DEVELOPMENT OF A MARINE MERCURY CYCLING MODEL
FOR PASSAMAQUODDY BAY, NEW BRUNSWICK
by Elsie M.H.A. Sunderland
B.Sc. McGill University, 1997
THESIS SUBMITTED IN PARTIAL FUFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
in the
School of Resource and Environmental Management
O Elsie M.H.A. Sunderland 2003
SIMON FRASER UNIVERSITY
March 2003
All rights reserved. This work may not be reproduced * in whole or in part, by photocopy or other means without the permission of the author. Approval
Name: Elsie M.H.A. Sunderland
Degree: Doctor of Philosophy Title of thesis: Development of a Marine Mercury Cycling Model for Passamaquoddy Bay, New Brunswick Examining Committee: Chair: Dr. Ken Lertzman
-
or, School of Resource and
weah~ewoung Associate Professor, Department of ~iolo# Supervisor
Dr. Brian Branfireun Assistant Professor, Department of Geography University of Toronto Supervisor
Ur. Peter ~flls Research ~clentis&onment Canada Supervisor
Dr. Margo Moore Associate Professor, Department of Biology -SFU Examiner
DY. David ~rabbinhbft Research Hydrologist/Geochemist U.S. Geological Survey External Examiner
Date Approved: PARTIAL COPYRIGHT LICENCE
I hereby grant to Simon Fraser University the right to lend my thesis, project or extended essay (the title of which is shown below) to users of the Simon Fraser University Library, and to make partial or single copies only for such users or in response to a request from the library of any other university, or other educational institution, on its own behalf or for one of its users. I further agree that permission for multiple copying of this work for scholarly purposes may be granted by me or the Dean of Graduate Studies. It is understood that copying or publication of this work for financial gain shall not be allowed without my written permission.
Title of Thesis/Project/Extended Essay
Development of a Marine Mercury Cycling Model for Passamaquoddy Bay, New Brunswick
Author: ABSTRACT
Despite large reductions in mercury emissions from anthropogenic sources, high mercury levels in wildlife, fish and avifauna in the Bay of Fundy region of Maritime Canada remain a problem. The objectives of this research were to: (i) develop a comprehensive
"ecosystem-based" understanding of the environmental fate of anthropogenic mercury in a temperate coastal marine environment; and (ii) formulate an empirical model that links anthropogenic mercury inputs to concentrations in sediments, water and benthic organisms that will aid in developing strategies for reducing the impacts of mercury in coastal ecosystems.
Analysis of sediment cores from the Bay of Fundy region revealed that approximately
65% of present mercury inputs from the atmosphere are composed of natural and recycled anthropogenic mercury. Thus, in the short term, the maximum decline in atmospheric mercury loading achievable through emissions reduction programs is approximately 35%. A field study of mercury speciation showed a proportional relationship between total mercury and methylmercury concentrations in Passamaquoddy
Bay (Bay of Fundy) sediments, but these concentrations are also affected by sediment geochemistry. These relationships indicate that declines in mercury loading should eventually translate into proportional declines in concentrations in organisms if the geochemical characteristics of sediments remain unchanged. Sediment core and composition analysis revealed a deep active sediment layer that provides a large pool of available and actively cycling total mercury and methylmercury. Mass budget calculations show that: (i) the sediment compartment contains the majority (>90%) of the mercury in Passamaquoddy Bay; and (ii) there is a large daily turnover of mercury through methylation and demethylation.
The principal findings of this study show that: (i) environmental changes that shift the present equilibrium between methylation and demethylation could result in rapid accumulation of methylmercury in the sediment compartment that could subsequently be taken up by marine organisms; and (ii) while concentrations in the water column of
Passamaquoddy Bay reach steady state rapidly in response to changes in mercury inputs, the overall dynamics of mercury in this system are governed by the slow rate of change in mercury concentrations in the sediments. These results help to explain why mercury concentrations in marine organisms in the Bay of Fundy region remain elevated despite large emissions reductions. Acknowledgements
The work presented in this thesis would not have been possible without the support and collaboration of many colleagues, friends and family. I would like to thank my senior supervisor, Frank Gobas, for his positive attitude and insights into environmental modeling. Each of my supervisory committee members, Leah Bendell-Young, Brian Branfireun and Peter Wells, provided invaluable guidance for this project and I thank all of them for their ongoing support. In its early stages this work also benefited from the input of Randall Peterman, who for me has always provided the example of how to push for rigorous, groundbreaking science. I would also like to acknowledge the critical role of the departmental staff, Rhonda, Bev, Mary Ann, Anissa and Laurence, in facilitating my completion and helping to maintain my sanity. I received financial support for this work from NSERC, the Gulf of Maine Council on the Marine Environment, Mountain Equipment Co-op, the Department of Fisheries and Oceans, and the Joseph & Rosalie Segal Foundation.
I would also like to especially thank two other de facto members of my supervisory committee: Ray Cranston from the Geological Survey of Canada and Andrew Heyes from Chesapeake Biological Laboratory (CBL) at the University of Maryland, who both assisted me in all aspects of this project. Thanks are also due to: Debby Heyes at CBL for training me in the often-incomprehensible world of mercury analysis; to Gail Chmura at McGill University who inspired me to begin this research many years ago; and to Brian and Mamie Branfireun for providing me with a surrogate family in Toronto and for continually reminding me that the incremental benefit of running one more distillation might not be worth lasting insanity.
The staff at Hunstman Marine Science Center in St. Andrews, NB, particularly Fred, Stephen and Mick, provided much needed logistical support for the majority of field work conducted as part of this study, and I thank the facility for in-kind contributions to the work. I also thank Hugh Akagi and Dave Wildish at the Biological Station in St. Andrews for continual use of their probes, lab and reagents while I was in the field and for not minding when I happened to break a piece of glassware. Thanks also to the crews ofthe coastguard vessels J.L. Hart and J. V. Navicula and to Bob "Murph" Murphy for being our favorite field technician and calling me "boss" in front of the crew. I am grateful to John Dalziel, Gareth Harding, Tim Milligan, Doug Loring, Mike Parsons and Peter Vass at Bedford Institute of Oceanography for countless hours of discussion, access to data, field equipment, and in-kind contributions that made this work possible. Yes, John I do owe you another bottle of rum. Angelika Bayer and Janice Weightman also deserve special mention as the two masters students that have contributed to the development of this project. "Red Rooster" will always be my favorite breakfast spot in New Brunswick and I promise never to insist on sampling during a hurricane again.
Thanks are also due to the people in the REM toxicology lab and good friends at REM who have been an integral part of this experience. Finally, my greatest debt of gratitude is owed to my parents, Julie, Laura and Mark who accepted and supported my choices and decisions over these past few years without question. I thank Julie for always providing me with the example of what was possible and for lying to me when necessary. Lastly to Mark, I know this process has been especially challenging and demanding for both of us and I am so grateful for your unyielding love and support. Table of Contents .. Approval Page ...... 11 Abstract ...... 111 Acknowledgements...... v Table of Contents ...... vii List of Tables ...... X List of Figures ...... Xlll List of Abbreviations ...... xvii 1.0 Introduction...... 1 1.1 General Background ...... 2 1.2 Mercury in Coastal Ecosystems ...... 5 1.3 The Problem of Regulation ...... 7 1.4 Major Scientific Questions Addressed in Thesis ...... 9 1.5 Thesis Outline ...... 10 1.6 Thesis Methods ...... 12 1.6.1 Field Research ...... 12 1.6.2 Contributions of Co-investigators ...... 13 1.7 Literature Cited ...... 20 2.0 Estimating the Anthropogenic Component of Atmospheric Mercury Loading in the Bay of Fundy, Canada ...... 25 2.1 Abstract ...... 25 2.2 Introduction ...... 26 2.3 Methods...... 29 2.4 Results and Discussion ...... 33 2.4.1 Physical and Biological Redistribution ...... 36 2.4.2 Redox Induced Changes ...... 42 2.4.3 Quantitative Retention of Mercury in the Sediments (Advective Fluxes of Hg) ...... 47 2.4.4 Anthropogenic Component of Atmospheric Deposition...... 51 2.4.5 Relationship between Mercury Emissions and Deposition... 54 2.5 Literature Cited ...... 62 3.0 Environmental Controls on the Speciation and Distribution of Mercury in Bay of Fundy Sediments...... 68 Abstract ...... 68 Introduction...... 69 Theory ...... 71 Methods ...... 74 3.4.1 Site Description ...... 74 3.4.2 Sediment Sample Collection ...... 77 3.4.3 Sediment Geochemistry...... 78 3.4.4 Sulfide Measurements ...... 79 3.4.5 Sediment Pore Waters ...... 80 3.4.6 Mercury Analyses ...... 80 3.4.7 Statistical Analysis ...... 81 Results and Discussion ...... 82
vii 3.5.1 Total and Methyl-mercury Distribution ...... 3.5.2 Role of Total Organic Carbon ...... 3.5.3 Redox Potential (Eh) ...... 3.5.4 Sulfides ...... 3 S.5 Mutivariate Analysis ...... 3.6 Conclusions...... 3.7 Literature Cited ...... 4.0 Speciation of Mercury in Well-Mixed Estuarine Sediments ...... 4.1 Abstract ...... 4.2 Introduction...... 4.3 Methods ...... 4.3.1 Study Site ...... 4.3.2 Samples Collected ...... 4.3.2.1 Push Cores ...... 4.3.2.2 Gravity Cores ...... 4.3.2.3 Sediment Pore Waters ...... 4.3.2.4 Isotope Cores ...... 4.3.2.5 Polychaetes ...... 4.3.3 Mercury. . Analyses 4.3.4 Stabsbcal analysis ...... 4.4 Results ...... 4.4.1 Speciation in the Active Sediment Layer ...... 4.4.2 Total mercury (Hg-T) ...... 4.4.3 Methylmercury (MMHg) ...... 4.4.4 Porewater-Solids Partitioning ...... 4.4.5 Geochemical Characteristics of the Sediment Column ...... 4.4.6 Anthropogenic Mercury in Passamaquoddy Bay ...... 4.5 Discussion...... 4.6 Literature Cited ...... 5.0 An Empirical Model of Mercury Cycling in Passamaquoddy Bay, NB . Abstract ...... Introduction...... Methods ...... 5.3.1 Study Site...... 5.3.2 Model. Development...... Model Description...... Treatment of Uncertainty in the Model ...... Model Parameter Estimates ...... 5.6.1 External Inputs of Mercury ...... 5.6.1.1 Atmospheric Deposition ...... 5.6.1.2 Riverborne Inputs of Mercury ...... 5.6.1.3 Tidal Mow...... 5.6.2 Water Column Dynamics ...... 5.6.2.1 Tidal Outflow ...... 5.6.2.2... Settling of Suspended Solids ...... 5.6.3 Volatihzabon ......
... Vlll 3.5.1 Total and Methyl-mercury Distribution ...... 3.5.2 Role of Total Organic Carbon ...... 3.5.3 Redox Potential (Eh) ...... 3.5.4 Sulfides ...... 3 S.5 Mutivariate Analysis ...... 3.6 Conclusions...... 3.7 Literature Cited ...... 4.0 Speciation of Mercury in Well-Mixed Estuarine Sediments...... 4.1 Abstract ...... 4.2 Introduction...... 4.3 Methods ...... 4.3.1 Study Site...... 4.3.2 Samples Collected ...... 4.3.2.1 Push Cores ...... 4.3.2.2 Gravity Cores ...... 4.3.2.3 Sediment Pore Waters ...... 4.3.2.4 Isotope Cores ...... 4.3.2.5 Polychaetes ...... 4.3.3 Mercury Analyses ...... 4.3.4 Statistical analysis ...... 4.4 Results ...... 4.4.1 Speciation in the Active Sediment Layer ...... 4.4.2 Total mercury (Hg-T) ...... 4.4.3 Methylmercury (MMHg) ...... 4.4.4 Porewater-Solids Partitioning ...... 4.4.5 Geochemical Characteristics of the Sediment Column ...... 4.4.6 Anthropogenic Mercury in Passamaquoddy Bay ...... 4.5 Discussion...... 4.6 Literature Cited ...... 5.0 An Empirical Model of Mercury Cycling in Passamaquoddy Bay, NB . 5.1 Abstract ...... 5.2 Introduction...... 5.3 Methods ...... 5.3.1 Study Site ...... 5.3.2 Model. Development...... 5.4 Model Description...... 5.5 Treatment of Uncertainty in the Model ...... 5.6 Model Parameter Estimates ...... 5.6.1 External Inputs of Mercury ...... 5.6.1.1 Atmospheric Deposition...... 5.6.1.2 Riverborne Inputs of Mercury ...... 5.6.1.3 Tidal Inflow ...... 5.6.2 Water Column Dynamics ...... 5.6.2.1 Tidal Outflow ...... 5.6.2.2 Settling of Suspended Solids ...... 5.6.3 Volatilization ......
viii 5.6.4 Sediment Budget ...... 5.6.5 Resuspension...... 5.6.6 Net Methylation of Mercury ...... Mass Budgets ...... 5.7.1 Total Mercury ...... 5.7.2 Methyl Mercury ...... 5.7.3 Demethylation in Sediments and Water ...... 5.7.4 Elemental Mercury Budget ...... Concentrations in Benthic Organisms ...... Application of the Empirical Model ...... Literature Cited ...... 6.0 Conclusions and Implication for Policies...... 6.1 Major Scientific Findings ...... 6.2 Implications for Policy and Management ...... 6.3 Future Research Directions...... 6.4 Literature Cited ...... 7.0 Appendices...... 7.1 Supporting Data Chapter 2 ...... 7.2 Supporting Data Chapter 3 ...... 7.3 Supporting Data Chapter 4 ...... 7.4 Supporting Data Chapter 5 ...... List of Tables
Table 2-1 Estimated tidal and atmospheric contributions to mercury loading in Dipper Harbour and Chance Harbour salt marsh sediments...... Table 2-2 Summary of mercury data obtained from dated sediment cores in the Bay of Fundy region of Canada. Data from ombrotrophic bog sediments are from Rutherford and Matthews (1998) and lake sediment data from Kainz et al. (1997). ASEF = anthropogenic sediment enrichment factors calculated by dividing mercury concentrations andlor loading rates in the surface horizons of sediment cores by the average pre-industrial concentrationsAoading in sediments that accumulated prior to 1880...... Table 2-3 Estimated mercury deposition rate in Maritime Canada based on historical emissions and precipitation data. These estimates are compared to sedimentary data from the Bay of Fundy region. Current deposition rates calculated from sediment cores should be in the range of the data presented if a reliable history of mercury loading is obtained...... Table 3- 1 Seasonal correlation matrices for monomethylmercury (MMHg) concentrations in Bay of Fundy sediments. Pearson correlation coefficients (r) are significant at the 95% level or greater. %MMHg = Fraction of total mercury (Hg) as MMHg. TOC = Total Organic Carbon. Eh = Redox Potential. NC = No Correlation...... Table 3-2 Correlation matrix for sediment redox potential (Eh), sulfide concentrations, and total organic carbon content (TOC) in sediments fiom Passamaquoddy Bay...... Table 3-3 Summary of regression equations developed in this study for mercury speciation and distribution in Bay of Fundy sediments...... Table 4-1 Porewater Hg-T and MeHg data fiom each sampling stations SC- 1, PB-1 through PB-5...... Table 4-2 Geochemical characteristics of gravity cores from Passamaquoddy Bay ...... Table 4-3 Variability in sulfide concentrations measured in Passamaquoddy Bay sediments...... Table 4-4 Concentrations of metals measured in sediment cores...... Table 5-1 Mass balance equations describing parameters and rate constants included in mercury cycling model...... Table 5-2 Summary of ecosystem parameters used in empirical model of mercury cycling in Passamaquoddy Bay...... Table 5-3 Concentrations and parameters used to estimate mercury speciation in water and sediment compartments...... Table 5-4 Total mercury (Hg-T) concentrations measured in sediments, water and benthic organisms in Passamaquoddy Bay ...... Table 5-5 Accumulation factors for biota between water (G)and sediment concentrations (C,)...... Table 7-1 Unsupported 210~band 13'cs data measured in sediment cores from the Bay of Fundy region. DH-A = Dipper Harbour salt marsh core A; DH-B = Dipper Harbour salt marsh core B; LL = Lily Lake core; SPL = St. Patrick's Lake core; CH = Chance Harbour salt marsh core ...... Table 7-2 Caribou Plains Bog data from Rutherford & Matthews (1998)...... Table 7-3 Lily Lake core data from Kainz et al. (1996)...... Table 7-4 St. Patrick's Lake core data from Kainz et al. (1996)...... Table 7-5 Chance Harbour salt marsh data. Note: 1830 sediment date estimated from pollen marker horizon. ASEF = Anthropogenic sediment enrichment factor calculated by dividing the mean loading rate to sediments that accumulated past 20.5 cm depth by respective loading rates in each subsequent sediment layer...... Table 7-6 Sediment data from Dipper Harbour salt marsh core A...... Table 7-7 Sediment data from Dipper Harbour salt marsh core B...... Table 7-8 Sediment data from Bocabec salt marsh...... Table 7-9 Spatial distribution of total mercury (Hg-T) in Passamaquoddy Bay.. T&le 7- 10 Correlation between total mercury (Hg-T) concentrations, total organic carbon and sediment grain size at selected sampling locations. Grain size data is from Loring et al. (1998)...... Table 7-1 1 Raw data for total (Hg-T) and methyl (MMHg) mercury concentrations, total organic carbon (TOC), redox (Eh) and sulfide concentrations measured in grab samples collected in Passamaquoddy Bay, the St. Croix River and outer Bay of Fundy.. ... Table 7-12 Bivariate correlation matrices for benthic sediment grab samples obtained in May 2001 ...... Table 7- 13 Bivariate correlation matrices for benthic sediment grab samples obtained in August 2000-2001 ...... Table 7-14 Bivariate correlation matrices for benthic sediment grab samples obtained in November 200 1...... Table 7- 15 Raw data for benthic sediment samples from Passamaquoddy Bay analyzed for porewater mercury concentrations in August 2001. Hg-I = Inorganic mercury concentration in the porewaters calculated as the difference between total mercury (Hg-T) and methylmercury (MMHg) concentrations in porewaters...... Table 7- 16 Raw Data Collected at Seasonally Monitored Stations...... Table 7- 17 Summary statistics for t-tests of seasonal differences in paired means...... Table 7- 18 Push core mercury data...... Table 7- 19 Raw data fiom gravity cores collected in Passamaquoddy Bay...... Table 7-20 Summary of annual data for mercury concentration and deposition in precipitation collected at weekly intervals between July 1996 and June 2000. Data from: Beauchamp (1998); Beauchamp (unpublished data, 2000). Concentrations are expressed as volume weighed means ('Vol Wt. Conc.') to reduce the effects of small samples with high concentrations and thus providing a more representative overall concentration. The standard deviation of mean concentration in precipitation in the "average" column is the standard deviation around the yearly mean values and does not take into account the deviation around each annual mean value...... 241 Table 7-21 Monthly freshwater discharges in P. Bay. Data adapted from: Gregory et al., (1993). Data presented are sum of freshwater discharges into P. Bay and the St. Croix River estuary...... 242 Table 7-22 Total mercury (Hg-T) concentrations measured in three main freshwater tributaries of Passamaquoddy Bay. Data from: Dalziel et al. (1998); Dalziel (unpublished data, 2002)...... 243 Table 7-23 Methylmercury (MMHg) concentrations measured in three main freshwater tributaries of Passamaquoddy Bay. Data from: Dalziel (unpublished data, 2002)...... 244 Table 7-24 Suspended particulate matter (SPM) mg L-' in freshwater tributaries of P. Bay. Data from: Dalziel et al. (1998). Measured concentrations are weighted by the respective discharges from each river for the means; annual averages are calculated from the fraction of discharge occurring during each sampling period. Seawater samples were filtered using a 0.4 pm Nucleopore filtration system...... 245 Table 7-25 Summary of annually averaged inputs of mercury to Passamaquoddy Bay from rivers, atmospheric deposition and tidal inflow ...... 246 Table 7-26 Mercury concentrations measured in Gammarus sp. collected from St. Andrews tidal flats in August 2001...... 247 Table 7-27 Mercury concentrations measured in Nephtys sp. at the head of the St. Croix River Estuary and in Passmaquoddy Bay, August 2001 ...... 247
xii List of Figures
Figure 1-1 Location of Passamaquoddy Bay at the mouth of the Bay of Fundy ...... Figure 1-2 The research vessel J.L. Hart shown here in St. Andrews, New Brunswick was used to collect sediment gravity cores, push cores and grab samples in May 2001...... Figure 1-3 The research vessel J. V. Navicula, moored in Brier Island, Nova Scotia during field sample collections in June 2001...... Figure 1-4 Deployment of gravity corer on board the J.L. Hart in May 2001 to sample bottom sediments in Passamaquoddy Bay, New Brunswick ...... Figure 1-5 Deployment of modified Van-Veen grab sampler for collection of benthic sediments in Passamaquoddy Bay ...... Figure 1-6 Deployment of Niskin sampler used to collect seawater samples for total and methyl-mercury (MMHg) analysis in Passamaquoddy Bay and the outer Bay of Fundy ...... Figure 2-1 Approximate locations of sediment coring sites in the Bay of Fundy region. Bog data are fiom Rutherford and Matthews (1998) and lake data are from Kainz et al. (1997) ...... Figure 2-2 Summary of mercury concentration data fiom dated sediment cores collected in the Bay of Fundy region. Compaction of the sediment with depth in Caribou Plains bog accounts for the discrepancy between mercury concentration data and loading rates that take into account the sediment bulk density...... Figure 2-3 Radionuclide data used to date salt marsh sediments also indicate disturbances through hysical andlor biological perturbations. The "ideal" decay of "Pb is shown by the straight line next to measured unsupported isotope activity...... Figure 2-4 Radionuclide and redox data for lake core sediments collected in the Bay of Fundy region. The "ideal" decay of *lOpbis shown by the straight line next to measured unsupported isotope activity. '37~sdata were not available for these cores. These data are used to indicate disturbances through physical and/or biological perturbations, as well as diffusion of mercury within the sediment cores. Significant correlations between mercury (Hg) and redox potential (Eh) are indicated by Spearman rank correlation coefficents (r,) with p-values of xiv (August) and fall (November) respectively...... Figure 4-1 Map of the study area showing sampling locations for gravity cores, push cores and isotope cores...... Figure 4-2 Speciation of mercury in contrasting depositional (SC-1) and well-mixed sediments (PB-3 & PB-4). Data include ambient total mercury (Hg-T), methylmercury (MMHg) and the fraction of total mercury in the sediments present as MMHg (%MMHg). Mercury isotope experiments were used to estimate the "potential" methylation rates at the head of the St. Croix River Estuary (SC- 1) and the center of Passamaquoddy Bay (PB-3)...... Figure 4-3 Porewater sulfate and ammonium concentrations in gravity cores fiom Passamaquoddy Bay and the St. Croix River Estuary. Sulfate concentrations <24 mM indicate the anoxic reduction of sulfate and ammonium concentrations >0.5 mM indicate the presence of fblly reduced buried sediments...... Figure 4-4 Anthropogenic sediment enrichment factors (ASEFs) for Hg-T at the head of the St. Croix River and the center of Passamaquoddy Bay. ASEFs are used to estimate modern/background Hg-T levels in P. Bay sediments...... Figure 4-5 Conceptual model of mercury speciation in contrasting depositional and well-mixed sediments...... Figure 5-1 Spatial grid of sedimentation rates generated for Passamaquoddy Bay ...... Figure 5-2 Contour map of total mercury concentrations (ng g" dry) in Passamaquoddy Bay. Concentration gradients are only valid for regions within Passamaquoddy Bay and the St. Croix River as field sampling did not cover the regions beyond Deer Island in the outer Bay of Fundy shown in the figure...... Figure 5-3 Methylmercury concentration distribution in August (pg g-' dry). Concentration gradients are only valid for regions within Passamaquoddy Bay and the St. Croix River as field sampling did not cover the regions beyond Deer Island in the outer Bay of Fundy shown in the figure ...... Figure 5-4 Contour map of methylmercury concentration distribution in May (pg g-' dry). Concentration gradients are only valid for regions within Passamaquoddy Bay and the St. Croix River as field sampling did not cover the regions beyond Deer Island in the outer Bay of Fundy shown in the figure ...... Figure 5-5 Contour map of sediment depositional flux (ng cmm2yr-') of total mercury in Passamaquoddy Bay. Concentration gradients are only valid for regions within Passamaquoddy Bay and the St. Croix River as field sampling did not cover the regions beyond Deer Island in the outer Bay of Fundy shown in the figure...... Figure 5-6 Contour map of sediment depositional flux (ng ~rn'~y-') of methylmercury in Passamaquoddy Bay. Concentration gradients are only valid for regions within Passamaquoddy Bay and the St. Croix River as field sampling did not cover the regions beyond Deer Island in the outer Bay of Fundy shown in the figure...... 168 Figure 5-7 Mass budget for total mercury (Hg-T) in Passamaquoddy Bay .. . .. 172 Figure 5-8 Steady state mass budget for methyl mercury (CH,Hg) in Passamaquoddy Bay...... 174 Figure 5-9 Steady state mass balance for divalent mercury (Hg(I1)) in Passamaquoddy Bay ...... 176 Figure 5-10 Hypothesized steady state mass balance for elemental mercury (~~4in Passamaquoddy Bay...... 178 xvi List of Abbreviations ASEF Anthropogenic Sediment Enrichment Factor Eh Redox Potential m0 Elemental Mercury Hg(II) Divalent Mercury Species Hg -I Inorganic Mercury Species (Hg(I1) + H~') Hg-T Total Mercury (Hg(I1) + H~O+ MMHg) MMHg/CH3Hg Monomethylmercury Species %MMHg Fraction of Total Mercury Concentration or Mass that is Methylmercury P. Bay Passamaquoddy Bay SRB Sulfate Reducing Bacteria TOC Total Organic Carbon xvii CHAPTER 1 Introduction I know not what I appear to the world, but to myself1 seem to have been only like a boy playing on the sea shore, and diverting myself in now and thenfinding a smoother pebble or a prettier shell than ordinav, whilst the great ocean of truth lay all undiscovered before me. Isaac Newton 1.1 General Background Mercury contamination is a well-known problem in most industrialized countries. In Sweden, for example, an estimated 10,300 lakes contain fish with mercury concentrations that are above the 1 mg kge' threshold that is generally considered to be the safe limit for humans who consume fish on a regular basis (Hakanson et al., 1990). In the United States, between 1-3% of women of childbearing age are exposed to potentially hard1 levels of mercury through the consumption of fish (USEPA, 1997). In 1998,39 States and 5 Canadian Provinces issued consumption advisories for freshwater fish to diminish the risks of mercury exposure (NESCAUM et al., 1998). Mercury causes neurological, developmental and reproductive problems in exposed organisms and humans (Clarkson, 1997, Ratcliffe and Swanson, 1996). Since the 1970s, managers have attempted to minimize the deleterious effects of mercury in the environment by regulating releases from large point sources and banning the inclusion of mercury in many consumer products. Throughout history, mercury has been a component of a variety of industrial and consumer products such as paint, pharmaceuticals, electrical goods, and various pest control products (Nriagu, 1979). Huge quantities of mercury were released in the late 1700s and 1800s during gold and silver mining activities in North and South America (Nriagu, 1994). Mercury is still commonly found in dental amalgams, fluorescent tubes, thermostats, thermometers, and other instrumentation. In addition, human activities such as fossil fuel combustion and base metal smelting result in the release of natural mercury deposits in sulfidic ores and disruption of natural biogeochemical cycles. In Canada, anthropogenic emissions of mercury have been reduced by approximately 85% since the 1970s. However, mercury levels in the tissues of fish and wildlife in many regions have not declined (CCME, 1999). The integration of mercury released as a byproduct of human activities with natural background concentrations makes it difficult to isolate and identify the effects of "antbropogenic" contamination. The relationship between anthropogenic mercury releases and resulting concentrations in the environment is confounded by the fact that mercury is a naturally occurring element and is now ubiquitous in the environment (Nriagu, 1979). In addition, the propensity of mercury for long-range transport and deposition can result in contamination of regions that are distant f5om point sources of pollution (Fitzgerald et al., 1998, Hermanson, 1993, Jackson, 1997). Mercury released to the atmosphere as the result of human activity becomes integrated with the natural pool of mercury in the environment and may then be continuously deposited and re-emitted, resulting in an ongoing legacy of mercury contamination (Nriagu, 1993, Nriagu, 1994). The re-emission of past sources of anthropogenic mercury is a significant component of ongoing pollution in many areas, accounting for up to one third of the current reservoir of mercury in the atmosphere (Pirrone et al., 1996). The total mass of mercury in a given ecosystem is thus a complex mix of local and long-range anthropogenic sources, natural sources, and recycled anthropogenic sources. The relationship between releases of mercury to the environment and accumulation of mercury in the food web is also complicated by the fact that concentrations in organisms are not necessarily a direct function of total mercury concentrations in the environment. Inorganic mercury species make up 95% to 99% of the mercury released fi-om anthropogenic sources (Lin and Pehkonen, 1999). Approximately 95% of the total mercury in the atmosphere is the gaseous elemental mercury species (Hg4 (Lindberg and Stratton, 1998). Elemental mercury is oxidized in the atmosphere to form the more soluble mercuric ion (Hg(I1)) (Schroeder et al., 1989). The mercuric ion is efficiently scavenged fiom the atmosphere in precipitation (wet and dry) and its complexes are the principle forms of mercury in aquatic systems (Jackson, 1998). Microorganisms in both the water and sediment transform a small fraction of the "pool" of mercuric ion (Hg(I1)) into the organic mercury species, monomethylmercury (cH~H~+o~MMHg). MMHg is the only species of mercury that significantly bioaccumulates in organisms (Lawson and Mason, 1998, Mason et al., 1996, Watras and Bloom, 1992) and is also a more potent toxin than inorganic mercury (Fitzgerald and Clarkson, 1991). Differences in the bioavailability and accumulation rates of inorganic and organic mercury species mean that concentrations of mercury in organisms are a function of the net rate of MMHg formation in addition to the amount of total mercury in a given ecosystem. Total mercury deposition is often a poor predictor of this rate because a number of physical and chemical factors can control the rate of MMHg formation in ecosystems, and inorganic mercury availability is only one of the important variables (Kelly et al., 1995). For example, the rate of MMHg formation is often proportional to the rate of sulfate reduction in freshwater systems (Gilmour and Henry, 1991, Gilmour et al., 1992). Accordingly, increases in sulfate deposition in many lakes throughout North America with acid rain problems have been linked to increases in mercury concentrations in organisms (Spry and Wiener, 1991). Mercury contamination problems can therefore occur even when there has been no change in the total amount of mercury deposited in the ecosystem. Thus, understanding the factors that affect the transport, fate and speciation of mercury is important for predicting the exposure and accumulation by organisms. 1.2 Mercury in Coastal Ecosystems There has been relatively little research on the dynamics of mercury in coastal and marine areas, despite their importance in terms of living resources. Studies such as Gill and Fitzgerald (1988) and Mason and Fitzgerald (1991) suggest that pelagic marine areas may be most susceptible to increases in atmospheric deposition of mercury. In many coastal and marine areas, production of MMHg appears to be limited by the supply of inorganic mercury (Benoit et al., 1998, Bloom et al., 1999, Mason and Lawrence, 1999). This contrasts mercury dynamics in many freshwater systems where environmental variables such as nutrient supply, pH and sulfate availability are more important than inorganic mercury supply (Gilmour et al., 1992, Kelly et al., 1995). Although the exact mechanism of mercury methylation is unknown, recent advances have been made in understanding the mechanisms of MMHg formation in marine and estuarine systems (Benoit et al., 1998, Benoit et al., 1999a, Benoit et al., 2001, Benoit et al., 2002, Bloom et al., 1999, Gilmour et al., 1998, King et al., 2001). In the marine environment, MMHg production is strongly influenced by the complexation of inorganic mercury by sulfides (Benoit et al., 1999b, Benoit et al., 1999a, Craig and Bartlett, 1978, Craig and Moreton, 1986), the uptake of dissolved neutral mercury species by sulfate reducing bacteria (Benoit et al., 2001, Benoit et al., 1999b, King et al., 2001), as well as by dissolved organic carbon concentrations, redox conditions and salinity (Compeau and Bartha, 1984, Compeau and Bartha, 1987, Laporte et al., 1997, Miskirnmin, 1991). Therefore, ecosystem changes and anthropogenic "stresses" that do not result in a direct increase in mercury loading to the ecosystem but alter the rate of MMHg formation may also affect mercury levels in organisms (e.g., Grieb et al., 1990). Environmental changes such as eutrophication, that may alter microbial activity and the chemical dynamics of mercury within an ecosystem, must therefore be considered together with emission control strategies to effectively manage mercury accumulation in the food web. Coastal areas in temperate latitudes are resource rich and provide habitat for a diverse range of plant and animal species. These areas also tend to be heavily populated and are therefore highly susceptible to anthropogenic influences. A number of industrialized estuarine regions such as the Scheldt estuary, flowing into the North Sea, and Kastela Bay in the Central Adriatic are already experiencing mercury contamination problems, even at low trophic levels (Mikac et al., 1985, Baeyens et al., 1998). In addition, Mason et al. (1994) speculate that over the last century anthropogenic emissions have tripled mercury concentrations in the surface waters of oceans. Mercury is a significant problem in Atlantic Canada and the New England States. The majority of anthropogenic mercury emissions in North America are released in the eastern half of the United States (Pai et al., 1997). Consumption advisories are in place throughout the Atlantic region due to elevated levels of mercury, and aboriginal peoples who consume large quantities of fish and pregnant women are believed to be at risk (NESCAUM et al., 1998). The Bay of Fundy is one of Canada's most distinct ecological regions. Both coastal fisheries and salmon and finfish aquaculture in this area are vital for the local economy. In their recent report on the status of the Fundy ecosystem, Wells et al. (1997) noted a number ofundesirable changes occurring in the region. Gaskin et al. (1973 and 1979) first noted elevated levels of mercury in porpoises (Phocoenaphocoena) and seals (Phoca vitula). High levels of mercury in the blood of common loons (Gavia immer) and seabirds (e.g. the cormorant, Phalacrocorax auritus) residing in the wider Fundy region have raised much concern, and may be threatening the viability of certain species (Elliott et al., 1992, Evers et al., 1998). Mercury is now a priority contaminant for study and further monitoring in the Bay of Fundy and Gulf of Maine (Burt and Wells, 1997, Chase et al. 2001, Pilgrim et al., 1998). Passamaquoddy Bay (Fig 1-I), the site of this study, is located at the mouth of the Bay of Fundy, and runs along the border between Canada and the United States. This region is located at the coalescence point for several continental airmasses fiom Ontario, Quebec, and the Eastern and Central United States, making study of atmospherically derived mercury in the area particularly relevant (Beauchamp, 1998). 1.3 The Problem of Regulation Regulation of atmospheric mercury emissions is a transboundary international problem. Elemental mercury (~~4is highly volatile and has an estimated atmospheric lifetime of one year (Pai et al., 1997, Iverfeldt and Lindqvist, 1986). Figure 1-1. Location of Passamaquoddy Bay at the mouth of the Bay of Fundy. Approximately half of the mercury released from anthropogenic sources is deposited near its point of origin, while the rest is transported long distances in the atmosphere (Seigneur et al., 1997, Seigneur et al., 1999). Annually, this results in approximately 3,000 tomes of anthropogenic mercury deposited globally in areas removed fiom point sources of pollution (Jackson, 1997). Accordingly, the United States Environmental Protection Agency has acknowledged a "plausible link" between elevated levels of mercury in the atmosphere and high levels in fish in remote regions (USEPA, 1997). An agreement between the New England Governors and Eastern Canadian Premiers to further reduce anthropogenic emissions by 50% was ratified in 1998 (NEG-ECP, 1998). However, the effectiveness of this policy at reducing concentrations of mercury in organisms is unknown, making it difficult for environmental managers and policy makers in Canada to negotiate management strategies to secure future fresh and marine water quality. 1.4 Major Scientific Questions Addressed in Thesis While considerable progress has been made in understanding the fate of mercury, it is not currently possible to predict how emissions reductions and other types of environmental change will affect mercury concentrations in organisms. The purpose of this study is to investigate the relationship between anthropogenic mercury loading and concentrations in marine ecosystems in Eastern Canada through a case study of Passamaquoddy Bay, a coastal embayment located at the mouth of the Bay of Fundy. Two major scientific questions that need to be ansewered to quantify this relationship for Passamaquoddy Bay are: 1) What is the anthropogenic component of total mercury deposition? 2) Which factors control MMHg production and subsequent accumulation in organisms? Therefore, the objectives of this study are to: (i) Develop a comprehensive "ecosystem-based" understanding of the environmental fate of anthropogenic mercury in coastal marine environments; (ii) Formulate and document this knowledge in an empirical model that links anthropogenic mercury inputs to concentrations in sediments, water and benthic marine organisms; and (iii) Consider the environmental management implications of the findings. 1.5 Thesis Outline The thesis consists of six chapters that address different components of these questions. This chapter (Chapter 1) serves as a general introduction to the problem of mercury contamination and provides the rationale for the research presented in subsequent chapters. Chapter 2 investigates the relationship between mercury emissions and deposition in the Bay of Fundy region by quantifying the component of atmospheric mercury loading that is anthropogenic in origin. Sediment core data from coastal salt marshes, lakes and bogs are used to investigate historical trends in atmospheric mercury deposition and to identify local, regional and long-range signals of contamination. This information is compared to regional and global scale mercury emissions inventories to assess the effectiveness of policies aimed at reducing emissions for lowering atmospheric loading of mercury in the Bay of Fundy region. 1 The purpose of Chapters 3 & 4 is to elucidate some of the main factors controlling the speciation and distribution of mercury in Passamaquoddy Bay. Chapter 3 investigates some of the key environmental and geochemical variables that control MMHg production in coastal systems. These variables include: redox status, sulfide concentrations, grain size, total organic carbon concentrations and total mercury levels in the sediments. This information is important for understanding the characteristics of Passamaquoddy Bay and other coastal ecosystems that control the fate, speciation of and bioavailability of mercury in benthic sediments. Chapter 4 provides insight into the variables that determine the time-response of Passamaquoddy Bay to overall declines in mercury loading and the component of the total mercury pool that is most likely to be accumulated by organisms. To do this, the effects of dynamic physical processes occurring in Passamaquoddy Bay on MMHg production in the sediment compartment are examined. Sediment core data are used to investigate stratigraphic variability in MMHg production in the sediment column and to identify the active layer in this system where MMHg production occurs. Because the depth of the active sediment layer determines the magnitude of total and MMHg reservoirs that are potentially available to organisms, this information is critical for the development of the mercury cycling model. Anthropogenic enrichment factors for total mercury loading in Passamaquoddy Bay are calculated fiom gravity cores to identify the natural and anthropogenic components of the total mass of mercury contained in the sediment compartment. * In Chapter 5, the information presented in previous chapters is used to develop an empirical model of mercury cycling in Passamaquoddy Bay. The model combines physical and hydrodynamic data for Passamaquoddy Bay with measured mercury concentrations in precipitation, river and tides to quantify the inputs and outputs of mercury in the sediment and water compartments. The model is used to quantify the relative importance of different mercury sources in Passamaquoddy Bay and provides insight into the physical mechanisms that are most important for mercury distribution, speciation and accumulation by organisms in this system. The results of mass budget calculations for inorganic and organic mercury species are compared to concentrations measured in benthic organisms to explore the relationship between mercury inputs and concentrations in organisms. In Chapter 6, the major scientific findings of each of the four main chapters are presented and the resulting advancements in understanding of the fate of mercury in coastal marine ecosystems are summarized. The broad implications of this study for policies regulating mercury contamination are discussed in this chapter. 1.6 Thesis Methods 1.6.1 Field Research I completed five cruises between August 2000 and November 2001 in Passamaquoddy Bay and the outer Bay of Fundy as part of this research. The cruises were coordinated in conjunction with researchers fiom the Department of Fisheries and Oceans (Gareth Harding and John Dalziel) in June 2001 and with Ray Cranston, Geological Survey of Canada in May 200 1 on the coast guard vessels J. V. Navicula and J.L. Hart respectively e (Fig 1-2 and 1-3). Cruises conducted in August 2000-200 1 and November 2001 made use of vessels at the Huntsman Marine Science Center in St. Andrews, New Brunswick. The objectives of the fieldwork were to: i) investigate ecosystem controls on the spatial and stratigraphic variability in mercury speciation in the sediment compartment and ii) collect data needed to parameterize the mercury cycling model for Passamaquoddy Bay. 1.6.2 Contributions of Co-investigators Field sampling was coordinated with other co-investigators to maximize the amount of data collected for development of the mercury cycling model. I collected and analyzed the coastal salt marsh sediment core from Chance Harbour discussed in Chapter 1. The remaining salt-marsh cores, and 2'0~band 137~~data were collected by Gail Chmura at McGill University. Ray Cranston analyzed sediment gravity cores (Fig 1-4) collected in May 2001 to determine sediment burial rates throughout Passamaquoddy Bay. Michael Parsons, Geological Survey of Canada (Bedford Institute of Oceanography), analyzed these cores for metals (excluding mercury). John Dalziel collected seawater data used to parameterize the mercury cycling model from the mouth of the Bay of Fundy in June 2001. In August 2001, Angelika Bayer (M.Sc. student, Brian Branfirem) collected cores for experimental work with mercury isotopes to determine "potential" MMHg production rates in Passamaquoddy Bay and at the head of the St. Croix River. These cores were analyzed by Andrew Heyes, Chesapeake Biological Laboratory, Maryland. Figure 1-2. The research vessel J.L. Hart shown here in St. Andrews, New Brunswick was used to collect sediment gravity cores, push cores and grab samples in May 2001. Figure 1-3. The research vessel J. V.Navicula, moored in Brier Island, Nova Scotia during field sample co1lections in June 200 1. Figure 1-4. Deployment of gravity corer on board the J.L. Hart in May 200 1 to sample bottom sediments in Passamaquoddy Bay, New Brunswick. Throughout four field seasons, I collected surface sediment samples using a modified Van Veen sampler (Fig 1-5) and push cores to investigate the spatial and stratigraphic variability in mercury in Passamaquoddy Bay sediments. Efforts were made to minimize the disturbance of surface sediments and preserve the sediment-water interface during the collection of all benthic sediment samples. This was done with the assistance of an experienced field technician (Bob Murphy) fiom Bedford Institute of Oceanography, employed by the Geological Survey of Canada. I extracted pore waters from sediment grab samples and push cores in August 2001 and collected seawater samples (Fig 1-6) fiom Passamaquoddy Bay in November 2001. The sample sizes of both sediment grabs and water samples in November 2001 were restricted by the occurrence of a hurricane during the sample collection period. I analyzed all sediment, seawater and pore-water samples in a Class 100 Clean Lab (Brian Branfireun) at the University of Toronto. The data analysis, modelling and interpretations presented in this thesis are all my own work, except where otherwise highlighted. Figure 1-5. Deployment of modified Van-Vm grab sampler for collection of benthic sediments in Passmaquoddy Bay. 1.7 Literature Cited Baeyens, W., Meuleman, C. and Leennakers, M. 1998. Behavior and speciation of mercury in the Scheldt estuary (water, sediment and benthic organisms). Hydrobiologia, 366,63-79. Beauchamp, S. 1998. Mercury in the atmosphere. In Mercury in Atlantic Canada: A Progress Report (Ed, Burgess, N.) Environment Canada, Atlantic Region, Bedford, NS, pp. 16-45. Benoit, J., Gilmour, C. C., Mason, R. P. and Heyes, A. 1999a. Sulfide controls on mercury speciation and bioavailability to methylating bacteria in sediment pore waters. Environmental Science and Technology, 33,95 1-957. Benoit, J. M., Gilmour, C. C., Heyes, A., Mason, R. P. and Miller, C. 2002. Geochemical and biological controls over methylmercury production and degradation in aquatic systems. In Biogeochemistry of Environmentally Important Trace Metals. ACS Symposium Series No. 835. (Eds, Cai, Y. and Braids, O.C.) Oxford University Press, New York, pp. 262-297. Benoit, J. M., Gilmour, C. C. and Mason, R. P. 2001. The influence of sulfide on solid- phase mercury bioavailability for methylation by pure cultures of DesuIfobulbus propionicus (lpr3). Environmental Science and Technology, 35,127-132. Benoit, J. M., Gilmour, C. C., Mason, R. P., Riedel, G. S. and Reidel, G. F. 1998. Behavior of mercury in the Patuxent River estuary. Biogeochemistry, 40,249- 265. Benoit, J. M., Mason, R. P. and Gilmour, C. C. 1999b. Estimation of mercury-sulfide speciation in sediment pore waters using octanol-water partitioning and implications for availability to methylating bacteria. Environmental Toxicology and Chemistry, 18,2138-2141. Bloom, N. S. 1992. On the chemical form of mercury in edible fish and marine invertebrate tissue. Canadian Journal of Fisheries and Aquatic Sciences, 49, 1010- 1017. Bloom, N. S., Gill, G. A., Cappellino, S., Dobbs, C., Mcshea, L., Driscoll, C., Mason, R. and Rudd, J. 1999. Speciation and cycling of mercury in Lavaca Bay, Texas, sediments. Environmental Science and Technology, 3 3,7- 13. Burt, M. D. B. and Wells, P. G. 1997. Coastal Monitoring and the Bay of Fundy. Proceedings of the Maritime Atlantic Ecozone Science Workshop. Huntsman Marine Science Center, St. Andrews, N.B., pp. 196. CCME. 1999. Canada-Wide Standards for Mercury. Canadian Council of Ministers of the Environment, Ottawa, pp. 13. Chase, M. E., Jones, S. H., Hennigar, P., Sowles, J., Harding, G. C. H., Freeman, K., Wells, P. G., Krahforst, C., Coombs, K., Crawford, R., Pederson, J. and Taylor, D. 2001. Gulfwatch: Monitoring spatial and temporal patterns of trace metal and organic contaminants in the Gulf of Maine (199 1- 1997) with the blue mussel, Mytilus edulis L. Marine Pollution Bulletin, 42,491-505. Clarkson, T. W. 1997. The toxicology of mercury. Critical Reviews of Clinical Laboratory Science, 34,369-403. Compeau, G. and Bartha, R. 1984. Methylation and demethylation of mercury under controlled redox, pH, and salinity conditions. Applied and Environmental Microbiology, 48, 1203-1207. Compeau, G. C. and Bartha, R. 1987. Effect of salinity on mercury-methylating activity of sufate reducing bacteria in estuarine sediments. Applied and Environmental Microbiology, 53,261 -265. Craig, P. J. and Bartlett, P. D. 1978. The role of hydrogen sulphide in environmental transport of mercury. Nature, 275,635-637. Craig, P. J. and Moreton, P. A. 1986. Total mercury, methyl mercury and sulphide levels in British estuarine sediments-In. Water Research, 20, 11 11- 11 18. Elliott, J. E., Scheuhammer, A. M., Leighton, F. A. and Pierce, P. A. 1992. Heavy metals and metallothionein concentration in Atlantic Canadian seabirds. Archives of Environmental Contamination and Toxicology, 22,63-73. Evers, D. C., Kaplan, J. D., Meyer, M. W., Reaman, P. S., Braselton, W. E., Major, A., Burgess, N. and Scheuhammer, A. M. 1998. Geographic trends measured in common loon feathers and blood. Environmental Toxicology and Chemistry, 17, 173-183. Fitzgerald, W. F. and Clarkson, T. W. 1991. Mercury and monomethyl mercury: present and future concerns. Environmental Health Perspectives, 96, 159- 166. Fitzgerald, W. F., Engstrom, D. R., Mason, R. P. and Nater, E. A. 1998. The case for atmospheric mercury contamination in remote areas. Environmental Science and Technology, 32, 1-7. Gaskin, D. E., Stonefield, K. I. and Suda, P. 1979. Changes in mercury levels in harbour porpoises fiom the Bay of Fundy, Canada and adjacent waters. Environmental Contamination and Toxicology, 8,733-762. Gaskin, G. E., Frank, R., Holdrinet, M., Ishida, K., Walton, C. J. and Smith, M. 1973. Mercury, DDT, and PCB in harbour seals (Phoca vitulina) fiom the Bay of Fundy and Gulf of Maine. Journal of Fisheries Research Board of Canada, 30,471-475. Gill, G. and Fitzgerald, W. F. 1988. Vertical mercury distributions in the oceans. Geochimica Cosmochimica Acta, 52, 1719- 1728. Gilmour, C. C., Henry, E. and Mitchell, R. 1992. Sulfate stimulation of mercury methylation in freshwater sediments. Environmental Science and ~echnology,26, 2281-2287. Gilmour, C. C. and Henry, E. A. 1991. Mercury methylation in aquatic systems affected by acid deposition. Environmental Pollution, 71, 131-1 69. Gilmour, C. C., Riedel, G. S., Ederington, M. C., Bell, J. T., Benoit, J. M., Gill, G. A. and Stordal, M. C. 1998. Methylmercury concentrations and production rates across a trophic gradient in the northern Everglades. Biogeochemistry, 40,327-345. Grieb, T. M., Driscoll, C. T., Gloss, S. P., Schofield, C. L., Bowie, G. L. and Porcella, D. B. 1990. Factors affecting mercury accumulation in fish in the upper Michigan peninsula. Environmental Toxicology and Chemistry, 9,919-930. Chase, M. E., Jones, S. H., Hennigar, P., Sowles, J., Harding, G. C. H., Freeman, K., Wells, P. G., Krahforst, C., Coombs, K., Crawford, R., Pederson, J. and Taylor, D. 2001. Gulfwatch: Monitoring spatial and temporal patterns of trace metal and organic contaminants in the Gulf of Maine (1 99 1- 1997) with the blue mussel, Mytilus edulis L. Marine Pollution Bulletin, 42,491-505. Hakanson, L., Andersson, T. and Nilsson, A. 1990. Mercury in fish in Swedish lakes - linkages to domestic and European sources of emission. Water, Air, and Soil Pollution, 50, 171 - 131. Hermanson, M. H. 1993. Historical accumulation of atmospherically derived pollutant trace metals in the arctic as measured in dated sediment cores. Water Science and Technology, 28,33-41. Iverfeldt, A. and Lindqvist, 0. 1986. Atmospheric oxidation of elemental mercury in the aqueous phase. Atmospheric Environment, 20,1567- 1573. Jackson, T. A. 1997. Long-range atmospheric transport of mercury to ecosystems, and the importance of anthropogenic emissions - a critical review and evaluation of the published evidence. Environmental Reviews, 5,99-120. Jackson, T. A. 1998. Mercury in aquatic ecosystems. In Metal Metabolism in Aquatic Environments (Eds, Langston, W. J. and Bebianno, M. J.) Chapman & Hall, London, pp. 77-158. Kelly, C. A., Rudd, J. W. M., St. Louis, V. L. and Heyes, A. 1995. Is total mercury concentration a good predictor of methylmercury concentration in aquatic systems? Water, Air, and Soil Pollution, 80,715-724. King, J. K., Kostka, J. E., Frischer, M. E., Saunders, F. M. and Jahnke, R. A. 2001. A quantitative relationship that demonstrates mercury methylation rates in marine sediments are based on community composition and activity of sulfate-reducing bacteria. Environmental Science and Technology, 35,249 1-2496. Laporte, J. M., Truchot, J. P., Ribeyre, F. and Boudou, A. 1997. Combined effects of water pH and salinity on the bioaccumulation of inorganic mercury and methylmercury in the shore crab Carcinus maenas. Marine Pollution Bulletin, 34, 880-893. Lawson, N. M. and Mason, R. P. 1998. Accumulation of mercury in estuarine food chains. Biogeochemistry, 40,235-247. Lin, C.-J. and Pehkonen, S. 0.1999. The chemistry of atmospheric mercury: a review. Atmospheric Environment, 33,2067-2079. Lindberg, S. E. and Stratton, J. E. 1998. Atmospheric mercury speciation: concentrations and behavior of reactive gaseous mercury in ambient air. Environmental Science and Technology, 32,49-57. Mason, R. P. and Fitzgerald, W. F. 1991. Mercury speciation in open ocean waters. Water, Air, and Soil Pollution, 56, 779-789. Mason, R. P., Fitzgerald, W. F. and Morel, F. M. M. 1994. The biogeochemical cycling of elemental mercury: Anthropogenic influences. Geochimica et Cosmochimica Act& 58, 3191-3198. Mason, R. P. and Lawrence, A. L. 1999. Concentration, distribution, and bioavailability of mercury and methylmercury in sediments of Baltimore Harbor and Chesapeake Bay, Maryland, USA. Environmental Toxicology and Chemistry, 18,2438-2447. Mason, R. P., Reinfelder, J. R. and Morel, F. M. 1996. Uptake, toxicity, and trophic transfer of mercury in a coastal diatom. Environmental Science and Technology, 30,1835-1845. Mikac, N., Picer, M., Stegnar, P. and Tusek-Znidaric, M. 1985. Mercury distribution in a polluted marine area, ratio of total mercury, methyl mercury and selenium in sediments, mussels and fish. Water Research, 19, 1387-1392. Miskirnmin, B. M. 1991. Effect of natural levels of dissolved organic carbon (DOC) on methyl mercury formation and sediment-water partitioning. Bulletin of Environmental Contamination and Toxicology, 47,743-750. New England Governors and Eastern Canadian Provinces (NEG-ECP). 1998. Mercury Action Plan. Conference of New England Governors and Eastern Canadian Premiers. Halifax, Nova Scotia. June 1998. Northeast States for Coordinated Air Use Management (NESCAUM), Northeast Waste Management Officials' Association (NEWMOA), New England Interstate Water Pollution Control Commission (NEIWPCC), Ecological Monitoring and Assessment Network (EMAN), 1998, Northeast States and Eastern Canadian Provinces Mercury Study, A Framework for Action: NESCAUM, pp. 350. Nriagu, J. 0. 1979. The Biogeochemistry of Mercury in the Environment, ElseviedNorth Holland Biomedical Press, Amsterdam, pp. 626. Nriagu, J. 0. 1993. Legacy of mercury pollution. Nature, 363,589. Nriagu, J. 0. 1994. Mercury pollution fiom the past mining of gold and silver in the Americas. The Science of the Total Environment, 149, 167- 18 1. Pai, P., Karamchandani, P. and Seigneur, C. 1997. Simulation of the regional atmospheric transport and the fate of mercury using a comprehensive Eulerian model. Atmospheric Environment, 3 l,27 17-2732. Pilgrim, W., Lucotte, M., Montgomery, S., Santos-Burgoa, C., Uriarte, M. I. I., Abascal- Garrido, F., Round, M. and Porcella, D. 1998. Meeting the Challenges of Continental Pollutant Pathways: Case Study of Mercury. Commission for Environmental Cooperation, Fredericton, New Brunswick, pp. 41. Pirrone, N., Keeler, G. J. and Nriagu, J. 0. 1996. Regional differences in worldwide emissions of mercury to the atmosphere. Atmospheric Environment, 30,2981- 2987. Ratcliffe, H. E. and Swanson, M. G. 1996. Human exposure to mercury: a critical assessment of the evidence of adverse health effects. Journal of Toxicology and Environmental Health, 49,221 -270. Schroeder, W. H., Munthe, J. and Lindqvist, 0. 1989. Cycling of mercury between water air and soil compartments of the environment. Water, Air, and Soil Pollution, 48, 337-347. Seigneur, C., Lohman, K., Pai, P., Heim, K., Mitchell, D. and Levin, L. 1999. Uncertainty analysis of regional mercury exposure. Water, Air, and Soil Pollution, 112, 151-162. Seigneur, C., Pai, P., Gerath, M., Mitchell, D., Hamby, G., Gong, G., Whipple, C. and Levin, L. 1997. Probabilistic assessment of regional mercury exposure. Water, Air, and Soil Pollution, 97, 159-168. Spry, D. J. and Wiener, J. G. 1991. Metal bioavailability and toxicity to fish in low- alkalinity lakes: a critical review. Environmental Pollution, 71,243-304. USEPA. 1997. Mercury Study Report to Congress. Office of Air Quality Planning and Standards, United States Environmental Protection Agency, Washington, D.C. Watras, C. J. and Bloom, N. S. 1992. Mercury and methylmercury in individual zooplankton: implications for bioaccumulation. Limnology and Oceanography, 37, 1313-1318. Wells, P.G., Keizer, P.D., Martin, J.L., Yeats, P.A., Ellis, K.M. and Johnston, D.W. 1997. The chemical environment of the Bay of Fundy. In: Bay of Fundy Issues: A Scientific Overview. Environment Canada - Atlantic Region, Occasional Report No. 8 (Eds, Percy, J.A., Wells, P.G. and Evans, A.) Environment Canada, Dartmouth, Nova Scotia, pp. 37-61. CHAPTER 2 Estimating the Anthropogenic Component of Atmospheric Mercury Loading in the Bay of Fundy, Canada 2.1 Abstract High levels of mercury in wildlife, fish and avifauna are an ongoing problem in the Bay of Fundy region of Canada, despite apparent declines in local and regional emissions. Bog, lake and coastal salt marsh sediment cores collected fiom the Bay of Fundy region were evaluated in this study for disturbance through diagenetic processes. Using the sediment cores that best preserved historical trends in mercury loading, a background deposition rate of approximately 4.1 pg mm2yil was calculated. Given the annual rate of mercury deposition measured in the region (12.8 pg m-2yr-'), this leaves an anthropogenic contribution of 8.7 pg m-2wl. Recent declines in mercury concentrations are evident in both salt marsh and lake sediment cores. Earlier peaks in the late 1800s and early l9OOs that correspond to peak emissions in North America from gold and silver amalgamation are also observed. Based on the sediment records, overall declines in 25 atmospheric loading of mercury are only achieved in the Bay of Fundy region when emissions are reduced on both the regional and continental scales. 2.2 Introduction Mercury is a potent environmental toxin that bioaccumulates in aquatic organisms (Jackson, 1998). Many human activities such as fossil fuel combustion, waste incineration, and base metal smelting release large quantities of mercury to the atmosphere. Over half of the mercury released globally fiom anthropogenic sources is in the gaseous H~Oform, which can be stable in the atmosphere for up to a year (Lin and Pehkonen, 1999, Pacyna and Pacyna, 2002). The propensity of this element for long- range transport and deposition results in contamination of regions far fiom point sources of mercury releases (Fitzgerald et al., 1998, Hermanson, 1998, Jackson, 1997). This problem is particularly relevant in the Bay of Fundy region of Canada where continental airmasses deposit mercury fiom industrialized regions of central Canada and the Northeastern United States (Beauchamp, 1998). Despite large reductions in local and regional scale emissions, high levels of mercury in fish, wildlife and avifauna are an ongoing problem in lakes, rivers and coastal regions surrounding the Bay of Fundy and represent a potential health threat to humans consuming fish from this region (NESCAUM et al., 1998, Peterson et al., 1989). The relationship between anthropogenic mercury releases and resulting concentrations in aquatic ecosystems is confounded by the fact that mercury is a naturally occurring element and is ubiquitous in the environment (Nriagu, 1979a). The potential significance of atmospheric deposition for determining concentrations in organisms is illustrated by recent experimental work in the Experimental Lakes Area of Northern Ontario, which showed that newly deposited mercury was more "bioavailable" than older or "native" mercury in the system (Hintelmanu et al., 2002). However, the continual deposition and re-emission of anthropogenic mercury released throughout history (e.g., the "grasshopper effect") is a significant source of ongoing pollution in many areas and is thought to account for up to one third of the total reservoir of mercury in the atmosphere (Pirrone et al., 1996). To manage the effects of mercury in the environment, it is therefore necessary to consider not only present sources of contamination but also the cumulative burden of mercury released throughout human history. A full understanding of how regulating existing sources of mercury contamination will affect the magnitude of inputs from the atmosphere is only possible when the relationship between emissions and deposition has been quantified. This relationship can be elucidated by identifying trends in historical deposition that covary with anthropogenic mercury emissions peaks and by quantifying the anthropogenic component of atmospheric mercury loading. In the past, examination of sedimentary records of historical deposition has been widely applied as a method for investigating changes in anthropogenic loading of mercury. Vertical profiles of mercury concentrations in sediments are often used to make inferences about long-term trends in mercury deposition and to discern the natural or "background" loading rate in a given region. However, mercury data from sediment cores may be incorrectly interpreted as records of historical deposition if internal processes controlling mercury accumulation and distribution are not taken into account. For instance, Rasmussen (1994) points out that enrichment of mercury in the surface layers of many sediment cores that has been attributed to anthropogenic contamination may actually be the result of remobilization and upward diffusion of mercury in the sediment column. Failure to correct for processes that disturb the sedimentary record results in a wide range of supposed "enrichment factors" for anthropogenic mercury within relatively small geographic regions. For example, in Maritime Canada, depending on the bog or lake sediment core studied, the estimated anthropogenic enrichment in atmospheric mercury loading ranges between a factor of two to fifteen times the natural background loading rate prior to human influences (Lamborg et al, 2002, NESCAUM et al., 1998, Rutherford and Matthews, 1998). This type of variability makes it difficult to constrain estimates of historical loading and reinforces the need for consistent methods and "careful scrutiny" of all sedimentary data used to interpret trends in mercury deposition over time (Electric Power Research Institute, 1996, Farmer, 1991, Landers et al., 1998, Rasmussen, 1994). Studies that do not systematically account for the physical, geochemical and biological processes that may alter the vertical profiles of mercury in sediments cannot distinguish changes in historical pollution fiom other sources of variability. The purpose of this paper is to: (i) estimate the anthropogenic component of atmospheric mercury deposition in the Bay of Fundy region of Canada, and (ii) evaluate trends in historical mercury emissions and deposition to determine the relationship between regional and continental scale emissions reductions and overall mercury loading in this region. To do this, mercury concentration data recorded in bog, lake and coastal salt marsh sediment cores fiom the Bay of Fundy region are evaluated. However, the utility of these data as archives of historical mercury loading is contingent on rejection of the hypothesis that the observed mercury profiles can be attributed to reworking of the sediments andlor redox induced changes in the distribution and mobility of mercury in the sediment column. In addition, mercury loading rates obtained from bog, lake and salt marsh sediments can only be extrapolated to infer absolute rates of atmospheric mercury loading when the advective losses of mercury are minimal and the atmospheric component of total deposition can be quantified. We propose that the consistent application of this fiarnework for evaluating sedimentary data would allow the relative significance of anthropogenic and natural sources of mercury to be quantified with more confidence. 2.3 Methods Bog and lake sediment data discussed in this study are from Rutherford and Matthews (1998) and Kainz et al. (1997) who detail their sampling and analytical procedures in their reports. Rutherford and Matthews (1998) collected triplicate cores in 1993 fiom Caribou Plains Bog, an ombrotrophic (raised) bog in Fundy National Park, New Brunswick (Fig 2- 1). Kainz et al. (1997) obtained sediment cores in 1996 fiom two headwater lakes located at the mouth of the Bay of Fundy: Lily Lake and St. Patrick's Lake (Fig 2-1). Both Lily Lake and St Patrick's Lake are in relatively undeveloped regions of Maritime Canada and have small drainage ratios of 4: 1 and 8: 1, respectively, indicating that the majority of mercury inputs should be atmospheric in nature (Kainz et al., 1997). L r C Coastal marsh sediment cores (n=4) were collected fkom Dipper Harbour (n=2), Chance w F Harbour and Bocabec Marsh between 1994 and 1996 (Fig 2- 1). All sampling locations ra S i are located on the New Brunswick coastline, near the mouth of the Bay of Fundy. Sediment accumulation in these marshes is determined by the productivity and decomposition of surface vegetation, and inorganic sediment inputs from tides (Chmura et al., 2001). The mean tidal range in this area is between 6-8 m (Gregory et al., 1993). The areas adjacent to the salt marshes are dominated by second growth spruce and pine forest. There is limited residential and commercial development in the region. Coring locations within Dipper Harbour and Chance Harbour salt marshes are described in detail by Daoust et al. (1996) and Chrnura et al. (2001). Based on vegetation indicators, all cores were collected in the high marsh zone according to the classification of Redfield (1972). At this elevation the sampling sites were exposed to <25% of the flood tides, thus providing better signals of atmospheric deposition than low marsh regions that are flooded diurnally (Chmura et al., 1997, Chrnura et al., 200 1). Salt marsh cores were obtained using a modified Hargis corer (Hargis and Twilley, 1994) to depths of 30-40 cm. Cores were sub-sectioned into 0.5-1.0 cm intervals, put on ice and returned to the laboratory for analysis. Samples were freeze-dried, homogenized using a mortar and pestle and analyzed for radionuclides, metals and organic carbon. Organic carbon content was estimated from loss on ignition (LOI) of duplicate or triplicate samples by heating the sample at 550•‹Covernight. Percent organic carbon content of sediments was calculated from LO1 using the relation reported by Craft et al. (1991). Details of dating techniques for cores from Chance Harbour and Dipper Harbour are outlined in Chrnura et al. (2001). Dating was determined through the combination of 2'0~band '37~sradionuclide analyses and pollen analysis. The Bocabec marsh core was dated using anthropogenic lead. This method is less accurate than isotope dating but has been applied successfully in other studies (Clegg, 2000, Weiss et al., 1999). ACME Analytical Laboratory of Vancouver performed trace metal analyses for the Dipper Harbour and Bocabec cores. The Bocabec core was analyzed for metals at coarser vertical intervals (5-10 cm) to a depth of 60 cm. Samples were digested with 30 m13: 1:2 HC1:HN03:H20solution at 95OC for one hour, then diluted with 100 ml of water. Total mercury was extracted with milbk-aliquat 336 and all metals were analyzed with inductively-plasma atomic emission (ICP) spectroscopy using mineral soil certified standards obtained from the National Research Council of Canada. The reported detection limit of this procedure for total mercury (Hg) was 10 ng g-', which is notably high relative to other methods that are in the picomolar range (Gill and Fitzgerald, 1987). To test the accuracy of these data, replicate samples were analyzed at random by Flett Research Ltd., Winnipeg, Manitoba who employ low level methods, to verify the precision of Hg analyses conducted by ACME Labs Inc. The coefficient of variation for Hg analysis between labs averaged 29%, which is reasonable agreement between techniques as overall trends in Hg fluxes are still discernable with this degree of uncertainty. Details of additional metals analysis (e.g., Fe, Mn) are reported by Clegg (2000). Chance Harbour samples were analyzed for total mercury by digestion in concentrated 5:2 nitric-sulfuric acid solution and oxidation with bromine monochloride (BrC1) under Class 100 conditions. Immediately prior to analysis, the excess bromine was neutralized with 10% hydroxylamine hydrochloride. Mercury was volatilized fi-om sample digests with stannous chloride, purged, and trapped on gold-coated sand columns. Quantification was by dual-stage gold amalgamation and cold-vapor atomic fluorescence spectroscopy (CVAFS). This procedure was based on EPA Method 1631, Gill and Fitzgerald (1987) and Bloom (1989). The method detection limit (MDL) for total mercury in sediment solids was 0.19 ng g-' (n=19), determined as three times the standard deviation of the mean of the sample blanks. Precision, measured as the relative percent difference (RPD) between digest duplicates, was 9.6% (n=24 pairs). Calibration curves of at least 3 = 0.99 were achieved daily or samples were re-run. Accuracy was measured both by spike recoveries and using the MESS-3 marine sediment certified reference material (91 * 9 ng g'') from the National Research Council of Canada. Recoveries averaged 103% + 10% (n=12) and 92 * 16 ng g-' for all MESS-3 samples (n=9). Samples from runs with poor recoveries (40%) were re-analyzed. 2.4 Results and Discussion The stratigraphic profiles of mercury concentrations in dated sediment cores from the Bay of Fundy region are presented in Figure 2-2. Data from Dipper Habour salt marsh, St. Patrick's Lake and Caribou Plain's bog all show a surficial maximum in loading/concentrations, suggesting that either inputs through atmospheric deposition are at their maximum level in recent years or that redox related mobility of mercury in these cores has resulted in diagenetic remobilization and precipitation at the sediment surface. In contrast, cores fiom Lily Lake, Chance Harbour and Bocabec salt marsh all have subsurface peaks and suggest that loading rates have declined in recent years. Distinct peaks in mercury concentrations that may indicate local, regional andfor global pollution events are distinguishable in the salt marsh and bog cores, but not in the lake sediments. Figure 2-2. Summary of mercury concentration data from dated sediment cores collected in the Bay of Fundy region. Compaction of the sediment with depth in Caribou Plains bog accounts for the discrepancy between mercury concentration data and loading rates that take into account the sediment bulk density. Dipper Harbour Chance Harbour Bocabec Dipper Harour Salt Marsh Core A Core B Lily Lake St. Patrick's Lake Caribou Plains Bog* - (mean of 3 cores) 1800 IIIII 25 0 8 16 0 80 160 Hg Loading (ug m-* yrl) Hg (ng g-'1 Overall, there is little consistent variability in mercury concentrations among cores and it is not possible to identify the cores that best reflect historical atmospheric deposition of mercury without further analysis. These examples reaffirm the need to test the hypothesis that sediments fiom lakes, bogs and salt marshes in the Bay of Fundy region are reliable indicators of historical pollution by looking at the evidence for diagenetic remobilization of mercury in the sediment column prior to making conclusions regarding historical pollution. In the following sections we analyze the effects of diagenesis in these sediments and identify cores where the stratigraphic profile of mercury concentrations has been significantly influenced by post-depositional mobility. Trends in historical deposition, ea- mercury loading rates and anthropogenic enrichment factors are interpreted from the r- remaining data and compared to global and regional mercury emissions inventories to firther evaluate the relationship between anthropogenic mercury emissions and deposition in the Bay of Fundy region. 2.4.1 Physical and Biological Redistribution Identifying anomalies in radionuclide profiles provides a simple method for determining the depth of physical and biological disturbance in sediment cores (Benniger et al., 1979, Charles and Hites, 1987, Nittrouer et al., 198311984). The "ideal" profile of '''~b is shown as the linear decrease in the natural logarithm of the unsupported activity with depth in Figure 2-3. Figure 2-3. Radionuclide data used to date salt marsh sediments also indicate disturbances through physical andor biological perturbations. The "ideal" decay of 210~b is shown by the straight line next to measured unsupported isotope activity. Dipper Harbour Core A 16 - - Core A 20 I IIIIIII 20 3.5 4 4.5 5 5.5 6 0 20 40 60 80 In excess 210Pb(Bq kg-') 137Cs(Bq kg-') Dipper Harbour Core B 0 + 4 h wg8 5 5 12 12 n n 16 16 4 4.4 4.8 5.2 5.6 6 0 20 40 60 80 In ex~ess~~OPb(Bq kg-') 137Cs(Bq kg-') Chance Harbour 123456 0 20 40 60 80 In excess2I0Pb(Bq kg-') 137Cs(Bq kgv1) A perfectly preserved 137~sprofile appears as a well-constrained peak beginning in 1950, with an overall maximum in 1963, followed by a rapid decline in concentrations corresponding to the phase-out of nuclear weapons testing (Charles and Hites, 1987). Cores from Dipper Harbour (core A), Chance Harbour and Lily Lake match the expected profile (Figs 2-3 and 2-4). In contrast, cores from St. Patrick's Lake and Dipper Harbour (core B) both show evidence of physical and/or biological disturbance (Figs 2-3 and 2-4). Mixing within the sediment column is indicated in the Dipper Harbour salt marsh (core B) by the more drawn out profile of 137~sactivity between the depths of 8-10 cm and constant activity in the unsupported 210~bprofile that does not match the predicted exponential decay (Fig 2- 3). The 210~bprofile from St. Patrick's Lake indicates redistribution of the sediments to a depth of -7 cm (Fig 2-4). Kainz et al. (1 997) attributed the observed pattern to bioturbation at the sediment water interface and/or physical disturbances that may have occurred during sampling. However, mixing of 210~bat lower depths in St. Patrick's Lake sediments suggests that either physical or biological turbation naturally occurs in this system. Overall, this analysis indicates that the sediment cores from Dipper Harbour salt marsh (core B) and St. Patrick's Lake likely do not accurately reflect the historical profile of mercury contamination in southwestern New Brunswick. Although bog sediments are subject to less disturbance through physical and biological mixing, a number of studies have noted that 210~bisotopes migrate in the sediment column (Pakarinen and Tolonen, 1977, Shotyk, 1988, Urban et al., 1990). Figure 2-4. Radionuclide and redox data for lake core sediments collected in the Bay of Fundy region. The "ideal" deca of 2'0~bis shown by the straight line next to measured unsupported isotope activity. "'Cs data were not available for these cores. These data are used to indicate disturbances through physical andlor biological perturbations, as well as diffusion of mercury within the sediment cores. Significant correlations between mercury (Hg) and redox potential (Eh) are indicated by Spearman rank correlation coefficents (r,) with p-values of ~0.05.All data presented were obtained from Kainz et al. (1997). St. Patrick's Lake Lilv Lake 0 2 4 6 8 2345678 In excess 210Pb(Bq kg-l) In excess 210Pb(Bq kg-l) Redox Potential (Eh) Bog cores from Maritime Canada were dated using 210~b(Rutherford and Matthews, 1998) but, because these data are not reported, we cannot comment on their reliability. However, based on the examples of lake and salt marsh cores given above, without this type of simple analysis it is difficult to distinguish anomalies in the mercury profiles caused by physical mixing from those caused by historical pollution. 2.4.2 Redox Induced Changes We tested the hypothesis that stratigraphic profiles of mercury in lake and salt marsh sediments are the result of redox induced changes in mercury mobility by looking at covariation of mercury with Fe, Mn and redox potential (Eh). Fe and Mn are good proxy indicators of the redox profile in salt marsh sediments because mercury is strongly sorbed to the surface of Fe and Mn hydroxides and oxides in oxic surface sediments, but is released under reducing conditions (Benoit et al., 1998, Bricker, 1993, Gobeil and Cossa, 1993). If the distribution of mercury in cores from the Bay of Fundy region can be corroborated with the major redox gradients within the sediment column indicated by profiles of Fe, Mn andor Eh, then the observed concentrations of mercury may be a function of fluctuations in the oxidation status of sediments. Accordingly, diffision of mercury would result in its upward movement and precipitation at the sediment surface, where it is re-associated with Fe and Mn oxides and hydroxides and can then be misconstrued as a peak in anthropogenic pollution. The results of this analysis indicate that there are significant correlations between redox status and mercury concentrations in cores from Dipper Harbour (core B) salt marsh, Bocabec salt-marsh and St. Patrick's Lake but there is no significant correlation between these variables in the Dipper Harbour (core A) and Lily Lake cores (Figs 2-4 and 2-5). Thus, we cannot reject the hypothesis that in St. Patrick's Lake, Bocabec marsh and Dipper Harbour (core B) sediments the observed profile of mercury concentrations is the result of redox related processes. Therefore, it is likely that these cores are not good indicators of historical pollution. No data were available on the redox profiles in cores from Chance Harbour and Caribou Plains Bog, limiting our confidence in these data because we cannot dismiss the hypothesis that diagenetic factors control the observed mercury distribution. Qualitatively, the concentration of mercury in surface sediments of Dipper Harbour salt marsh (core A) appears to be associated with precipitation of Fe and Mn hydroxides, however, this correlation is not statistically significant (Fig 2-5). Where possible, it is important to develop and apply methods that actually quantify the significance of such processes. The method described by Gobeil and Cossa (1993) was used to quantify the diffusive flux of mercury (J) in the Dipper Harbour salt marsh (core A) by applying Fick's first law of diffusion: Where: z (cm) = depth (+ downwards); C (ng ~rn-~)= concentration of Hg (mass/volume); Zl& = maximum concentration gradient of Hg in sediment; 0 = porosity of the sediments; and D, = whole sediment diffusion coefficient. It was also assumed that: Ds = m2Do (Gobeil and ~ossa,1993, Ullman and Aller, 1982) where: Do, the molecular diffusion coefficient for mercury in seawater. Figure 2-5. Variability in mercury (Hg) profiles in salt marsh sediments as a function of redox related transformations in the sediment column. DH-A = Dipper Harbour salt marsh core A; DH-B = Dipper Harbour salt marsh core B. Correlations between Hg and Fe/Mn suggest that changes in the oxidation status of sediments may be affecting the mobility of mercury within the sediment core. Significant correlations are indicated by Spearman rank correlation coefficents (r,) with p-values of Fe (mg g-'1 Bocabec Salt Marsh Core (BM) A value of 5 x 10" cm2 s" for Do for total mercury was selected based on measurements in Bellingham Bay sediments (Bother et al., 1980). This value is in reasonable agreement with more recent work in Lavaca Bay, Texas where Do for inorganic mercury species was estimated at 9.5 x loa cm2 s-' and 2 x cm2 s-' for inorganic mercury associated with colloidal matter (Gill et al., 1999). The maximum Hg gradient in this core occurs between the surface and 3 cm depth. The concentration of mercury in pore waters was estimated from the partition coefficient &= 3.6+ 0.2) for inorganic mercury measured in Bay of Fundy sediments (Chapter 3). The porosity of marsh sediments was estimated to be approximately 60% (T. Moore, Personal Communications, 1996). A number of assumptions were necessary to complete this calculation for the salt marsh sediments. The limitations of this analysis are determined by uncertainty regarding the value of the molecular diffusion coefficient for mercury in seawater (Do), averaging of depth intervals over which diffusion occurs, and the approximation of sediment porosity values. Because of these uncertainties, the calculations presented should only be treated as approximate estimates of potential fluxes through diffusion relative to total deposition. Based on these calculations, the upward diffusion of mercury in the surface sediments of Dipper Harbour (core A) was estimated to be 3 pg m-*yr-'. This amount likely does not make a significant contribution to elevated concentrations of mercury at the sediment surface, as it is only 4% of the total loading rate measured in this core (76 pg mm2yt'). Such types of simple calculations should be incorporated into future studies that use sediments as records of historical deposition to systematically dismiss alternate hypotheses regarding the factors controlling the observed profile of mercury concentrations. 2.4.3 Ouantitative Retention of Mercury in the Sediments (Advective Fluxes of HpJ An analysis of advective processes controlling mercury inputs and removal fkom lakes, coastal salt marsh and bog systems is necessary to determine whether absolute rates of atmospheric mercury loading can be reliably estimated fkom the sedimentary data. Calculations that estimate the fiaction of mercury in the sediments derived fkom atmospheric deposition can help to establish the utility of these cores for indicating trends in atmospheric pollution. It is therefore necessary to evaluate whether mercury is "quantitatively retained" within the sediment column (Benoit et al., 1998) by looking at differences between the expected fluxes of mercury to the sediments and the observed rate of mercury accumulation. The quantitative retention of mercury within the sediments of the lakes sampled in southwestern New Brunswick was not evaluated by Kainz (1997) and loading rates for these sediments cannot be calculated without sediment bulk density values. Other studies show that absolute loading rates can only be estimated fkom lake sediments when multiple cores are collected throughout the lake basin to account for differences in accumulation rates or sediment "focusing" among regions (Charles and Hites, 1987). However, temporal trends should be preserved in lake sediments that do not show significant evidence of bioturbation and redistribution of the sediments (e.g., Lily Lake). The relative contributions of tidal and atmospheric inputs of mercury to Dipper Harbour (core A) and Chance Harbour salt marsh sediments are estimated by comparing accumulation rates of inorganic and organic materials (Table 2-1). Assuming that the majority of mercury in flooding tidal waters is partitioned onto solid particles, then the contribution from tidal inputs can be calculated fiom the average mercury concentration on suspended sediments and the inorganic deposition rate. Following the method of Bricker (1993), the input of inorganic sediment to each marsh is calculated fiom the product of average bulk density in surface sediments, sediment accretion rate and fraction of inorganic materials in the sediments indicated by loss on ignition values. The annual inputs of mercury from tidal sources are estimated by multiplying inorganic bulk density by the sediment accretion rate for each core and the average mercury concentration in suspended sediments of the tidal waters (60 ng g-'). Atmospheric inputs are estimated from measured inputs of mercury in precipitation (8.5 pg m-* yr -') and by assuming that dry deposition is approximately equal to 50% of wet deposition (Fitzgerald et al., 1994, Lindqvist et al., 1991). Mercury inputs fkom tidal and atmospheric sources are then compared to measured rates of mercury loading in the upper 5-10 cm of the marsh cores. The results of these calculations (Table 2-1) show that in both the Dipper Harbour and Chance Harbour marsh cores, tidal and atmospheric inputs comprise over 90% of the total loading. However, atmospheric inputs are relatively smaller than tidal sources, making up between 26% and 39% of the total loading in Dipper Harbour (core A) and Chance Harbour cores respectively. Table 2-1. Estimated tidal and atmospheric contributions to mercury loading in Dipper Harbour and Chance Harbour salt marsh sediments. Dipper Harbour Chance (core DH-A) Harbour Marsh Averaged depth (cm) - z 5.0 5.5 Bulk density (g ~rn'~)- BD 0.46 0.29 Loss on ignition (%) - LOI Sediment accretion rate (cm yr-I)- AR Inorganic bulk density (g ~m-~) BDI = BD * LOI * AR Est. Hg conc. suspended sediment ' (ng g-') - Css Estimated inorganic Hg deposition (ng ~rn-~yr-') Inorg = Css* BDI Measured atmospheric inputs2 (ng cm-2yr-') - Atm Hg in marsh surface sediments (ng.g-') - CBS Estimated total deposition (ng cm-2yr-') Total = CBs* BD * AR Tidal contributions (%) Tidal% = Inorg/Total Atmospheric contributions (%) Ah% = Atm/Total Remaining Hg (%) = (Total - Inorg - Ah) /Total 9 2 Data from Loring (1 98 1) Data fiom Beauchamp (1998) The fraction of total mercury loading that is not accounted for in these cores (noted as "remaining" Hg in Table 2-1) is small and could be the result of measurement error and/or variability in the concentration of mercury in surface sediments, suspended sediments and precipitation and/or assumptions regarding the magnitude of dry deposition in the region. Despite large contributions from tidal waters, the remote locations of these marshes combined with minimal historic sources in the region suggests that the salt marsh sediments would still preserve the relative trends in atmospheric mercury deposition if the main source of mercury to estuarine waters is from the atmosphere. However, this analysis indicates that loading rates are sensitive to historic variability in mercury concentrations in incoming tidal waters. Rutherford and Matthews (1998) did not analyze or present the data needed to assess the effects of diagenetic processes and advective losses on the profiles of mercury loading observed in the Caribou Plains Bog. Hence, the reliability of these data as indicators of historical atmospheric deposition is equivocal. There is significant variability in mercury concentration data within bogs in Maritime Canada that do not correspond to changes in local or regional emissions sources (Rutherford and Matthews, 1998). Past research shows that in general approximately 15% of atmospheric mercury inputs in bog ecosystems are lost in runoff and surface volatilization and that differences in accumulation rates between sites do not necessarily indicate differences in atmospheric deposition (Roos-Barraclough and Shotyk, 2003). Thus, sediment data from Caribou Plains bog must only be used to indicate relative trends in atmospheric deposition rather than absolute loading rates. The results of this analysis suggest that sediment cores from Lily Lake, Dipper Harbour (core A), Chance Harbour and Caribou Plains Bog are all potentially usell for interpreting trends in historical mercury deposition in the Bay of Fundy region. The validity of the observed mercury profiles as records of historical pollution in some of these cores, particularly the core from Caribou Plains Bog, remains uncertain because few of the data needed to assess the effects of diagenetic remobilization of mercury in the sediments were available. This uncertainty limits the strength of conclusions that can be drawn from these cores. 2.4.4 Anthropogenic Component of Atmospheric De~osition The mercury concentration or loading of mercury in surface sediments divided by the mean value prior to human influence is often represented as an "anthropogenic sediment enrichment factor" (ASEF). ASEFs calculated from both mercury concentrations and loading rates in cores from the Bay of Fundy region ranged from 2.8 to 15 (Table 2-2). The high value of 15 was calculated in Caribou Plains Bog, where a complete analysis of controls on stratigraphic variability in mercury concentrations was not completed. Thus, this value must be treated as uncertain. ASEFs are more consistent among the remaining cores (Lily Lake, Dipper Harbour Core A and Chance Harbour), ranging between 2.1 and 4.8 (Table 2-2). These values correspond with other studies that show a global scale increase in atmospheric mercury deposition of 2-4 times the pre-industrial levels (Engstrom and Swain, 1997, Johnson et al., 1986, Lucotte et al., 1995, Swain et al., 1992) and regional studies of bog and lake sediments in central Nova Scotia that showed a five-fold enrichment over pre-industrial Table 2-2. Summary of mercury data obtained from dated sediment cores in the Bay of Fundy region of Canada. Data fiom ombrotrophic bog sediments are from Rutherford and Matthews (1998) and lake sediment data fiom Kainz et al. (1997). ASEF = anthropogenic sediment enrichment factors calculated by dividing mercury concentrations andlor loading rates in the surface horizons of sediment cores by the average pre-industrial concentrations/loading in sediments that accumulated prior to 1880. System Year ASEF Pre-industrial Earliest Max. Peak Sampled (modern 1 pre- Loading Rate increase Loading Yr(s) industrial) (pg.m-2.w1) (I% In-2 -1 Caribou 1993 -1 5 (loading) 1.4 -1900 19 -1 992 Plains Bog Lily Lake 1996 2.8 (concentration) N/A -1880 N/A -1988 Dipper 1994 2.8 (loading) 27.4 -1900 76.2 1988-94 Harbour 4.8 (concentration) 68.1 1938-44 ( core A) Chance 1996 2.1 (loading) 10.5 -1860 37.6 1988-92 Harbour 3.6 (concentration) 44.9 1936-42 122 1875-80 mercury deposition rates (Lamborg et al., 2002). Visual inspection of the remaining data reveals that the earliest observable increase in mercury concentrations and loading rates is also fairly consistent among these cores, ranging between the years 1860 and 1900 (Fig 2-2). The actual magnitude of atmospheric deposition in this region is more difficult to estimate from the sedimentary data. Data needed to estimate loading rates were not available for Lily Lake and loading data from Caribou Plains bog are tenuous as recent studies show significant losses of mercury occur through advective fluxes in bog ecosystems (Roos-Barraclough and Shotyk, 2003). Pre-industrial loading rates were calculated fiom both Dipper Harbour and Chance Harbour salt marsh data, however, tidal sources contribute substantial amounts of mercury to these cores (Table 2-1). Loading also appears to be somewhat higher in Dipper Harbour than Chance Harbour. However, this is likely a reflection of the detection limit of the analytical technique used to measure concentrations in the Dipper Harbour, which is less reliable at lower mercury concentrations. Hence, the Chance Harbour data likely provide a more accurate reflection of absolute loading rates. The pre-industrial loading rate in Chance Harbour salt marsh multiplied by the atmospheric contribution to total loading of 39% (Table 2-I), gives a "background" deposition rate of approximately 4.1 pg m-2 yf'. This calculation relies on the assumption that the relative contributions of tidal and atmospheric sources have remained relatively constant over time. This loading rate is only slightly higher than the estimate based on lake core data from central Nova Scotia of 3.0 pg m-2yr-' (Lamborg et al. 2002) and other studies throughout North America that have estimated natural fluxes of mercury to be between 2.0 and 3.7 pg m'2yr-' (Hermanson, 1993, Hermanson, 1998, Lockhart et al., 1995, Swain et al., 1992). Subtracting the estimated natural component of mercury inputs in the Bay of Fundy region fiom the measured annual mercury deposition rate (Table 2-3) leaves an anthropogenic contribution of approximately 8.7 pg m-2wl. The estimated rate of total atmospheric mercury deposition (12.8 pg m-2 yr-') in the Bay of Fundy region was based on atmospheric monitoring data and the assumption that dry deposition comprises approximately 50% of wet deposition. This rate is slightly greater than the total annual deposition estimated fiom bog and lake sediments from Nova Scotia of 11 pg m-2 yr-' (Lamborg et al., 2002). These differences are likely a reflection of uncertainty in both the magnitude of dry deposition and spatial variability in deposition rates between study sites. 2.4.5 Relationship between Mercurv Emissions and Deposition ASEFs calculated in Lily Lake, Dipper Harbour (core A) and Chance Harbour sediments between 1880 and 2000 are presented in Figures 2-6 and 2-7. In Lily Lake there is a relatively steady increase in mercury concentrations/enrichment beginning in the late 1800s up to a maximum around 1988, followed by a decline in recent years (Fig 2-6). In Dipper Harbour and Chance Harbour salt marshes the data are more variable showing several local maxima in the late 1930s and 1940s. Table 2-3. Estimated mercury deposition rate in Maritime Canada based on historical emissions and precipitation data. These estimates are compared to sedimentary data from the Bay of Fundy region. Current deposition rates calculated from sediment cores should be in the range of the data presented if a reliable history of mercury loading is obtained. Description Value Source Wet deposition -8.5 pg m-2yr-' (Beauchamp, 1998) Surface area Maritime Canada (Stanford, 1977) Total annual deposition ~stimated' Anthropogenic emissions in Maritime Canada (Sunderland (1995) and Chmura, 2000) Estimated contribution of local sources to (Sunderland deposition in Maritime canada2 and Chmura, 2000) Natural fluxes of mercury This study Recycled fluxes and long-range sources (Sunderland and Chmura, 2000) Estimated anthropogenic component of 8.7 pg m-2yr-' This study atmospheric deposition in Bay of Fundy region 'Assuming dry deposition is 50% of wet deposition (Lindqvist and Rodhe, 1991, Fitzgerald et al., 1994) 2~ssumingthat 50% of anthropogenic emissions are deposited locally/regionally and 50% is transported outside of the source region (Mason et al., 1994, Expert Panel on Mercury Atmospheric Processes, 1994) 3~ssumingnatural fluxes are constant throughout history and are represented by the loading of mercury to sediments prior to human influence. Figure 2-6. Anthropogenic sediment enrichment factors (ASEFs) calculated for Dipper Harbour Salt Marsh (core A) and Lily Lake sediment cores. Enrichment factors were calculated by dividing the sediment mercury concentration in successive sediment layers by the "background" level represented by the average concentration in sediments that accumulated prior to 1880. Differences among cores likely reflect variability in the temporal resolution of data obtained fiom lake and salt marsh sediments. ASEF (dern/b&gm, 1 2000 1 2000 1 Dipper Harbour 1 Salt Marsh (Core A) Figure 2-7. Correspondence between mercury loading measured in Chance Harbour salt marsh sediments and anthropogenic mercury emissions peaks. The dashed line in the figure shows historical emissions in Maritime Canada, while loading rates measured in the sediments are depicted by the solid line. Mercury loading in these sediments is calculated as the product of measured mercury concentrations (Fig 2-2), sediment bulk density and the average sediment accretion rate (0.17 cm yt') calculated fiom '37~sand 210~bdata. Mercury emissions data for Maritime Canada are fiom Sunderland and Chrnura (2000), while North American emissions data shown as the table in the far right comer of the figure are fiom Pirrone et al. (1998). Atmospheric Emissions (kg yrl) ASEF (modem/pd,,,,,,~, ) 0 1000 2000 3000 I I 2000 I I I >\f__Declines in recent I emissions and deposition - 1960 increase in - emissions 1920 Overall maximum emissions Mar. - / Canada, mainly from Chlor-alkali 1 industry ca. 1970 1 Peak is likely the result of globallnatural emissions peak from gold & silver amalgamation (Peak in NA emissions ca. I887 at 1708 t.yrl) and natural volcanic Hg from Krakatau (1883) 0 40 80 1 20 Hg Loading (ug .m-*.yrl) Year t.pl Chance Harbour Salt Marsh 1879 1780 NA Emissions from 920 940 HH~Loading Pirrone et al., 1998 1947 - 247 ------Anthropogenic Emissions 1970 325 in Maritime Canada 1989 330 The highest ASEF in Dipper Harbour (core A) is observed between 1988 and 1994 (Fig 2-6). In Chance Harbour, there is also evidence of a large subsurface peak in Hg loading around 1880, followed by local peaks in the 1930s and late 1980s and a decline in recent years (Fig 2-7). The degree to which historical pollution events are distinguishable in these cores is largely determined by the temporal resolution of mercury concentration data. Differences in the temporal resolution of mercury data also help to explain variability in the enrichment profiles among cores. For example, concentration peaks in the late 1800s shown in Figure 2-7 were only picked up in the Chance Harbour sediments as they may have been missed or integrated within the depth intervals studied in the Dipper Harbour and Lily Lake cores. The temporal resolution of sedimentary data is a function of the sediment accretion rate and the depth interval homogenized for analysis. In this study, the resolution of data is highest (max. -3 years) in Chance Harbour and lowest in Lily Lake. In Chance Harbour, sediments were divided into 0.5 cm depth intervals for analysis and the accretion rate in this system was approximately 0.17 cm yr-I. In comparison, Dipper Harbour sediments were homogenized over 1 cm intervals and the accumulation rate was approximately 0.15 cm yr-I,which allows a maximum resolution of 7 years. In Lily Lake, the accumulation rate of sediments was somewhat lower than in the marsh systems (0.059 cm yr-I) and sediments were homogenized over 1-2 cm intervals (Kainz et al., 1997). Thus, the maximum temporal resolution in Lily Lake sediments is between 17-34 years, which may in-part explain why the profile of historical deposition appears relatively smoother in Lily Lake relative to the salt-marsh systems studied. In reality, the resolution of both lake and marsh data may be even lower than these estimates due to small surface disturbances and slight compaction of sediments during core extrusion. This analysis emphasizes the importance of maximizing the resolution of data in all sedimentary studies, which is most easily accomplished by analyzing relatively narrow depth intervals of sediments. Comparing trends in historical mercury loading recorded in sediments to historical emissions data fiom Maritime Canada and North America (Fig 2-7) helps to distinguish the relative importance of local and long-range pollutant sources in the Bay of Fundy region. Recent declines in mercury concentrations/loading in surface sediments were apparent in both Chance Harbour (Fig 2-7) and Lily Lake (Figs 2-2; 2-6). These data correspond to declines in emissions in Maritime Canada fiom their peak levels in 1970 (Fig 2-7) and are consistent with data fiom other regions (Engstrom and Swain, 1997). All cores show a relative maximum in concentrations/loading between 1988 and 1992 (Table 2-2), which corresponds to a peak in North American emissions in 1989 of 330 t.wl (Fig 2-7). It is therefore likely that enhanced atmospheric deposition fiom long- range sources of pollution offset the relatively earlier declines in regional emissions in Maritime Canada that reached their maximum level of just under 3000 kg.yr-' in 1970 (Fig 2-7). In both the Dipper Harbour (Fig 2-6) and Chance Harbour (Fig 2-7) sediment cores additional peaks are observed ca. 1936-1944 (Table 2-2) that correspond to a large increase in world-wide production of mercury with the second world-war (Nriagu, 1993, Nriagu, 1979b) and a relative increase in local and regional scale emissions fiom coal combustion, agricultural products and household products in Maritime Canada (Fig 2-7). There is a third distinguishable enrichmenthcrease in loading in the salt marsh sediments ca. 1920 in the Dipper Harbour core and in the late 1800s in the Chance Harbour sediments. Both dates correspond to relative maxima in North America emissions from large quantities of mercury released during gold and silver amalgamation in 1879 (1708 t wl), and again in 1920 (940 t w') (Pirrone et al., 1998). The sharp peak in mercury loading in the late 1800s in Chance Harbour sediments also corresponds to a large increase in natural mercury emissions in 1883 as the result of a major volcanic eruption in Krakatau (Schuster et al., 2002) but was not observed in either Lily Lake or Dipper Harbour. As discussed above, this may be a function of the resolution of data collected in the respective cores. The relative trends of mercury deposition in salt-marsh and lake cores from the Bay of Fundy region appear to reflect a combination of local, regional and global sources of mercury. The sediment core data suggest that the relative significance of global vs. regional inputs has varied throughout history, but that overall declines in loading are only achieved when emissions are reduced throughout North America. This means that international transboundary attempts to regulate mercury emissions will be most effective at reducing overall loading rates in the Bay of Fundy. 2.5 Literature Cited Beauchamp, S. 1998. Mercury in the atmosphere. In: Mercury in Atlantic Canada: A Progress Report (Ed, Mercury Team Regional Science Coordinating Committee) Environment Canada, Atlantic Region, Bedford, NS, pp. 16-45. Benniger, L. K., Aller, R. C., Cochran, J. K. and Turekian, K. K. 1979. Effects of biological and sediment mixing on the Pb-2 10 chronology and trace metal distribution in a Long Island Sound sediment core. Earth and Planetary Science Letters, 43,241-259. Benoit, J. M., Fitzgerald, W. F. and Damman, A. W. H. 1998. The biogeochemistry of an ombrotrophic peat bog: Evaluation of use as an archive of atmospheric mercury deposition. Environmental Research, 78, 118-133. Bloom, N. S. 1989. Determination of picogram levels of methylmercury by aqueous phase ethylation, followed by cryogenic gas chromatography with cold vapor atomic fluorescence detection. Canadian Journal of Fisheries and Aquatic Sciences, 46, 113 1 - 1140. Bother, M. H., Jahnke, R. A., Peterson, M. L. and Carpenter, R. 1980. Rate of mercury loss from contaminated estuarine sediments. Geochimica et Cosmochimica Acta, 44,273-285. Bricker, S. B. 1993. The history of Cu, Pb, and Zn inputs to Narragansett Bay, Rhode Island as recorded by salt marsh sediments. Estuaries, 16,589-607. Charles, M. J. and Hites, R. A. 1987. Sediments as archives of environmental pollution trends. In Sources and Fates of Aquatic Pollutants. Advances in Chemistry Series 216. (Eds, Hites, R. A. and Eisenreich, S. J.) American Chemical Society, Washington, DC, pp. 365-389. Chmura, G. L., Chase, P. and Bercovitch, J. 1997. Climatic controls of the middle marsh zone in the Bay of Fundy. Estuaries, 20,689-699. Chrnura, G.L., Helmer, L. L., Beecher, C.B. and Sunderland, E. M. 2001. Historical rates of salt marsh accretion on the outer Bay of Fundy. Canadian Journal of Earth Sciences, 38, 1081-1092. Clegg, Y. 2000. Historical Inventory of Sedimentary Carbon and Metals in a Bay of Fundy Salt Marsh. M.Sc. Thesis. Department of Geography. McGill University, Montreal, PQ. Craft, C. B., Seneca, E. D. and Broome, S. W. 1991. Loss on ignition and Kjeldahl digestion for estimating organic carbon and total nitrogen in estuarine marsh soils: calibration with dry combustion. Estuaries, 14, 175-179. Daoust, R. J., Moore, T. R., Chmura, G. L. and Magenheimer, J. F. 1996. Chemical evidence of environmental changes and anthropogenic influences in a Bay of Fundy saltmarsh. Journal of Coastal Research, 12,420-433. Electric Power Research Institute. 1996. Protocol for Estimating Historic Atmospheric Mercury Deposition. EPRI TR-106768 3297. Electric Power Research Institute, Palo Alto, California, pp. 60. Engstrom, D. R. and Swain, E. B. 1997. Recent declines in atmospheric mercury deposition in the upper Midwest. Environmental Science and Technology, 3 1, 960-967. Expert Panel on Mercury Atmospheric Processes. 1994. Mercury Atmospheric Processes: A Synthesis Report - Workshop Proceedings. EPRI ITR- 104214. Electric Power Research Institute, Tampa, Florida, pp.23. Farmer, J. G. 1991. The perturbation of historical pollution records in aquatic sediments. Environmental Geochemistry and Health, 13,76-83. Fitzgerald, W. F., Engstrom, D. R., Mason, R. P. and Nater, E. A. 1998. The case for atmospheric mercury contamination in remote areas. Environmental Science and Technology, 32, 1-7. Fitzgerald, W. F., Mason, R. P., Vandal, G. M. and Dulac, F. 1994. Air-water cycling of mercury in lakes. In: Mercury Pollution: Integration and Synthesis (Eds, Watras, C. J. and Huckabee, J. W.) Lewis Publishers, Chelsea, MI, pp. 203-220. Gill, G. A. and Fitzgerald, W. F. 1987. Picomolar mercury measurements in seawater and other material using stannous chloride reduction and two-stage gold amalgamation with gas phase detection. Marine Chemistry, 20,227-243. Gill, G.A., Bloom, N.S., Cappellino, S., Driscoll, C.T., Dobbs, C., McShea, L., Mason, R. and Rudd, J.W.M. 1999. Sediment-water fluxes of mercury in Lavaca Bay, Texas. Environmental Science and Technology, 33,663-669. Gobeil, C. and Cossa, D. 1993. Mercury in sediments and sediment pore water in the Laurentian trough. Canadian Journal of Fisheries and Aquatic Sciences, 50, 1794- 1800. Gregory, D., Petrie, B., Jordan, F. and Langille, P. 1993. Oceanographic, geographic and hydrological parameters of Scotia-Fundy and southern Gulf of St. Lawrence inlets. Canadian Technical Report of Hydrographic Ocean Sciences, ,248. Hargis, T. G. and Twilley, R. R. 1994. Improved coring device for measuring soil bulk density in a Louisiana deltaic marsh. Journal of Sedimentary Research, A64,681- 683. Hermanson, M. H. 1993. Historical accumulation of atmospherically derived pollutant trace metals in the arctic as measured in dated sediment cores. Water Science and Technology, 28,33-41. Hermanson, M. H. 1998. Anthropogenic mercury deposition to Arctic lake sediments. Water, Air and Soil Pollution, 101,309-321. Hintelmann, H., Harris, R., Heyes, A., Hurley, J. P., Kelly, C. A., Krabbenhofi, D. P., Lindberg, S., Rudd, J. W. M., Scott, K. J. and St. Louis, V. L. S. 2002. Reactivity and mobility of new and old mercury deposition in a boreal forest ecosystem during the first year of the METAALICUS study. Environmental Science and Technology, 36, 5034-5040. Jackson, T. A. 1997. Long-range atmospheric transport of mercury to ecosystems, and the importance of anthropogenic emissions - a critical review and evaluation of the published evidence. Environmental Reviews, 5,99-120. Jackson, T. A. 1998. Mercury in aquatic ecosystems. In: Metal Metabolism in Aquatic Environments (Eds, Langston, W. J. and Bebianno, M. J.) Chapman & Hall, London, pp. 77-158. Johnson, M. G., Culp, L. R. and George, S. E. 1986. Temporal and spatial trends in metal loadings to sediments of the Turkey Lakes, Ontario. Canadian Journal of Fisheries and Aquatic Sciences, 43,754-762. Kainz, M., Martineau, P., Canuel, R. and Lucotte, M. 1997. Final Report of the New Brunswick Mercury Project. Ecological Monitoring and Assessment Network (EMAN), Montreal, QC, pp. 23. Lamborg, C.H., Fitzgerald, W.F., Damman, A.W.H., Benoit, J.M., Balcom, P.H. and Engstrom, D.H. 2002. Modern and historic atmospheric mercury fluxes in both hemispheres: Global and regional mercury cycling implications. Global Biogeochemical Cycles, 16,51-1- 1 1. Landers, D. H., Gubala, C., Verta, M., Lucotte, M., Johansson, K., Vlasova, T. and Lockhart, W. L. 1998. Using lake sediment mercury flux ratios to evaluate the regional and continental dimensions of mercury deposition in arctic and boreal ecosystems. Atmospheric Environment, 32,919-928. Lin, C.-J. and Pehkonen, S. 0. 1999. The chemistry of atmospheric mercury: a review. Atmospheric Environment, 33,2067-2079. Lindqvist, O., Johansson, K., Aastrup, M., Anderson, A., Bringmark, L., Hovenius, G., Hakanson, L., Iverfelt, A., Meili, M. and Timm, B. 1991. Mercury in the Swedish environment: recent research on causes, consequences and corrective methods. Water, Air and Soil Pollution, 55, 1-26 1. Lindqvist, 0. and Rodhe, H. 1991. Regional and global atmospheric budgets: Mercury in the Swedish environment. Water, Air and Soil Pollution, 55, 65-71. Lockhart, W. L., Wilkinson, P., Billeck, B., Hunt, R. V., Wageman, R. and Brunskill, G. 1995. Current and historical inputs of mercury to high-latitude lakes in Canada and Hudson Bay. Water, Air and Soil Pollution, 80,603-615. Loring, D. H. 1981. Geochemical factors controlling the accumulation and dispersal of heavy metals in the Bay of Fundy region. Canadian Journal of Earth Sciences, 19, 930-944. Lucotte, M., Mucci, A., Hillaire-Marcel, C., Pichet, P. and Grondin, A. 1995. Anthropogenic mercury enrichment in remote lakes of Northern Quebec, (Canada). Water, Air, and Soil Pollution, 80,467-476. Mason, R.P., Fitzgerald, W.F. and Morel, F.M.M. 1994. The biogeochemical cycling of elemental mercury: Anthropogenic influences. Geochimica et Cosmochimica Acta, 58,3191-3198. Moore, T. 1996. Personal Communications, Department of Geography, McGill University, Montreal, PQ. Nittrouer, C. A., DeMaster, D. J., McKee, B. A., Cutshall, B. A. and Larsen, I. L. 198311984. The effect of 2'0~bmixing on 210~baccumulation rates for the Washington continental shelf. Marine Geology, 54,201-221. Northeast States for Coordinated Air Use Management (NESCAUM), Northeast Waste Management Officials' Association (NEWMOA), New England Interstate Water Pollution Control Commission (NEIWPCC), Ecological Monitoring and Assessment Network (EMAN), 1998, Northeast States and Eastern Canadian Provinces Mercury Study, A Framework for Action: NESCAUM, pp. 350. Nriagu, J. 0. 1979. The Biogeochemistry of Mercury in the Environment, Elsevier/North Holland Biomedical Press, Amsterdam, pp. 626. Nriagu, J. 0. 1979b. Global inventory of natural and anthropogenic emissions of trace metals to the atmosphere. Nature, 279,409-410. Nriagu, J. 0. 1993. Legacy of mercury pollution. Nature, 363,589. Pacyna, E. G. and Pacyna, J. M. 2002. Global emission of mercury from anthropogenic sources in 1995. Water, Air, and Soil Pollution, 137, 149-165. Pakarinen, P. and Tolonen, K. 1977. Distribution of lead in Sphagnum fuscum profiles in Finland. Oikos, 28,69-73. Peterson, A., Sreedharan, A. and Ray, S. 1989. Accumulation of trace metals in three species of fish from lakes in New Brunswick and Nova Scotia (Canada): influence of pH and other chemical parameters. Water Pollution Research Journal of Canada, 24,101-117. Pirrone, N., Allegrini, I., Keeler, G., J., Nriagu, J. O., Rossman, R. and Robbins, J. A. 1998. Historical atmospheric mercury emissions and depositions in North America compared to mercury accumulations in sedimentary records. Atmospheric Environment, 32,929-940. Pirrone, N., Keeler, G. J. and Nriagu, J. 0. 1996. Regional differences in worldwide emissions of mercury to the atmosphere. Atmospheric Environment, 30,2981- 2987. Rasmussen, P. E. 1994. Current methods of estimating atmospheric mercury fluxes in remote areas. Environmental Science and Technology, 28,2233-2241. Redfield, A. C. 1972. Development of a New England salt marsh. Ecological Monographs, 42,201-237. Roos-Barraclough, F. and Shotyk, W. 2003. Millennial-scale records of atmospheric mercury deposition obtained from ombrotrophic and minerotrophic peatlands in the Swiss Jura mountains. Environmental Science and Technology, 37,235-244. Rutherford, L. A. and Matthews, S. L. 1998. Mercury Deposition in Ombrotrophic Bogs in New Brunswick, Nova Scotia, and Prince Edward Island. Surveillance Report EPS-5-AR-98-4. Environmental Protection Branch, Environment Canada, Atlantic Region, pp. 15. Schuster, P. F., Krabbenhoft, D. P., Naftz, D. L., Cecil, L. D., Olson, M. L., Dewild, J. F., Susong, D. D., Green, J. R. and Abbott, M. L. 2002. Atmospheric mercury deposition during the last 270 years: a glacial ice core record of natural and anthropogenic sources. Environmental Science and Technology, 36,2303-23 10. Shotyk, W. 1988. Review of the inorganic geochemistry of peats and peatland waters. Earth Science Reviews, 25,95- 176. Stanford, Quentin. 1977. Canadian Oxford School Atlas, 4th Edition. Oxford University Press, Don Mills Ontario, Canada. Sunderland, E. M. and Chrnura, G. L. 2000. An inventory of historical mercury pollution in Maritime Canada: Implications for present and hture contamination. The Science of the Total Environment, 256,39-57. Swain, E. B., Engstrom, D. R., Brigham, M. E., Henning, T. A. and Brezonik, P. L. 1992. Increasing rates of atmospheric mercury deposition in Midcontinental North America. Science, 257,784-786. Ullman, W. J. and Aller, R. 1982. Diffusion coefficients in nearshore marine sediments. Lirnnology and Oceanography, 27,552-556. Urban, N. R., Eisenreich, S. J., Gringal, D. F. and Schurr, K. T. 1990. Mobility and diagenesis of Pb and Pb-2 10 in peat. Geochimica et Cosmochimica Acta, 54, 3329-3346. Weiss, D., Shotyk, W., Appleby, P. G., Kramers, J. D. and Cheburkin, A. K. 1999. Atmospheric Pb deposition since the industrial revolution recorded by five peat profiles: Enrichment factors, fluxes, isotopic composition, and sources. Environmental Science and Technology, 33, 1340-1352. CHAPTER 3 Environmental Controls on the Speciation and Distribution of Mercury in Bay of Fundy Sediments 3.1 Abstract High levels of mercury in fish and wildlife in the Bay of Fundy pose both a human and ecological health threat. In coastal ecosystems, the amount of methylmercury (MMHg) produced in the sediments is critical for anticipating the bioaccumulation of mercury by organisms. In this paper, some of the main factors known to affect the conversion of inorganic mercury to MMHg in aquatic systems are investigated. The hypothesis is tested that net MMHg production in the Bay of Fundy is enhanced in sediments with elevated total mercury and organic carbon content and inhibited in more oxic sediments and in sediments with higher sulfide concentrations. Total mercury and MMHg concentrations in sediments from this system range between 10 to 150 ng g-l and 0.05 to 1.48 ng g-', respectively. Variability in total mercury and MMHg concentrations were measured as a function of sediment grain size, total organic carbon content (TOC), redox potential (Eh), and sulfide concentrations. The results show that total mercury is a reasonable predictor of MMHg concentrations across all stations sampled. However, within the more homogeneous fine-grained muds found in Passamaquoddy Bay, geochemical characteristics of the sediments indicated by TOC, Eh and sulfide concentrations are significant determinants of variability in MMHg concentrations. Ongoing inputs of atmospheric mercury to the wider Bay of Fundy region and organic inputs through commercial activities such as fish farming in P. Bay (leading to elevated sulfide levels and declines in surface sedimsnt Eh) may both result in increased mercury concentrations in organisms. The results are used to produce a simple model of mercury distribution in P.Bay that will be usefid for developing a mercury cycling model for this system. 3.2 Introduction Mercury is a potent metal that accumulates in the tissues of aquatic organisms and can cause serious health problems in exposed wildlife and humans. In the Bay of Fundy region of Canada, high levels of mercury in lakes, fish and birds are an ongoing problem (NESCAUM et al., 1998). Mercury exists as several different chemical forms in the environment. Most of the mercury released as a byproduct of human activities and present in air, water, soils and sediments is in the inorganic form (Hg-I). However, in most organisms the predominant form of mercury is monomethylmercury (MMHg). MMHg is the most toxic mercury species and accounts for over 90% of the total burden in higher organisms (Bloom, 1992). The conversion of Hg-I to MMHg is therefore a key process that determines the availability and subsequent bioaccumulation of mercury by aquatic organisms. Understanding the characteristics of an ecosystem that control the rate of MMHg production is essential for predicting concentrations in aquatic organisms and developing effective management strategies for control and remediation. Mechanistic information on the variables controlling MMHg production in estuarine sediments is currently incomplete, However, it is recognized that total mercury concentrations alone are not sufficient to reliably forecast MMHg concentrations (Kelly et al., 1995). The predominant hypothesis of MMHg formation presented in the literature is that the conversion of inorganic mercury in the water and sediments is a microbially mediated reaction carried out by sulfate reducing bacteria (SRB) (Compeau and Bartha, 1984, Gilmour et al., 1992, King et al., 1999). SRB thrive in the redoxocline, where the maximum gradient between oxic and anoxic conditions exists (Hintelmam et al., 2000). Thus, in addition to the presence of bioavailable Hg-I, MMHg production and accumulation in aquatic systems is a function of the geochemical parameters that enhance or inhibit the activity of methylating microbes, especially sulfide concentrations, redox potential (Eh) and the composition and availability of organic carbon. The importance of these variables for determining MMHg production and bioavailability has been demonstrated in several marine systems. Other studies have found linear relationships between Eh, sulfide concentrations and ambient MMHg concentrations, where surface sediments with low Eh (greater degree of anoxia) and low sulfide concentrations tend to have higher ambient MMHg concentrations and produce relatively more MMHg (Benoit et al., 1998, Benoit et al., 1999b, Compeau and Bartha, 1984). The significance of total mercury concentrations (Hg-T) and TOC as predictors of variability in MMHg concentrations in marine sediments has also been demonstrated by studies that found higher MMHg concentrations were strongly correlated with elevated levels of TOC and Hg-T (Baeyens et al., 1998, Mason and Lawrence, 1999). In this paper, some of the main factors known to affect the conversion of inorganic mercury to MMHg in Bay of Fundy sediments are investigated. Specifically, the hypothesis is tested that net MMHg production in the Bay of Fundy is enhanced in sediments with elevated total mercury and organic carbon content and inhibited in more oxic sediments and in sediments with higher sulfide concentrations. To do this, variability in total and MMHg concentrations in the surface sediments are explored as a function of sediment grain size, TOC, sulfide concentrations, and Eh. These data are used to develop a better understanding of geochemical factors affecting the production of MMHg in Bay of Fundy sediments. The latter is required for understanding the relationship between mercury loading in the Bay of Fundy and concentrations in organisms. 3.3 Theory When interpreting empirical measurements of Hg-T and MMHg concentrations in terms of methylation rates and methylmercury production, it is useful to consider the conceptual and theoretical framework that is applied. The inherent assumption in the design of this study is that, in Bay of Fundy sediments, ambient concentrations of MMHg reflect the competing rates of in situ MMHg formation and demethylation (e.g., Benoit et al., 2002, Gilmour et al., 1998). MMHg formation can be represented as a function of a rate constant (k)that describes and quantifies the activity of sulfate reducing bacteria (SRB) methylating mercury, the sediment volume over which methylation is taking place (V3and the "bioavailable" fraction of inorganic mercury (H~-I*)or: Where: c(H~-I*)represents the concentration of bioavailable inorganic mercury. This concentration is difficult to determine as it comprises some unknown fraction of the pool of inorganic mercury present in the ecosystem and is affected by numerous physical and biological parameters. Similarly, demethylation is a function of the rate constant for MMHg degradation &), which represents the breakdown of MMHg by demethylating microbes and abiotic degradation, the sediment volume over which this breakdown occurs (Vsed)and the ambient concentrations of MMHg. Thus: (2) Demethylation = k, * V,, * C(MMHg) Where: C(MMHg) represents the ambient MMHg concentration. The amount of MMHg formed or net MMHg production (in units of grams per day) is therefore: (3) dXIMmgl = km * Ysd * C(Hg - I*)- ke * Vsed* C(M34H.g) dt At steady state (i.e., dX[MMHg]/dt = 0), equation (3) becomes: If the physical transport and loss of MMHg is relatively constant among stations, ambient MMHg concentration data should provide a reasonable first approximation of net in situ production. This assumption has been verified in a number of other systems (e.g., Benoit et al., 2002, Gilmour et al., 1998). Equation (4) illustrates that net MMHg production in Bay of Fundy sediments is a function of the methylation rate (b),the demethylation rate (k,J and the amount of inorganic mercury (H~I*)that can potentially be converted to MMHg. Significant relationships between MMHg concentrations and the parameters investigated in this study (Hg-T, TOC, Eh and sulfides) therefore indicate changes in MMHg production through changes in the overall methylation rate as a function of microbial activity andfor the supply of bioavailable inorganic mercury that is available for conversion to MMHg. Total mercury (Hg-T) levels measured in this study can be used to approximate concentrations of inorganic mercury (Hg-I) in the sediments because MMHg comprises -4%of Hg-T. However, Hg-I and thus Hg-T are not equivalent to the fraction of inorganic mercury that is available to methylating microbes (H~-I*)in equation (4). Based on past research, one definition of "bioavailable" Hg-I (Hg-I*) presented is the concentration of neutral inorganic mercury species in the porewaters that readily cross the membranes of methylating bacteria (SRB) (Benoit et al., 1999a, Benoit et al., 1999b). Thus, at station "i" some fraction of Hg-T (e.g. chi) is comprised of bioavailable inorganic mercury that can potentially be converted to MMHg, e.g.: Equation (5) can be substituted into (4) to give MMHg as a fimction of Hg-T: Equation (6) can be rearranged to express MMHg as a fraction of the total mercury pool (%MMHg): Thus, %MMHg represents the product of the rate constant quotient (k&) and the fraction of Hg-T potentially available for conversion to MMHg ($i). When the bioavailable fraction of total mercury available to methylating microbes ($i) remains relatively constant among stations (e.g., gi = $, =. ..$,), %MMHg provides a reasonable first approximation of the net methylation rate in the sediment compartment (k&). These equations will be revisited in the discussion to help explain the mechanistic implications of correlations observed in this study. 3.4 Methods 3.4.1 Site Description Samples were collected from Passamaquoddy Bay, the St. Croix River Estuary and the outer Bay of Fundy (Fig 3-1). Figure 3-1. Map of the study area depicting sampling locations in Passamaquoddy Bay, the St. Croix River Estuary and the outer Bay of Fundy. Stations SC-1 and PB-1 through PB-6 were monitored seasonally. Passamaquoddy Bay (P. Bay) is a semi-enclosed coastal embayrnent that is located at the mouth of the Bay of Fundy on the southwestern coast of New Brunswick, Canada. It has a surface area of 172.3 km2 and a maximum water depth of 67 m (Gregory et al., 1993, Loring et al., 1998). The St. Croix River is the major freshwater inflow to P. Bay and forms the border along the waters between the province of New Brunswick and the state of Maine. The river estuary is long (16.7 lun), narrow (0.1-2.4 km) and shallow (0.3-10 m) relative to P. Bay (Loring et al., 1998). The entire region is subject to the extreme tidal range of the Bay of Fundy, averaging between 6-8 m in P. Bay and up to 16 m at the mouth of the Bay of Fundy (Gregory et al., 1993). Historically, the main point sources of mercury in this area were several pulp and paper facilities and a chlor-alkali facility located at the head of the St. Croix River that operated in the 1970s (Fink et al., 1976). Present sources of mercury include difhse discharges from municipal wastewater, as well as atmospheric deposition from local and long-range sources of contamination. 3.4.2 Sediment Sample Collection Integrated 15-20 cm surface sediment samples (n=95) were collected fiom P. Bay, the St. Croix River estuary and the outer Bay of Fundy on five cruises between July 2000 and November 2001 using a modified Van Veen grab sampler. Sampling locations (Fig 3-1) included Canadian Department of Fisheries and Oceans (Loring et al., 1998) and Gulfwatch monitoring sites (Chase et al., 2001) that we analyzed for total mercury (n=28). All sampling equipment and storage containers were prepared following standard trace metal ultra-clean techniques (Gill and Fitzgerald, 1987, Mason et al., 1998). Triplicate grabs were obtained at 17 sampling stations to provide an estimate of both inter- and intra-site variability. In addition, between August 2000 and November 2001 selected sites highlighted in Figure 1 (n=7) were monitored in the spring (May), summer (August) and fall (November) to investigate seasonal variability in total mercury and MMHg concentrations. Samples collected in November 2001 were limited (n=9) by the occurrence of a humcane during the collection period. 3.4.3 Sediment Geochemistry Redox potential (Eh) of the surface sediments was measured using an Orion platinum redox electrode and a calomel reference electrode at the sediment-water interface preserved in the grab sampler. Efforts were made to minimize disturbances of the sediment-water interface during measurement of redox status and sulfide concentrations. Sub-samples for grain size (August 2000 only) and sulfide (all grab samples) analyses were obtained from the top layer of the sediments using a cutoff 5 cm3 syringe. The particle dynamics laboratory at Bedford Institute of Oceanography conducted disaggregated inorganic grain size spectral analyses using a Coulter Multisizer IIE on samples collected in August 2000, as well as sediments from stations sampled by Loring et al. (1998). Organic carbon content of the samples was estimated through loss on ignition (LOI), by heating each sample at 550•‹C overnight. Selected samples (n=27) were analyzed for total organic carbon (TOC) by combustionlnondispersive infrared gas analysis using a Shimadzu 5050A TOC analyzer. TOC is calculated as the difference between measured total carbon and total inorganic carbon. For all grab samples, LO1 was converted to organic carbon content by fitting the relationship between independently measured LO1 and TOC (3= 0.87, p<0.001). 3.4.4 Sulfide Measurements Sediment samples collected for sulfide analysis were put on ice until they were returned to the laboratory. Sulfide antioxidant buffer (SAOB) was added to the wet sediments in the lab on shore immediately after sampling and sulfide concentrations were determined using an ion-specific electrode according to the method developed by Wildish et al. (1999). The SAOB was prepared from 20 g NaOH and 17.9 g EDTA diluted with deionized water to 250 mi volume. Just before addition to the sediment sample 8.75 g of L-ascorbic acid was added to the SAOB solution. The maximum time between sampling and sulfide measurement was four hours. Although past studies have found significant relationships between pore water sulfide concentrations and ambient MMHg concentrations (e.g., Benoit et al., 1999b), the intensive labor and handling procedures associated with pore-water extractions made it infeasible to analyze the dissolved phase within the window of time known to be most reliable for sulfide analyses (Mason et al., 1998, Wildish et al., 1999). The measurement technique used in this study uses the EDTA in the SAOB solution as an auxiliary ligand to preferentially complex metals that are weakly bound in both the dissolved and solid phases. Weakly bound solid phase metal-sulfides are outcompeted by EDTA following addition of the SAOB and re-dissolved but strongly bound complexes (e.g., thiol complexes: H~-Rs+, Hg-RSH) are not affected (Haitzer et al., 2002). The sulfide concentrations reported in this study therefore include both dissolved sulfide species in the porewaters and some solid phase sulfides that are weakly associated with metals and/or the organic carbon matrix. Accordingly, the reported values are much higher than those measured in porewaters (e.g., Benoit et al. 1998). 3.4.5 Sediment Pore Waters Sediment pore waters were separated by transferring the bulk phase sample into 50 ml acid washed polycarbonate centrifuge tubes under a nitrogen atmosphere. Tubes were purged with NZprior to transfer, centrifuged at 3000 RPM for 30 minutes, followed by vacuum filtration with disposable 0.2 pm cellulose nitrate filter units. All filters were rinsed with 1% HC1 and deionized distilled (1 8 R Millipore filtration system) water immediately prior to use. Pore water samples for total mercury analysis were preserved in 0.5% ultrapure HCl, while MMHg samples were immediately frozen until analysis. 3.4.6 Mercury Analyses Samples for total and MMHg analyses were placed in 125 ml acid washed polypropylene specimen jars, and were immediately cooled to <4OC and frozen upon return to the laboratory until analysis. Wet sediments samples (bulk phase) were analyzed for total mercury by digestion in concentrated 5:2 nitric-sulfuric acid solution and oxidation with bromine monochloride (BrC1) under Class 100 conditions. Immediately prior to analysis, the excess bromine was neutralized with 10% hydroxylamine hydrochloride. Aqueous samples were digested with BrCl for at least 12 hours prior to analyses and then neutralized with an equivalent volume of hydroxylamine hydrochloride immediately before analysis. All samples were then reduced with stannous chloride, purged with nitrogen gas, and trapped on gold packed columns. Quantification was by dual-stage gold amalgamation and cold-vapor atomic fluorescence spectroscopy (CVAFS). This procedure was based on EPA Method 163 1, Gill and Fitzgerald (1987) and Bloom (1 989). MMHg was determined by steam distillation, aqueous phase ethylation using sodium tetraethylborate, purging onto Tenax packed columns, gas chromatography separation and CVAFS detection following a technique by Bloom and Fitzgerald (1988) and Horvat et al. (1993), modified by Branfireun et al. (1999). The method detection limit (MDL) for total mercury in sediment solids was 0.19 ng g-' (n=19), determined as three times the standard deviation of the mean of the sample blanks. For aqueous samples, the MDL based on a 150 ml sample volume was 0.041 ng L-' (n=18). Precision, measured as the relative percent difference (RPD) between digest duplicates (sediment solids), and analytical duplicates (aqueous phase) was 9.6% (n=24 pairs) and 5.7% (n=2 pairs) respectively. Calibration curves of at least = 0.99 were achieved daily or samples were re-run. Accuracy was measured both by spike recoveries and using the MESS-3 marine sediment certified reference material (91 + 9 ng g-') fiom the National Research Council of Canada. Recoveries averaged 103% k 10% (n= 12) and 92 + 16 ng g-' for all MESS-3 samples (n=9). Samples fiom runs with poor recoveries (40%) were reanalyzed. For MMHg, the MDL was 0.041 ng L-' (n=6) in the aqueous phase and 0.007 ng g-' (n=16) for sediment solids. The RPD for distillation duplicates was 18.1% (n=2 I), while the average recovery of spikes between 100-500 pg per gram of wet sediment was 106 * 26% (n=12). Some of this variability can be attributed to uncertainty as to the true concentration of the sediment sample being spiked, as reflected in the RPD of distillation duplicates. A wet to dry weight conversion was calculated for each sample analyzed by oven drying sub-samples of wet sediments for at least 24 hours at 60' C. 3.4.7 Statistical Analysis Bivariate correlation matrices between Hg, MMHg, Eh, sulfide, TOC and grain size were developed for each sampling period to identify the predominant variables controlling production and accumulation of MMHg in Fundy sediments. Regression analysis was used to test the hypothesis regarding variability in MMHg concentrations with Eh, sulfide, TOC and total mercury levels. This was followed by stepwise regression analysis to isolate the most important factors explaining variability in MMHg concentrations toward the development of a model for MMHg distribution in Bay of Fundy sediments. All correlations presented are significant at the 95% confidence level. Analyses were performed using SPSS statistical software. 3.5 Results and Discussion 3.5.1 Total and Methyl-mercury Distribution Mercury concentrations in Bay of Fundy sediments ranged from 40ng g-' Hg-T and 0.05 ng g-' MMHg in the sandy regions of P.Bay and the outer Bay of Fundy to >15O ng Hg-T and 1.48 ng gj-' MMHg in the upper reaches of the St. Croix river that were affected by historical discharges of mercury from a chlor-alkali facility (Fig 3-2). Because P. Bay and the outer Bay of Fundy stations sampled have not been highly impacted by any historical point sources of mercury pollution, factors influencing MMHg production in P.Bay are distinguished from those affecting all stations sampled in this study in subsequent analyses and discussion. There were no significant seasonal differences in Hg-T concentrations between August 2000 and November 2001. However, both MMHg concentrations and the fraction of total mercury in methylated form (%MMHg) were significantly elevated in the spring and fall relative to summer (t-test, paired two sample for means, p<0.05). Hence, MMHg concentration data for each sampling period were analyzed separately (Table 3-1). Figure 3-2. Spatial distribution of total mercury (Hg-T) in Passamaquoddy Bay and the St. Croix River Estuary. , 8 45 82120 157 Scale in Kilometers Hg-T (ng.g-I dry) Table 3- 1. Seasonal correlation matrices for monomethylmercury (MMHg) concentrations in Bay of Fundy sediments. Pearson correlation coefficients (r) are significant at the 95% level or greater. %MMHg = Fraction of total mercury (Hg) as MMHg. TOC = Total Organic Carbon. Eh = Redox Potential. NC = No Correlation. May August ~ovember' % MMHg % MMHg MMHg % MMHg Hg-T All Stations X X da X Without River X X da X All P.Bay Stns. X X nla X PB-I to PB-6 X X NC X TOC All Stations NC NC nla da Without River NC 0.384 da da All P.Bay Stns. NC 0.412 da da PB-1 to PB-6 NC NC NC NC Eh All Stations NC NC da nla Without River NC NC da da All P.Bay Stns. NC NC n/a nla PB-1 to PB-6 NC NC NC NC Sulfides All Stations NC 0.685 nla nla Without River NC 0.664 da n/a All P.Bay Stns. NC 0.663 nla nla PB- 1 to PB-6 NC 0.524 NC 0.810 'sample size for November is small (n=9) and does not include the same range in Hg and MMHg data (Hg: 37-60 ng g-l; MMHg: 0.24-0.36 ng g-') observed in May (Hg: 10-113 ng g-l MMHg:0.07-1.48 ng g-l) and August (Hg:13-148 ng g-l; MMHg:0.05-0.73 ng g-l). As anticipated in our original hypothesis, there is a significant linear relationship between Hg-T and MMHg over all stations sampled and in P. Bay (Fig 3-3f; Table 3-1). These results I indicate that MMHg production in P. Bay sediments may be limited by the supply of total mercury as expressed in equation (6). Thus, as total mercury levels increase or decrease in this system, a corresponding increase or decline in ambient MMHg is expected. This relationship is especially significant for anticipating the response of this system to changes in mercury inputs due to emissions controls, as the majority of anthropogenic emissions are comprised of inorganic mercury species and only MMHg bioaccumulates in organisms. 3.5.2 Role of Total Organic Carbon Strong and significant correlations were found between Hg-T, grain size and TOC for all stations (Fig 3-3a-b). TOC was also linearly correlated with particle size (P), expressed as the fraction of sediments by weight with a grain size less than 63 pm in P.Bay and the St. Croix River (Fig 3-3c). This correlation reflects the tendency of fine-grained sediments to contain relatively higher TOC than large grained (sandy) sediments. As a result, Hg-T concentrations are generally higher in depositional areas with a greater portion of muddy fine-grained sediments than those in more sandy erosional regions (Fig 3-2). There are no significant differences among Hg-T concentrations normalized to TOC throughout P. Bay using a t-test for paired mean concentrations at sampling locations (excluding the St. Croix river and the outer Bay of Fundy). Thus, normalizing sediment Hg- T to TOC contents can account for much of the variability among sites in P. Bay. TOC is also an important parameter controlling both Hg-T and MMHg concentrations in porewaters. Figure 3-3. Controls on mercury speciation and distribution in Bay of Fundy sediments. Panel: (a) shows mean total mercury concentrations (Hg-T) at all stations sampled as a function of total organic carbon (TOC); (b) shows Hg-T as a function of sediment grain size (P) as reflected by the fraction of sediments by weight <63pm in diameter; (c) shows co- variation between sediment grain size and TOC; (d) illustrates methylmercury (MMHg) as a both a linear and exponential function of TOC; (e) depicts partitioning of mercury between the dissolved and solid phases (log Kd)as a linear function of total organic carbon (TOC) in P. Bay porewaters, e.g., Log Kd (Hg-T), r = 0.77, pC0.01; Log &(MMHg), r = 0.76, pc0.01; and (f) shows ambient concentrations of MMHg as a function of Hg-T. (hp ,.6-6u) k6~ The relative proportion of solid-phase to porewater Hg-T and MMHg increases linearly with TOC in sediments fiom this system (Fig 3-3e). We hypothesized that increasing TOC in Bay of Fundy sediments would correspond to an increase in MMHg production. The results show strong linear relationships between TOC and MMHg concentrations both among all stations sampled and when P.Bay is considered separately (Fig 3-3d, Table 3-1). However, the relationship between %MMHg and TOC among all stations sampled in May was not significant and only weakly positive correlations were observed in August in P. Bay (Table 3-1). Theoretically, the potential effects of TOC on net production of MMHg as laid out in equation (6) are threefold. First, previous studies indicate that because suspended organic matter has a high affinity for inorganic mercury, mercury tends to be co-deposited with TOC to the benthic sediment layer in coastal ecosystems (Cossa and Noel, 1987, Gagnon et al., 1996, Mason and Lawrence, 1999). Secondly, TOC provides a substrate for microbial activity (Lord and Church, 1983) and may therefore stimulate microbial activity, potentially increasing the effective methylation rate (k,) in equation (6). Finally, both spectroscopic studies and partitioning experiments show that inorganic mercury can form strong bonds with the functional groups on the organic carbon matrix (eg thiol- RSH complexes) influencing its reactivity and effectively reducing the pool of H~-I* present in the sediments (Haitzer et al., 2002, Le Roux et al., 2001). The relationships between ambient Hg-T, MMHg and TOC shown in this study support the premise that inorganic mercury in this system may be co-deposited with TOC, and shows that TOC may effectively increase the supply of Hg-T, increase the bioavailable 88 pool size, or stimulate microbial activity yielding greater MMHg production. The role of TOC on the effective methylation rate in these sediments described in equation (7) is less clear, as the relationship between %MMHg and TOC is likely confounded by the counteracting effects of a potential increase in the effective methylation rate (L)and a decline in the bioavailable fraction of Hg-T in the sediments (9) with increasing TOC. Overall, the dominant effect of TOC in this system appears to be on the supply of inorganic mercury for methylation. However, the literature indicates that different types of organic carbon may be relatively more labile and decomposed more readily by methylating microbes or contain fewer "strong" binding sites for inorganic mercury (e.g., Xia et al., 1999, Haitzer et al., 2002). Hence, variability in net MMHg production rates that are due to differences in the composition of organic carbon will likely not be distinguishable using the TOC measurement reported in this study. 3.5.3 Redox Potential (Eh) We hypothesized that MMHg production would be enhanced in sediments with the lowest Eh because mechanistically more reducing sediments provide a more favorable environment for anaerobic microbes such as SRB, thereby increasing net MMHg production (Hintelmam et al., 2000). In this study, a significant inverse relationship between Eh and MMHg was observed in August among all stations and in P. Bay but there is no significant relationship between %MMHg and Eh (Fig 3-4a-b, Table 3-1). When P. Bay is considered separately, Eh is also inversely related to TOC (Table 3-2). Figure 3-4. Covariation of MMHg concentrations with redox potential and sulfide concentrations. Panels (a) and (b) show relationship between redox potential of the surface sediments and MMHg concentrations at all Bay of Fundy stations and in P.Bay only respectively. Panels (c) and (d) show the fraction of Hg-T present as MMHg (%MMHg) as a function of sulfides in the summer (August) and fall (November) respectively. 0.80 (a) :gust (all stns.) (b) August (p.Bay)' 0m407 + 1 (c) Summer (all stns.) 1 (d) Fall (P. Bay) 0 1000 2000 3000 0 1000 2000 3000 4000 Bulk Sulfides (pM) Bulk Sulfides (pM) 'Sub-set of samples from P. Bay analyzed for porewaters in August 2001 Table 3-2. Correlation matrix for sediment redox potential (Eh), sulfide concentrations, and total organic carbon content (TOC) in sediments from Passamaquoddy Bay. May 2001 August 2000/2001 Sulfide TOC Sulfide TOC Eh All Stations NC NC NC NC P. Bay -0.363 -0.671 NC -0.508 Sulfide All Stations X NC X 0.408 P. Bay X 0.715 X 0.566 This relationship is consistent with other studies in the region that show one of the geochemical indicators of organic enrichment in sediments is declining surface redox potential because the chemical oxygen demand in sediments is proportional to the concentration of organic carbon undergoing mineralization (Hargrave et al., 1997, Wildish et al., 1999, Wildish et al., 2001). Thus, the inverse correlation between Eh and TOC suggests that the relationship between Eh and MMHg is driven by the effects of TOC on MMHg production rather than a direct effect of Eh on the effective methylation rate (k&). 3.5.4 Sulfides In Bay of Fundy sediments, %MMHg is linearly correlated with sulfides (Fig 3-4) suggesting a proportional increase in the effective methylation rate with increasing sulfide concentrations. This is an interesting finding given that past studies have found elevated dissolved sulfide concentrations inhibit MMHg production (Gilmour et al., 1998, Benoit et al., 2001). Sulfide concentrations reported in this study are also positively correlated with TOC and inversely correlated with Eh (Table 3-2), reflecting the general tendency of sediments enriched in organic carbon tend to have higher sulfide concentrations and lower redox potentials (Wildish et al., 1999). Thus, these data are consistent with other studies in the region that show that both sulfide concentrations and Eh are sensitive indicators of organic enrichment (Hargrave et al., 1997, Wildish et al., 2001). Mechanistically, other studies have demonstrated that the inhibitory effect of increasing inorganic sulfide concentrations in interstitial waters on MMHg production is caused by a shift in the speciation of inorganic mercury away from neutral complexes that readily cross the cell membranes of methylating bacteria (e.g., H~s~,toward charged polysulfide complexes that do not readily diffuse (Benoit et al., 1999a, Benoit et al., 2001). These studies observed that at higher sulfide concentrations the amount of inorganic mercury in the dissolved phase increases, but the fraction of inorganic mercury in the porewaters available to SRB decreases, reflected by a decline in ambient MMHg concentrations (Gilmour et al., 1998, Benoit et al., 1999b, Benoit et al., 2001, Benoit et al., 1999a, Benoit et al., 1998). Thus, increasing concentrations of dissolved inorganic sulfides in these studies effectively lowered the amount of Hg-I* available to methylating microbes or as represented in equation (6),lowered the effective fraction of the total mercury pool (9) that crosses the cell membranes of methylating microbes. However, it is likely that the relationship between %MMHg and sulfides reported in this study reflects an increase in the effective methylation rate (k,,&-) with certain types of organic enrichment rather than the direct role of inorganic sulfides on the speciation of inorganic mercury that affects the pool of inorganic mercury available to methylating microbes (4). The apparent contrast between the effects of increasing sulfide concentrations in this study and other investigations can be explained by differences in the "pools" of sulfides measured. The sulfide measurement technique used in this study releases "labile" sulfides associated with the organic carbon matrix (see methods) in sediments whereas other studies have measured concentrations in filtered porewaters, thus reported the pool of dissolved inorganic sulfide ions (Benoit et al., 1999a, Benoit et al., 1999b, Benoit et al., 2001). The correlation between TOC and sulfides (Table 3-2) suggests that sulfide concentrations measured in this study are partially a function of the degradation of organic matter by microbes as seen in other systems (Lord and Church, 1983). In addition, sulfate reduction rates have been shown to be proportional to the activity of SRB responsible for MMHg production in other estuarine systems (King et al., 1999). Thus, this type of sulfide measurement may be a proxy indicator of the activity of SRB responsible for methylation (k&), potentially providing another method to empirically estimate relative methylation rates in marine sediments. 3 S.5 Multivariate Analysis The data collected in this study can be used to develop a simplified semi-empirical model for speciation and distribution of mercury in Bay of Fundy sediments (Table 3-3). A multivariate analysis of Hg-T, TOC, Eh and sulfides as predictors of MMHg concentrations selected TOC in May and Hg-T and sulfide concentrations in August (Table 3-3). The distribution of Hg-T in this system can be forecasted as a function of both TOC and particle size (P). Partitioning of both Hg-T and MMHg between the solid and dissolved phases can be modeled empirically as a function of TOC (Table 3-3). The results of this study suggest that when the bioavailable fraction of mercury ($i) is relatively constant among regions, MMHg production represented in equation (6) can be modeled as a function of Hg-T (reflecting the supply of inorganic mercury), and the effective methylation rate (k,&) or %MMHg. In turn, the effective methylation rate can be forecasted as a function of sulfide concentrations measured in this study. Table 3-3. Summary of regression equations developed in this study for mercury speciation and distribution in Bay of Fundy sediments. Linear Models Dependent Variable Model Log &[H~-~(L kg-') 0.345 * TOC(%) + 2.92 ;2 = 0.59, p<0.01 Multivariate Analysis (Stepwise Regression) for MMHg concentrations Sampling Period Model Excluded Statistics Variables May 2001 0.742 * TOC(%) - 0.172 Hg-T; Eh; 2~0.55; Sulfide p August 2000-2001 4.1 3 x 1o-~ * Hg - T (ng / g) + Eh, TOC 2=0.75; p Note: Hg-T = total mercury concentration; MMHg = methylmercury concentration; TOC = total organic carbon content of sediments; P = particles with grain sizes <63 pm; Sulfide = sulfide concentrations measured in the wet sediments. 3.6 Conclusions The interrelationships among TOC, Eh and bulk sulfides observed in this study likely result from strong coupling of the biogeochemical cycles of carbon, sulfur and mercury and can be used to anticipate the effects of organic enrichment on mercury speciation and bioaccumulation in P. Bay. In P.Bay, there was a positive correlation between sulfide concentrations and TOC and an inverse relationship between Eh and TOC. As discussed above, elevated sulfide concentrations and lower Eh in the surface sediments covary with higher MMHg concentr&ions in these sediments. Both increased sulfide concentrations and declining Eh in benthic sediments have been established as sensitive measures of organic enrichment (Wildish et al., 1999). Thus, one of the major management implications of this study is that organic enrichment from aquaculture farming taking place in the region, that can cause increases in sulfide concentrations in the sediments and declines in Eh (e.g., Hargrave et al., 1997, Wildish et al., 2001, Wildish et al., 1999), could promote MMHg exposure in these sediments, potentially increasing the amount of mercury accumulating in the food-web in this system. 3.7 Literature Cited Baeyens, W., Meuleman, C. and Leermakers, M. 1998. Behavior and speciation of mercury in the Scheldt estuary (water, sediment and benthic organisms). Hydrobiologia, 366,63-79. Benoit, J., Gilmour, C. C., Mason, R. P. and Heyes, A. 1999a. Sulfide controls on mercury speciation and bioavailability to methylating bacteria in sediment pore waters. Environmental Science and Technology, 33,95 1-957. Benoit, J. M., Gilmour, C. C., Heyes, A., Mason, R. P. and Miller, C. 2002. Geochemical and biological controls over methylmercury production and degradation in aquatic systems. In: Biogeochemistry of Environmentally Important Trace Metals. ACS Symposium Series No. 835. (Eds, Cai, Y. and Braids, O.C.) Oxford University Press, New York, pp. 262-297. Benoit, J. M., Gilmour, C. C. and Mason, R. P. 2001. The influence of sulfide on solid- phase mercury bioavailability for methylation by pure cultures of DesuIfobulbus propionicus (1pr3). Environmental Science and Technology, 35, 127-132. Benoit, J. M., Gilmour, C. C., Mason, R. P., Riedel, G. S. and Reidel, G. F. 1998. Behavior of mercury in the Patuxent River estuary. Biogeochemistry, 40,249- 265. Benoit, J. M., Mason, R. P. and Gilmour, C. C. 1999b. Estimation of mercury-sulfide speciation in sediment pore waters using octanol-water partitioning and implications for availability to methylating bacteria. Environmental Toxicology and Chemistry, 18,2138-2141. Bloom, N. S. 1992. On the chemical form of mercury in edible fish and marine invertebrate tissue. Canadian Journal of Fisheries and Aquatic Sciences, 49, 1010- 1017. Bloom, N. S. 1989. Determination of picogram levels of methylmercury by aqueous phase ethylation, followed by cryogenic gas chromatography with cold vapor atomic fluorescence detection. Canadian Journal of Fisheries and Aquatic Sciences, 46, 113 1-1 140. Bloom, N.S. and Fitzgerald, W. F. 1988. Determination of volatile mercury species at the picogram level by low-temperature gas chromatography with cold-vapor atomic fluorescence detection. Analytica Chemica Acta, 208, 15 1- 161. Branfireun, B., Roulet, N. T., Kelly, C. A. and Rudd, J. W. M. 1999. In situ stimulation of mercury methylation in a boreal peatland: Toward a link between acid rain and methylmercury contamination in remote environments. Global Biogeochemical Cycles, 13,743-750. Chase, M. E., Jones, S. H., Hennigar, P., Sowles, J., Harding, G. C. H., Freeman, K., Wells, P. G., Krahforst, C., Coombs, K., Crawford, R., Pederson, J. and Taylor, D. 2001. Gulfwatch: Monitoring spatial and temporal patterns of trace metal and organic contaminants in the Gulf of Maine (1 99 1- 1997) with the blue mussel, ik&tilus edulis L. Marine Pollution Bulletin, 42,491-505. Compeau, G. and Bartha, R. 1984. Methylation and demethylation of mercury under controlled redox, pH, and salinity conditions. Applied and Environmental Microbiology, 48, 1203-1207. Cossa, D. and Noel, J. 1987. Concentrations of mercury in near shore surface waters of the Bay of Biscay and in the Gironde Estuary. Marine Chemistry, 20,389-396. Fink, L. K Jr., Pope, D. M., Hams, A. B. and Schick, L. L. 1976. Heavy Metal Levels in Suspended Particulates, Biota and Sediments of the St. Croix Estuary in Maine. Land and Water Resources Institute, University of Maine at Orono, Orono, Maine, pp. 59. Gagnon, C., Pelletier, E., Mucci, A. and Fitzgerald, W. F. 1996. Diagenetic behavior of methylmercury in organic-rich coastal sediments. Limnology and Oceanography, 41,428-434. Gill, G. A. and Fitzgerald, W. F. 1987. Picomolar mercury measurements in seawater and other material using stannous chloride reduction and two-stage gold amalgamation with gas phase detection. Marine Chemistry, 20,227-243. Gilmour, C. C., Henry, E. and Mitchell, R. 1992. Sulfate stimulation of mercury methylation in freshwater sediments. Environmental Science and Technology, 26, 228 1-2287. Gilmour, C. C., Riedel, G. S., Ederington, M. C., Bell, J. T., Benoit, J. M., Gill, G. A. and Stordal, M. C. 1998. Methylmercury concentrations and production rates across a trophic gradient in the northern Everglades. Biogeochemistry, 40,327-345. Gregory, D., Petrie, B., Jordan, F. and Langille, P. 1993. Oceanographic, geographic and hydrological parameters of Scotia-Fundy and southern Gulf of St. Lawrence inlets. Canadian Technical Report of Hydrographic Ocean Sciences No. 143. Department of Fisheries and Oceans, Scotia-Fundy Region, Dartmouth, NS, pp. 248. Hargrave, B.T., Phillips, G.A., Doucette, L.I., White, M.J., Milligan, T.M., Wildish, D.J. and Cranston, R.E. 1997. Assessing benthic impacts of organic enrichment from marine aquaculture. Water, Air and Soil Pollution, 99, 641-650. Haitzer, M., Aiken, G. R. and Ryan, J. N. 2002. Binding of mercury (II) to dissolved organic matter: the role of mercury-to-DOM concentration ratio. Environmental Science and Technology, 36,3564-3570. Hintelmanu, H., Keppel-Jones, K. and Evans, R. D. 2000. Constants of mercury methylation and demethylation rates in sediments and comparison of tracer and ambient mercury availability. Environmental Toxicology and Chemistry, 19, 2204-22 11. Horvat, M., Liang, L. and Bloom, N. S. 1993. Comparison of distillation with other current isolation methods for the determination of methyl mercury compounds in low level environmental samples. Analytica Chemica Acta, 282, 153-168. Jay, J. A., Morel, F. M. M. and Hemond, H. F. 2000. Mercury speciation in the presence of polysulfides. Environmental Science and Technology, 34,2196-2200. Kelly, C. A., Rudd, J. W. M., St. Louis, V. L. and Heyes, A. 1995. Is total mercury concentration a good predictor of methylmercury concentration in aquatic systems? Water, Air, and Soil Pollution, 80, 715-724. King, J. K., Saunders, M., Lee, R. F. and Jahnke, R. A. 1999. Coupling mercury methylation rates to sulfate reduction rates in marine sediments. Environmental Toxicology and Chemistry, 18, 1362-1369. Le Row, S., Turner, A., Millward, G. E., Ebdon, L. and Appriou, P. 2001. Partitioning of mercury onto suspended sediments in estuaries. Journal of Environmental Monitoring, 3,37-42. Lord, C. J. I. and Church, T. M. 1983. The geochemistry of salt marshes: sedimentary ion difision, sulfate reduction, and pyritization. Geochimica et Cosmochimica Acta, 47, 1381-1391. Loring, D. H., Milligan, T. G., Willis, D. E. and Saunders, K. S. 1998. Metallic and Organic Contaminants in Sediments of the St. Croix Estuary and Passamaquoddy Bay. Canadian Technical Report of Fisheries and Aquatic Sciences No. 2245, Dartmouth, N.S., pp. 38. Mason, R., Bloom, N., Cappellino, S., Gill, G., Benoit, J. and Dobbs, C. 1998. Investigation of porewater sampling methods for mercury and methylmercury. Environmental Science and Technology, 32,403 1-4040. Mason, R. P. and Lawrence, A. L. 1999. Concentration, distribution, and bioavailability of mercury and methylmercury in sediments of Baltimore Harbor and Chesapeake Bay, Maryland, USA. Environmental Toxicology and Chemistry, 18,2438-2447. Northeast States for Coordinated Air Use Management (NESCAUM), Northeast Waste Management Officials' Association (NEWMOA), New England Interstate Water Pollution Control Commission (NEIWPCC), Ecological Monitoring and Assessment Network (EMAN), 1998, Northeast States and Eastern Canadian Provinces Mercury Study, A Framework for Action: NESCAUM, pp. 350. Wildish, D. J., Akagi, H. M., Hamilton, N. and Hargrave, B. T. 1999. A Recommended Method for Monitoring Sediments to Detect Organic Enrichment fiom Mariculture in the Bay of Fundy. Canadian Technical Report of Fisheries and Aquatic Sciences No. 2286, St. Andrews, N.B., pp. 30. Wildish, D. J., Hargrave, B. T. and Pohle, G. 2001. Cost-effective monitoring of organic enrichment resulting fiom salmon mariculture. ICES Journal of Marine Science, 58,469-476. Xia, K., Skyllberg, U. L., Bleam, W. F., Bloom, P. R., Nater, E. A. and Helmke, P. A. 1999. X-ray absorption spectroscopic evidence for the complexation of Hg(II) by reduced sulfur in soil humic substances. Environmental Science and Technology, 33,257-261. CHAPTER 4 Speciation and Bioavailability of Mercury in Well-Mixed Estuarine Sediments 4.1 Abstract Despite stringent regulations controlling anthropogenic mercury sources in North America, high levels of mercury in coastal fish and shellfish are an ongoing problem in Maritime Canada and the Northeastern United States. This study presents sediment core data fiom a macrotidal estuary located at the mouth of the Bay of Fundy showing stratigraphic profiles of total and methylmercury concentrations and "potential" methylation rates measured using stable mercury isotopes. The results show that in contrast to the classic profile typically observed in unmixed sediments, methylmercury production occurs throughout the estimated 15-cm-thick active surface layer of these well-mixed sediments. The resulting large reservoir of total and methylmercury in these sediments helps to explain why mercury concentrations in organisms in this system remain high despite large emissions reductions. Current management policies need to take into account the expected delay in the response time of well-mixed estuarine systems to declines in mercury loading due to the greater reservoir of historic mercury available in these sediments that can potentially be converted to methylmercury and biomagnify in the coastal food-chains. 4.2 Introduction Mercury emissions from anthropogenic sources in Maritime Canada and the Northeastern United States have declined by more that 50% from peak levels in the 1970s as a result of pollution control measures (Sunderland and Chmura, 2000). However, there is no evidence of similar declines in the concentrations of mercury in marine birds, fish and shellfish fiom the Bay of Fundy region of Canada, which is a catchment for atmospheric mercury contamination fiom industrialized regions of Central Canada and the United States (Chase et al., 2001, Evers et al., 1998, NESCAUM et al., 1998). Methylmercury levels in fish and invertebrates surpass Environment Canada's tissue residue guideline of 0.033 ppm for protection of all aquatic life (Canadian Council of Ministers of the Environment, 2000) and, at the highest trophic levels, total mercury concentrations exceed the Health Canada guideline of 0.5 ppm for safe consumption by humans (NESCAUM et al., 1998). A better understanding of the relationship between concentrations in organisms and emissions reductions is needed to develop effective strategies for reducing potential human health impacts of mercury in the Bay of Fundy. The conversion of inorganic mercury to methylmercury (MMHg) is a critical process affecting the relationship between mercury inputs and concentrations in biota. The majority of mercury released as a byproduct of human activities and present in the environment is inorganic mercury but only the most toxic organic form, MMHg, bioaccumulates in organisms (Bloom, 1992). There is considerable evidence that production of MMHg is principally a biologically mediated reaction carried out by sulfate-reducing bacteria (SRB) in marine sediments (Benoit et al., 1999, King et al., 1999). Because these microbes are anaerobes, benthic sediments in coastal systems often provide the most suitable environment for MMHg production (Choi and Bartha, 1994). In coastal systems, bottom sediments act as a reservoir for past and present mercury inputs and provide the most favorable geochemical environment for the conversion of inorganic mercury to MMHg. The exposure of coastal organisms to mercury is therefore determined by the amount of inorganic mercury in the sediments that is converted to MMHg and the geochemical conditions that affect the rate of MMHg production by methylating microbes. To gain insight into the temporal response of mercury concentrations in response to changes in mercury inputs, it is important to understand both of these factors. The Bay of Fundy has the world's largest tides, reaching up to 16 m at the mouth of the Bay (Gregory et al., 1993). Tseng et al. (2001) found that mixing of sediments in the fluid mud profile of a turbid macrotidal estuary in France creates a distinct geochemical environment that can facilitate the activity of microbial populations that convert inorganic mercury to MMHg. Based on these data we hypothesized that physical mixing in the Bay of Fundy will change the geochemical characteristics of the surface sediments and increase the depth of the active sediment layer where methylation takes place. The potential net effects of these changes are an increased reservoir of total mercury and MMHg in the sediment compartment, thereby increasing the amount of mercury available to organisms. In this paper, the effects of sediment mixing are explored by investigating ambient profiles of total mercury and MMHg, and potential MMHg production rates measured in sediment cores from contrasting sites at the mouth of a freshwater tributary and in well- mixed regions of a coastal embayment located at the mouth of the Bay of Fundy. Evidence is presented that demonstrates the effects of mixing on the depth and geochemistry of the active sediment layer in this system and the implications for mercury uptake at the base of the food-chain and the temporal response of this system to reductions in anthropogenic mercury emissions are discussed. 4.3 Methods 4.3.1 Study Site Sediment cores were collected in Passamaquoddy Bay (P. Bay) and the St. Croix River Estuary between May and August 2001 (Fig 4-1). P. Bay is a semi-enclosed macrotidal estuary located at the mouth of the Bay of Fundy. The mean tidal range in P. Bay is 6-8 m and reaches up to 16 m at the mouth of the Bay of Fundy (Gregory et al., 1993). Thus, P. Bay is a highly turbid, tidally dominated system. The St. Croix River is the main freshwater inflow to P. Bay, and runs along the border between Canada and the United States. Related investigations (Chapter 3) have shown that total mercury concentrations in surface sediments in P. Bay and the St. Croix River Estuary range between 10 ng g-' and 150 ng g-' dry weight, with the highest concentrations found at the head of the St. Croix River Estuary. Figure 4-1. Map of the study area showing sampling locations for gravity cores, push cores and isotope cores. Scale in Kilometers I 0 10 20 30 40 New Brunswick, Canada Maine, USA i Gravity Cores tj push cores A lsotope cores Elevated mercury concentrations in this region are most likely due to residual contamination associated with historical discharges of mercury from a chlor-alkali facility that operated in the 1970s (Fink et al., 1976). 4.3.2 Samples Collected 4.3.2.1 Push Cores Eight 15 cm-long surface sediment push cores were obtained from a large volume, modified Van Veen grab sampler at selected locations in May and August 2001 using acid-washed PVC/plexiglass core tubes (Fig 4-1). Cores were extruded at two-centimeter intervals in the lab on shore and all samples were frozen until analysis. Sulfide concentrations in the wet sediments of push core sub-samples were analyzed using an ion specific electrode after addition of sulfide antioxidant buffer according to the method outlined by Wildish et al. (1999). Redox potential (Eh) was measured at the sediment surface of gravity core and push core sampling locations using an Orion platinum redox electrode and a calomel reference electrode. 4.3.2.2 Gravity Cores Multiple gravity cores (n=20) were collected to characterize the geochemical profile of sediments in P. Bay and the St. Croix River using a 1.5 m long gravity corer (Fig 4-1). Gravity cores were sectioned into 5 cm vertical intervals to a depth of 20 cm, after which they were divided into 10 cm increments. Sediments were placed in polyethylene sample containers and cooled at temperatures <4OC to minimize chemical transformations following extrusion. Vertical gradients of dissolved ammonium and sulfate were measured in sediment pore waters within 24 hrs of sampling to estimate present day sediment accumulation rates following the method developed by Cranston (1991, 1997). This method estimates sediment burial rates within a factor of two of radiometric dating (Cranston, 1991) and is particularly useful in erosion and transport regions of the estuarine basin where sediments cannot be dated using traditional methods. All cores were then freeze-dried and archived for Meranalysis. In addition to total mercury, gravity cores were analyzed for a spectrum of metals including Fe, Li, Mn and Pb. Subsamples for metal analysis were prepared by digesting 1.0 g of freeze-dried sediment in 5.0 mL of concentrated nitric acid for approximately 24 hours at 60•‹C. Flame atomic absorption analyses were carried out using a Varian 220FS spectrometer for Fe, Li and Mn. Electrothermal atomic absorption analyses of Pb were carried out using a Varian 220FS spectrometer fitted with a Varian GTAl10 graphite furnace. Relative precision and accuracy limits, estimated Erom replicate analyses of CANMET certified reference materials (STSDs 1 to 4) and an internal GSC standard (EMG-017), were determined to be f 3% for Fe, Li and Mn and * 5% for Pb. Total and organic carbon was determined in 0.5 g of freeze-dried sediment using a Leco WR- 112 carbon analyzer. Inorganic carbon was removed using 1 M hydrochloric acid prior to organic carbon measurements. Precision and accuracy were estimated to be f 0.03 wt.% based on replicate analyses of calibration standards. 4.3.2.3 Sediment Pore Waters Sediment pore water samples were separated from selected push cores (n=2) and surface sediment samples (n=23) collected at the push core and gravity core sampling locations. Pore waters were extracted under a nitrogen atmosphere by transferring the bulk sample into 50 mL acid-washed polycarbonate centrifuge tubes. Tubes were purged with N2 prior to transfer, centrifuged at 3000 RPM for 30 minutes, followed by vacuum filtration with disposable 0.2 pm cellulose nitrate filter units. All filters were rinsed with 1% HCl and distilled deionized (18 R Millipore filtration system) water immediately prior to use. Pore water samples for total mercury analysis were preserved in 0.5% ultrapure HC1, while MMHg samples were immediately frozen until analysis. 4.3.2.4 Isotope cores At two stations, duplicate intact sediment cores were spiked at 1 cm intervals with inorganic mercury isotope (9 1.95% Ig9Hg(11)from Oak Ridge Batch # 168490) that was pre-equilibrated with seawater before injection. Cores were incubated at ambient seawater temperature for 4 hours, extruded and frozen until analysis for conversion of isotope into methylated mercury. The purpose of this work was to determine a methylation "potential" for these sediments rather than using the isotope as a tracer because the added '99~g(~~)is likely more available for methylation than the in situ Hg(I1). 4.3.2.5 Polvchaetes Polychaete worms (Nephtys sp.) were collected from benthic sediments at the same sampling locations as push cores, gravity cores and isotope cores. Samples were obtained by immediately sieving the wet sediments collected using a modified Van Veen grab sampler on board the sampling vessel. Polychaetes were identified in the lab and then flushed with deionized water for 24 hours before freezing until analysis for mercury. Biological samples were analyzed for total mercury using the same methodology as sediment samples. 4.3.3 Mercurv Analyses Samples for total and MMHg analyses were placed in 125 mL acid washed polypropylene specimen jars and were immediately cooled to <4OC and fiozen upon return to the laboratory until analysis. Wet sediment samples were analyzed for total mercury by digestion in 5:2 concentrated nitric-sulfiuic acid solution and oxidation with bromine monochloride (BrC1) under Class 100 clean room conditions. Immediately prior to analysis, the excess bromine was neutralized with 10% hydroxylamine hydrochloride. Aqueous samples were digested with BrCl for at least 12 hours prior to analyses and then neutralized with an equivalent volume of hydroxylamine hydrochloride immediately before analysis. All samples were then reduced with stannous chloride, purged with nitrogen gas, and trapped on gold packed columns. Quantification was by dual-stage gold amalgamation and cold-vapor atomic fluorescence spectroscopy (CVAFS). This procedure was based on EPA Method 1631, Gill and Fitzgerald (1 987) and Bloom (1989). Methylmercury was determined by steam distillation, aqueous phase ethylation using sodium tetraethylborate, purging onto TenaxTMpacked columns, gas chromatography separation and CVAFS detection following a technique by Bloom and Fitzgerald (1988) and Horvat et al. (1993), modified by Branfueun et al. (1999). The same methods were used for samples spiked with 199~g(~~)isotope except detection was made using a CP and HP4500 ICP-MS (Gill and Fitzgerald, 1987, Hintelman and Evans, 1997). The method detection limit (MDL) for total mercury in sediment solids was 0.19 ng g-' (n=19), determined as three times the standard deviation of the mean of the sample blanks. For aqueous samples, the MDL based on a 150 rnL sample volume was 0.041 ng L-' (n=18). Precision, measured as the relative percent difference (RPD) between digest duplicates (sediment solids), and analytical duplicates (aqueous phase) was 9.6% (n=24 pairs) and 5.7% (n=2 pairs) respectively. Calibration curves of at least 3 = 0.99 were achieved daily or samples were re-run. Accuracy was measured both by spike recoveries and using the MESS3 marine sediment certified reference material (91 * 9 ng g-l) fiom the National Research Council of Canada. Recoveries averaged 103% * 10% (n=12) and 92 * 16 ng g-' for all MESS-3 samples (n=9). Samples fiom runs with poor recoveries (~80%) were re-analyzed. For MMHg, the MDL was 0.041 ng L-' (n=6) in the aqueous phase and 0.007 ng g-' (n=16) for sediment solids. For ICP-MS, the detection limit was 0.015 ng g-l and sample reproducibility was 10% for MMHg concentration and 23% for the CH~'"H~ isotope concentration. Isotope detection was further constrained to 0.5% of the ambient concentration. The RPD for distillation duplicates was 18.1% (n=2 I), while the average recovery of spikes between 100-500 pg per gram of wet sediment was 106 * 26% (n=12). Some of this variability can be attributed to uncertainty as to the true concentration of the sediment sample being spiked, as reflected in the RPD of distillation duplicates. A wet to dry weight conversion was determined for each sample analyzed by oven drying sub- samples of wet sediments for at least 24 hours at 60•‹C. 4.3.4 Statistical Analysis Non-parametric bivariate correlation matrices (Spearman rank correlation coefficients, r,) were developed for gravity core and push core data to investigate covariation between total mercury (Hg-T), MMHg, %MMHg and potential methylation rates. For gravity core data, Hg-T profiles were analyzed as a function of other metals with known anthropogenic origins such as Pb and other geochemical data including pore water sulfate and ammonium concentrations. 4.4 Results 4.4.1 S~eciationin the Active Sediment Laver Figure 4-2 shows mercury speciation in sediment cores from contrasting physical regions. The St. Croix River station (SC-1) is located at the head of the St. Croix River Estuary where the sediment accumulation rate was estimated to be between 1-2 rnm w' based on the gradients in dissolved ammonium found in the gravity cores using the method developed by Cranston (199 1,1997). In contrast, stations PB-3 and PB-4 are located on opposite sides of the main basin of P. Bay (Fig 4-I), and are both located near the center of tidally dominated circulation gyres (Greenberg et al., 1997) that are expected to result in significant mixing of these sediments. 4.4.2 Total Mercurv (HE-T) At station SC-1 (Fig 4-2a) there is a pronounced subsurface peak in total mercury (Hg-T) that may be the result of historic mercury discharges from the chlor-alkali facility that operated along the river in the 1970s. Concentrations of Hg-T in the sediments decrease from the head of the river estuary into the center of P. Bay. At station PB-3 (Fig 4-2b), located directly downstream of the mouth of the river estuary in the center of P. Bay, there is also evidence of subsurface increase in Hg-T that is likely the result of the same source of historical pollution. However, there are no distinct peaks in Hg-T concentrations with depth at station PB-3, which supports the effects of greater physical mixing in this region. The uniform Hg-T profile at station PB-4 on the opposite side of P. Bay (Fig 4-2c) provides further evidence that these sediments are also well-mixed. Figure 4-2. Speciation of mercury in contrasting depositional (SC-1) and well-mixed sediments (PB-3 & PB-4). Data include ambient total mercury (Hg-T), methylmercury (MMHg) and the fraction of total mercury in the sediments present as MMHg (%MMHg). Mercury isotope experiments were used to estimate the "potential" methylation rates at the head of the St. Croix River Estuary (SC-1) and the center of Passamaquoddy Bay (PB-3). (a) Head of St Croix River Estuary (SC-1) 100 200 300 0 400 8001200 0 0.4 0.8 0 0.08 0.16 Hs-T (ng g" dry) MMHg (pg g-' dry) % MMHg Potential methylation (b) Center of Passamaquoddy Bay (PB-3) rate (day1) 40 60 80 240 320 400 0 0.4 0.8 0 0.08 0.16 Hs-T (ng g" dry) MMHg (pg g" dry) % MMHg Potential methylation rate (day1) (c) Center of Passamaquoddy Bay (PB-4) 30 40 50 60 240 320 400 0 0.4 0.8 6 8 101214 Hs-T (ng g" dry) MMHg (pg g-I dry) % MMHg Porewater Hg-T (ng L-I) 4.4.3 Methvlrnercuw MMHn) The effects of differing physical dynamics on MMHg production are illustrated by the contrasting MMHg profiles at the head of the river estuary (SC- 1) relative to the well- mixed areas in the center of P. Bay (PB-3 and PB-4). The fraction of total mercury present as methylmercury (%MMHg) in these sediments is strongly correlated with potential methylation rates measured in cores spiked with mercury isotopes (r, = 0.88, p supporting the premise that %MMHg is a reasonable approximation of the relative rates of Hg methylation in these sediments. This relationship has also been seen in other estuarine sediments (Benoit et al., 2002, Heyes et al., Submitted). At station SC-1, MMHg production is taking place in a narrowly constrained subsurface zone. There is a subsurface peak between 2-4 cm in ambient MMHg concentrations, %MMHg and "potential" methylation rates measured in isotope cores. Ambient MMHg concentrations decline rapidly in the oxic surface layer and beyond depths of several centimeters, while methylation rates indicated by both the %MMHg and the isotope core data are low to non-detectable, respectively, in these depth intervals. For example, the %MMHg ranges between 0.54% and 0.76% in the 2-4 cm subsurface peak in MMHg production at station SC-1, compared to 0.28% in the surface layer and between 0.1 1% and 0.34% at depths greater than 6 cm. This profile is typical of those observed in other studies of lake and estuarine sediments that showed the %MMHg in estuarine sediments is generally less than 0.5%, particularly in the oxic surface sediments and at depth, and that MMHg production occurs in a relatively narrow subsurface zone within the sediments (e.g., Benoit et al., 1998a, Bloom et al., 1999, Gagnon et al., 1996). In the mixed sediments at stations PB-3 and PB-4 the ambient MMHg profiles and %MMHg suggest that MMHg conversion is taking place at all depths in the surface sediment layer. Methylmercury concentrations and %MMHg are both relatively uniform at all depths (Fig 4-2b,c) and potential methylation rates measured at station PB-3 are detectable throughout the entire profile studied. At station PB-4, the %MMHg ranges between 0.62% and 0.92%, which is in the same range as the 2-4 cm subsurface peak in %MMHg at site SC- 1 that corresponds to maximum methylation rates. Additionally, the mean %MMHg in integrated surface samples (0-10 cm depth) fiom multiple locations throughout P. Bay (n= 45) was 0.88%, again characteristic of sediments where MMHg is being actively produced in situ in the sediment column. These data support the premise that in the well-mixed sediments of P. Bay, MMHg production is taking place both at the sediment water interface and throughout the 15 cm surface layer. 4.4.4 Porewater-Solids Partitioning Porewater samples in this study represent "operationally defined" dissolved concentrations of total mercury and methylmercury since colloidal matter binds mercury in the c0.2 pm size fraction (Guentzel et al., 1996). Porewater concentrations of Hg-T ranged between 10 and 30 ng L-' (Table 4-1) and are significantly elevated above the levels expected on the basis of solid phase concentrations, as they are in the same range as sites heavily impacted by historical pollution such as Lavaca Bay, Texas (Bloom et al., 1999). Accordingly, the range in partition coefficients (log Kd) for Hg-T between 3.12 - 3.76 (L kg-') in P. Bay is lower than those observed in other systems (Bloom et al., 1999, Leermakers et al., 1995, Turner et al., 2001). Table 4-1. Porewater Hg-T and MeHg data from each sampling stations SC-1, PB-1 through PB-5. Station Hg-T (ng L-') MMHg (ng L-') SC- 1 27 0.47 PB-1 30" 0.94a PB-2 11 NIA PB-3 16" 0.55" PB-4 10 0.68~ PB-5 1 6b 1 .44" "Means of triplicate samples %4eans of duplicate samples Elevated levels of Hg-T in porewaters of P. Bay sediments are consistent with the effects of mixing in this system that is causing substantial recycling of Hg in the surface sediments and an increased fraction of colloidally bound Hg-T in the porewaters. Porewater MMHg concentrations are also elevated relative to systems with similar Hg-T concentrations in sediments such as the Patuxent River, Maryland (Benoit et al., 1998a), but partition coefficients (log Kd) ranging between 2.20-3.00 (L kg-') are in the same range as values found in other studies (Benoit et al., 1998a, Bloom et al., 1999). These results suggest that MMHg concentrations are enriched in both the solid and dissolved phases and support the hypothesized increase in MMHg production in well-mixed sediments. 4.4.5 Geochemical Characteristics of the Sediment Column Figure 4-3 shows porewater ammonium (N&3 and sulfate levels measured in gravity cores collected in P. Bay. Porewater N&+ concentrations that are consistently >0.5 mh4 occur below a depth of 30 cm at station PB-3 (Fig 4-3) and below 15 cm at station PB-4 (Table 4- 2). Porewater NH~+concentrations >0.5 mM indicate fully reduced anoxic sediments that are below the active sediment layer due to lack of oxidation or ammonium produced by decomposition of organic matter (Buckley, 1991, Buckley and Cranston, 1988) and allow an upper boundary to be placed on the depth of the active sediment layer. These sediments may be considered truly buried and effectively removed from fbrther interaction with the sediment water interface. Data fiom other cores collected throughout P. Bay further confirm that the depth of the active sediment layer based on porewater ammonium data is generally between 15-30 cm (Table 4-2). This relatively large volume 118 Figure 4-3. Porewater sulfate and ammonium concentrations in gravity cores from Passamaquoddy Bay and the St. Croix River Estuary. Sulfate concentrations <24 rnM indicate the anoxic reduction of sulfate and ammonium concentrations >0.5 mM indicate the presence of fully reduced buried sediments. Table 4-2. Geochemical characteristics of gravity cores from Passamaquoddy Bay. Station Length Organic m+> 0.5 mM ~04'-<24mM Range in ID (cm) Carbon (%) depth interval(s) depth interval(s) ~04~-mMb (cm)" (cm) SC- 1 SC-2 PB- 1 PB-2 PB-3 PB-4 PB-5 PB-6 PB-7 PB-8 PB-9 PB- 10 PB-11 PB-12 PB- 13 PB-14 PB-15 PB-16 PB- 17" PB- 18 . - 0.44- 1.43 NI NI "Concentrations of NH~+>o.~mM measured in pore waters indicates lack of oxidation of ammonium produced by decomposition of organic matter. In a number of cores, the presence of consistently >0.5 mM m'at depth suggests truly reduced buried sediments. Minimum to maximum ~0~~-concentrations throughout core measured in pore waters are reported. Note that the range in SO:- concentrations are all very close to the geochemical cutoff (24 rnM) indicating anoxic reduction of sulfate. This supports the supposition that most of these sediments are in the oxic-transitory range where there is a large gradient in redox potential, ideal for the activity of sulfate-reducing bacteria. "Top 30 cm of core PB-17 missing. NI = no indication of reduction. There is evidence of reduction in all cores except PB-14 and PB- 18, and even in these cores the sulfate levels are very close to the geochemical cutoff characterizing reducing conditions. of sediments in the active sediment layer results in a large pool of historic Hg-T that can potentially be converted to MMHg. Data on porewater sulfate levels in gravity cores from P. Bay indicate that sulfate reduction is taking place at all depths within these sediments, including at the sediment surface (Fig 4-3, Table 4-2). Anoxic reduction of sulfate based on a geochemical threshold established in the literature is indicated by porewater sulfate concentrations <24 mM (Buckley, 1991, Buckley and Cranston, 1998). In all P. Bay sediments, porewater sulfate levels are very close to this level, supporting the idea that these sediments are in the oxic-transitory range that is most conducive for the activity of SRB. Although porewater sulfate data (Fig 4-3) indicate anaerobic reduction of sulfate, which typically corresponds to a redox potential of approximately -200 mV (Wildish et al., 1999), redox potentials (Eh) measured in surface sediments varied between -2 13 and 5 12 mV (Chapter 3). The large variability in redox measurements and length of time (>5 minutes) needed in the field for the redox probe to achieve equilibrium may both signify the presence of Eh microniches in the surface layer (Wildish et al., 1999), however, these data should be interpreted with caution as variability may also reflect measurement errors associated with these types of Eh measurements. Sulfide concentrations measured in the surface sediments using an ion specific electrode were also highly variable, ranging between 10 and 4000 pM in replicate samples taken at a single sampling station (Table 4- 3)- Table 4-3. Variability in sulfide concentrations measured in P. Bay sediments. Sulfide Concentrations (uM) Station Mean N min max PB- 1 213 9 23 1500 PB-2 659 7 190 1300 PB-3 554 12 32 1100 PB-4 667 6 480 800 PB-5 1471 9 10 4000 PB-6 363 6 33 1000 PB-8 167 3 42 370 PB-9 393 4 30 1200 PB-10 92 3 36 150 The large variability in sulfide concentrations and the presence of Eh microniches show rapid transitions in the sediment geochemistry in the surface layer that provide conditions most favorable to methylating microbes and are indicative of the effects of organic rich "mottles" or "pockets" in the surface sediments. Visual inspection of sediment cores fi-om P. Bay in the field revealed the presence of black, organic rich mottles interspersed with the dominant light brown oxic clay muds. There was also detectible H2S odour in the presence of these mottles, indicating anoxia. Based on these data we hypothesize that the anoxic organic rich "pockets" in P. Bay sediments effectively increase the volume of sediments suitable for formation of MMHg by SRB by increasing the transition zone between oxic and anoxic sediments. The resulting geochemical environment would be comparable to that observed by Tseng et al. (2001) of a turbid macrotidal estuary in France where methylation of mercury was measured in analogous "anoxic pockets" occurring within a fluid mud profile. These organic carbon "pockets" facilitate MMHg production throughout the relatively deep active sediment layer in P. Bay, thereby accounting for the observed enrichment in MMHg in the sediment porewaters and the high %MMHg in the bulk phase. 4.4.6 Anthroponenic Mercurv in Passamaauoddv Bay Anthropogenic sediment enrichment factors (ASEFs) were calculated fi-om Hg-T concentrations measured in sediment cores at stations SC- 1 and PB-3. Mixing of these sediments means it is not possible to obtain detailed information on historical Hg-T loading fi-om these cores and that traditional dating methods using 2'0~band 131cscould not be applied. However, ASEFs calculated from the difference between mercury concentrations in the sediments that accumulated prior to human influence and those at the surface provide a simple method for estimating the overall enrichment in this system resulting fiom anthropogenic pollution. Present day burial rates of approximately 1-2 mm per year measured in this study suggest that sediments below 40 cm depth in gravity cores should represent Hg-T concentrations in the sediments prior to significant human sources of mercury, while allowing a wide margin for integration of the sediment column due to mixing. Diagenetic remobilization of mercury in sediment cores can cause peaks in mercury concentrations in the surface sediments that are not indicative of anthropogenic pollution (Benoit et al., 1998b, Walton-Day et al., 1990). However, the lack of significant correlations between Hg-T concentrations and redox sensitive metals Fe and Mn (Table 4-4) suggests that redistribution of mercury through difhsion and co-precipitation is not a significant factor in gravity cores fiom P. Bay. To isolate the natural and anthropogenic components of metal enrichment, iron (Fe) and lithium (Li) can both be used as normalizing factors to correct for the mineralogical and granulometric variability in the sediments (Loring, 1991). However, there were no significant correlations between Hg-T and Fe or Li in these sediments, thus it can be assumed that grain size does not vary appreciably throughout the sediment profile. The range of "background" concentrations (i.e., those that are naturally occurring) of Hg-T in P. Bay cores was estimated fiom 95% confidence limits around Hg-T concentrations below 40 cm in the gravity cores. Anthropogenic sediment enrichment factors shown in Figure 4-4 include the surface of gravity cores and push cores collected at stations SC-1 and PB-3. Table 4-4. Concentrations of metals measured in sediment cores. Figure 4-4. Anthropogenic sediment enrichment factors (ASEFs) for Hg-T at the head of the St. Croix River and the center of Passamaquoddy Bay. ASEFs are used to estimate modernhackground Hg-T levels in P. Bay sediments. Head of St. Croix River Center of Passamaauoddv Wua? !SC-11 Bav (PB-3) 4 8 12 2.4 3.2 4 ASEF ASEF 95% CLs 95% CLs I I I I I 1 1 I I I For surface sediments (0- 14 cm), Hg-T concentrations measured in the upper horizon of the sediments were divided by the upper 95% confidence limit value for background mercury levels to produce the plots of ASEFs shown in Figure 4-4. Visual inspection of the gravity core collected at station SC-1 in the field indicated the loss of the upper 10 cm of the surface layer and ASEF calculations have been adjusted accordingly in Figure 4-4. The results presented in Figure 4-4 show that ASEFs in P. Bay range between 1.9 and 4.0, and up to 12.4 at the head of the St. Croix River where localized historical discharges are likely the most important source of contamination. The enrichment factors for P. Bay are consistent with other studies that show a global scale increase in atmospheric mercury deposition of 2-4 times the pre-industrial levels (Engstrom et al., 1994, Swain et al., 1992), suggesting that the majority of Hg-T in these sediments is derived from atmospheric sources. The high degree of correlation between anthropogenic lead and Hg-T concentrations (r, = 0.75, p<0.01; Table 4-4) provides further evidence of anthropogenic enrichment in these cores. The results of this analysis indicate the majority of the mercury in the active sediment layer of P. Bay is from historic inputs of anthropogenic mercury. 4.5 Discussion We hypothesized that mixing of the active sediment layer in P. Bay results in geochemical changes in the sediment column that enhance the activity of methylating microbes. In marine sediments, net methylation rates are highest in the transition zone between oxic and anoxic conditions because these conditions are most conducive to the activity sulfate reducing bacteria (SRB) (King et al., 2001, Hintelmann et al., 2000). In addition, these microbes require organic matter as a substrate for microbial activity (Mason and Lawrence, 1999). Physical mixing in the Bay of Fundy may enhance the transfer of sulfate and carbon and introduces more bioavailable inorganic mercury into the deeper sediment, potentially stimulating the methylating activity of SRB. Regular disturbances through mixing also appear to create a unique geochemical environment more likely to exhibit microzonal redox gradients, compared to the classic down profile gradients of sediments in which more static conditions may limit the activity of microbial populations that methylate mercury. The observed effects of mixing on mercury speciation profiles and potential methylation rates in the active sediment layer and the geochemistry of sediments are contrasted with the classic profile of mercury speciation in more static sediments in the conceptual diagram presented in Figure 4-5. As illustrated in Figure 4-5, production of MMHg in sediments fiom unmixed depositional systems must be modeled as a two-compartment system, taking into account that MMHg production occurs in a narrow zone in the redoxcline below the oxic surface layer of these sediments. In well-mixed systems such as P. Bay, the sediment compartment is better described by System 2 (Fig 4-5) where MMHg production occurs throughout a relatively deep active sediment layer and is facilitated by the presence of organic-rich anoxic "pockets" or mottles. Additionally, MMHg production occurs at the sediment-water interface in the well-mixed sediments, potentially providing a vector for MMHg entry into the water column and resulting in exposure of organisms feeding at the sediment surface. The dynamic physical mixing occurring in P. Bay results in the creation of an approximately 15-cm-thick active sediment layer. Burial provides the main removal Figure 4-5. Conceptual model of mercury speciation in contrasting depositional and well- mixed sediments. mechanism for historic mercury that has accumulated over time. In P. Bay, the sediment burial rate estimated from this study is approximately 1 mm w' . This means that the 15 cm active sediment layer is comprised of a relatively large reservoir of Hg-T consisting of mainly historical pollution (Fig 4-4) that has accumulated in the sediments over many decades. In contrast, in a depositional system, the active layer is typically much shallower (e.g., 3-5 cm) resulting in a much smaller reservoir of Hg-T that can be potentially converted to MMHg by methylating microbes. The physical mechanism of mixing may also provide a vector for MMHg entry into the water column and food-web through organisms feeding at the sediment-water interface. In the well-mixed sediments, MMHg production occurs throughout the active sediment layer, including at the sediment-water interface as illustrated in the push core profiles for P. Bay (Fig 4-2b,c). In contrast, the oxic sediment layer in depositional systems can act as a geochemical barrier to diffising MMHg through the precipitation of MMHg with Fe and Mn hydroxides (Gagnon et al., 1996). In depositional areas, this oxic layer can inhibit the entry of MMHg to the water column and limit exposure of all but burrowing benthic organisms. In the well-mixed sediments of P. Bay, the enhanced levels of Hg-T and MMHg in the porewaters of surface sediments are consistent with rapid cycling of mercury in this system at the sediment-water interface that is facilitated by the physical dynamics of the area. Preliminary evidence for increased availability of mercury to organisms in well-mixed sediments is provided by the observed differences in the mean concentrations of total mercury measured in polychaetes collected from sediment grabs at the head of the St. Croix River Estuary (n=3) compared to well-mixed sediments in P. Bay (n=10). These data show that concentrations of mercury in polychaetes are higher in the well-mixed sites (1 1 + 3 ng g" wet wt.) than at the head of the river estuary (8+ lng g"), despite the fact that concentrations of Hg-T in P. Bay sediments are approximately five times lower than concentrations at the head of the river estuary. The results of this study help to elucidate why concentrations of mercury in organisms fiom this coastal system are still high despite large emissions reductions. In P. Bay, there is a linear relationship between Hg-T and MMHg in the sediments (Chapter 3), which means that increases or decreases in Hg-T emissions should eventually translate into corresponding changes in MMHg concentrations in both sediments and ultimately organisms. The integration of historic Hg inputs over the 15 cm active layer that is slowly being converted to MMHg and the relatively slow removal via sediment burial could explain the large lag time between emissions reductions and changes in ambient concentrations. This hypothesis is currently being further tested through the application of a mercury cycling model for the region. 4.6 Literature Cited Benoit, J., Gilmour, C. C., Mason, R. P. and Heyes, A. 1999. Sulfide controls on mercury speciation and bioavailability to methylating bacteria in sediment pore waters. Environmental Science and Technology, 33,95 1-957. Benoit, J. M., Gilmour, C. C., Heyes, A., Mason, R. P. and Miller, C. 2002. Geochemical and biological controls over methylmercury production and degradation in aquatic systems. In: Biogeochemistry of Environmentally Important Trace Metals. ACS Symposium Series No. 835. (Eds, Cai, Y. and Braids, O.C.) Oxford University Press, New York, pp. 262-297. Benoit, J. M., Gilmour, C. C., Mason, R. P., Riedel, G. S. and Reidel, G. F. 1998a. Behavior of mercury in the Patuxent River estuary. Biogeochemistry, 40,249- 265. Benoit, J. M., Fitzgerald, W. F. and Damman, A. W. H. 1998b. The biogeochemistry of an ombrotrophic peat bog: Evaluation of use as an archive of atmospheric mercury deposition. Environmental Research, 78, 118- 133. Bloom, N. and Fitzgerald, W. F. 1988. Determination of volatile mercury species at the picogram level by low-temperature gas chromatography with cold-vapor atomic fluorescence detection. Analytica Chemica Acta, 208, 151-1 61. Bloom, N. S. 1989. Determination of picogram levels of methylmercury by aqueous phase ethylation, followed by cryogenic gas chromatography with cold vapor atomic fluorescence detection. Canadian Journal of Fisheries and Aquatic Sciences, 46, 113 1-1 140. Bloom, N. S. 1992. On the chemical form of mercury in edible fish and marine invertebrate tissue. Canadian Journal of Fisheries and Aquatic Sciences, 49, 10 10- 1017. Bloom, N. S., Gill, G. A., Cappellino, S., Dobbs, C., Mcshea, L., Driscoll, C., Mason, R. and Rudd, J. 1999. Speciation and cycling of mercury in Lavaca Bay, Texas, sediments. Environmental Science and Technology, 33,7- 13. Branfueun, B., Roulet, N. T., Kelly, C. A. and Rudd, J. W. M. 1999. In situ stimulation of mercury methylation in a boreal peatland: Toward a link between acid rain and methylmercury contamination in remote environments. Global Biogeochemical Cycles, 13,743-750. Buckley, D.E. 1991. Deposition and diagenetic alteration of sediment in Emerald Basin, the Scotian Shelf. Continental Shelf Research, 11, 1099-1 122. . Buckley, D. E. and Cranston, R.E. 1988. Early diagenesis in deep sea turbidites: the imprint of paleo-oxidation zones. Geochimica et Cosmochimica Acta, 52,2925- 2939. Canadian Council of Ministers of the Environment. 2000. Canadian tissue residue guidelines for the protection of wildlife consumers of aquatic biota: Methylmercury. In: Canadian Environmental Quality Guidelines, 1999. Winnipeg, Manitoba. Chase, M. E., Jones, S. H., Hennigar, P., Sowles, J., Harding, G. C. H., Freeman, K., Wells, P. G., Krahforst, C., Coombs, K., Crawford, R., Pederson, J. and Taylor, D. 2001. Gulfwatch: Monitoring spatial and temporal patterns of trace metal and organic contaminants in the Gulf of Maine (1 99 1- 1997) with the blue mussel, Mytilus edulis L. Marine Pollution Bulletin, 42,49 1-505. Choi, S.-C. and Bartha, R. 1994. Environmental factors affecting mercury methylation in estuarine sediments. Bulletin of Environmental Contamination and Toxicology, 53,805-812. Cranston, R. E. 1991. Sedimentation rate estimates from sulfate and ammonia gradients. Proceedings of the Ocean Drilling Program, Scientific Results, 119,40 1-405. Cranston, R. E. 1997. Organic carbon burial rates across the Arctic Ocean fiom the 1994 Arctic Ocean Section expedition. Deep Sea Research 11, 44, 1705-1723. Engstrom, D. R., Swain, E. B., Henning, T. A., Brigham, M. E. and Brezonlk, P. L. 1994. Atmospheric mercury deposition to lakes and watersheds: A quantitative reconstruction fiom multiple sediment cores. In: Environmental Chemistry of Lakes and Reservoirs (Ed, Baker, L. A.) American Chemical Society, Washington, DC, pp. 33-66. Evers, D.C., Kaplan, J.D., Meyer, M.W., Reaman, P.S., Braselton, W.E., Major, A., Burgess, N. and Scheuharnrner, A.M. 1998. Geographic trends in mercury measured in common loon feathers and blood. Environmental Toxicology and Chemistry, 17, 173-183. Fink, L. K. J., Pope, D. M., Harris, A. B. and Schick, L. L. 1976. Heavy Metal Levels in Suspended Particulates, Biota, and Sediments of the St. Croix Estuary in Maine. Land and Water Resources Institute, University of Maine at Orono, Orono, Maine, pp. 59. Gagnon, C., Pelletier, E., Mucci, A. and Fitzgerald, W. F. 1996. Diagenetic behavior of methylmercury in organic-rich coastal sediments. Limnology and Oceanography, 41,428-434. Gill, G. A. and Fitzgerald, W. F. 1987. Picomolar mercury measurements in seawater and other material using stannous chloride reduction and two-stage gold amalgamation with gas phase detection. Marine Chemistry, 20,227-243. Greenberg, D., Shore, J. and Shen, Y. (1997). Modelling tidal flows in Passamaquoddy Bay. In: Coastal Monitoring and the Bay of Fundy: Proceedings of the Maritime Atlantic Ecozone Science Workshop (Eds, Burt, M. D. B. and Wells, P. G.) Huntsman Marine Science Centre, St. Andrews, NB. Gregory, D., Petrie, B., Jordan, F. and Langille, P. 1993. Oceanographic, geographic and hydrological parameters of Scotia-Fundy and southern Gulf of St. Lawrence inlets. Canadian Technical Report of Hydrographic Ocean Sciences No. 143. Department of Fisheries and Oceans, Scotia-Fundy Region, Dartmouth, NS, pp. 248. Guentzel, J. L., Powell, R. T., Landing, W. M. and Mason, R. P. 1996. Mercury associated with colloidal material in an estuarine and open-ocean environment. Marine Chemistry, 55,177-188. Heyes, A., Gilmour, C. C., Clark, M., Heyes, D. B. and Mason, R. P. Submitted. The effect of sediment disturbances on methylmercury production in near shore estuarine sediment. Estuaries. Hintelmam, H. and Evans, R. D. 1997. Application of stable isotopes in environmental tracer studies - measurement of monomethylmercury by isotope dilution ICP-MS and detection of species transformation. Fresenius Journal of Analytical Chemistry, 358,378-385. Hintelmam, H., Keppel-Jones, K. and Evans, R. D. 2000. Constants of mercury methylation and demethylation rates in sediments and comparison of tracer and ambient mercury availability. Environmental Toxicology and Chemistry, 19, 2204-22 11. Horvat, M., Liang, L. and Bloom, N. S. 1993. Comparison of distillation with other current isolation methods for the determination of methyl mercury compounds in low level environmental samples. Analytica Chemica Acta, 282,153-168. King, J. K., Kostka, J. E., Frischer, M. E., Saunders, F. M. and Jahnke, R. A. 2001. A quantitative relationship that demostrates mercury methylation rates in marine sediments are based on community composition and activity of sulfate-reducing bacteria. Environmental Science and Technology, 35,2491 -2496. King, J. K., Saunders, M., Lee, R. F. and Jahnke, R. A. 1999. Coupling mercury methylation rates to sulfate reduction rates in marine sediments. Environmental Toxicology and Chemistry, 18, 1362-1369. Leermakers, M., Meuleman, C. and Baeyens, W. 1995. Mercury speciation in the Scheldt Estuary. Water, Air, and Soil Pollution, 80,64 1-652. Loring, D. H. 1991. Normalization of heavy-metal data from estuarine and coastal sediments. ICES Journal of Marine Science, 48, 101 - 1 15. Mason, R. P. and Lawrence, A. L. 1999. Concentration, distribution, and bioavailability of mercury and methylmercury in sediments of Baltimore Harbor and Chesapeake Bay, Maryland, USA. Environmental Toxicology and Chemistry, 18,2438-2447. Northeast States for Coordinated Air Use Management (NESCAUM), Northeast Waste Management Officials' Association (NEWMOA), New England Interstate Water Pollution Control Commission (NEIWPCC), Ecological Monitoring and Assessment Network (EMAN), 1998, Northeast States and Eastern Canadian Provinces Mercury Study, A Framework for Action: NESCAUM, pp. 350. Sunderland, E. M. and Chmura, G. L. 2000. An inventory of historical mercury pollution in Maritime Canada: Implications for present and future contamination. The Science of the Total Environment, 256,39-57. Swain, E. B., Engstrom, D. R., Brigham, M. E., Henning, T. A. and Brezonik, P. L. 1992. Increasing rates of atmospheric mercury deposition in midcontinental North America. Science, 257,784-786. Tseng, C. M., Amouroux, D., Abril, G. and Donard, 0. F. X. 2001. Speciation of mercury in a fluid mud profile of a highly turbid macrotidal estuary (Gironde, France). Environmental Science and Technology, 35,2627-2633. Turner, A., Millward, G. E. and Row, S. M. L. 2001. Sediment-water partitioning of inorganic mercury in estuaries. Environmental Science and Technology, 35,4648- 4654. Walton-Day, K., Filipek, L. H. and Papp, C. S. E. 1990. Mechanisms controlling Cu, Fe, Mn, and Co profiles in peat of the Filson Creek Fen, northeastern Minnesota. Geochimica et Cosmochima Acta, 54,2933-2946. Wildish, D. J., Akagi, H. M., Hamilton, N. and Hargrave, B. T. 1999. A Recommended Method for Monitoring Sediments to Detect Organic Enrichment from Mariculture in the Bay of Fundy. Canadian Technical Report of Fisheries and Aquatic Sciences No. 2286, St. Andrews, N.B., pp. 30. CHAPTER 5 An Empirical Model of Mercury Cycling in Passamaquoddy Bay, New Brunswick Everything should be made as simple as possible, but not simpler Albert Einstein 5.1 Abstract A major problem faced by environmental protection managers in Atlantic Canada is that, despite large reductions in mercury emissions, there is no evidence to indicate that mercury concentrations in aquatic organisms have fallen. The goal of this study is to develop an analytical framework for understanding and predicting the behavior of mercury in a coastal ecosystem by developing a model for mercury cycling in Passamaquoddy Bay (P. Bay). Mass budgets for mercury show that: (i) the sediment compartment contains over 95% and 90% of the total and methyl-mercury respectively; (ii) concentrations in the sediments can be expected to respond very slowly to changes in mercury inputs; (iii) there is a large turnover of mercury on a daily basis through methylation and demethylation, equivalent to approximately 40% of the methylmercury reservoir in the sediments. The findings of this study indicate that: (i) a shift in the present equilibrium between methylation and demethylation could result in rapid methylmercury accumulation in the sediments and subsequent increases in uptake by organisms; (ii) while concentrations in the water column of P. Bay reach steady state rapidly in response to changes in mercury inputs, the temporal response of concentrations in this system is governed by the slow rate of change in sediment mercury concentrations. The model suggests it will take on the order of centuries for P. Bay to reach steady state with respect to "natural" mercury inputs if emissions from anthropogenic sources are virtually eliminated. These results emphasize that short-term management strategies must focus on minimizing anthropogenic changes that stimulate methylmercury production, while long-term management strategies must continue initiatives aimed at reducing mercury emissions with eventual reductions in sediment and organism mercury concentrations. 5.2 Introduction A major problem faced by environmental protection managers in Atlantic Canada is that, despite reductions in mercury emissions of more than 50% since the 1970s (Sunderland and Chmura, 2000), there is no evidence to indicate that concentrations in aquatic organisms have fallen. In the Bay of Fundy and Gulf of Maine, mercury concentrations in shark, swordfish and tuna exceed the Health Canada safe consumption limit of 0.5 ppm (Burns, 1998, Harding, unpublished data, 2002) and high concentrations of mercury in marine mammals (>I ppm) and avifauna have been observed (Gaskin et al., 1973, Gaskin et al., 1979, Elliott et al., 1992). Consumption advisories for human populations eating fish are in place throughout the Atlantic region due to elevated levels of mercury (NESCAUM et al., 1998) and sensitive sub-populations such as aboriginal communities and pregnant women that consume large quantities of fish are believed to be at risk. These data suggest that mercury may soon be a problem at lower trophic levels in the marine food-web, potentially affecting 20,000 U.S. and Canadian commercial fishers that depend on the marine resources in this region (Gulf of Maine Council on the Marine Environment, 1994). To alleviate this problem, an agreement between the New England Governors and Eastern Canadian Premiers to hrther reduce anthropogenic emissions by 50% was ratified (NEG-ECP, 1998). However, the effectiveness of this policy remains unknown. Understanding how changes in present-day mercury inputs affect current and future concentrations in water, sediments and biota is critical for evaluating the success of current initiatives regulating mercury releases from human sources. The goal of this study is to develop an analytical framework for understanding and predicting the behavior of mercury in a coastal ecosystem. In this paper, mass budgets are presented for total and methyl-mercury in Passamaquoddy Bay (P. Bay), an embayment located at the mouth of the Bay of Fundy. Field data are used to generate contour maps for total and methyl-mercury concentrations and burial fluxes in P. Bay and the St. Croix River Estuary. Modeling efforts were limited to P. Bay and do not include the St. Croix River Estuary because the dynamics of mercury in the river are distinct from those in P. Bay due to the influence of significant historical contamination of mercury at the head of the river (Chapter 3). Data fiom P. Bay are incorporated into spatially weighted mass budget calculations. Reservoirs of mercury in the sediment and water compartments as well as mass fluxes between media are quantified using measured concentrations of mercury and physicaVhydrodynamic data for the region. Measured concentrations of total mercury and methyl mercury in the water and sediments are compared to those in benthic organisms to estimate biota-sediment accumulation factors for P. Bay. The integration of all the empirical data collected in P. Bay to describe the fate of mercury and concentrations in the benthic food-web provides insight into the relationship between inputs of mercury and concentrations in organisms. The mass balance equations presented here describe the movement of mercury in and out of water and sediments using rate constants for mercury transport that are derived fiom the empirical data collected in P. Bay (see previous chapters). There are three principal forms of mercury considered in the model presented: elemental mercury (H~'),divalent mercury (Hg(I1)) and monomethylmercury (MMHg). Although dimethylmercury is hypothesized to occur as a significant byproduct of the demethylation process (Mason and Fitzgerald, 1991), it is not included in the model because past studies of coastal systems indicate that this species is not present in detectable concentrations in coastal waters or sediments (Mason et al., 1993). Some of the most useful ecosystem based mercury cycling models are those developed by Harris et al. (1 996) for lakes and Diamond et al. (2000) for freshwater reservoirs. There are a number of other related, more specific models (e.g., Baeyens et al., 1991, Henry et al., 1995, Meili, 1991). However, a comparable model for the marine environment that can relate mercury loadings to concentrations in a marine ecosystem is a novel contribution. The existing freshwater models are not easily extended to the marine environment because of differences in the factors controlling the rate of MMHg formation in marine sediments, the principal species of mercury that bioaccumulates in the food web (Spry and Wiener, 1991, Weiner et al., 1990, Wang et al., 1998). By developing our understanding of the fate of mercury in a coastal ecosystem, the empirical model of mercury cycling in a marine ecosystem presented in this study contributes to the development of a mechanistic model for marine ecosystems that can be used to forecast the relationship between mercury inputs and concentrations in marine biota. 5.3 Methods 5.3.1 Study Site Passamaquoddy Bay (P. Bay) is a semi-enclosed macrotidal estuary. Presently, there are no large point sources of mercury in P. Bay. However, in the 1970's several pulp and paper plants and one chlor-alkali facility discharged mercury into the catchment area of the St. Croix Estuary (Fink et al., 1976). 5.3.2 Model Develovment The model presented in this paper is based on empirical measurements of mercury concentrations and fluxes in P. Bay collected in the field over the course of this study. As such, the model represents a generalized "snapshot" of mercury dynamics in P. Bay. The model was not developed to capture the mechanistic variables affecting the spatial and temporal dynamics in P. Bay and therefore has restricted applicability as a tool used to forecast the long-term temporal dynamics of mercury in this system. However, integration of field data for development of the empirical mercury cycling model allow the relative magnitude of mercury fluxes among the sediments, water and atmosphere to be quantified, thereby providing insight into the response time of P. Bay to changes in external inputs of mercury. A summary of calculation methods used to estimate each parameter in the empirical model of mercury cycling developed for P. Bay is given below and a more complete description of assumptions made and data used are shown in the appendix. Field investigations were conducted to measure spatial variability in sediment mercury concentrations, net sedimentation rates and "potential" methylation rates using surface sediment samples, gravity cores and incubations of cores spiked with mercury isotopes respectively. Burial fluxes were estimated from gravity core data using gradients in dissolved ammonia and carbon with depth according to the method outlined by Cranston (1999). Further details of sediment sampling and analytical techniques for total and methyl-mercury are given in other papers (Chapters 3 and 4). Concentrations of Hg-T (n=4) and MMHg (n=2) in the water column of P. Bay were measured in November 2001. Seawater sample sizes in P. Bay were limited by the occurrence of a hurricane during the collection period resulting in some uncertainty associated with these numbers. Samples were integrated over water depths of 0-10 m using a Niskin water sampling device following ultraclean handling protocols (Mason et al., 1998, Bloom, 1989) and analyzed for total and methyl-mercury using the procedure for aqueous samples outlined in Chapters 3 and 4. Concentrations of mercury in polychaetes and amphipods in P. Bay were also measured in this study using the same analytical methods as for sediment samples. The remaining data used to compile the mercury mass budgets were obtained from other studies and are detailed in the description of model parameters. 5.4 Model Description The mercury cycling model developed for P. Bay describes physical and biological transport and transformation of mercury in the water column and benthic sediments based on steady state mass balance equations (Table 5-1) adapted fiom Gobas et al. (1998). Individual models are developed for total mercury (Hg-T), methylmercury (CH3Hg or MMHg), elemental mercury (~~4and divalent inorganic mercury (Hg(I1)). Inorganic mercury in the sediments and water is operationally defined as the difference between Hg-T and MMHg and consists of both divalent inorganic mercury species (Hg(I1)) and elemental mercury (H~'). In the empirical model, the benthic sediment compartment is divided into an "active layer" and a "truly buried" inaccessible sediment layer. The active sediment layer can potentially exchange mercury with the water column and buried sediments through resuspension, diffusion and burial. Truly buried sediments act as a sink for mercury, removing it from further interaction with the sediment-water interface and benthic organisms. The active sediment layer is also where conversion of inorganic mercury to methylmercury takes place, represented as a net loss of Hg(I1) and a net input of MMHg in individual models for mercury species. A mass balance is assumed for the active sediment layer, meaning that the settling flux is equal to the combined fluxes of solids resuspension and burial at all times. Although this is a simplification of the processes occurring in the active sediment layer, it is expected to be a reasonable representation of chemical dynamics associated with the net deposition and resuspension of sediment solids (Gobas et al., 1998). In the mercury model for P. Bay, measured mercury fluxes are converted to rate constants that describe the rate of reaction and transport of mercury species listed in Table 5-1. Specifically, the model describes transport and reaction as a function of: (i) direct inputs of mercury (e.g., atmospheric deposition), (ii) water inflow (rivers and tides) and outflow (tides), (iii) volatilization of elemental mercury (H~')fi-om the water column, (iv) sorption to suspended sediments, (v) deposition of suspended sediments, (vi) resuspension of benthic sediments, (vii) burial of sediments, and (viii) species interconversions in the sediments and water (e.g., methylatioc of Hg(I1) and reduction of Hg(I1) and CH3Hg to H~'). Ecosystem parameters for Passamaquoddy Bay are summarized in Table 5-2. 5.5 Treatment of Uncertainty in the Model Although both spatial and temporal variability in mercury concentrations and fluxes are present in all parameters discussed in the mercury cycling model, the deterministic solutions presented in this study are meant to represent the "most likely" state of nature based on measures of central tendency in the empirical data (mean and median sample values). On a temporal basis, this means model parameters have been averaged over an annual time step. Treatment of uncertainty in this manner is justifiable when the nature of the model presented is considered to be "empirical" rather than "predictive" in nature as discussed above, thereby representing a snapshot of the real dynamics of mercury in P. Bay. In terms of the overall reliability of the deterministic mercury cycling model, efforts were made to resolve the uncertainty associated with the most important model parameters throughout the course of this study. Table 5-1. Mass balance equations describing parameters and rate constants included in mercury cycling model. Mass Balance Equations for Total Mercury (Hg-T) /1, = Total mass of mercury in water (g) A, = Total mass of mercury in sediments (g) nput = combined external inputs of mercury or specified species) (g d-') iverage daily inputs of mercury from rivers g d-I) iverage daily inputs of mercury from direct itmospheric deposition (g d-I) iverage daily inputs of mercury from tidal nflow (g d-') Volatilization of elemental mercury (~~4 i-om the water column (d-I) htflow of Hg-T from the water column (d-') Settling of mercury associated with suspended ;olids in the water column (d-') Xesuspension of mercury associated with ~enthicsediments (d") 3urial of mercury associated with benthic jediments (d-') Mass Balance Equations for Methyl Mercury (CH3Hg) M,.,cH,H~= Total mass of methylmercury in water (g) M, CHIH~= Total mass of methylmercury in sediments (g) Combined external inputs of methylmercury (g d-'1 Outflow of methylmercury from the water column (d-I) Settling of mercury associated with suspendec solids in the water column (d-') Resuspension of methylmercury associated with benthic sediments (6') Burial of mercury associated with benthic sediments (6') 20xidativehehethylation of CH3Hg in water to form H~O(d-') Net methylation (methylation-demethylation) of Hg(I1) in benthic sediments to form CH3Hg ' Estimated from mass budgets for all mercury species assuming steady state in water column * A rate constant for reduction of HgOto form H~Ois included in the individual species sub-models for H~Oand Hg(I1) described as the reaction term in the text. Preliminary calculations and sensitivity analysis at early stages in model development revealed that the distribution and fluxes of mercury in the sediment compartment were ' the most significant factor limiting confidence in the mass budget calculations for mercury in P. Bay. Accordingly, empirical measurements and experimental work conducted as part of this study were focused on the dynamics of mercury in the sediment compartment and care was taken to incorporate spatial variability in mercury concentration data into the mass budget calculations (see section 5.6.5). The effect of variability in individual parameter estimates on confidence in the deterministic solutions presented in the mass budgets for each mercury species is discussed in further detail below. 5.6 Model parameter Estimates 5.6.1 External Inputs of Mercury External sources of mercury enter P. Bay through freshwater inflow, tides and direct atmospheric deposition. Each source is quantified in the empirical model by combining measured mercury concentrations with hydrological records. 5.6.1.1 Atmospheric Deposition The majority of mercury deposited from the atmosphere is divalent inorganic mercury (Hg(II)) (Pai et al., 1997). Although the amount of MMHg in precipitation has not been measured in the P. Bay region, this fraction is estimated as -1% of total deposition based on Hultberg et al. (1994) and Fitzgerald et al. (1991). Table 5-2. Summary of ecosystem parameters used in empirical model of mercury cycling in Passamaquoddy Bay. Ecosystem Parameters for Passamaquoddy Bay Volume of water (m3) Water surface area (m2) "Netr' seawater inflow (m3day) Daily inputs through precipitation (m3 day) Daily freshwater discharge (m3 day) Water outflow (m3day) Depth of active sediment layer (m) Surface area sediment (m2) Volume of sediment in the active layer (m3) Concentration of solids in sediments (kg L-') Density of benthic sediments (kg L-') Data obtained from Gregory et al. (1993) representing the surface area and water volume in Passamaquoddy Bay below the zero contour line (chart datum). Freshwater discharge rates data were estimated from river gauges where reported standard deviations in monthly discharge rates were large (28-87%). The empirical model does not capture mercury dynamics associated with these types of seasonal fluctuations. Calculated from flushing time of 16 days reported in Ketchum and Keen (1953) and assuming outflow through tides (To) is equal to the sum of inputs through tidal inflow (Ti), rivers (R) and precipitation (P). Natural variability in inputs through precipitation is large (0.0- 18.9 mm d-'). The value reported was estimated by multiplying the daily precipitation rate (obtained by averaging total precipitation rates measured between 1996 and 2000 and converting the annual total into a daily rate) by the water surface area in P. Bay. Precipitation data were from Beauchamp (1998). Data obtained from field measurements (Chapter 4) and calculations reported in this study (see section 5.5.5). The fraction of elemental mercury (H$) relative to other mercury species is very small in both wet and dry deposition (Fitzgerald et al., 1991, Lamborg et al., 1999), thus net deposition of H~Ois assumed to be insignificant. The flux of mercury from atmospheric deposition is calculated as the average of Hg-T deposition measured in precipitation between 1996 and 2000 (7.2 k1.4 pg m-2 yr -1), the water surface area of P. Bay (1.32 x lo8m2) and by assuming dry deposition is approximately 50% the magnitude of wet deposition (Chapter 2). Hg-T deposition data were collected daily at the mercury deposition network (MDN) monitoring site in St. Andrews, NB (45' 05' 12" N, 67' 04', 30" W) between 1996 and 2000 (Table 7-20, Appendix). The assumption that dry deposition is approximately 50% of the magnitude of wet deposition is based on other studies (e.g., Mason et al., 1994, St-Louis et al., 1996), as well as RELMAP atmospheric deposition modeling in the Northeastern United States (NESCAUM et al., 1998). Although there is some uncertainty in this number as discussed in Chapter 2, it likely does not account for more than -110% of the dry deposition estimate as corroborated by other studies in central Nova Scotia (Lamborg et al., 2002). There were no significant trends in measured deposition between 1996 and 2000 (Beauchamp, 1998). Based on these data, total direct Hg-T and MMHg inputs through atmospheric inputs were estimated to be 1430 g yr-' and 150 g yr" respectively. 5.6.1.2 Riverborne Invuts ofMercury The St. Croix River, Digdegaush River and Magaguadavic River account for the majority of freshwater discharges into Passamaquoddy Bay (Tables 7-2 1, Appendix). Unfiltered mercury concentrations were measured seasonally (May, August, November) in these three rivers between November 1994 and May 2002 (Tables 7-22 and 7-23, Appendix). Among the three rivers sampled, Hg-T concentrations ranged between 1.15 ng L-' and 9.26 ng L-' depending on the year, season sampled and discharge rate, while MMHg concentrations varied between 137 pg L-' and 3 17 pg L-'. To account for some of this variability, for each season an average concentration in freshwater inflow was estimated by multiplying the percent of total discharge of each river by the measured concentrations of mercury. Annually averaged concentrations (Hg-T: 4.17 f 0.41 ng L-'; MMHg: 241 f 42 pg L-') were calculated by weighting average seasonal concentrations of mercury in discharges by the fraction of total annual discharges occurring during each season (Gregory et al., 1993). The total mass of mercury entering P. Bay from rivers was estimated from the product of the annually averaged concentration and total discharge (4.68 x 109 m3 yr-'). Concentrations of suspended particulate matter measured in freshwater discharges ranged between 1.19 and 6.19 mg L-' (Dalziel et al. 1998). A weighted annual average concentration of suspended particulates (3.56 f 1.03 mg L") was developed from seasonal data on sediment discharges collected over three years (Dalzeil et al., 1998) using the same methodology outlined above for mercury concentration data. The flux of dissolved and solid phase mercury in riverborne mercury was estimated using partition coefficients of 4.9 and 4.6 (log K, L kg-') for Hg-T and MMHg empirically determined in other riverine systems (Babiarz et al., 1998). There is some uncertainty in these values as they were not measured directly in P. Bay and are known to vary as a function of total organic carbon content (Chapter 3). Total sediment fluxes through riverborne inputs were estimated to be approximately 1.67 x lo7 kg w'. Therefore, dissolved and solid phase inputs from freshwater inflow were estimated at 15.2 kg yil and 4.3 kg yf' Hg-T and 980 g yr'land 140 g yr-' for MMHg. 5.6.1.3 Tidal Inflow Concentrations of Hg-T and MMHg were measured at multiple locations in the outer Bay of Fundy between August 2000 and June 2001, where the coefficient of variation in mercury concentrations spatially and stratigraphically was approximately 40% for both Hg-T and MMHg (Dalziel, unpublished data, 2002). The intertidal volume of P. Bay (8.40 x lo8m3), calculated as the product of mean tide (6 m) and the average of the surface area of P. Bay at low water and at high water (Gregory et al., 1993), comprises approximately 30% of the average water volume. However, the intertidal volume is not a good estimate of the flushing of contaminants from the system because currents along the coast do not always displace tidal waters far enough to prevent it from cycling back into the Bay (Page, 2001). Nonetheless, over several tidal cycles a net translocation of water results. It can be generally assumed that the rates of water entering and leaving P. Bay are equal, so the volume of water in the system does not change over the time unit of interest. Ketchum and Keen (1953) estimated the flushing time of P. Bay to be 16 days based on salinity measurements and freshwater inputs. The net volume of seaward flow (To) is estimated from this flushing time, which represents the length of time required for the estuary to exchange its volume (e.g., Vw/To= flushing time; where, Vwrepresents the average water volume of the estuary). Assuming Vwis constant, the net exchange of seawater with P. Bay or tidal inflow (Ti) is estimated fiom To (e.g., To= Ti + R + P ) assuming outflow is equal to the sum of tidal inflow (Ti ), freshwater inflow (R) and inputs through precipitation (P). The concentration of suspended solids in tidal water was obtained fiom suspended particulate matter (SPM) concentrations measured at the mouth of the Bay of Fundy and the outer Quoddy region (Showell and Gaskin, 1992). Solid and dissolved phase inputs were estimated using partition coefficients (log K, L kg-') of 4.5 and 3.8 for Hg-T and MMHg respectively (Bloom et al., 1999, Benoit et al., 1998, Tseng et al., 2001). Again, there is some uncertainty in these values, as they were not measured directly in this study. Total inputs of suspended solids through tidal fluxes are estimated to be in the range of 1.73 x lo8kg yi' . Accordingly, dissolved and solid phase inputs are approximately 13.8 and 1.3 kg yr-' for Hg-T and 3003 and 60 g yr" for MMHg, respectively. Again, these estimates are measures of "most likely" annually averaged conditions and do not capture inter-annual and seasonal variability in concentrations and fluxes. In summary, inputs of Hg-T and MMHg to P. Bay through tidal and riverborne inflow are much greater than direct atmospheric deposition, although the majority of mercury in tidal waters and river catchments is likely also derived from atmospheric deposition (Chapter 4). MMHg contributions from tidal inflow are greater than the riverborne sources as a result of the relatively greater volume of tidal water entering P. Bay. 5.6.2 Water Column Dynamics The total mass of mercury in the water column is estimated fiom steady state concentrations of Hg-T and MMHg in water (Table 5-3) and the mean volume (V,) of the estuary (2.8 1 x 109). The magnitude of solid and dissolved mercury was estimated using partition coefficients of 3.8 for MMHg and 4.9 for Hg-T (Bloom et al., 1999, Benoit et al., 1998, Tseng et al., 2001). These coefficients are higher than those measured in the benthic sediments to account for the higher organic fraction (16-20%) of SPM in the water column (Showell and Gaskin, 1992) relative to the benthic sediments (1-6%), which significantly influences partitioning between dissolved and solid phases (Chapter 3). 5.6.2.1 Tidal Outflow Mercury contained in the water column is also removed from Passamaquoddy Bay through outflow of tidal waters. The net loss of mercury through outgoing tidal waters is estimated from To, calculated above, and measured concentrations of Hg-T and MMHg. A rate constant for tidal outflow for both Hg-T and MMHg is calculated as the volume of water leaving the Bay (To) divided by the total mean volume of P. Bay (Gregory et al., 1993). The flux of mercury is calculated as the product of the outflow volume and mercury concentrations in the water column of P. Bay. 5.6.2.2 Settling of Suspended Solids Settling of suspended solids in the water column results in a net loss of mercury fiom the water column and a net input to the sediment layer. Settling fluxes were estimated fiom suspended sediment concentrations in P. Bay (Showell and Gaskin, 1992) and an estimated particle settling velocity of 1.25 m day" (Gobas et al., 1995). 5.6.3 Volatilization There were no direct measurements of volatilization of elemental mercury in P. Bay. Fluxes measured in other studies of coastal and marine systems indicate that volatilization results in significant losses of mercury fiom the water column (Baeyens et al., 1991, Mason et al., 1993, Quemerais et al., 1999). 153 Table 5-3. Concentrations and parameters used to estimate mercury speciation in water and sediment compartments. Sediment Speciation: Partition coefficient benthic sediments (log Kd L kg-') Fraction of mercury dissolved in pore water Mean concentration in porewater (ng L-I) Mean concentration in solids (ng g-' dry) Water Speciation: Water column concentration of mercury at steady state (pg L-I) Partition coefficient seawater (log K, L kg) Fraction of dissolved phase mercury in water column Calculated by assuming the fraction of mercury in porewater (f,,) is equal to: 1 ,where C,, is the concentration of solids in sediments (Table 5-2). l+(C, *K,) Spatially weighted concentrations of mercury (described in section 5.6.5) were used to incorporate spatial variability in mercury concentrations into mass budget calculations. Concentration of Hg-T in water column at steady-state. Sample sizes of data used to estimate values were limited by the occurrence of a hurricance during field collection, thus there is some uncertainty in the reported values. However, empirical values are within the expected range of values based on other studies in the region (Dalziel et al., 1998, Dalziel, unpublished data, 2002). Estimated values based on measurements reported in Bloom et al. (1999), Benoit et al., (1998), and Tseng et al. (2001). Calculated values using the same relationship reported above for sediments, partition coefficients for the water column and a suspended sediment (C,,) concentration of 1.97 mg L-' reported by Showell and Gaskin (1992). For example, flux rates of elemental mercury in Halifax Harbour measured under calm conditions were slightly above 16 ng m'2 day-' (Beauchamp et al., 2000). In the Scheldt Estuary, UK, volatilization rates reached a maximum of 5208 ng m-2day-' during stormy conditions (Baeyens et al., 1991). An initial flux rate of 50 ng m-2day" was specified in the model for P. Bay based on the range of empirical values presented in the literature. This term was treated as uncertain in mass budget calculations and, when subject to the constraints necessary to achieve steady state in the water column for all mercury species, an estimate of 87 ng m-2 day-' was obtained. The corresponding mean concentration of HgOin the water column is 44 pg L-', calculated using the two-layer, gas transfer model by Liss and Slater (1974). In this model, the flux of elemental mercury [F (ng m-2day-')] is related to the concentration of HgOdissolved in the water column (C,) at equilibrium by the equation: Where Henry's Law coefficient (H) for elemental mercury is 0.29 at 20•‹C(Sanemasa, 1975) and k, and kwrepresent overall air-side and water-side mass transfer coefficients of 216 m day-' and 2.16 m day-' respectively (Schroeder et al., 1992). In this model, negative values denote an "upward" flux and Carepresents the concentration of H~Oin air above the water column, estimated fiom measured total gaseous mercury (TGM) concentrations in St. Andrews (Beauchamp, 1998) and by assuming approximately 50% of TGM is present as HgO(Baeyens et al., 1991, Fitzgerald, 1986). 5.6.4 Sediment Budnet Typically, large spatial and stratigraphic heterogeneity in mercury concentrations measured in sediments and interstitial waters make it diffkult to constrain mass budget and flux calculations, as discussed in several other studies (Gill et al., 1999; Mason et al., 1998). In this study, preliminary sensitivity analysis highlighted the importance of resolving the uncertainty associated with sediment mass budget calculations, stemming from the fact that the majority of the mercury is contained in this compartment in P. Bay. To account for some of these challenges, extensive field investigations and spatially explicit calculations were used to refine calculations of the reservoir and fluxes of different mercury species in the sediments. This is a considerable improvement over "box" models that use the mean value of a sample population because such models are unable to represent areas with non-detectable mercury concentrations and do not take into account spatial variability in mercury concentrations. Burial rates measured at multiple stations (n=27) were used to construct a spatial grid of sediment accumulation rates in P. Bay using Kriging as a spatial interpolation method (Fig 5- 1). This geostatistical gridding method employs several algorithms to spatially weight measured fluxes using the Variogram, Drift Type and Nugget Effect models to determine the best fit to the empirical data (Cressie, 1991). Volume calculations were used to estimate the total mass of sediment buried on an annual basis by summing volume estimates for each cell defined in the spatial grid using the Trapezoidal Rule, Simpson's Rule and Simpson's 318 Rule, where the difference between methods was used to check the accuracy of volume calculations (Press, 1986). The same methodology was used to spatially grid mercury concentrations at multiple stations in P. Bay that were sampled for Hg-T (n=56) and MMHg in May (n=24) and August (n=27). The resulting spatially interpolated grids of mercury concentrations were used to construct contour maps of concentration gradients throughout P. Bay by drawing contours as a series of straight-line segments between adjacent grid lines representing concentrations (Figs 5-2 to 5-4). The spatial grid of burial fluxes described above was multiplied by concentration grids for mercury to estimate the annual flux of Hg-T and MMHg in P. Bay through sediment burial (Figs 5-5 and 5-6). The actual magnitude of burial fluxes (g yr-l) and the overall mass (g) of Hg-T and MMHg contained in the sediment compartment of P. Bay were quantified using the same spatially weighted volume calculations applied to estimate total sediment burial and by assuming an active sediment layer depth of 15 cm based on field measurements (Chapter 4). 5.6.5 Resuspension Resuspension of sediment solids was determined as the difference between the burial flux and settling flux. Resuspension rate constants for Hg-T and MMHg were calculated by dividing the total flux of sediments through resuspension by the product of the fraction of Hg-T and MMHg in the solid phase sediments and the total volume of the sediment compartment. The fraction of mercury in the benthic sediments associated with solids was estimated fiom partition coefficients for total mercury and methylmercury measured in P. Bay sediments of 3.6 + 0.2 L kgm1and 2.6 + 0.3 L kg-' (log Kd),respectively (Chapter 3). Figure 5-2. Contour map of total mercury concentrations (ng f1 dry) in Passamaquoddy Bay. Concentration gradients are only valid for regions within Passamaquoddy Bay and the St. Croix River as field sampling did not cover the regions beyond Deer Island in the outer Bay of Fundy shown in the figure. Figure 5-3. Methylmercury concentration distribution in August (pg 6' dry). Concentration gradients are only valid for regicns within Passamaquoddy Bay and the St. Croix River as field sampling did not cover the regions beyond Deer Island in the outer Bay of Fundy shown in the figure. Figure 5-4. Contour map of methylmercury concentration distribution in May (pg g-' dry). Concentration gradients are only valid for regions within Passamaquoddy Bay and the St. Croix River as field sampling did not cover the regions beyond Deer Island in the outer Bay of Fundy shown in the figure. Figure 5-5. Contour map of sediment depositional flu (ng ~rn-~yr'')of total mercury in Passamaquoddy Bay. Concentration gradients are only valid for regions within Passamaquoddy Bay and the St. Croix River as field sampling did not cover the regions beyond Deer Island in the outer Bay of Fundy shown in the figure. Figure 5-6. Contour map of sediment depositional flux (ng ~rn-~yr") of methylmercury in Passamaquoddy Bay. Concentration gradients are only valid for regions within Passamaquoddy Bay and the St. Croix River as field sampling did not cover the regions beyond Deer Island in the outer Bay of Fundy shown in the figure. 5.6.6 Net Methylation of Mercury The conditions most conducive to production of MMHg by anaerobic microbes in P. Bay are found in the sediment compartment (Chapter 4). It was assumed that the net microbial production of MMHg in the water column of P. Bay is not significant relative to fluxes fiom the sediments and via rivers and tides. "Potential" methylation rates were measured in duplicate push cores fiom P. Bay by pre-equilibrating the stable mercury isotope, '99~g(~~),with seawater, injecting the solution at 1 cm vertical intervals and incubating the sediment core at ambient seawater temperature for four hours. Core incubations were used to provide empirical values for the potential of these sediments to convert inorganic mercury (Hg(I1)) to MMHg (CH3Hg), but may not identically reflect the absolute rates of methylation due to differences in the bioavailable fraction of injected inorganic mercury '99~g(~~)and ambient Hg(I1). However, keeping in mind the uncertainty associated with this experimental technique, the relative magnitude of methylation taking place in the sediment column can be obtained using this method. The "potential" methylation rate in P. Bay sediments was estimated fiom the conversion of 199~gisotope to 199~~3~g+during the four-hour incubation experiments. Specifically, methylation is represented as a first order rate function: Where A is the amount of '99~gisotope remaining after a four hour incubation (t) and Ao is the initial amount of 199~gisotope injected into the sediments calculated fiom the sum of 199~gand 199~~3~g+mea~uredin the sediment core. Integrating the values measured in the surface sediments provided a value for the potential methylation rate (h)of 0.02 ng g-'hr'l or 2.71 x lo5 day-'. The net methylation rate, is the difference between microbially mediated conversion of inorganic mercury Hg(I1) to MMHg (methylation) and MMHg coversion to Hg(I1) (demethylation). Because no adequate measurements of demethylation rates were available for P. Bay, the "net" methylation rate was estimated by assuming steady state with respect to Hg(I1) in the sediment column and deriving the reaction rate of Hg(I1) to MMHg from the associated mass balance. Using this method, the realized net reaction of Hg(I1) to MMHg in the sediments is very small (approximately 150 g yr-') relative to the daily turnover of Hg(I1) and MMHg through methylation and demethylation suggested by the core incubation experiments. The significance of these findings will be revisited in the discussion of mass budget calculations below. 5.7 Mass Budgets Mass budgets for all mercury species are shown in Figures 5-7 through 5-10. Inorganic Hg(I1) fluxes are calculated as the difference between concentrations of Hg-T and the sum of MMHg and H~O(Fig 5-9). Because H~Owas not measured directly in this study (see section 5.6.4), the mass budget presented (Fig 5-10) is a hypothetical description of fluxes that assumes that the losses of H~Othrough volatilization and tidal outflow calculated from the model presented should be roughly equal to the sum of inputs through tidal and riverine inflow and Hg(II)/MMHg reduction in the water column. Figure 5-7. Mass budget for total mercury (Hg-T) in Passamaquoddy Bay. Figure 5-8. Steady state mass budget for methylmercury (CH3Hg) in Passamaquoddy Bay. Figure 5-9. Steady state mass balance for divalent mercury (Hg(I1)) in Passamaquoddy Bay. Figure 5-10. Hypothesized steady state mass balance for elemental mercury (H~@)in Passamaquoddy Bay. /i.7.1 Total Merc Mass budget calculations indicate that greater than 99% of the reservoir of Hg-T in P. Bay is contained in the active sediment compartment. Based on the empirical data, Hg-T fluxes (Fig 5-7) in the sediment compartment appear to be close to steady state (inputs = losses). In contrast, rate constants for Hg-T cycling developed in this study (Table 5-1) indicate that Hg-T levels in the water column will continue to rise over a period of approximately two months from the empirically determined concentration of 267 pg L-' until a steady state concentration of 464 pg L-' is achieved. Given the range of uncertainty present in the measured water concentration data due to the limited sample size and spatial variability in aqueous mercury concentrations on the order of 40% in inflowing tidal waters (discussed in section 5.6.1.3), it is likely that the water column in P. Bay is also at or very close to the estimated steady state values for Hg-T. Deterministic values for total mercury fluxes at steady state are shown in Figure 5-7. 5.7.2 Methvlmercurv (MMHg) The mass budget for MMHg (Fig 5-8) developed from the empirical data shows that there is a large turnover of MMHg in both the sediments and water compartments in P. Bay. Similar to the Hg-T budget, the majority of MMHg (>go%) is contained in the sediment compartment and exchange of MMHg occurs rapidly in the water column, reflecting the magnitude of present inputs from rivers, tides and direct atmospheric deposition. In contrast, concentrations in the sediments respond more slowly to changes in external inputs, due to the relatively small fluxes of MMHg through settling, resuspension and burial compared to the overall mass of MMHg contained in this compartment. The empirical model indicates that although the "net" accumulation of MMHg in the sediment compartment though reaction of Hg(I1) is very small, the daily turnover of MMHg through methylation and demethylation is almost 40% of the total reservoir. This estimate is based on the estimated methylation and demethylation rates discussed in section 5.6.1. Although the absolute value of the potential methylation rate measured using stable mercury isotopes is somewhat uncertain, comparison of the experimentally measured methylation rates with the estimated net accumulation of MMHg in P. Bay sediments provides strong evidence for a large daily turnover of Hg(I1) and MMHg in P. Bay sediments. For example, deterministic solutions calculated fiom the isotope data for k,[Hg(II)] and k,JCH3Hg] were approximately 1230 g day-'. This large daily turnover of MMHg in the sediment compartment relative to the small "net" rate of MMHg production (shown in Figure 5-8) indicates that environmental perturbations such as organic enrichment, declining redox status and increased sulfide concentrations that shift the present equilibrium between methylation and demethylation in the sediments (Chapter 3) could potentially result in rapid accumulation of MMHg in the sediment compartment. Mass budget calculations for MMHg show that net accumulation of MMHg in the sediments fiom net methylation of inorganic mercury and settling of suspended solids exceeds daily losses through resuspension and burial by a rate of 167 g day-' or about 5% of the standing pool. Thus, MMHg may currently be accumulating in the sediments over time or there may be another removal mechanism present. A number of other studies have shown that diffusion of total and MMHg from the sediments into the water column is negligible relative to other transport mechanisms (Gobeil and Cossa, 1993, Gobeil et al., 1999, Benoit et al., 1998). However, Gill et a1 (1999) found that diffision estimates can vary by 2-3 orders of magnitude depending on the time of day measured. Thus, it is possible that diffusing MMHg accounts for a small component of the net losses of MMHg from the sediments in P. Bay, although the total mass of MMHg in the porewaters is small (-12 g) making it unlikely that diffusion alone accounts for the full 167 g yr-' needed to reach steady state for MMHg in the sediment compartment. 5.7.3 Demethvlation in Sediments and Water Alternately, degradation of MMHg through oxidative demethylation in the sediment compartment (e.g., Ramlal et al., 1986, Marvin-Dipasaquale et al., 2000) could also account for some of imbalance between MMHg inputs and losses in the sediment compartment of P. Bay (Fig 5-8). Elemental mercury is quickly volatilized and would therefore likely diffuse into the water column where it would eventually be released to the atmosphere at the air-sea interface (Fig 5-10). The rate of degradation of MMHg in the water column is a critical factor determining exposure of pelagic organisms. Based on mass balance calculations, demethylation of MMHg to form Hg(1I) through photodegradation or microbially mediated reaction in the water column likely accounts for up to approximately 190 g w'. The rate at which MMHg entering P. Bay in rivers and tides is converted to Hg(II) in the water column is also important for determining the degree to which P. Bay acts as a net source of MMHg to the larger Bay of Fundy and Gulf of Maine region, currently estimated at approximately 1 kg MMHg yr-'. The significance of oxidative demethylation in P. Bay sediments and demethylation of MMHg in the water column on the overall dynamics of mercury in this system warrants further investigation. 5.7.4 Elemental Mercury Budpet Reduction of Hg(I1) to form HgOin the water column of coastal an marine systems is known to be important for overall mass budgets, accounting for up to 35% of the total mercury deposited in other coastal systems (Rolfhus and Fitzgerald, 2001). Reduction of Hg(I1) and formation of dissolved gaseous mercury (DGM) in the water column can occur both abiotically through photoreduction and as the result of microbially mediated reactions (Amyot et al., 1994, Krabbenhoft et al., 1998). Because no direct measurements of HgOwere included in the field component of this study, estimates of HgOfluxes in P. Bay are highly uncertain. Loss of HgOat the air-sea interface through volatilization was modeled in section 5.6.4, allowing the truly dissolved concentration of H~Oto be estimated. Preliminary calculations of inputs and outputs of HgOin the water column were based on this model and the constraints associated with reaction of other mercury species (Fig 5-10). The resulting mass budget indicates that reduction of Hg(I1) to form HgO in the water column is likely an important process and that the rate of reaction needed to achieve steady state with respect to HgOconcentrations is much higher than anticipated from the mass balance for Hg(II) (Figure 5-9). This difference may be accounted for by volatilization of Hg(I1) in outflowing tidal waters in P. Bay, making up an addition flux of up to 1050 g of Hg(I1). 5.8 Concentrations in Benthic Organisms Benthic and intertidal organisms are a useful indicator of the present status of contamination in P. Bay. Concentrations of Hg-T in water, sediments, polychaetes and amphipods measured in this study and biological data from other sources are compiled in Table 5-4. The relative enrichment of mercury in the organisms relative to concentrations in the sediments and waters is shown in Table 5-5. Comparing measured concentrations of Hg-T in polychaetes and amphipods collected in the early 1980's by Braune et al. (1987) to those collected for this study in 2001 supports the premise that concentrations in sediment-dwelling organisms in this system are not declining in response to emissions reductions. This result is consistent with the mass fluxes of mercury represented in Figures 5-7 and 5-8, which suggest that the large reservoir of mercury in the sediment compartment will result in a slow temporal response of concentrations in both sediments and organisms to changes in external inputs. Hg-T concentrations in all species were 4-5 fold the concentrations measured in seawater and ranged from slightly less than the mean concentration of Hg-T in the wet sediments (polychaetes) to an estimated biota-sediment accumulation factor of 10 and 381 for Hg-T and MMHg respectively in mussels. Hg-T concentrations in blue mussels (Mytilus edulis) from P. Bay measured by the Gulfwatch environmental monitoring program (Chase et al., 2001) are the highest of the benthic species shown (156 ng g" wet wt.). This is consistent with other studies that showed that mussels accumulate the majority of their mercury burden from sediment bound/solid phase mercury in their diet (Gagnon and Fisher, 1997). Table 5-4. Total mercury (Hg-T) concentrations measured in sediments, water and benthic organisms in Passamaquoddy Bay. Std. Date Collected Location n1 ng g" wet Dev. Source Passamaquoddy Water Nov. 2001 Bay 4 3~10-~- This study Passamaquoddy Sediments 2000/200 1 Bay 94 - This study Amphipod St. Andrews Tidal (Gammarus sp.) Aug. 2001 Flats 5 3.1 This study Amphipod St. Andrews Tidal (Gammarus sp.) 1983-1984 Flats 1 (Braune, 1987) Polychaete Passamaquoddy (Nephtys sp.) Aug. 2001 Bay 13 2.9 This study Poly chaete Deer Island (Neiris sp. ) 1984 water's edge 2 6.7 (Braune, 1987) Common Periwinkle Magaguadavic (Littorina littorea) 1998 River 1 Health canada2 Blue Mussels (Chase et al., (Mytilus edulis) 1991-1997 St. Croix Estuary 10 62 2001) Blue Mussels Magaguadavic (Mytilus edulis) 1998 River 1 Health canada2 Sofi Shelled Clam Magaguadavic (Mya areanaria) 1998 River 1 Health canada2 Lobster (Homarus americanus) 1990 Back Bay 5 2 1.7 Health canada2 'n refers to number of samples analyzed not number of organisms Unpublished data Health Canada Data (Burns, unpublished data, 1998) Table 5-5. Accumulation factors for biota between water (Cw) and sediment concentrations (C,). ' Polychaete 3.6 x lo4 0.7 2.5 x lo4 16 Amphipod 8.4 x lo4 1.6 n/a n/a Periwinkle 1.8 x 10' 3.3 6 x lo5 381 Mussel 5.5 lo5 10 n/a n/a Clam 1.7 x lo5 3.3 n/a n/a Lobster 3.6 x lo5 6.7 n/a nla Methylmercury (MMHg) concentrations estimated assuming the fraction of MMHg is -1 7% of total mercury (Hg-T) based on Wang et al. (1 998). Methylmercury concentrations estimated assuming the fraction of MMHg is -27% of Hg-T based on Mikac et al. (1985). In contrast, the mercury concentrations in polychaete species that are thought to achieve equilibrium with the dissolved MMHg concentrations in sediment porewaters (Wang et al., 1998) are much lower (-10 ng Hg-T g-'; 1.3 ng MMHg g-'). Assuming the proportion of MMHg in blue mussels is approximately 27% (e.g., Mikac et al., 1985), the mean concentration of MMHg in blue mussels is expected to be approximately 42 ng g-', which slightly exceeds the Environment Canada guideline for MMHg intended to protect all life stages of 0.033 ppm (CCME, 1999). 5.9 Application of the Empirical Model The purpose of the mercury cycling model for P. Bay presented in this study was to capture a "snapshot" of mercury dynamics in this system based on the field data compiled in this study. This is the first step toward development of a mechanistic mercury cycling model for this system that will have more applicability as a tool for forecasting future states of nature. As an exercise, recognizing the limits of this model as a predictive tool imposed by the scope and uncertainty in the empirical data collected and described above, the model for P. Bay can be used to investigate the temporal response of mercury concentrations in this system to reductions in anthropogenic mercury emissions. As discussed above, though concentrations in the water column respond rapidly to changes in loading, the water column contains only a small fraction of the total mass of mercury in P. Bay. Therefore, the temporal response of this system to changes in inputs is governed by dynamics in the sediment compartment. The overall anthropogenic mercury enrichment factor measured in P. Bay sediments was slightly greater than a factor of two, suggesting that the majority of inputs are atmospherically derived (Chapter 4). As discussed in Chapter 1, approximately 65% of the atmospheric inputs of mercury in the Bay of Fundy are accounted for by fluxes of mercury from natural and recycled historic sources. Virtual elimination of present-day anthropogenic emissions would therefore result in immediate reductions in Hg-T loadings to P. Bay of only 35%. Given this level of emissions reductions, application of the empirical model shows that it will take several hundred years for Hg-T concentrations in the sediments to reach steady state with respect to inputs levels. These results suggest that in certain ecosystems with similar physical and hydrodynamic attributes to P. Bay, there will be a large lag time between reductions in anthropogenic emissions and declines in mercury concentrations. This large lag time does not discount the importance of emissions reduction initiatives, which are critical for the long-term management of mercury contamination and for minimizing recycled historic inputs of mercury. The shorter-term risks of enhanced MMHg production fiom changes to the geochemical characteristics of aquatic ecosystems must also be addressed in all mercury management strategies. As discussed above and in Chapter 3, anthropogenic changes such as organic enrichment that lead to declines in sediment redox status and increased sulfide concentrations could potentially shift the present equilibrium between methylation and demethylation resulting in relatively rapid increases in the concentrations of MMHg in the sediments thereby increasing the exposure of organisms. Future management strategies aimed at securing environmental quality should therefore take into account the indirect effects of anthropogenic perturbations like eutrophication that may stimulate MMHg production and facilitate more rapid accumulation of mercury in organisms. 5.10 Literature Cited Amyot, M., Mierle, G., Lean, D.R.S., and McQueen, D. J. 1994. Sunlight-induced formation of dissolved aqueous mercury in lake waters. Environmental Science and Technology, 28,2366-2371. Babiarz, C. L., Hurley, J. P., Shafer, M. M., Andren, A. W. and Webb, D. A. 1998. Seasonal influences on partitioning and transport of total and methylmercury in rivers from contrasting watersheds. Biogeochemistry, 41,237-257. Baeyens, W., Leermakers, M., Dedeunvaerder, H. and Lansens, P. 1991. Modelization of the mercury fluxes at the air-sea interface. Water, Air and Soil Pollution, 56, 73 1- 744. Beauchamp, S., Tordon, R., Phinney, L., Pinette, A., Rencz, A., Dalziel, J. and Wong, H. T. K. 2000. Air-surface exchange of mercury over natural and impacted surfaces in Atlantic Canada. In: 25th International Conference on Heavy Metals in the Environment. August 6-10,2000, Ann Arbor, Michigan, USA. Beauchamp, S. 1998. Mercury in the atmosphere. In Mercury in Atlantic Canada: A Progress Report (Ed, Burgess, N.) Environment Canada, Atlantic Region, Bedford, NS, pp. 16-45. Benoit, J. M., Gilmour, C. C., Mason, R. P., Riedel, G. S. and Reidel, G. F. 1998. Behavior of mercury in the Patuxent River estuary. Biogeochemistry, 40,249- 265. Bloom, N. S. 1989. Determination of picogram levels of methylmercury by aqueous phase ethylation, followed by cryogenic gas chromatography with cold vapor atomic fluorescence detection. Canadian Journal of Fisheries and Aquatic Sciences, 46, 113 1-1 140. Bloom, N. S., Gill, G. A., Cappellino, S., Dobbs, C., Mcshea, L., Driscoll, C., Mason, R. and Rudd, J. 1999. Speciation and cycling of mercury in Lavaca Bay, Texas, sediments. Environmental Science and Technology, 33,7- 13. Braune, B. M. 1987. Mercury accumulation in relation to size and age of Atlantic herring Clupea clupea harengus from the southwestern Bay of Fundy, Canada. Archives of Environmental Contamination and Toxicology, 16,3 11-320. Burns, G. 1998. Canadian Food Inspection Agency, Personal Communications, Unpublished Health Canada Data. CCME. 1999. Canada-Wide Standards for Mercury. Canadian Council of Ministers of the Environment, Ottawa, pp. 13. Chase, M. E., Jones, S. H., Hennigar, P., Sowles, J., Harding, G. C. H., Freeman, K., Wells, P. G., Krahforst, C., Coombs, K., Crawford, R., Pederson, J. and Taylor, D. 200 1. Gulfwatch: Monitoring spatial and temporal patterns of trace metal and organic contaminants in the Gulf of Maine (199 1- 1997) with the blue mussel, Mytilus edulis L. Marine Pollution Bulletin, 42,49 1-505. Cranston, R. E. 1999. Estimating marine sediment accumulation rates fiom geochemical pore water gradients. GeoResearch Forum, 5,57-66. Cressie, N. A. C. 1991. Statistics for Spatial Data, John Wiley and Sons, New York, pp. 900. Dalziel, J.A. 2002. Canadian Department of Fisheries and Oceans, Unpublished Data. Dalzeil, J. A., Yeats, P. A. and Amirault, B. P. 1998. Inorganic chemical analysis of major rivers flowing into the Bay of Fundy, Scotian Schelf, and Bras dtOr Lakes. Canadian Technical Report of Fisheries and Aquatic Sciences, 2226,140. Diamond, M., Ganapathy, M., Peterson, S. and Mach, C. 1999. Mercury dynamics in the Lahantan Reservoir, Nevada: Application of QWASI Fugacity/Aquivalence multispecies model. Water, Air, and Soil Pollution, 1 17, 133-156. Elliott, J. E., Scheuhammer, A. M., Leighton, F. A. and Pierce, P. A. 1992. Heavy metals and metallothionein concentration in Atlantic Canadian seabirds. Archives of Environmental Contamination and Toxicology, 22,63-73. Fink, L. K. J., Pope, D. M., Harris, A. B. and Schick, L. L. 1976. Heavy Metal Levels in Suspended Particulates, Biota, and Sediments of the St. Croix Estuary in Maine. Land and Water Resources Institute, University of Maine at Orono, Orono, Maine, pp. 59. Fitzgerald, W. F. 1986. Cycling of mercury between the atmosphere and oceans. In: The Role of Air-Sea Exchange in Geochemical Cycling, NATO Advanced Science Institute Series (Ed, Buat-Manard, P.) Reidel, Netherlands, pp. 363-408. Fitzgerald, W. F., Mason, R. P. and Vandal, G. M. 1991. Atmospheric cycling and air- water exchange of mercury over mid-continental lacustrine regions. Water, Air, and Soil Pollution, 56,745-767. Gagnon, C. and Fisher, N. S. 1997. Bioavailability of sediment-bound methyl and inorganic mercury to a marine bivalve. Environmental Science and Technology, 31,993-998. Gaskin, D. E., Stonefield, K. I. and Suda, P. 1979. Changes in mercury levels in harbour porpoises from the Bay of Fundy, Canada and adjacent waters. Environmental Contamination and Toxicology, 8,733-762. Gaskin, G. E., Frank, R., Holdrinet, M., Ishida, K., Walton, C. J. and Smith, M. 1973. Mercury, DDT, and PCB in harbour seals (Phoca vitulina) from the Bay of Fundy and Gulf of Maine. Journal of Fisheries Research Board of Canada, 30,471 -475. Gobas, F. A. P. C., Pasternak, J. P., Lien, K. and Duncan, R. K. 1998. Development and field validation of a multimedia exposure model for waste load allocation in aquatic ecosystems: application to 2,3,7,8-tetrachloro-p-dioxinand 2,3,7,8- tetrachlorodibenzofuran in the Fraser River watershed. Environmental Science and Technology, 32,2442-2449. Gobas, F. A. P. C., ZtGraggen, M. N. and Zhang, X. 1995. Time response of the Lake Ontario ecosystem to virtual elimination of PCBs. Environmental Science and Technology, 29,2038-2046. Gobeil, C. and Cossa, D. 1993. Mercury in sediments and sediment pore water in the Laurentian trough. Canadian Journal of Fisheries and Aquatic Sciences, 50, 1794- 1800. Gobeil, C., Macdonald, R. W. and Smith, J. N. 1999. Mercury profiles in sediments of the Arctic Ocean basins. Environmental Science and Technology, 33,4194-4198. Gregory, D., Petrie, B., Jordan, F. and Langille, P. 1993. Oceanographic, geographic and hydrological parameters of Scotia-Fundy and southern Gulf of St. Lawrence inlets. Canadian Technical Report of Hydrographic Ocean Sciences No. 143. Department of Fisheries and Oceans, Scotia-Fundy Region, Dartmouth, NS, pp. ' 248. Gulf of Maine Council on the Marine Environment. 1994. Marine Environmental Quality in the Gulf of Maine: Fact Sheet 94-1. Gulf of Maine Council on the Marine Environment, Boston, MA, pp. 16. Harding, G.C.H. 2002. Canadian Department of Fisheries and Oceans, Unpublished Data. Harris, R., Gherini, S. and Hudson, R. 1996. Regional Mercury Cycling Model: A Model for Mercury Cycling in Lakes, R-MCM Version 1.0 Draft User Guide and Technical Reference. Electric Power Research Institute, Wisconsin Department of Natural Resources, Lafayette, California. Henry, E. A., Dodge-Murphy, L. J., Bigham, G. N. and Klein, S. M. 1995. Modeling the transport and fate of mercury in an urban lake (Onondaga Lake, NY). Water, Air, and Soil Pollution, 80,489-498. Hultberg, H., Iverfeldt, A. and Lee Y. H. 1994. Methylmercury inputfoutput and accumulation from forested catchments and critical loads for lakes in southwestern Sweden In: Mercury Pollution: Integration and Synthesis (Eds, Watras, C. and Huckabee, J.) CRC Press, Boca Raton, pp. 313-322. Ketchum, B. H. and Keen, D. J. 1953. The exchanges of fresh and salt waters in the Bay of Fundy and in Passamaquoddy Bay. Journal of Fisheries Research Board of Canada, 10,97-122. Krabbenhoft, D.P., Hurley, J.P., Olson, M.L. and Cleckner, L.B. 1998. Die1 variability of mercury phase and species distributions in the Florida Everglades. Biogeochemistry, 40,3 11-325. Lamborg, C.H., Fitzgerald, W.F., Damman, A.W.H., Benoit, J.M., Balcom, P.H. and Engstrom, D.H. 2002. Modem and historic atmospheric mercury fluxes in both hemispheres: Global and regional mercury cycling implications. Global Biogeochemical Cycles, 16,5 1- 1- 11. Lamborg, C. H., Rolfhus, K. R., Fitzgerald, W. F. and Kim, G. 1999. The atmospheric cycling and air-sea exchange of mercury species in the South and equatorial Atlantic Ocean. Deep-sea Reasearch 11, 46,957-977. Liss, P. and Slater, P. 1974. Flux of gases across the air-sea interface. Nature, 247, 181- 184. Marvin-Dipasaquale, M., Agee, J., McGowan, C., Oremland, R. S., Thomas, M., Krabbenhoft, D. and C.G.Gilmour. 2000. Methyl-mercury degradation pathyways: A comparison among three mercury-impacted ecosystems. Environmental Science and Technology, 34,4908-49 16. Mason, R., Bloom, N., Cappellino, S., Gill, G., Benoit, J. and Dobbs, C. 1998. Investigation of porewater sampling methods for mercury and methylmercury. Environmental Science and Technology, 32,403 1-4040. Mason, R. P. and Fitzgerald, W. F. 1991. Mercury speciation in open ocean waters. Water, Air, and Soil Pollution, 56,779-789. Mason, R. P., Fitzgerald, W. F., Hurley, J., Hanson, A. K. H. Jr., Donaghay, P. L. and Sieburth, J. M. 1993. Mercury biogeochemical cycling in a stratified estuary. Limnology and Oceanography, 38,1227- 1241. Mason, R. P., Fitzgerald, W. F. and Morel, F. M. M. 1994. The biogeochemical cycling of elemental mercury: Anthropogenic influences. Geochimica et Cosmochimica Act& 58, 3191-3198. Meili, M. 1991. The coupling of mercury and organic matter in the biogeochemical cycle, towards a mechanistic model for the boreal forest zone. Water, Air, and Soil Pollution, 56,333-347. Mikac, N., Picer, M., Stegnar, P. and Tusek-Znidaric, M. 1985. Mercury distribution in a polluted marine area, ratio of total mercury, methyl mercury and selenium in sediments, mussels and fish. Water Research, 19, 1387- 1392. New England Governors and Eastern Canadian Provinces (NEG-ECP). 1998. Mercury Action Plan. Conference of New England Governors and Eastern Canadian Premiers. Halifax, Nova Scotia. June 1998. Northeast States for Coordinated Air Use Management (NESCAUM), Northeast Waste Management Officials' Association (NEWMOA), New England Interstate Water Pollution Control Commission (NEIWPCC), Ecological Monitoring and Assessment Network (EMAN), 1998, Northeast States and Eastern Canadian Provinces Mercury Study, A Framework for Action: NESCAUM, pp. 350. Page, F. 2001. An overview of circulation and mixing in the Bay of Fundy and adjacent areas. In: Opportunities and Challenges for Protecting, Restoring and Enhancing Coatal Habitats in the Bay of Fundy. September 19-21,2000. (Eds, Chopin, T. and Wells, P. G.) Environment Canada, Atlantic Region Occasional Report No. 17, Environment Canada, Dartmouth, Nova Scotia, Saint John, New Brunswick, pp. 37-40. Pai, P., Karamchandani, P. and Seigneur, C. 1997. Simulation of the regional atmospheric transport and the fate of mercury using a comprehensive Eulerian model. Atmospheric Environment, 3 l,2717-2732. Press. W.H. 1986. Numerical Recipes: The Art of Scientific Computing. Cambridge University Press, Cambridge, pp. 818. Quemerais, B., Cossa, D., Rondeau, B., Pham, T. T., Gagnon, P. and Fortin, B. 1999. Sources and fluxes of mercury to the St. Lawrence River. Environmental Science and Technology, 33,840-849. Ramlal, P. S., Rudd, J. W. M. and Hecky, R. E. 1986. Methods for measuring specific rates of mercury methylation and degradation and their use in determining factors controlling net rates of mercury methylation. Applied and Environmental Microbiology, 5 1, 110-1 14. Rolfhus, K.R. and Fitzgerald, W.F. 2001. The evasion and spatiaVtempora1 distribution of mercury species in Long Island Sound, CT-NY. Geochimica et Cosmochimica Acta, 65,407-41 8. Sanemasa, I. 1975. The solubility of elemental mercury vapor in water. Bulletin of the Chemical Society of Japan, 48, 1795-1798. Schroeder, W., Lindqvist, O., Munthe, J. and Xiao, Z. 1992. Volatilization of mercury from lake surfaces. The Science of the Total Environment, 125,47-66. Showell, M. A. and Gaskin, D. E. 1992. Partitioning of cadmium and lead within seston of coastal marine waters of the western Bay of Fundy. Archives of Environmental Contamination and Toxicology, 22,322-333. Spry, D. J. and Wiener, J. G. 1991. Metal Bioavailability and Toxicity to Fish in Low- Alkalinity Lakes: A Critical Review. Environmental Pollution, 71,243-304. St-Louis, V. L., Rudd, J. W. M., Kelly, C. A., Beaty, K. G., Flett, R. J. and Roulet, N. T. 1996. Production and loss of methylmercury and loss of total mercury from boreal forest catchments containing different types of wetlands. Environmental Science and Technology, 30,2719-2729. Sunderland, E. M. and Chmura, G. L. 2000. An inventory of historical mercury pollution in Maritime Canada: Implications for present and future contamination. The Science of the Total Environment, 256,39-57. Tseng, C. M., Arnouroux, D., Abril, G. and Donard, 0. F. X. 2001. Speciation of mercury in a fluid mud profile of a highly turbid macrotidal estuary (Gironde, France). Environmental Science and Technology, 35,2627-2633. Wang, W.-X., Stupakoff, I., Gagnon, C. and Fisher, N. S. 1998. Bioavailability of inorganic and methylmercury to a marine deposit-feeding polychaete. Environmental Science and Technology, 32,2564-2571. Weiner, J. G., Fitzgerald, W. F., Watras, C. J. and Rada, R. G. 1990. Partitioning and bioavailability of mercury in an experimentally acidified Wisconsin Lake. Environmental Toxicology and Chemistry, 9,909-918. CHAPTER 6 Conclusions and Implications for Policies 6.1 Major Scientific Findings The main objectives of this study were: (i) to develop a comprehensive "ecosystem- based" understanding of the environmental fate of anthropogenic mercury in a temperate coastal marine environment; and (ii) to formulate and document this knowledge in an empirical model that links anthropogenic mercury inputs to concentrations in sediments, water and benthic organisms. The main purpose of this research was to address the issue of why mercury concentrations in organisms in the Bay of Fundy region of Canada have remained high despite large reductions in anthropogenic mercury releases. In chapter two of the thesis the relationship between mercury emissions and deposition in the Bay of Fundy region was investigated by evaluating sediment archives of historical atmospheric loading and historical emissions data. The results of this analysis show that modern atmospheric mercury loading in the Bay of Fundy region is two to five times the magnitude of pre-industrial mercury deposition. The natural rate of atmospheric mercury deposition in the Bay of Fundy region prior to human influence was estimated to be approximately 4.1 pg mm"yfl based on coastal salt marsh sediment data. The remaining anthropogenic component of modern atmospheric inputs is estimated at 8.7 pg m-2yr-I. With recycled anthropogenic mercury that is continually released and re-deposited accounting for an estimated one-third of total deposition (Pirrone et al., 1998), the maximum reduction in total loading achievable through control of anthropogenic emissions based on the above estimates would be approximately 35%. In other words, to accomplish an immediate decline in total mercury inputs of roughly half of the current anthropogenic component of deposition requires the virtual elimination of present day anthropogenic emissions. These results help to illustrate why large reductions in anthropogenic mercury emissions in Maritime Canada have only resulted in small overall declines in mercury inputs to the Bay of Fundy region where high levels of mercury contamination are an ongoing problem in both freshwater and coastal ecosystems. This analysis also reinforces the importance of resolving differences in the bioavailability of newly deposited and historic mercury in aquatic ecosystems (Hintelmann et al., 2002) to more accurately forecast the effectiveness of emissions reductions strategies at controlling overall rates of accumulation in organisms. In chapter three, some of the key variables known to control methylmercury production in estuarine sediments were investigated. The results show a proportional relationship between total mercury and methylmercury concentrations in Passamaquoddy Bay, suggesting that methylation in this system is in part limited by the supply of inorganic mercury. Thus, as total mercury levels increase or decrease in Passamaquoddy Bay, a corresponding increase or decline in ambient methylmercury is expected. This relationship is especially significant for anticipating the response of this system to changes in mercury inputs due to emissions controls, as the majority of anthropogenic emissions are comprised of inorganic mercury species and only methylmercury bioaccumulates in organisms. The field data also indicate that the geochemical characteristics of the sediments including total organic carbon (TOC) content, redox potential (Eh) and sulfide concentrations can significantly influence methylmercury production, with the highest concentrations of methyl mercury found in sediments with elevated total organic carbon content, lower sediment redox potential and elevated sulfide concentrations. One of the major management implications of this study is that significant organic enrichment through commercial activities such as fish farming, which is known to cause increased sediment sulfide concentrations and lower redox potential (Hargrave et al., 1997, Wildish et al., 2001, Wildish et al., 1993), may also enhance exposure of organisms to methylmercury through either enhanced methylmercury production or greater persistence of methylmercury in the sediments. Thus, changes in methylmercury dynamics and potential increases in mercury concentration in organisms as the result of organic enrichment requires further investigation. The relationships developed in this chapter were also used to produce a simple semi-empirical model of mercury methylation and production in the Bay of Fundy that will be useful for developing a mass-balance model of mercury distribution. The research presented in chapter four involving sediment core and compositional analysis provides new insight into the effects of mixing on methylmercury production and mercury bioavailability in coastal sediments. The results show that, in contrast to the classic profile typically observed in unmixed sediments, methylmercury production occurs throughout the estimated 15-cm-thick active surface layer of the well-mixed sediments found in Passamaquoddy Bay. The resulting large reservoir of total mercury and methylmercury in these sediments helps to explain why concentrations in organisms in the Bay of Fundy have not fallen concurrently with emissions reductions. Current management policies need to take into account the expected delay in the response time of well-mixed estuarine systems to declines in mercury loading due to the greater reservoir of historic mercury available in these sediments that can potentially be converted to methylmercury and biomagnify in the coastal food-chains. The results of chapters two through four are integrated in chapter five to complete the development of an empirical mercury cycling model for Passmaquoddy Bay. The major findings from mass budget calculations are that: (i) the sediment compartment contains over 95% and 90% of the total and methyl mercury in Passamaquoddy Bay respectively; (ii) concentrations in the sediments can be expected to respond very slowly to changes in mercury inputs; and (iii) there is a large turnover of mercury on a daily basis through methylation and demethylation, equivalent to approximately 40% of the methylmercury reservoir in the sediments. The mercury cycling model also indicates that: (i) environmental changes that shift the present equilibrium between methylation and demethylation could result in a rapid accumulation of methylmercury in the sediment compartment that could subsequently be taken up by organisms and (ii) while concentrations in the water column of Passamaquoddy Bay reach steady state rapidly in response to changes in mercury inputs, the overall dynamics of mercury in this system are governed by the slow rate of change in mercury concentrations in the sediments. The latter implies that mercury concentrations in fish and other marine life in Passamaquoddy Bay and the larger Bay of Fundy will respond slowly to declines in atmospheric inputs. 6.2 Implications for Policy and Management Mercury is currently a "Track 2" substance under the Canadian Environmental Protection Act and is targeted for reduction to naturally occurring levels in the environment. This study has provided several key findings that can help environmental managers achieve realistic policy goals. First, the information provided by this study on the background or "natural" rate of mercury loading in the Bay of Fundy region is critical for evaluating the progress made toward reducing mercury concentrations in this region to "naturally occurring levels" and evaluating the cycling of mercury prior to human influences. Second, to achieve declines in mercury loading, the study indicates that environmental managers must address emissions on the both regional and continental scales. A comparison of the historical emissions data fiom Maritime Canada and North America with sediment archives fiom lakes, bogs and coastal salt marshes indicates that overall declines in atmospheric mercury loading are only attained in the Bay of Fundy region when emissions are reduced on both the regional and continental scales. Third, the study shows that there is a limit to the immediate effectiveness of emissions reductions initiatives in terms of achievable declines in mercury loading. The study indicates that in the short-term up to 65% of atmospheric inputs of mercury to the Bay of Fundy region are fiom recycled anthropogenic and natural sources and will not be affected by declines in present-day mercury releases. Thus, the maximum overall decline in mercury loading that can be achieved immediately in the Bay of Fundy region through emissions reductions programs is approximately 35%. Fourth, the study shows that there is a temporal response of ecosystems to declines in mercury loading that must be taken into account in designing effective management programs. At the projected 35% level of emissions reductions, preliminary application of the mercury cycling model for Passamaquoddy Bay revealed that it would take hundreds of years for total mercury concentrations to reach steady state with respect to loading rates. Long-term management strategies remain critical for reducing mercury contamination and for minimizing recycled historic inputs of mercury but their effectiveness must be assessed with this time-lag of centuries in mind. Finally, the results of this study emphasize the importance of considering not only the total amount of mercury released from anthropogenic sources but the other geochemical factors affecting the speciation and bioavailability of mercury in designing effective management strategies. In Passamaquoddy Bay, the dynamic and actively cycling pool of methylmercury in the sediment compartment could result in a rapid increase in concentrations in the sediments and organisms if the present equilibrium between methylation and demethylation is disturbed by changes in sediment geochemistry. Geochemical factors in sediments such as organic enrichment leading to changes in sulfide concentration and redox potential could have a significant short-term impact on the accumulation of mercury in organisms. By addressing these factors in mercury management strategies, managers can reduce the short-term risk of rapid mercury accumulation in the food-web and more effectively protect human and ecosystem health. By using an integrated approach to understanding mercury contamination problems in Passamaquoddy Bay and the Bay of Fundy, this study provides valuable insight into the varied and complex issues facing environmental managers of mercury contamination. The results of this study will help managers better understand how to design and negotiate programs to maximize the long-term impact of emissions reductions initiatives and what to expect over time from the programs. Also as important, managers will better understand how to contain the short-term risk of mercury contamination from other anthropogenic changes. 6.3 Future Research Directions The empirical model presented in this study provides an analytical framework with which to build and test a mechanistic mercury cycling model for coastal and marine systems that can be used to forecast the relationship between changes in anthropogenic mercury emissions and concentrations in organisms. Further evaluation of the bioavailability of newly deposited mercury and its role in determining concentrations in organisms across trophic levels is of critical importance for future research. In this study, there remained some uncertainty regarding the cycling of elemental mercury in the water column, including losses through volatilization and degradation of methylmercury through oxidative demethylation in both the water and sediments. This uncertainty should be resolved with additional study. A mechanistic food-web bioaccumulation model for Passamaquoddy Bay that can be tested using the empirical data presented in this study is the next step. It will allow the relationship between mercury emissions and concentrations in organisms to be filly developed. The resulting model will provide managers with an invaluable tool for anticipating the effects of hture policies and environmental changes on mercury concentrations in organisms and will be a valuable component of human and ecological risk assessments for mercury in coastal and marine systems. 6.4 Literature Cited Hargrave, B.T., Phillips, G.A., Doucette, L.I., White, M.J., Milligan, T.M., Wildish, D.J. and Cranston, R.E. 1997. Assessing benthic impacts of organic enrichment from marine aquaculture. Water, Air and Soil Pollution, 99,641-650. Hintelmam, H., Harris, R., Heyes, A., Huley, J. P., Kelly, C. A., Krabbenhoft, D. P., Lindberg, S., Rudd, J. W. M., Scott, K. J. and St. Louis, V. L. S. 2002. Reactivity and mobility of new and old mercury deposition in a boreal forest ecosystem during the first year of the METAALICUS study. Environmental Science and Technology, 36,5034-5040. Northeast States for Coordinated Air Use Management (NESCAUM), Northeast Waste Management Officials' Association (NEWMOA), New England Interstate Water Pollution Control Commission (NEIWPCC), Ecological Monitoring and Assessment Network (EMAN), 1998, Northeast States and Eastern Canadian Provinces Mercury Study, A Framework for Action: NESCAUM, pp. 350. Pirrone, N., Allegrini, I., Keeler, G., J., Nriagu, J. O., Rossman, R. and Robbins, J. A. 1998. Historical atmospheric mercury emissions and depositions in North America compared to mercury accumulations in sedimentary records. Atmospheric Environment, 32,929-940. Wildish, D. J., Hargrave, B. T. and Pohle, G.2001. Cost-effective monitoring of organic enrichment resulting from salmon mariculture. ICES Journal of Marine Science, 58,469-476. Wildish, D. J., Keizer, P. D. and Wilson, A. J. 1993. Seasonal changes of dissolved oxygen and plant nutrients in seawater near Salmonid net pens in the macrotidal Bay of Fundy. Canadian Journal of Fisheries and Aquatic Sciences, 50,303-3 11. APPENDIX 7.1 Supporting Data Chapter 2 Table 7-1. Unsupported 210~band 137~~data measured in sediment cores from the Bay of Fundy region. DH-A = Dipper Harbour salt marsh core A; DH-B = Dipper Harbour salt marsh core B; LL = Lily Lake core; SPL = St. Patrick's Lake core; CH = Chance Harbour salt marsh core. Core Depth (cm) Pb-210 (Bqkg) Cs-137 (Bqkg) -- DH-A 0.5 403 7 DH-A DH-A DH-A DH-A DH-A DH- A DH-A DH-A DH-A 9.8 49 13 DH-B 0.5 18 6 DH-B 3.0 DH-B 5.0 DH-B 6.5 DH-B 7.5 DH-B 8.7 DH-B 10.0 DH-B 12.5 DH-B 15.0 DH-B 19.5 LL 0 LL 0.5 LL 1.5 LL 2.5 LL 3.5 LL 5 LL 6 LL 7 LL 19 17 SPL 0 1198 SPL SPL SPL SPL SPL SPL SPL SPL SPL SPL SPL SPL SPL SPL SPL SPL SPL 19 114 CH 0.25 32 1 11 Table 7-2. Caribou Plains Bog data from Rutherford & Matthews (1998). ~g-T(ng g-' Hg-T Loading Depth (cm) Est.Year dry) (pg m-2yr") Table 7-3. Lily Lake core data fiom Kainz et al. (1996). Depth (cm) Est Year Hg-T (ng g-' dry) Eh ASEF Table 7-4. St. Patrick's Lake core data from Kainz et al. (1996). Depth (cm) Est Year Hg-T (ng g-l dry) Eh 0.0 1996 270 Table 7-5. Chance Harbour salt marsh data. Note: 1830 sediment date estimated from pollen marker horizon. ASEF = Anthropogenic sediment enrichment factor calculated by dividing the mean loading rate to sediments that accumulated past 20.5 cm depth by respective loading rates in each subsequent sediment layer. Depth (cm) Est. Year Bulk Density (g Hg-T (ng g-' Hg-T Loading ASEF ~m-~) 1 Table 7-6. Sediment data from Dipper Harbour salt marsh core A. Bulk Hg-T Depth Density Hg-T Fe Mn Loading (cm) Est Year (g cm- ) (ng g; - 1 dry) (mg g - 1) (pg g - 1) (pg m-2wl) ASEF Table 7-7. Sediment data from Dipper Harbour salt marsh core B. - Hg -T Fe Mn Depth (cm) Est Year (ng g-' dry) (mg g- 1 ) (pg g - 1) Table 7-8. Sediment data from Bocabec salt marsh. Bulk Hg-T Depth Pb Dens$ (ng g-l Fe Mn Hg-T Loading 0(pg R-l) Est. Year (g cm- ) dry) (mg g-1 ) (pg g-1 ) 21.3 1996 8 1 2.35 26 1 104 7.2 Supporting Data Chapter 3 Table 7-9. Spatial distribution of total mercury (Hg-T) in Passamaquoddy Bay. Latitude (decimal degrees) Long (decimal degrees) N Hg-T (ng g" dry) stdev 117 Table 7-10. Correlation between total mercury (Hg-T) concentrations, total organic carbon and sediment grain size at selected sampling locations. Grain size data is fiom Loring et al. (1998). % Sediment Hg-T TOC <63 um - 1 (ng g dry) (%) (Equiv. wt.) Table 7- 1 1. Raw data for total (Hg-T) and methyl (MMHg) mercury concentrations, total organic carbon (TOC), redox (Eh) and sulfide concentrations measured in grab samples collected in Passamaquoddy Bay, the St. Croix River and outer Bay of Fundy. Hg -T % Latitude Longitude TOC (ng g-' EFk MMHe/ Eh Sulfide Sample ID (dd) (dd) (%) Hg-T (mV) (uM) 4Wa 8da 1Oda 117da 81da 18da 17da 21da 18da 19da 38da 48da 48da 18da 19da 28da 37da 80da ES03-012 ES04-0 12 ES05-012 ES06-0 12 ES07-0 12 ES08-012 ES09-0 12 ES10-012 ESll-012 ES12-012 ES23-012 ES24-0 12 ES25-012 ES27-012 ES28-0 12 ES01-013 ES02-013 ES03-013 ES04-0 13 ES05-013 ES06-0 13 ES07-013 ES08-013 ES09-0 13 TM18 TM19 TM13 TM14 TM2 TM3 TM4 TM6 TM7 GHA GHB GHl GH2 GH3 GH4 GH5 da '~a~have been a problem with redox probe calibration at these stations. Table 7-12. Bivariate correlation matrices for benthic sediment grab samples obtained in May 2001. All Stations MMHg Hg-T %MMHg TOC EH SULFIDE MMHg Pearson Correlation 1 ,669" .675" .748" .331' .319 Sig. (2-tailed) ,000 .OOO ,000 .042 .OM N 41 41 41 38 38 37 Hg-T Pearson Correlation .669** 1 -.017 .804" .319 .086 Sig. (2-tailed) .OOO .918 .OOO .051 .612 N 41 41 41 38 38 37 %MMHg Pearson Correlation .675" -.017 1 .278 ,184 ,257 Sig. (2-tailed) .OOO ,918 .091 .268 .I25 N 41 41 41 38 38 37 TOC Pearson Correlation .748" .804" .278 1 .233 .303 Sig. (2-tailed) .OOO ,000 .091 .I59 .068 N 38 38 38 38 38 37 EH Pearson Correlation .331' .319 .I84 .233 1 -.238 Sig. (2-tailed) .042 ,051 .268 .I59 .I56 N 38 38 38 38 38 37 SULFIDE Pearson Correlation .319 .086 .257 .303 -.238 1 Sig. (2-tailed) .054 .612 .I25 ,068 .I56 N 37 37 37 37 37 37 .* Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). -Without St. Crolx River Stations - %MMHG TOC EH SULFIDE MMHg Pearson Correlation .389' .752" ,089 ,247 Sig. (2-tailed) .016 .OOO .612 ,160 N - 38 35 35 34 Hg-T Pearson Correlation -.356* .808*" .094 -.014 Sig. (2-tailed) .028 ,000 .592 .937 N - 38 35 35 34 %MMHg Pearson Correlation 1 -.OM .010 .249 Sig. (2-tailed) .971 .952 .I56 N 38 35 35 34 TOC Peanon Correlation -.OM 1 .045 .223 Sig. (2-tailed) .971 .797 .205 N - 35 35 35 34 EH Peanon Correlation .010 .045 1 -.321 Sig. (2-tailed) .952 .797 .064 N 35 35 35 35 35 34 SULFIDE Pearson Correlation .247 -.014 .249 .223 -.321 1 Sig. (2-tailed) .I60 .937 .I56 .205 .064 N 34 34 34 34 34 34 .* . Correlation is significant at the 0.01 level (2-tailed). '. Correlation is significant at the 0.05 level (2-tailed). All Passamaquoddy Bay Stations I MMHg 7 Hg-T %MMHg I TOC I EH I SULFIDE MMHg Pearson Correlation ( 1 I .540* .48t"l .6821 -.048 1 .215 Sig. (2-tailed) I . I .OOl N I 33 1 33 Hg-T Pearson Correlation I 540.1 1 N 33 33 %MMHg Pearson Correlation .481" -.451' Sig. (2-tailed) .005 .009 N 33 33 TOC Pearson Correlation .682*' .730' Sig. (2-tailed) .OOO ,000 N 33 33 EH Pearson Correlation -.048 -.075 Sig. (2-tailed) .791 ,679 N I 33 1 33 SULFIDE Pearson Correlation I .215 1 -.076 Sig. (2-tailed) .229 .676 N 33 33 .* Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). Passarnaquoddy Bay Stations PB-1 to PBI MMHg Hg-T %MMHg TOC EH SULFIDE MMHg Pearson Correlation 1 .187 .616*' .393 -.533* .322 Sig. (2-tailed) ,472 .008 .I19 .028 .208 N 17 17 17 17 17 17 Hg-T Pearson Correlation .I87 1 -.624* .215 -.362 .276 Sig. (2-tailed) .472 .007 ,406 .I54 .283 N 17 17 17 17 17 17 %MMHg Pearson Correlation .616" -.624" 1 .I78 -.202 .014 Sig. (2-tailed) .008 .007 ,494 .437 .956 N 17 17 17 17 17 17 TOC Pearson Correlation .393 .215 ,178 1 -.671" .715" Sig. (2-tailed) .I19 .406 .494 .003 .001 N 17 17 17 17 17 17 EH Pearson Correlation -.533* -.362 -.202 -.67lW 1 -.415 Sig. (2-tailed) .028 .I54 .437 .003 .098 N 17 17 17 17 17 17 - SULFIDE Pearson Correlation .322 .276 ,014 .715*' -.415 1 Sig. (2-tailed) .208 .283 .956 .001 .098 N 17 17 17 17 17 17 .. Correlation is significant at the 0.01 level (2-tailed). '. Correlation is significant at the 0.05 level (2-tailed). Table 7- 13. Bivariate correlation matrices for benthic sediment grab samples obtained in August 2000-200 1. All Passamaquoddy Bay Stations -- MMHg Hg-T %MMHG TOC EH SULFIDE MMHg Pearson Correlation 1 .261 ,936" .455' -.459' .529' Sig. (2-tailed) .lo4 .OOO .003 .016 .002 N I 40 1 40 1 40 1 40 27 33 Hg-T Pearson Correlation I .261 1 1 I -.OM 1 .I96 -.333 -.3811 Sig. (2-tailed) 1 .A04 1 . 1 .605 1 .224 ,089 .029 N I 40 1 40 1 40 ( 40 27 33 %MMHg Pearson Correlation I .936'1 -.OM 1 1 I .412' -.308 ,663' Sig. (2-tailed) 1 .000 1 .605 1 . I .008 .I18 .ooo N I 40 1 40 ( 40 1 40 27 33 TOC Pearson Correlation I .455'1 .I96 I .4 12'1 1 -.219 .566' Sig. (2-tailed) .003 .224 .008 .272 .001 N 40 40 40 40 27 33 EH Pearson Correlation -.459' -.333 -.308 -.219 1 -.328 I Sig. (2-tailed) .016 .089 .I18 .272 .ow i N 27 27 27 27 27 27 SULFIDE Pearson Correlation .529" -.381' .663" .566' -.328 1 Sig. (2-tailed) .002 .029 .OOO .001 .ow N 33 33 33 33 27 33 .. Correlation is significant at the 0.01 level (2-tailed). '-Correlation is significant at the 0.05 level (2-tailed). Passamaquoddy Bay Stations PB-1 to PB6 I MMHg I Hg-T 1 %MMHg I EH MMHg Pearson Correlation ) 11 .I55 1 .9Ole1 -.389 Sig. (2-tailed) 566 ,000 .I36 N 16 16 16 16 Hg-T Pearson Correlation .I55 1 -.281 -.317 Sig. (2-tailed) .566 ,292 .232 N I 16 I 16 1 16 1 16 %MMHg Pearson Correlation I .901q -.281 1 1 I -.256 Sig. (2-tailed) 1 .000 1 .292 ( . 1 .338 N I 16 I 16 1 16 1 16 EH Pearson Correlation I -.389 1 -.317 1 -.256 1 1 Sig. (2-tailed) I .I36 1 .232 1 .338 1 N I 16 1 16 I 16 1 16 TOC Pearson Correlation I ,348 1 .332 1 .208 1 -.508* Sig. (2-tailed) .I86 .209 ,440 .045 N 16 16 16 16 SULFIDE Pearson Correlation 347 -346 .524' -.353 Sig. (2-tailed) .188 .I90 .037 .I80 N 16 16 16 16 t. t. . Correlation is significant at the 0.01 level (2-tailed). '. Correlation is significant at the 0.05 level (2-tailed). Table 7-14. Bivariate correlation matrices for benthic sediment grab samples obtained in November 200 1. Passamaquoddy Bay Sig. (2-tailed) .200 .029 ,178 .320 N 9 9 8 9 9 %MMHG Pearson Correlation .721' -.814" ,274 -.I01 .787* Sig. (2-tailed) .028 .008 511 ,795 .012 N 9 9 8 9 9 *. Correlation is significant at the 0.05 level (2-tailed). t. t. Correlation is significant at the 0.01 level (2-tailed). Table 7- 15. Raw data for benthic sediment samples from Passamaquoddy Bay analyzed for porewater mercury concentrations in August 2001. Hg-I = Inorganic mercury concentration in the porewaters calculated as the difference between total mercury (Hg- T) and methylmercury (MMHg) concentrations in porewaters. TOC Hg-T MMHg Hg-T MMHg Hg-I % &- Kd- PA) (ng g-') (ng g-') (ng L-') (ng L-') (ng L") MMHg Hg-T MMHg (L kg-') (L kg-') 1.82 46.54 0.36 11.33 n/a 10.77 n/a 4107 n/a 2.49 114.24 0.38 da 0.47 n/a 1.73 4236 8 17 1.83 55.54 0.24 21.27 0.45 20.82 2.11 2611 525 2.05 53.12 0.26 11.96 0.26 11.69 2.18 4443 992 1.61 56.17 0.32 15.52 0.95 14.57 6.13 3620 340 1.80 42.55 0.33 nla 2.06 da n/a nla 159 1.71 41.64 0.34 16.35 1.21 15.14 7.41 2547 277 1.43 39.61 0.25 14.98 1.06 13.92 7.08 2643 232 1.03 39.28 0.16 29.55 0.94 28.60 3.19 1329 171 1.88 46.08 0.31 9.72 0.56 9.16 5.74 4740 548 1.90 47.40 0.28 n/a 0.81 n/a da nla 342 1.69 47.50 0.28 8.76 nla 8.32 n/a 5423 n/a 2.16 49.07 0.36 8.43 n/a 8.01 nla 5820 da Table 7-1 6. Raw Data Collected at Seasonally Monitored Stations. Spring (May) Hg -T % TOC Eh Sulfide SC- 1 SC-1 SC- 1 PB-1 PB- 1 PB-1 PB-2 PB-2 PB-2 PB-3 PB-3 PB-3 PB-4 PB-4 PB-4 PB-5 PB-5 PB-5 PB-6 PB-6 August (Summer) MMHg Hg-T % TOC Eh Sulfide Station 0% 8) (ng g-') MMHg (mV) (mv) SC-1 0.42 83 0.50% 2.18 133 210 SC-1 0.38 114 0.33% 2.49 70 200 SC-1 0.73 149 0.49% 3.14 94 250 PB-1 0.2 1 46 0.47% 1SO 131 83 PB- 1 0.16 39 0.41% 1.03 280 80 PB- 1 0.22 44 0.50% 1.10 364 23 PB-2 0.26 45 0.58% 1.78 118 1300 PB-2 0.20 43 0.45% 1.90 138 190 PB-2 0.36 47 0.77% 1.82 133 450 PB-3 0.24 56 0.42% 1.83 134 870 PB-3 0.26 53 0.49% 2.05 129 150 PB-3 0.32 56 0.58% 1.61 171 32 PB-4 0.3 1 46 0.66% 1.88 132 590 PB-4 0.28 47 0.58% 1.90 124 480 PB-4 0.30 39 0.76% 2.12 155 690 PB-5 0.33 43 0.77% 1.80 133 810 PB-5 0.34 42 0.81% 1.71 116 1400 PB-5 0.25 40 0.62% 1.43 140 1900 PB-6 0.24 41 0.57% 2.13 298 580 PB-6 0.26 44 0.59% 2.15 278 1000 PB-6 0.22 42 0.53% 2.14 264 400 Fall (November) MMHg Hg-T % TOC Eh Sulfide MMHg (ng g-'1 (ng g-') MMHg (%I (mv) (mv) (ng g-'1 PB- 1 0.25 44 0.56% 1.46 124 1500 0.26 PB- 1 0.32 6 1 0.53% 1.25 195 5 8 0.23 PB- 1 0.28 48 0.60% 1.69 214 74 0.35 PB-3 0.3 1 56 0.55% 2.86 57 560 0.32 PB-3 0.3 1 59 0.52% 2.05 139 990 0.33 PB-3 0.36 49 0.73% 2.16 164 1000 0.38 PB-5 0.34 47 0.72% 2.42 114 2000 0.35 PB-5 0.36 37 0.97% 1.41 134 2500 0.38 PB-5 0.36 41 0.87% 2.52 124 4000 0.34 Spring (May) MMHg Hg-T % TOC Eh Sulfide (ns g-'1 (ng g-'1 MMHg ("/.I mv) (mv) PB-1 0.26 47 0.55% 1.30 512 42 PB-1 0.23 28 0.82% 1.38 410 25 PB- 1 0.35 42 0.84% 1.55 475 34 PB-3 0.32 58 0.55% 1.96 117 34 PB-3 0.33 57 0.58% 1.97 134 680 PB-3 0.38 49 0.79% 1.70 129 85 PB-5 0.35 49 0.71% 1.47 117 220 PB-5 0.38 43 0.89% 1.68 125 10 PB-5 0.34 45 0.77% 1.71 154 400 Summer (August) MMHg Hg-T % TOC Eh Sulfide (ng g-'1 (ng g-9 MMHg PB- 1 0.2 1 45.51 PB- 1 0.16 39.28 PB- 1 0.22 43.92 PB-3 0.24 55.54 PB-3 0.26 53.12 PB-3 0.32 56.17 PB-5 0.33 42.55 PB-5 0.34 41.64 PB-5 0.25 39.61 Table 7-17. Summary statistics for t-tests of seasonal differences in paired means. Total Mercury Concentrations Hvoothesized--,r - mean difference = 0 t-Test: Paired Two Sample for Means Mav August Mean Variance Observations Pearson Correlation Hypothesized Mean Difference df t Stat P(T<==t) one-tail t Critical one-tail P(T<=t) two-tail t Critical two-tail Reject null? no May November Mean 46.35575 49.0123025 Variance Observations Pearson Correlation Hypothesized Mean Difference df t Stat P(T<=t) one-tail t Critical one-tail P(T<=t) two-tail t Critical two-tail Reject null? no November August Mean 49.0123 46.3724093 Variance 66.78718 45.6844008 Observations 9 9 Pearson Correlation 0.407355 Hypothesized Mean Difference 0 df 8 t Stat 0.964171 P(T<=t) one-tail 0.181598 t Critical one-tail 1.859548 P(T<=t) two-tail 0.363196 t Critical two-tail 2.306006 Reject null? no Total Organic Carbon Hypothesized mean difference = 0 t-Test: Paired Two Sample for Means May August Mean 1.997787 1.889471 Variance 0.241212 0.206504 Observations 2 1 21 Pearson Correlation 0.824554 Hypothesized Mean Difference 0 df 20 t Stat 1.758654 P(T<=t) one-tail 0.046965 t Critical one-tail 1.7247 18 P(T<=t) two-tail 0.093929 t Critical two-tail 2.085962 Reject null? no November May Mean 1.979459 1.637718 Variance 0.314188 0.055214 Observations 9 9 Pearson Correlation 0.610509 Hypothesized Mean Difference 0 df 8 t Stat 2.244815 P(T+=t) one-tail 0.027505 t Critical one-tail 1.859548 P(T<=t) two-tail 0.05501 1 t Critical two-tail 2.306006 Reject null? November August Mean 1.979459 1.561974 Variance 0.314188 0.1 14303 Observations 9 9 Pearson Correlation 0.519017 Hypothesized Mean Difference 0 df 8 t Stat 2.601524 P(T<=t) one-tail 0.015772 t Critical one-tail 1.859548 P(T<=t) two-tail 0.03 1543 t Critical two-tail 2.306006 Reject null? no Methyl Mercury Concentrations t-Test: Paired Two Sample for Means Hypothesized mean difference = 0 Stations SC-1 through PB-6 May August Mean 0.4229807 0.29841 18 Variance 0.08 18352 0.01 38902 Observations 2 1 21 Pearson Correlation 0.9158425 Hypothesized Mean Difference 0 df 20 t Stat 3.0972 194 P(T<=t) one-tail 0.0028403 t Critical one-tail 1.724718 P(T<=t) two-tail 0.0056806 t Critical two-tail 2.0859625 Reject null? Yes -- -- Without Station SC-1 (River) May August Mean 0.3311418 0.2631816 Variance 0.0069168 0.0028809 Observations 18 18 Pearson Correlation 0.4840591 Hypothesized Mean Difference 0 df 17 t Stat 3.8963217 P(T<=t) one-tail 0.0005805 t Critical one-tail 1.7396064 P(T<=t) two-tail 0.00 11609 t Critical two-tail 2.1098 185 Reject null? Yes Stations PB-1, PB-3, PB-5 November May Mean 0.3202257 0.327198 Variance 0.0014666 0.0027983 Observations Pearson Correlation Hypothesized Mean Difference df t Stat P(T<=t) one-tail t Critical one-tail P(T<=t) two-tail t Critical two-tail Reject null? no Stations PB-1, PB-3, PB-5 November August Mean 0.3202257 0.2575472 Variance 0.0014666 0.0036073 Observations 9 9 Pearson Correlation 0.5906538 Hypothesized Mean Difference 0 df 8 t Stat 3.8733038 P(T<=t) one-tail 0.0023596 t Critical one-tail 1.8595483 P(T<=t) two-tail 0.0047 192 t Critical two-tail 2.3060056 Reject null? Yes % Methylmercurv (%MMH& Hypothesized mean difference = 0 t-Test: Paired Two Sample for Means May August Mean 0.00817 0.00566 Variance 1.38E-05 1.69E-06 Observations 2 1 2 1 Pearson Correlation -0.026 Hypothesized Mean Difference 0 df 20 t Stat 2.898326 P(T<=t) one-tail 0.004444 t Critical one-tail 1.7247 18 P(T<=t) two-tail 0.008887 t Critical two-tail 2.085962 Reject null? Yes November May Mean 0.006728 0.0072 15 Variance 2.6E-06 1.73E-06 Observations 9 9 Pearson Correlation 0.606763 Hypothesized Mean Difference 0 df 8 t Stat -1.10363 P(T<=t) one-tail 0.150921 t Critical one-tail 1.859548 P(T<=t) two-tail 0.301842 t Critical two-tail 2.306006 Reject null? no November August Mean 0.006728 0.005618 Variance Observations Pearson Correlation Hypothesized Mean Difference df t Stat P(T<=t) one-tail t Critical one-tail P(T<=t) two-tail t Critical two-tail Reject null? Yes May August Mean 0.007376 0.005866 Variance 5.28E-06 1.57E-06 Observations 18 18 Pearson Correlation 0.092301 Hypothesized Mean Difference 0 d f 17 t Stat 2.548667 P(T<=t) one-tail 0.010383 t Critical one-tail 1.739606 P(T<=t) two-tail 0.020767 t Critical two-tail 2.109819 Reject null? no 7.3 Supporting Data Chapter 4 Table 7-18. Push core mercury data. Core ID: ES29-012 (August 2001) Hg-T-mean H -T (ng MeHg (ngKd (L kg- Log Kd (ng g-') Hg-T stdev L-9 ) g-'1 ' 1 (L kg-') % MMHg Core 1 0-2 cm 39.41 12.26 0.300 3216 3.51 0.76% 2-4 cm 38.74 2.53 9.49 0.298 4083 3.61 0.77% 4-6 cm 41 .07 5.21 12.30 0.280 3340 3.52 0.68% 6-8 cm 38.67 0.72 12.43 0.338 3111 3.49 0.87% 8-10 cm 41.71 3.28 11.32 0.361 3684 3.57 0.87% 10-12 cm 40.65 2.10 10.64 0.360 3821 3.58 0.89% 12-14 cm 39.16 2.68 12.83 0.361 3053 3.48 0.92% Core 2 0-2 cm 35.51 2-4 cm 43.29 4-6 cm 44.05 6-8 cm 36.83 8-10 cm 43.5 1 10-12 cm 43.9 1 Core ID: ES30-01 (May 2001) Depth Hg-T (ng g-') LO1 (%) TOC (%) 0-2 cm 41.39 7.28% 1.63 2-4 cm 43.92 6.62% 1.39 4-6 cm 44.20 7.10% 1.56 6-8 cm 44.02 6.95% 1.51 8-10 cm 43.16 7.23% 1.61 10-12 cm 43.90 5.42% 1.OO Core ID: ES42-01 Nay2001) Depth Hg-T ng g-' LO1 (%) TOC (%) SC- 1 (Head of River) - Core ES 19-0 12 Core 1 z wet:dry ratio Hg-T (ng g-l) MMHg (ng g-') %MMHg LO1 (%) 0 1.75 216.61 0.347 0.16% 14.56 2 1.83 117.12 1.485 1.27% 12.67 4 1.70 195.94 1.049 0.54% 11.86 6 1.89 585.70 0.819 0.14% 12.75 8 2.08 186.54 0.450 0.24% 12.51 10 1.86 153.69 0.231 0.15% 11.26 12 2.30 135.36 0.229 0.17% 14.78 14 2.20 172.77 0.162 0.09% 12.09 Core 2 0 1.72 96.94 0.385 0.40% 7.52 2 2.1 1 245.57 0.619 0.25% 10.15 4 149.06 6 2.36 76.98 0.411 0.53% 13.18 8 2.31 142.53 0.275 0.19% 12.84 10 2.31 98.62 0.087 0.09% 14.13 12 2.48 91.27 0.104 0.11% 14.61 14 2.21 92.72 0.066 0.07% 11.97 ES07-012 (Same location as Gravity Core 21) Core 1 z Hg-T (ng MMHg (ng g'l) %MMHg LO1 (%) 0 46.32 2 60.38 4 44.55 6 59.95 8 57.46 10 69.38 12 65.94 14 46.32 Core 2 0 43.95 2 53.39 4 54.81 6 63.94 8 59.20 10 72.74 Table 7-19. Raw data fiom gravity cores collected in Passamaquoddy Bay. core Z Salinity NH4 SO4 T.C. O.C. 1.C Cd Cu Fe Li Mn Ni Pb Zn Cm ppt mM rnM % % % pprn pprn pprn pprn pprn pprn pprn pprn 5 0 30 0.1 25 1.05 1.05 0.00 0.06 17 2.07 31 243 20 17 56 core Z Salinity NH4 SO4 T.C. O.C. LC Cd Cu Fe Li Mn Ni Pb Zn cm ppt mM mM % % % pprn pprn pprn pprn pprn pprn pprn pprn 10 5 31 0.2 24 1.27 1.11 0.16 0.08 12 1.95 31 290 21 11 63 core Z Salinity NH4 SO4 T.C. O.C. 1.C Cd Cu Fe Li Mn Ni Pb Zn cm ppt mM rnM O/O '10 '10 ppm ppm ppm ppm ppm ppm ppm ppm core z Salinity NH4 SO4 T.C. O.C. 1.C Cd Cu Fe Li Mn Ni Pb Zn cm ppt mM mM O/o O/o O/o ppm ppm ppm ppm ppm ppm ppm ppm core z Salinity NH4 SO4 T.C. O.C. LC Cd Cu Fe Li Mn Ni Pb Zn cm ppt mM mM Oh '10 Oh ppm ppm ppm ppm ppm ppm ppm ppm 0.4 25 1.30 1.30 0.00 0.09 12 2.43 39 332 32 core z Salinity NH4 SO4 T.C. O.C. 1.C Cd Cu Fe Li Mn Ni Pb Zn cm ppt mM mM O/O O/O O/O pprn pprn pprn pprn pprn pprn pprn pprn - core z Salinity SO4 T.C. O.C. 1.C Cd -cm ppt mM % % % ppm 55 5 31 24 1.83 1.68 0.15 0.08 27 1.83 1.65 0.18 0.09 25 1.72 1.60 0.12 0.09 25 1.73 1.57 0.16 0.08 25 1.58 1.44 0.14 0.08 25 1.40 1.20 0.20 0.08 24 1.22 1.07 0.15 0.08 27 1.26 1.13 0.13 0.10 25 1.23 1.05 0.18 0.08 25 1.24 1.07 0.17 0.08 core z Salinity NH4 SO4 T.C. O.C. 1.C Cd Cu Fe Li Mn Ni Pb Zn cm ppt mM mM % % % ppm ppm ppm ppm ppm ppm ppm ppm 0.4 25 1.65 1.13 0.52 0.13 12 1.96 32 321 21 18 70 Min. Max. Mean n Std Dev 7.4 Supporting Data Chapter 5 Table 7-20. Summary of annual data for mercury concentration and deposition in precipitation collected at weekly intervals between July 1996 and June 2000. Data from: Beauchamp (1 998); Beauchamp (unpublished data, 2000). Concentrations are expressed as volume weighed means ('Vol Wt. Conc.') to reduce the effects of small samples with high concentrations and thus providing a more representative overall concentration. The standard deviation of mean concentration in precipitation in the "average" column is the standard deviation around the yearly mean values and does not take into account the deviation around each annual mean value. Year 96-97 97-98 98-99 99-00 Average Vol Wt Conc. ng L-' 7.89 6.84 5.68 7.53 6.98 stdev mean concent 11.5 14.1 13.9 7.9 11.84 Mean Concentration ng L-I 10.2 11.8 12.4 9.1 10.88 Total Deposition ng.m-2 yr-' 79 16 6700 5465 872 1 720 1 Total Deposition pg m-2 w' 7.9 6.7 5.5 8.7 7.20 Total precipitation1 mrn yr-' 1 15 111 170 965IlO 19 95319 19 1l6Olll82 1065 Min 2.3 0.9 1.6 0.2 0.23 Max 70.0 89.1 77.7 50.4 89.12 ' Two monitors apart were used to measure precipitation rates, the numbers to the left are fiom Belfort weather station located 5 meters from the mercury deposition monitor in St. Andrews, NB. This monitor is part of the Mercury Deposition Network (MDN) component of the North American Deposition Network (NADP). Table 7-21. Monthly freshwater discharges in P. Bay. Data adapted from: Gregory et al., (1993). Data presented are sum of fi-eshwater discharges into P. Bay and the St. Croix River estuary. Monthly Discharge Rates (m3 mo-' x lo8): Winter Spring Summer Fall Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec Percent Total Discharge (%): Table 7-22. Total mercury (Hg-T) concentrations measured in three main freshwater tributaries of Passamaquoddy Bay. Data from: Dalziel et al. (1998); Dalziel (unpublished data, 2002). Hg-T (ng L-') Digdegaush Magauadavic St. Croix Weighted mo. avg. Nov-94 5.67 6.97 9.26 Weighted am. avg. 4.87 3.74 4.20 4.17 Table 7-23. Methylmercury (MMHg) concentrations measured in three main frcshwtcr tributaries of Passamaquoddy Bay. Data fiom: Dab1(\mpublishcd da& 2002). St. Weighted mo. Digdegaush Magauadavic Croix avg. Weighted am. avg. (Pg L-l)* 262 297 214 241 'NOV. concentrations were used for fall and winter; May for spring and summer to calculate annual mean MMHg concentration. Table 7-24. Suspended particulate matter (SPM) mg L*' in freshwater tributaries of P. Bay. Data from: Dalziel et al. (1998). Measured concentrations are weighted by the respective discharges from each river for the means; annual averages are calculated from the fraction of discharge occurring during each sampling period. Seawater samples were filtered using a 0.4 pm Nucleopore filtration system. (SW mg L-' Digdegaush Magaguadavic St. Croix Weighted avg. May 5.09 2.25 3.24 3.23 3.23 2.26 3.46 3.13 1.35 2.06 4.27 3.33 Mean 3.22 2.19 3.66 3.23 August 2.75 2.54 3.02 2.86 2.39 2.30 3.91 3.31 2.3 1 2.14 3.51 3.01 Mean 2.48 2.33 3.48 3.06 November 1.19 2.21 4.46 3.46 2.27 3.85 6.19 5.08 1.78 2.34 4.33 3.49 Mean 1.75 2.80 4.99 4.01 Annual 2.44 2.49 4.24 3.56 Table 7-25. Summary of annually averaged inputs of mercury to Passamaquoddy Bay from rivers, atmospheric deposition and tidal inflow. Flux of Hg-T from freshwater discharges 19.5 1.13 Dissolved mercury inputs fiom rivers 14.2 0.99 Solid phase mercury inputs from rivers 4.16 0.14 Flux of HgT from atmospheric inputs 1.43 0.0143 Wet deposition flux 0.950 0.0095 Dry deposition flux 0.480 0.0048 ...... Flux of HgT fiom incoming tidal waters 15.1 3.08 Flux of dissolved phase tidal inputs 13.9 3 .03 Flux of solid phase tidal inputs 1.28 0.056 ------.---- Total External Inputs of Mercury 36.1 4.23 Table 7-26. Mercury concentrations measured in Gammarus sp. collected from St. Andrews tidal flats in August 2001. Hg-T (ng g-') wet wt. g wet/ g dry Hg-T (ng g-l) dry wt. 24.3 5.57 135 27.5 5.99 165 19.3 5.57 108 22.6 5.57 126 25.3 5.99 152 Table 7-27. Mercury concentrations measured in Nephtys sp. at the head of the St. Croix River Estuary and in Passmaquoddy Bay, August 2001. Hg -T Hg-T Hg-T-sed MMHg-sed - 1 (ng g ) wet wt. g wet/ g dry (ng g-') dry wt (ng g-l) wet wt (ng g'l) wet wt Passamaquoddy Bay 10.99 9.54 105 21.0 0.10 12.48 9.54 119 21.0 0.10 10.12 8.38 85 8.91 8.38 75 8.14 25.12 205 8.17 25.12 205 11.OO 8.82 97 22.6 0.10 10.91 8.82 96 22.6 0.10 17.90 8.82 158 10.92 8.82 96 15.44 0.10 St. Croix River Estuary 6.45 7.77 50 35.9 0.14 8.08 7.77 63 35.9 0.14