scaling of osmotic regulation in mammals and birds. Am. J.
Physiol 244:R601-R606.
Charbonneau, S.M., I.C. Munro, E.A. Nera, R.F. Willes, T. Kuiper-Goodman, F. Iverson, C.A.
Moodie, D.R. Stoltz, F.A.J. Armstrong, J.F. Uthe, B.C. Grice (1974). Subacute toxicities of
methylmercury in the adult cat Toxic. Appl. Phann. 27:569-581.
Charbonneau, S.M., I.C. Munro, E.A. Nera, F.AJ. Armstrong, R.F. Willes, F. Bryce and R.F. Nelson
(1976). Chronic toxicity of methylmercury hi the adult cat. Interim report. Toxicol. 5:337-349.
Clark, ICE., F.A.P.C. Gobas and D. Mackay (1990). Model of organic chemical\iptake and clearance
by fish from food and water. Environ. Sci. Technol. 24:1203-1213.
Clarkson, T.W. (1990). Human health risks from methylmercury in fish. Environ. Toxicol. Chem.
9:957-961.
Colborn, T.I. (1991). Epidemiology of Great Lakes bald eagles. /. Environ. Health Toxicol.
4:395-453.
Cope, W.G., J.G. Wiener and R.G. Rada (1990). Mercury accumulation in yellow perch in Wisconsin
seepage lakes: Relation to lake characteristics. Environ. Toxicol. Chem. 9:931-940.
Crowder, A. (1991). Acidification, metals and macrophytes. Environ. Pqllut. 71:171-203.
Crowder, A.A., W. Dushenko and J. Grieg (1988). Metal contamination of wetland food chains in the
Bay of Quinte, Ontario. Environment Ontario, Nov. 28-29, 1988. Toronto, Canada, pp. 133-153.
Cumbie, P.M. (1975). Mercury levels in Georgia otter, mink, and freshwater fish. Bull. Environ.
Contam. Toxicol. 14:193-196.
Dillon, T.M. (1977). Mercury and the estuarine marsh clam, Rangia cuneata Gray. I. Toxicity. Arch.
Environ. Contain. Toxicol. 6:249-255.
Dorfman, D. (1977). Tolerance of Fundulus heteroclitus to different metals in salt waters. Bull. New
Jersey Acad. Sci. 22:21.
Dourson, M.L. and J.F. Stara (1983). Regulatory history and experimental support of uncertainty
(safety) factors. Reg. Toxicol. Pharmacol. 3:224-238.
Dukerschein, J.T., J.G. Wiener, R.G. Rada and M.T. Steingraeber (1992). Cadmium and mercury in
emergent mayflies (Hexagenia bilineatd) from the upper Mississippi River. Arch. Environ. Contam.
Toxicol 23:109-116.
Eisler, R. (1987). Mercury hazards to fish, wildlife, and invertebrates: A synoptic review. Publication
No. 85 (1.10), U.S. Fish and Wildlife Service, Department of the Interior, Washington, DC.
June 1996 7-3 SAB REVIEW DRAFT
-------
Eisler, R. and RJ. Hennekey (1977). Acute toxicities of Cd+2, Cr+6, Hg+2, Ni+2, and Zn+2 to
estuarine macrofauna. Arch. Environ. Contam. Toxicol. 6:315-323.
Ensor, K.L., D.D. Helwig and L.C. Wemmer (1992). Environmental mercury and lead in Minnesota
common loons (Gavia immef). Minnesota Pollution Control Agency, Water Quality Division, St. Paul,
MM.
Environment Canada (1991). Toxic chemicals in the Great Lakes and associated effects: Volume I -
Contaminant levels and trends. Department of Fisheries and Oceans, Health and Welfare Canada,
Toronto, Canada.
Evans, R.D. (1986). Sources of mercury contamination hi the sediments of small headwater lakes hi
south-central Ontario, Canada. Arch. Environ. Contain. Toxicol. 15:505-512.
Evans, H.L., R. Carman and B. Weiss. 1977. Methylmercury: Exposure duration and regional
distribution as determinants of neurotoxicity in nonhuman primates. Toxicol. Appl. Pharmacol.
41:15-33.
Fimreite, N. (1970). Effects of methylmercury treated feed on the mortality and growth of leghorn
cockerels. Can. J. Anim. Sci. 50:387-389.
Fimreite, N. (1971). Effects of methylmercury on ring-necked pheasants. Canadian Wildlife Service
Occasional Paper Number 9. Department of the Environment. 39 pp.
Fimreite, N. (1974). Mercury contamination of aquatic birds hi northwestern Ontario. J. Wildl.
Manage. 38:120-131.
Fimreite, N. (1979). Accumulation and effects of mercury on birds. In: The Biogeochemistry of
Mercury in the Environment. J.O. Nriagu (Ed.), Elsevier/North-Holland Biomedical Press, Amsterdam,
The Netherlands, pp. 601-628.
Finley, M.T. and R.C. Stendell (1978). Survival and reproductive success of black ducks fed methyl
mercury. Environ. Pollu. 16:51-64.
Finley, M.T., W.H. Stickel and RE. Christensen (1979). Mercury residues in tissues of dead and
, surviving birds fed methylmercury. Bull. Environ. Contam. Toxicol. 21:105-110.
Fjeld, E. and S. Rognerud (1993). Use of path analysis to investigate mercury accumulation in brown
trout (Salmo truttd) in Norway and the influence of environmental factors. Can. J. Fish. Aquat. Sci.
50:1158-1167.
Fleming, W.J., J.A. Rodgers, J.A., Jr., and C.J. Stafford, C.J. (1984). Contaminants in wood stork
eggs and their effects on reproduction, Florida, 1982. Colonial Waterbirds 7:88-93.
Florida Department of Environmental Regulation (FDER, 1990). Mercury, largemouth bass, and water
quality: A preliminary report. Department of Environmental Regulation, Florida.
Florida Panther Interagency Committee (FPIC, 1989). Mercury contamination in Florida panthers.
Status Report of the Technical Subcommittee.
June 1996 7-4 SAB REVIEW DRAFT
-------
Fowler, B.A. and J.S. Woods. 1977. The transplacental toxicity of methylmercury to fetal rat liver
mitochondria. Lab. Invest. 36:122-130.
Putter, M.N. (1994). Pelagic food-web structure influences probability of mercury contamination in
lake trout (Salvelinus namaycush). Sci. Tot. Environ. 145:7-12.
Gilbertson, M., T. Kubiak, J. Ludwig and G. Fox (1991). Great Lakes embryo mortality, edema, and
deformities syndrome (GLEMEDS) in colonial fish-eating birds: Similarity to chick-edema disease. J.
Toxicol. Environ. Health 33:455-520.
Gilmour, C.C. and E.A. Henry (1991). Mercury methylation in aquatic systems affected by acid
deposition. Environ. Pollut. 71:131-169.
Glass, G.E., J.A. Sorensen, K.W. Schmidt, J.K. Huber and G.R. Rapp, Jr. (1993). Mercury sources
and distribution in Minnesota's aquatic resources: Precipitation, surface water, sediments, plants,
plankton, and fish. Final report to Minnesota Pollution Control Agency and Legislative Commission
on Minnesota Resources, 1989-1991 (Contract Nos. 831479 and WQ/PDS020).
Gobas, F.A.P.C. (1993). A model for predicting the bioaccumulation of hydrophobic organic
chemicals in aquatic food webs: Application to Lake Ontario. Ecol. Model. 69:1-17.
Godbold, D.L. (1991). Mercury-induced root damage in spruce seedlings. Water Air Soil Pollut.
56:823-831.
Grieb, T.M., C.T. Driscoll, S.P. Gloss, C.L. Schofield, G.L. Bowie and D.B. Porcella (1990). Factors
affecting mercury accumulation in fish in the upper Michigan peninsula. Environ. Toxicol. Chem.
9:919-930.
Grier, J.W. (1974). Reproduction, organochlorines, and mercury in northwestern Ontario bald eagles.
Can. Field Nat. 88:469-475.
Grubb, T.G., S.N. Wiemeyer and L.F. Kiff (1990). Eggshell thinning and contaminant levels in bald
eagle eggs from Arizona, 1977 to 1985. The Southwestern Naturalist 35:29*8-301.
Gutenmann, W.H., J.G. Ebel, Jr., H.T. Kuntz, K.S. Yourstone and D.J. Lisk (1992). Residues of
p,p'-DDE and mercury in lake trout as a function of age. Arch. Environ. Contain. Toxicol. 22:452-455.
Halbrook, R.S. (1978). Environmental pollutants in the river otter of Georgia. M.S. thesis, University
of Georgia, Athens, GA.
Harper, R.G., D.S. Hopkins and T.C. Dunstan (1988). Nonfish prey of wintering bald eagles in
Illinois. Wilson Bull. 100:688-690.
Harrison, S.E., J.F. Klaverkamp and R.H. Hesslein (1990). Fates of metal radiotracers added to a
whole lake: accumulation in fathead minnow (Pimephales promelas) and lake trout (Salvelinus
namaycush). Water Air Soil Pollut. 52:277-293.
Heinz, G.H. (1974). Effects of low dietary levels of methylmercury on mallard reproduction. Bull.
Environ. Contain. Toxicol. 11:386-392.
June 1996 7-5 SAB REVIEW DRAFT
-------
Heinz, G.H. (1975). Effects of methylmercury on approach and avoidance behavior of mallard
ducklings. Bull Environ. Contain. ToxicoL 13:554-564.
Heinz, G.H. (1976a). Methylmercury: Second-year feeding effects on mallard reproduction and
duckling behavior. J. Wildl. Manag. 40(1):82-90.
Heinz, G.H. (1976b). Methylmercury: Second-generation reproductive and behavioral effects on
mallard ducks. J Wildl Manag. 40(4):710-715.
Heinz, G.H. (1979). Methylmercury: Reproductive and behavioral effects on three generations of
Mallard ducks. J. Wildl. Manage. 43:394-401.
Helmke, P.A., W.P. Robarge, Rl. Korotev and P.J. Schomberg (1979). Effects of soil-applied sewage
sludge on concentrations of elements in earthworms. J. Environ. Qual 8:322-327.
Hildebrand, S.G., R.H. Strand and J.W. Huckabee (1980). Mercury accumulation in fish and
invertebrates of the North Fork Holston River, Virginia and Tennessee. /. Environ. Qual 9:393-400.
Hirano, M., K. Mitsumori, K. Maita, et al. 1986. Further carcinogenicity study on methylmercury
chloride in ICR mice. Jap. J. Vet Sci. 48(1): 127-135.
Hongve, D., O.K. Skogheim, A. Hindar and H. Abrahamsen (1980). Effects of heavy metals in
combination with NTA, humic acid, and suspended sediment on natural phytoplankton photosynthesis.
Bull Environ. Contam. ToxicoL 25:594-600.
Huckabee, J.W., J.W. Elwood and S.G. Hildebrand (1979). Accumulation of mercury in freshwater
biota. In: The Biogeochemistry of Mercury in the Environment. J.O. Nriagu (Ed.), Elsevier/North
Holland Biomedical Press, Amsterdam, The Netherlands, pp. 277-302.
Hudson, R.J.M., S.A. Gherini, C.J. Watras, and D.B. Porcella (1994). Modelling the biogeochemical
cycle of mercury in lakes: the mercury cycling model (MCM) and its application to the MTL study
lakes. In: Mercury Pollution: Integration and Synthesis. C.J. Watras and J.W. Huckabee (Eds.),
Lewis Publishers, Boca Raton, FL, pp. 473-523.
Jackson, T.A. (1991). Biological and environmental control of mercury accumulation by fish in lakes
and reservoirs of northern Manitoba, Canada. Can. J. Fish. Aquat. Sci. 48:2449-2470.
Johansson, K., M. Aastrup, A. Andersson, L. Bringmark and A. Iverfeldt (1991). Mercury in Swedish
forest soils and waters - Assessment of critical load. Water Air Soil Pollut. 56:267-281.
Johnston, T.A., R.A. Bodaly and J.A. Mathias (1991). Predicting fish mercury levels from physical
characteristics of boreal reservoirs. Can. J. Fish. Aquat. Sci. 48:1468-1475.
Jordan, D. (1990). Mercury contamination: Another threat to the Florida panther. Fish and Wildlife
Service Endangered Species Technical Bulletin 15(2): 1.
Joslin, J.D. (1994). Regional differences in mercury levels in aquatic ecosystems: A discussion of
possible causal factors with implications for the Tennessee River system and the northern hemisphere.
Environ. Manage. 18:559-567.
June 1996 7-6 SAB REVIEW DRAFT
-------
Khera, S. 1973. Reproductive capability of male rats and mice treated with methyl mercury. Toxicol.
Appl. Pharm. 24:167-177.
Khera, K.S. and S.A. Tabacova. 1973. Effects of methylmercuric chloride on the progeny of mice
and rats treated before or during gestation. Food. Cosmet Toxicol. 11:245-254.
Klaassen, C.D., M.O. Amdur and J. Doull (1986). Casarett and Doull's Toxicology: The Basic
Science of Poisons. MacMillan Publishing Company, New York, NY.
Klaunig, J., S. Koepp, and M. McCormick (1975). Acute toxicity of a native mummichog population
(Fundulus heteroclitus) to mercury. Bull Environ. Contain. Toxicol 14:534-536.
Koller, L.D., J.H. Exon and B. Arbogast. 1977. Methylmercury: Effect on serum enzymes and
humoral antibody. Journal of Toxicology and Environmental Health 2:1115-1123.
Kozie, K.D. and R.K. Anderson (1991). Productivity, diet, and environmental contaminants in bald
eagles nesting near the Wisconsin shoreline of Lake Superior. Arch. Environ. Contain. Toxicol 20:41-
48.
Kucera, E. (1983). Mink and otters as indicators of mercury hi Manitoba waters. Can. J. Zool.
61:2250-2256.
Kudo, A., H. Nagase and Y. Ose (1982). Proportion of methylmercury to the total amount of mercury
hi river waters in Canada and Japan. Water Res. 16:1011-1015.
Kuiper, J. (1981). Fate and effects of mercury in marine plankton communities in experimental
enclosures. Ecotoxicol Environ. Safety 5:106-134.
Lange, T.R., H.E. Royals and Li. Connor (1993). Influence of water chemistry on mercury
concentration in largemouth bass from Florida lakes. Trans. Am. Fish. Soc. 122:74-84.
•
Lathrop, R.C., P.W. Rasmussen and D.R. Knauer (1991). Mercury concentrations in walleyes from
Wisconsin (USA) lakes. Water Air Soil Pollut. 56:295-307.
Lee, Y.H., H. Hultberg and I. Andersson (1985). Catalytic effect of various metal ions on the
methylation of mercury hi the presence of humic substances. Water Air Soil Pollut. 25:391-400.
Lee, Y.H. and H. Hultberg (1990). Methylmercury in some Swedish surface waters. Environ. Sci.
Technol. 9:833-841.
Lindqvist, O. (1991). Mercury in the Swedish environment. Recent research on causes, consequences
and corrective measures. Water Air Soil Pollut. 55:1-261.
MacCrimmon, H.R., C.D. Wren and B.L. Gots (1983). Mercury uptake by lake trout, Salvelinus
namaycush, relative to age, growth, and diet hi Tadenac Lake with comparative data from other
precambrian shield lakes. Can. J. Fish. Aquat. Sci. 40:114-120.
Maserti, B.E. and R. Ferrara (1991). Mercury in plants, soil and atmosphere near a chlor-alkali
complex. Water Air Soil Pollut. 56:15-20.
June 1996 7-7 SAB REVIEW DRAFT
-------
Mathers, R.A. and P.H. Johansen (1985). The effects of feeding ecology on mercury accumulation in
walleye (Stizostedion vitrewri) and pike (Esox lucius) in Lake Simcoe. Can. J. Zool. 63:2006-2012.
May, K., M. Stoeppler and K. Reisinger (1987). Studies hi the ratio total mercury/methylmercury in
the aquatic food chain. Tax. Environ. Chem. 13:153-159.
McGrath, J.T. (1960). Neurological Examination of the Dog with Clinicopathological Observations.
2nd ed. Lea and Febiger, Philadelphia, PA.
•
McKim, J.M., G.F. Olson, G.W. Holcombe and E.P. Hunt (1976). Long-term effects of
methylmercuric chloride on three generations of brook trout (Salvelinus fontinalis): Toxicity,
accumulation, distribution, and elimination. J. Fish. Res. Bd. Can. 33:2726-2739.
McMurtry, M.J., D.L. Wales, W.A. Schneider, G.L. Beggs and P.E. Diamond (1989). Relationship of
mercury concentrations hi lake trout (Salvelinus namaycush) and smallmouth bass (Micropterus
dolomieui) to the physical and chemical characteristics of Ontario lakes. Can. J. Fish. Aquat. Sci.
46:426^34. '
Meili, M., A. Iverfeldt and L. Hakanson (1991). Mercury in the surface water of Swedish forest lakes -
concentrations, speciation and controlling factors. Water Air Soil Pollut. 56:439-453.
Mhatre, G.N. and S.B. Chaphekar (1985). The effect of mercury on some aquatic plants. Environ.
Pollut. 39:297-216.
Michigan Department of Natural Resources (MDNR, 1993). Mercury hi Michigan's environment:
Environmental and human health concerns. A science report to Governor John Engler. R.D. Sills,
Michigan Environmental Science Board, Lansing, MI.
Miskunnun, B.M., J.W.M. Rudd and C.A. Kelly (1992). Influence of dissolved organic carbon, pH
and microbial respiration rates on mercury methylation and demethylation hi lake water. Can. J. Fish.
Aquat. Sci. 49:17-22.
Mitsumori, K., M. Hirano, H. Ueda, et al. 1990. Chronic toxicity and carcinogenicity of
methylmercury chloride hi B6C3F1 mice. Fundamental and Applied Toxicology 14:179-190.
Mohamed, M., T. Burbacher and N. Mottet. 1987. Effects of methyl mercury on testicular functions
in Macaca fascicularis monkeys. Pharmacol. Toxicol. 60(l):29-36.
Mosbaek, H., J.C. Tjell and T. Sevel (1988). Plant uptake of airborne mercury in background areas.
Chemosphere 17:1227-1236.
Munro, I.C., E.A. Nera, S.M. Charbonneau, et al. 1980. Chronic toxicity of methylmercury in the rat
J. Environ. Path. Toxicol. 3:437-447.
Muramoto, S. and Y. Oki (1984). Influence of anionic surface-active agents on the uptake of heavy
metals by water hyacinth (Eichornia crassipes). Bull. Environ. Contam. Toxicol. 33:444-450.
Nagase, H., Y. Ose, T. Sato and T. Ishikawa (1984). Mercury methylation by compounds in humic
material. Sci. Tot. Environ. 32:147-156.
June 1996 7-8 SAB REVIEW DRAFT
-------
Nagy, K.A. (1987). Field metabolic rate and food requirement scaling in mammals and birds. Ecol
Monogr. 57:111-128.
National Acid Precipitation Assessment Program (NAPAP, 1990). Acidic deposition: State of science
and technology, volume n, aquatic pocesses and effects. National Acid Precipitation Program,
Washington, D.C.
Nature Conservancy (1994). Heritage database. Eastern Heritage Task Force, Boston, MA.
Developed under U.S. EPA appropriation 683V40108 for the Office of Air Quality Planning and
Standards, Emissions Standards Division, Research Triangle Park, NC.
Niimi, A.J. and G.P. Kissoon (1994). Evaluation of the critical body burden concept based on
inorganic and organic mercury toxicity to rainbow trout (Oncorhynchus mykiss). Arch. Environ,
Contam. Toxicol. 26:169-178.
Nilsson, A. and L. Hakanson (1992). Relationships between mercury in lake water, water color and
mercury in fish. Hydrobiologia 235/236:675-683.
Nolen, G.A., E.V. Buchler, R.G. Geil, et al. 1972. Effects of trisodium nitrotriacetate on cadmium
and methylmercury toxicity and teratogenicity in rats. Toxicol. Appl. Pharmacol 23:222-237.
O'Connor, D.J. and S.W. Nielsen (1980). Environmental survey of methylmercury levels in wild mink
(Mustela visori) and otter (Lutra canadensis) from the northeastern United States and experimental
pathology of methylmercurialism in the otter. Worldwide Furbearer Conference Proceedings, pp.
1728-1745.
Odsjo', T. (1982). Eggshell thinning and levels of DDT, PCB and mercury in the eggs of osprey
(Pandion haliaetus L.) and marsh harrier (Circus aeruginosus L.) in relation to their breeding success
and population status in Sweden. Ph.D. dissertation, University of Stockholm, Sweden.
Olson, K.R. and R.C. Harrel (1973). Effect of salinity on acute toxicity of mercury, copper, and
chromium'for Rangia cuneata (Pelecypoda, Mactridae). Contrib. Mar. Sci. 17:9-13.
Olson, K.R., K.S. Squibb and R.J. Cousins (1978). Tissue uptake, subcellular distribution, and
metabolism of 14CH3HgCl and CH3203HgCl by rainbow trout, Salmo gairdneri. J. Fish. Res. Bd. Can.
35:381-390.
Parks, J.W., A. Lutz and J.A. Sutton (1989). Water column methylmercury in the Wabigoon/English
River-Lake system: Factors controlling concentrations, speciation, and net production. Can. J. Fish.
Aquat. Sci. 46:2184-2202.
Peakall, D.B. (1988). Known effects of pollutants on fish-eating birds hi the Great Lakes of North
America. In: Toxic Contamination in Large Lakes. Vol. I: Chronic Effects of Toxic Contaminants in
Large Lakes. N.W. Schmidtke (Ed.), Lewis Publishers, Inc., Chelsea, MI, pp. 39-54.
Porcella, D.B., C.J. Watras and N.S. Bloom (1991). Mercury species in lake water. In: The
deposition and fate of trace metals in our environment. Gen. Tech. Rep. NC-150, S. Verry and S.J.
Vermette, U.S. Dept Agric., Forest Service, North Central Forest Exp. Station, St. Paul, MN, pp.
127-138.
June 1996 7-9 SAB REVIEW DRAFT
-------
Rada, R.G., DE. Powell and J.G. Wiener (1993). Whole-lake burdens and spatial distribution of
mercury in surficial sediments in Wisconsin seepage lakes. Can. J. Fish. Aquat Sci. 50:865-873.
Ribeyre, R. and A. Boudou (1984). Bioaccumulation et repartition tissulaire du mercure - HgCl2 et
CH3HgCl - chez Salmo gairdneri apres contamination par voie directe. Water Air Soil Pollut. 23:169-
186.
Roelke, M.E., D.P. Schultz, C.F. Facemire, S.F. Sundlof and HE. Royals (1991). Mercury
contamination in Florida panthers. A report of the Florida Panther Technical Subcommittee to the
Florida Panther Interagency Committee.
Saouter, E., L. Hare, P.G.C. Campbell, A. Boudou and F. Ribeyre (1993). Mercury accumulation in
the burrowing mayfly (Hexagenia rigidd) (ephemeroptera) exposed to CH3HgCl or HgCl2 in water and
sediment Water Res. 27:1041-1048.
Sarkar, A. and S. Jana (1986). Heavy metal pollutant tolerance ofAzolla pinnata. Water Air Soil
Pollut. 27:15-18.
Sato, T. and F. Ikuta (1975). Long-term studies on the neurotoxicity of small amount of
methylmercury in monkeys (first report). In: Tsubaki T, ed. Studies on the Health Effects of
Alkylmercury in Japan. Japan: Environment Agency, 63-70.
Scheuhammer, A.M (1987). The chronic toxicity of aluminum, cadmium, mercury, and lead hi birds:
A review. Environ. Pollut. 46:263-295.
Scheuhammer, AM. (1988). Chronic dietary toxicity of methylmercury in the zebra finch, Poephila
guttata. Bull Environ. Contam. Toxicol. 40:123-130.
Scheuhammer, A.M. (1991). Effects of acidification on the availability of toxic metals and calcium to
wild birds and mammals. Environ. Pollut. 71:329-375.
•
Scheuhammer, A.M. and P.J. Blancher (1994). Potential risk to common loons (Gavia immer) from
methylmercury exposure in acidified lakes. Hydrobiologia 279-289:445-455.
Schlegel, H., D.L. Godbold and A. Huttermann (1987). Whole plant aspects of heavy metal induced
changes in CO2 uptake and water relations of spruce (Picea abies) seedlings. Physiol. Plant.
69:265-270.
Schmidt, M. (1987). Atmospharischer eintrag und interner umsatz von schwermetallen in
waldokosystemen. ber. forschungszentr. WaldokosysTWaldst. A 34/37; Gottingen.
Schreiner, G., B. Ulbrich and R. Bass. 1986. Testing strategies hi behavioral teratology: II.
Discrimination learning. Neurobehav. Toxicol. Teratol. 8:567-572.
Scott, M.L. (1977). Effects of PCBs, DDT, and mercury compounds in chickens and Japanese quail.
Fed. Proc. 36:1888-1893.
Scott, D.P. and F.A.J. Armstrong (1972). Mercury concentration hi relation to size in several species
of freshwater fishes from Manitoba and Northwestern Ontario. J. Fish. Res. Ed. Can. 29:1685-1690.
June 1996 7-10 SAB REVIEW DRAFT
-------
Shefry, T.B. and J.R. St. Amant (1982). Mercury burdens in furbearers in Wisconsin. J. Wildl.
Manage. 46:1117-1120.
Siegel, S.M., B.Z. Siegel, N. Puerner and T. Speitel (1975). Water and soil biotic relations in mercury
distribution. Water Air Soil Pollut. 4:9-18.
Siegel, S.M., B.Z. Siegel, C. Lipp, A. Kruckeberg, O.H.N. Towers and H. Warren (1985). Indicator
plant-soil mercury patterns in a mercury-rich mining area of British Columbia. Water Air Soil Pollut.
25:73-85.
Siegel, S.M., B.Z. Siegel, C. Barghigiani, K. Aratani, P. Penny and D. Penny (1987). A contribution
to the environmental biology of mercury accumulation in plants. Water Air Soil Pollut. 33:65-72.
Singleton, F.L. and R.K. Guthrie (1977). Aquatic bacterial populations and heavy metals -1.
Composition of aquatic bacteria in the presence of copper and mercury salts. Water Res. 11:639-642.
Skurdal, J., T. Qvenild and O.K. Skogheim (1985). Mercury accumulation in five species of
freshwater fish in Lake Tyrifjorden, southeast Norway, with emphasis on their suitability as test
organisms. Environ. BioL Fish. 14:233-237.
Sorensen, J.A., GJE. Glass, K.W. Schmidt, J.K. Huber and G.R. Rapp, Jr. (1990). Airborne mercury
deposition and watershed characteristics in relation to mercury concentrations in water, sediments,
plankton, and fish of eighty northern Minnesota lakes. Environ. Sci. Technol. 24:1716-1727.
Spry, D.J. and J.G. Wiener (1991). Metal bioavailability and toxicity to fish in low-alkalinity lakes: A
critical review. Environ. Pollut. 71:243-304.
Stanley, R.A. (1974). Toxicity of heavy metals and salts to Eurasian watermilfoil (Myriophyllwn
spicatum L.). Arch. Environ. Contam. Toxicol. 2:331-341.
State of Wisconsin (1989). Technical support document for NR 105, 1988. Wisconsin Administrative
Code NR 105.07, 1989.
Stickel, L.F., W.H. Stickel, M.A.R. McLane and M. Bruns (1977). Prolonged retention of methyl
mercury by mallard drakes. Bull. Environ. Contam. Toxicol. 18:393-400.
Stoewsand, G.S., C.A. Bache and D.J. Lisk (1974). Dietary selenium protection of methylmercury
intoxication of Japanese quail. Bull. Environ. Contam. Toxicol. 11:152-156.
Suchanek, T.H., P.J. Richerson, L.A. Woodward, D.G. Slotton, L.J. Holts and C.E.E. Woodmansee
(1993). A survey and evaluation of mercury in: sediment, water, plankton, periphyton, benthic
invertebrates and fishes within the aquatic ecosystem of Clear Lake, California. Preliminary Report,
Prepared for the U.S. Environmental Protection Agency, Region 9: Superfund Program, by the
Institute of Ecology, University of California at Davis, Davis, CA.
Suns, K. and G. Hitchin (1990). Interrelationships between mercury levels in yearling yellow perch,
fish condition and water quality. Water Air Soil Pollut. 50:255-265.
June 1996 7-11 SAB REVIEW DRAFT
-------
Talmage S.S. and B.T. Walton (1993). Food chain transfer and potential renal toxicity of mercury to
small mqmmais as a contaminated terrestrial field site. Ecotoxicol. 2:243-256.
Thomann, R.V. (1989). Bioaccumulation model for organic chemical distribution in aquatic food
chains. Environ. Sci. Technol. 23:699-707.
U.S. Environmental Protection Agency (U.S. EPA, 1985). Ambient water quality criteria for mercury
- 1984. EPA/440/5-84/026. Office of Water, Washington, D.C.
U.S. Environmental Protection Agency (U.S. EPA, 1989). Ecological assessment of hazardous waste
sites: A field and laboratory reference. Corvallis, Oregon: Environmental Research Laboratory.
EPA/600/3-89/013.
U.S. Environmental Protection Agency (U.S. EPA; 1992a). Peer review workshop on a framework for
ecological risk assessment Risk assessment forum. EPA/625/3-91/002.
U.S. Environmental Protection Agency (U.S. EPA, 1992b). A national study of chemical residues in
fish. EPA 823/R-92/008. Office of Water Regulations and Standards, Washington, D.C.
U.S. Environmental Protection Agency (U.S. EPA, 1993a). Wildlife exposure factors handbook.
EPA/600/R-93/187a. U.S. EPA Office of Research and Development, Washington, DC.
U.S. Environmental Protection Agency (U.S. EPA, 1993b). Great Lakes water quality initiative
criteria documents for the protection of wildlife (proposed) DDT; Mercury; 2,3,7,8-TCDD; PCBs.
EPA/822/R-93/006. U.S. EPA Office of Science and Technology, Washington, DC.
U.S. Environmental Protection Agency (U.S. EPA, 1993c). Water quality guidance for the Great
Lakes system and correction: Proposed rules. Fed. Regist. 58(72):20802-21047 (April 16, 1993).
U.S. Environmental Protection Agency (U.S. EPA, 1994). Draft proceedings of the national wildlife
criteria methodologies meeting, April 13-16, 1992, Charlottesville, VA. U.S. EPA Office of Water
and Office of Science and Technology, Washington, DC.
U.S. Environmental Protection Agency (U.S. EPA, 1995a). Trophic level and exposure analyses for
selected piscivorous birds and mammals. Volume I. Analysis for species of the Great Lakes Basin
(Draft). U.S. EPA Office of Science and Technology, Washington, DC.
U.S. Environmental Protection Agency (U.S. EPA, 1995b). Final water quality guidance for the Great
Lakes system: Final Rule. Fed. Regist. 60(56): 15366-15425 (March 23, 1995).
U.S. Fish and Wildlife Service (U.S. FWS, 1993). Mercury contamination in tissues of Florida bald
eagles. Final Project Report. Prepared for the U.S. Fish and Wildlife Service, Department of the
Interior, Washington, DC. (December, 1993)
Vermeer, K., F.A.J. Armstrong and D.R.M. Hatch (1973). Mercury in aquatic birds at Clay Lake,
Western Ontario. J. Wildl. Manage. 37:58-61.
Watras, C.J. and N.S. Bloom (1992). Mercury and methylmercury in individual zooplankton:
Implications for bioaccumulation. Limnol. Oceanogr. 37:1313-1318.
June 1996 7-12 SAB REVIEW DRAFT
-------
Weber, J.H. (1993). Review of possible paths for abiotic methylation of mercury (II) in the aquatic
environment. Chemosphere 26:2063.
Weil, C.S. and D.D. McCollister (1963). Relationship between short- and long-term feeding studies in
designing an effective toxicity test. Agric. Food Chem. 11:486-491.
Weis, J.S. and P. Weis (1989). Tolerance and stress in a polluted environment. BioScience 39:89-95.
Weis, J.S. and P. Weis (1995). Effects of embryonic exposure to methylmercury on larval prey-
capture ability in the mummichog, Fundulus heteroclitus. Env. Toxicol. Chem. 14:153-156.
Welch, L.J. (1994). Contaminant burdens and reproductive rates of bald eagles breeding in Maine.
Ph.D. dissertation, University of Maine, Orono, ME.
Wells, J.R., P.B. Kaufman and J.D. Jones (1980). Heavy metal contents in some macrophytes from
Saginaw Bay (Lake Huron, USA). Aquat. Bot. 9:185-193.
Westermark, T., T. Odsjo, T. and A.G. Johnels (1975). Mercury in bird feathers before and after the
Swedish ban on alkyl mercury in agriculture. Ambio 4:87-97.
Wiemeyer, S.N., T.G. Lament, CM. Bunck, C.R. Sindelar, F.J. Gramlich, J.D. Fraser and M.A. Byrd
(1984). Organochlorine pesticide, polychlorobiphenyl, and mercury residues in bald eagle eggs -
1969-1979 - and their relationships to shell thinning and reproduction. Arch. Environ. Contain.
Toxicol 13:529-549.
Wiemeyer, S.N., D.M. Bunck and C.J. Stafford (1993). Environmental contaminants in bald eagle
eggs - 1980-1984 - and further interpretations of relationships to productivity and shell thickness.
Arch. Environ. Contain. Toxicol. 24:213-227.
Wiener, J.G., G.A. Jackson, T.W. May and B.P. Cole (1982). Longitudinal distribution of trace
elements (As, Cd, Cr, Hg, Pb, and Se) in fishes and sediments in the upper Mississippi River. In:
Contaminants in the Upper Mississippi River. J.G. Wiener, R.V. Anderson and D.R. McConville
(Eds.). Butterworth Publishers, Boston, MA. p. 139-170.
Wiener, J.G.; R.E. Martini, T.B. Sherry and G.E. Glass (1990). Factors influencing mercury
concentrations in walleyes in northern Wisconsin lakes. Trans. Amer. Fish. Soc. 119:862-870.
Wiener, J.G. and D.J. Spry (1995). Toxicological significance of mercury in freshwater fish. In:
Interpreting Environmental Contaminants in Animal Tissues. G. Heinz and N. Beyer (eds.), Lewis
Publishers, Chelsea, MI. (in press).
Winfrey, M.R. and J.W.M. Rudd (1990). Review - Environmental factors affecting the formation of
methylmercury in low pH lakes. Environ. Toxicol. Chem. 9:853-869.
Wobeser, G. (1973). Aquatic mercury pollution: studies of its occurrence and pathologic effects on
fish and mink. Ph.D. dissertation, University of Saskatchewan, Saskatoon, Saskatchewan, Canada.
Wobeser, G. and M. Swift III (1976). Mercury poisoning in a wild mink. J. Wildl. Diseas. 12:335-
340.
June 1996 7-13 SAB REVIEW DRAFT
-------
Wobeser, G., N.D. Nielsen and B. Schiefer (1976a). Mercury and mink I: The use of mercury
contaminated fish as a food for ranch mink. Can. J. Comp. Med. 40:30-33.
Wobeser, G., N.D. Nielsen and B. Schiefer (1976b). Mercury and mink II: Experimental methyl
mercury intoxication. Can. J. Comp. Med. 40:34-45.
World Health Organization (WHO, 1989). Environmental health criteria 86: Mercury - environmental
aspects.
Wren, C.D. (1985). Probable case of mercury poisoning in a wild otter, Lutra canadensis, in
northwestern Ontario. Can. Field Nat. 99:112-114.
Wren, C.D. (1986). A review of metal accumulation and toxicity in wild mammals. I. Mercury.
Environ. Res. 40:210-244.
Wren, C.D. (1991). Cause-effect linkages between chemicals and populations of mink (Mustela visori)
and otter (Lutra canadensis) in the Great Lakes basin. J. Toxicol. Environ. Health 33:549-585.
Wren, C.D., H.R. MacCrimmon and B.R. Loescher (1983). Examination of bioaccumulation and
biomagnification of metals in a precambrian shield lake. Water Air Soil Pollut. 19:277-291.
Wren, C.D. and H.R. MacCrimmon (1986). Comparative bioaccumulation of mercury in two adjacent
freshwater ecosystems. Water Res. 20:763-769.
Wren, C.D,, P.M. Stokes and K.L. Rscher (1986). Mercury levels in Ontario mink and otter relative
to food levels and environmental acidification. Can. J. Zool 64:2854-2859.
Wren, C.D. and G.L. Stephenson (1991). The effect of acidification on the accumulation and toxicity
of metals to freshwater invertebrates. Environ. Pollu. 71:205-241.
Wright, D.R. and R.D. Hamilton (1982). Release of methyl mercury from sediments: effects of
mercury concentration, low temperature, and nutrient addition. Can. J. Fish Aquat. Sci. 39:1459-1466.
Xun, L., N.E.R. Campbell and J.W.M. Rudd (1987). Measurements of specific rates of net methyl
• mercury production in the water column and surface sediments of acidified and circumneutral lakes.
Can. J. Fish. Aquat. Sci. 44:750-757.
Yannai, S., I. Berdicevsky and L. Duek (1991). Transformations of inorganic mercury by Candida
albicans and Saccharomyces cerevisiae. Appl. Environ. Micro. 57:245-247.
Zelles, L., I. Scheunert and F. Korte (1986). Comparison of methods to test chemicals for side effects
on soil microorganisms. Ecotoxicol. Environ. Safety 12:53-69.
Zillioux, E.J., D.B. Porcella and J.M. Benoit (1993). Mercury cycling and effects in freshwater
wetland ecosystems. Environ. Toxicol. Chem. 12:2245-2264.
June 1996 7-14 SAB REVIEW DRAFT
-------
APPENDIX A
ESTIMATION OF BIOACCUMULATION FACTORS
FOR MERCURY IN FISH
-------
TABLE OF CONTENTS
Page
A.1 INTRODUCTION A-l
A.2 MONTE CARLO ANALYSIS A-l
A.2.1 Sources of Uncertainty and Their Treatment A-l
A.2.2 Probabilistic Simulation Using the Monte Carlo Method A-2
A.2.3 Selection of Distributions for Input Variables A-2
A.3 ESTIMATION OF BAFs FOR MERCURY A-3
A.3.1 Data Quality Objectives A-4
A.3.2 Estimation of BAFs For Mercury Using Methods Presented in Proposed Guidance
for the Great Lakes Water Quality Initiative . .%. A-4
A.3.3 Estimation of a BAF4 for Mercury Using Measured Values for Trophic Level 3
and a Field-Derived Food Chain Multiplier A-22
A.3.4 Specification of a Distribution for BAF4 Directly from Field Data A-23
A.3.5 Selection of Bioaccumulation Factors for Trophic Levels 3 and 4 A-23
A.3.6 Sensitivity Analysis A-25
A.3.7 Interpretation and Discussion of Sensitivity Analysis A-27
A.4 REFERENCES A-30
June 1996 A-i SAB REVIEW DRAFT
-------
LIST OF TABLES
Page
A-l Methylmercury as a Percentage of Total Mercury in Water A-7
A-2 Methylmercury as a Percentage of Total Mercury in Fish A-9
A-3 Predator-Prey Factor for Trophic Level 4 A-13
A-4 Bioconcentration Factor for Inorganic Mercury in Fish A-14
A-5 Bioconcentration Factor for Methylmercury in Fish A-16
A-6 Statistics for BAF3 Calculated Using the GLWQI Methodology A-23
A-7 Statistics for BAF4 Calculated Using the GLWQI Methodology A-23
A-8 Bioaccumulation Factor for Methylmercury in Trophic Level 3 Fish A-20
A-9 Statistics for Field-derived BAF3 A-23
A-10 Statistics for BAF4 Based on BAF3 and PPF4 A-23
A-11 Bioaccumulation Factors for Methylmercury in Trophic Level 4 Fish : A-C2
A-12 Statistics for Reid-derived BAF4 A-23
A-13 Summary of Bioaccumulation Factors for Trophic Levels 3 and 4 A-24
A-14 Sensitivity of BAF4 Estimate to Correlation of Input Variables A-26
A-15 BAF4 Simulations with Alternate PPF4 Distributions A-27
June 1996 A-ii SAB REVIEW DRAFT
-------
LIST OF FIGURES
Page
A-l Input Distribution for MeHgw A-23
A-2 Input Distribution for MeHgT A-23
A-3 Input Distribution for PPF2 A-23
A-4 Input Distribution for PPF3 A-23
A-5 Input Distribution for PPF4 A-23
A-6 Input Distribution for BCFfflg A-23
A-7 Input Distribution for BCF^ A-23
A-8 Distribution of BAF3 Values from the Monte Carlo Simulation
of the GLWQI Methodology A-23
A-9 Distribution of BAF4 Values from the Monte Carlo Simulation
of the GLWQI Methodology -. . A-17
A-10 Distribution of Field-derived BAF3 Values A-23
A-l 1 Distribution of BAF4 Values output from the Monte Carlo
Simulation of the Product of BAF3 and PPF4 A-21
A-12 Distribution of Field-derived BAF4 Values A-23
June 1996 A-iii SAB REVIEW DRAFT
-------
A.1 Introduction
This appendix describes efforts to estimate bioaccumulation factors (BAFs) for mercury in fish.
Following the food chain structure described in Section 3.3 of Volume V, BAFs were estimated for
fish that occupy trophic levels 3 and 4. These values are referred to as BAF3 and BAF4, respectively.
These BAFs serve as inputs to the calculation of wildlife criterion values (WC) and are also used to
characterize human exposure from consumption of contaminated fish. Emphasis was placed on
evaluating uncertainties associated with these values.
Measures of mercury accumulation that are not treated in this appendix include the following:
(1) BAFs for trophic levels 1 and'2; (2) biota-sediment bioaccumulation factors (BSAFs) for trophic
levels 1 - 4; and (3) predator-prey factors (PPFs) for piscivorous wildlife (that is, the concentration of
mercury in piscivorous birds and mammals divided by that of their prey). Information on these
parameters, including summaries of field data from which estimated values can be derived, can be
found in Section 2.3.1 of Volume V.
For the purposes of this analysis, BAFs for mercury are defined as the concentration of
methylmercury in whole fish divided by the concentration of total mercury in filtered water. This
definition is used so that reference dose values (RfDs) for methylmercury fed to wildlife can be related
directly to water concentration. A more common definition of BAF is the total concentration in fish
divided by the total concentration in water; the degree of error introduced by the definition used in this
Report is minimal, as it is generally agreed that 95% or more of the total mercury in fish is
methylmercury (Bloom, 1994).
The methods used to derive BAFs in the Great Lakes Water Quality Initiative (GLWQI) served
as a starting point for the present analysis. The analysis was then extended to include an examination
of field data from which BAFs could be estimated. It was recognized that considerable natural
variability exists with respect of the accumulation of mercury in aquatic food chains. An effort was,
therefore, made to incorporate this variability into the analysis. This was accomplished by using a
Monte Carlo simulation approach. The Monte Carlo simulation method is described in Section A.2.2.
Important modeling assumptions were evaluated by performing a sensitivity analysis, the results of
which are presented in Section A.3.6.
A.2 Monte Carlo Analysis
A.2.1 Sources of Uncertainty and Their Treatment
Models of environmental phenomena must deal with two basic sources of uncertainty. The first
is uncertainty arising from natural variability, such as the size of individuals in a population. The
second is uncertainty around the value of a parameter or variable when it is known that there is a
single value. These two sources of uncertainty are formally referred to as "variability" and
"uncertainty". There is no attempt in the current analysis to separate variability and uncertainty fully
in the current analysis. In this appendix the term "variability" is used in a general context, comprising
both variability and uncertainty. The term "variable" is used to describe model variables treated as
random variates, while the term "parameter" refers to a fixed parameter of the mathematical form of a
specific distribution.
In dealing with the issue of uncertainty, it is important to distinguish between qualitative and
quantitative models. A qualitative analysis can only make descriptive value judgment statements about
the magnitude of the uncertainty or about the general confidence in the model output, such as "high",
"medium" or "low" and cannot address the statistical properties of the model. A quantitative analysis
allows for a more precise expression of the overall variability, is essential for comparing the results of
June 1996 A-l SAB REVIEW DRAFT
-------
different models and is necessary to determine which of the input parameters have the greatest effect
on the model output. The latter procedure, called a sensitivity analysis, allows the model developers
to focus future efforts on the most important aspects of the model and gives the risk assessor or risk
manager valuable perspective for interpreting the results.
A.2.2 Probabilistic Simulation Using the Monte Carlo Method
There are a number of methods for expressing uncertainty in a quantitative fashion, the
discussion of which is beyond the scope of this document; see Morgan and Henrion (1990) for a
description of these techniques. A Monte Carlo simulation approach was used as a means of treating
the variability in the input variables (Rubinstein, 1981). The Monte Carlo method is an iterative
random sampling technique that mathematically combines specified distributions (rather than single
numbers) and allows for the propagation of variability in each input variable throughout the model;
that is, the variability in each input will be reflected in the output Implicit in this treatment is the
assumption that all parameters vary independently of one another. Thus, during a single iteration, a
"high" value for one variable is equally likely to be combined with a "high" or "low" value of a
second variable.
In a probabilistic Monte Carlo approach, a distribution for each variable in a mathematical
equation is prescribed. The equation is then repeatedly evaluated by random drawing of a value from
each distribution for each iteration and placing the result in the output distribution. The process is
typically repeated 10,000 times or more and produces a distribution of values that can be described in
terms of percentiles. The focus of this analysis was on the 50th percentile and the spread of the
distribution between the 5th and 95th percentiles. The central tendency (median or mean) of the output
distribution is primarily dependent on the median values of the input distributions. The spread of the
output distribution will be determined by the spread of each of the input distributions and by the
nature of the mathematical operation applied to each input.
Calculation of a WC value requires that a single BAF value be established for each trophic level
contributing to the analysis. The same is true for estimating human exposure due to ingestion of
contaminated fish. It should be noted, however, that the Monte Carlo approach yields both a mean
and distribution of BAF values. Although mean values were used for the calculation of WC values, it
is useful to characterize the" statistical variation about this mean as it may reflect actual variation in
natural systems. The possible significance of these distributions is discussed in Volume VI.
A.2.3 Selection of Distributions for Input Variables
Input distributions were based on an analysis of published data and data from two unpublished
reports (see Section A.3.I., Data Quality Objectives). In general, an empirical distribution representing
both the central tendency and the extremes of the given data was determined. It was decided to
represent the actual input distributions in parameterized form for this analysis. That is, formal analytic
distributions that are expressed by a mathematical equation (with specific defining parameters) were
assigned to the inputs rather than using the empirical data. This decision was based on the general
scarcity of data and on theoretical considerations. The particular set of parameters chosen for each of
the variables is only one choice of a number of possible choices. The choice of the form, location and
scale for each of the variables is largely a matter of judgment
The general form of the distribution, whether normal, lognormal, beta, or otherwise, was
determined by examination of the shape of a histogram (distribution) of empirical the data and by
consideration of the underlying biological and physical processes.
June 1996 A-2 SAB REVIEW DRAFT
-------
Generally speaking, distributions are characterized by location and scale parameters. The location
of a distribution is its central tendency, either the mean, median or mode. The mean is the average of
all possible values. The median is the value below which 50% of all possible values fall. The mode
is the value that occurs most frequently. The scale parameter generally represents the spread of the
distribution. The standard deviation (s.d.) is the scale parameter for the normal and lognormal
distributions, which are both open-ended; that is, they have no absolute limits. For closed-end
distributions, such as the triangular and beta, scale is represented by the smallest and largest values
that the function can take on. For any given variable, extreme values are generally considered to be
less likely than values near the median. In these cases distributions with higher probabilility densities
in the middle than in the tails (either end of the distribution) is assigned. All distributional forms
except the uniform belong to this class. In some cases, when this assumption is unwarranted, a
uniform distribution, in which all values are equally likely, was assigned.
Many of the variables reflect underlying exponential processes and are distributed as the
logarithm of the nominal values. Hereafter, such variables will be denoted as being distributed in log
space. In these cases, and when the empirical data suggest such a distribution, a lognormal or log-
uniform distribution was chosen to represent the variable. Triangular or beta distributions .were used
when a judgement was made that the value of the variable will fall within identifiable absolute limits.
A triangular distribution, which requires fewer assumptions than a beta, could be used in those cases
where a beta is assigned. The beta distribution, however, allows for greater uncertainty in the central
tendency and is an appropriate alternative for nonsymmetrical triangular distributions (Evans et al.,
1993). The standard beta variate, with limits of 0 and 1 and a range of shapes, is an appropriate
choice for modeling fractions and percentiles.
Determination of the distribution parameters focuses on identifying the median and extreme
percentiles. The location parameter was estimated from the data. Scale was determined as a function
of the number of observations and the observed extreme values. The empirically-determined
percentiles corresponding to the lowest and highest observations, determined by dividing the rank
order of the observation by n + 1 (where n is the total number of observations), were preserved in the
mathematically defined distribution whenever possible. Calculating the percentiles on the basis of n +
1, rather than n, allows for possibility of obtaining more extreme values in larger sample sizes.
The fundamental data unit hi all analyses was defined as the study. That is, each published
study was treated as an independent unit. High and low values from each study were included in the
empirical distribution if reported. Otherwise, the study was represented by the reported mean value.
Monte Carlo simulations were performed on Intel® 486 DX2/66 CPUs in Crystal Ball® (version
3.0) for Excel® (version 4.0) and in S-PLUS® in the MS Windows® (version 3.1) environment.
A.3 Estimation of BAFs for Mercury
BAF values for mercury in aquatic food chains were calculated hi three different ways. The first
method of calculation was identical to that used to support WC development in the GLWQI (U.S.
EPA, 1995) and involved multiplication of a weighted bioconcentration factor (BCF) by appropriate
food chain multipliers (FCMs). This method yields BAFs for trophic levels 3 and 4. The second
method involved estimation of a BAF for trophic level 3 from field data, which was then multiplied by
a predator-prey factor (PPF) for trophic level 4 to yield a BAF for trophic level 4. A BAF for trophic
level 4 was also directly estimated from field data. The results of all three analyses were then
compared.
It is recognized that other, more detailed approaches have been proposed to estimate BAFs for
mercury in aquatic food chains (e.g., the mercury cycling model (MCM), developed by the Electric
June 1996 A-3 SAB REVIEW DRAFT
-------
Power Research Institute; Hudson et al., 1994). Such approaches were considered to be inappropriate,
however, hi view of the general lack of understanding of mercury accumulation and the broad
geographical focus of this Report In particular, it was determined that models requiring calibration to
specified food chains and lake water characteristics were unlikely to yield information that could be
applied with confidence to a different food chain, or in a lake with different biogeochemical
characteristics. Instead, the decision was made to accept that considerable variability exists in mercury
bioaccumulation in fish and to employ statistical methods that treat this variability quantitatively.
A.3.1 Data Quality Objectives
Preference was given to data published in the peer-reviewed literature. In some instances, due to
advances in analytical methods, preference was also given to the most recently published values. An
attempt was made to characterize the data as necessary to permit comparisons to be made between
studies. For example, mean values were estimated, even if the original authors did not do so. Every
effort was made to report BAFs hi terms of both the age and/or size of the fish involved and the
mercury species. In general, BAFs are expressed hi the literature as total mercury in fish divided by
total mercury in filtered water from which the fish were obtained. Exceptions to this rule were noted
where they occured and were included only if they provided other pertinent information (such as, a
predator-prey factor for two adjacent trophic levels).
Field data from two unpublished reports were also included (Suchanek et al., 1993; Parsons and
Bigham, 1994). Both of these studies are notable for their extent and quality and include data
collected from organisms hi all four trophic levels.
A.3.2 Estimation of BAFs For Mercury Using Methods Presented hi Proposed Guidance for the
Great Lakes Water Quality Initiative
A.3.2.1 BAFs Published hi the Proposed Guidance (GLWQI)
BAFs for mercury were estimated to support the development of WC values in the GLWQI
(U.S. EPA, 1993). The approach and assumptions used hi these calculations were subsequently
modified to incorporate new information (U.S. EPA, 1995). The following is a description of the
modified approach.
BAFs were calculated in support of the GLWQI to relate methylmercury concentrations in fish
to total mercury concentrations hi filtered water. The formula for the calculation of the BAF for
trophic level 4 (BAF^) is given in equation 1.
B AF4 = BCFHg x FCM4 x MeHgT (1)
where
BCFjjg is ti16 weighted-average bioconcentration factor (BCF) for total mercury at trophic level
1 and
FCM4 is the food-chain multiplier representing the cumulative biomagnification of mercury
from trophic level 2 to trophic level 4.
MeHgT is the fraction of total mercury in fish flesh that is in the methylated form.
The formula for calculating BCFHg is given in equation 2.
BCFHg = (BCFMHg x MeHgw) + (BCF^ x (1 - MeHgw)) (2)
June 1996 ' A-4 SAB REVIEW DRAFT
-------
where
BCFMHg is the bioconcentration factor for methylmercury at trophic level 1,
BCFIHg is the bioconcentration factor for inorganic mercury at trophic level 1 and
MeHgw is the fraction of total mercury in the water column that is in the methylated form.
The formula for FCM4 is given in equation 3.
FCM4 = PPF2 x PPF3 x PPF4 • (3)
where
PPF2 is the predator-prey factor at trophic level 2 representing the biomagnification of
mercury in zooplankton as a result of feeding on contaminated phytoplankton,
PPF3 is the predator-prey factor for forage fish feeding on contaminated zooplankton, and
PPF4 is the predator-prey factor piscivorous fish feeding on forage fish.
The estimated BAFs for trophic levels 3 and 4 are as follows:
BAF for trophic level 4 = 140,000
BAF for trophic level 3 = 27,900
Several assumptions were made to permit estimation of these values. These assumptions include
the following.
1. 17% of total mercury in the water column exists as methylmercury.
2. 97.5% of total mercury in fish exists as methylmercury.
3. The predator-prey factors (PPFs) for trophic levels 2, 3 and 4 are 2.00, 1.26 and 5.00,
respectively. Thus, the FCM for trophic level 2 is 2.0, that for trophic level 3 is 2.0 x 1.26,
or 2.52, and that for trophic level 4 is 2.0 x 1.26 x 5.0, or 12.6.
4. The mercury concentration at trophic level 1 is determined by the extent to which mercury
bioconcentrates during an aqueous exposure.
5. The BCF for inorganic mercury in aquatic biota is 2,998.
6. The BCF for methylmercury in aquatic biota is 52,175.
BAFs for mercury at trophic levels 3 and 4 were then determined as follows.
1. A weighted BCF for total mercury (inorganic and methyl) at trophic level 1 was calculated,
based on the assumption that 17% of the total mercury in water exists as the methylated
form, and 83% as the inorganic form.
2. The weighted BCF for total mercury at trophic level 1 was multiplied by FCMs for trophic
levels 2, 3 and 4 to obtain BAFs for total mercury.
3. BAFs for total mercury were multiplied by 0.975 to obtain BAFs for methylmercury in fish.
A.3.2.2 Inputs and Assumptions for the Present Analysis
In the present analysis, data used to calculate BAFs for the GLWQI were combined with
additional information from both published and unpublished sources. Input variable distributions are
presented individually along with the data from which they were derived. In addition, several of the
assumptions made to support the WC calcuation in the GLWQI were evaluated in light of recent
findings.
June 1996 A-5 SAB REVIEW DRAFT
-------
Mercury Speciation in Water
Variable: MeHgw
Definition: Percent of total mercury in water existing as the methylated form
Units: %
Technical Basis:
Kudo et al. (1982) reported that methylmercury in water from two Japanese rivers constituted
"around 30%" of total mercury (range of 26% to 46%). Similar levels were found in three rivers in
Canada and Japan. The two Japanese rivers were known to be polluted by mercury from point source
discharges.
A value of < 10% was reported by Parks et al. (1989) at sampling sites downstream of a chlor-
alkali plant on the Wabigoon-English River Lake system.
Organomercury as a percentage of total mercury was reported to range from 1 to 89% in twelve
lakes and eight rivers located throughout the United States (Gill and Bruland, 1990). Paniculate
mercury comprised a high, but variable (10% - 92%), percentage of the total mercury present but was
not divided into mercury species.
Bloom and Effler (1990) reported that methylmercury constituted from 3.6% to 27.3% of total
mercury in samples from Onondaga Lake in New York State, which is known to have been polluted
with mercury by a chlor-alkali plant Values are for unfiltered water, although values for filtered
water are also available and yield similar or higher percentages for methylmercury. Data varied
considerably with season and depth indicating the presence of a dynamic system. Additional data
from other sources were presented for comparison, but they were not presented in a manner that
allows calculation of maximum values. Based upon these data, it appears that methylmercury as a
percent of total mercury ranges up to 10% in both polluted and pristine waters.
Based upon their analysis of surface waters from Sweden, Lee and Hultberg (1990) concluded
that the percentage of organic-bound mercury that is methylmercury is relatively low (6% to 13%).
The percentage of total mercury existing as methylmercury in these water samples varied from 3.6% to
5.3%.
Methylmercury constituted from 5% to 12% of total mercury in surface water samples obtained
from Little Rock Lake (Watras and Bloom, 1992). The lower figure corresponds to an untreated
reference basin, the higher figure to an experimentally acidified treatment basin.
Summary:
Speciation data for mercury in water remain relatively scarce, and comparisons are difficult to
make because the different mercury species tend to be operationally defined. This reflects the fact that
methods for measuring the different mercury species and the interpretation of these values are in a
state of evolution. This, in turn, impacts the derivation of BAF values for methylmercury, since it is
not always clear which of these operationally defined measurements be used in calculation.
Speciation data from Gill and Bruland (1990) suggest that the range of possible methylmercury
values may be large, although there is a question of how much organomercury exists as the methylated
form. Gill and Bruland (1990) assumed that virtually all the organic mercury fraction is
methylmercury. Actual measurements of methylmercury in this fraction range from 6-56% (Lee and
Hultberg, 1990; Meili et al., 1991). In a majority of published studies, methylmercury as a percentage
June 1996 A-6 SAB REVIEW DRAFT
-------
of total mercury in water is reported to range from 5% to 15%. Thus for many systems, the figure
(17%) used in the GLWQI may be too high. It may also be important to distinguish between surface
and subsurface waters. Measurements made by Bloom and Effler (1990) suggest that methylmercury
levels in surface waters may be considerably lower than those at depth and would, therefore, not
accurately reflect the situation that exists throughout a water body.
The values defining the empirical distribution for MeHgw are given in Table A-l. The beta
distribution was chosen as best representative of this highly skewed closed-end distribution. The
probability density function is graphically illustrated in Figure A-l. Figure A-l plots probability
density on the y:axis against the the values of MeHgw (in %) on the x-axis.
Table A-l
Methylmercury as a Percentage of Total Mercury in Water
Values
3.6, 5.3
5,12
8, 10
3.6, 27.3
3.5, 50
26,46
26,51
Reference
Lee and Hultberg, 1990
Watras and Bloom, 1992
Parks et al., 1989
Bloom and Effler, 1990
Gill and Bruland, 1990
Kudo et al., 1982
Meili et al., 1991
Figure A-l
Input Distribution for MeHgw
protdm
10
SSMeHg
60
June 1996
A-7
SAB REVIEW DRAFT
-------
Mercury Speciation in Fish Tissues
Variable: MeHgT
Definition: Percentage of total mercury in fish tissues existing as the methylated form.
Units: %
Technical Basis:
Based upon their review of data obtained in earlier studies, Huckabee et al. (1979) concluded
that methylmercury constitutes from 80% to 95% of total mercury in fish.
Hildebrand et al. (1980) obtained an average value of 91.7% in muscle from 86 rock bass and
hog suckers (range from 78.9% to 103.8%). The authors also cite numerous older references
corroborating this value.
Values ranging from 59% to 96% were reported by Cappon and Smith (1981) hi muscle tissue
from seven species of freshwater fish, including walleye and northern pike, small and largemouth bass,
yellow perch, bullhead and muskellunge; however, the mean of these values tended toward the high
end of the range. Literature cited by the authors suggests that methylmercury values in marine fish
tend to vary somewhat more (38.4% - 92.9%, with numerous values around 60%) than in their
freshwater counterparts; total mercury concentrations were similar. Additional data cited from several
sources suggest that methylmercury comprises > 95% of the total hi northern pike from Finland and
Sweden.
Paasivirta et al. (1981) reported values ranging from 82% to 84% in muscle tissue from northern
pike.
Methylmercury as a percentage of total mercury was reported to range from 79.2% to 94.8% in
pike and roach from four lakes in Finland (Paasivirta et al., 1983).
Baluja et al. (1983) reported values ranging from 75% to 94% hi sand smelt, carp and eel, with
the mean tending toward the high end of the range.
Values ranging from 62.9% to 79.2% were measured by Cappon (1984) in muscle tissues from
coho and Chinook salmon, and brown and lake trout from Lake Ontario.
Methylmercury was reported to comprise 99% of total mercury in muscle tissue of yellow perch,
northern pike and white suckers from lakes hi northern Wisconsin and the upper peninsula of
Michigan (Grieb et al., 1990).
Jackson (1991) reported that methylmercury constitutes 80% to 90% of total mercury in muscle
tissue from several species of fish in lakes and reservoirs hi Manitoba.
Bloom (1992) reported that virtually all (>95%) of the mercury in muscle tissues from
largemouth bass, yellow perch, northern pike and white suckers existed as methylmercury. The author
suggested that lower values reported in earlier literature are probably erroneous due to inadequate
sampling and analytical techniques.
June 1996 A-8 SAB REVIEW DRAFT
-------
Summary:
Because of continued refinement in mercury analysis methods, more confidence should be placed
in recent values (see Bloom (1992) for a discussion of factors that can result in lower estimates than
are actually present). Collectively, the most recent values suggest that the percentage of mercury in
fish that exists as the methylated form is close to and perhaps even exceeds 95%. Minor differences
in reported values may be due the different types of samples evaluated (e.g., whole fish vs. skin-on
fillets vs. skin-off fillets) but are unlikely to be important in the calculation of a BAF. The values
used to define MeHgT are given in Table A-2. The beta distribution was chosen as best representative
of this highly skewed closed-end distribution. The probability density function for this distribution is
shown in Figure A-2.
Table A-2
Methylmercury as a Percentage of Total Mercury in Fish
Value
71
83
85
87
87
90
92
95
99
t
Reference
Cappon, 1984
Paasivirta et al., 1981
Jackson, 1991
Cappon and Smith, 1981
Paasivirta et al., 1983
Baluja et al., 1983
Hildebrand et al., 1980
Bloom, 1992
Grieb et al., 1990
Figure A-2
Input Distribution for MeHgT
June 1996
A-9
SAB REVIEW DRAFT
-------
Predator-Prey Factor for Trophic Level 2
Variable:
PPF,
Definition: Factor by which mercury concentrations in trophic level 2 organisms exceed those in
trophic level 1 organisms upon which they prey.
Units: Unifless
Technical Basis:
Estimates of PPF2 and PPF3 used to derive BAFs in the GLWQI were based on a single study
(Watras and Bloom, 1992) in which the concentration of total mercury was measured in phytoplankton
(trophic level 1), zooplankton (trophic level 2) and age-1 (year old) yellow perch (trophic level 3).
For the present analysis, a distribution of PPF2 values was developed based on the logs (base 10) of
^ single high and low values of 1.2 and 3.07, defining the 33rd and 67th percentiles, respectively. These
values were calculated from the reported low and high Hg concentrations of 36 and 92 ug/g,
respectively, in zooplankton divided by the single reported value T)f 30 ng/g in phytoplankton (Watras
and Bloom, 1992). The uniform distribution was chosen because there is no reason to believe that
either of the reported values for zooplankton are more likely to occur. The probability density
function for PPF2 is shown in Figure A-3.
Figure A-3.
Input Distribution for PPF2
(U
pnt dm
0.47
O.S
W
778
5.0 100
PPF
Predator-Prey Factor for Trophic Level 3
Variable:
PPF,
Definition: Factor by which mercury concentrations in trophic level 3 organisms exceed those in
trophic level 2 organisms upon which they prey.
Units: Unitless
June 1996
A-10
SAB REVIEW DRAFT
-------
Technical Basis:
A single value of about 1.2 was calculated for PPF3 from Figure 3 in Watras and Bloom (1992),
by dividing the visually-interpolated BCF for yellow perch by that for zooplankton. The form and
relative scale of this distribution was assumed to be same as that for PPF2. The probability density
function for PPF3 is shown hi Figure A-4.
Figure A-4.
Input Distribution for PPF3
0.4
proton
0.1 .
03 0.4 0.6 0.8
tat
IS 3d 4.0 SO
PPF
Predator-Prey Factor for Trophic Level 4
Variable: PPF4
Definition: Factor by which mercury concentrations in trophic level 4 organisms exceed those in
trophic level 3 organisms upon which they prey.
Units: Unitless
Technical Basis:
PPF4s ranging from 2.4 to 7.5 were estimated for "standardized" lake trout (60 cm) and rainbow
smelt (15 cm) from nine Ontario lakes (MacCrimmon et al., 1983). Values from very old (20+ years)
lake trout from Tadenac Lake exceeded those of age 2+ year-old rainbow smelt by a factor of 12.3.
Levels in trout appeared to increase dramatically when they became large enough (about 6 years old)
to switch from a diet of benthic invertebrates to smelt
PPF4s ranging from 1.2 to 8.4 were calculated from data reported by Wren et al. (1983). These
estimates were computed by dividing the average values for three predators (smallmouth bass, northern
pike, and lake trout) by average values for two forage fish (bluntnose minnow and rainbow smelt).
The maximum value was obtained by dividing the value for pike by that for minnows.
Data presented by Mathers and Johansen (1985) were used to calculate PPF4s of 5.9 and 4.9 for
northern pike and walleye, respectively. Each value was calculated by dividing the mercury residue in
eight-year-old fish by the weighted average mercury content of the diet for each species.
June 1996
A-ll
SAB REVIEW DRAFT
-------
Corresponding values for four-year-old fish were 2.8 and 2.2, respectively. Values for both species
tended to increase with fish age and in some very old walleye exceeded 10.0.
PPF4s ranging from 2.7 to 15.1 were computed by dividing the average mercury residues in two
predators (northern pike and brown trout) by average values for two forage fish (whitefish and smelt)
(Skurdal et al., 1985). The maximum value was obtained by dividing the mercury level in brown trout
by that in whitefish; however, brown trout were reported to leave this lake system to forage in the
ocean, thus complicating the comparison. The maximum PPF4 for pike was 4.46 (pike -f whitefish).
PPF4s from 3.0 to 3.2 were calculated from data presented by Wren and MacCrimmon (1986).
Concentrations in "standardized" northern pike were divided by those in "standardized" yellow perch.
The range represents values calculated for two adjacent freshwater systems.
A PPF4 of 2.5 was obtained from average values for 1+ year-old pike and 1+ year-old yellow
perch collected from several locations on the English/Wabigoon system (Parks, 1988). This value may
be lower than others calculated for similar systems due to the small size of the pike sampled (mean =
26 cm).
A PPF of 6.3 was calculated from data presented by Cope et al. (1990). Data for age 5 walleye
were regressed against data from age 2 yellow perch. All fish were collected from northern Wisconsin
seepage lakes. The PPF was calculated from the regression equation for a perch containing 0.1
micrograms/g of total mercury. It should be noted that in this study mercury levels in muscle from
walleye were compared with whole-body levels in perch.
Data collected by Sorenson (1990) from northern Minnesota lakes suggest that the PPF across
two trophic levels (1 kg northern pike to zooplankton) is approximately 5. Assuming the same
increase between trophic levels, the increase from trophic level 3 to trophic level 4 would be 2.25.
Residue data given by Jackson (1991) was used to calculate PPF4s ranging from 5.2 (average of
walleye and pike/shiners) to 15.5. Estimates were computed for two lakes in Manitoba by dividing the
average values for two predators (northern pike and walleye) by average values for two forage fish
(yellow perch and spottail shiners). The maximum value of 15.5 was obtained by dividing the value
for pike by that for perch. The maximum value for walleye was 12.0 (walleye -r perch).
A value of 6.8 was obtained by regressing data for 1 kg northern pike against that from 8 to 10
cm yellow perch (Lindqvist, 1991). The data are from 43 lakes and are remarkable for the consistency
of the relationship.
A PPF4 of 7.4 was calculated by comparing 1 kg northern pike with 5 to 10 g yellow perch
(Meili et al., 1991).
Average concentrations of total mercury were calculated for largemouth bass and shiners in
Clear Lake, California (Suchanek, 1993). The PPF4 estimated from these data was 5.4. Interestingly,
chemical analysis on the same samples suggests that the PPF4 for methylmercury is considerably
higher (17.7). The reason for this discrepancy is not presently known.
Summary:
Predator-prey factors reflecting the increase in mercury concentration between trophic levels 3
and 4 range from 1.2 to 15.1; the mean is close to 5.0. Interpretation of predator-prey factors is
complicated by the fact that piscivorous fish accumulate mercury throughout their lifetime; thus,
calculation of this value for a given species and system depends to a large extent upon the age of the
June 1996 A-12 SAB REVIEW DRAFT
-------
fish sampled. In addition, it is well known that the diet of a piscivorous fish changes with age,
tending in many cases to be dominated by invertebrates until fish reach a critical size that allows them
to prey efficiently upon small fish. In general, therefore, the mercury concentration in prey of a
piscivorous fish can be expected to increase with the age or size of the predator. Additional
considerations, including sexual reproduction, prey selection and availability and seasonal changes in
bioenergetics due to changes in water temperature are also likely to be important determinants of
bioaccumulation.
Overall, it can be shown that for many, if not most, piscivorous fish, mercury concentrations
increase in a nearly linear fashion with age. For a given system, therefore, it may be possible to
extrapolate data for small predators to larger members of the same species. Furthermore, the statistical
distribution of BAFs determined from a random sampling of a population of predators would be
expected to reflect the relative abundance of the different size classes and would have an upper limit
reflecting the fact that fish have finite life spans.
The values defining the empirical distribution for PPF4 are given in Table A-3. Five of the
twelve studies permitted a calculation of both low and high values; the high values pertain to older
fish. Both the low and high values from these studies were included in the analysis along with
average values reported in the other seven studies. The PPF, being a ratio of two positive values,
cannot be lower than 0. The distribution was truncated at the practical minimum of 1 because fish do
not appear to be capable of eliminating mercury at any appreciable rate. The probability density
function for this distribution is shown in Figure A-5. The beta function was chosen because the
distribution is bounded at the lower end, and the general shape of the distribution matches that of the
empirical distribution more closely than a lognormal form does. The upper limit of 54 was not
predetermined but is a result of preserving the empirically-determined percentiles of the maximum
observed values (15.1-15.5 at the 91st percentile) in the closed-form (beta) representation.
Table A-3
Predator-Prey Factor for Trophic Level 4
Value
2.25
2.5
3.1
2.7, 15.1
2.2, 11
2.4, 12.3
1.2, 8.4
5.4
6.3
6.8
7.4
4.5, 15.5
Reference
Sorenson et al., 1990
Parks, 1988
Wren and MacCrimmon, 1986
- Skurdal et al., 1985
Mathers and Johansen, 1985
MacCrimmon et al., 1983
Wren et al., 1983
Suchanek et al., 1993
Cope et al., 1990
Lindqvist, 1991
Meili et al., 1991
Jackson, 1991
June 1996
A-13
SAB REVIEW DRAFT
-------
Figure A-5
Input Distribution for PPF4
pnbdn
fi
PPf
30
Bioconcentration Factor for Inorganic Mercury in Fish
Variable:
BCF
fflg
Definition: Total mercury concentration in fish divided by that in water following a waterborne
exposure to inorganic mercury.
Units: Unitless
Technical Basis:
Snarski and Olson (1982) reported a BCF of 4994 for fathead minnows exposed to mercuric
chloride for 287 days.
A BCF of 1800 was measured in rainbow trout fry exposed to mercuric chloride for 60 days
(Boudou and Ribeyre, 1984).
Ribeyre and Boudou (1984) obtained a BCF of 7000 in rainbow trout exposed-to mercuric
chloride for 30 days.
Summary:
Laboratory exposures (water-only) conducted using small fish yielded values ranging from 1,800
to 7,000. The BCF used to support BAF development in the GLWQI was based on two of the three
values listed above (Snarski and Olson (1982) and Boudou and Ribeyre (1984)). Importantly, the data
collected in these studies did not permit an evaluation of progress to chemical steady-state. It is
possible, therefore, that if exposure times were extended the fish would have continued to accumulate
mercury, thereby resulting in higher BCF values. In addition, exposures conducted with fast-growing
fish have the potential to yield concentration estimates that are lower than those which would have
been observed in fish that were not growing, due to the effect of growth dilution. The values defining
the empirical distribution for BCFIH are given in Table A-4. These values are assumed to be
distributed in log space and are assigned a lognormal distribution. The probability density function for
this distribution is shown in Figure A-6.
June 1996
A-14
SAB REVIEW DRAFT
-------
Table A-4
Bioconcentration Factor for Inorganic Mercury in Fish
Value
1,800
4,994
7,000
Reference
Boudou and Ribeyre, 1984
Snarski and Olson, 1982
Ribeyre and Boudou, 1984
Figure A-6.
Input Distribution for BCFIHg
pn**m
0.20
;<>•«.
OJB.
00 .
togttKF
Bioconcentration Factor for Methylmercury in Fish
Variable:
BCF,
v oiiauit. ^ MHe
Definition: Total mercury concentration in fish divided by that in water following a waterborne
exposure to methylmercury.
Units: Unitless
Technical Basis:
BCFs ranging from 44,130 to 84,670 were reported by Olson et al. (1975) in fathead minnows
exposed to methyl mercuric chloride for 336 days. Interestingly, BCF values appeared to vary
inversely with test concentration.
McKim et al. (1976) obtained BCFs ranging from 10,000 to 33,333 in brook trout exposed to
methyl mercuric chloride for 273 days. Methylmercury concentrations in muscle were essentially the
same as those in whole body and were directly proportional to water concentration. A kinetic analysis
showed that although mercury concentration in tissues tended toward a "steady-state" value, the total
amount in tissues continued to increase. Progress to "steady-state", therefore, reflected growth dilution
June 1996
A-15
SAB REVIEW DRAFT
-------
more than an actual decline in uptake. In fact, uptake remained constant or increased throughout the
exposures.
Ribeyre and Boudou (1984) reported a BCF of 36,000 in rainbow trout exposed to methyl
mercuric chloride for 30 days. A somewhat lower BCF of 11,000 was reported by Boudou and
Ribeyre in rainbow trout fry exposed for 60 days, possibly due to growth dilution in the smaller, faster
growing fish.
A BCF of 85,700 was reported for rainbow trout exposed for 75 days (Niimi and Lowe-Jinde,
1984).
Summary:
BCF values reported in the literature range from 11,000 to approximately 86,000. It has been
argued that because the BCF appears to vary inversely with water concentration, priority should be
given in each study to the BCF calculated for the lowest level tested. This is reasonable, since, in
general, laboratory exposure concentrations tend to be higher than those measured in the environment.
For this reason, it has also been suggested that a regression equation be applied to extrapolate to an
"environmentally relevant" value. Existing data are not sufficient, however, to support such an
extrapolation due both to a lack of data and to differences in exposure duration, species, etc. There is
also a question as to whether chemical steady-state conditions ever occurs in fish exposed in a
laboratory setting. Data collected from the two studies of longest duration (McKim et al. (1976) and
Olson et al. (1975)) suggest that the appearance of a constant concentration during the latter phase of
each study was due to growth dilution and not to a steady-state condition. The values defining the
empirical distribution for BCF^g are given in Table A-5. These values were assumed to be
distributed in log space and are assigned a lognormal distribution. The probability density function for
this distribution is shown in Figure A-7.
Table A-5
Bioconcentration Factor for Methylmercury in Fish
Value
11,000
33,333
36,000
84,670
85,700
Reference
Boudou and Ribeyre, 1984
McKim et al., 1976
Ribeyre and Boudou, 1984
Olson et al., 1975
Niimi and Lowe-Jinde, 1984
June 1996
A-16
SAB REVIEW DRAFT
-------
Figure A-7
Input Distribution for BCFMHg
06
0.4
00 .
4.0
4.5
5J1
5.5
toglOBCF
Geometric Increase in Food Chain Multiplier
A geometric increase in the food chain multiplier occurs when the predator-prey factors at each
trophic level are approximately equal. This assumption was part of the original set of assumptions
made to calculate a BAF for mercury in the GLWQI (U.S. EPA, 1993), but was later dropped in favor
of incorporating measured PPFs (U.S. EPA, 1995). An attempt was made to evaluate the validity of
this assumption on the basis of existing literature.
Technical Basis:
Lindqvist (1991) reported that mercury concentrations in pike (trophic level 4) were 6.8 times
those in perch (trophic level 3) and 25 times those in zooplankton (trophic level 2). Predator-prey *
factors of 6.8 for trophic level 4 and 3.7 for trophic level 3 were calculated from this study.
Data presented by Jackson (1991) suggested that methylmercury concentrations in walleye and
pike were about 5.2 times those of shiners, while total mercury concentrations in shiners were about
twice those of zooplankton.
Data published by Watras and Bloom (1992) suggested that the PPF between trophic levels 1
and 2 was approximately two, while total mercury concentrations in age-1 yellow perch were only
slightly higher (approximately 1.25 times) than those in zooplankton.
Total mercury data from Clear Lake, CA (Suchanek et al., 1993) suggested that the PPF for
trophic level 4 (largemouth bass) was 5.4, while that for trophic level 3 (silversides) was about 3.6. In
contrast residue concentrations in zooplankton (trophic level 2) were similar to or perhaps even
slightly lower than those in phytoplankton (trophic level 1) suggesting a PPF for trophic level 2 of
approximately 1.0. A separate analysis of methylmercury residues suggested that PPFs increased from
about 2.2 between trophic levels 1 and 2, to about 20 between trophic levels 3 and 4. This resulted in
an exponential increase in the food chain multiplier and was apparently due to differences in the
percentage of total mercury existing as methylmercury at each trophic level. In general, the percentage
June 1996
A-17
SAB REVIEW DRAFT
-------
of total mercury existing as methylmercury increased with trophic level, ranging from about 50% at
trophic level 1, to 100% at trophic levels 3 and 4.
PPFs reported by Parsons and Bigham (1994) declined with successively higher trophic level;
the PPF between trophic levels 1 and 2 was approximately 8.0, that between trophic levels 2 and 3
was approximately 5.8, while that between trophic levels 3 and 4 was about 1.8.
Summary:
The few data that can be used to evaluate this assumption do not support a scientific consensus.
In some systems the PPF for both total and methylmercury appears to increase with trophic level (eg.,
Undqvist, 1991; Jackson, 1991; Suchanek et al., 1993), while in others the PPF tends to decrease
(Watras and Bloom, 1992; Parsons and Bigham, 1994).
Each of the studies examined supports the conclusion that mercury is biomagnified in aquatic
food chains. There is considerable uncertainty, however, concerning the extent to which this occurs,
particularly at the_lower trophic levels. This uncertainty has a large potential to impact subsequent
calculation of BAFs due to the multiplicative manner in which PPFs are used.
A.3.2.3 Results of Monte Carlo Analysis of GLWQI Methodology
The distribution for BAF3 generated from the Monte Carlo simulation of the GLWQI
methodology is shown in Figure A-8, with selected statistics given in Table A-6. The direct output of
the distribution is given as the log (base 10) of the statistic because the distribution is based on the
logarithms of the original observations. The quantities in the 'Value' column are the antilogs of the
distribution output
Figure A-8
Distribution of BAF3 Values from the Monte Carlo Simulation of the GLWQI Methodology
20000 iterations
2000
1500
frequency
1000
SOD
toglOBAF
June 1996
A-18
SAB REVIEW DRAFT
-------
Table A-6
Statistics for BAF3 Calculated Using the GLWQI Methodology
Statistic
Value (loglO) Value
Mean
Standard Deviation
4.437
0.644
27,360 (GM)
4.41 (GSD)
Percentiles:
5th
25th
50th
75th
95th
3.363
4.000
4.446
4.879
5.488
2,307
10,040
27,930
75,750"
307,530
GM = geometric mean; GSD = geometric standard deviation
The distribution for BAF4 generated from the GLWQI methodology is shown in Figure A-9,
with selected statistics given in Table A-7. The direct output of the distribution is given as the log
(base 10) of the statistic as the distribution is based on the logarithms of the original observations.
The quantities in the 'Value' column are the antilogs of the distribution output.
Figure A-9
Distribution of BAF4 Values from the Monte Carlo Simulation of the GLWQI Methodology
20000 Itaratlons
2000
1500
frequency
1000
500.
toglOBAF
June 1996
A-19
SAB REVIEW DRAFT
-------
Table A-7
Statistics for BAF4 from the GLI Approach
Statistic
Value (loglO) Value
Mean
Standard Deviation
5.134
0.720
136,070 (GM)
5.25 (GSD)
Percentiles:
5th
25th
50th
75th
95th
3.942
4.647
5.135
5.635
6.316
8,757
44,330
136,540
431,750
2,069,200
A.3.3
GM = geometric mean; GSD = geometric standard deviation
Estimation of a B AF^ for Mercury Using Measured Values for Trophic Level 3 and a Field-
Derived Food Chain Multiplier
In this analysis a BAF for trophic level 4 was calculated using a BAF value derived from field
data for trophic level 3 and published predator-prey factors for trophic level 4. The distribution of
BAFs for trophic level 3 is presented with the data from which it was derived. The distribution of
predator-prey factors for trophic level 4 was defined previously in Section A.3.2.2. In contrast to the
GLWQI methodology, the simplifying assumption was made that 100% of total mercury in fish exists
as methylmercury. This decision was based upon a review of published (see Section A.3.2.2) and
unpublished data (Bloom, 1994). Here, as is the case throughout the document, BAF3 and BAF4 refer
to the concentrations of methylmercury in fish divided by the concentration of total mercury in filtered
water.
A.3.3.1 Bioaccumulation Factors Determined From Field Data - Total Mercury in Forage
Fish
Variable: BAF3
Definition: Methylmercury in forage fish (trophic level 3) divided by total mercury in filtered
water, accumulated by all possible routes of exposure.
Units: Unitless.
Technical Basis:
BAF3s ranging from 25,000 to 68,000 were reported by Glass et al. (1993) for yellow perch
ranging in size from 3.8 to 10.5 cm. Data were obtained from 215 individuals collected from Crane
and Sand Point Lakes (MN) and were averaged across season and site. BAF3s ranging from 1,400 to
35,000 were measured in a variety of forage fish collected from the St. Louis River estuary. These
data were not included in the analysis, however, because high mercury levels in the estuary (due to an
emissions point source) were thought to result in anomalously low BAF3 estimates.
Watras and Bloom (1992) reported BAF3s for year-1 yellow perch ranging from approximately
63,000 to 250,000. The fish were collected from Little Rock Lake (WI), which has been the subject
June 1996
A-20
SAB REVIEW DRAFT
-------
of investigations concerning the effects of experimental acidification. The lower figure corresponds to
perch collected from the reference basin (pH of 6.1), while the higher value is for fish collected from
the acidified basin (pH of 4.7). Most fish will not survive if pH is reduced to a value much less than
that of the acidified basin of Little Rock Lake. Lakes approaching this value, and which contain
viable populations of perch and other fish, are common in areas that are impacted by acid
precipitation, including large portions of the upper midwest and northeast. Such fish populations are
likely to contribute substantially to the diet of piscivorous wildlife living in these areas. The entire
range of values reported in this study is, therefore, relevant to the present analysis.
Yellow perch (5 to 10 g) collected from 25 lakes in Sweden were shown to have BAF3s ranging
from 17,000 to about 65,000 (mean 40,400; Mieli et al., 1991). These estimates were obtained by
dividing minimum and maximum residue values by the mean water value. Water samples, however,
do not appear to have been filtered before analysis. The total mercury value for water, therefore,
includes an unknown quantity of mercury adsorbed to particulate matter. If the samples had been
filtered, it is likely that total mercury concentrations in water would have declined, resulting in
correspondingly higher BAF3 values.
Parsons and Bigham (1994) reported BAF3s of approximately 28,300, 31,500, and 178,000 for
bluegill, gizzard shad and white perch collected from Onondaga Lake (NY). It can be argued that
large white perch occupy a trophic position intermediate to levels 3 and 4. Unfortunately, the sizes of
the fish sampled were not given. Lacking this information, it was assumed that the animals from
which these values were derived represent the complete range of reported year classes (bluegill and
shad - 0 to 5 years; white perch - 0 to 7 years). The extent of the error introduced by including both
large and small individuals is, therefore, assumed to be small.
BAF3s determined for silversides from Clear Lake (CA) ranged from 83,000 to 483,000 (mean
of 208,400; Suchanek et al., 1993). Silversides are almost entirely planktivorous and form the forage
base for the piscivores (e.g., largemouth bass) in this system.
Summary:
Because of recent advances in analytical techniques for measurement of mercury in natural
waters, consideration was given only to those values reported since 1990. BAFs calculated in earlier
literature tend to be lower due to higher reported water concentrations; see for example, Parks (1988).
BAFs for forage fish (trophic level 3) ranged from 10,000 to nearly 500,000. This range brackets the
BAF3 derived for the GLWQI (27,900; U.S. EPA, 1995), although in three out of five systems the
GLWQI estimate appears to be low.
In defining the distribution, the log transformations of the data points were used because the
untransformed data are highly skewed to the right, and the underlying physical and biological
processes are more likely to be multiplicative than additive. The values defining the empirical
distribution for BAF3 are given in Table A-8 and are the limits of the ranges given under Technical
Basis above. These values are assumed to be distributed in log space and are assigned a lognormal
distribution. The probability density function for BAF3 derived from field data is shown in Figure A-
10, with selected statistics given in Table A-9. The direct output of the distribution is given as the log
(base 10) of the statistic as the distribution is based on the logarithms of the original observations.
The quantities in the 'Value' column are the antilogs of the distribution output.
June 1996 A-21 SAB REVIEW DRAFT
-------
Table A-8
Bioaccumulation Factor for Methylmercury in Trophic Level 3 Fish
Values
25,000; 68,000
63,000; 250,000
17,000; 65,000
28,300; 178,000
83,000; 483,000
Reference
Glass et al., 1993
Watras and Bloom, 1992
Mieli et al., 1991
Parsons and Bigham, 1994
Suchanek et al., 1993
Figure A-10
Distribution of Field-derived BAF3 Values
("•"••
OiS
020.
'o.s.
OJH.
33
4.0
4.5 3.0
BAF
U
SJ)
Table A-9
Statistics for Field-derived BAF3
Statistic
Value (loglO) Value
Mean
Standard Deviation
4.820
0.617
66,170 (GM)
4.14 (GSD)
Percentiles:
5th
25th
50th
75th
95th
3.806
4.405
4.820
5.237
5.835
6,400
25,390
66,170
172,400
684,000
June 1996
A-22
SAB REVIEW DRAFT
-------
CM = geometric mean; GSD = geometric standard deviation
A.3.3.2 Results of Monte Carlo Analysis Using Field-derived BAF3 and PPF4 Estimates
The formula for the calculation of BAF4 by this method is given in equation 4.
BAF4 = BAF3 x PPF4 (4)
where
BAF3 is the field-derived distribution for the BAF at trophic level 3
PPF4 is the predator-prey factor at trophic level 4 representing the biomagnification of mercury in
piscivorous fish feeding on forage fish
The distribution of BAFs calculated by Monte Carlo simulation using field-derived BAF3 values
and PPF4 estimates is shown in Figure A-ll. Table A-10 shows selected statistics from this
distribution. Results are expressed both as the logarithms (base 10) of the value and the BAF value
itself.
Figure A-ll
Distribution of BAF4 Values Output from the Monte Carlo Simulation
of the Product of BAF3 and PPF4
20000 Iterations
2000
1500
frequency
1000
500
tofflOBAF
June 1996
A-23
SAB REVIEW DRAFT
-------
Table A-10
Statistics for BAF4 based on BAF3 and PPF4
Statistic
Value (loglO) Value
Mean
Standard Deviation
5.525
0.703
335,000 (GM)
5.053 (GSD)
Percentiles:
5th
25th
50th
75th
95th
4.356
5.046
5.526
6.001
6.672
22,700
111,000
336,000
1,000,000
4,700,000
GM = geometric mean; GSD = geometric standard deviation
A.3.4 Specification of a Distribution for BAF^ Directly from Field Data
A.3.4.1 Bioaccumulation Factors Determined From Field Data - Total Mercury in Piscivorous Fish
Variable:
BAF,
Definition: Methylmercury in piscivorous fish (trophic level 4) divided by that in water,
accumulated by all possible routes of exposure.
Units: Unitless.
Technical Basis:
Mercury concentrations in standardized (1 kg) northern pike collected from 25 lakes in Sweden
were reported by Meili et al. (1991). These data were divided by the mean concentration of total
mercury in water to obtain BAF4s ranging from 128,000 to approximately 425,500 (mean of 297,900).
As noted in the discussion of B AF3 values, water samples analyzed for these calculations do not
appear to have been filtered. BAF4s calculated on a dissolved total mercury basis are, therefore, likely
to be higher.
BAF4s reported by Suchanek et al. (1993) for largemouth bass ranged from 440,000 to nearly
2,900,000 (mean of 1,161,000). The ages of the bass analyzed were not given; however, their weights
were listed (ranging from 340 g to 4.44 kg) and suggest that a broad range of year classes was
represented.
Mercury residues were computed by Sorenson et al. (1990) for 65 standardized (1 kg) pike
collected from lakes located throughout northern Minnesota. Dividing by the average mercury level in
water from these study sites yielded BAF4s ranging from 56,700 to 615,400 (mean of 178,100).
BAF4s of 123,900 and 249,300 were reported by Parsons and Bigham (1994) for smallmouth
bass and walleye from Onondaga Lake (NY). These estimates were obtained from adult fish
representing a broad range of age classes (walleye - 4 to 10, bass - 4 to 11). Ranges for individual
species were not given.
June 1996
A-24
SAB REVIEW DRAFT
-------
Summary:
BAFs for piscivorous fish (trophic level 4) range from 130,000 to nearly 3,000,000 and, thus,
bracket the BAF4 derived for the GLWQI (140,000; U.S. EPA, 1995); in three out of four systems the
GLWQI estimate appears to be low.
Even though data for trophic level 4 are few, support for these numbers is provided by
combining observations for trophic level 3 and the PPF for trophic level 4. Thus, taking 70,000 as an
estimate for tropic level 3 and multiplying by a PPF4 of 5.0 yields an estim'ated BAF for trophic level
4 of 350,000.
Additional support for BAFs in this range is provided by measured BAFs for methylmercury in
fish and water. For example, Jackson (1991) reported methylmercury BAFs ranging from 3,000,000 to
greater than 9,000,000 in walleye from northern Manitoba. Similarly, Parsons and Bigham (1994)
reported a BAF for methylmercury in walleye of approximately 5,500,000. Dividing these values by
20 (a reasonable figure if about 5% of total mercury in water is assumed to be methylated) yields
BAFs ranging from 150,000 to 450,000.
The values defining the empirical distribution for BAF4 are given in Table A-11 and are the
limits of the ranges given under Technical Basis above. These values are assumed to be distributed in
log space and are assigned a lognormal distribution. The probability density function for this
distribution is shown in Figure A-12. Statistics and percentiles derived from this distribution are given
in Table A-12. Results are expressed both as the logarithms (base 10) of the value and the BAF value
itself.
Table A-11
Bioaccumulation Factors for Methylmercury in Trophic Level 4 Fish
Value
128,000; 425,000
123,900; 249,300
56,700; 615,400
440,000; 2,900,000
Reference
Meili et al., 1991
Parsons and Bigham, 1994
Sorenson et al., 1990
Suchanek et al., 1993
June 1996 A-25 SAB REVIEW DRAFT
-------
Figure A-12
Distribution of Field-Derived BAF4 Values
0.20
O.fl .
pnbdn
OJB.
BAF
Table A-12
Statistics for Field-Derived BAF4
Statistic
Value (loglO) Value
Mean
Standard
Deviation
5.602
0.747
400,000 (GM)
5.585 (GSD)
Percentiles:
5th
25th
50th
75th
95th
4.374
5.099
5.602
6.106
6.831
23,600
125,000
400,000
1,280,000
6,780,000
GM = geometric mean; GSD = geometric standard deviation
A.3.5 Selection of Bioaccumulation Factors for Trophic Levels 3 and 4
Bioaccumulation factors calculated using each of the three previously described methods are
given in Table A-13. Field data for trophic levels 3 and 4 yield mean BAF4 estimates that differ by
less than 20%, while methods employed in the GLWQI yield a mean BAF4 value that is between one-
half and one-third of those derived from field data. No attempt was made to estimate a BAF for
trophic level 3 from field data for trophic level 4.
June 1996
A-26
SAB REVIEW DRAFT
-------
It should be noted that throughout this discussion most of the BAF estimates reported to date
originate from studies of similar types of freshwater systems; that is, those that are at risk due to acid
deposition and that, by virtue of average temperature, etc., support similar fish assemblages (typically
northern pike, walleye, yellow perch, spottail shiners). This has the potential for two problems of
interpretation.
1) Because data from the reported systems tend to be similar, relationships could be
overlooked that would be important under a different set of circumstances.
2) The relative abundance of these data introduce a regional bias into any type of analysis
that is intended for nationwide application.
In this regard, data from Suchanek et al. (1993) are notable because they suggest that in Clear Lake
(CA), BAFs for trophic levels 3 and 4 are considerably higher than those reported by other authors for
yellow perch, walleye, and northern pike. Clear Lake is warm, highly eutrophic and receives
considerable agricultural runoff from nearby orchards and vineyards. Lacking any corroborating
information, it is not possible to determine how representative the Clear Lake data are of warm water
systems generally. Clearly, however, there is a need for additional residue data collected from a
broader spectrum of freshwater systems.
Table A-13
Summary of Bioaccumulation Factors for Trophic Levels 3 and 4
(mean, 5% and 95% values)
Trophic
Level
Recommended
Method
Mean
5th pctl
95th pctl
BAF3
66,200
Field-derived
66,200
6,400
684,000
GLWQI
25,200
2,310
308,000
BAF4
335,000
BAF3x
PPF4
335,000
22,700
4,700,000
Field-derived
400,000
23,600
6,780,000
GLWQI
136,000
8,760
2,070,000
In the Proposed Guidance to the GLWQI it is stated that field-derived BAFs should take
precedence over values estimated using laboratory data, when such data are deemed to be sufficient
(U.S. EPA, 1993; p. 20859). Based upon a review of the foregoing analyses, it was judged that BAFs
calculated from existing field data are more accurate and reflect better the true level of natural
variability than BAFs calculated using the GLWQI methodology. Specifically, the recommendation
that field data for trophic levels 3 and 4 (ie., the BAF3 x PPF4 method) be used to estimate BAFs for
fish was based on the following considerations. The GLWQI method involves several variables and is
dependent upon a number of assumptions, some of which are poorly supported. The direct application
of field data to the estimation of BAF4 yields a value similar to the BAF3 x PPF4 method, but it
results in wide variability about the mean, due to the small number of measured values, one of which
(Suchanek et al., 1993) appears as a relative outlier. The BAF3 x PPF4 approach results in an
intermediate degree of variability about the mean BAF4 estimate, and the inclusion of PPF4 allows for
a better representation of fish size (age) distribution.
June 1996
A-27
SAB REVIEW DRAFT
-------
The mean BAF values estimated using the BAF3 x PPF4 method were, therefore, recommended
in this Report for subsequent calculation of a WC for mercury (Volume V) and for an evaluation of
human exposure due to consumption of contaminated fish (Volume in). These BAFs are the
following.
For trophic level 3, the BAF for mercury was estimated to be 66,200 L/kg.
For trophic level 4, the BAF for mercury was estimated to be 335,000 L/kg.
A.3.6 Sensitivity Analysis
Sensitivity analyses examining the effect of changes in assumptions have been conducted. Three
factors have been varied in this analysis: (1) dependency of one input variable on another; (2)
distributional assumptions for BAF3; and (3) disaggregation of PPF4.
A.3.6.1 Sensitivity of Output to Correlation of Input Variables (Dependency of on Variable
on Another Variable)
Sensitivity analyses of the dependence of the variance (spread) of BAF4 have been performed
with varying assumptions about the correlations between input variables. Both BAF3 and PPF4 depend
on the concentration of methylmercury in trophic level 3 fish ([MeHg3]). The BAF3 is directly
proportional to [MeHg3], and PPF4 is inversely proportional to [MeHg3]. The equation for BAF4,
which is the product of BAF3 and PPF4, thus, has the same term in the numerator and denominator,
the variance or which will be counted twice. As a result, the apparent variance of the BAF4
simulation will be inflated unless the correlation is taken into account. This can be accomplished by
inversely correlating BAF3 and PPF4 in the Monte Carlo simulation. The magnitude of the correlation
depends on the data overlap (defining each variable) and the extent of the contribution of [MeHg3] to
the overall variance of each variable (BAF3 and PPF^. BAF3 and PPF4 are based on the same data
for 3 out of 9 and 2 out of 17 data points, respectively (Meili et al., 1991, and Suchanek et al., 1993
studies). Assuming that the two implicit variables in each of BAF3 and PPF4 contribute equally to the
variance, and that each data point contributes equally to the variance of the simulated distribution, the
overall correlation is unlikely to be greater (more negative) than -0.25, and very likely to be less. The
standard simulation assumed no correlation.
Table A-14 shows the effect of different correlation assumptions on the spread of the BAF4
distribution and on the relative contributions of each variable to the overall variance. The results of
the simulations show that a moderately strong correlation (50%) between BAF3 and PPF4 would have
a significant effect on the spread of the output, reducing it by a factor of 3.6. If the correlation were
weaker (10%), the spread of the distribution would be reduced by only 30%.
June 1996 A-28 SAB REVIEW DRAFT
-------
Table A-14
Sensitivity of BAF4 Estimate to Correlation of Input Variables
Correlation
none
-0.1
-0.25
-0.5
5*
22,700
25,800
30,500
42,300
95th
4,700,000
4,210,000
3,430,000
2,570,000
logic SP^
2.32
2.21
2.05
1.78
Contribution to variance
BAF3
64%
67%
74%
94%
PPF4
36%
33%
26%
6% •
A.3.6.2 Simulation of BAF4 With Disaggregated PPF4 Distributions
Alternate PPF4 distributions were constructed from restricted ranges of the data to represent
specific fish size (age) distributions. The distribution definitions are the following:
Name: mid predator-prey factor for trophic level 4
Label: PPF4M
Form: Uniform {min = 2.5, max = 7.5}
Basis: Standardized pike/yellow perch ratios; n = 4
Description: PPF4M is based on four data points representing two- to four-year-old pike. This
distribution is meant to represent a more homogeneous sample of average size
trophic level 4 fish. The form of the distribution reflects the lack of information
pertaining to the estimation of the relative probability of any given value in this
range. The median of the distribution corresponds to the median of the four data
points. The tails were unweighted.
Name: low predator-prey factor for trophic level 4
Label: PPF4L
Form: Uniform {min = 1, max = 4}
Basis: Lowest six values for PPF4
Description: PPF4L is based on the smallest values for PPF4 representing smaller tier 4 fish.
The rationale is the same as for PPF4M
Name: high predator-prey factor for trophic level 4
Label: PPF4H
Form: Uniform {min =11, max = 17}
Basis: Highest four values for PPF4
Description: PPF4H is based on the values reported for older (larger) tier 4 fish for PPF4.
The rationale is the same as for PPF4M
Table A-15 shows the results of BAF4 Monte Carlo simulations with alternate PPF4
distributions. These simulations represent scenarios in which the consumption patterns of the final
consumer can be more narrowly defined in terms of fish size. No correlation between BAF3 and each
of the alternate PPF4s was assumed for these simulations. The corresponding percentile of the
standard BAF3 output distribution is given for the median values of the alternate scenarios.
June 1996
A-29
SAB REVIEW DRAFT
-------
Table A-15
BAF4 Simulations with Alternate PPF4 Distributions
Distribution
low PPF4
midPPF4
high PPF4
Percentiles
5«h
13,460
27,790
85,160
50th
151,500 (31st)
308,100 (48th)
904,800 (73rd)
95th
1,590,000
3,297,000
9,106,000
A.3.7 Interpretation and Discussion of Sensitivity Analysis
The BAF4 output distribution attempted to simulate the selection of a random tier 4 fish from a
random lake hi a random location. It is meant to be used to estimate the concentration of
methylmercury in such a randomly-selected fish when multiplied by the total mercury (inorganic and
organic combined) concentration in filtered water. The BAF3 performs the same function for trophic
level 3 fish. Because of the large variance hi the distributions, and due to lack of a distinction
between variability and uncertainty, the recommendation was to apply the mean values from each
distribution in the appropriate situations.
Except as discussed below, there were no distinctions, for either distribution, as to size of fish,
lake trophic status, lake pH, absolute mercury concentrations (in fish or water) or relative
methylmercury concentrations hi the water column. The data, however, are heavily biased towards
northern oligotrophic lakes and somewhat towards smaller (younger) fish. There was also no
distinction made between variability and uncertainty in the BAF3 and BAF4 distributions. Thus, it
cannot be determined where natural variability stops and uncertainty starts; the percentiles of the
distribution cannot be interpreted as the likelihood of the true mean value. The bounding distributions
are meant to give a rough idea of uncertainty hi the mean itself but are hypothetical in part. The
actual distribution of the mean is unknown. The bounding distributions do have potential significance
hi real-world scenarios, however. The mid and high PPF4 distributions, for example, represent
estimates of PPF4 for standardized two- to four-year-old and older tier 4 fish (eg., large pike),
respectively. These estimates could be applied in situations where the size of the consumed fish is
known to be at one end of the distribution; for example with recreational anglers, who tend to catch
larger individuals within a given population.
The large amount of variability evidenced by the data and reflected in the output distributions
arises from several identifiable but, as yet, unquantified sources. A primary source of variability in
both BAF3 and PPF4 is the dependence of methylmercury bioaccumulation on the age of the fish. For
example, it has been repeatedly shown that mercury hi fish accumulates throughout the lifetime of the
individual (Scott and Armstrong, 1972; MacCrimmon et al., 1983; Wren et al., 1983; Mathers and
Johansen, 1985; Skurdal et al., 1985, Wren and MacCrimmon, 1986; Sorenson et al., 1990; Jackson,
1991; Gutenmann et al., 1992; Glass et al., 1993, Suchanek et al., 1993; Lange et al., 1993). Reported
BAF values for a given species may, therefore, vary as a function of the ages of the animals
examined. As a result, some researchers have suggested that comparisons between lakes should be
made using "standardized" fish values (for example, a value for a hypothetical 1 kg northern pike),
typically derived by linear regression of residue data collected from individuals of varying size and/or
age (Wren and MacCrimmon, 1986; Sorenson et al., 1990; Meili et al., 1991).
June 1996
A-30
SAB REVIEW DRAFT
-------
Although the fish age factor is probably a major contributor to the variance of PPF4, the
influence of this factor on BAF3 is probably much less. Because of the lesser sensitivity of BAF4 to
PPF4 than to BAF3, the total reduction in variability of the BAF4 by accounting for fish age may not
be large.
Perhaps the greatest source of variability is that of model uncertainty; that is, uncertainty
introduced by failure of the model to account for significant real-world processes. The simple linear
BAF model relating methylmercury in fish to total mercury in water masks a number of nonlinear
processes leading to the formation of bioavailable methylmercury in the water column. Much of the
variability in field data applicable to the estimation of mercury BAFs can be attributed to differences
between aquatic systems. For example, in lake surveys conducted within a relatively restricted
geographic region, large differences can exist between lakes with respect to mercury concentrations in
a given species of fish (see for example Cope et al., 1990; Grieb et al., 1990; Sorenson et al., 1990;
Jackson, 1991; Lange et al., 1993). These observations have led to the suggestion that much of this
variability is due to differences in within-lake processes that determine the percentage of total mercury
t&at exists as the methylated form. Limited data also^ suggest that within a given water body
concentrations of methylmercury are likely to vary with depth and season. Unfortunately, while the
concentration of methylmercury in fish flesh is presumably a function of these varying concentrations,
published BAFs are generally estimated from a small number of measured water values, the
representativeness of which is poorly known.
June 1996 A-31 SAB REVIEW DRAFT
-------
A.4 References
Baluja, G., L.M. Hernandez, J. Gonzalez and R. Clavero (1983). Mercury distribution in an ecosystem
of the "Parque Nacional de Donana," Spain. Bull. Environ. Contam. Toxicol. 30:544-551.
Bloom, N.S (1992). On the chemical form of mercury in edible fish and marine invertebrate tissue.
Can. J. Fish. Aquat. Sci. 49:1010-1017.
Bloom, N.S. (1994). Considerations in the Analysis of Water and Rsh for Mercury. Presented at the
National Forum on Mercury in Rsh, New Orleans, LA, September 27, 1994.
Bloom, N.S. and S.W. Effler (1990). Seasonal variability in the mercury speciation of Onondaga Lake
(New York). Water Air Soil Pollu. 53:251-265.
Boudou A. and F. Ribeyre (1984). Influence of exposure length on the direct bioaccumulation of two
mercury compounds by Salmo gairdnqi (fry) and the relationship between organism weight and
mercury concentrations. Water Res. 18:81-86.
Cappon, C.J. (1984). Content and chemical form of mercury and selenium in Lake Ontario salmon
and trout J. Great Lakes Res. 10:429-434.
Cappon, C.J. and J.C. Smith (1981). Mercury and selenium content and chemical form in fish muscle.
Arch. Environ. Contam. Toxicol. 10:305-319.
Cope, W.G., J.G. Wiener and R.G. Rada (1990). Mercury accumulation in yellow perch in Wisconsin
seepage lakes: Relation to lake characteristics. Environ. Toxicol. Chem. 9:931-940.
Evans, M., N. Hastings and B. Peacock (1993). Statistical Distributions. 2nd Edition. John Wiley
and Sons, Inc., New York, NY, pp. 42-44.
Gill, G.A. and K.W. Bruland (1990). Mercury speciation in surface freshwater systems in California
and other areas. Environ. Sci. Technol. 24:1392-1400.
Glass, G.E., J.A. Sorensen, K.W. Schmidt, J.K. Huber and G.R. Rapp, Jr. (1993). Mercury sources
and distribution in Minnesota's aquatic resources: Precipitation, surface water, sediments, plants,
plankton, and fish. Final report to Minnesota Pollution Control Agency and Legislative Commission
on Minnesota Resources, 1989-1991 (Contract Nos. 831479 and WQ/PDS020).
Grieb, T.M., C.T. Driscoll, S.P. Gloss, C.L. Schofield, G.L. Bowie and D.B. Porcella (1990). Factors
affecting mercury accumulation in fish in the upper Michigan peninsula. Environ. Toxicol. Chem.
9:919-930.
Gutenmann, W.H., J.G. Ebel, Jr., H.T. Kuntz, K.S. Yourstone and D.J. Lisk (1992). Residues of
p,p'-DDE and mercury in lake trout as a function of age. Arch. Environ. Contam. Toxicol.
22:452-455.
Hildebrand, S.G., R.H. Strand and J.W. Huckabee (1980). Mercury accumulation in fish and
invertebrates of the North Fork Holston River, Virginia and Tennessee. J. Environ. Qual. 9:393- 400.
Huckabee, J.W., J.W. Elwood and S.G. Hildebrand (1979). Accumulation of mercury in freshwater
biota. In: The Biogeochemistry of Mercury in the Environment. J.O. Nriagu (Ed.), Elsevier/North
Holland Biomedical Press, Amsterdam, The Netherlands, pp. 277-302.
June 1996 A-32 SAB REVIEW DRAFT
-------
Hudson, R.J.M, S.A. Gherini, C.J. Watras and D.B. Porcella (1994). Modeling the biogeochemical
cycle of mercury in lakes: the mercury cycling model (MCM) and its application to the MTL study
lakes. In: Mercury Pollution, Integration and Synthesis. C.J. Watras and J.W. Huckabee (Eds.),
Lewis Publishers, Boca Raton, FL, pp. 473-523.
Jackson, T.A. (1991). Biological and environmental control of mercury accumulation by fish in lakes
and reservoirs of northern Manitoba, Canada. Can. J. Fish. Aquat. Sci. 48:2449-2470.
Kudo, A., H. Nagase and Y. Ose (1982). Proportion of methylmercury to the total amount of mercury
in river waters in Canada and Japan. Water Res. 16:1011-1015.
Lange, T.R., H.E. Royals and Li. Connor (1993). Influence of water chemistry on mercury
concentration in largemouth bass from Florida lakes. Trans. Am. Fish. Soc. 122:74-84.
Lee, Y.H. and H. Hultberg (1990). Methylmercury in some Swedish surface waters. Environ. Sci.
Technol. 9:833-841.
Lindqvist, O. (1991). Mercury in forest lake ecosystems - Bioavailability, bioaccumulation and
biomagnification. Water Air Soil Pollu. 55:131-157.
MacCrimmon, H.R., C.D. Wren and B.L. Gots (1983). Mercury uptake by lake trout, Salvelinus
namaycush, relative to age, growth, and diet in Tadenac Lake with comparative data from other
precambrian shield lakes. Can. J. Fish. Aquat. Sci. 40:114-120.
Mathers, R.A. and P.M. Johansen (1985). The effects of feeding ecology on mercury accumulation in
walleye (Stizostedion vitreum) and pike (Esox lucius) in Lake Simcoe. Can. J. Zool. 63:2006-2012.
McKim, J.M., G.F. Olson, G.W. Holcombe and E.P. Hunt (1976). Long-term effects of
methylmercuric chloride on three generations of brook trout (Salvelinus fontinalis): Toxicity,
accumulation, distribution, and elimination. /. Fish. Res. Ed. Can. 33:2726-2739.
Meili, M., A. Iverfeldt and L. Hakanson (1991). Mercury in the surface water of Swedish forest lakes
- Concentrations, speciation, and controlling factors. Water Air Soil Pollu. 56:439-453.
•
Morgan, M.G. and M. Henrion (1990). Uncertainty. Cambridge University Press, Cambridge,
England.
Niimi, A.J. and L. Lowe-Jinde (1984). Differential blood cell ratios of rainbow trout (Salmo
gairdneri) exposed to methylmercury and chlorobenzenes. Arch. Environ. Contain. Toxicol.
13:303-311.
Olson, G.F., D.I. Mount, V.M. Snarski and T.W. Thorslund (1975). Mercury residues in fathead
minnows, Pimephales promelas Rafinesque, chronically exposed to methylmercury in water. Bull.
Environ. Contain. Toxicol. 14:129-134.
Paasivirta, J., J. Sarkka, K. Surma-Aho, T. Humppi, T. Kuokkanen and M. Marttinen (1983). Food
chain enrichment of organochlorine compounds and mercury in clean and polluted lakes of Finland.
Chemosphere 12:239-252.
Paasivirta, J., J. Sarkka, M. Aho, K. Surma-Aho, J. Tarhanen and A. Roos (1981). Recent trends of
biocides in pikes of the Lake Paijanne. Chemosphere 10:405-414.
June 1996 A-33 SAB REVIEW DRAFT
-------
Parks, J.W. (1988). Selected ecosystem relationships in the mercury contaminated Wabigoon-English
River system, Canada, and their underlying causes. Water Air Soil Pollu. 42:267-279.
Parks, J.W., A. Lutz and J.A. Sutton (1989). Water column methylmercury in the Wabigoon/English
River-Lake system: Factors controlling concentrations, speciation, and net production. Can. J. Fish.
Aquat. Sci. 46:2184-2202.
Parsons, A. and G. Bigham (1994). Comments on the Proposed Water Quality Guidance for the Great
Lakes System. Report presented to the U.S. EPA Great Lakes Task Force, Washington, DC. PTI
Environmental Services , Waltham, MA.
Ribeyre, F. and A. Boudou (1984). Bioaccumulation et repartition tissulaire du mercure - HgC^ et
CH3HgCl - chez Salmo gairdneri apres contamination par voie directe. Water Air Soil Pollu.
23:169-186.
Rubinstein, R.Y..(1981). Simulation and the Monte Carlo Method. John Wiley and Sons, Inc., New
York, NY.
Scott, D.P. and F.A.J. Armstrong (1972). Mercury concentration in relation to size in several species
of freshwater fishes from Manitoba and Northwestern Ontario. J. Fish. Res. Bd. Can. 29:1685-1690.
Skurdal, J., T. Qvenild and O.K. Skogheim (1985). Mercury accumulation in five species of
freshwater fish in Lake Tyrifjorden, southeast Norway, with emphasis on their suitability as test
organisms. Environ. BioL Fish. 14:233-237.
Snarski, V.M. and G.F. Olson (1982). Chronic toxicity and bioaccumulation of mercuric chloride in
the fathead minnow (Pimephales promelas). Aquat. Toxicol. 2:143-156.
Sorensen, J.A., GE. Glass, K.W. Schmidt, J.K. Huber and G.R. Rapp, Jr. (1990). Airborne mercury
deposition and watershed characteristics in relation to mercury concentrations in water, sediments,
plankton, and fish of eighty northern Minnesota lakes. Environ. Sci. Technol. 24:1716-1727.
Suchanek, T.H., P.J. Richerson, L.A. Woodward, D.G. Slotton, L.J. Holts and C.E.E. Woodmansee
(1993). A survey and evaluation of mercury. In: Sediment, water, plankton, periphyton, benthic
invertebrates and fishes within the aquatic ecosystem of Clear Lake, California. Preliminary lake study
report prepared for the U.S. Environmental Protection Agency, Region 9, Superfund Program.
U.S. Environmental Protection Agency (1993). Water Quality Guidance for the Great Lakes System
and Correction; Proposed Rules. Fed. Regist. 58(72):20802-21047 (April 16, 1993).
U.S. Environmental Protection Agency (1995). Great Lakes Water Quality Initiative Technical
Support Document for the Procedure to Determine Bioaccumulation Factors. EPA-820-B-95-005.
U.S. EPA, Office of Water, Washington, D.C.
Watras, C.J. and N.S. Bloom (1992). Mercury and methylmercury in individual zooplankton:
Implications for bioaccumulation. Umnol. Oceanogr. 37:1313-1318.
Wren, C.D. and H.R. MacCrimmon (1986). Comparative bioaccumulation of mercury in two adjacent
freshwater ecosystems. Nat. Resour. 20:763-769.
Wren, C.D., H.R. MacCrimmon and B.R. Loescher (1983). Examination of bioaccumulation and
biomagnification of metals in a precambrian shield lake. Water Air Soil Pollu. 19:277-291.
June 1996 A-34 SAB REVIEW DRAFT
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO.
EPA-452/R-96-001e
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Mercury Study Report to Congress. Draft Submitted to U.S.
EPA's Science Advisory Board. Volume V. An Ecological
Assessment of Anthropogenic Mercury Emissions in the United
States.
5. REPORT DATE
1996
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Dr. John W. Nichols
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Mid-Continent Ecology Division
Office of Research and Development
U.S. Environmental Protection Agency
Deluth, MN 55804
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Director
Mid-Continent Ecology Division
Office of Research and Development
U.S. Environmental Protection Agency
Deluth, MN 55804
13. TYPE OF REPORT AND PERIOD COVERED
Draft. June, 1996.
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
U.S. EPA Project Officer: Martha H. Keating
16. ABSTRACT
In this volume of the draft Mercury Study Report to Congress an ecological assessment for anthropogenic
mercury emissions is developed. The assessment follows the U.S.EPA Framework for Ecological
Assessment: problem formulation; presentation of a conceptual model describing airborne mercury
accumulation in aquatic biota, biomagnification in the aquatic food chain; and analysis of exposure of
wildlife species to methylmercury through consumption of fish and shellfish. Exposures of wildlife species
to methylmercury through the aquatic food chain are compared with toxicity data calculated in the
development of criteria for the protection of fish-eating avian and mammalian wildlife. Descriptions of
mercury impacts on biota are provided in the problem formulation chapter. Estimates are provided of
mercury deposition on a local scale in areas near emissions point sources. The distributions of selected fish-
consuming wildlife species are overlaid with predicted high mercury areas of high concern (e.g., areas with
low-pH surface water) and compared with predicted deposition of anthropogenic mercury emissions. This
volume of the draft Report analyzes sources of variability and uncertainty in these estimates and identifies
data that would strengthen the certainty of these findings.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Mercury; Methylmercury; Mercury Poisoning;
Animals, wild; Wildlife; Bioaccumulate; Aquatic
food chain; Fishes, poisonous; Neurotoxin;
Piscivorous; Toxicology; Indirect exposure..
Air Pollution Control
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (Report)
Unclassified
21. NO. OF PAGES
184 nn.
20. SECURITY CLASS (Page)
Unclassified
22. PRICE
------- |