Climate Change Impacts on Mercury Cycling in …...ii Climate Change Impacts on Mercury Cycling in...

Climate Change Impacts on Mercury Cycling in Peatlands

by

Kristine Marie Haynes

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy

Department of Geography and Planning University of Toronto

© Copyright by Kristine Marie Haynes, 2017

ii

Climate Change Impacts on Mercury Cycling in Peatlands

Kristine Marie Haynes

Doctor of Philosophy

Department of Geography and Planning

University of Toronto

2017

Abstract

Climate change, through hydrological impacts and shifts in vascular plant communities, may

significantly affect the strength of peatlands as sinks of inorganic mercury (Hg) and downstream

sources of neurotoxic methylmercury (MeHg). Four complementary studies were conducted at

the PEATcosm (Peatland Experiment at the Houghton Mesocosm Facility) and SPRUCE

(Spruce and Peatland Responses Under Climatic and Environmental change) sites to investigate

the anticipated effects on Hg cycling. Lowered, more variable water tables and the removal of

Ericaceae shrubs in the PEATcosm mesocosms significantly enhance Hg and MeHg mobility in

peat pore waters as well as export from the peat in runoff during the snowmelt period. Mercury

is mobilized in association with dissolved organic carbon leached from the peat as a result of

increased aerobic decomposition. Enhanced Hg and MeHg solid phase peat concentrations are

observed within the zone of the lowered, fluctuating water tables due to translocation within the

peat profile. This shift in peat concentrations does not correspond to any significant differences

in Hg methylation or MeHg demethylation. The strongest trend of Hg deposition to peat occurs

with increased sedge cover, possibly as a result of coincidental shuttling of Hg to the peat by

aerenchymous tissues. Different plant functional groups alter the pathway of gaseous Hg

exchange, with greater Hg sorption to leaf surfaces and accumulation in leaf tissues by stomatal

uptake for the Ericaceae shrubs as compared to the sedges. Total gaseous Hg fluxes increase

iii

marginally with deep soil warming. Isotopic analyses demonstrate differences in Hg depositional

sources to the two peatland sites, likely due to differences in snow accumulation, and provide

insight into the mechanisms governing Hg cycling in peatlands. Collectively, these studies

demonstrate that anticipated climate change may significantly affect peatland Hg mobility and

cycling within and among the terrestrial, atmospheric and hydrological pools.

iv

Acknowledgments

I would like to express my gratitude to all those who gave of their valuable time on my journey

to the completion of my research.

Thank you to my supervisor, Dr. Carl Mitchell, for his comprehensive, capable guidance and

support. I am also grateful for the professional development opportunities provided to me by Dr.

Mitchell.

Thank you to my committee members, Dr. George Arhonditsis and Dr. Bridget Bergquist, for

their advice with regard to data analysis and interpretation, and their constructive suggestions

over the course of this research.

I appreciate the financial assistance provided through the Natural Sciences and Engineering

Research Council of Canada (NSERC) Alexander Graham Bell Canada Graduate Scholarship

(CGS-Doctoral), an Ontario Graduate Scholarship, and the University of Toronto awards

program.

I am grateful to Dr. Evan Kane (Michigan Technological University), Lynette Potvin, Dr. Erik

Lilleskov (USDA Forest Service Northern Research Station, Houghton, MI) and Dr. Randy

Kolka (USDA Forest Service Northern Research Station, Grand Rapids, MN) for facilitating

access and collaboration with the PEATcosm experiment.

Thank you to Dr. Paul Hanson (Oak Ridge National Laboratory) and Dr. Randy Kolka for

granting access to the SPRUCE site, and Robert Nettles for assistance in field sampling. Thank

you to Deacon Kyllander at the Marcell Experimental Forest Northern Research Station (USDA

Forest Service), Minnesota for logistical field assistance.

Within the Mitchell lab group, thank you to Kevin Ng for hours of extensive field and laboratory

assistance and Planck Huang for considerable laboratory analysis. I also appreciate the

assistance of Raymond Co, Brent Perron, and Ilana Tavshunsky in the lab.

Thank you to the Bergquist lab for the use of their mercury isotope analytical equipment under

the capable guidance of Dr. Wang Zheng and the assistance of Laura Zimmermann.

v

Thank you to my external appraiser Dr. Scott Brooks, Oak Ridge National Laboratory and

external examiners Dr. Sarah Finkelstein and Dr. Igor Lehnherr.

Thanks also to Jessica Finlayson, Geography Graduate Program Administrator for her assistance

dealing with academic matters.

To my family, my sincere gratitude for their ongoing and unwavering support through these

academic years to achieve the end towards my new beginning.

vi

Table of Contents

Acknowledgments.......................................................................................................................... iv

Table of Contents ........................................................................................................................... vi

List of Tables ................................................................................................................................. xi

List of Figures .............................................................................................................................. xiii

List of Appendices .........................................................................................................................xx

Chapter 1 Introduction .....................................................................................................................1

1.1 Mercury Cycling in the Environment ..................................................................................1

1.1.1 Overview of the Mercury Cycle ..............................................................................1

1.1.2 Mercury Methylation and MeHg Demethylation in Wetlands ................................3

1.1.3 Mercury - Organic Matter Relationships .................................................................6

1.1.4 Soil-Air Hg Exchange ..............................................................................................7

1.1.5 Mercury Stable Isotope Fractionation ......................................................................9

1.2 Peatland Hydrology and Hydrogeology.............................................................................10

1.3 Peatland Ecosystems and Global Climate Change ............................................................12

1.3.1 Global Climate Change and Peatland Hydrological Processes..............................12

1.3.2 Peatland Ecological Effects due to Climate Change ..............................................14

1.3.3 Potential Impacts on Mercury Cycling ..................................................................17

1.4 Objectives, Research Questions and Study Sites ...............................................................20

1.5 Thesis Structure and Publication Information ...................................................................25

1.5.1 Chapter 1 ................................................................................................................25

1.5.2 Chapter 2 ................................................................................................................25

1.5.3 Chapter 3 ................................................................................................................26

1.5.4 Chapter 4 ................................................................................................................27

1.5.5 Chapter 5 ................................................................................................................27

vii

1.5.6 Chapter 6 ................................................................................................................28

1.6 References ..........................................................................................................................28

Chapter 2 Mobility and Transport of Mercury and Methylmercury in Peat as a Function of

Changes in Water Table Regime and Plant Functional Groups ................................................45

2.1 Abstract ..............................................................................................................................45

2.2 Introduction ........................................................................................................................45

2.3 Materials and Methods .......................................................................................................48

2.3.1 Study Site and Experimental Design .....................................................................48

2.3.2 Water Sampling .....................................................................................................51

2.3.3 Pre-Treatment Peat.................................................................................................52

2.3.4 Peat Decomposition Assays ...................................................................................52

2.3.5 Analytical Methods ................................................................................................53

2.3.6 Statistical Analyses ................................................................................................53

2.4 Results and Discussion ......................................................................................................54

2.5 Conclusions ........................................................................................................................63

2.6 Acknowledgements ............................................................................................................64

2.7 References ..........................................................................................................................65

Chapter 3 Gaseous Mercury Fluxes in Peatlands and the Potential Influence of Climate

Change.......................................................................................................................................72

3.1 Abstract ..............................................................................................................................72

3.2 Introduction ........................................................................................................................73

3.3 Methods..............................................................................................................................76

3.3.1 PEATcosm Site Description ..................................................................................76

3.3.2 SPRUCE Site Description......................................................................................77

3.3.3 Experimental design and mercury flux measurements ..........................................78

3.3.4 PEATcosm vegetation and surface peat Hg ...........................................................81

3.3.5 Analytical Methods ................................................................................................82

viii

3.3.6 Statistical Analyses ................................................................................................82

3.4 Results ................................................................................................................................83

3.4.1 PEATcosm – water table and plant functional groups influence ...........................83

3.4.2 SPRUCE bog TGM fluxes .....................................................................................89

3.5 Discussion ..........................................................................................................................92

3.5.1 PEATcosm peat monoliths – influence of vascular plant community on TGM

fluxes ......................................................................................................................92

3.5.2 SPRUCE bog fluxes – deep soil warming influence on TGM fluxes....................96

3.5.3 Peatland TGM fluxes compared to other wetland environments ...........................97

3.6 Conclusions ........................................................................................................................98

3.7 Acknowledgements ............................................................................................................98

3.8 References ..........................................................................................................................99

Chapter 4 Impacts of Experimental Alteration of Water Table Regime and Vascular Plant

Community Composition on Solid Phase Peat Mercury Profiles and Methylmercury

Production ...............................................................................................................................109

4.1 Abstract ............................................................................................................................109

4.2 Introduction ......................................................................................................................110

4.3 Methods............................................................................................................................113

4.3.1 Study Site and Experimental Design ...................................................................113

4.3.2 Peat Sampling ......................................................................................................114

4.3.3 Pore Water Sampling ...........................................................................................116

4.3.4 Analytical Methods ..............................................................................................117

4.3.5 Statistical Analyses ..............................................................................................118

4.4 Results ..............................................................................................................................119

4.4.1 Hg Methylation and MeHg Demethylation .........................................................122

4.4.2 Hg(II) and MeHg Partitioning .............................................................................125

4.5 Discussion ........................................................................................................................126

ix

4.5.1 Conclusions and Implications ..............................................................................130

4.6 Acknowledgements ..........................................................................................................132

4.7 References ........................................................................................................................132

Chapter 5 Contrasting Mercury Isotopic Compositions of Two Sub-boreal Peatlands ...............141

5.1 Abstract ............................................................................................................................141

5.2 Introduction ......................................................................................................................142

5.3 Methods............................................................................................................................145

5.3.1 S1 Peatland Site Description ................................................................................145

5.3.2 PEATcosm Site Description ................................................................................145

5.3.3 Peat Sampling and THg Analysis ........................................................................147

5.3.4 Total Gaseous Mercury Flux Measurements .......................................................148

5.3.5 Sample Preparation and Mercury Isotope Analysis .............................................149

5.3.6 Data Analyses ......................................................................................................150

5.4 Results and Discussion ....................................................................................................151

5.4.1 S1 Peat Isotopic Composition ..............................................................................151

5.4.2 S1 Peat MIF .........................................................................................................156

5.4.3 PEATcosm Peat Isotopic Signature .....................................................................157

5.4.4 PEATcosm Treatment Effects – Water Table and Plant Functional Groups .......159

5.4.5 PEATcosm Peat MIF ...........................................................................................161

5.4.6 Conclusions and Implications ..............................................................................162

5.5 Acknowledgements ..........................................................................................................163

5.6 References ........................................................................................................................164

Chapter 6 Summary and Synthesis ..............................................................................................172

6.1 Summary ..........................................................................................................................172

6.1.1 Hg and MeHg Mobility in Peat Pore Water and Runoff .....................................173

6.1.2 Solid Phase Hg and MeHg – Translocation and Net MeHg Production..............174

x

6.1.3 Vascular Plant Communities – Interactions with Atmospheric Hg Deposition ..176

6.1.4 Spatial Variability in Hg Isotopic Signatures ......................................................177

6.1.5 Impacts of Climate Change on Peatland Hg Cycling ..........................................177

6.2 Limitations and Future Research Directions ....................................................................178

6.3 References ........................................................................................................................181

Appendix A. Supplementary Information for Chapter 2 .............................................................182

Appendix B. Supplementary Information for Chapter 3 .............................................................194

Appendix C. Supplementary Information for Chapter 4 .............................................................197

Appendix D. Supplementary Information for Chapter 5 .............................................................200

xi

List of Tables

Table 2-1 PEATcosm full-factorial experimental design. N=4 mesocosm bins per crossed

treatment. ...................................................................................................................................... 49

Table 2-2 Mean ± standard deviation THg and MeHg concentrations (in ng g-1

) in solid-phase

peat prior to experimental manipulations (n = 8 samples per treatment and depth increment). ... 56

Table 5-1 Contributions of Hg(II) (fHg(II)) and Hg(0) (fHg(0)) to peat isotopic signatures for the

PEATcosm 2014 treatment peat and S1 bog 2014 peat. Results are the mean of 1000 iterations

of the isotopic mixing model with Monte Carlo simulation based on the Δ201

Hg isotopic values.

..................................................................................................................................................... 152

Table A-2-1 Volume of water drained (mean ± standard deviation, in L) from the mesocosm

bins during spring snowmelt 2014 to achieve the water table treatment positions (n=4 per

treatment). ................................................................................................................................... 193

Table D-5-1 Mercury isotope signatures of reference materials and standards. ........................ 204

Table D-5-2 Mercury isotope signatures of all peat samples. Each sample was analyzed twice

for isotopes. 2σ is the higher of either 2 standard deviation (2SD) of the JTBaker standard or 2

standard error (2SE) of the two replicate results of each sample. .............................................. 205

Table D-5-3 Mean and standard deviation values of Δ199

Hg, Δ200

Hg and Δ201

Hg (all expressed in

‰) for Hg(II) and Hg(0) used in the binary mixing model with Monte Carlo simulation. ........ 206

Table D-5-4 Contributions of Hg(II) (fHg(II)) and Hg(0) (fHg(0)) to peat isotopic signatures for the

PEATcosm 2014 treatment peat and S1 2014 peat. Results are the mean of 1000 iterations of the

isotopic mixing model with Monte Carlo simulation based on the Δ199

Hg isotopic values. ...... 207

xii




Hg isotopic values. ...... 208

xiii

List of Figures

Figure 1-1 Dominant aspects of the mercury cycle in a peatland system (Source: C. Mitchell). .. 2

Figure 1-2 Potential synergistic and antagonistic implications of climate-induced peatland

hydrologic changes and hydraulic parameters (white boxes) and their potential impacts on Hg

cycling and transport (grey boxes). Solid arrows represent direct effects and dashed arrows

represent indirect or potential effects. (Mercury implications added to climate-induced peatland

hydrological changes that are modelled after Whittington and Price 2006). ................................ 18

Figure 1-3 PEATcosm Mesocosm Facility in Houghton, Michigan with 24 cubic metre peat

mesocosm bins. ............................................................................................................................. 22

Figure 1-4 Climate-controlled tunnel beneath Mesocosm Facility allowing access to the 24

mesocosm bins. Each bin was equipped with overflow tubing and collection bin to allow for

runoff sampling from each mesocosm, particularly during the snowmelt period. ....................... 23

Figure 1-5 Aerial view (left) of the SPRUCE plot footprints and boardwalks in the S1 bog,

Marcell Experimental Forest, Minnesota (September 2013 aerial photo from Paul Hanson, Oak

Ridge National Laboratory (ORNL)). One SPRUCE experimental plot (right). ......................... 25

Figure 2-1 Mean water table positions (in cm below the peat surface) in the low and high WT

treatment mesocosms and precipitation (in mm) over the course of this study from 2013 to 2015.

Water table manipulations were conducted in the summer months, while water table levels were

left to stabilize during the winter months. Dashed lines around the low and high WT treatment

mean water table positions represent the 95% confidence interval. Pore water and snowmelt

sampling events are denoted. ........................................................................................................ 50

Figure 2-2 Total Hg and MeHg concentrations and %MeHg in pore water and snowmelt runoff

in relation to water table and vascular plant functional group manipulations (treatment means of

all depths and sampling events). (a) pore water THg, (b) pore water MeHg, (c) pore water

%MeHg, (d) snowmelt runoff THg, (e) snowmelt runoff MeHg, and (f) %MeHg in snowmelt

runoff. Letters denote statistically similar groups based on transformed data. No significant

differences were observed in snowmelt runoff %MeHg across treatments. Significance of

xiv

treatment effects both individual (water table “WT” and plant functional group “Veg”) and

interactive (“WT*Veg”) for THg, MeHg, and %MeHg for both pore waters and snowmelt runoff

are noted. n = 318 pore water samples total (n = 50–56 per treatment). ...................................... 55

Figure 2-3 Relationships between shallow pore water (20 and 40 cm below the peat surface) Hg

and mean % mass loss of cellulose decomposition assays in the top 40 cm of the peat. (a) THg

concentrations, (b) MeHg concentrations, (c) %MeHg. Error bars represent standard deviation.

Decomposition assays were harvested in 2014 and represent potential peat decomposition among

the treatments. Pore water Hg data were averaged by treatment from all five sampling events. . 59

Figure 2-4 Hg-DOC relationships in 2014 and 2015 snowmelt runoff for all 24 mesocosms (n =

48 total). (a) THg in relation to DOC concentrations and (b) MeHg in relation to DOC

concentrations. Data are plotted on log-transformed axes. ........................................................... 63

Figure 3-1 PEATcosm daily TGM fluxes (in ng m-2

d-1

) across the four crossed water table

(WT) and plant functional group treatments (8 mesocosms in total monitored). * signifies that

sedge vegetation was present beneath the footprint of the DFC placed on the peat surface. ....... 84

Figure 3-2 Relationship between daily TGM fluxes (ng m-2

d-1

) and the number of sedge stems

present beneath the DFC at both the PEATcosm (closed red circles) and SPRUCE (open black

circles) sites. .................................................................................................................................. 85

Figure 3-3 PEATcosm 24-h TGM flux measurements (in ng m-2

h-1

) for the four crossed water

table and plant functional group treatments (two replicates each) in relation to surface soil

temperature (in °C) and solar radiation (kW m-2

). ........................................................................ 86

Figure 3-4 Relationship between PEATcosm hourly TGM fluxes (ng m-2

h-1

) and a) soil

temperature (in °C) and b) solar radiation (kW m-2

) among the four crossed water table and

vascular plant functional group treatments. .................................................................................. 87

Figure 3-5 Total Hg concentrations of PEATcosm vegetation a) leaf rinses expressed per unit

leaf tissue mass (ng g-1

dry wt.), b) leaf rinses per unit leaf area (ng m-2

) and c) leaf tissue mass

(ng g-1

dry wt.). Letters denote statistically similar values. No significant differences between

THg concentrations expressed on a per leaf surface area basis. ................................................... 88

xv


d-1

) in relation to a) estimates of total leaf

area (in cm2) of all Ericaceae vegetation present beneath the DFC footprint, b) estimates of THg

sorbed to the surfaces of Ericaceae leaves under each DFC and c) surface (0–10 cm) peat THg

concentrations. .............................................................................................................................. 89

Figure 3-7 SPRUCE daily Hg fluxes (in ng m-2

d-1

) for both of the two replicate DFCs across the

range of deep peat heating temperature treatment plots (+0, +4.5 and +9 °C) in May 2014,

August 2014 and June 2015. Flux data for Plots 4 and 6 in June 2015 not available. ................. 90

Figure 3-8 Relationship between SPRUCE daily TGM fluxes and mean daily surface peat

temperatures for the May 2014 (pre-warming-solid circles), August 2014 (deep warming

achieved at depth-open circles) and June 2015 (prolonged deep warming-stars) measurement

periods. Deep warming target temperature differential treatments (+0, +4.5, +9 °C) are colour-

coded in blue, orange and red, respectively. ................................................................................. 92

Figure 4-1 Mean peat MeHg and THg concentration (ng g-1

) profiles for the six crossed water

table and vascular plant functional group treatments throughout the PEATcosm experiment from

2011 to 2014 and associated mean June to October water table positions for the High and Low

WT prescriptions. For clarity, error bars for each 10 cm depth increment have been omitted. 120

Figure 4-2 Relationship between mean treatment peat MeHg concentrations (ng g-1

) at 40 cm

below the peat surface (mean of the 30-40 and 40-50 cm depth increments) and pore water

acetate concentrations at 40 cm depth. ....................................................................................... 122

Figure 4-3 Mean kmeth (d-1

) in the upper 15 cm of the peat profile and 35-50 cm below the peat

surface for the a) High WT and b) Low WT treatments. Mean water table levels for the month

prior to sampling are denoted by the dashed lines. Letters denote statistically similar depths for

each of the High and Low WT data. ........................................................................................... 123

Figure 4-4 Relationship between the fraction of THg as MeHg (%MeHg) in the peat and a) kmeth

(% d-1

), b) kdemeth (% d-1

). ............................................................................................................ 123


) among the six experimental treatments in the upper 15 cm of the peat

profile (left column) and 35-50 cm below the surface (situated beneath the water table of the

Low WT treatments) (right column). Letters denote statistically similar groups among the six

xvi

treatments in the 10-15 cm and 35-40 cm depth increments. No significant differences among

treatments for any other peat depths. .......................................................................................... 124

Figure 4-6 Soil-water partition coefficients (logKD) for MeHg at 20 and 40 cm below the peat

surface among the six crossed water table and plant functional group treatments in 2013 and

2014 (n = 4 per treatment). Letters denote statistically similar groups. .................................... 126

Figure 5-1 Relationship between Δ200

Hg, Δ201

Hg and the surface peat THg concentrations (in ng

g-1

) for the S1 peatland and PEATcosm sites. Note different scale for Δ201

Hg for S1 bog and

PEATcosm data. ......................................................................................................................... 154

Figure 5-2 Relationship between δ202

Hg values (in ‰) and the mean daily TGM fluxes (in ng m-

2 d

-1) for the a) S1 and b) PEATcosm peatlands. Note different δ

202Hg and mean daily TGM flux

scales for both plots. ................................................................................................................... 155

Figure 5-3 Relationship between a) Δ199

Hg and Δ201

Hg and b) Δ199

Hg and δ202

Hg (all expressed

in ‰) for the surface 0-10 cm peat from both the S1 and PEATcosm peatland sites. ............... 157

Figure 5-4 Δ199

Hg and Δ201

Hg values (in ‰) among the six crossed water table and plant

functional group treatments of the PEATcosm peat monoliths. Analyses were on pooled samples

from replicate treatments, thus error bars represent analytical precision, not sample replication

variability. ................................................................................................................................... 160


Hg and Δ201

Hg (expressed in ‰) for the surface 0-10 cm

peat from both the S1 and PEATcosm peatland sites. ................................................................ 162

Figure 6-1 Influence of water table lowering on the partitioning of Hg and MeHg from the solid

phase into the aqueous pore water phase, translocation down the peat profile and re-sorption to

the solid phase within the zone of lowered water table fluctuation. ........................................... 175

Figure A-2-1 Photos of mesocosm bins. .................................................................................... 184

xvii

Figure A-2-2 Runoff total Hg and MeHg loads across the six crossed water table and vascular

plant functional group treatments during the 2014 spring snowmelt period (n=4 per treatment). a)

Total Hg and b) MeHg. Letters denote statistically similar groups based on transformed data. 185

Figure A-2-3 Mean pore water MeHg concentrations (on log scale) during the five sampling

events considered for this study. a) through e) denote specifically sampling periods. Each box

represents the mean pore water concentration of all three sampling depths across the four

replicate mesocosms (n=10 - 12). ............................................................................................... 186

Figure A-2-4 Mean pore water THg concentrations (on log scale) during the five sampling


represents the mean pore water concentrations of all three sampling depths across the four

replicate mesocosms (n=10 - 12). ............................................................................................... 187

Figure A-2-5 Pore water Hg-DOC relationships. a) THg (log scale) in relation to DOC (square

root transformed) concentration, and b) MeHg (square root transformed) in relation to DOC

(square root transformed) concentration. The data includes all three pore water depths from each

of the four replicates per each of the six treatments. All five sampling events are included. ... 188

Figure A-2-6 Relationships between shallow pore water (20 and 40 cm below the peat surface)

Hg and % mass loss of cellulose decomposition assays in the top 40 cm of the peat. a) THg

concentrations, b) MeHg concentrations, c) %MeHg. Decomposition assays were harvested in

2014 and represent potential peat decomposition among the treatments. From all five sampling

events, pore water Hg data collected at 20 and 40 cm depths was averaged by mesocosm. n=18-

20 per treatment. ......................................................................................................................... 189

Figure A-2-7 Locally-weighted scatterplot smoothing (LOWESS) relationships between shallow

pore water (20 and 40 cm below the peat surface) Hg and % mass loss of cellulose

decomposition assays in the top 40 cm of the peat. a) THg concentrations, b) MeHg

concentrations, c) %MeHg. Decomposition assays were harvested in 2014 and represent

potential peat decomposition among the treatments. From all five sampling events, pore water

Hg data collected at 20 and 40 cm depths was averaged by mesocosm. n=18-20 per treatment.

..................................................................................................................................................... 190

xviii

Figure A-2-8 Pore water Hg-total phenolics relationships. a) THg (log transformed) in relation

to total phenolics (square root transformed) concentrations, and b) MeHg (square root

transformed) in relation to total phenolics (square root transformed) concentrations. The data

includes all three pore water sampling depths from each of the four replicates per each of the six

treatments. All five sampling events are included. .................................................................... 191

Figure A-2-9 Sulfate concentrations among the six crossed WT and plant functional group

treatments including all three sampling depths for the 2013 and 2014 sampling events. Letters

denote statistically similar groups. .............................................................................................. 192

Figure B-3-1 a) Relationship between mean inlet TGM concentrations (ng m-3

) and

corresponding 20-min TGM fluxes measured during daylight hours across the PEATcosm

experimental treatments. b) Boxplots of mean inlet Hg concentrations across the four PEATcosm

treatments. Letters denote statistically similar values. ............................................................... 194

Figure B-3-2 Piecewise regressions of PEATcosm hourly TGM fluxes (in ng m-2

h-1

) and

surface soil temperatures (in °C) for the a) High WT Control, b) Low WT Control, c) Low WT

Sedge, and d) Low WT Ericaceae treatments. Temperature thresholds are noted for each

treatment by the vertical dashed lines. The equations of the relationships before and after each

threshold temperature are specified. ........................................................................................... 195

Figure B-3-3 a), c) and e) Boxplots of mean inlet TGM concentrations measured during daylight

hours across the SPRUCE plots for each of the three sampling events. Letters denote statistically

similar values. b), d) and f) Relationship between mean inlet TGM concentrations (ng m-3

) and

corresponding 20-min TGM fluxes (daylight only) across the SPRUCE plots during each of the

three samplings. .......................................................................................................................... 196

Figure C-4-1 Peat MeHg and THg stock (in ng cm-3

) profiles for the six combined water table

and vascular plant functional group treatments throughout the PEATcosm experiment from 2011

to 2014 and associated mean June to October water table positions for the High and Low WT

prescriptions. For clarity, error bars for each 10 cm depth increment have been omitted. ........ 197

xix

Figure C-4-2 Mean kdemeth (d-1

) in the upper 15 cm of the peat profile and 35-50 cm below the

peat surface for the a) High WT and b) Low WT treatments. Mean water table levels for the

month prior to sampling are denoted by the dashed lines. No statistically significant differences

among the depths for each of the High and Low WT data. ........................................................ 198

Figure C-4-3 Soil-water partition coefficients (logKD) for Hg(II) at 20 and 40 cm below the peat


2014 (n = 4 per treatment). Letters denote statistically similar groups. .................................... 199

Figure D-5-1a S1 peatland of the Marcell Experimental Forest in north-central Minnesota

showing one SPRUCE plot footprint. ......................................................................................... 200

Figure D-5-1b PEATcosm Mesocosm Facility, Houghton, Michigan, with the 24 peat monoliths

(top) and underlain by a climate-controlled tunnel (bottom)…………………………………...200

Figure D-5-2 Relationship between Δ200

Hg and Δ204

Hg (all expressed in ‰) for the surface 0-10

cm peat from both the S1 and PEATcosm peatland sites. .......................................................... 202


Hg and Δ201


peat from both the S1 and PEATcosm peatland sites as compared to reported values of Hg(II)

deposition in the literature (Gratz et al. 2010; Sherman et al. 2012; Demers et al. 2013). ......... 203

xx

List of Appendices

Appendix A. Supplementary Information for Chapter 2.............................................................182

Appendix B. Supplementary Information for Chapter 3.............................................................194

Appendix C. Supplementary Information for Chapter 4.............................................................197

Appendix D. Supplementary Information for Chapter 5.............................................................200

1

Chapter 1 Introduction

1

1.1 Mercury Cycling in the Environment

1.1.1 Overview of the Mercury Cycle

Mercury (Hg) is a toxic global pollutant that poses a severe toxicological and teratological threat

to both wildlife and humans (Morel et al. 1998; Mergler et al. 2007; Scheuhammer et al. 2007).

This threat is evidenced by the extent of fish consumption advisories due to high Hg levels

across both Canada and the United States (Driscoll et al. 2007). In addition to natural sources

such as volcanic emissions, anthropogenic activities such as coal burning, artisanal gold mining

and waste incineration have contributed to global mercury pools in the atmospheric, terrestrial

and aquatic environments (Selin 2009). Mercury exists predominantly in three forms in the

environment: elemental mercury (Hg(0)), inorganic divalent mercury (Hg(II)) and organic

methylmercury (CH3Hg+; MeHg) (Zillioux et al. 1993). A potent neurotoxin, MeHg is known to

bioaccumulate in aquatic organisms and biomagnify in higher trophic levels of the food web.

Sinks, sources and fluxes involved in Hg biogeochemical cycling have been generally

established on a global scale (Mason et al. 1994). Novel techniques are emerging that are greatly

advancing the level of understanding of Hg cycling processes across a range of scales. Some of

these recent advancements include the determination of genes hgcAB involved in Hg

methylation (Parks et al. 2013; Gilmour et al. 2013), an important step in the direction of

developing a Hg gene probe, the role of nanoparticles in controlling the association of Hg with

dissolved organic matter (e.g. Graham et al. 2012) and the application of high-resolution mass

spectroscopy techniques to track natural stable isotope fractionation (e.g. Demers et al. 2013;

Kritee et al. 2013; Zheng et al. 2016) allowing further insight into the sources and processes

governing Hg cycling.

Following emission from either natural or anthropogenic sources, elemental Hg in the

atmosphere has an average residence time of 0.5 to 2 years; sufficient time for global-scale

transport from point sources to remote landscapes (Mason et al. 1994; Schroeder and Munthe

2

1998). This elemental Hg in the atmosphere may be oxidized by strong oxidants such as

halogens resulting in the production of reactive gaseous Hg(II) as well as particulate Hg(II); both

with shorter atmospheric residence times than that of elemental Hg(0) prior to deposition

(Lehnherr 2014). Atmospheric deposition, via either wet or dry deposition, represents the

dominant input of Hg pollution to watersheds (Fitzgerald et al. 1998; Driscoll et al. 2007), with a

large portion of this deposited Hg being incorporated into the soil and vegetation pools

(Hintelmann et al. 2002). However, while acting as sinks of Hg, terrestrial watersheds can also

be significant sources of MeHg to downstream aquatic ecosystems via runoff (Munthe et al.

2007; Harris et al. 2007). Wetlands as well as other aquatic environments such as lake-bottom

sediments are known to be sites of Hg methylation due to favourable redox and hydrological

conditions (Branfireun and Roulet 2002; Tjerngren et al. 2012; Figure 1-1). Deposited Hg is

subjected to numerous influences in soils following deposition which may incorporate it into the

soil and vegetation pools, including accumulation in organic peat soils (Figure 1-1) or this

deposited Hg may be subsequently re-emitted from soils to the atmosphere through reduction to

Hg(0). Inorganic Hg and MeHg may also be mobilized in pore waters, exported from peatlands

in runoff (Figure 1-1) and transported to downstream aquatic environments. In such

environments, Hg and MeHg may enter the aquatic food chain.

Figure 1-1 Dominant aspects of the mercury cycle in a peatland system (Source: C. Mitchell).

3

1.1.2 Mercury Methylation and MeHg Demethylation in Wetlands

The process of Hg methylation which transforms inorganic Hg into the organic form MeHg

occurs predominately via biological mediation. Chemical abiotic methylation of Hg is also

known to occur in sediments via biochemical pathways not involving microbial communities.

However, abiotic Hg methylation is relatively minor in terms of environmental significance as

compared to that of biological methylation (Berman and Bartha 1986). Abiotic methylation may

play an increasingly important role in certain environments such as wetlands due to the potential

role of dissolved organic matter, specifically recalcitrant humic substances, as has been

suggested by Ravichandran (2004). An important benchmark study by Compeau and Bartha

(1985) was first to identify sulfate-reducing bacteria (SRB) as the principal methylators of Hg(II)

in anoxic sediments. This critical finding was further supported by research conducted by

Gilmour et al. (1992) in which sulfate additions stimulated MeHg production while specific

inhibition of SRB using molybdate prevented methylation. Sulfate additions have also been

observed to enhance MeHg production at the wetland scale (Jeremiason et al. 2006). However,

the accumulation of sulfides may act to inhibit MeHg production by reducing the bioavailability

of Hg for methylating bacterial communities (Benoit et al. 1999). As well, research has found

that iron-reducing bacteria may play a role in Hg methylation (Warner et al. 2003; Mitchell and

Gilmour 2008). Early studies investigating the effect of sulfate amendments on the methylation

potential of SRB employed only Desulfovibrio desulfuricans in culture experiments as being

representative of all SRB species (Compeau and Bartha 1985). However, more recent studies

have revealed 19 genera of SRB, each of which have the potential to methylate Hg to varying

extents and rates (King et al. 2002), while some SRB do not methylate Hg (Benoit et al. 2003;

Gilmour et al. 2013). Despite the observation of methylation in artificial cultures, some bacteria

such as Desulfovibrio desulfuricans and Desulfobulbus propionicus are not prominent

methylators of Hg within the natural environment (Benoit et al. 2003). Similarly,

Desulfobacterium studied by King et al. (2000) required sulfate reduction to be occurring in

order to methylate Hg. In contrast, Desulfobulbus propionicus methylates Hg while growing

fermentatively (Benoit et al. 2003).

Not all sulfate- and iron-reducing bacteria have the ability to methylate Hg (Gilmour et al. 2011).

Recent novel research has determined the gene pair (hgcAB) necessary for MeHg production

(Parks et al. 2013). Further testing has confirmed that the presence of this gene pair predicts an

4

organism’s ability to methylate Hg (Gilmour et al. 2013). The study by Gilmour et al. (2013)

extends the knowledge of Hg methylators beyond the Deltaproteobacteria. In addition to sulfate-

and iron-reducers, Gilmour et al. (2013) determined that some methanogens and syntrophic,

acetogenic and fermentative Firmicutes that possess the hgcAB gene pair are capable of Hg

methylation. The experimental confirmation that other microbial communities than sulfate- and

iron-reducers can methylate Hg may have important implications for different habitats from

which the Hg-methylating strains were isolated; including northern peatlands (Gilmour et al.

2013). The linkage between Hg methylation and discrete gene markers in microbial communities

is an important step in order to develop gene probes for Hg methylation (Gilmour et al. 2013;

Parks et al. 2013).

Given the nature of Hg methylation as a reaction facilitated by anaerobic SRB and other

heterotrophic, redox-sensitive organisms, highly specific conditions are required in order for the

process to occur. The synthesis and stability of MeHg is favoured by negative redox potentials

(Compeau and Bartha 1984) and as such MeHg production is known to occur in anaerobic

environments such as lake-bottom sediments (Winfrey and Rudd 1990; Ramlal et al. 1993;

Miskimmin et al. 1992) and wetlands including peatlands (St. Louis et al. 1994; Tjerngren et al.

2012). Temperature is also a major contributing factor controlling Hg methylating activity.

Mercury methylation is limited by low temperatures; occurring optimally at a temperature of

35ºC (Winfrey and Rudd 1990). This influence of temperature on MeHg production suggests

seasonality in which methylation is highest in late summer (Winfrey and Rudd 1990), as well as

the potential for climate change-induced alteration of this transformation with increasing soil and

atmospheric temperatures. However, the influence of climate change on Hg methylation has not

been explored in the literature. The role of pH in Hg methylation is dependent upon the

conditions of the system under study. Some research reports increased methylation with

decreasing pH; while others conclude the opposite (i.e. Miskimmin et al. 1992; Ramlal et al.

1985). The pH level may play a role in the association of Hg with solid phases rendering it

unavailable for methylation; with greater partitioning to the liquid, mobile phase in acidic

conditions (Benoit et al. 2003; Haitzer et al. 2003).

Microbial communities also require an available labile carbon source in order to carry out the

methylation process (Mitchell et al. 2008). The source of available carbon could originate in

upland systems and be delivered to wetlands via runoff in natural environments (Mitchell et al.

5

2008), while enhanced organic matter decomposition in newly-flooded reservoirs may provide

this source of labile carbon (Winfrey and Rudd 1990; Heyes et al. 2000). As part of the

Experimental Lakes Area Reservoir Project (ELARP), Heyes et al. (2000) reported an increase in

net MeHg production as a result of impoundment of a boreal peatland. The observed spike in

MeHg production was attributed to the increasing areal extent of anoxia due to flooding as well

as the release of carbon and other nutrients due to increased decomposition of newly-inundated

vegetation. A field study in which MeHg production in a wetland decreased as a result of

vegetation removal further highlights the need for an organic carbon source in order for

methylation to occur (Windham-Myers et al. 2009). This study suggested that the exudates from

roots including acetate provide labile carbon for the SRB thereby stimulating Hg methylation in

the rhizosphere. Mitchell et al. (2008) amended peatland mesocosms with sulfate and an array of

carbon sources with the goal of simulating nutrient fluxes from the upland to the peatland that

would occur naturally and then assessed the resultant MeHg production. This mesocosm study

demonstrated that the addition of both a labile carbon source and sulfate in combination

stimulates the greatest net MeHg production. Ravichandran (2004) suggested that predominantly

labile and easily decomposed organic matter may stimulate microbial growth and therefore

promote Hg methylation. However, recalcitrant, refractory portions of dissolved organic matter

(including humic and fulvic acids) may contribute to abiotic Hg methylation (Ravichandran

2004).

In natural systems the concentration of MeHg present depends upon the balance between the

competing processes of methylation and demethylation. Methylmercury degradation may occur

via a number of both abiotic and biotic pathways (Marvin-Dipasquale et al. 2000). Abiotic

pathways include photodegradation of MeHg (Sellers et al. 1996) as well as the production of

dimethylmercury through a reaction with sulfides, particularly in marine waters (Jensen and

Jernelöv 1969; Mason et al. 1995; Lehnherr 2014; Jonsson et al. 2016). The most widely studied

biotic degradation pathways involve the investigation of microbes present in mercury-

contaminated systems that have developed a resistance to mercury with the evolution of genes

encoded by the mer operon (Marvin-Dipasquale et al. 2000). The mer-A gene results in the

production of the mercuric reductase enzyme which reduces inorganic Hg2+

to volatile elemental

Hg. This occurs following the cleavage of MeHg into CH4 and Hg2+

by the mer-B gene which

encodes for the organomercurial-lyase enzyme (Barkay et al. 1991; Marvin-Dipasquale et al.

6

2000; Parks et al. 2009). The relative contributions of oxidative degradation (OD) and reductive

demethylation pathways including mer-mediated and non-mer-mediated methylmercury

degradation were suggested by Marvin-Dipasquale et al. (2000) to be a function of the level of

site Hg contamination; with OD dominating in less-contaminated sites. Further research is

required to decipher the controls on demethylation and how this process influences net MeHg

production. This may be particularly important for wetland environments, as this antagonistic

process influences the amount of MeHg available for export.

1.1.3 Mercury - Organic Matter Relationships

Organic matter is important in both the transport of Hg among different watershed

compartments such as from upland soils to wetlands and water bodies (i.e. Mierle and Ingram

1991), as well as in the retention of Hg in soils (Skyllberg et al. 2000). Several studies have

found mercury fluxes during high-flow events exhibit strong correlations to dissolved organic

carbon (DOC) fluxes (i.e. Bishop et al. 1995; Dittman et al. 2010; Schuster et al. 2008). Mercury

has a tendency to complex with dissolved organic matter (DOM) (e.g. Driscoll et al. 1995;

Schuster 1991) and as a result Hg mobilization may be enhanced during high-flow events due to

this complexation.

The extent of Hg binding to organic matter depends upon the available functional groups. Thiol

and other reduced sulfur groups (including those with either one oxygen or nitrogen atom) in

humic and fulvic acids in both soil and fresh waters have been observed to exhibit high affinity

for inorganic mercury (Hg2+

) as well as MeHg (Hintelmann et al. 1997; Karlsson and Skyllberg

2003; Skyllberg et al. 2000). It has been suggested that the affinity of Hg to complex with DOC

may act to limit its bioavailability for transformation processes such as Hg methylation

facilitated by microbial communities capable of performing this reaction (Graham et al. 2012;

2013). Graham et al. (2012) highlights the potential role of metal nanoparticles, such as HgS,

from preventing the binding of Hg with organic matter in sulfidic environments. By inhibiting

binding with organic matter these small HgS particles remain bioavailable to Hg methylating

bacteria (Graham et al. 2012). The presence of such particles may account for the propensity of

organic-rich environments such as wetlands and lake sediments to facilitate the production of

MeHg.

7

1.1.4 Soil-Air Hg Exchange

Depending upon the speciation of atmospheric Hg (elemental Hg(0), reactive gaseous Hg and

particulate-bound ionic Hg(II)), residence times of Hg in the atmosphere may range on the order

of hours to days for particulate-bound and reactive gaseous Hg to between six months to two

years for Hg(0) (Schroeder and Munthe 1998; Driscoll et al. 2013). These variable atmospheric

residence times account for both observed hot spots of Hg contamination near emission sources

as well as the propensity of Hg to be a global pollutant (Lindberg et al. 2007). Mercury can be

deposited from the atmosphere by both wet and dry processes. Local or regional deposition via

both wet and dry depositional processes is most common with both particulate and reactive Hg

forms whereas elemental Hg may be transported long distances prior to being either deposited

directly via dry deposition or oxidized and deposited through either wet or dry mechanisms to

the soil or water surface. Wet deposition involves scavenging of gas-phase and aerosol-phase Hg

by precipitation and is delivered to the soil surface (Zhang et al. 2009). In addition to the direct

net flux of gaseous Hg to the soil surface, dry deposition may involve plant uptake of

atmospheric Hg followed by deposition to soil via litterfall or deposition of Hg to vegetation

surfaces and delivery to soil via washing of foliage during precipitation events (Rea et al. 2000;

Laacouri et al. 2013; Driscoll et al. 2013).

Soils may act both as a sink or source of Hg to the atmosphere. Once deposited, Hg enters

various soil pools and undergoes numerous reactions which affect the speciation and therefore

determine the fate of deposited Hg. Following deposition, Hg particularly in the divalent form,

has the potential to be retained in the soil due to its ability to bind to organic and mineral particle

surfaces (Gabriel and Williamson 2004). However, both biotic and abiotic processes can act to

reduce this soil-bound Hg for re-emission to the atmosphere (e.g. Xin et al. 2007; Lin et al.

2010). Adsorption to organic, particularly humic, material and to a lesser extent inorganic

material may be important in indirectly controlling the potential reduction and re-emission of

deposited Hg. The strong tendency of Hg to bind to reduced sulfur functional groups (e.g. thiol

groups) in organic matter (Graham et al. 2012; Khwaja et al. 2006; Ravichandran 2004;

Skyllberg et al. 2006) may act to reduce the pool of Hg in soil available for re-emission.

However, processes which enhance desorption may counteract such adsorption and shift the

balance of whether a soil acts as a source or sink of Hg.

8

Both biotic and abiotic processes have been observed to influence desorption and subsequent

emission from a variety of soils (Zhang and Lindberg 1999). Biologically-mediated reduction of

Hg(II) may act to increase Hg(0) evasion from soils. Soil temperature has been observed to be a

significant control on Hg fluxes to the atmosphere with enhanced fluxes from warmer soils

(Edwards et al. 2001), as increased temperatures effectively lower the activation energy required

for Hg reduction (Carpi and Lindberg 1998). As well, ultraviolet radiation is a strong influence

on the enhancement of evasion from soils (Carpi and Lindberg 1998; Xin et al. 2007). Strong

diel variations in Hg fluxes are often observed with enhanced evasion during midday and

diminished evasion or net uptake reported during the overnight hours. Soil moisture may also

play an important role in influencing the adsorption and desorption from soil particle surfaces

(Gustin and Stamenkovic 2005; Song and Van Heyst 2005). Soils, particularly mineral soils,

often show a greater affinity for water molecules than Hg. As such, adsorbed Hg may be

dislodged from soil surfaces and emitted to the atmosphere. This process has been observed with

enhanced Hg fluxes following precipitation events (Song and Van Heyst 2005), which

subsequently diminish as Hg diffusion is suppressed in saturated soils (Bahlmann et al. 2004;

Gustin and Stamenkovic 2005; Song and Van Heyst 2005). This concept requires further

investigation as it has been suggested that antecedent soil moisture conditions may be a

contributing influence (Gustin and Stamenkovic 2005). Other factors which may influence the

reduction of divalent Hg in terrestrial environments and therefore play a role in controlling

emissions include soil organic matter content and soil Hg concentrations (Lin et al. 2010). The

individual controls of certain processes influencing Hg evasion from soils warrant further

investigation particularly with controlled experimental treatments or at the laboratory scale as

possible synergisms may exist between the factors governing soil-air Hg fluxes.

Recent research has examined atmospheric Hg fluxes from forest and wetland ecosystems

including northern boreal peatlands using both chamber methods (Kyllönen et al. 2012; Fritsche

et al. 2014) and relaxed eddy accumulation techniques (Osterwalder et al. 2016). In order to

effectively model the potential global Hg sources to the atmosphere and how the evasion

processes may be impacted by climatic and other environmental changes, further monitoring of

soil-air Hg fluxes is warranted across a range of environments. Mercury fluxes and the processes

that govern Hg evasion from climate change-sensitive northern peatlands are understudied at the

9

present time despite the large stocks of Hg stored in these systems that represent a potentially

significant source of Hg to the atmosphere (Kolka et al. 2001; Grigal 2003).

1.1.5 Mercury Stable Isotope Fractionation

The fractionation of stable isotopes of light elements such as carbon, oxygen and nitrogen has

been applied as a means of tracing the transformations involved in the biogeochemical cycles of

these elements. With the advent of multiple collector inductively coupled plasma mass

spectrometry (MC-ICP-MS), the natural fractionation of stable isotopes of heavier elements can

be utilized to elucidate the processes controlling the cycling of Hg through the environment

(Bergquist and Blum 2007; Yin et al. 2010). Mercury has seven stable isotopes: 196

Hg, 198

Hg,

199Hg,

200Hg,

201Hg,

202Hg and

204Hg with a mass difference of 4% (Bergquist and Blum 2009).

Mercury also has an active redox chemistry, a volatile form (Hg(0)), and a tendency to form

covalent bonds thereby facilitating ample opportunities for isotopic fractionation to occur

(Bergquist and Blum 2007). Recent and on-going research is investigating the natural

fractionation of Hg stable isotopes in order to trace Hg sources (due to anthropogenic pollution,

for example) and pathways involved in Hg biogeochemical cycling (i.e. Hg methylation, Hg2+

reduction to elemental Hg) (Yin et al. 2010; Donovan et al. 2014). Atmospheric Hg reaction

mechanisms are an area of study to which Hg isotopic analyses, investigating the atmospheric

processes that fractionate Hg, are making significant strides (e.g. Gratz et al. 2010; Chen et al.

2012; Sun et al. 2016).

Mercury isotopes are useful in tracking and characterizing Hg transformations due to the

significant variations observed in the isotope ratios in the natural environment. In studying these

isotopic variations of Hg in nature it is difficult to discern the signature of the source of Hg from

that resulting from an array of processes which transform Hg from one species to another

(Bergquist and Blum 2009). Several processes, both biotic and abiotic, have been found to cause

mass-dependent fractionation (MDF), while mass-independent fractionation (MIF) has been

determined for only a limited number of processes such as photochemical reduction (Bergquist

and Blum 2007; Bergquist and Blum 2009; Zheng et al. 2016). It is important to understand the

mechanisms that govern the natural fractionation of Hg isotopes in order to accurately trace Hg

fluxes through the environment.

10

Recent studies have begun to investigate Hg isotopic signatures in peat soils (Jiskra et al. 2015;

Enrico et al. 2016). These studies have taken an important first step in modelling relative Hg(0)

and Hg(II) contributions to peatlands as well as discerning the mechanisms controlling peat Hg

cycling including non-photochemical abiotic reduction of Hg facilitated by natural organic

matter reduction as well as photochemical reduction (Jiskra et al. 2015). Further research

applying Hg isotopic analyses in peatland ecosystems is required to better understand the sources

of Hg deposition to these systems. Isotopic analyses may also provide insight into how Hg

cycling in peatlands may be affected in a changing climate.

1.2 Peatland Hydrology and Hydrogeology

Peatland ecosystems are predominantly located in boreal and subarctic regions globally and

cover approximately 12% of the Canadian landscape (Tarnocai 2009). Through their unique

biogeochemical and hydrological characteristics these ecosystems play an important role in

global climate; acting as seemingly perpetual sinks of atmospheric carbon dioxide (CO2)

throughout the Holocene epoch via accumulation of immense quantities of organic matter

(Gorham 1991). Accumulation of organic matter in thick peat soils occurs in peatlands due to a

net imbalance between plant net primary production and microbial decomposition. Hydrological

processes are a significant control on maintaining peatland water tables and therefore the delicate

balance between carbon accumulation through the maintenance of anoxia and microbial

mineralization (Whittington and Price 2006). The slow rates of organic matter decomposition are

primarily attributed to waterlogged, anoxic soils, cold temperatures and chemically-recalcitrant

organic matter substrate for microbial decomposers (Moore and Basiliko 2006). Despite their

small global land area, covering only approximately 3%, these ecosystems store between 10%

(Gorham 1991) to 30% (Turunen et al. 2002) of the global soil carbon pool. Northern peatlands

generally comprise a spectrum from nutrient-poor, ombrotrophic bogs to rich, minerotrophic fens

(Glaser et al. 1997). Ombrotrophic bogs are generally isolated from groundwater inputs as well

as from significant surface water inputs. As a result, the hydrology of these systems is

predominantly controlled by the balance between precipitation and evapotranspiration. In

contrast, minerotrophic fens are connected to groundwater inputs and consequently receive

mineral-rich groundwater resulting in higher pH values than bog systems (Glaser et al. 1997). As

11

a result of the disparity between bogs and fens in terms of hydrological inputs and nutrient status,

the vegetation communities present in these ecosystems differ with fens dominated by vascular

plant communities (such as sedges) while acidophilic Sphagnum mosses are predominant in bogs

(Glaser et al. 1997).

The structure of accumulated peat in these systems plays an important role in controlling

flowpaths through peatlands. The upper peat layers, known as the acrotelm, consist of fresh,

minimally-decomposed organic matter and have higher hydraulic conductivities than that of the

deeper peat, known as the catotelm. The catotelm consists of well-humified organic material

with low porosity and hydraulic conductivity (Clymo 1984). Due to the highly decomposed and

compressed peat in the deeper catotelm layer, the majority of subsurface flow is considered to

occur through the acrotelm. However, research by Chason and Siegel (1986) suggests that the

connection between surface water and groundwater may be significant. Field and laboratory

measurements of vertical and horizontal conductivity demonstrated the ability of peat at various

stages of humification to transmit groundwater quickly (Chason and Siegel 1986). Such a result

calls into question the idea that well-humified peat does not effectively transmit groundwater due

to low hydraulic conductivity. In large patterned peatland systems such as the Glacial Lake

Agassiz Peatlands in Minnesota, groundwater contributions have been observed to play an

important role in the water balance of the ecosystem as well as the pore water chemistry and

consequently the dominant vegetation communities (i.e. Siegel and Glaser 1987).

Although peatlands are important zones of groundwater recharge in the landscape they are also

considerable discharge zones (i.e. Siegel and Glaser 1987; Siegel et al. 1995). By monitoring

groundwater observation wells during the winter in a number of peatlands, Nichols and Verry

(2001) calculated the rate of water table decline without any additional recharge due to

precipitation. Deviations from these background winter water table recession rates facilitate the

determination of groundwater recharge throughout the rest of the year. Nichols and Verry (2001)

concluded that approximately 40% of total water yield in these watersheds was the result of

groundwater recharge and that this recharge varied linearly with precipitation. Nichols and Verry

(2001) suggested that the water table depths in the S1 peatland did not exhibit a predictable

decline during the winter season such as those recorded for nearby watersheds in the Marcell

Experimental Forest because S1 is an area of groundwater discharge. In some cases groundwater

flow reversals have been observed to occur as a result of drought conditions; resulting in

12

groundwater upwelling due to a change in hydraulic gradient (i.e. Romanowicz et al. 1993;

Siegel et al. 1995).

1.3 Peatland Ecosystems and Global Climate Change

Global climate change has the potential to significantly impact the hydrology of mid- to high-

latitude northern peatland ecosystems due to increasing temperatures and changes in

precipitation patterns (Bridgham et al. 2008). Within Canada, peatlands in the Boreal and

Subarctic regions are forecasted to exhibit severe to extremely severe sensitivity to climatic

changes (Tarnocai 2006). The susceptibility of peatlands to climatic changes may depend upon

the relative contributions of groundwater inputs versus those of precipitation; with precipitation-

dominated wetlands exhibiting particular vulnerability to climate change (Winter 2000).

Climate-induced alterations in hydrology may have subsequent impacts on vegetation

communities present across peatland types as a result of fluctuating water tables (Weltzin et al.

2000; Strack et al. 2006). Soil microbial communities present in these peatland ecosystems may

in turn be affected as a result of the hydrological and redox conditions as well as potential

changes in available organic matter substrate for decomposition (Peltoniemi et al. 2009). Due to

the coupled relationship between the cycling of Hg and DOC (Driscoll et al. 1995; Skyllberg et

al. 2000), the cycling of Hg and transport in peatlands will likely be impacted by alterations in

carbon cycling and redox conditions resulting from climate-induced changes in hydrological

processes.

1.3.1 Global Climate Change and Peatland Hydrological Processes

At mid- to high-latitudes in Canada, mean atmospheric temperatures are forecasted to increase 3-

4°C by 2020 and 5-10°C by 2050; relative to the 1960-1990 period (Tarnocai 2009). These

predicted temperature increases in conjunction with altered precipitation patterns with potentially

less frequent, more intense events (Groisman et al. 2012; Janssen et al. 2014; Yu et al. 2016)

may affect peatland water tables. Evapotranspiration may significantly increase as a result of

enhanced temperatures. Consequently, peatland water tables are predicted to decline due to

enhanced water losses via evapotranspiration exceeding precipitation inputs; particularly in

13

ombrotrophic bog systems due to the dominance of precipitation in maintaining water levels

(Winter 2000). Depending upon the severity and duration of water table drawdown, the carbon

storage function of peatlands may be lost as a result of feedback mechanisms facilitating the

release of carbon to the atmosphere as CO2 due to enhanced peat aeration (Ise et al. 2008),

becoming weaker carbon sinks under drier conditions (Chimner et al. 2016). This decline in

peatland water tables due to enhanced evapotranspiration may be further exacerbated by the

predicted shift in vegetation communities present in peatlands from Sphagnum- to vascular-

dominated as a result of prolonged climatic changes. Modelling of evapotranspiration by

Admiral and Lafleur (2007) suggested that vascular plants contribute the majority of the

observed total latent heat flux (~60-80%) when compared to both hummock and hollow

Sphagnum mosses. However, Sphagnum moss present in peatlands may be able to control energy

and water fluxes resulting in a potential cooling effect (Blok et al. 2011), through their ability to

blanch during periods of drought which in turn increases their albedo. The amount of energy

available for evapotranspiration by vascular plants may be affected and therefore Sphagnum

moss may provide a self-regulating feedback mechanism in the maintenance of higher water

table levels. However, this resiliency mechanism may be lost if the changes in peatland

hydrology result in the successful recruitment of vascular plants at the expense of moss species

(Waddington et al. 2015).

Climate change may also dramatically impact both surface and groundwater flowpaths and rates

in peatland ecosystems as a result of alterations in important hydraulic parameters which govern

the nature of hydrological processes and subsequent feedback mechanisms. With deeper water

tables as a result of climate change-induced hydrological variations, the character of the peat

may change due to effects on the bulk density, hydraulic conductivity and specific yield

(Waddington et al. 2015; Whittington and Price 2006). Due to its compressibility and large

potential capacity for water storage, long-term or seasonal changes in peatland water table

position may affect the storage capacity of peatlands. As the water pressure in the surface peat

soils decreases due to water table recession, subsidence occurs as a result of enhanced

compression and decomposition due to increased oxidation (Whittington and Price 2006);

potentially providing a shallower relative water table. Whittington and Price (2006) observed a

significant increase in peat bulk density as a result of drawdown-induced peat compression. This

result was observed with a concurrent decrease in hydraulic conductivity across the studied

14

peatland microforms (ridge, mat, lawn). These changes in hydraulic parameters of peat may

consequently affect the magnitude, direction and rates of groundwater and surface water flows

(Siegel et al. 1995; Ferone and Devito 2004). Peatlands that experience prolonged periods of

drawdown also exhibit considerable fluctuations in water table position upon precipitation inputs

(Whittington and Price 2006). As peat becomes more compressed due to water table drawdown it

loses its ability to swell and subside with changes in water table position as the reduction of pore

water pressure results in irreversible compression (Shantz and Price 2006). Consequently, water

table variations are amplified relative to the surface due to this increase in peat rigidity

(Waddington et al. 2015; Whittington and Price 2006). These amplifications in peatland water

table and the increase in episodic periods of peat aeration may have subsequent impacts on

oxidation-reduction reactions that are controlled by water table fluctuations (such as carbon

mineralization and Hg methylation) as well as the mobility and transport of dissolved organic

matter (Fenner et al. 2007) and contaminants (including inorganic Hg and MeHg) from peatlands

to downstream aquatic ecosystems.

1.3.2 Peatland Ecological Effects due to Climate Change

In addition to the potential effects on hydrological parameters and processes, climate-induced

changes in hydrology may also significantly impact the ecology of peatland ecosystems with

regard to vegetation community composition and structure as well as the types and activity of

microbial populations responsible for organic matter decomposition. Although the potential

development of shallower relative water tables as a result of peat decomposition and

compression is possible with climate change, the dramatic fluctuations in water levels with

precipitation inputs following prolonged dry periods may exert considerable water stress on

Sphagnum mosses (Weltzin et al. 2000; Waddington et al. 2015). Vascular plants may have a

competitive advantage over Sphagnum moss species in such situations of prolonged water stress

given their physiological attributes (Chimner et al. 2016). Due to their vascular structures with

extensive rooting systems, vascular plants can access deeper stores of water whereas bryophytes

must be in close contact with water for survival. The rooting systems of vascular plants such as

sedges may also act to shuttle oxygen to the rhizosphere, increasing the zone of aeration, and act

as a conduit for gaseous carbon emissions providing a positive feedback mechanism to enhanced

15

water table recession (Bridgham et al. 2008; Strack et al. 2006; Waddington et al. 2015; Weltzin

et al. 2000). Although mosses may be able to regulate available water for vascular plants by

energy balance feedback mechanisms, increased shading by vascular plants over Sphagnum

mosses may further exacerbate the stress placed upon moss species and reduce their resilience

(Turetsky et al. 2012). The timing and coverage of such changes in vegetation communities may

be governed by spatial heterogeneity as a result of natural peatland microforms such as

hummocks, lawns and hollows (Strack et al. 2006). In currently wet pool locations in peatlands it

has been hypothesized that the forecasted drawdown in water table levels may act to increase

carbon accumulation in these microforms over time as a result of ecological succession and

colonization (Strack et al. 2006). Fenner et al. (2007) observed that increasing temperatures and

enhanced atmospheric CO2 concentrations exhibited an additive effect on vascular plant

dominance as a result of water table drawdown and CO2 fertilization which favoured the

colonization and growth of vascular plants due to their ability to access deeper water stores

during periods of water stress as compared to Sphagnum species. Both graminoids, including

sedges, and Ericaceae shrubs may increase in productivity and abundance under drier conditions

depending upon the level of water stress (e.g. Dieleman et al. 2015; Potvin et al. 2015).

Succession of plant communities may also play an important role in carbon biogeochemical and

hydrological processes under future climate scenarios. Upon senescence and through root

exudation, vascular plants may provide more labile organic matter to the peat surface and

subsurface as compared to Sphagnum plant material; which is more chemically-recalcitrant for

microbial mineralization (Scheffer et al. 2001; Haynes et al. 2015). The resiliency mechanisms

inherent in peatland ecosystems which act to maintain slow rates of carbon mineralization such

as the recalcitrant Sphagnum-derived organic matter substrate and the anoxic soils resulting from

high water tables may be significantly impacted if there were to be a shift in the dominant

peatland plant communities from Sphagnum-to vascular-dominated. In addition to potentially

enhancing fluxes of CO2 to the atmosphere, the delivery of organic matter derived from vascular

plants to the underlying peat may alter the hydrological characteristics of the peat over time.

Changes in vegetation type over the long-term may affect seasonal water storage as a result of

the reduced compressibility of peat containing woody roots or rhizomes (Whittington et al.

2007). The reduced compressibility of vascular-derived peat may therefore also yield higher

hydraulic conductivities as compared to Sphagnum-derived peat (Limpens et al. 2008).

16

Consequently, hydrological flowpaths may be affected and flow rates increased, which may in

turn impact the redox conditions through enhanced aeration. As a result, these hydrological

changes could have important positive feedbacks on nutrient and contaminant mobility and

transport from these systems as well as impacting transformations of these constituents within

the peat by potentially reducing the residence time of nutrients and contaminants in the peat

soils.

Climate-induced changes in peatland hydrology significantly impact microbial community

structure, which subsequently affects peat decomposition and the ability of peatlands to sequester

carbon (Nunes et al. 2015; Peltoniemi et al. 2015). Microbial (bacterial, fungal and archaeal)

populations present in peatlands may also be altered due to the organic matter quality of the

substrate in both the surface and subsurface. Changes in temperature and ambient CO2

concentrations may govern the activity and community composition of microbial populations

given their dependence on temperature for optimal functioning. Peltoniemi et al. (2009)

determined that over the long-term, fungal and actinobacterial communities become more similar

between peatland types of different nutrient status and hydrological conditions. Actinobacteria

appear to be less sensitive to hydrological changes, such as water table drawdown (Peltoniemi et

al. 2009). With regard to the mineralization of organic matter substrate, there may be minimal

change in microbial usage of different quality substrates due to the potential observed pattern of

functional redundancy (Strickland and Rousk 2010), suggesting no change in carbon

mineralization and CO2 production with changing vegetation across peatland types. Recent

research suggests that microbial communities can rapidly adapt to utilizing the available organic

matter substrate (Haynes et al. 2015).

The shift toward the dominance of vascular plants at the expense of Sphagnum in peatlands over

time due to climate change may lead to enhanced litter decomposition, increased heterotrophic

respiration and greater hydraulic conductivity of the upper surface peat layers (Limpens et al.

2008). The increase in vascular plants is expected to enhance carbon mineralization through the

provision of a more labile carbon source, as compared to Sphagnum, to the peat surface and to

the subsurface through root exudates (Ise et al. 2008; Waddington et al. 2015). Increasing

microbial activity with the concurrent increasing temperatures and greater peat aeration through

water table fluctuations will also contribute to this greater expected carbon turnover. The loss of

negative feedback mechanisms that facilitate the carbon-accumulating nature of peatland

17

ecosystems may be both the direct and indirect results of the impacts of a changing climate on

the hydrological processes and ecological function of these systems (Limpens et al. 2008;

Waddington et al. 2015).

1.3.3 Potential Impacts on Mercury Cycling

The hydrological and redox conditions of peatlands facilitate the production of MeHg (Tjerngren

et al. 2012). In concert with the potential hydro-climatic impacts of global climate change, the

process of Hg methylation and the export of MeHg to downstream aquatic ecosystems may be

significantly impacted (Krabbenhoft and Sunderland 2013). Multiple synergistic and antagonistic

processes impacted by a changing climate may differentially affect Hg cycling in peatlands

(Figure 1-2). Therefore, forecasting the potential direction and magnitude of such hydrological,

biogeochemical and ecological changes on Hg cycling is challenging, but warranted given that

peatlands accumulate on the order of 7 to 44 µg total Hg m-2

y-1

including litterfall (Kolka et al.

2011).

18

Figure 1-2 Potential synergistic and antagonistic implications of climate-induced peatland

hydrologic changes and hydraulic parameters (white boxes) and their potential impacts on Hg

cycling and transport (grey boxes). Solid arrows represent direct effects and dashed arrows

represent indirect or potential effects. (Mercury implications added to climate-induced peatland

hydrological changes that are modelled after Whittington and Price 2006).

Increased water table drawdown has the potential to reduce the zone of suitable anoxic

conditions for Hg methylation as a result of increased peat aeration. However, self-regulating

negative feedback mechanisms which alter peat hydraulic properties to minimize the impact of

external forces on peatlands, such as peat surface subsidence, may act to lessen the degree of

water table recession (Waddington et al. 2015) and therefore the zone of aerated peat. Lowered,

fluctuating water tables may in fact stimulate MeHg production due to the regeneration of

terminal electron acceptors, such as sulfate, required for Hg methylating microbial communities

(Coleman Wasik et al. 2015). Increased atmospheric and soil temperatures may also facilitate

greater production of MeHg in redox-suitable environments within the peat through the

stimulation of microbial activity (Benoit et al. 2003). The increasing contribution of vascular

plant material to the peat surface and subsurface over time due to a shift in plant communities

19

may act to increase MeHg production by providing a more labile source of carbon for microbial

methylators; which is required for methylation to occur (Mitchell et al. 2008). The increasing

prevalence of vascular vegetation may also positively feedback to cause further water table

recession as a result of oxygen transport to the rhizosphere through vascular tissues and

enhanced evapotranspiration (Waddington et al. 2015). Vascular plants with aerenchymous

tissues, such as sedges, may also prime aerobic organic matter decomposition as well as

anaerobic respiration by regenerating terminal electron acceptors via root oxygen loss (Mueller

et al. 2016) and in turn enhance MeHg production. Although likely to be strongly influenced by

the potential loss or interrupted periods of anoxia, these changes in the physical and chemical

conditions of the peat may be important in influencing the manner in which the process of Hg

methylation is impacted. Another consideration, although minimally explored in the literature, is

whether bacterial communities capable of Hg methylation may be impacted by the anticipated

climatic changes of increased temperatures and enhanced CO2 concentrations. New, cutting-

edge research in the field of microbial ecology by Parks et al. (2013) and Gilmour et al. (2013)

signals the potential future development of gene probe techniques to target Hg methylating

bacterial activity.

Greater fluctuations in water table levels due to changes in the frequency and intensity of

precipitation events as well as those resulting from climate-induced alterations in the hydraulic

characteristics of the peat, (e.g. hydraulic conductivity, bulk density and storage volume)

(Whittington and Price 2006) may considerably affect the mobility and transport of Hg and

MeHg to receiving systems downstream of peatland-dominated watersheds. Mercury mobility is

known to be enhanced from initially drier soils upon input of large precipitation events in both

terrestrial upland (Haynes and Mitchell 2015) and peatland environments (Grondin et al. 1995;

Gustin et al. 2006). Given the tightly coupled relationship between Hg and DOC (Driscoll et al.

1995; Skyllberg et al. 2000), Hg fluxes may be enhanced due to the increased export of DOC

(Fenner et al. 2007). Greater amplitude fluctuating water tables and the increase in episodic

periods of peat aeration may promote greater decomposition of organic matter (Ise et al. 2008;

Whittington and Price 2006). More labile organic matter as a result of enhanced decomposition

as well as that delivered from vascular root exudates may facilitate Hg partitioning from the solid

to dissolved phase. Collectively, these influences may lead to enhanced mobility and flushing of

Hg from peatlands (Figure 1-2).

20

Increasing temperatures and fluctuating water tables may act to increase soil-air Hg emissions

from peatlands due to increased soil drying, as has been observed from mineral soils (Gustin and

Stamenkovic 2005). Enhanced organic matter decomposition may also contribute to greater

exchange of Hg from the soil surface to the atmosphere (Figure 1-2). Shifting vegetation

communities may be instrumental in promoting Hg fluxes from the soil surface due to the

influence of vascular physiology. However, it is possible that the stomata of the vascular

vegetation, as well as non-stomatal mechanisms such as sorption to leaf surfaces, may influence

the exchange of Hg with the atmosphere (Lee et al., 2000; Rea et al., 2002; Ericksen et al., 2003;

Marsik et al., 2005; Stamenkovic and Gustin, 2009; Laacouri et al., 2013). Stomatal uptake and

other non-stomatal pathways may represent a competing mechanism resulting in potentially

greater Hg deposition from the atmosphere.

Climate change-induced hydrological impacts and the anticipated corresponding shifts in plant

community composition could significantly affect all aspects of the Hg cycle in peatlands;

including MeHg production, soil-air Hg evasion as well as Hg transport and mobility. Both the

hydrological and subsequent ecological impacts could influence peatland Hg cycling in different,

and potentially competing, manners in terms of magnitude and direction. Despite the

vulnerability of peatland ecosystems to climatic change and the significant stores of Hg in the

accumulated peat soils, very little field or experimental research in the literature is focused upon

the consequences of these changes on Hg cycling. This research will contribute in beginning to

close this considerable gap in the literature.

1.4 Objectives, Research Questions and Study Sites

The overall objective of this research is to assess the potential impacts of climate change;

specifically water table drawdown, shifting plant functional groups to more vascular-dominated

communities and increasing soil temperatures on Hg cycling in peatlands through mesocosm-

and field-scale experimental research. To address this objective, research was conducted at two

peatland sites at the southern extent of the boreal ecosystem. The primary location was the

Mesocosm Facility, termed PEATcosm (Peatland Experiment at the Houghton Mesocosm),

established in Houghton, Michigan by the United States Department of Agriculture (USDA)

Forest Service in association with Michigan Technological University. Research was also

21

conducted in a spruce-Sphagnum bog (locally known as S1) at the Marcell Experimental Forest

in north-central Minnesota, which is now part of the SPRUCE (Spruce and Peatland Responses

Under Climatic and Environmental change) experiment; funded and managed by Oak Ridge

National Laboratory (ORNL), the United States Department of Energy, Office of Science,

Biological and Environmental Research and the USDA Forest Service.

In this thesis, I set out to address the following questions/objectives:

1) How will the accumulation of MeHg and inorganic Hg, within peatland pore waters and

export from peatlands during spring snowmelt, be affected by the combined effects of, and

interaction between, altered plant functional groups and greater amplitude fluctuating water

tables as are anticipated with climate change?

2) Characterize total gaseous Hg fluxes in an ombrotrophic bog peatland. How will soil-air Hg

fluxes in a peatland be impacted by deep (~2 m) peat warming across a range of temperatures up

to +9°C?

3) How will soil-air Hg fluxes be affected by altered plant functional groups and lowered

peatland water tables as are anticipated with climate change?

4) What role will different plant functional groups play in controlling the direction and

magnitude of total gaseous Hg (TGM) exchange between peatlands and the atmosphere?

5) How will solid peat MeHg and inorganic Hg concentrations, throughout the peat profile, be

impacted by lowered water tables, altered plant functional groups as well as the interactive

effects of such hydrological and ecological changes?

6) How will climate change-induced fluctuations in water table regime and the resultant shift in

vascular vegetation impact peat Hg methylation and MeHg demethylation potentials?

7) Characterize and compare the Hg isotopic signatures in surface peat from two sub-boreal

peatlands.

8) What processes govern Hg deposition and re-emission in two contrasting sub-boreal

peatlands?

22

Chapter 2 addresses Question #1.

Chapter 3 addresses Questions #2, 3, 4.

Chapter 4 addresses Questions #5, 6.

Chapter 5 addresses Questions #7, 8.

All four research chapters are based on data collected from the PEATcosm climate change

experiment. The PEATcosm experiment involved manipulations of water table position and

vascular plant communities to simulate anticipated climate change impacts on twenty-four cubic

metre (1 m x 1 m x 1 m) peat monoliths. The peat was harvested from a bog peatland in

Meadowlands, Minnesota in May 2010 and placed in Teflon-lined stainless steel mesocosm bins

in Houghton, Michigan (Figure 1-3). Each bin was insulated and the Mesocosm Facility was

underlain by a climate-controlled tunnel which facilitated the maintenance of natural peat

temperature profiles (Figure 1-4).

Figure 1-3 PEATcosm Mesocosm Facility in Houghton, Michigan with 24 cubic metre peat

mesocosm bins.

23

Figure 1-4 Climate-controlled tunnel beneath Mesocosm Facility allowing access to the 24

mesocosm bins. Each bin was equipped with overflow tubing and collection bin to allow for

runoff sampling from each mesocosm, particularly during the snowmelt period.

The PEATcosm study comprises a full-factorial experimental design with two water table (WT)

prescriptions crossed with three plant functional group treatments in a randomized complete

block design with four replicates per treatment combination (24 experimental units). The two

WT treatments include:

1) low variability, relatively high WT position (referred to as ‘high WT’ and used as the

control)

2) high variability, low WT position (referred to as ‘low WT’).

The water table treatments were based on long-term (approximately 50 years) data from the

Marcell Experimental Forest, with the two target water table seasonal profiles modelled after

typical low variability, average WT years (‘high WT’) and typical high variability, low water

table years (‘low WT’). These target WT profiles were maintained through a combination of

artificial precipitation additions (formulated to mimic rainwater chemistry at the nearby National

Atmospheric Deposition Program monitoring site at Chassell, MI), rain-exclusion covers and

regulated outflow from each of the bins during the spring snowmelt periods from the

24

approximate acrotelm-catotelm boundary. A piezometer nest was installed in each of the 24 bins

to allow for water sampling at three depths – 20, 40 and 70 cm below the peat surface.

The three plant functional group treatments were as follows:

1) all Ericaceae removed (referred to as ‘sedge only’ treatment)

2) all sedge removed (‘Ericaceae only’ treatment)

3) both sedge and Ericaceae present (Unmanipulated Control)

Vegetation manipulations were achieved and maintained by clipping stems of the excluded

species as deeply into the peat as possible while attempting to minimize peat disturbance. The

PEATcosm vascular plant functional group manipulations were initiated in 2011 and water table

experimental treatments were implemented in 2012. The experiment concluded in 2015 with the

destructive harvest of the peat monoliths.

Chapters 3 and 5 also include work conducted at the SPRUCE experiment in the Marcell

Experimental Forest in north-central Minnesota (Figure 1-5). During this research, an array of

large open-top enclosures, approximately 12 m in diameter, were installed in the S1 peatland

(Figure 1-5). In June 2014 through to August 2015, a preliminary study of deep soil warming,

approximately 2 m below the peat surface, was conducted within each enclosure plot.

Temperatures were manipulated in ten enclosures according to a regression design (duplicates of

+0, +2.25, +4.5, +6.75, +9°C above ambient temperatures). This deep warming experiment was

a pre-cursor to the full SPRUCE experiment, which now involves whole-ecosystem warming and

elevated atmospheric CO2 concentrations to simulate climate change within the bog.

25

Figure 1-5 Aerial view (left) of the SPRUCE plot footprints and boardwalks in the S1 bog,

Marcell Experimental Forest, Minnesota (September 2013 aerial photo from Paul Hanson, Oak

Ridge National Laboratory (ORNL)). One SPRUCE experimental plot (right).

1.5 Thesis Structure and Publication Information

1.5.1 Chapter 1

Chapter 1 provides background information on the aspects of the mercury cycle investigated in

this thesis in relation to climate change impacts. A brief overview on peatland hydrology is

given and the anticipated effects of climate change on peatland hydrology and ecology are

discussed. The potential implications of these climate-induced changes on mercury cycling are

hypothesized. Information on the publication status of each of the thesis chapters is also

provided.

Contributions: Written by K.M.H. with comprehensive review by C.P.J.M.

1.5.2 Chapter 2

Chapter 2 focuses on the anticipated climate change impacts of lowered water tables and shifting

vascular plant functional groups on Hg and MeHg mobility in peat pore waters and transport in

snowmelt runoff. This paper is currently in press at Global Biogeochemical Cycles (31, DOI:

10.1002/2016GB005471). The co-authors on this paper are Evan S. Kane (USDA Forest Service

Northern Research Station Houghton, MI and Michigan Technological University), Lynette

26

Potvin, Erik A. Lilleskov (USDA Forest Service Northern Research Station Houghton, MI),

Randall K. Kolka (USDA Forest Service Northern Research Station Grand Rapids, MN) and

Carl P. J. Mitchell.

Contributions: The PEATcosm experiment was designed and implemented by E.A.L., E.S.K.,

L.P. and R.K.K. Research questions and experimental approach for this chapter were articulated

by K.M.H. with input from C.P.J.M. Pore water and snowmelt runoff samples were collected by

L.P. with assistance from E.S.K. and K.M.H. All mercury laboratory analyses were performed

by K.M.H. Sulfate and phenolics pore water data and cellulose decomposition assay data were

provided by L.P. and E.S.K. Data analyses and data interpretation were performed by K.M.H.

with input from C.P.J.M. The manuscript was prepared by K.M.H. with comprehensive review

by C.P.J.M. All co-authors L.P., E.S.K., E.A.L. and R.K.K. provided revisions to the manuscript

prior to submission to Global Biogeochemical Cycles.

1.5.3 Chapter 3

Chapter 3 investigates the potential impacts of deep peat warming as well as lowered water

tables and altered vascular plant communities on total gaseous Hg exchange with peat soils. This

paper is published in Atmospheric Environment (2017, Vol. 154, pg. 247-259,

http://dx.doi.org/10.1016/j.atmosenv.2017.01.049). The co-authors on this paper are Evan S.

Kane (USDA Forest Service Northern Research Station Houghton, MI and Michigan

Technological University), Lynette Potvin, Erik A. Lilleskov (USDA Forest Service Northern

Research Station Houghton, MI), Randall K. Kolka (USDA Forest Service Northern Research

Station Grand Rapids, MN) and Carl P. J. Mitchell.


L.P. and R.K.K. Access to the SPRUCE site was facilitated by R.K.K. Research questions and

experimental approach for this chapter were articulated by K.M.H. with input from C.P.J.M.

Total gaseous mercury flux measurements at PEATcosm were performed by K.M.H. with

assistance from L.P. and E.S.K. SPRUCE flux measurements were performed by K.M.H.

Vegetation samples from PEATcosm were collected by L.P. All mercury laboratory analyses

were performed by K.M.H. Data analyses and data interpretation were performed by K.M.H.

27

with input from C.P.J.M. The manuscript was prepared by K.M.H. with comprehensive review

by C.P.J.M. All co-authors L.P., E.S.K., E.A.L. and R.K.K. provided revisions to the manuscript

prior to submission to Atmospheric Environment.

1.5.4 Chapter 4

Chapter 4 examines the impacts of lowered, fluctuating water tables and shifting vascular plant

communities on solid phase peat Hg and MeHg including partitioning to pore water as well as

methylation and demethylation rate potentials. This paper has been submitted to Biogeochemistry

and is currently under review.. The co-authors on this paper will be Evan S. Kane (USDA Forest

Service Northern Research Station Houghton, MI and Michigan Technological University),

Lynette Potvin, Erik A. Lilleskov (USDA Forest Service Northern Research Station Houghton,

MI), Randall K. Kolka (USDA Forest Service Northern Research Station Grand Rapids, MN)

and Carl P. J. Mitchell.


L.P. and R.K.K. Research questions and experimental approach for this chapter were articulated

by K.M.H. with input from C.P.J.M. Peat samples were collected by E.A.L., L.P. and E.S.K.

Peat coring and isotopic incubations for methylation-demethylation assays were performed by

K.M.H. with input from C.P.J.M. All mercury laboratory analyses were performed by K.M.H.

Data analyses and data interpretation were performed by K.M.H. with input from C.P.J.M. The

manuscript was prepared by K.M.H. with comprehensive review by C.P.J.M. All co-authors

L.P., E.S.K., E.A.L. and R.K.K. provided revisions to the manuscript prior to submission to

Biogeochemistry.

1.5.5 Chapter 5

Chapter 5 applies Hg isotopic analyses to compare the sources of Hg to the two sub-boreal

peatlands study sites and the processes influencing the fractionation of natural abundance stable

Hg isotopes in the surface peat. This paper is currently being prepared for submission to Global

Biogeochemical Cycles. The co-authors on this paper will be Evan S. Kane (USDA Forest

28

Service Northern Research Station Houghton, MI and Michigan Technological University),

Lynette Potvin, Erik A. Lilleskov (USDA Forest Service Northern Research Station Houghton,

MI), Randall K. Kolka (USDA Forest Service Northern Research Station Grand Rapids, MN),

Wang Zheng (University of Toronto), Bridget A. Bergquist (University of Toronto) and Carl P.

J. Mitchell.


L.P. and R.K.K. Access to the S1 (SPRUCE) site was facilitated by R.K.K. Research questions

and experimental approach for this chapter were articulated by K.M.H. with input from C.P.J.M.

Peat samples were collected by E.A.L., L.P. and E.S.K. at PEATcosm and K.M.H. at the S1

(SPRUCE) peatland. Isotopic analyses were performed in B.A.B.’s laboratory at the University

of Toronto. All mercury laboratory analyses were performed by W.Z. and K.M.H. Data

analyses were performed by W.Z. and data interpretation was performed by K.M.H. with input

from C.P.J.M. and B.A.B. The manuscript was prepared by K.M.H. with comprehensive review

by C.P.J.M.

1.5.6 Chapter 6

Chapter 6 summarizes and synthesizes the results of the thesis. Areas of future research to

further explore the impacts of climate change on peatland Hg cycling are also addressed.

Contributions: Written by K.M.H. with comprehensive review by C.P.J.M.

1.6 References

Admiral, S. A. and P. M. Lafleur (2007), Modelling of latent heat flux partitioning a bog

peatland. Agriculture and Forest Meteorology, 144, 213-229.

Bahlmann, E., R. Ebinghaus, and W. Ruck (2004), The effect of soil moisture on the emissions

of mercury from soils. Materials and Geoenvironment, 51, 791–794.

29

Barkay, T., R. R. Turner, A. VandenBrook, and C. Liebert (1991), The relationships of Hg (II)

volatilization from a fresh-water pond to the abundance of mer-genes in the gene pool of

the indigenous microbial community. Microbial Ecology, 21, 151-161.

Benoit, J. M., C. C. Gilmour, A. Heyes, R. P. Mason, and C. L. Miller (2003), Geochemical and

biological controls over methylmercury production and degradation in aquatic

ecosystems. ACS Symposium Series, 835, 262-297.

Benoit, J. M., C. C. Gilmour, R. P. Mason, and A. Heyes (1999), Sulfide controls on mercury

speciation and bioavailability to methylating bacteria in sediment pore waters.

Environmental Science and Technology, 33, 951-957.

Bergquist, B. A. and J. D. Blum (2007), Mass-dependent and –independent fractionation of Hg

isotopes by photoreduction in aquatic systems. Science, 318, 417–420.

Bergquist, B. A. and J. D. Blum (2009), The odds and evens of mercury isotopes: Applications

of mass-dependent and mass-independent isotope fractionation. Elements, 5, 353–357.

Berman, M. and R. Bartha (1986), Levels of chemical versus biological methylation of mercury

in sediments. Bulletin of Environmental Contamination and Toxicology, 36, 401-404.

Bishop, K., Y. –H. Lee, C. Pettersson, and B. Allard (1995), Methylmercury output from the

Svartberget catchment in Northern Sweden during spring flood. Water, Air, and Soil

Pollution, 80, 445-454.

Blok, D., M. M. P. D. Heijmans, G. Schaepman-Strub, J. van Ruijven, F. J. W. Parmentier, T. C.

Maximov, and F. Berendse (2011), The cooling capacity of mosses: Controls on water

and energy fluxes in a Siberian tundra site. Ecosystems, 14, 1055-1065.

Branfireun, B. A. and N. T. Roulet (2002), Controls on the fate and transport of methylmercury

in a boreal headwater catchment, Northwestern Ontario, Canada. Hydrology and Earth

System Sciences, 6, 785-794.

Bridgham, S. D., J. Pastor, B. Dewey, J. F. Weltzin, and K. Updegraff (2008), Rapid carbon

response of peatlands to climate change. Ecology, 89, 3041-3048.

30

Carpi, A. and S. E. Lindberg (1998), Application of a Teflon dynamic flux chamber for

quantifying soil mercury flux: Tests and results over background soil. Atmospheric

Environment, 32, 873-882.

Chason, D. B. and D. I. Siegel (1986), Hydraulic conductivity and related physical properties of

peat, Lost River Peatland, Northern Minnesota. Soil Science, 142, 91-99.

Chen, J., H. Hintelmann, X. Feng, and B. Dimock (2012), Unusual fractionation of both odd and

even mercury isotopes in precipitation from Peterborough, ON, Canada. Geochimica et

Cosmochimica Acta, 90, 33–46.

Chimner, R. A., T. G. Pypker, J. A. Hribljan, P. A. Moore, and J. M. Waddington (2016), Multi-

decadal changes in water table levels alter peatland carbon cycling. Ecosystems,

DOI:10.1007/s10021-016-0092-x.

Clymo, R. S. (1984), The limits to peat bog growth. Philosophical Transactions of the Royal

Society of London Series B: Biological Sciences, 303, 605-654.

Coleman Wasik, J. K., D. R. Engstrom, C. P. J. Mitchell, E. B. Swain, B. A. Monson, S. J.

Balogh, J. D. Jeremiason, B. A. Branfireun, R. K. Kolka, and J. E. Almendinger (2015),

The effects of hydrologic fluctuation and sulfate regeneration on mercury cycling in an

experimental peatland. Journal of Geophysical Research Biogeosciences, 120,

doi:10.1002/2015JG002993.

Compeau, G. and R. Bartha (1984), Methylation and demethylation of mercury under controlled

redox, pH, and salinity conditions. Applied and Environmental Microbiology, 48, 1203-

1207.

Compeau, G. and R. Bartha (1985), Sulfate-reducing bacteria: Principal methylators of mercury

in anoxic estuarine sediment. Applied and Environmental Microbiology, 50, 498-502.

Demers, J. D., J. D. Blum, and D. R. Zak (2013), Mercury isotopes in a forested ecosystem:

Implications for air-surface exchange dynamics and the global mercury cycle. Global

Biogeochemical Cycles, 27, 222-238.

31

Dieleman, C. M., B. A. Branfireun, J. W. McLaughlin, and Z. Lindo (2015), Climate change

drives a shift in peatland ecosystem plant community: Implications for ecosystem

function and stability. Global Change Biology, 21, 388-395.

Dittman, J. A., J. B. Shanley, C. T. Driscoll, G. R. Aiken, A. T. Chalmers, J. E. Towse, and P.

Selvendiran (2010), Mercury dynamics in relation to dissolved organic carbon

concentration and quality during high flow events in three Northeastern U.S. streams.

Water Resources Research, 46, W07522.

Donovan, P. M., J. D. Blum, J. D. Demers, B. Gu, S. C. Brooks, and J. Peryam (2014),

Identification of multiple mercury sources to stream sediments near Oak Ridge, TN,

USA. Environmental Science and Technology, 48, 3666-3674.

Driscoll, C. T., V. Blette, C. Yan, C. L. Schofield, R. Munson, and J. Holsapple (1995), The role

of dissolved organic carbon in the chemistry and bioavailability of mercury in remote

Adirondack Lakes. Water, Air, and Soil Pollution, 80, 499–508.

Driscoll, C. T., Y. –J. Han, C. Y. Chen, D. C. Evers, K. F. Lambert, T. M. Holsen, N. C.

Kamman, and R. K. Munson (2007), Mercury contamination in forest and freshwater

ecosystems in the northeastern United States. BioScience, 57, 17-28.

Driscoll, C. T., R. P. Mason, H. M. Chan, D. J. Jacob, and N. Pirrone (2013), Mercury as a

global pollutant: Sources, pathways, and effects. Environmental Science and Technology,

47, 4967-4983.

Edwards, G. C., P. E. Rasmussen, W. H. Schroeder, R. J. Kemp, G. M. Dias, C. R. Fitzgerald-

Hubble, E. K. Wong, L. Halfpenny-Mitchell, and M. S. Gustin (2001), Sources of

variability in mercury flux measurements. Journal of Geophysical Research, 106, 5421-

5435.

Enrico, M., G. L. Roux, N. Marusczak, L. –E. Heimbürger, A. Claustres, X. Fu, R. Sun, and J. E.

Sonke (2016), Atmospheric mercury transfer to peat bogs dominated by gaseous

elemental mercury dry deposition. Environmental Science and Technology, 50, 2405–

2412.

32

Ericksen, J. A., M. S. Gustin, D. E. Schorran, D. W. Johnson, S. E. Lindberg, and J. S. Coleman

(2003), Accumulation of atmospheric mercury in forest foliage. Atmospheric


Fenner, N., C. Freeman, M. A. Lock, H. Harmens, B. Reynolds, and T. Sparks (2007),

Interactions between elevated CO2 and warming could amplify DOC exports from

peatland catchments. Environmental Science and Technology, 41, 3146-3152.

Ferone, M. and K. J. Devito (2004), Variation in groundwater-surface water interactions of pond-

peatland complexes along a Boreal Plain landscape gradient. Journal of Hydrology, 292,

75-95.

Fitzgerald, W. F., D. R. Engstrom, R. P. Mason, and E. A. Nater (1998), The case for

atmospheric mercury contamination in remote areas. Environmental Science and

Technology, 32, 1-7.

Fritsche, J., S. Osterwalder, M. B. Nilsson, J. Sagerfors, S. Åkerblom, K. Bishop, and C. Alewell

(2014), Evasion of elemental mercury from a boreal peatland suppressed by long-term

sulfate addition. Environmental Science and Technology Letters, 1, 421-425.

Gabriel, M. C. and D. G. Williamson (2004), Principal biogeochemical factors affecting the

speciation and transport of mercury through the terrestrial environment. Environmental

Geochemistry and Health, 26, 421-434.

Gilmour, C. C., D. A. Elias, A. M. Kucken, S. D. Brown, A. V. Palumbo, C. W. Schadt, and J.

D. Wall (2011), Sulfate-reducing bacteria Desulfovibrio desulfuricans ND132 as a model

for understanding bacterial mercury methylation. Applied and Environmental

Microbiology, 77, 3938-3951.

Gilmour, C. C., E. A. Henry, and R. Mitchell (1992), Sulfate stimulation of mercury methylation

in freshwater sediments. Environmental Science and Technology, 26, 2281-2287.

Gilmour, C. C., M. Podar, A. L. Bullock, A. M. Graham, S. D. Brown, A. C. Somenahally, A.

Johs, R. A. Hurt Jr., K. L. Bailey, and D. A. Elias (2013), Mercury methylation by novel

33

microorganisms from new environments. Environmental Science and Technology, 47,

11810-11820.

Glaser, P. H., D. I. Siegel, E. A. Romanowicz, and Y. P. Shen (1997), Regional linkages between

raised bogs and the climate, groundwater, and landscape of north-western Minnesota.

Journal of Ecology, 85, 3-16.

Gorham, E. (1991) Northern peatlands: role in the carbon cycle and probable responses to

climatic warming. Ecological Applications, 1, 182-195.

Graham, A. M., G. R. Aiken, and C. C. Gilmour (2012), Dissolved organic matter enhances

microbial mercury methylation under sulfidic conditions. Environmental Science and

Technology, 46, 2715-2723.

Graham, A. M., G. R. Aiken, and C. C Gilmour (2013), Effect of dissolved organic matter source

and character on microbial Hg methylation in Hg-S-DOM solutions. Environmental

Science and Technology, 47, 5746-5754.

Gratz, L. E., G. J. Keeler, J. D. Blum, and L. S. Sherman (2010), Isotopic composition and

fractionation of mercury in great lakes precipitation and ambient air. Environmental

Science and Technology, 44, 7764–7770.

Grigal, D. F. (2003), Mercury sequestration in forests and peatlands: A review. Journal of

Environmental Quality, 32, 393-405.

Groisman, P. Y., R. W. Knight, and T. R. Karl (2012), Changes in intense precipitation over the

central United States. Journal of Hydrometeorology, 13, 47-66.

Grondin, A., M. Lucotte, A. Mucci, and B. Fortin (1995), Mercury and lead profiles and burdens

in soils of Quebec (Canada) before and after flooding. Canadian Journal of Fisheries and

Aquatic Sciences, 52, 2493-2506.

Gustin, M. S., P. J. Chavan, K. E. Dennett, E. A. Marchand, and S. Donaldson (2006),

Evaluation of wetland methyl mercury export as a function of experimental

manipulations. Journal of Environmental Quality, 35, 2352-2359.

34

Gustin, M. S. and J. Stamenkovic (2005), Effect of watering and soil moisture on mercury

emissions from soils. Biogeochemistry, 76, 215-232.

Haitzer, M., G. R. Aiken, and J. N. Ryan (2003), Binding of mercury(II) to aquatic humic

substances: Influence of pH and source of humic substances. Environmental Science and

Technology, 37, 2436-2441.

Harris, R. C., J. W. M. Rudd, M. Amyot, C. L. Babiarz, K. G. Beaty, P. J. Blanchfield, et al.

(2007), Whole-ecosystem study shows rapid fish-mercury response to changes in

mercury deposition. Proceedings of the National Academy of Sciences, 104, 16586-

16591.

Haynes, K. M. and C. P. J. Mitchell (2015), Precipitation input and antecedent soil moisture

effects on mercury mobility in soil – Laboratory experiments with an enriched stable

isotope tracer. Hydrological Processes, 29, 4161-4174.

Haynes, K. M., M. D. Preston, J. W. McLaughlin, K. Webster, and N. Basiliko (2015),

Dissimilar bacterial and fungal decomposer communities across rich to poor fen

peatlands exhibit functional redundancy. Canadian Journal of Soil Science, 95, 219-230.

Heyes, A., T. R. Moore, J. W. M. Rudd, and J. J. Dugoua (2000), Methyl mercury in pristine and

impounded boreal peatlands, Experimental Lakes Area, Ontario. Canadian Journal of

Fisheries and Aquatic Sciences, 57, 2211-2222.

Hintelmann, H., R. Harris, A. Heyes, J. P. Hurley, C. A. Kelly, D. P. Krabbenhoft, S. Lindberg,

J. W. M. Rudd, K. J. Scott, and V. L. St. Louis (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.

Hintelmann, H., P. M. Welbourn, and R. D. Evans (1997), Measurement of complexation of

methylmercury(II) compunds by freshwater humic substances using equilibrium dialysis.


35

Ise, T., A. L. Dunn, S. C. Wofsy, and P. R. Moorcroft (2008), High sensitivity of peat

decomposition to climate change through water-table feedback. Nature Geoscience, 1,

763-766.

Janssen, E., D. J. Wuebbles, K. E. Kunkel, S. C. Olsen, and A. Goodman (2014), Observational-

and model-based trends and projections of extreme precipitation over the contiguous

United States. Earth’s Future, 2, 99–113.

Jensen, S. and A. Jernelöv (1969), Biological methylation of mercury in aquatic organisms.

Nature, 223, 753-754.

Jeremiason, J. D., D. R. Engstrom, E. B. Swain, E. A. Nater, B. M. Johnson, J. E. Almendinger,

B. A. Monson, and R. K. Kolka (2006), Sulfate addition increases methylmercury

production in an experimental wetland. Environmental Science and Technology, 40,

3800–3806.

Jiskra, M., J. G. Wiederhold, U. Skyllberg, R. –M. Kronberg, I. Hajdas, and R. Kretzschmar

(2015), Mercury deposition and re-emission pathways in boreal forest soils investigated

with Hg isotope signatures. Environmental Science and Technology, 49, 7188–7196.

Jonsson, S., N. M. Marui, and R. P. Mason (2016), Dimethylmercury formation mediated by

inorganic and organic reduced sulfur surfaces. Scientific Reports, 6, 27958. DOI:

10.1038/srep27958.

Karlsson, T. and U. Skyllberg (2003), Bonding of ppb levels of methyl mercury to reduced sulfur

groups in soil organic matter. Environmental Science and Technology, 37, 4912-4918.

Khwaja, A. R., P. R. Bloom, and P. L. Brezonik (2006), Binding constants of divalent mercury

(Hg2+

) in soil humic acids and soil organic matter. Environmental Science and


King, J. K., S. M. Harmon, T. T. Fu, and J. B. Gladden (2002), Mercury removal, methylmercury

formation, and sulfate-reducing bacteria profiles in wetland mesocosms. Chemosphere,

46, 859-870.

36

King, J. K., J. E. Kostka, M. E. Frischer, and F. M. Saunders (2000), Sulfate-reducing bacteria

methylate mercury at variable rates in pure culture and in marine sediments. Applied and

Environmental Microbiology, 66, 2430-2437.

Kolka, R. K., D. F. Grigal, E. A. Nater, and E. S. Verry (2001), Hydrologic cycling of mercury

and organic carbon in a forested upland-bog watershed. Soil Science Society of America

Journal, 65, 897-905.

Kolka, R. K., C. P. J. Mitchell, J. D. Jeremiason, N. A. Hines, D. F. Grigal, D. R. Engstrom, J. K.

Coleman Wasik, E. A. Nater, E. B. Swain, B. A. Monson, J. A. Fleck, B. Johnson, J. E.

Almendinger, B. A. Branfireun, P. L. Brezonik, and J. B. Cotner (2011), Mercury cycling

in peatland watersheds. Chapter 11. In: Kolka, R. K., S. D. Sebestyen, E. S. Verry, K. N.

Brooks, eds. Peatland biogeochemistry and watershed hydrology at the Marcell

Experimental Forest. Boca Raton, FL: CRC Press: 349-370.

Krabbenhoft, D. P. and E. M. Sunderland (2013), Global change and mercury. Science, 341,

1457-1458.

Kritee, K., J. D. Blum, J. R. Reinfelder, and T. Barkay (2013), Microbial stable isotope

fractionation of mercury: A synthesis of present understanding and future directions.

Chemical Geology, 336, 13-25.

Kyllönen, K., H. Hakola, H. Hellén, M. Korhonen, and M. Verta (2012), Atmospheric mercury

fluxes in a southern boreal forest and wetland. Water Air and Soil Pollution, 223, 1171-

1182.

Laacouri, A., E. A. Nater, and R. K. Kolka (2013), Distribution and uptake dynamics of mercury

in leaves of common deciduous tree species in Minnesota, U.S.A. Environmental Science

and Technology, 47, 10462-10470.

Lee, X., G. Benoit and X. Hu (2000), Total gaseous mercury concentration and flux over a

coastal saltmarsh vegetation in Connecticut, USA. Atmospheric Environment, 34, 4205-

4213.

37

Lehnherr, I. (2014), Methylmercury biogeochemistry: a review with special reference to Arctic

aquatic ecosystems. Environmental Reviews, 22, 229-243.

Limpens, J., F. Berendse, C. Blodau, J. G. Canadell, C. Freeman, J. Holden, N. Roulet, H. Rydin,

and G. Schaepman-Strub (2008), Peatlands and the carbon cycle: from local processes to

global implications – a synthesis. Biogeosciences, 5, 1475-1491.

Lin, C. -J., M. S. Gustin, P. Singhasuk, C. Eckley, and M. Miller (2010), Empirical models for

estimating mercury flux from soil. Environmental Science and Technology, 44, 8522-

8528.

Lindberg, S., R. Bullock, R. Ebinghaus, D. Engstrom, X. B. Feng, W. Fitzgerald, N. Pirrone, E.

Prestbo, and C. A. Seigneur (2007), A synthesis of progress and uncertainties in

attributing the sources of mercury in deposition. Ambio, 36, 19-32.

Marsik, F. J., G. J. Keeler, S. E. Lindberg, and H. Zhang (2005), Air-surface exchange of

gaseous mercury over a mixed sawgrass-cattail stand within the Florida Everglades.


Marvin-DiPasquale, M., J. Agee, C. McGowan, R. S. Oremland, D. Krabbenhoft, and C. C.

Gilmour (2000), Methyl-mercury degradation pathways: A comparison among three

mercury-impacted ecosystems. Environmental Science and Technology, 34, 4908-4916.

Mason, R. P., W. F. Fitzgerald, and F. M. M. Morel (1994), The biogeochemical cycling of

elemental mercury: Anthropogenic influences. Geochimica et Cosmochimica Acta, 58,

3191-3198.

Mason, R. P., K. R. Rolfhus, and W. F. Fitzgerald (1995), Methylated and elemental mercury

cycling in surface and deep ocean waters of the north Atlantic. Water, Air, and Soil

Pollution, 80, 665- 677.

Mergler, D., H. A. Anderson, L. H. M. Chan, K. R. Mahaffey, M. Murray, M. Sakamoto, and A.

H. Stern (2007), Methylmercury exposure and health effects in humans: A worldwide

concern. Ambio, 36, 3-11.

38

Mierle, G. and R. Ingram (1991), The role of humic substances in the mobilization of mercury

from watersheds. Water, Air, and Soil Pollution, 56, 349-357.

Miskimmin, B. M., J. W. M. Rudd, and C. A. Kelly (1992), Influence of dissolved organic

carbon, pH, and microbial respiration rates on mercury methylation and demethylation in

lake water. Canadian Journal of Fisheries and Aquatic Sciences, 49, 17-22.

Mitchell, C. P. J., B. A. Branfireun, and R. K. Kolka (2008), Assessing sulfate and carbon

controls on net methylmercury production in peatlands: An in situ mesocosm approach.

Applied Geochemistry, 23, 503-518.

Mitchell, C. P. J. and C. C. Gilmour (2008) Methylmercury production in a Chesapeake Bay salt

marsh. Journal of Geophysical Research, 113, G00C04, doi:10.1029/2008JG000765.

Moore, T. and N. Basiliko (2006), Decomposition in boreal peatlands. Boreal Peatland

Ecosystems (eds R. K. Wieder & D. H. Vitt), pp. 125-143. Springer-Verlag, Berlin,

Heidelberg.

Morel, F. M. M., A. M. L. Kraepiel, and M. Amyot (1998), The chemical cycle and

bioaccumulation of mercury. Annual Review of Ecology and Systematics, 29, 543-566.

Mueller, P., K. Jensen, and J. P. Megonigal (2016), Plants mediate soil organic matter

decomposition in response to sea level rise. Global Change Biology, 22, 404-414.

Munthe, J., R. A. Bodaly, B. A. Branfireun, C. T. Driscoll, C. C. Gilmour, R. Harris, M. Horvat,

M. Lucotte, and O. Malm (2007), Recovery of mercury-contaminated fisheries. Ambio,

36, 33-44.

Nichols, D. S. and E. S. Verry (2001), Stream flow and ground water recharge from small

forested watersheds in north central Minnesota. Journal of Hydrology, 245, 89-103.

Nunes, F. L. D., L. Aquilina, J. de Ridder, A.-J. Francez, A. Quaiser, J.-P. Caudal, P.

Vandenkoornhuyse, and A. Dufresne (2015), Time-scales of hydrological forcing on the

geochemistry and bacterial community structure of temperate peat soils. Scientific

Reports, 5, 14612. DOI: 10.1038/srep14612.

39

Osterwalder, S., J. Fritsche, C. Alewell, M. Schmutz, M. B. Nilsson, G. Jocher, J. Sommar, J.

Rinne, and K. Bishop (2016), A dual-inlet, single detector relaxed eddy accumulation

system for long-term measurement of mercury flux. Atmospheric Measurement

Techniques, 9, 509-524.

Parks, J. M., H. Guo, C. Momany, L. Liang, S. M. Miller, A. O. Summers, and J. C. Smith

(2009), Mechanism of Hg-C protonolysis in the organomercurial lyase merB. Journal of

the American Chemical Society, 131, 13278-13285.

Parks, J. M., A. Johs, M. Podar, R. Bridou, R. A. Hurt, S. D. Smith, et al. (2013), The genetic

basis for bacterial mercury methylation. Sciencexpress, DOI: 10.1126/science.1230667.

Peltoniemi, K., H. Fritze, and R. Laiho (2009), Response of fungal and actinobacterial

communities to water-level drawdown in boreal peatland sites. Soil Biology and

Biochemistry, 41, 1902-1914.

Peltoniemi, K., R. Laiho, H. Juottonen, O. Kiikkilä, P. Mäkiranta, K. Minkkinen, T. Pennanen,

T. Penttilä, T. Sarjala, E.-S. Tuittila, T. Tuomivirta, and H. Fritze (2015), Microbial

ecology in a future climate: effects of temperature and moisture on microbial

communities of two boreal fens. FEMS Microbiology Ecology, 91, fiv062.

doi:10.1093/femsec/fiv062.

Potvin, L. R., E. S. Kane, R. A. Chimner, R. K. Kolka, and E. A. Lilleskov (2015), Effects of

water table position and plant functional group on plant community, aboveground

production, and peat properties in a peatland mesocosm experiment (PEATcosm). Plant

and Soil, 387, 277-294.

Ramlal, P. S., C. A. Kelly, J. W. M. Rudd, and A. Furutani (1993), Sites of methylmercury

production in remote Canadian Shield lakes. Canadian Journal of Fisheries and Aquatic

Sciences, 50, 972-979.

Ramlal, P. S., J. W.M. Rudd, A. Furutani, and L. Xun (1985), The effect of pH on methyl

mercury production and decomposition in lake sediments. Canadian Journal of Fisheries

and Aquatic Sciences, 42, 685-692.

40

Ravichandran, M. (2004), Interactions between mercury and dissolved organic matter – a review.

Chemosphere, 55, 319-331.

Rea, A. W., S. E. Lindberg, and G. J. Keeler (2000), Assessment of dry deposition and foliar

leaching of mercury and selected trace elements based on washed foliar and surrogate

surfaces. Environmental Science and Technology, 34, 2418-2425.

Rea, A. W., S. E. Lindberg, T. Scherbatskoy, and G. J. Keeler (2002), Mercury accumulation in

foliage over time in two northern mixed-hardwood forests. Water Air and Soil Pollution,

133, 49-67.

Romanowicz, E. A., D. I. Siegel, and P. H. Glaser (1993), Hydraulic reversals and episodic

methane emissions during drought cycles in mires. Geology, 21, 231-234.

Scheffer, R. A., R. S. P. Van Logtestijn, and J. T. A. Verhoeven (2001), Decomposition of Carex

and Sphagnum litter in two mesotrophic fens differing in dominant plant species. OIKOS,

92, 44-54.

Scheuhammer, A. M., M. W. Meyer, M. B. Sandheinrich, and M. W. Murray (2007), Effects of

environmental methylmercury on the health of wild birds, mammals and fish. Ambio, 36,

12-18.

Schroeder, W. and J. Munthe (1998), Atmospheric mercury – an overview. Atmospheric


Schuster, E. (1991), The behaviour of mercury in the soil with special emphasis on complexation

and adsorption processes – A review of the literature. Water, Air, and Soil Pollution, 56,

667-680.

Schuster, P. F., J. B. Shanley, M. Marvin-Dipasquale, M. M. Reddy, G. R. Aiken, D. A. Roth, H.

E. Taylor, D. P. Krabbenhoft, and J. F. DeWild (2008), Mercury and organic carbon

dynamics during runoff episodes from a northeastern USA watershed. Water, Air, and

Soil Pollution, 187, 89-108.

Selin, N.E. (2009), Global biogeochemical cycling of mercury: A review. Annual Review of

Environment and Resources, 34, 43-63.

41

Sellers, P., C. A. Kelly, J. W. M. Rudd, and A. R. MacHutchon (1996), Photodegradation of

methylmercury in lakes. Nature, 380, 694-697.

Shantz, M. A. and J. S. Price (2006), Hydrological changes following restoration of the Bois-des-

Bel Peatland, Quebec, 1999-2002. Journal of Hydrology, 331, 543-553.

Siegel, D. I. and P. H. Glaser (1987), Groundwater flow in a bog fen complex, Lost River

Peatland, Northern Minnesota, Journal of Ecology, 75, 743-754.

Siegel, D. I., A. S. Reeve, P. H. Glaser, and E. A. Romanowicz (1995), Climate-driven flushing

of pore water in peatlands. Nature, 374, 531-533.

Skyllberg, U., P. R. Bloom, J. Qian, C. –M. Lin, and W. F. Bleam (2006), Complexation of

mercury(II) in soil organic matter: EXAFS evidence for linear two-coordination with

reduced sulfur groups. Environmental Science and Technology, 40, 4174-4180.

Skyllberg, U., K. Xia, P. R. Bloom, E. A. Nater, and W. F. Bleam (2000), Binding of mercury(II)

to reduced sulfur in soil organic matter along upland-peat soil transects. Journal of


Song, X. and B. Van Heyst (2005), Volatilization of mercury from soils in response to simulated

precipitation. Atmospheric Environment, 39, 7494-7505.

Stamenkovic, J. and M. S. Gustin (2009), Nonstomatal versus stomatal uptake of atmospheric

mercury. Environmental Science and Technology, 43, 1367-1372.

St. Louis, V. L., J. W. M. Rudd, C. A. Kelly, K. G. Beaty, N. S. Bloom, and R. J. Flett (1994),

Importance of Wetlands as Sources of Methyl Mercury to Boreal Forest Ecosystems.

Canadian Journal of Fisheries and Aquatic Sciences, 51, 1065-1076.

Strack, M., J. M. Waddington, L. Rochefort, and E. –S. Tuittila (2006), Response of vegetation

and net ecosystem carbon dioxide exchange at different peatland microforms following

water table drawdown. Journal of Geophysical Research, 111, G02006,

DOI:10.1029/2005JG000145.

42

Strickland, M. S. and J. Rousk (2010), Considering fungal:bacterial dominance in soils –

Methods, controls, and ecosystem implications. Soil Biology and Biochemistry, 42,

1385-1395.

Sun, G., J. Sommar, X. Feng, C. –J. Lin, M. Ge, W. Wang, R. Yin, X. Fu, and L. Shang (2016),

Mass-dependent and -independent fractionation of mercury isotope during gas-phase

oxidation of elemental mercury vapor by atomic Cl and Br. Environmental Science and


Tarnocai, C. (2006), The effect of climate change on carbon in Canadian peatlands. Global and

Planetary Change, 53, 222-232.

Tarnocai, C. (2009), The impact of climate change on Canadian peatlands. Canadian Water

Resources Journal, 34, 453-466.

Tjerngren, I., M. Meili, E. Björn, and U. Skyllberg (2012), Eight boreal wetlands as sources and

sinks for methyl mercury in relation to soil acidity, C/N ratio, and small-scale flooding.


Turetsky, M. R., B. Bond-Lamberty, E. Euskirchen, J. Talbot, S. Frolking, A. D. McGuire, and

E. –S. Tuittila (2012), The resilience and functional role of moss in boreal and arctic

ecosystems. New Phytologist, 196, 49-67.

Turunen, J., E. Tomppo, K. Tolonen, and A. Reinikainen (2002), Estimating carbon

accumulation rates of undrained mires in Finland – application to boreal and subarctic

regions. The Holocene, 12, 69-80.

Waddington, J. M., P. J. Morris, N. Kettridge, G. Granath, D. K. Thompson, and P. A. Moore

(2015), Hydrological feedbacks in northern peatlands. Ecohydrology, 8, 113-127.

Warner, K. A., E. E. Roden, and J. –C. Bonzongo (2003), Microbial mercury transformation in

anoxic freshwater sediments under iron-reducing and other electron-accepting conditions.


43

Weltzin, J. F., J. Pastor, C. Harth, S. D. Bridgham, K. Updegraff, and C. T. Chapin (2000),

Response of bog and fen plant communities to warming and water-table manipulations.

Ecology, 81, 3464-3478.

Whittington, P. N. and J. S. Price (2006), The effects of water table drawdown (as a surrogate for

climate change) on the hydrology of a fen peatland, Canada. Hydrological Processes, 20,

3589-3600.

Whittington, P. N., M. Strack, S. Kaufman, R. van Haarlem, P. Stoesser, J. Maltez, J. S. Price,

and M. Stone (2007), The role of peat volume change and vegetation community on the

hydrology of a kettle-hole wetland in southern Ontario, Canada. Mires and Peat, 2, 1-14.

Windham-Myers, L., M. Marvin-DiPasquale, D. P. Krabbenhoft, J. L. Agee, M. H. Cox, P.

Heredia-Middleton, et al. (2009), Experimental removal of wetland emergent vegetation

leads to decreased methylmercury production in surface sediment. Journal of

Geophysical Research, 114, G00C05.

Winfrey, M. R. and J. W. M. Rudd (1990), Environmental factors affecting the formation of

methylmercury in low pH lakes. Environmental Toxicology and Chemistry, 9, 853-869.

Winter, T.C. (2000), The vulnerability of wetlands to climate change: A hydrologic landscape

perspective. Journal of the American Water Resources Association, 36, 305-311.

Xin, M., M. Gustin, and D. Johnson (2007), Laboratory investigation of the potential for re-

emission of atmospherically derived Hg from soils. Environmental Science and

Technology, 41, 4946-4951.

Yin, R., X. Feng, and W. Shi (2010), Application of the stable-isotope system to the study of

sources and fate of Hg in the environment: A review. Applied Geochemistry, 25, 1467-

1477.

Yu, L., S. Zhong, L. Pei, X. Bian, and W. E. Heilman (2016), Contribution of large-scale

circulation anomalies to changes in extreme precipitation frequency in the United States.

Environmental Research Letters, 11, 044003.

44

Zhang, H. and S. E. Lindberg (1999), Processes influencing the emission of mercury from soils:

A conceptual model. Journal of Geophysical Research, 104, 21889-21896.

Zhang, L., L. P. Wright, and P. Blanchard (2009), A review of current knowledge concerning dry

deposition of atospheric mercury. Atmospheric Environment, 43, 5853−5864.

Zheng, W., D. Obrist, D. Weis, and B. A. Bergquist (2016), Mercury isotope compositions

across North American forests. Global Biogeochemical Cycles, 30, 1475-1492.

Zillioux, E. J., D. B. Porcella, and J. M. Benoit (1993), Mercury cycling and effects in freshwater

wetland ecosystems. Environmental Toxicology and Chemistry, 12, 2245-2264.

45

Chapter 2 Mobility and Transport of Mercury and Methylmercury in Peat as a Function of Changes in Water Table Regime and Plant Functional

Groups

2

2.1 Abstract

Climate change is likely to significantly affect the hydrology, ecology and ecosystem function of

peatlands, with potentially important but unclear impacts on mercury mobility within and

transport from peatlands. Using a full-factorial mesocosm approach we investigated the potential

impacts on mercury mobility of water table regime changes (high and low) and vegetation

community shifts (sedge-dominated, Ericaceae-dominated or unmanipulated control) in peat

monoliths at the PEATcosm mesocosm facility in Houghton, Michigan. Lower and more

variable water table regimes and the loss of Ericaceae shrubs act significantly and independently

to increase both total Hg (THg) and methylmercury (MeHg) concentrations in peat pore water

and in spring snowmelt runoff. These differences are related to enhanced peat decomposition and

internal regeneration of electron acceptors which are more strongly related to water table regime

than to plant community changes. Loss of Ericaceae shrubs and an increase in sedge cover may

also affect Hg concentrations and mobility via oxygen shuttling and/or the provision of labile

root exudates. Altered hydrological regimes and shifting vegetation communities, as a result of

global climate change, are likely to enhance Hg transport from peatlands to downstream aquatic

ecosystems.

2.2 Introduction

Boreal peatlands store vast amounts of carbon, are critical to hydrological cycling, and are

landscape scale “hot spots” for the storage and biogeochemical transformations of pollutants

such as mercury (Hg) in the landscape (Gorham, 1991; Branfireun et al., 1998; Limpens et al.,

2008). Peatland systems act as strong sinks of atmospherically deposited inorganic Hg (Kolka et

al., 2001; Grigal, 2003). With predominantly saturated, anoxic, organic peat soils and the

availability of electron acceptors such as sulfate, the hydrological and redox conditions of these

46

systems also provide an optimal setting for the production of methylmercury (MeHg), a potent

neurotoxin which bioaccumulates and biomagnifies up the food chain (Branfireun et al., 1996;

Scheuhammer et al., 2007; Tjerngren et al., 2012). Considerable research has focused on the

influence of enhanced sulfate deposition on the process of Hg methylation in peatland

ecosystems as sulfate-reducing bacteria are known to actively methylate Hg (Gilmour et al.,

1992; Bergman et al., 2012; Coleman Wasik et al., 2015). Recent research has confirmed that in

addition to sulfate- and iron-reducing bacteria, some methanogens and syntrophic, acetogenic

and fermentative Firmicutes are capable of Hg methylation, including those that have been

isolated from northern peatland ecosystems (Gilmour et al., 2013).

Peatlands, particularly those located at high latitudes, are sensitive to increasing temperatures

and altered precipitation regimes resulting from global climate change (Bridgham et al., 1995;

Ise et al., 2008; Limpens et al., 2008). The susceptibility of peatlands to climatic change may

depend upon the relative contributions of groundwater versus precipitation inputs. Bogs,

wherein water table position and peat accumulation capacity are governed strongly by the

balance between precipitation and evapotranspiration, are particularly vulnerable to climate

change (Winter, 2000). Climate-induced changes in the hydrological regimes of peatlands

significantly influence microbial community structure, which may have profound effects on peat

decomposition and carbon sequestration (Nunes et al., 2015; Peltoniemi et al., 2015). Changes in

bacterial community structure also have important links to methylmercury production in

peatlands (Strickman et al., 2016). Alterations to the carbon storage abilities of peatlands due to

climate-induced changes in the hydrology and ecology of these systems may affect the

biogeochemical cycling and mobility of Hg within, and export from, these wetland systems.

With increased variability in precipitation patterns predicted to occur in the northern continental

United States, prolonged periods of water table recession particularly during the summer months

may have important ecological implications (Kunkel et al., 2003; Groisman et al., 2005;

Thomson et al., 2005). For example, these hydrological changes may result in a shift in peatland

vegetation communities toward vascular-dominated functional groups (Weltzin et al., 2003;

Strack et al., 2006; Breeuwer et al., 2009; Dieleman et al., 2015). Different plant functional

groups including graminoids and Ericaceae shrubs may be more suited to the hydrological

conditions resulting from climate change. Although the development of shallower relative water

tables as a result of peat decomposition and compression is possible, dramatic fluctuations in

47

water levels when precipitation inputs follow prolonged dry periods may exert considerable

water stress on bryophytes including Sphagnum mosses (Weltzin et al., 2000; Whittington and

Price, 2006; Waddington et al., 2014). Vascular plants may have a competitive advantage over

Sphagnum moss species in such situations of prolonged water stress because of their ability to

access deeper stores of water through their more extensive rooting systems (Dieleman et al.,

2015). Vascular plants can also regulate water loss through their stomata, while Sphagnum

mosses do not possess stomata (Breeuwer et al., 2009). The ecophysiological traits of the

established plant functional groups may influence the biogeochemistry of the peat and the

regulation of ecosystem carbon dynamics. The rooting systems of some vascular plants such as

sedges may also act to shuttle oxygen to the rhizosphere, increasing the zone of peat aeration and

acting as a conduit for gaseous emissions (Weltzin et al., 2000; Strack et al., 2006; Bridgham et

al., 2008; Waddington et al., 2014). Shallowly-rooted ericaceous shrubs lack the adaptive

structures to survive in flooded conditions (Chapin et al., 1996) and therefore may thrive under

drier conditions. However, growth of shrubs may be hindered during periods of prolonged water

stress with significant water table recession, particularly in association with increasing

temperatures (Armstrong et al., 1991; Weltzin et al., 2003). In contrast, the deep root systems of

sedges may allow this plant functional group to survive and out-compete Ericaceae shrubs during

prolonged drought periods (Dieleman et al., 2015). The complex synergistic and antagonistic

feedback mechanisms limit the ability to predict the response of peatland vegetation

communities to climate change-induced water table recession. Therefore, both the sedge and

Ericaceae plant functional groups should be considered, both individually and coexisting, when

investigating the effects of shifting plant communities.

Water table recession and fluctuation, increased peat aeration, and shifting plant communities

may significantly affect Hg cycling and the process of Hg methylation due to alterations in redox

conditions of the peat and the internal recycling of sulfur species (Coleman Wasik et al., 2015).

The potential increased liberation of dissolved organic matter (DOM) may act to augment Hg

methylation through the provision of labile carbon substrates needed as electron donors for

methylating microbes (Mitchell et al., 2008a; Graham et al., 2013). The establishment of deeply-

rooted vascular plants such as sedges may also contribute an additional labile carbon source

through root exudates, stimulating methylating microbial communities (Bridgham et al., 1995;

Windham-Myers et al., 2009). Increased Hg mobilization via peat decomposition, partitioning,

48

and/or desorption into the pore water phase as well as Hg complex formation with DOM may

impact downstream aquatic ecosystems with the potential for enhanced transport and loading

particularly during high-flow events such as spring snowmelt (Bishop et al., 1995; Mitchell et al.,

2008c; Haynes and Mitchell, 2012). Moreover, the complexation of Hg with DOM, particularly

on thiol functional groups, increases Hg solubility and mobility, and possibly the bioavailability

of Hg for methylation (Skyllberg et al., 2000; Graham et al., 2013). Despite the propensity of

peatland ecosystems to methylate and export Hg to downstream ecosystems, minimal research

has been conducted to explore how Hg mobility and transport from peatlands may be affected by

climate change.

The purpose of this study was to investigate the potential effects of climate change-induced shifts

in water table regime and plant community composition on the mobility/movement of Hg within

and transport from peatland ecosystems. A novel peatland mesocosm experiment known as

PEATcosm (Peatland Experiment at the Houghton Mesocosm Facility) allowed for the

manipulation of water tables and plant community composition to simulate some of the potential

effects of climate change on peatland ecosystem functioning, including Hg cycling. Investigation

at the mesocosm scale provides a process-level understanding, which may be applied to the

landscape-scale, of the influences of hydrological and vegetation changes on the movement of

Hg within and export from peatlands.

2.3 Materials and Methods

2.3.1 Study Site and Experimental Design

The PEATcosm Mesocosm Facility is located in Houghton, Michigan, USA (47.11469° N,

88.54787° W) on the property of the United States Department of Agriculture (USDA) Forest

Service Northern Research Station - Forestry Sciences Laboratory. The regional climate is humid

continental with typical annual precipitation of approximately 870 mm at the site. Mean

temperatures in this area range from -13°C in January to 24°C in July. The average growing

season lasts approximately 132 days (30 year means at Houghton County Airport; NOAA

National Climatic Data Center) (Potvin et al., 2015).

Twenty-four intact 1-m3 (1 m x 1 m x 1 m) peat monoliths were extracted from an ombrotrophic

peatland located in Meadowlands, Minnesota, USA in May 2010. Monoliths were placed into

49

individual mesocosm bins, transported to and subsequently installed in the PEATcosm facility.

The stainless steel interior of each bin was Teflon-coated to prevent direct contact and potential

metal transfer between the bin and the peat. The top of each bin was open, exposing the peat to

ambient climate conditions. The mesocosm bins were inserted into a climate-controlled tunnel

and insulated on the sides, which allowed belowground access to one face of each of the bins as

well as to facilitate a vertical temperature gradient, as would be observed in a natural peatland

profile. Potvin et al. (2015) provides a detailed and comprehensive overview of the PEATcosm

experiment including the peat harvest, experimental design and treatment maintenance. The

PEATcosm study was a full-factorial experimental design with two water table (WT)

prescriptions crossed with three vascular plant functional group treatments in a randomized

complete block design, with four replicates per treatment combination, resulting in a total of 24

experimental units (Table 2-1). The plant functional group treatments were designed to

distinguish between effects of Ericaceae (non-aerenchymous shallowly-rooted shrubs with

enzymatically competent mycorrhizal root symbionts) and sedges (aerenchymous herbaceous

graminoids with non-mycorrhizal roots that penetrate deeper into the saturated peat).

Table 2-1 PEATcosm full-factorial experimental design. N=4 mesocosm bins per crossed

treatment.

WATER TABLE (WT) POSITION

PLANT

FUNCTIONAL

GROUP

HIGH WT LOW WT

SEDGE ONLY High WT Sedge Low WT Sedge

ERICACEAE ONLY High WT Ericaceae Low WT Ericaceae

CONTROL

(UNMANIPULATED) High WT Control Low WT Control

The water table treatments were based on long-term (approximately 50 years) data from the

Marcell Experimental Forest in north-central Minnesota (47.51907° N, 93.45966° W), located

50

near the peat monolith harvest site. The two target water table seasonal profiles were modelled

after typical variability, average water table years (‘high WT’) and typical high variability, low

water table years (‘low WT’). The mean difference between the high and low WT positions was

approximately 20 cm throughout the experiment. These target water table profiles were

maintained via a combination of artificial precipitation additions, rain-exclusion covers, and

regulated outflow (spring-only, from ~25 cm depth, roughly at the acrotelm-catotelm boundary)

from each of the bins (Potvin et al., 2015). Water table manipulations were performed only

during the summer months and water table positions were allowed to stabilize throughout the

winter months. Figure 2-1 displays the mean water table positions for the low and high WT

treatments as well as the precipitation throughout the course of this study from 2013 to 2015.

Figure 2-1 Mean water table positions (in cm below the peat surface) in the low and high WT

treatment mesocosms and precipitation (in mm) over the course of this study from 2013 to 2015.

Water table manipulations were conducted in the summer months, while water table levels were

left to stabilize during the winter months. Dashed lines around the low and high WT treatment

mean water table positions represent the 95% confidence interval. Pore water and snowmelt

sampling events are denoted.

51

The three vascular plant functional group treatments simulated anticipated climate change-

induced community composition (Chapin et al., 1996; Weltzin et al., 2000; Strack et al., 2006)

and included 1) sedge only (all Ericaceae removed), 2) Ericaceae only (all sedge removed), and

3) unmanipulated control (includes both sedge and Ericaceae) (see Figure A-2-1 for

photographs). The plant functional group treatments were initiated in June 2011 and maintained

weekly by clipping the stems of the excluded species. The dominant sedge species present was

Carex oligosperma Michx., while the dominant ericaceous shrubs included Chamaedaphne

calyculata (L.) Moench., Kalmia polifolia Wangenh., and Vaccinium oxycoccos L.. The mosses

Sphagnum rubellum Wilson, S. magellanicum Brid., S. fuscum (Schimp.) Klinggr and

Polytrichum strictum Brid. comprised the dominant bryophyte species in all of the 24

mesocosms. To a lesser extent, P. commune Hedw., Eriophorum vaginatum L., Andromeda

polifolia L. var. glaucophylla (Link) DC., Rhododendron groenlandicum (Oeder) Kron and Judd,

and Drosera rotundifolia L. were also present.

2.3.2 Water Sampling

Pore waters were sampled repeatedly throughout the 2013, 2014 and early 2015 growing seasons

to monitor the influence of the simulated climate change effects on Hg and MeHg mobility. Pore

waters were collected from a micro-piezometer nest (ultra-high-density polyethylene casing with

Teflon tubing) installed in each of the 24 mesocosms at three depths – 20 cm, 40 cm and 70 cm

below the peat surface (see Romanowicz et al. (2015) for details on piezometer construction).

Sampling was conducted in June, August and November 2013, May, July and September 2014,

and in May 2015 prior to the mid-summer destructive harvest of the mesocosms at the

conclusion of the experiment. Due to low mid-summer water tables preventing the collection of

samples for analysis from the 20 and 40 cm depths in the Low WT bins, only the spring and fall

data is considered throughout this analysis (see Text A-2-1 in Appendix A for full details).

Spring snowmelt runoff was also collected in 2014 and 2015 to examine impacts on the potential

export of Hg from peatlands to downstream aquatic ecosystems. Outflow or runoff from each

mesocosm was initiated by sufficiently high water tables sensed via a pressure transducer and

controlled via an outflow line located at about 25 cm depth below the peat surface (approximate

acrotelm-catotelm boundary), draining into plastic tubs in the belowground tunnel. Large rain

events and spring snowmelt triggered the production of runoff. However, only runoff produced

as a result of snowmelt was collected for the purposes of this study.

52

Pore water samples were collected from the micro-piezometers using a portable peristaltic pump

equipped with a Teflon line rinsed with dilute hydrochloric acid (HCl) prior to sampling. Acid-

cleaned Teflon in-line filtering units (Savillex, Eden Prairie, MN) were attached to the pump line

to filter all water samples to a level of 0.7 μm using ashed glass fiber filters (Lewis and Brigham,

2004; Shanley et al., 2008). Samples were collected in new polyethylene terephthalate glycol

(PETG) bottles, acidified to 0.5% by volume with trace metal grade HCl and stored at 4°C in

darkness until analysis. Method blanks were collected by running deionized water through the

cleaned pump line and filter units during each sampling event to ensure the cleanliness of the

sampling equipment and sample handling. Mesocosm runoff was similarly collected during the

snowmelt period from each bin outflow line, preserved and stored until analysis. Ultra-clean

trace metal techniques were followed during sample collection, laboratory handling and analyses

(Shanley et al., 2008).

Pore waters from 20, 40 and 70 cm depths and mesocosm runoff were individually analyzed for

both total Hg (THg) and MeHg concentrations. Ancillary chemical analyses including dissolved

organic carbon (DOC), chloride, sulfate and total phenolic concentrations were determined for

pore water and runoff samples collected at the same time as those for Hg analyses.

2.3.3 Pre-Treatment Peat

Peat was collected from each of the 24 mesocosms in 2011 prior to the initiation of the water

table and plant functional group prescriptions to assess any pre-treatment trends in solid phase

THg and MeHg concentrations among the monoliths. Peat cores were collected with a stainless

steel corer from the surface to 60 cm depth and were sectioned in 10 cm increments.

2.3.4 Peat Decomposition Assays

To assess peat decomposition potential for each of the crossed water table and plant functional

group treatments, wooden dowels with strips of cellulose filter paper (pre-dried at 55°C and

weighed) attached at 10 cm increments using heat shrink tubing and encased in Nylon mesh

netting (20x20, 67% open area) were installed to a depth of 80 cm. Two rods were deployed in

each of the peat monoliths for the duration of the 2014 growing season (n=48 total). They were

removed and kept frozen until processed. Peat adhering to the rods upon removal was gently

53

rinsed off. The cellulose strips were removed, dried at 55°C, and weighed to determine the

percentage mass lost in relation to the initial weight.

2.3.5 Analytical Methods

Total Hg concentrations of the pore water and runoff samples were determined with a Tekran

Model 2600 automated Total Mercury Analyzer using cold vapor atomic fluorescence

spectroscopy (CVAFS) according to US EPA Method 1631 (US EPA Method 1631, 2002).

Recovery of a THg spike was 94.5 ± 7.9% (mean ± standard deviation, n=33), replication of

duplicates was 1.7 ± 1.4% (n=32) and the detection limit, calculated as three standard deviations

of matrix blanks, was 0.25 ng L-1

(n=140). Freeze-dried peat samples were digested in hot nitric

acid and diluted digestates were similarly analyzed by CVAFS. Methylmercury analysis was

conducted by isotope dilution-gas chromatography-inductively coupled plasma mass

spectrometry (ID-GC-ICPMS; Hintelmann and Evans, 1997) on samples, both water and solid

phase peat, that were distilled in Teflon vessels according to US EPA Method 1630 (US EPA

Method 1630, 1998). During the distillation process a trace amount of enriched stable Me199Hg

isotope was added to each sample as an internal standard (Hintelmann and Evans, 1997).

Recovery of standard reference material (estuarine sediment ERM CC580) was 99.8 ± 8.6%

(n=50), replication of duplicates was 5.6 ± 4.9% (n=45) and the MeHg detection limit was

calculated (n=48 matrix blanks) to be 0.04 ng L-1

. Field blanks were below detection limits for

both THg and MeHg concentrations.

Sulfate and chloride were determined with an ICS-2000 ion chromatograph with an IonPac AS11

separator column (Dionex Corporation, Bannockburn, IL, USA). Sulfate and chloride pore water

analyses were conducted only for the 2013 and 2014 growing seasons. Dissolved organic carbon

was determined from piezometer samples filtered to 0.45 μm, acidified to pH 2, and analyzed

with a TOC-V Analyzer (Shimadzu Scientific Instruments, Columbia, MD, USA). Pore water

collection and analysis for total phenolics by microplate technique (with tannic acid standard)

were as previously described in Romanowicz et al. (2015).

2.3.6 Statistical Analyses

All statistical analyses were performed using R statistical software (R Development Core Team,

2014) with α = 0.05. The influence of water table regime and vascular plant functional group on

54

THg and MeHg concentrations, as well as %MeHg over the three sampling depths of pore waters

collected over the course of the experiment, was assessed using repeated measures analyses of

variance (ANOVA); with water table, plant functional group and sampling depth treated as main

factors and sampling event as the repeated factor. A repeated measures ANOVA was similarly

performed for the snowmelt runoff Hg concentration data collected from the bins over the two

sampling years. One-way ANOVAs were performed on the pre-treatment solid phase peat THg

and MeHg concentrations individually at 10-30 cm and 30-50 cm to assess any significant

differences among assigned treatment bins prior to implementation of the treatments. These peat

depths were selected for statistical analysis to coincide with the pore water sampling depths of 20

and 40 cm. Additional details regarding statistical analyses are provided in the Supplementary

Information (Text A-2-1).

2.4 Results and Discussion

Both THg and MeHg concentrations in pore waters and snowmelt runoff were significantly

affected by the experimental manipulation of water table (both p < 0.0001) and vascular plant

functional group (both p < 0.0001), with no significant interaction (p = 0.11 to 0.35; Figure 2-2)

between these factors indicating that changes in either are likely to have an impact on Hg

accumulation in pore waters and subsequently, for mobility to downstream regions. Specifically,

lower, more variable water table regimes significantly increased pore water and snowmelt runoff

THg and MeHg concentrations, as did the change to a sedge-dominated vascular plant

community composition with no Ericaceae shrubs present. The combined low WT – sedge only

treatment resulted in the highest pore water and runoff concentrations of both THg and MeHg

(Figure 2-2). One-time mass transfer loads were determined for the 2014 spring snowmelt period

by multiplying the THg and MeHg concentrations of the runoff by the amount of water drained

from the mesocosms acrotelm-catotelm boundaries (Table A-2-1). The trends in THg and MeHg

loads among the six crossed WT and vascular plant functional group treatments (Figure A-2-2)

were consistent with those observed for THg and MeHg pore water and runoff concentrations

(Figure 2-2). No significant, systematic differences were observed in solid phase peat THg or

MeHg concentrations at either the 10 – 30 cm or 30 – 50 cm depth intervals as these values were

quite variable among the six assigned experimental treatment combinations prior to the

experiment (Table 2-2). The differences observed in pore water and runoff THg and MeHg

concentrations can therefore be confidently attributed to the experimental manipulation.

55

Figure 2-2 Total Hg and MeHg concentrations and %MeHg in pore water and snowmelt runoff

in relation to water table and vascular plant functional group manipulations (treatment means of

all depths and sampling events). (a) pore water THg, (b) pore water MeHg, (c) pore water

%MeHg, (d) snowmelt runoff THg, (e) snowmelt runoff MeHg, and (f) %MeHg in snowmelt

runoff. Letters denote statistically similar groups based on transformed data. No significant

differences were observed in snowmelt runoff %MeHg across treatments. Significance of

treatment effects both individual (water table “WT” and plant functional group “Veg”) and

interactive (“WT*Veg”) for THg, MeHg, and %MeHg for both pore waters and snowmelt runoff

are noted. n = 318 pore water samples total (n = 50–56 per treatment).

56

Table 2-2 Mean ± standard deviation THg and MeHg concentrations (in ng g-1

) in solid-phase

peat prior to experimental manipulations (n = 8 samples per treatment and depth increment).

THg Concentration (ng g-1

) MeHg Concentration (ng g-1

)

10 – 30 cm 30 – 50 cm 10 – 30 cm 30 – 50 cm

High WT

Ericaceae 91.8 ± 63.4 87.8 ± 38.4 5.4 ± 6.6 5.9 ± 5.7

High WT

Sedge 82.8 ± 27.0 104.7 ± 40.5 5.0 ± 3.6 8.1 ± 6.7

High WT

Control 80.9 ± 33.3 91.3 ± 21.3 4.5 ± 6.0 5.8 ± 4.1

Low WT

Ericaceae 65.3 ± 15.5 103.3 ± 28.2 1.0 ± 0.7 6.6 ± 6.4

Low WT

Sedge 92.0 ± 33.3 83.2 ± 22.6 5.6 ± 3.9 3.5 ± 1.8

Low WT

Control 75.9 ± 23.8 100.2 ± 28.9 2.8 ± 3.0 6.8 ± 5.9

For both THg and MeHg pore water concentrations, the effect of water table significantly

interacted with both the depth (p = 0.0002 to 0.008) and the timing of the sampling event (p <

0.0001 to 0.002). Total Hg and MeHg concentrations decreased from the 20 cm through to the 70

cm sampling depths in the low WT bins whereas THg and MeHg concentrations at the three

sampling depths were not significantly different from one another in the high WT bins. Greater

fluctuations in water table position in the low WT bins (mean position approximately 35 cm

below the peat surface with a standard deviation of 12 cm), likely resulted in enhanced leaching

and mobility of Hg into the pore waters due to greater aerobic decomposition of the surface peat

layers, and may account for the observed differences in Hg concentrations with depth. These

differences are not the result of evaporative concentration of Hg under low water table conditions

as there was no significant difference in chloride pore water concentrations between the two

water table treatments (p = 0.76). In contrast, the comparatively minimal water table variability

in the high WT bins (mean position approximately 14 cm below the peat surface with a standard

deviation of 5 cm) reduced this mobilization, likely as a result of reduced leaching, yielding both

57

lower and more temporally consistent pore water THg, MeHg and DOC concentrations at all

sampling depths. In association with the effect of WT position, THg and MeHg concentrations

were slightly higher during the early growing season sampling events as compared to the fall

samplings. This may be due to considerable water table variability as a result of draining at the

beginning of each growing season to re-establish the water table treatments following spring

snowmelt. Despite this seasonal variability, a consistent trend in mean THg and MeHg

concentrations was observed among the six crossed WT and vascular plant functional group

treatments in each of the five sampling events from 2013 through to 2015 (Figures A-2-3 and A-

2-4). A marginally significant interaction (p = 0.02) was observed for THg concentrations only

between the vascular plant functional group and the sampling depth. The influence of vegetation

on Hg mobility appears to be more evident nearer to the rooting zone; particularly with the

removal of Ericaceae shrubs.

Water table position (p < 0.0001) and vascular plant functional group (p < 0.0001) also

significantly affected the percentage of THg found as MeHg in pore waters (Figure 2-2c). The

observed differences in %MeHg may be the result of the greater solubility of MeHg in water as

compared to inorganic Hg (Morel et al., 1998). However, the significant differences in pore

water %MeHg among treatments suggest that differences in Hg methylation may also be caused

by the water table and plant functional group treatments. In contrast to the pore waters, no

significant effect of either water table (p = 0.62) or vascular plant functional group treatment (p =

0.77) was observed for the %MeHg in mesocosm runoff collected during the spring snowmelt

periods. This trend was also consistent over the two sampling periods as no significant direct or

interactive influence of time was observed. There were no significant differences in runoff

%MeHg between any of the six crossed treatments combinations (p = 0.87) (Figure 2-2f). This

suggests that a similar process is responsible for the transport of both THg and MeHg to runoff

during spring snowmelt.

One contributing factor controlling the mobilization and accumulation of THg and MeHg into

peat pore waters as well as transport in snowmelt runoff may be the degradation of peat through

decomposition. Pore water THg and MeHg concentrations were both positively and significantly,

but weakly correlated with DOC concentrations (r = 0.32 for THg, r = 0.30 for MeHg; both p <

0.0001) (Figure A-2-5), which would be expected to similarly increase as peat decomposes. Pore

water DOC concentrations significantly increased in the low WT treatment bins throughout the

58

sampling period (p < 0.001), while those in the high WT treatments did not significantly change

over time. Mean THg and MeHg concentrations as well as %MeHg in the shallow pore waters

(20 and 40 cm below the peat surface) were positively correlated with the amount of cellulose

decomposition (expressed as percent loss) in the top 40 cm of peat, near the mean water table

position in the low WT mesocosms (Figure 2-3). These relationships between pore water THg

and MeHg concentrations and the percentage of mass lost from the cellulose decomposition

assays are significant when all pore water sampling events are considered (Figure A-2-6). The

summary Figure 2-3 is included to demonstrate the role of Ericaceae removal on pore water THg

and MeHg concentrations beyond the influence of peat decomposition. The decomposition data

could only be used in a correlative manner, as it is to be published separately as part of the

PEATcosm experiment. Although the decomposition data was only from 2014, the percentage

of mass lost during this time should still be representative of potential peat decomposition as a

result of the water table and plant functional group treatment effects. The low WT – sedge only

treatment which yielded the highest concentrations of THg and MeHg in pore waters also

exhibited the greatest potential for peat decomposition through the cellulose decomposition

assays. The observed relationships are similar when examining the trends fitted using locally-

weighted scatterplot smoothing (LOWESS; Figure A-2-7). Elevated pore water THg and MeHg

in the sedge treatments under both water table prescriptions resulted in a curved relationship with

potential peat decomposition. This illustrates that the removal of Ericaceae shrubs may further

enhance THg and MeHg accumulation in pore waters beyond that predicted by peat

decomposition alone.

59

Figure 2-3 Relationships between shallow pore water (20 and 40 cm below the peat surface) Hg

and mean % mass loss of cellulose decomposition assays in the top 40 cm of the peat. (a) THg

concentrations, (b) MeHg concentrations, (c) %MeHg. Error bars represent standard deviation.

Decomposition assays were harvested in 2014 and represent potential peat decomposition among

the treatments. Pore water Hg data were averaged by treatment from all five sampling events.

The prolonged periods of water table drawdown sustained in the low WT treatments likely

caused the enhanced observed peat decomposition by increasing the zone of aerated peat,

resulting in greater aerobic respiration as compared to the relatively high water table position in

60

the high WT treatments (Whittington and Price, 2006; Ise et al., 2008; Waddington et al., 2014).

The degradation of soil organic matter leads to increased DOC mobilization and accumulation in

pore waters as reflected by higher concentrations in the mesocosm pore waters, with repeated

fluctuations in water table position resulting from precipitation inputs likely leaching carbon

from the peat into the pore waters following periods of drawdown (Fenner et al., 2007). This

may also account for the greater mobilization of THg and MeHg at the upper depths of the peat

monoliths, with higher pore water THg and MeHg concentrations in the 20 and 40 cm samples

within the zone of fluctuation of the low WT treatments. Significant, but relatively weak,

positive correlations with total phenolics in pore water were observed for MeHg (r = 0.34 p <

0.0001) and THg (r = 0.32; p < 0.0001) (Figure A-2-8). Romanowicz et al. (2015) found a

positive correlation between pore water total phenolics and peroxidase enzyme activity.

Oxidative enzymes including peroxidase and phenol oxidase contribute to the degradation of

lignin and polymeric complexes (Romanowicz et al. 2015). As pore water concentrations of

THg and MeHg increased with total phenolics, which were positively correlated with peroxidase

enzyme activity in the PEATcosm pore waters (Romanowicz et al. 2015), organic matter

decomposition was likely the source of increased pore water Hg mobility (Figure A-2-8).

Mercury forms complexes with organic matter (Driscoll et al., 1995) predominantly with thiol

moieties (Graham et al., 2012). Therefore, enhanced peat decomposition and degradation are

likely responsible for much of the observed increase of both THg and MeHg, as well as DOC

accumulation in the mobile pore water phase. Given that solid phase peat THg and MeHg

concentrations among bins were not significantly different prior to initiation of the experiment,

and THg concentrations in particular are not expected to change significantly over time, the

observed pore water THg and MeHg trends are a clear indication that Hg mobility has increased.

In addition to the peat decomposition effect on Hg and MeHg concentrations in pore water, the

water table manipulations may also act to enhance MeHg production in the peat. Prolonged

periods of water table drawdown followed by rewetting of the peat through precipitation input

may act to reset the oxidation-reduction conditions in the zone of aeration (Coleman Wasik et al.,

2015). Mercury methylation is a redox-sensitive process, facilitated by a diverse set of anaerobic

bacteria and Archaea (Gilmour et al., 2013). The periodic oxidation of peat through water table

variability may stimulate MeHg production via regeneration of the oxidized forms of terminal

electron acceptors, such as the oxidation of sulfide to regenerate sulfate (Coleman Wasik et al.,

61

2015). The highest sulfate concentrations in pore waters were indeed observed in the low WT

treatment mesocosms (Figure A-2-9).

In addition to the clear influence of the hydrological treatments, the dominant vascular plant

functional group also significantly affects the accumulation of Hg in pore waters. The

correlations between Hg concentrations and decomposition demonstrate the likely influence of

vascular plant functional groups affecting Hg mobility by mechanisms other than decomposition.

This is exemplified by the increased explanation of variability in the relationships when the

sedge only treatments for both water table prescriptions are removed (r = 0.96 for THg, r = 0.83

for MeHg, r = 0.49 for %MeHg). The removal of Ericaceae and the establishment of sedges may

lead to greater Hg methylation and the amount of Hg mobilized in pore waters, as sedge roots

exude labile carbon compounds and oxygen in the rhizosphere (Crow and Wieder 2005;

Bridgham et al. 2013). Under both water table treatments, the sedge-dominated peat monoliths

had the highest concentrations of THg and MeHg in pore water (Figure 2-2a, b), higher than may

be explained by potential peat decomposition and leaching into pore waters (Figure 2-3). Sedges

increase peat aeration and subsequent oxidation via their aerenchyma tissues (Bridgham et al.,

2008; Waddington et al., 2014) and may, similarly to the water table fluctuation, prime Hg

methylation by stimulating the regeneration of the terminal electron acceptors required for

methylating microbial communities (Coleman Wasik et al. 2015). The leakage of oxygen from

the roots of salt marsh sedges has been observed to stimulate aerobic respiration by acting as a

terminal electron acceptor as well as facilitating anaerobic respiration by regenerating alternative

electron acceptors including nitrate, ferric iron and sulfate (Mueller et al., 2016). Pore water

sulfate concentrations were slightly higher for the sedge-dominated treatments subjected to both

water table prescriptions (Figure A-2-9). Thus, in addition to supplying the necessary terminal

electron acceptors for soil organic matter decomposition (Mueller et al. 2016), sedges may prime

MeHg production through the provision of oxygen and labile organic compounds via root

exudation to the rhizosphere, which are required for methylation to occur (Mitchell et al. 2008a).

Conversely, the presence of Ericaceae shrub roots in the Ericaceae only and unmanipulated

control treatment bins (both of which had lower Hg concentrations than the sedge treatment) may

compete with the heterotrophs for available oxygen (Romanowicz et al., 2015), thereby reducing

sulfate regeneration (slightly, although not significantly; Figure A-2-9) and suppressing MeHg

production. Ericaceous shrubs form a symbiotic relationship with mycorrhizal fungi, which may

62

mediate changes in soil microbial communities, compete for available oxygen thereby limiting

terminal electron acceptor regeneration and decrease peat decomposition through the suppression

of heterotrophs. This has been demonstrated in the PEATcosm experimental monoliths wherein a

negative correlation was observed between Ericaceae biomass and peroxidase as well as β-

glucosidase enzyme activity (Romanowicz et al., 2015). Inhibition of MeHg production in the

upper rooting zone depths of the Ericaceae only as well as the unmanipulated control mesocosms

as compared to the sedge only monoliths may contribute to the peat pore water Hg concentration

differences observed.

The trends in THg and MeHg concentrations in snowmelt runoff are similar to those observed in

the pore waters (Figure 2-2). Snowmelt runoff THg and MeHg concentrations both exhibit a

significant (p < 0.001), positive correlation with DOC concentrations (r = 0.68 for THg, r = 0.66

for MeHg; Figure 2-4). Increased Hg export in association with organic carbon has similarly

been observed in other systems (St. Louis et al., 1994; Driscoll et al., 1995). The statistically

similar %MeHg across the treatments along with the positive correlation with DOC

concentrations provides further support for peat degradation being the principal driving factor

controlling not only the accumulation of Hg in pore waters, but also its transport within and

potential export from peatlands. The concentrations of THg and MeHg observed in snowmelt are

likely the result of peat decomposition liberating Hg in association with organic matter, which

has accumulated in pore waters throughout the growing season and flushed from the peat during

the spring snowmelt period.

63

Figure 2-4 Hg-DOC relationships in 2014 and 2015 snowmelt runoff for all 24 mesocosms (n =

48 total). (a) THg in relation to DOC concentrations and (b) MeHg in relation to DOC

concentrations. Data are plotted on log-transformed axes.

2.5 Conclusions

Climate-induced changes in the hydrological regime and vegetation communities of peatland

systems have the potential to significantly enhance both THg and MeHg accumulation in peat

pore water and transport downstream during spring snowmelt. Although peatlands are strong

sinks of inorganic Hg in the landscape, peatlands are also well-documented sources of MeHg (St.

Louis et al., 1994). By potentially increasing the amount of MeHg exported to aquatic

ecosystems for uptake into the food chain, as well as the amount of inorganic Hg available for in-

lake methylation, wildlife and human populations which rely upon fish consumption for

subsistence may be exposed to increased Hg levels (Scheuhammer et al., 2007; Mergler et al.,

64

2007). Given that peatlands cover approximately 3% of the global land area and are located

predominantly in the subarctic and boreal regions which are vulnerable to the effects of climatic

change (Bridgham et al., 2008), significant stores of Hg sequestered with vast stocks of carbon

(Grigal, 2003) may be liberated from either impacts on water table regime, plant community

changes, or a combination of both.

The small scale of the peat monoliths in this study facilitated a direct assessment of the influence

of water table position and vascular plant functional groups on Hg mobility by removing

potential confounding factors such as heterogeneities in microtopography and hydrological flow

paths and sources. At larger scales, spatial heterogeneities in Hg methylation and accumulation

have been observed, including MeHg hot spots at the outer edges of peatlands, which tend to

increase MeHg export from peatland systems (Mitchell et al., 2008b). The results of this study

are most applicable to the predominant interior peatland area and suggest that higher MeHg

concentrations observed at the outer edges of peatlands may be even further enhanced by climate

change-driven increases in Hg methylation and/or mobility from the peatland interior. Certainly,

results from this experimental work may be scaled to larger peatland ecosystems and may aid in

developing modeling approaches to forecast the response of peatlands under future climate

change scenarios. Overall, the results of this study suggest that deeper, more fluctuating water

tables and shifts toward sedge-dominated plant functional groups will measurably enhance Hg

and MeHg concentrations in peatland pore waters and likely lead to greater Hg and MeHg export

to downstream ecosystems during spring snowmelt.

2.6 Acknowledgements

We would like to acknowledge the laboratory assistance of P. Huang, K. Ng, R. Co and B.

Perron as well as the field assistance of K. Ng, S. Rao, K. Griffith, and E. Grupido. We thank T.

Veverica for pore water carbon chemistry data. Funding was provided through a Natural

Sciences and Engineering Research Council of Canada (NSERC) Alexander Graham Bell

Canada Graduate Scholarship (CGS-Doctoral) to K.M.H and a NSERC Discovery Grant to

C.P.J.M. The PEATcosm experiment was funded by the USDA Forest Service Northern

Research Station Climate Change Program and the National Science Foundation (DEB-

1146149).

65

2.7 References

Armstrong, W., S. H. F. W. Justin, P. M. Beckett, and S. Lythe (1991), Root adaptation to soil

waterlogging. Aquatic Botany, 39, 57–73.

Bergman, I., K. Bishop, Q. Tu, W. Frech, S. Åkerblom, and M. Nilsson (2012), The influence of

sulphate deposition on the seasonal variation of peat pore water methyl Hg in a boreal

mire. PLoS ONE, 7(9), e45547.

Bishop, K., Y.-H. Lee, C. Pettersson, and B. Allard (1995), Methylmercury output from the

Svartberget catchment in northern Sweden during spring flood. Water Air and Soil

Pollution, 80, 445-454.

Branfireun, B. A., A. Heyes, and N. T. Roulet (1996), The hydrology and methylmercury

dynamics of a Precambrian Shield headwater peatland. Water Resources Research, 32,

1785-1794.

Branfireun, B. A., D. Hilbert, and N. T. Roulet (1998), Sinks and sources of methylmercury in a

boreal catchment. Biogeochemistry, 41, 277-291.

Breeuwer, A., B. J. M. Robroek, J. Limpens, M. M. P. D. Heijmans, M. G. C. Schouten, and F.

Berendse (2009), Decreased summer water table depth affects peatland vegetation. Basic

Applied Ecology, 10, 330-339.

Bridgham, S. D., C. A. Johnston, J. Pastor, and K. Updegraff (1995), Potential feedbacks of

northern wetlands on climate change. BioScience, 45, 262–274.



Bridgham, S. D., H. Cadillo-Quiroz, J. K. Keller, and Q. Zhuang (2013), Methane emissions

from wetlands: biogeochemical, microbial, and modeling perspectives from local to

global scales. Global Change Biology, 19, 1325-1346.

66

Chapin, F. S., M. S. Bret-Harte, S. E. Hobbie, and H. Zhong (1996), Plant functional types as

predictors of transient responses of arctic vegetation to global change. Journal of

Vegetation Science, 7, 347–358.





doi:10.1002/2015JG002993.

Crow, S. E. and R. K. Wieder (2005), Sources of CO2 emission from a northern peatland: root

respiration, exudation, and decomposition. Ecology, 86, 1825-1834.




Driscoll, C. T., V. Blette, C. Yan, C. L. Schofield, R. Munson, and J. Holsapple (1995), The role

of dissolved organic carbon in the chemistry and bioavailability of mercury in remote

Adirondack lakes. Water Air and Soil Pollution, 80, 499–508.



peatland catchments. Environmental Science and Technology, 41, 3146-3152.

Gilmour, C. C., E. A. Henry, and R. Mitchell (1992), Sulfate stimulation of mercury methylation

in freshwater sediments. Environmental Science and Technology, 26, 2281-2287.




11810-11820.

Gorham, E. (1991), Northern peatlands: role in the carbon cycle and probable responses to


67

Graham, A. M., G. R. Aiken, and C. C. Gilmour (2012), Dissolved organic matter enhances

microbial mercury methylation under sulfidic conditions. Environmental Science and

Technology, 46, 2715-2723.

Graham, A. M., G. R. Aiken, and C. C. Gilmour (2013), Effect of dissolved organic matter

source and character on microbial Hg methylation in Hg-S-DOM solutions.




Groisman, P. Y., R. W. Knight, D. R. Easterling, T. R. Karl, G. C. Hegerl, and V. N. Razuvaev

(2005), Trends in intense precipitation in the climate record. Journal of Climate, 18,

1326–1350.

Haynes, K. M. and C. P. J. Mitchell (2012), Inter-annual and spatial variability in hillslope runoff

and mercury flux during spring snowmelt. Journal of Environmental Monitoring, 14,

2083-2091.

Hintelmann, H. and R. D. Evans (1997), Application of stable isotopes in environmental tracer

studies-Measurement of monomethylmercury (CH3Hg+) by isotope dilution ICP-MS and

detection of species transformation. Fresenius Journal of Analytical Chemistry, 358, 378

– 385.



763-766.



Journal, 65, 897-905.

Kunkel, K. E., D. R. Easterling, K. Redmond, and K. Hubbard (2003), Temporal variations of

extreme precipitation events in the United States: 1895–2000. Geophysical Research

Letters, 30 (17), 1900, doi:10.1029/2003GL018052.

68

Lewis, M. E. and M. E. Brigham (2004), Low-level mercury. In: Wilde, F.D., Radtke, D.B.,

Gibs, J., Iwatsubo, R.T. (Eds.), Processing of Water Samples. US Geological Survey

Techniques of Water-Resources Investigations, Book 9, Chapter A5. US Geological

Survey, Reston, VA.



global implications – a synthesis. Biogeosciences, 5, 1475-1491.




Mitchell, C. P. J., B. A. Branfireun, and R. K. Kolka (2008a), Assessing sulfate and carbon



Mitchell, C. P. J., B. A. Branfireun, and R. K. Kolka (2008b), Spatial characteristics of net

methylmercury production hot spots in peatlands, Environmental Science and

Technology, 42, 1010–1016.

Mitchell, C. P. J., B. A. Branfireun, and R. K. Kolka (2008c), Total mercury and methylmercury

dynamics in upland-peatland watersheds during snowmelt. Biogeochemistry, 90, 225-

241.

Morel, F. M. M., A. M. L. Kraepiel, and M. Amyot (1998), The chemical cycle and

bioaccumulation of mercury. Annual Review of Ecology and Systematics, 29, 543-566.



Nunes, F. L. D., L. Aquilina, J. de Ridder, A.-J. Francez, A. Quaiser, J.-P. Caudal, P.

Vandenkoornhuyse, and A. Dufresne (2015), Time-scales of hydrological forcing on the

geochemistry and bacterial community structure of temperate peat soils. Scientific

Reports, 5, 14612. DOI: 10.1038/srep14612.

69

Peltoniemi, K., R. Laiho, H. Juottonen, O. Kiikkilä, P. Mäkiranta, K. Minkkinen, T. Pennanen,

T. Penttilä, T. Sarjala, E.-S. Tuittila, T. Tuomivirta, and H. Fritze (2015), Microbial

ecology in a future climate: effects of temperature and moisture on microbial

communities of two boreal fens. FEMS Microbiology Ecology, 91, fiv062.

doi:10.1093/femsec/fiv062.




and Soil, 387, 277-294.

R Development Core Team (2014). R: A language and environment for statistical computing. R

Foundation for Statistical Computing, Vienna, Austria. URL: http://www.R-project.org/.

Romanowicz, K. J., E. S. Kane, L. R. Potvin, A. L. Daniels, R. K. Kolka, and E. A. Lilleskov

(2015), Understanding drivers of peatland extracellular enzyme activity in the

PEATcosm experiment: mixed evidence for enzymic latch hypothesis. Plant and Soil,

397, 371-386.



12-18.

Shanley, J. B., M. A. Mast, D. H. Campbell, G. R. Aiken, D. P. Krabbenhoft, R. J. Hunt, J. F.

Walker, P. F. Schuster, A. Chalmers, B. T. Aulenbach, N. E. Peters, M. Marvin-

DiPasquale, D. W. Clow, and M. M. Shafer (2008), Comparison of total mercury and

methylmercury cycling at five sites using the small watershed approach. Environmental

Pollution, 154, 143–154.

Strickman, R.J.S., R.R. Fulthorpe, J.K. Coleman Wasik, D.R. Engstrom, and C.P.J. Mitchell

(2016), Experimental sulfate amendment alters peatland bacterial community structure.

Science of the Total Environment, doi: 10.1016/j.scitotenv.2016.05.189.

70





Importance of wetlands as sources of methyl mercury to boreal forest ecosystems.


Strack, M., J. M. Waddington, L. Rochefort, and E.-S. Tuittila (2006), Response of vegetation


water table drawdown. Journal of Geophysical Research, 111, G02006.

Thomson, A. M., R. A. Brown, N. J. Rosenberg, R. C. Izaurralde, and V. Benson (2005),

Climate change impacts for the conterminous USA: An integrated assessment. Part 4:

Water resources. Climatic Change, 69, 67–88.




U.S. EPA, Method 1631, Revision E: Mercury in water by oxidation, purge and trap, and cold

vapour atomic fluorescence spectrometry, U.S. Environmental Protection Agency:

Washington, D.C., 2002.

U.S. EPA, Method 1630: Methyl Mercury in Water by Distillation, Aqueous Ethylation, Purge

and Trap, and Cold Vapor Atomic Fluorescence Spectrometry, U.S. Environmental

Protection Agency: Washington, D.C., 1998.



Weltzin, J. F., S. D. Bridgham, J. Pastor, J. Chen, and C. Harth (2003), Potential effects of

warming and drying on peatland plant community composition. Global Change Biology,

9, 141–151.

71



Ecology, 81, 3464-3478.



3589-3600.


Heredia-Middleton, C. Coates, and E. Kakouros (2009), Experimental removal of

wetland emergent vegetation leads to decreased methylmercury production in surface

sediment. Journal of Geophysical Research, 114, G00C05.

Winter, T.C. (2000), The vulnerability of wetlands to climate change: A hydrologic landscape

perspective. Journal of the American Water Resources Association, 36, 305-311.

72

Chapter 3 Gaseous Mercury Fluxes in Peatlands and the Potential Influence

of Climate Change

3

3.1 Abstract

Climate change has the potential to significantly impact the stability of large stocks of mercury

(Hg) stored in peatland systems due to increasing temperatures, altered water table regimes and

subsequent shifts in vascular plant communities. However, the Hg exchange dynamics between

the atmosphere and peatlands are not well understood. At the PEATcosm Mesocosm Facility in

Houghton, Michigan, total gaseous Hg (TGM) fluxes were monitored in a subset of 1-m3 peat

monoliths with altered water table positions (high and low) and vascular plant functional groups

(sedge only, Ericaceae only or unmanipulated control) above the Sphagnum moss layer. At the

SPRUCE bog in north-central Minnesota, TGM fluxes were measured from plots subjected to

deep peat soil warming (up to +9 °C above ambient at a depth of 2 m). At PEATcosm, the

strongest depositional trend was observed with the Low WT – sedge only treatment mesocosms

with a mean TGM flux of −73.7 ± 6.3 ng m-2

d-1

, likely due to shuttling of Hg to the peat at depth

by aerenchymous tissues. The highest total leaf surface and tissue Hg concentrations were

observed with the Ericaceae shrubs. A negative correlation between TGM flux and Ericaceae

total leaf surface area suggests an influence of shrubs in controlling Hg exchange through

stomatal uptake, surface sorption and potentially, peat shading. Surface peat total Hg

concentrations are highest in treatments with greatest deposition suggesting deposition controls

Hg accumulation in surface peat. Fluxes in the SPRUCE plots ranged from −45.9 ± 93.8 ng m-2

d-1

prior to the implementation of the deep warming treatments to −1.41 ± 27.1 ng m-2

d-1

once

warming targets were achieved at depth and +10.2 ± 44.6 ng m-2

d-1

following prolonged deep

soil warming. While these intervals did not differ significantly, a significant positive increase in

the slope of the regression between flux and surface temperature was observed across the pre-

treatment and warming periods. Shifts in vascular vegetation cover and peat warming as a result

73

of climate change may significantly affect the dynamics of TGM fluxes between peatlands and

the atmosphere.

3.2 Introduction

In addition to natural sources such as volcanic emissions, anthropogenic sources including coal

burning, gold mining, and waste incineration have contributed to significantly increase mercury

(Hg) in global terrestrial, aquatic, and atmospheric pools (Selin, 2009). Atmospheric deposition

via either wet or dry processes represents the dominant input of Hg to the Earth's surface

(Fitzgerald et al., 1998; Driscoll et al., 2007). Wetlands are particularly important sinks of

atmospherically-deposited Hg in the landscape (Wallschläger et al., 2000; Zillioux et al., 1993).

Following deposition, Hg is subjected to numerous hydrological and biogeochemical influences,

which might either incorporate it into soil and vegetation stocks (Hintelmann et al., 2002), or

lead to its re-emission from soil, water and/or vegetation back to the atmosphere following

reduction to its elemental form, Hg(0) (Poissant et al., 2004a). Total gaseous mercury (including

all forms of gaseous Hg, the majority of which is typically Hg(0)) fluxes have been monitored in

freshwater wetlands (Poissant et al., 2004a, b; Zhang et al., 2006; Lindberg and Zhang, 2000;

Marsik et al., 2005). However, only a few more recent studies have focused on soil-air exchange

dynamics in boreal peatlands (Kyllönen et al., 2012; Fritsche et al., 2014; Osterwalder et al.,

2016) despite the large amounts of Hg stored in these systems (Grigal, 2003).

Peatlands are immense stores of organic matter in the landscape (Gorham, 1991) and may

represent a significant source of Hg to the atmosphere (Grigal, 2003). Climate change has the

potential to significantly impact the hydrology, biogeochemistry and ecology of peatland

ecosystems due to increasing temperatures and changes in precipitation patterns (Kunkel et al.,

2003; Groisman et al., 2005; Bridgham et al., 2008). Enhanced variability in precipitation, with

less frequent and more intense events anticipated across the northern continental United States

(Kunkel et al., 2003; Groisman et al., 2005), may have direct impacts on the hydrology of

peatlands with the potential for prolonged periods of water table recession and greater variability

throughout the summer months (Thomson et al., 2005; Whittington and Price, 2006). Subsequent

changes in the dominant vascular plant functional groups may occur as a result of prolonged dry

periods and warming temperatures in peatlands (Weltzin et al., 2003; Strack et al., 2006;

Breeuwer et al., 2009; Dieleman et al., 2015). Due to their ability to reach deeper water stores,

74

regulate water loss through stomata and oxygenate the rooting zone by shuttling oxygen via

aerenchymous tissue, sedges may have a competitive advantage over shrub vegetation and

Sphagnum mosses during periods of water stress (Dieleman et al., 2015). However, several

studies (i.e. Weltzin et al., 2000; Breeuwer et al., 2009) including the PEATcosm mesocosm

experiment (Potvin et al., 2015) have observed a shift towards increasing Ericaceae shrub

abundance and productivity under drier conditions. Therefore, plant communities in both bogs

and fens are likely to change with the increasing temperatures and enhanced water stress

resulting from climate change, although the direction of this change is not clear (Weltzin et al.,

2003). Changes in plant community composition in concert with alterations to the hydrology and

redox condition of the peat may have profound effects on peat decomposition and therefore the

carbon storage abilities of peatlands (Waddington et al., 2015). Given the strong relationship

between Hg and organic matter, the potential changes in water table recession and variability,

dominant vascular plant communities, peat mineralization and shifting carbon balances resulting

from hydrological, ecological and biogeochemical changes may affect the dynamic exchange of

total gaseous Hg (TGM) fluxes between the peat and the atmosphere.

Total gaseous Hg exchange with background soils is often well correlated with meteorological

variables (Gustin et al., 2006; Lin et al., 2010). Air and soil temperatures are important

thermodynamic controls on Hg reduction and emission from soils (Edwards et al., 2001), as is

solar radiation because of the importance of photoreduction processes in controlling Hg

emissions (Moore and Carpi, 2005). These factors act to influence Hg emissions by effectively

lowering the activation energy required for Hg reduction (Carpi and Lindberg, 1998). Changes in

soil moisture significantly impact Hg fluxes from mineral soils (Gustin and Stamenkovic, 2005;

Song and Van Heyst, 2005). Wetting of the soil surface by precipitation increases Hg emissions

from initially dry soils, with spikes in fluxes following the rain event and decreasing fluxes as

the soil dries (Gustin and Stamenkovic, 2005; Mazur et al., 2015). The mechanisms governing

this increase with wetting may involve desorption of Hg from the soil particles and subsequent

displacement of the Hg from the soil pores. However, emissions from saturated soils are often

suppressed, potentially due to the very slow diffusion of Hg(0) through water (Bahlmann et al.,

2004; Gustin and Stamenkovic, 2005; Song and Van Heyst, 2005). Enhanced evaporation

coincident with increasing atmospheric temperatures may also affect the dynamic exchange of

Hg by transporting Hg to the soil surface via capillary action (Briggs and Gustin, 2013).

75

Therefore, altered water table position and variability in precipitation patterns may have a direct

impact on seasonal Hg emission and deposition patterns in peatlands resulting from changes in

soil moisture and redox conditions. Assuming similar mechanisms to those observed in mineral

soils (Bahlmann et al., 2004; Gustin and Stamenkovic, 2005; Song and Van Heyst, 2005),

episodic re-wetting of the peat following prolonged water table drawdown may enhance TGM

fluxes from peatlands. Once saturated however, significant emission of Hg to the atmosphere

may be inhibited. Periods of wetting may provide the necessary redox conditions to promote

Hg(II) reduction (Moore and Castro, 2012) to Hg(0) and subsequent release from the peat. Taken

together, the direct and indirect influences of increasing temperatures and altered precipitation

regimes on peatland TGM fluxes clearly warrant further investigation.

The shifting composition of the vascular plant community may also affect the exchange of Hg

through stomata and the potential liberation of Hg from soil via the roots (Rea et al., 2002;

Leonard et al., 1998a, b). Sedges with deep roots and aerenchymous tissues (Breeuwer et al.,

2009) may affect the magnitude and direction of TGM fluxes depending upon root depth and Hg

concentrations within the soil profiles, as has been observed with other wetland vegetation

(Lindberg et al., 2005). Lindberg et al. (2005) determined that the source of TGM emissions

from aquatic macrophytes was the rhizosphere. However, translocation of Hg between the roots

and the foliage of forest canopies is minimal (Ericksen et al., 2003). Although several studies of

soil-air Hg dynamics have been conducted over bare soil (e.g. Carpi and Lindberg, 1997;

Poissant et al., 2004b; Miller et al., 2011), vegetation in both forests and wetland systems plays

an important role in controlling the exchange of Hg with the atmosphere via both stomatal and

non-stomatal pathways including leaf surface sorption (Lee et al., 2000; Rea et al., 2002;

Ericksen et al., 2003; Marsik et al., 2005; Stamenkovic and Gustin, 2009; Laacouri et al., 2013).

Evaporation of morning dew from the surface of vegetation has also been identified as a

potential source of Hg to the atmosphere (Marsik et al., 2005). Stomatal exchange was observed

to be an instrumental process governing net TGM emission from the canopy of a mixed

sawgrass-cattail wetland in the Everglades as Hg emission was coincident with stomatal uptake

of carbon dioxide (Marsik et al., 2005). The manner in which climate change-induced shifts in

vascular plant community composition affects the direction and magnitude of TGM fluxes

between peatlands and the atmosphere remains unclear and therefore requires further study.

76

The overall objectives of this study were to characterize diurnal TGM fluxes in peatlands and to

investigate impacts on TGM fluxes due to changes in vascular plant community composition,

water table position, and peat warming, using two field-scale peatland climate change simulation

experiments. Diurnal fluxes were measured in a subset of the PEATcosm (Peatland Experiment

at the Houghton Mesocosm Facility) peat monoliths, manipulated to simulate climate-induced

changes in vascular plant functional groups and water table position and variability. To assess

the potential effects of soil warming, TGM fluxes were also measured in the SPRUCE (Spruce

and Peatland Responses Under Climatic and Environmental Change) experimental peatland

prior to and throughout a large-scale manipulation of deep (2 m) peat temperatures. It was

hypothesized that 1) peatland plant functional groups would differentially affect TGM fluxes due

to differences in vascular structures 2) increased sedge cover would increase TGM deposition to

peatlands due to incidental shuttling of Hg via aerenchymous tissue and 3) increased soil

temperatures would enhance TGM fluxes from the peat to the atmosphere.

3.3 Methods

3.3.1 PEATcosm Site Description

The PEATcosm experiment was located at the United States Forest Service Houghton Mesocosm

Facility, at the Forestry Sciences Laboratory in Houghton, Michigan, USA (47.11469° N,

88.54787° W). The regional climate is humid continental with typical annual precipitation of

approximately 870 mm. Mean temperatures range from −13 °C in January to +24 °C in July (30

year means at Houghton County Airport; Potvin et al., 2015).

The Mesocosm Facility houses 24 intact 1 m3 (1 m × 1 m × 1 m) peat monoliths open to the air

on top and inserted into a climate-controlled tunnel which allows belowground access to each of

the bins and the simulation of natural peat temperature gradients with depth. The monoliths were

harvested from an ombrotrophic peatland located in Meadowlands, MN, USA in May 2010 and

transferred into individual Teflon-coated, stainless steel mesocosm bins, preventing metal

transfer to the peat. Vegetation manipulations were initiated in 2011, and full water table

manipulations were initiated in 2012. At the conclusion of the experiment in July 2015, the

monoliths were destructively harvested. The full description of the PEATcosm experiment can

be found in Potvin et al. (2015), including peat monolith harvest, facility details and

experimental design.

77

The PEATcosm study comprises a full-factorial experimental design, with two water table (WT)

prescriptions crossed with three different plant functional group treatments, simulating potential

climate change outcomes. Each treatment combination is replicated across four mesocosms, in a

randomized complete block design, for a total of 24 experimental units. The water table

treatments were based on long-term (approximately 50 years) data from the Marcell

Experimental Forest in north-central Minnesota (47.51907° N, 93.45966° W), located near the

peat monolith harvest site. The two WT treatments were modelled after years with: 1) typical

variability and average WT position (referred to as ‘High WT’) and 2) comparably high

variability and low WT position (referred to as ‘Low WT’). These target WT profiles were

maintained through a combination of artificial precipitation additions, rain-exclusion covers and

regulated outflow from approximately the acrotelm-catotelm boundary during the snowmelt

period from each of the bins (see Potvin et al., 2015 for details). The three plant functional group

treatments, simulating likely community composition alterations resulting from climate change

(Strack et al., 2006; Weltzin et al., 2000; Chapin et al., 1996), are as follows: 1) all Ericaceae

removed (referred to as ‘sedge only’ treatment), 2) all sedge removed (‘Ericaceae only’

treatment) and 3) both sedge and Ericaceae present (‘unmanipulated control’ treatment). The

dominant sedge species present in the bins was Carex oligosperma Michx., while the dominant

Ericaceae shrubs included Chamaedaphne calyculata (L.) Moench., Kalmia polifolia Wangenh.,

and Vaccinium oxycoccos L. The dominant moss species in the peat monoliths were Sphagnum

rubellum Wilson, S. magellanicum Brid., S. fuscum (Schimp.) Klinggr and Polytrichum strictum

Brid. Polytrichum commune Hedw., Eriophorum vaginatum L., Andromeda polifolia L. var.

glaucophylla (Link) DC., Rhododendron groenlandicum (Oeder) Kron and Judd, and Drosera

rotundifolia L. were also present in the mesocosms.

3.3.2 SPRUCE Site Description

The SPRUCE experimental peatland is an 8.1 ha ombrotrophic, spruce-Sphagnum bog (known

locally as “S1”), located within the Marcell Experimental Forest (MEF) in north-central

Minnesota (47.51907° N, 93.45966° W). Typical annual precipitation for the MEF is

approximately 780 mm. Mean temperatures at the MEF range from −15 °C in January to +19 °C

in July (Sebestyen et al., 2011). The overstory is dominated by two tree species: Picea mariana

(Mill.) B.S.P and Larix laricina (Du Roi) K. Koch. The understory consists primarily of

Ericaceae shrubs including Chamaedaphne calyculata, Kalmia polifolia, Vaccinium

78

angustifolium Aiton, Vaccinium oxycoccos, and Rhododendron groenlandicum as well as sedges

Eriophorum spp. and the lily Maianthemum trifolium (L.) Sloboda. The dominant bryophyte

present on hummocks is Sphagnum magellanicum, while hollows are mainly colonized by S.

angustifolium (C.E.O. Jensen ex Russow).

The SPRUCE study (http://mnspruce.ornl.gov/) is a climate change field experiment that

involves the long-term (10-year) manipulation of both soil and air temperature, and atmospheric

carbon dioxide (CO2) concentrations within open-top, approximately 12 m diameter, enclosures

along three transects within the S1 bog . Prior to the full initiation of the SPRUCE experiment in

2016, an approximate one-year, deep (2 m below the peat surface) soil warming experiment was

conducted from June 2014 through the end of July 2015. This study describes TGM fluxes

before (May 2014), during (August 2014), and following prolonged deep peat warming (June

2015). The heating infrastructure utilized to warm the peat at depth within the experimental

enclosures is similar to that described in Hanson et al. (2011).

3.3.3 Experimental design and mercury flux measurements

Subsets of the total number of experimental treatments in each experiment were monitored for

TGM flux using dynamic flux chambers (DFCs). Despite their limitations (Wallschläger et al.,

1999; Gillis and Miller, 2000), DFCs were the only appropriate method given the size and scale

of the experimental mesocosms and plots. These measurements were taken over individual 24-h

periods. This was necessary as similar meteorological conditions (i.e. no rain, with similar cloud

cover and incoming solar radiation) could only be achieved within relatively short (4–6 days)

timeframes. Similar ambient conditions were necessary to discern potential treatment effects. At

PEATcosm, TGM fluxes were monitored on a subset of 8 of the mesocosm bins in July 2014 at

the peak of the growing season. Simultaneous measurements were taken on two mesocosms in

each 24-h period. In all, duplicate mesocosms of the following experimental treatment pairings

were monitored for TGM fluxes: High WT – control vegetation, Low WT – control vegetation,

Low WT – Ericaceae only, and Low WT – sedge only. These treatments span the range of

vascular plant functional group treatments with the Low WT prescription representing the

greatest anticipated hydrological changes in peatlands due to climate change. The High WT –

control vegetation treatment acts as the unmanipulated, control scenario. The duplicate

mesocosms for each treatment were not measured during the same 24-h period. Mean water table

79

positions during the flux measurement period were 37 ± 5 cm below the peat surface (mean ±

standard deviation) for the Low WT treatments and 15 ± 0.1 cm below the peat surface for the

High WT treatment.

At the SPRUCE site, TGM fluxes were measured in six experimental enclosures, focusing on

duplicates of deep soil warming treatments with target temperature differentials of +0, +4.5 and

+9 °C above ambient soil temperatures at 2 m depth. These target differentials were achieved at

2 m depth within 60 days of the June 2014 initiation, but were not similarly observed nearer the

surface. Due to the considerable distance between SPRUCE plots and equipment availability,

two replicate locations within each experimental plot were measured simultaneously. Flux

measurements at SPRUCE were performed in May 2014 prior to the initiation of the deep soil

warming treatments, August 2014 once soil warming target temperatures were achieved at depth

and June 2015 following prolonged deep soil warming. Due to an equipment malfunction, one

+0 and one +4.5 °C plot (Plots 4 and 6, respectively) could not be monitored in June 2015.

Ambient TGM fluxes were measured using two transparent Teflon DFCs (0.036 m2 footprint, 6.5

cm height and 2.0 L volume with 1 cm diameter inlet holes every 2.5 cm around the perimeter

and a 5 cm wide bottom flange as per Eckley et al., 2010) placed on the peat surface and

connected to a Tekran 2537A Gaseous Mercury Analyzer. The DFCs were placed on hummock

microtopographic forms within each plot and mesocosm. Standing vegetation was present

beneath the footprint of each DFC. Weighted rings (clear plastic tubing filled with small ball

bearings) were placed on the DFC bottom flange to ensure a tight seal to the peat surface. The

sampling flow rate of the analyzer was 1.5 L min-1

. A soda lime trap was connected in the

sampling inlet line just prior to introduction into the analyzer to prevent passivation of the gold

traps by acid aerosols (Ericksen et al., 2006; Miller et al., 2011). A four-port sampling

manifold/switching unit (Tekran 1115 Synchronized Multi-Port Sampling System) connected to

the analyzer facilitated the alternation of air accumulation among four Teflon lines, an inlet and

outlet for each of the two DFCs, each fitted with a 0.2 μm disposable particulate syringe filter

(PTFE-membrane in a polypropylene housing, Cole-Parmer). The manifold allowed nearly

simultaneous duplication of measurements (Carpi and Lindberg, 1997). One complete flux

measurement for one DFC was achieved every 20 min. Fluxes were calculated using the

equation:

80

𝐹 = 𝑄 ∗ (𝐶𝑜 − 𝐶𝑖)/𝐴 (3.1)

where F is the total flux (ng m-2

h-1

), Q is the chamber flushing flow rate (m3 h

-1), Co is the Hg

concentration measured at the chamber outlet (ng m-3

), Ci is the Hg concentration measured

adjacent to one of the chamber inlet holes (ng m-3

), and A is the footprint area of the chamber

(m2). Quality control of the calculated fluxes was based on the determination of the inter-

cartridge concentration difference for each of the TGM fluxes measured using the Tekran 2537A

according to the method described by Eckley et al. (2011). Calculation of daily TGM fluxes was

achieved by integrating the area under the 24-h curve of all valid measurements for each DFC.

Positive flux values represent emission from the soil surface to the atmosphere, while negative

values signify TGM deposition to the soil.

To ensure consistent conditions in each DFC and its inlet and outlet lines while not being

measured, a secondary bypass vacuum pump was attached to the four-port manifold to draw

ambient air through the lines at the 1.5 L min-1

sampling rate (Mazur et al., 2014). This flushing

flow rate was applied to prevent artificial concentration gradients over the soil and vegetation

surfaces within the DFCs and as a consequence induce higher fluxes due to exceedingly high

flow rates, erring on the side of a lower flow rate as suggested by Eckley et al. (2010). A bubble

flow calibrator (mini-Buck Calibrator M-5) was connected to each of the inlet and outlet lines

prior to each sampling campaign to ensure the rate of air flow facilitated by the secondary pump.

The analyzer was calibrated prior to each 24-h cycle measurement using the internal mercury

permeation source. Each of the two DFCs were soaked in a 10% hydrochloric acid bath for 24 h

and rinsed thoroughly with deionized water prior to sampling. To ensure the cleanliness of the

DFCs prior to and throughout each monitoring campaign, blank measurements were performed

in the field for approximately 24-h by placing each DFC on clean polycarbonate sheeting on the

ground surface adjacent to the plots or mesocosms (0.04 ± 0.16 ng m-2

h-1

, mean ± standard

deviation, n = 123 20-min fluxes) and were not subtracted from flux measurements as is

consistent with other published work (e.g. Eckley et al., 2010; Mazur et al., 2014).

Condensation occurred on the inside of the DFCs during the experiments, which could not be

mitigated through an increase in DFC flushing flow rate. To assess if Hg(0) accumulation

occurred in the condensation, samples were collected by rinsing each DFC following the

measurement cycle using a known volume of deionized water (18 MΩ cm-1

) as has been

81

previously done by Briggs et al. (2012). The collected rinse water was acidified with 0.5% trace-

metal grade HCl and stored in darkness at 4 °C until total Hg analysis. Mean THg concentrations

of these unfiltered DFC condensation rinses were low, representing approximately 1.2 ± 1.2%

(mean ± standard deviation) of the mean daily TGM fluxes.

Incoming solar radiation as well as air and surface peat temperature were recorded at 5-min

intervals on a Campbell Scientific CR1000 data logger during the Hg flux measurements. A

pyranometer (Kipp and Zonen SP-Lite 2 Silicon Pyranometer) was placed mid-way between the

two DFCs being monitored for each 24-h cycle. Surface peat temperature adjacent to each of the

monitored DFCs (to prevent disturbance under the DFC) was measured using temperature probes

(109 Temperature Probes) inserted approximately 2 cm below the peat surface. Ambient air

temperature outside of the chambers was also measured using a shielded thermocouple

throughout the flux measurements.

3.3.4 PEATcosm vegetation and surface peat Hg

To determine the relative influences of TGM uptake through the stomata of vascular vegetation

as compared to sorption to leaf surfaces, foliar rinsing was performed similar to the methods

conducted by Rea et al. (2000), Ericksen et al. (2003) and Laacouri et al. (2013). Following the

completion of the PEATcosm experiment in July 2015, vascular vegetation was collected from

each of the mesocosm bins that were monitored for TGM fluxes. The vegetation from each bin

was separated according to species: Carex oligosperma (n = 5), Eriophorum vaginatum (n = 3),

Kalmia polifolia (n = 6), Chamaedaphne calyculata (n = 6). The leaves of each species collected

from each bin (ranging from approx. 170–500 leaves) were placed in a known volume of

deionized water and gently shaken for five minutes using a horizontal shaker. The rinse solution

was decanted into a new polyethylene terephthalate glycol (PETG) bottle and a second DI water

rinse was sequentially performed on each leaf sample. These rinses were followed by two

sequential rinses using a dilute (pH = 4) trace-metal grade nitric acid solution to ensure all

surface-deposited Hg was removed. All samples were stored in darkness at 4 °C until THg

analysis the following day. Surface Hg loads are expressed both by dry weight of leaves washed

(ng g-1

dry wt.) and by one-sided surface area (ng m-2

) similar to Rea et al. (2000) and Laacouri

et al. (2013). An average leaf surface area for each species was determined by measuring the

length and width of a subset of leaves and assuming the leaves to be oval in shape (one-sided

82

surface area = (½ length)*(½ width)*π). This average was multiplied by the number of leaves

counted for each species to estimate a total area of leaves rinsed. To estimate the amount of Hg

sorbed to the vascular vegetation under each DFC, the number of leaves and an average leaf

length and width for each species present under the footprint of each DFC was counted in the

field following the TGM flux measurement period. The coverage of vascular vegetation was not

consistent across the bins and therefore by counting the leaves under each DFC and applying a

mean leaf area, an estimate of total leaf area (in cm2) could be made. After rinsing, the leaf

samples were frozen and subsequently lyophilized. The dried samples were ground prior to THg

analysis to determine the amount of Hg found within the leaf tissue following surface rinsing as

compared to that resulting from the leaf rinses and therefore sorbed to the leaf surface. To assess

the possible relationship between TGM fluxes and peat Hg concentrations, surface peat (0–10

cm) samples were collected from each of the eight monitored mesocosms using a clean stainless

steel corer. Peat samples were immediately frozen and lyophilized prior to analysis for THg

concentrations.


Total Hg concentrations of the DFC condensation rinses, foliar rinses, lyophilized vegetation and

peat were determined using a Tekran Model 2600 automated Total Mercury Analyzer, using cold

vapor atomic fluorescence spectroscopy (CVAFS) according to US EPA Method 1631. Freeze-

dried vegetation and peat samples were digested in hot nitric acid. All Hg in the condensation

and foliar rinses as well as the diluted peat and vegetation digestates was oxidized by reaction

with bromine monochloride (BrCl) overnight prior to analysis. Recovery of a THg spike was 96

± 5% (mean ± standard deviation, n = 6), replication of duplicates was 2.0 ± 1.6% (n = 6) and the

detection limit for water samples, calculated as three standard deviations of matrix blanks, was

0.16 ng L-1

(n = 23). Recovery of standard reference material (MESS-3) following digestion was

97 ± 2% (n = 2). The detection limit for solid material was 0.01 ng g-1

(n = 6).



2014) with α = 0.05. For the PEATcosm measurements, only a qualitative assessment of TGM

flux trends among the four crossed water table and plant functional group treatments were

considered since only two replicate mesocosms per treatment were measured and statistical

83

power of duplicates is low. Correlative relationships using Pearson correlation were investigated

between hourly TGM fluxes and incoming solar radiation as well as surface peat soil

temperatures among the four treatments. Only non-zero solar radiation values (i.e. daylight) were

included in these correlative relationships. Further examination of the relationships between

TGM fluxes and surface peat temperatures was performed using piecewise regressions. The

correlative relationships between daily TGM fluxes and surface peat THg concentrations, total

leaf area beneath each DFC, and Hg sorbed to the vegetation surfaces in each of the eight

monitored mesocosms were also examined. The relationship between the number of sedge stems

present beneath each DFC and the daily TGM fluxes was investigated collectively for the

PEATcosm and SPRUCE sites. Significant differences in THg concentrations on the surface of

the different vascular plant species, as well as that within the vascular plant tissues were assessed

using a one-way analysis of variance (ANOVA), followed by post-hoc testing with Tukey

adjustments. Differences in daily TGM fluxes among SPRUCE treatment plots throughout the

course of the deep soil warming experiment were assessed using regression analysis with mean

daily surface peat temperatures for each of the three sampling campaigns. Significant changes in

the slope of this regression over the three sampling events of the deep warming experiment were

assessed using an analysis of covariance (ANCOVA) with the surface peat temperature as the

covariate.

3.4 Results

3.4.1 PEATcosm – water table and plant functional groups influence

The crossed water table position and vascular plant functional group treatments influenced the

dynamic exchange of TGM between the peat monoliths and the atmosphere (Figure 3-1).

Consistent evasion of TGM from the peat was observed in both of the High WT control

mesocosms over the course of the 24-h periods with a mean (±standard deviation) daily flux of

+35.2 ± 10 ng m-2

d-1

. In contrast, strong deposition of TGM to the peat was observed in the Low

WT sedge only bins, with a mean daily TGM flux of −73.7 ± 6.3 ng m-2

d-1

. Total gaseous Hg

was deposited to one of the Low WT Ericaceae only bins (−69.1 ng m-2

d-1

), while slight

emission of TGM was observed from the other bin (+4.2 ng m-2

d-1

). Two contrasting flux

directions were also observed for the Low WT control mesocosms with one strongly emitting

TGM to the atmosphere (+144.5 ng m-2

d-1

) while TGM deposition occurred in the other

84

mesocosm (−35.2 ng m-2

d-1

; Figure 3-1). There was a significant, positive, but weak correlation

(r = 0.32, p < 0.001) between the mean inlet TGM concentrations and the corresponding 20-min

TGM fluxes measured during daylight hours across the treatments (Figure B-3-1).


d-1

) across the four crossed water table

(WT) and plant functional group treatments (8 mesocosms in total monitored). * signifies that

sedge vegetation was present beneath the footprint of the DFC placed on the peat surface.

Sedges were present beneath the DFC placed on the peat surface of the Low WT – control

monolith exhibiting deposition throughout the 24-h monitoring cycle. Although sedges were

present in the other monitored Low WT – control mesocosm, which emitted TGM to the

atmosphere over a 24-h period, no sedge stems were located beneath the footprint of the DFC

(Figure 3-1). No sedges were present beneath the footprints of the DFCs for either of the High

WT – control mesocosms. A significant negative correlation (r = - 0.91, p < 0.001) was observed

between the TGM fluxes measured at both the PEATcosm and SPRUCE sites and the number of

sedge stems present beneath the DFC (Figure 3-2). Stronger depositional fluxes were observed as

the number of sedges under the DFC increased across the plots at both peatland sites.

85

Figure 3-2 Relationship between daily TGM fluxes (ng m-2

d-1

) and the number of sedge stems

present beneath the DFC at both the PEATcosm (closed red circles) and SPRUCE (open black

circles) sites.

The trend and magnitude of incoming solar radiation were similar across the four measurement

days (Figure 3-3), with no significant differences in radiation between the treatments (p = 0.92).

Therefore, the trend in TGM fluxes among the treatments is not the result of differences in

incoming solar radiation. Diel variation was observed across the treatments with early-afternoon

peaks in TGM fluxes coincident with peaks in solar radiation and surface soil temperatures

(Figure 3-3). Daylight solar radiation is strongly correlated with TGM flux, although to varying

degrees depending upon the treatment (Figure 3-4b). Similarly, positive, second-order

polynomial correlations were also observed between the hourly TGM fluxes and surface peat

temperatures (Figure 3-4a). These relationships may be suggestive of temperature thresholds in

each treatment, beyond which TGM fluxes increased (Figure B-3-2).

86

Figure 3-3 PEATcosm 24-h TGM flux measurements (in ng m-2

h-1

) for the four crossed water

table and plant functional group treatments (two replicates each) in relation to surface soil

temperature (in °C) and solar radiation (kW m-2

).

87

Figure 3-4 Relationship between PEATcosm hourly TGM fluxes (ng m-2

h-1

) and a) soil

temperature (in °C) and b) solar radiation (kW m-2

) among the four crossed water table and

vascular plant functional group treatments.

When considering the foliar rinses of the Ericaceae and sedge vegetation, the highest surface

THg concentrations were washed from the Chamaedaphne calyculata (leatherleaf) leaves. This

shrub had significantly higher THg sorbed to the leaf surfaces on a dry mass basis than Kalmia

polifolia (bog laurel) as well as both sedges, Carex oligosperma and Eriophorum vaginatum

(Figure 3-5a). Carex oligosperma had the lowest surface THg concentrations; significantly lower

than either of the Ericaceae shrubs. These patterns were also consistent, although not significant

88

on a per area basis, wherein C. calyculata and K. polifolia had slightly higher THg

concentrations than both of the sedges (Figure 3-5b). When considering the amount of THg in

the leaf tissue, both C. calyculata and K. polifolia had significantly higher tissue THg

concentrations than either of the sedges (Figure 3-5c).

Figure 3-5 Total Hg concentrations of PEATcosm vegetation a) leaf rinses expressed per unit

leaf tissue mass (ng g-1

dry wt.), b) leaf rinses per unit leaf area (ng m-2

) and c) leaf tissue mass

(ng g-1

dry wt.). Letters denote statistically similar values. No significant differences between

THg concentrations expressed on a per leaf surface area basis.

89

Among leaf area, the amount of Hg sorbed to the vegetation under each DFC, and THg in

surface peat, only peat THg concentrations were significantly correlated with mean daily TGM

fluxes (Figure 3-6). A negative, but not significant correlation was observed between the

measured TGM daily fluxes and the estimated Ericaceae total leaf area present beneath each

DFC (Figure 3-6a; r = - 0.66, p = 0.15). A marginally significant negative correlation was

observed between the estimated amount of leaf-sorbed Hg on Ericaceae shrubs under each DFC

and measured TGM fluxes (Figure 3-6b; r = - 0.74, p = 0.09). A significant negative correlation

(r = - 0.90, p < 0.01) was observed between the PEATcosm TGM fluxes and the surface (0–10

cm) peat THg concentrations (Figure 3-6c).


d-1

) in relation to a) estimates of total leaf

area (in cm2) of all Ericaceae vegetation present beneath the DFC footprint, b) estimates of THg

sorbed to the surfaces of Ericaceae leaves under each DFC and c) surface (0–10 cm) peat THg

concentrations.

3.4.2 SPRUCE bog TGM fluxes

In May 2014 before the onset of experimental warming, 24-h TGM fluxes measured across all

plots had an average daily flux of −45.9 ± 93.8 ng m-2

d-1

(mean ± standard deviation). These

fluxes ranged from deposition at an average rate of −158.8 ng m-2

d-1

to evasion of +199.1 ng m-2

d-1

(Figure 3-7). In August 2014 once the deep soil warming treatments had been achieved at

depth, the minimum TGM flux recorded was −38.5 ng m-2

d-1

while the maximum TGM flux

90

observed was +59.2 ng m-2

d-1

with a mean TGM daily flux rate of −1.41 ± 27.1 ng m-2

d-1

.

Following a prolonged period under deep soil warming conditions, the mean daily TGM flux

measured in June 2015 was +10.2 ± 44.6 ng m-2

d-1

, ranging from deposition at a rate of −42.6 ng

m-2

d-1

to evasion at a rate of +107.4 ng m-2

d-1

(Figure 3-7). Significant, positive, but relatively

weak correlations were observed between the mean TGM inlet concentrations and the

corresponding TGM fluxes measured during daylight hours during the May 2014 (r = 0.47, p <

0.0001) and June 2015 (r = 0.58, p < 0.0001) samplings (Figure B-3-3).

Figure 3-7 SPRUCE daily Hg fluxes (in ng m-2

d-1

) for both of the two replicate DFCs across the

range of deep peat heating temperature treatment plots (+0, +4.5 and +9 °C) in May 2014,

August 2014 and June 2015. Flux data for Plots 4 and 6 in June 2015 not available.

3.4.2.1 Influence of deep peat warming

In May 2014, prior to the implementation of the deep warming treatments, no significant

correlation was observed between the mean TGM fluxes and the mean daily surface peat

91

temperatures (p = 0.23). Similarly, no relationships were observed in August 2014 (p = 0.55)

once the warming targets in the deep peat layers had been achieved nor in June 2015 (p = 0.69)

following prolonged deep warming. No significant correlations were observed between the TGM

fluxes and the peat temperatures at a depth of 2 m for the August 2014 (p = 0.78) or June 2015 (p

= 0.46) samplings, both of which achieved the target temperature differentials.

When comparing the slopes for the relationships between surface peat temperature and TGM

flux, the slope of the May 2014 data collected prior to deep soil warming is significantly

different from those collected in August 2014 and June 2015 under warming conditions (Figure

3-8; p = 0.04). Deep warming may therefore influence the amount of Hg emitted from the peat as

there is a transition in the slope of the relationship from strongly negative in May 2014 prior to

the initiation of the deep warming treatments, to moderately negative once warming target

differentials were achieved at depth in August 2014, and to slightly positive in June 2015

following prolonged deep warming (Figure 3-8). This trend in TGM flux may not be strongly

reflected in relation to surface peat temperature as target temperature differentials were not

achieved nearer the surface of the peat profile as air warming was not initiated for this one-year

experiment. However, this progressive change in slope throughout the deep warming experiment

under warming conditions suggests some influence of deep soil warming on TGM flux from the

peatland.

92

Figure 3-8 Relationship between SPRUCE daily TGM fluxes and mean daily surface peat

temperatures for the May 2014 (pre-warming-solid circles), August 2014 (deep warming

achieved at depth-open circles) and June 2015 (prolonged deep warming-stars) measurement

periods. Deep warming target temperature differential treatments (+0, +4.5, +9 °C) are colour-

coded in blue, orange and red, respectively.

3.5 Discussion

3.5.1 PEATcosm peat monoliths – influence of vascular plant community on TGM fluxes

The relatively strong depositional trends observed when sedges were present under DFCs, both

as the sole surface vascular cover and with other vegetation present (Figures 3-1 and 3-2),

suggest a significant role for vegetation type in controlling TGM fluxes. Foliage may interact

with atmospheric Hg in multiple ways including exchange of Hg through stomata (Ericksen and

Gustin, 2004; Hanson et al., 1995; Lindberg et al., 1998), wet and dry deposition of Hg onto

foliar surfaces (Stamenkovic and Gustin, 2009) as well as uptake of Hg into vascular vegetation

from soil water transported to the foliage via transpiration (Graydon et al., 2006; Rea et al.,

2002). Rooting structures of vascular vegetation may also be important conduits for the dynamic

exchange of Hg between the atmosphere and soil (Leonard et al., 1998a, b). Since the sedge

species (C. oligosperma and E. vaginatum) present in the mesocosm bins had generally both the

lowest surface-sorbed and leaf tissue THg concentrations, it is unlikely that sorption to sedges or

uptake by sedges is the mechanism by which they enhance TGM deposition. Rather, the impact

93

is more likely a function of increased shuttling of TGM from the atmosphere, coincident with

oxygen transport to the peat via aerenchymous tissues (Breeuwer et al., 2009) under low water

table conditions.

Relative to the role of sedges, Ericaceae shrubs appear to influence gaseous Hg exchange

dynamics through sorption onto foliar surfaces and/or uptake through stomata and accumulation

in leaf tissue. Greater sorption of Hg to the foliar surfaces of Ericaceae shrubs as well as stomatal

uptake into the shrub leaf tissue is associated, albeit weakly, with reduced emission and some

depositional fluxes (Figure 3-6a and b). Mercury sorbed to the surface of Ericaceae leaves,

particularly that which is exposed to incoming solar radiation on the top of the leaves, may likely

be photoreactive (Laacouri et al., 2013). Once deposited, this surface-sorbed Hg may undergo

considerable re-emission to the atmosphere and contribute to the predominantly emissive fluxes

from the Ericaceae-dominated treatments. Increased leaf area will also result in more stomata

present and therefore the potential for greater Hg uptake, likely in the form of gaseous Hg(0),

while stomata are open during daylight hours for the process of photosynthesis (Lindberg et al.,

2002; Marsik et al., 2005; Laacouri et al., 2013). Increased uptake of Hg through stomata may

account for the increased accumulation of Hg within the leaf tissues of the Ericaceae shrubs, as

translocation of Hg from the roots to the leaves is often observed to be negligible in vascular

plants (Ericksen et al., 2003) and therefore contribute to the diminished emissions from the

monoliths with greater Ericaceae leaf area coverage.

The relative importance of surface Hg sorption for different plant functional groups may depend

upon such factors as surface roughness (Rea et al., 2000) as well as any potential chemical

compounds present on the vegetation or in the atmosphere which may promote TGM sorption

and the subsequent potential for Hg reduction and re-emission back to the atmosphere (Hanson et

al., 1995; Marsik et al., 2005). Sedge leaves are sleek and glabrous as compared to the scabrous

and leathery surfaces of C. calyculata and K. polifolia leaves. These differences may influence

the strength of Hg deposition and the potential re-emission from the plant surface. In addition to

exposure to solar radiation, leaf surface texture may contribute to the observed differences in

TGM flux dynamics between treatments as well as the leaf surface Hg concentrations between

vascular functional groups. Rea et al. (2000) suggested that the rough surface of birch leaves

with small hairs may more efficiently trap dry deposition as compared to the smooth surface of

maple leaves. The form of Hg deposited to the leaf surfaces may also be influenced by the

94

texture of the leaves. Reactive and water soluble Hg(II) and aerosol Hg were the likely forms of

Hg rinsed from foliar surfaces in the study by Rea et al. (2000). Comparatively, gaseous

elemental Hg(0) is more likely to be quickly re-emitted to the atmosphere following deposition

to foliar surfaces due to its minimal solubility in water, but may also be taken up through stomata

or oxidized to soluble Hg(II) (Hanson et al., 1995; Rea et al., 2000). Leaves are considered

dynamic surfaces for Hg exchange acting as either sites of deposition or emission depending

upon the prevailing conditions such as temperature and moisture (Hanson et al., 1995).

Therefore, leaves may contribute to the predominant emission observed in the Ericaceae-

dominated PEATcosm peat mesocosms, although this flux may vary depending upon the amount

of leaf area.

The observed differences in the strength of the relationships between TGM fluxes and solar

radiation and soil temperature (Figures 3-4 and B-3-2) among the treatments may be due to

differences in vegetative coverage among the plant functional groups, which results in

differential peat shading. Reduced emissions occurred in mesocosms with greater cover of

Ericaceae shrubs which, in addition to the influence of TGM sorption to and potential re-

emission from foliar surfaces, may be the result of diminished solar radiation reaching the peat

surface because of higher leaf area (Figure 3-6a). In forested environments, canopy cover

influences Hg fluxes from the forest floor due to shading and shade impacts on solar radiation

and soil temperatures; with soil temperature becoming an increasingly stronger control as canopy

cover inhibits solar radiation penetration to the forest floor (Choi and Holsen, 2009a). Among the

PEATcosm treatments, peat temperature thresholds between 20.2°C to 24.5°C were observed,

beyond which fluxes increased (Figure B-3-2). A reduction of incoming solar radiation reaching

the peat surface may reduce the amount of photoreduction of Hg(II) to Hg(0) available for re-

emission as seen in a harvested forest system with varying degrees of biomass removal (Mazur et

al., 2014). Shading variations as a result of seasonal progression from leaf-out to leaf-off periods

in a forested system significantly affect the magnitude of Hg flux with the highest fluxes

occurring prior to leaf-out (Choi and Holsen, 2009a). Therefore, shading by the Ericaceae

vegetation may also contribute to the decreased emissions of Hg from the peat surface by

potentially inhibiting photoreduction and Hg re-emission. Shading by shrub cover in bog systems

is therefore an important potential control on TGM fluxes and may influence the relative

contributions of soil temperature and solar radiation in governing Hg exchange dynamics.

95

The trend in TGM fluxes across the PEATcosm treatments is reflected in the THg concentrations

in the surface 0–10 cm of peat, with higher peat THg in monoliths with stronger deposition,

suggesting that the deposited Hg is retained in the surface peat. This is a rather novel finding

since higher surface soil concentrations typically correspond to higher TGM emissions,

particularly in polluted sites (Gustin et al., 2003; Miller et al., 2011), but also less distinctively in

background soils (Gustin et al., 2006; Kuiken et al., 2008). In some low Hg soils (i.e. less than

100 ng g-1

) no relationship between soil Hg concentrations and TGM fluxes may be observed

(Ericksen et al., 2006). The observed negative relationship between flux and surface peat Hg

provides further support for the shuttling of Hg from the atmosphere and its deposition to the

peat by the sedge vascular vegetation. The strength of the correlation may also suggest that there

is minimal re-emission of Hg from the peat surface once deposited.

3.5.1.1 Potential climate change implications for TGM fluxes

Global climate change has the potential to significantly affect the dominant vascular plant

functional groups present in peatlands due to changes in hydrology (Weltzin et al., 2003; Strack

et al., 2006; Breeuwer et al., 2009; Dieleman et al., 2015; Potvin et al., 2015). Regardless of the

direction of shifting plant communities, the exchange of TGM with the atmosphere will likely be

affected. Dominance of Ericaceae shrubs may result in a greater proportion of atmospheric Hg

being sorbed to the leaf surfaces of these plants. This pool of Hg sorbed to the surface of

vegetation may be susceptible to reduction mechanisms including photochemical and subsequent

re-emission to the atmosphere, although Ericaceae leaf coverage will likely reduce emissions

from the peat due to shading. If increased temperatures and elevated atmospheric CO2

concentrations in concert with prolonged periods of water table recession do indeed promote the

establishment of sedges in peatlands (Fenner et al., 2007; Dieleman et al., 2015), enhanced

deposition of TGM to the peat may occur. This deposited Hg may then be bound to the soil

organic matter or contributed to pore water due to the favourable redox environment of the soil,

particularly with aerated surface peat as a result of lowered, fluctuating water tables

(Waddington et al., 2015), promoting the oxidation of gaseous Hg(0) to Hg(II) (Moore and

Castro, 2012). Increased provision of Hg to peatlands by sedge vegetation may have important

implications in terms of both availability for inorganic Hg methylation and inorganic Hg and

methylmercury (MeHg) mobility. Peatlands are ideal environments for MeHg production with

predominantly saturated soils and favourable redox conditions for sulfate and iron-reducing as

96

well as methanogenic bacteria to carry out this reaction (Branfireun et al., 1996; Tjerngren et al.,

2012). Previous research at the PEATcosm experimental mesocosms demonstrated that the

highest THg and MeHg concentrations in pore water and snowmelt runoff were consistently in

the Low WT – sedge only monoliths (Haynes et al., 2017). With greater Hg deposition in these

sedge-dominated wetlands, more Hg is available for the conversion to MeHg and subsequent

export to downstream aquatic ecosystems. The observation of temperature thresholds controlling

TGM fluxes may have important implications for the increasing temperatures associated with

climate change. Prolonged periods above these threshold temperatures have the potential to

increase TGM fluxes from peat, regardless of the dominant vascular plant community. Although

each of the four treatment combinations were only monitored for TGM fluxes on two separate

24-h periods, the mechanisms postulated to account for the observed magnitude and direction of

these fluxes provide an important first look at how these influences simulating climate change

may impact TGM fluxes within peatland ecosystems.

3.5.2 SPRUCE bog fluxes – deep soil warming influence on TGM fluxes

Warming of peat soils at depth influenced TGM fluxes in the SPRUCE bog with a statistically

significant increase in the slope of the surface peat temperature-TGM flux relationship in

response to warming conditions (Figure 3-8). Although consistent temperature differentials of

+4.5 and +9 °C above ambient throughout the peat profile could not be achieved in this initial

SPRUCE deep warming manipulation experiment due to the lack of added air warming,

differences were observed in the surface peat temperatures among the treatment plots (Figure 3-

8). Gustin et al. (1997) concluded that heating of the air above the soil had a greater impact on

Hg flux than heating of the soil itself in contaminated tailings soil cores. Even without added air

warming to the enclosures, a slight warming of the surface peat does appear to result in an

increase in TGM emission to the atmosphere. Previous studies have observed correlative

relationships between both air and soil temperatures and TGM fluxes from background soils

including wetlands (e.g. Poissant and Casimir, 1998; Edwards et al., 2001; Marsik et al., 2005).

Increased temperatures result in enhanced TGM fluxes due to a lowering of the activation energy

for Hg reduction (Carpi and Lindberg, 1998; Kim et al., 2012; Moore and Castro, 2012). Biotic

influences on TGM emissions including microbially-mediated reduction of Hg may also be

accelerated by warming influences by augmenting microbial activity, although they exert less

control than abiotic factors (Choi and Holsen, 2009b). Deep warming within the peat profile does

97

indeed result in enhanced TGM fluxes from the peatland, suggesting either that the slight

warming observed at the surface enhanced reduction processes or that greater warming at depth

resulted in Hg reduction at depth, with only small proportions able to diffuse to the surface for

evasion. With increasing air and soil temperatures as a result of global climate change, TGM

fluxes from peatland ecosystems may be enhanced, but confidence in this conclusion would be

enhanced by further, longer term research at the SPRUCE project, including surface peat

warming.

3.5.3 Peatland TGM fluxes compared to other wetland environments

Only two published studies have previously examined soil-air Hg dynamics in boreal peatland

systems using DFCs (Kyllönen et al., 2012; Fritsche et al., 2014). However, the measurements in

these studies were not conducted continuously throughout 24-h diurnal cycles, which integrate

both daytime and nighttime fluxes. The magnitude of the fluxes recorded for the ombrotrophic,

spruce-Sphagnum, SPRUCE bog is similar to that of other studies of boreal peatlands as well as

other freshwater wetlands. For example, Kyllönen et al. (2012) observed mean hourly fluxes of

0.2 ng m-2

h-1

with values ranging from −0.3 to +0.6 ng m-2

h-1

in a boreal Sphagnum and grass-

dominated wetland. Measured in August under shaded plots within a mixed acid peatland,

Fritsche et al. (2014) observed TGM fluxes 1.39 ± 2.3 ng m-2

h-1

. In a marsh with vegetation

removed from beneath the DFCs, TGM fluxes ranged from 0.35 ± 0.15 ng m-2

h-1

measured in

cloudy conditions to 1.03 ± 0.79 ng m-2

h-1

recorded in sunny conditions (Zhang et al., 2006).

Using a relaxed eddy accumulation system, Osterwalder et al. (2016) observed mean fluxes of

3.0 ± 3.8 ng m-2

h-1

from a boreal peatland during the spring snowmelt period. Collectively, these

studies demonstrate that peatlands and some freshwater wetlands represent nearly neutral fluxes.

The diel pattern in TGM fluxes observed for both the PEATcosm (Figure 3-3) and SPRUCE

peatland sites, with peak fluxes occurring in concert with peak solar radiation and surface soil

temperatures shortly after midday due to radiative heating, is similar to the trends observed over

both mineral soil and open water sites (Poissant and Casimir, 1998) as well as boreal forest soils

(Kyllönen et al., 2012). This trend has similarly been observed for most soils with both

contaminated and background Hg levels (e.g. Poissant and Casimir, 1998), as well as in a boreal

wetland environment (Kyllönen et al., 2012). Despite the moderate TGM fluxes observed in

peatlands as compared to upland, mineral soils, the congruence in diurnal trends suggests that

98

influences such as solar radiation and soil and air temperatures may similarly affect peat soils.

Other meteorological variables such as precipitation patterns and soil moisture (Gustin and

Stamenkovic, 2005) as well as wind speed and direction (Marsik et al., 2005), which are

important controls on TGM exchange dynamics in other environments, may also play key roles

in peatland TGM flux. Given the likely impacts of climate change on the hydrology, ecology and

biogeochemistry of peatland systems, further study on the influence of these factors is warranted.

3.6 Conclusions

Combined changes in water table position and variability, and vascular plant functional groups

may significantly alter TGM deposition to peatlands. Aboveground and belowground biomass of

vascular vegetation is instrumental in controlling TGM dynamics via stomatal exchange, surface

sorption and vascular aerenchyma tissues. Increased Ericaceae shrub abundance with climate

change (Potvin et al., 2015) may alter the dynamic exchange of gaseous Hg via leaf surface

sorption and reduction, leading to more positive (surface to air) emission fluxes. A shift in

dominant vascular plant functional group towards sedges under a changing climate may result in

greater deposition of gaseous Hg to these systems via vascular shuttling under low water table

conditions. This deposited Hg may sorb to the organic matter increasing the amount of Hg

available for methylation or be mobilized in peat pore waters (Haynes et al., 2017). From this

study, the potential impact of increasing peat temperatures on TGM flux is still equivocal. Inter-

site variability was too great to make a strong conclusion about temperature impacts. The deep-

peat-only warming in the current experiment is unlikely to adequately represent future peat

temperature scenarios, particularly at the surface. Still, the significantly shifting trend between

flux and soil temperatures throughout the deep-peat warming experiment is suggestive of a

developing temperature influence on increasing TGM flux. Given the potential antagonistic

impacts of climate change-induced increases in soil and air temperatures combined with altered

water table regimes and shifts in plant functional groups on TGM fluxes, further integrated

investigation into how these factors may, in combination, affect the exchange of gaseous Hg

between the atmosphere and peatland systems is warranted.


We would like to acknowledge the field assistance of K. Ng and S. Rao and the laboratory

assistance of P. Huang. We would like to thank P. Hanson, W.R. Nettles, Oak Ridge National

99

Laboratory, U.S. Department of Energy, and U.S. Forest Service for access to the SPRUCE site.

We thank E.B. Swain for use of his Tekran 2537A Gaseous Mercury Analyzer. Funding was

provided through a Natural Sciences and Engineering Research Council of Canada Alexander

Graham Bell Canada Graduate Scholarship (CGS-Doctoral) to K.M.H and a NSERC Discovery

Grant (Fund #355866) to C.P.J.M. The PEATcosm experiment was funded by the USDA Forest

Service Northern Research Station and the National Science Foundation (DEB-1146149).

3.8 References



Branfireun, B. A., A. Heyes, and N. T. Roulet (1996), The hydrology and methylmercury

dynamics of a Precambrian Shield headwater peatland. Water Resources Research, 32,

1785-1794.



and Applied Ecology, 10, 330-339.



Briggs, C. W., R. Fine, M. Markee, and M. S. Gustin (2012), Investigation of the potential for

mercury release from flue gas desulfurization solids applied as an agricultural

amendment. Journal of Environmental Quality, 43, 253-262.

Briggs, C. and M. S. Gustin (2013), Building upon the conceptual model for soil mercury flux:

evidence of a link between moisture evaporation and Hg evasion. Water Air and Soil

Pollution, 224, 1744.

Carpi, A. and S. E. Lindberg (1997), Sunlight-mediated emission of elemental mercury from soil

amended with municipal sewage sludge. Environmental Science and Technology, 31,

2085–2091.

100

Carpi, A. and S. E. Lindberg (1998), Application of a Teflon dynamic flux chamber for

quantifying soil mercury fluxes: tests and results over background soils. Atmospheric

Environment, 32, 873–882.




Choi, H. D. and T. M. Holsen (2009a), Gaseous mercury fluxes from the forest floor of the

Adirondacks. Environmental Pollution, 157, 592–600.

Choi, H. D. and T. M. Holsen (2009b), Gaseous mercury emissions from unsterilized and

sterilized soils: The effect of temperature and UV radiation. Environmental Pollution,

157, 1673-1678.







Eckley, C. S., M. Gustin, C. –J. Lin, X. Li, and M. B. Miller (2010), The influence of dynamic

chamber design on operating parameters on calculated surface-to-air mercury fluxes.

Atmospheric Environment, 44, 194−203.

Eckley, C. S., M. Gustin, F. Marsik, and M. B. Miller (2011), Measurement of surface mercury

fluxes at active industrial gold mines in Nevada (USA). Science of the Total

Environment, 409, 514–22.


Hubble, E. K. Wong, L. Halfpenny-Mitchell, and M. S. Gustin (2001), Sources of

variability in mercury flux measurements. Journal of Geophysical Research –

Atmospheres, 106, 5421-5435.

101

Ericksen, J. A., M. S. Gustin, D. E. Schorran, D. W. Johnson, S. E. Lindberg, and J. S. Coleman

(2003), Accumulation of atmospheric mercury in forest foliage. Atmospheric


Ericksen, J. A. and M. S. Gustin (2004), Foliar exchange of mercury as a function of soil and air

mercury concentrations. Science of the Total Environment, 324, 271-279.

Ericksen, J. A., M. S. Gustin, M. Xin, P. J. Weisberg, and G. C. J. Fernandez (2006), Air-soil

exchange of mercury from background soils in the United States. Science of the Total




peatland catchments. Environmental Science and Technology, 41, 3146–3152.

Fitzgerald, W. F., D. R. Engstrom, R. P. Mason, and E. A. Nater (1998), The case for

atmospheric mercury contamination in remote areas. Environmental Science and





Gillis, A. and D. R. Miller (2000), Some potential errors in the measurement of mercury gas

exchange at the soil surface using a dynamic flux chamber. Science of the Total




Graydon, J. A., V. L. St. Louis, S. E. Lindberg, H. Hintelmann, and D. P. Krabbenhoft (2006),

Investigation of mercury exchange between forest canopy vegetation and the atmosphere

using a new dynamic chamber. Environmental Science and Technology, 40, 4680-4688.

Grigal, D.F. (2003), Mercury sequestration in forests and peatlands: A review. Journal of


102



1326–1350.

Gustin, M. S., G. E. Taylor, and R. A. Maxey (1997), Effect of temperature and air movement on

the flux of elemental mercury from substrate to the atmosphere. Journal of Geophysical

Research – Atmospheres, 102, 3891–3898.

Gustin, M. S., M. F. Coolbaugh, M. A. Engle, B. C. Fitzgerald, R. E. Keislar, S. E. Lindberg, et

al. (2003), Atmospheric mercury emissions from mine wastes and surrounding

geologically enriched terrains. Environmental Geology, 43, 339–351.


emissions from soils. Biogeochemistry, 76, 215-232.

Gustin, M. S., M. Engle, J. Ericksen, S. Lyman, J. Stamenkovic, and M. Xin (2006), Mercury

exchange between the atmosphere and low mercury containing substrates. Applied

Geochemistry, 21, 1913–1923.

Hanson, P. J., S. E. Lindberg, T. Tabberer, J. G. Owens, and K. –H. Kim (1995), Foliar exchange

of mercury vapor: evidence for a compensation point. Water Air and Soil Pollution, 80,

373–382.

Hanson, P. J., K. W. Childs, S. D. Wullschleger, J. S. Riggs, W. K. Thomas, D. E. Todd, and J.

M. Warren (2011), A method for experimental heating of intact soil profiles for

application to climate change experiments. Global Change Biology, 17, 1083-1096.

Haynes, K. M., E. S. Kane, L. Potvin, E. A. Lilleskov, R. K. Kolka, and C. P. J. Mitchell (2017),

Mobility and transport of mercury and methylmercury in peat as a function of changes in

water table regime and plant functional groups. Global Biogeochemical Cycles, 31. DOI:

10.1002/2016GB005471.

Hintelmann, H., R. Harris, A. Heyes, J. P. Hurley, C. A. Kelly, D. P. Krabbenhoft, S. Lindberg,

J. W. M. Rudd, K. J. Scott, and V. L. St. Louis (2002), Reactivity and mobility of new

103

and old mercury deposition in a boreal forest ecosystem during the first year of the

METAALICUS study. Environmental Science and Technology, 36, 5034-5040.

Kim, K. H., H. O. Yoon, M. C. Jung, J. M. Oh, and R. J. C. Brown (2012), A simple approach

for measuring emission patterns of vapor phase mercury under temperature-controlled

conditions from soil. The Scientific World Journal, 940413. DOI:10.1100/2012/940413.

Kuiken, T., M. Gustin, H. Zhang, S. Lindberg, and B. Sedinger (2008), Mercury emission from

terrestrial background surfaces in the eastern USA. II: air/surface exchange of mercury

within forests from South Carolina to New England. Applied Geochemistry, 23, 356–

368.



Letters, 30, 1900.



1182.




Lee, X., G. Benoit and X. Hu (2000), Total gaseous mercury concentration and flux over a

coastal saltmarsh vegetation in Connecticut, USA. Atmospheric Environment, 34, 4205-

4213.

Leonard, T. L., G. E. Jr. Taylor, M. S. Gustin, and G. C. J. Fernandez (1998a), Mercury and

plants in contaminated soils: 1. Uptake, partitioning, and emission to the atmosphere.

Environmental Toxicology and Chemistry, 17, 2063–2071.

Leonard, T. L., G. E. Jr. Taylor, M. S. Gustin, and G. C. J. Fernandez (1998b), Mercury and

plants in contaminated soils: 2. Environmental and physiological factors governing

104

mercury flux to the atmosphere. Environmental Toxicology and Chemistry, 17, 2072–

2079.

Lin, C. -J., M. S. Gustin, P. Singhasuk, C. Eckley, and M. Miller (2010), Empirical models for

estimating mercury flux from soil. Environmental Science and Technology, 44, 8522-

8528.

Lindberg, S. E., P. J. Hanson, T. P. Meyers, and K. –H. Kim (1998), Air/surface exchange of

mercury vapor over forests – The need for a reassessment of continental biogenic

emissions. Atmospheric Environment, 32, 895–908.

Lindberg, S. E. and H. Zhang (2000), Air/water exchange of mercury in the Everglades II:

measuring and modeling evasion of mercury from surface waters in the Everglades

Nutrient Removal Project. Science of the Total Environment, 259, 135-143.

Lindberg, S. E., W. Dong, and T. Meyers (2002), Transpiration of gaseous elemental mercury

through vegetation in a subtropical wetland in Florida. Atmospheric Environment, 36,

5207-5219.

Lindberg, S. E., W. Dong, J. Chanton, R. G. Qualls, and T. Meyers (2005), A mechanism for

bimodal emission of gaseous mercury from aquatic macrophytes. Atmospheric





Mazur, M., C. P. J. Mitchell, C. S. Eckley, S. L. Eggert, R. K. Kolka, S. D. Sebestyen, and E. B.

Swain (2014), Gaseous mercury fluxes from forest soils in response to forest harvesting

intensity: A field manipulation experiment. Science of the Total Environment, 496, 678-

687.

Mazur, M. E. E., C. S. Eckley, and C. P. J. Mitchell (2015), Susceptibility of soil bound mercury

to gaseous emission as a function of source depth: An enriched isotope tracer

investigation. Environmental Science and Technology, 49, 9143-9149.

105

Miller, M. B., M. S. Gustin, and C. S. Eckley (2011), Measurement and scaling of air–surface

mercury exchange from substrates in the vicinity of two Nevada gold mines. Science of

the Total Environment, 409, 3879-3886.

Moore, C. and A. Carpi (2005), Mechanisms of the emission of mercury from soil: role of UV

radiation. Journal of Geophysical Research, 110, D24302.

Moore, C. W. and M. S. Castro (2012), Investigation of factors affecting gaseous mercury

concentrations in soils. Science of the Total Environment, 419, 136−143.

Poissant, L. and A. Casimir (1998), Water–air and soil–air exchange rate of total gaseous

mercury measured at background sites. Atmospheric Environment, 32, 883–893.

Poissant, L., M. Pilote, X. Xu, H. Zhang, and C. Beauvais (2004a), Atmospheric mercury

speciation and deposition in the Bay St. François wetlands. Journal of Geophysical

Research 109, D11301.

Poissant, L., M. Pilote, P. Constant, C. Beauvais, H. H. Zhang, and X. Xu (2004b), Mercury gas

exchange over selected bare soil and flooded sites in the Bay St. Francois Wetlands

(Quebec, Canada). Atmospheric Environment, 38, 4205–4214.




and Soil, 387, 277-294.

Osterwalder, S., J. Fritsche, C. Alewell, M. Schmutz, M. B. Nilsson, G. Jocher, J. Sommar, J.

Rinne, and K. Bishop (2016), A dual-inlet, single detector relaxed eddy accumulation

system for long-term measurement of mercury flux. Atmospheric Measurement

Techniques, 9, 509-524.


Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.

106





foliage over time in two northern mixed-hardwood forests. Water Air and Soil Pollution,

133, 49-67.

Sebestyen, S. D., C. Dorrance, D. M. Olson, E. S. Verry, R. K. Kolka, A. E. Elling, et al. (2011),

Long-term monitoring sites and trends at the Marcell Experimental Forest. In: Kolka,

R.K., S. D. Sebestyen, E. S. Verry, and K. N. Brooks, editors. Peatland biogeochemistry

and watershed hydrology at the Marcell Experimental Forest. Boca Raton: CRC Press. p.

15–71.

Selin, N. E. (2009), Global biogeochemical cycling of mercury: A review. Annual Review of

Environment and Resources, 34, 43-63.



SPRUCE – Spruce and Peatland Responses Under Climatic and Environmental Change, Oak

Ridge National Laboratory, http://mnspruce.ornl.gov/. Last accessed July 2016.



Strack, M., J. M. Waddington, L. Rochefort, E. –S. Tuittila (2006), Response of vegetation and

net ecosystem carbon dioxide exchange at different peatland microforms following water

table drawdown. Journal of Geophysical Research, 111, G02006.


Climate change impacts for the conterminous USA: an integrated assessment. Part 4:

water resources. Climate Change, 69, 67–88.

107









Wallschläger, D., H. H. Kock, W. H. Schroeder, S. E. Lindberg, R. Ebinghaus, and R. –D.

Wilken (2000), Mechanism and significance of mercury volatilization from contaminated

floodplains of the German river Elbe. Atmospheric Environment, 34, 3745-3755.

Wallschläger, D., R. R. Turner, J. London, R. Ebinghaus, H. H. Kock, J. Sommar, and Z. Xiao

(1999), Factors affecting the measurement of mercury emissions from soils with flux

chambers. Journal of Geophysical Research, 104, 21859-21871.



9, 141-151.



Ecology, 81, 3464-3478.



3589-3600.

Zhang, H. H., L. Poissant, X. Xu, M. Pilote, C. Beauvais, M. Amyot, E. Garcia, and J.

Laroulandie (2006), Air-water gas exchange of mercury in the Bay Saint François

wetlands: Observation and model parameterization. Journal of Geophysical Research,

111, D17307.

108

Zillioux, E. J., D. B. Porcella, and J. M. Benoit (1993), Mercury cycling and effects in freshwater

wetland ecosystems. Environmental Toxicology and Chemistry, 12, 2245-2264.

109

Chapter 4 Impacts of Experimental Alteration of Water Table Regime and Vascular Plant Community Composition on Solid Phase Peat

Mercury Profiles and Methylmercury Production

4

4.1 Abstract

Climate change has the potential to significantly alter the hydrology and vascular plant

community composition in peatland ecosystems. These changes are likely to have as yet

unexplored impacts on mercury (Hg) cycling and net methylmercury (MeHg) production in the

solid phase peat. In this study, peat was collected from the PEATcosm peat monoliths, an

outdoor, controlled experiment where water table positions (High and Low) and vascular plant

functional groups (sedge-dominated, Ericaceae-dominated and unmanipulated control) were

manipulated to simulate potential climate change effects. Potential inorganic Hg methylation

and MeHg demethylation rate constants using enriched stable isotope incubations were assessed

in 2015, and ambient peat THg and MeHg concentrations were tracked annually from 2011

(prior to manipulation) through to the end of the PEATcosm study in 2015. Water table position

significantly affects THg, MeHg and the percentage of THg as MeHg (%MeHg), with elevated

concentrations in the lowered water table treatments within the zone of water table fluctuation.

In mesocosms where Ericaceae shrubs were removed and sedges became the dominant vascular

plant, MeHg concentrations and %MeHg within the rooting zone become significantly elevated.

At approximately 40 cm below the peat surface, where the lowered water tables were located

during the experiment, mean MeHg concentrations are positively and strongly correlated

(r=0.96) with pore water acetate concentrations, which are derived from aerobic peat

decomposition under a lowered water table. Potential Hg methylation rate constants in the

depths coinciding with the lowered water table treatment are similar under both water table

treatments. No significant treatment effects on demethylation are observed. Enhanced

partitioning of inorganic Hg and MeHg from the solid phase peat into pore waters occur with a

lowered, fluctuating water table and sedge vegetation. Lowered fluctuating water tables and

110

sedge-dominated vascular plant communities with a changing climate may enhance inorganic Hg

and MeHg accumulation within the zone of water table fluctuation via partitioning and

translocation, more so than net MeHg production.

4.2 Introduction

Boreal peatland ecosystems store large amounts of atmospherically-deposited mercury (Hg) in

accumulated organic soils (Kolka et al. 2001; Grigal 2003), on the order of 7 to 44 µg total Hg

m-2

y-1

including litterfall (Kolka et al. 2011). These systems are also considerable sources of

methylmercury (MeHg) to downstream aquatic ecosystems (Heyes et al. 2000), with wetland-

dominated catchments exporting approximately 26 to 79 times higher loads than upland

environments (St. Louis et al. 1994). A potent neurotoxin, MeHg poses a significant threat to

vulnerable wildlife and human populations as it biomagnifies and accumulates to significant

concentrations at higher trophic levels (Mergler et al. 2007; Scheuhammer et al. 2007). With

predominantly saturated, anoxic soils, peatland systems are optimal sites for the production of

MeHg (Tjerngren et al. 2012); a process mediated by a relatively wide array of anaerobic

microbes such as sulfate- and iron-reducing bacteria and fermentative archaea (Gilmour et al.

2013). Sulfate-reducing and methanogenic microbes are also known to facilitate oxidative

demethylation in freshwater ecosystems (Oremland et al. 1991; Marvin-DiPasquale and

Oremland 1998; Marvin-DiPasquale et al. 2000). Demethylation may also occur via abiotic

mechanisms including photo-decomposition (Sellers et al. 2001; Lehnherr and St. Louis 2009;

Lehnherr et al. 2012a), which may be influenced by the presence of dissolved organic matter and

solutes such as ferric iron and nitrate (Chen et al. 2003; Hammerschmidt and Fitzgerald 2010;

Kim and Zoh 2013). Overall, MeHg concentrations in natural systems reflect the net balance

between methylation and demethylation processes.

Climate change has the potential to significantly affect the ecosystem function and carbon

storage abilities of peatlands through changes in hydrology (Ise et al. 2008) and as a

consequence, significantly impact peatland mercury cycling. Increased evaporation associated

with increasing temperatures, as well as enhanced variability in precipitation patterns in the

northern continental United States (Groisman et al. 2012; Janssen et al. 2014; Yu et al. 2016)

will likely result in prolonged water table drawdown and considerable fluctuations in water table

position particularly during the summer months (Thomson et al. 2005; Whittington and Price

111

2006). In peatlands, greater fluctuations in water table position may drastically alter

biogeochemical cycling and transformations of several elements (Limpens et al. 2008), including

Hg (Coleman Wasik et al. 2015), as well as peatland vegetation community structure (Dieleman

et al. 2015; Weltzin et al. 2003; Breeuwer et al. 2009). Redox-sensitive processes such as Hg

methylation are likely to be affected by longer sustained peat aeration as a result of periodic

water table drawdown and fluctuation (Coleman Wasik et al. 2015). Regeneration of terminal

electron acceptors, such as the oxidation of sulfide to sulfate during periods of peat aeration,

enhance MeHg production (Coleman Wasik et al. 2015) and MeHg mobilization in peatland pore

waters (Haynes et al., 2017a). However, factors that affect Hg methylation may also impact the

process of MeHg demethylation, thereby influencing net MeHg production (Lehnherr 2014). For

example, microbial demethylation is likely to exceed methylation under oxidative conditions in

aquatic environments (Ullrich et al. 2001). Warmer temperatures may enhance both Hg

methylation (Yang et al. 2016a) and MeHg demethylation (Matilainen et al. 1991) by stimulating

microbial activity. In Arctic soils where warming temperatures are resulting in permafrost thaw,

Hg methylation under anoxic conditions was strongly associated with methane (CH4) fluxes

(Yang et al. 2016b). Both methylation and methanogenesis were enhanced due to accelerated

degradation of labile soil organic matter (Yang et al. 2016b). The hydrological controls on the

balance between Hg methylation and MeHg demethylation in peatlands have not been previously

investigated. Therefore, the effect of more drastically fluctuating water tables, as may be

observed due to global climate change, on the THg and MeHg storage capacity in peatlands is

unclear.

With prolonged periods of water stress due to less frequent but more intense precipitation events

(Groisman et al. 2012; Janssen et al. 2014; Yu et al. 2016), the dominant vascular plant

functional groups present in peatlands may shift (Weltzin et al. 2003; Strack et al. 2006;

Breeuwer et al. 2009; Dieleman et al. 2015; Potvin et al. 2015). The degree to which plant

communities change may depend upon the severity of water stress and the adaptive abilities of

the different plant functional groups to survive in such conditions (Potvin et al. 2015). Sedges

may thrive under conditions where the water table position becomes more variable. Typically

located on hummocks and lawns due to the inability of shallowly-rooted Ericaceae shrubs to

survive in flooded conditions, shrub coverage and productivity are anticipated to increase with a

warmer and drier climate (Potvin et al. 2015; Strack et al. 2006; Weltzin et al. 2003). Prolonged

112

water stress however, may suppress Ericaceae shrub growth (Potvin et al. 2015). The

physiological features of sedges, such as deeper roots compared to other vascular vegetation,

allow sedges to access deeper water stores during periods of water deprivation (Strack et al.

2006; Dieleman et al. 2015). Sedges also regulate water losses via stomata (Breeuwer et al.

2009) and expand the rooting zone by actively shuttling oxygen via aerenchymous tissues in

waterlogged conditions (Armstrong et al. 1991). Therefore, altered plant communities may

influence both Hg methylation and demethylation processes. A potential shift toward sedge-

dominated plant communities may enhance MeHg production by altering the peat redox potential

through oxygen shuttling and the availability of terminal electron acceptors within the rooting

zone (Mueller et al. 2016). The provision of labile sources of carbon, such as acetate via root

exudates, may stimulate sulfate and iron reduction, and thus also augment methylation

(Windham-Myers et al. 2009; 2014). In contrast, Ericaceae shrubs, the roots of which form a

symbiotic relationship with mycorrhizal fungi, may affect the soil microbial communities within

the rooting zone. The presence of Ericaceae shrubs may influence the redox potential of the peat

by outcompeting heterotrophs for available oxygen (Romanowicz et al. 2015) and therefore may

reduce peat net MeHg production. The potential effect of the dominance of either sedges or

Ericaceae shrubs under warmer and drier conditions on peat MeHg production has not been

previously examined.

Given the large stocks of Hg in peatlands, the connection of these wetlands to downstream

aquatic ecosystems and the vulnerability of these systems to a changing climate, it is important to

understand how combined changes in hydrology and plant communities may affect the net

production of MeHg in peatlands. In this study, peat at several depths was collected throughout

the course of the PEATcosm (Peatland Experiment at the Houghton Mesocosm Facility)

experiment, annually from 2011 through 2015, to assess the influence of simulated climate

change effects on solid phase peat THg and MeHg. PEATcosm is an outdoor, controlled

experiment that manipulated water table positions and vascular plant functional groups in peat

monoliths to simulate potential climate change impacts. Pore waters were also collected to

determine the effect of the experimental treatments on Hg and MeHg partitioning. Enriched Hg

isotope-based measurements of Hg methylation and MeHg demethylation potential rate constants

were completed at the conclusion of the overall PEATcosm study, once larger-scale destructive

sampling was possible. The main hypotheses of this experiment are: 1) The water table and

113

plant functional group treatments will increase solid phase peat THg and MeHg accumulation

throughout the peat profile , 2) Hg methylation will be enhanced within the zone of the lowered

water tables due to the regeneration of terminal electron acceptors (i.e. sulfate) and increased

labile carbon availability as a result of peat decomposition and 3) MeHg and inorganic Hg

partitioning from the solid phase into the peat pore waters will be enhanced with a lowered,

fluctuating water table and dominant sedge vegetation.

4.3 Methods

4.3.1 Study Site and Experimental Design

The PEATcosm experiment is located at the United States Forest Service Mesocosm Facility at

the Forestry Sciences Laboratory in Houghton, Michigan, USA (47.11469° N, 88.54787° W).

The regional climate is humid continental with typical annual precipitation of approximately 870

mm, approximately 50% of this falling as snow. Mean temperatures in this area range from -

13°C in January to +24°C in July (30 year means at Houghton County Airport; Potvin et al.

2015).

Twenty-four intact 1-m3 (1 m x 1 m x 1 m) peat monoliths were harvested from an ombrotrophic

peatland located in Meadowlands, MN, USA in May 2010 and transferred into individual

mesocosm bins. The interior of each stainless steel bin was coated with Teflon to prevent

potential metal transfer between the bin and the peat. The top of each mesocosm bin was open,

exposing the peat to the ambient climate. The bins were insulated and installed in a climate-

controlled tunnel, allowing belowground access to each of the bins as well as the simulation of a

natural vertical temperature gradient. Potvin et al. (2015) provides a detailed and comprehensive

explanation of the peat monolith harvest and experimental set-up. The experiment concluded in

July 2015 with the destructive harvest of the peat monoliths.

The PEATcosm study comprises a full-factorial experimental design with two water table (WT)

prescriptions crossed with three different plant functional group treatments simulating potential

climate change outcomes. Each treatment combination is replicated across four mesocosms, in a

randomized complete block design, for a total of 24 experimental units. Vascular plant

functional group manipulations were initiated in 2011 and water table experimental treatments

were implemented in 2012. The water table treatments were based on long-term (approximately

114

50 years) data from the Marcell Experimental Forest in north-central Minnesota (47.51907° N,

93.45966° W), located near the peat monolith harvest site. The two water table treatments were

modelled after years with: 1) typical variability and average water table position (referred to as

‘High WT’) and 2) comparably high variability and lower average water table position (referred

to as ‘Low WT’). These target water table positions were maintained through a combination of

artificial precipitation additions, rain-exclusion covers and regulated outflow from approximately

the acrotelm-catotelm boundary (Potvin et al. 2015). The three plant functional group treatments

simulating likely community composition alterations resulting from climate change (Strack et al.

2006; Weltzin et al. 2000; Chapin et al. 1996), were as follows: 1) all Ericaceae removed

(referred to as the ‘sedge only’ treatment), 2) all sedge removed (‘Ericaceae only’ treatment) and

3) both sedge and Ericaceae present (‘unmanipulated control’ treatment). The dominant sedge

species present in the bins was Carex oligosperma Michx., while the dominant Ericaceae shrubs

included Chamaedaphne calyculata (L.) Moench., Kalmia polifolia Wangenh., and Vaccinium

oxycoccos L. The dominant moss species in the peat monoliths were Sphagnum rubellum

Wilson, S. magellanicum Brid., S. fuscum (Schimp.) Klinggr and Polytrichum strictum Brid.

Polytrichum commune Hedw., Eriophorum vaginatum L., Andromeda polifolia L. var.

glaucophylla (Link) DC., Rhododendron groenlandicum (Oeder) Kron and Judd, and Drosera

rotundifolia L. were also present in the mesocosms.

In 2011 mean water table positions between the beginning of June and the end of October were 7

± 5 cm (mean ± standard deviation) below the peat surface for the High WT treatments and 11 ±

4 cm for the Low WT bins. Mean 2012 water table positions were 11 ± 4 cm for the High WT

mesocosms and 19 ± 8 cm for the Low WT treatments. In 2013 a mean differential of

approximately 20 cm was imposed between the High WT (15 ± 5 cm) and the Low WT

treatments (35 ± 11 cm). Similarly in 2014 the High WT (12 ± 5 cm) and Low WT bins (33 ± 12

cm) had an approximate mean differential of 20 cm from June through to the end of October.

4.3.2 Peat Sampling

Peat samples were collected for ambient THg and MeHg analyses from each of the 24

mesocosms in August 2011, 2012, 2013 and 2014. Cores were also collected in July 2015 when

the mesocosms were destructively harvested and THg and MeHg concentrations were obtained

simultaneously with methylation and demethylation potential assays (see explanation next

115

paragraph). Cores were collected using a stainless steel corer from which the peat was extruded

and divided in 10 cm increments. Core lengths were typically 60 cm. The corer was rinsed with

deionized water between sampling each mesocosm. Peat samples were immediately frozen

following sub-sectioning. In 2012, peat samples were only analyzed down to 40 cm below the

peat surface as the water table treatments were only initiated during this growing season and

were not dramatically drawn down.

To determine potential Hg methylation and MeHg demethylation rate constants (kmeth and kdemeth,

respectively), cores collected from each mesocosm in July 2015 were injected with enriched,

stable inorganic Hg and MeHg isotopes (Hintelmann et al. 1995; Hintelmann and Ogrinc 2003).

Surface peat (0-20 cm) and deeper peat cores (20-60 cm) were sampled using clear

polycarbonate corers (4.8 cm internal diameter) with silicone septa (1/16”) spaced at 1 cm

intervals. Once extracted from the peat monolith, the core tubes were capped immediately at

both ends. Capped cores remained upright and stored at ambient peat temperatures in an

incubator until the isotopic injections were completed (within approximately 2 hours). Potential

mercury methylation rate constants were assessed using spike additions of enriched 200

Hg

(94.3% abundance), while potential demethylation rate constants were measured using additions

of enriched Me201

Hg (84.7% abundance). Spike solutions were made by diluting the 200

HgCl+

and Me201

HgCl+ stock solutions with filtered pore water collected from each peat mesocosm on

the same day as peat collection. The solutions were allowed to equilibrate for approximately one

hour prior to injection into the peat cores, with the assumption that enriched isotopes would

equilibrate with the dissolved organic carbon present in the pore waters. Through each of the 1

cm-spaced silicone septa from 0-15 cm and 35-50 cm, 100 µL of the prepared spike solution (~2

µg mL-1

200

Hg and ~0.07 µg mL-1

Me201

Hg) was injected using a gas-tight borosilicate glass

syringe into the peat. As the added isotopic tracers are likely more bioavailable than ambient

inorganic Hg and MeHg, the values of kmeth and kdemeth represent potential methylation and

demethylation rate constants, respectively (Mitchell and Gilmour 2008). Following the spike

additions the cores were incubated in the dark in an incubator for six hours at 18°C (ambient peat

temperature measured at approximately 5 cm below the peat surface in the mesocosms). After

the six-hour interval, the peat was extruded from the tubes and sectioned into 5 cm increments.

Each peat sample was homogenized using a stainless steel hand blender, which was thoroughly

116

washed with deionized water between each sample. Samples were then immediately frozen to

terminate the methylation and demethylation incubations.

4.3.3 Pore Water Sampling

Pore waters were collected from micro-piezometer nests; ultra-high-density polyethylene casings

with Teflon tubing, located at 20, 40 and 70 cm below the peat surface in each of the 24

mesocosms (see Romanowicz et al. 2015 for complete piezometer construction details). Both

THg and MeHg analyses were performed on pore waters collected in June, August and

November 2013, May, July and September 2014 and in May 2015 prior to the destructive harvest

of the mesocosms at the conclusion of the experiment. Detailed description of the pore water

sampling and analytical methods, including quality control, can be found in Haynes et al.

(2017a). For other chemical analyses, including sulfate and acetate concentrations, pore waters

were collected monthly throughout the growing season of each study year. For this experiment,

only pore water acetate concentrations collected at 40 cm below the peat surface were considered

in association with solid phase peat THg and MeHg concentrations at corresponding depths (30-

40 and 40-50 cm).

Soil-water partition coefficients (KD, expressed in units of L kg-1

) were determined according to

the following equation:

𝐾𝐷 = [𝐻𝑔]𝑆

[𝐻𝑔]𝑃𝑊 (4.1)

where [Hg]S is the ambient solid phase peat Hg (either inorganic Hg or MeHg) concentration (in

ng kg-1

) and [Hg]PW is the ambient pore water Hg (inorganic Hg or MeHg) concentration (in ng

L-1

). Values of KD are expressed in log form. Partition coefficients were calculated for both

experimental years of 2013 and 2014 at 20 cm and 40 cm below the peat surface to coincide with

the zone of greatest water table variability. Mean solid phase peat concentrations at 10-20 cm

and 20-30 cm, as well as 30-40 cm and 40-50 cm were determined to correspond with the pore

waters collected at 20 cm and 40 cm below the peat surface, respectively. Mean annual pore

water Hg concentrations for each of 2013 and 2014 were used to determine the KD coefficients

since pore waters were not collected at the same time as the solid phase peat.

117


All frozen peat samples were freeze-dried prior to analysis. For MeHg analysis, both ambient

and tracer incubated peat samples were distilled in Teflon vessels according to US EPA Method

1630 (US EPA Method 1630, 1998). Prior to distillation a known, trace amount of enriched

stable Me199

Hg isotope was added to each sample as an internal standard (Hintelmann and

Evans, 1997). Both ambient and excess MeHg concentrations were assessed by gas

chromatography-inductively coupled plasma mass spectrometry (GC-ICP-MS) using an Agilent

Technologies 7700x ICP-MS for Hg isotope detection. Ambient and excess tracer

concentrations were calculated as per the isotope dilution calculations of Hintelmann and Ogrinc

(2003). Recovery of standard reference material (estuarine sediment ERM CC580) was 100 ±

6% (n=42), replication of duplicates was 3.9 ± 3.3% (n=37) and the MeHg detection limit was

calculated (n=42 matrix blanks) to be 0.05 ng g-1

. For THg analysis, the peat samples were

digested in hot nitric acid and the diluted digestates were analyzed by cold vapor atomic

fluorescence spectroscopy (CVAFS) according to US EPA Method 1631 (US EPA Method

1631, 2002) using a Tekran 2600 automated Total Mercury Analyzer. For detection of the added

inorganic 200

Hg tracer in the 2015 incubated peat samples, the Tekran 2600 analyzer was directly

hyphenated to the ICP-MS. Recovery of a THg spike was 101 ± 4% (mean ± standard deviation,

n=38), replication of duplicates was 3.4 ± 2.8% (n=41) and the detection limit was 0.04 ng g-1

(n=40 matrix blanks). Recovery of standard reference material (MESS-3) following digestion

was 101 ± 5% (n=42).

Potential rate constants for inorganic Hg methylation (kmeth) were calculated using the

concentration of the isotopic tracer solution that was methylated over the course of the six hour

incubation period with respect to the abundance of the isotope tracer (Hintelmann et al. 1995;

Hintelmann and Ogrinc 2003). Time zero cores for Hg methylation were not collected, which is

consistent with other studies of methylation rate potential determination (e.g. Mitchell and

Gilmour 2008). Potential MeHg demethylation rate constants (kdemeth) were determined

assuming first-order reaction kinetics according to Lehnherr et al. (2012b). The initial

concentrations of excess Me201

Hg were assumed to be those of excess total 201

Hg at the end of

the incubation as time zero peat cores were not collected. This approach facilitated the

determination of MeHg degradation over the course of the incubation period, as the Me201

Hg

spike solution contained no 201

Hg prior to injection. Detection limits for both kmeth and kdemeth

118

potential rate constants were determined based on the error associated with isotope ratio

measurements, ambient peat MeHg and inorganic Hg concentrations and the tracer spike levels,

as calculated by Mitchell and Gilmour (2008). The detection limit for kmeth ranged from 2 x 10-5

to 0.028 d-1

(mean 0.0028 d-1

). The detection limit for kdemeth ranged from 0.09 to 0.16 d-1

(mean

0.12 d-1

).

Pore water acetate concentrations were determined using a Dionex ICS 2000 ion chromatograph

(Dionex Corp.). Analytical methods and quality control information for pore water THg and

MeHg concentration determination are provided in Haynes et al. (2017a).



2014) with α = 0.05. All data were tested for normality (Shapiro-Wilk W test) and

heteroscedasticity and were successfully log-transformed to achieve normality when parametric

assumptions were not met. Repeated measures analyses of variance (ANOVA) were performed

with water table, plant functional group and depth as the main factors to explore the direct and

interactive treatment effects on THg and MeHg concentrations and %MeHg from the initial

sampling in 2011, following preliminary exposure to the imposed treatments in 2012, through to

the full experimental treatment years of 2013 and 2014. In addition to concentrations of THg

and MeHg expressed in units of ng g-1

, solid phase concentrations in units of ng cm-3

were also

considered for this analysis to account for the changes in peat bulk density that occurred with

time throughout the experiment and that naturally occur down the peat profile. The relationship

between mean peat MeHg concentrations and mean pore water acetate concentrations at 40 cm

below the peat surface was assessed using a Pearson correlation.

For the kmeth and kdemeth data from the 2015 peat core incubations, two-way ANOVAs were

performed for each individual isotopically-spiked 5 cm depth increment from 0-15cm and 35-

50cm to assess significant influences of water table position, plant functional groups and

potential interactions between these factors at each spiked depth. When no significant

interactions between the two treatment factors were observed, one-way ANOVAs were also

performed to determine significant differences in potential methylation and demethylation rate

constants among the six crossed water table and vascular plant functional group treatments. The

119

correlative relationships between kmeth, kdemeth and the percentage of THg present as MeHg in the

peat were investigated using Pearson correlations.

Significant differences in soil-water partition KD coefficients for each sampling depth and year

were assessed individually with two-way ANOVAs to determine any significant direct and

interactive influences of the water table and plant functional group treatments. As no interactive

effects were observed at any depth in both 2013 and 2014, one-way ANOVAs were also applied

to the KD values to determine significant differences among the six crossed treatments. The

correlative relationship between pore water acetate concentrations collected at 40 cm depth and

mean peat MeHg concentrations from 30-40 and 40-50 cm below the peat surface for the six

treatments was examined in 2014 using Pearson correlation. These depths were selected as the

mean lowered water tables were located around 40 cm below the peat surface and the peak peat

MeHg concentrations in the Low WT mesocosms were observed at this depth.

4.4 Results

Over the course of the PEATcosm experiment both peat MeHg concentrations and the

percentage of THg as MeHg (%MeHg) were significantly affected by the altered water table

positions (p = 0.02 and < 0.01, respectively). Peat MeHg concentrations and %MeHg

significantly changed with depth throughout the peat profile (both p < 0.0001). The differences

observed among the treatments primarily occurred within the zone of water table fluctuation of

the lowered water table treatments with elevated MeHg concentrations at these depths in the Low

WT monoliths as compared to the High WT treatments (Figure 4-1). A significant interaction

between peat depth and water table treatment was observed for peat MeHg concentrations and

%MeHg (p ≤ 0.01), suggesting that the depth at which peak MeHg concentrations were located

changed in response to the implementation of a lowered water table. The depth at which the

peak MeHg concentrations occurred within the peat profile changed over time (significant

interaction between sampling depth and year; p < 0.0001), as the depth profile of the peat MeHg

concentrations became increasingly similar with prolonged exposure to the water table and

vascular plant functional group treatments (Figure 4-1). The influence of water table position on

MeHg concentrations was not significant at 30-40 and 40-50 cm below the peat surface in 2013

(p = 0.07-0.08). In 2014 MeHg concentrations and %MeHg were significantly elevated at 30-40

cm below the surface in the Low WT treatment peat (p < 0.01 and 0.05, respectively). The

120

observed changes in the peat THg and MeHg concentrations throughout the peat profile can be

attributed to the imposed treatments as no systematic trends with depth were observed prior to

the experimental manipulations. In 2011, prior to the implementation of the plant functional

group treatments and with a mean difference of approximately 4 cm between the water table

positions of the High and Low WT bins, the profiles of THg concentrations were very similar

among the mesocosms assigned to each of the six treatment combinations (Figure 4-1).

Methylmercury concentrations also did not have any systematic pattern among the assigned

treatments in 2011 (Figure 4-1). In 2012 with the initiation of the water table treatments and a

mean growing season differential of approximately 8 cm between the Low and High WT

prescriptions, all treatments had similar THg and MeHg concentrations down to the water table

level regardless of the treatment prescription (Figure 4-1), likely as a function of the maintenance

of the water table positions. Similar trends were observed in MeHg and THg stocks, which

accounted for changes in solid phase bulk density over time, expressed in ng cm-3

, throughout

the peat profile (Figure C-4-1).

Figure 4-1 Mean peat MeHg and THg concentration (ng g-1

) profiles for the six crossed water

table and vascular plant functional group treatments throughout the PEATcosm experiment from

2011 to 2014 and associated mean June to October water table positions for the High and Low

WT prescriptions. For clarity, error bars for each 10 cm depth increment have been omitted.

121

Solid phase THg concentrations significantly changed with peat depth (p < 0.0001). A

significant interaction was observed between the water table treatments and the depth of peat

sampled (p < 0.0001), suggesting that the influence of the imposed water table treatments

changed throughout the peat profile. Similar to the peat MeHg concentrations, the depth at

which the peak THg concentrations occurred in the peat changed over the course of the

experiment (p < 0.0001), with a deepening of the peak THg concentrations from pre-treatment to

experimental conditions (Figure 4-1). The impact of the water table treatments became apparent

over the course of the experiment (significant interaction between water table treatment and the

sampling year; p < 0.01), with the peat within the zone of water table fluctuation of the Low WT

treatment increasing in THg concentrations with time as compared to the High WT treatment.

Peat THg concentrations at both 30-40 and 40-50 cm below the peat surface in 2013 were

significantly affected by the water table treatments (both p < 0.05), with higher peat THg

concentrations in the Low WT bins at these depths in 2013. In 2014, water table position

significantly affected THg concentrations in the peat 30-40 cm below the surface (p < 0.01), with

greater concentrations near the mean water table positions in the Low WT prescription

mesocosms (Figure 4-1).

The vascular plant functional group treatments also significantly affected the peat MeHg

concentrations and %MeHg (both p < 0.001). A significant interaction between the plant

functional groups and the depth of the peat sampled for both MeHg concentrations and %MeHg

(p < 0.01 and p = 0.03, respectively) suggests that only certain depths within the peat profile

were influenced by the vegetation communities. Within the top 30 cm of the peat profile, MeHg

concentrations and %MeHg were significantly enhanced in the sedge only treatments as

compared to the Ericaceae only and control vegetation bins (Figure 4-1). No significant impact

of vascular plant functional group was observed directly on peat THg concentrations (p = 0.31)

or in association with peat depth (p = 0.07).

A strong, positive correlation was observed between pore water acetate concentrations and mean

peat MeHg concentrations around the depth of the mean water table (r = 0.96, p = 0.002; Figure

4-2). This depth of 40 cm below the peat surface was selected to correspond with the zone of

lowered water table fluctuation where the greatest influence of the water table treatments was

observed on peat MeHg and THg concentrations (Figure 4-1).

122

Figure 4-2 Relationship between mean treatment peat MeHg concentrations (ng g-1

) at 40 cm

below the peat surface (mean of the 30-40 and 40-50 cm depth increments) and pore water

acetate concentrations at 40 cm depth.

4.4.1 Hg Methylation and MeHg Demethylation

Based on the isotopic incubations of peat collected in 2015, both water table and plant functional

group treatments significantly influenced potential Hg methylation rate constants at depths

throughout the peat profile (Figure 4-3). The potential Hg methylation rate constants in the High

WT treatment bins were relatively consistent down the peat profile, with the kmeth of the aerated

upper 10 cm of the peat profile significantly lower than that in the 35-40 cm peat (Figure 4-3).

In the Low WT treatment mesocosms, the kmeth values in the aerated upper 0-15cm peat were

significantly lower than all depth increments below the water table at 35-50 cm below the peat

surface (Figure 4-3). In terms of MeHg demethylation, no significant differences were observed

with depth for either the High WT or Low WT treatments (Figure C-4-2). No strongly

significant influences of either water table position or vascular plant functional group on

demethylation were observed throughout the individual 5 cm increments located 0-15 cm and

35-50 cm below the peat surface. A significant (p < 0.0001), albeit weak (r = 0.50) positive

correlation was observed between the Hg methylation potential and the percentage of THg

present as MeHg (%MeHg) in the peat (Figure 4-4a). No relationship was observed between the

MeHg demethylation potential and %MeHg (p = 0.55; Figure 4-4b).

123


) in the upper 15 cm of the peat profile and 35-50 cm below the peat

surface for the a) High WT and b) Low WT treatments. Mean water table levels for the month

prior to sampling are denoted by the dashed lines. Letters denote statistically similar depths for

each of the High and Low WT data.

Figure 4-4 Relationship between the fraction of THg as MeHg (%MeHg) in the peat and a) kmeth

(% d-1

), b) kdemeth (% d-1

).

124

When looking at the individual 5 cm peat increments, a significant influence of water table

position on kmeth values was observed at both 5-10 cm and 10-15 cm below the peat surface, with

higher potential methylation rate constants in the High WT monoliths (p ≤ 0.01; Figure 4-5). No

significant effect of either water table or vascular plant functional group was observed in the

surface 0-5 cm as both treatment water table positions were maintained lower than 5 cm (Figure

4-5).


) among the six experimental treatments in the upper 15 cm of the peat

profile (left column) and 35-50 cm below the surface (situated beneath the water table of the

Low WT treatments) (right column). Letters denote statistically similar groups among the six

treatments in the 10-15 cm and 35-40 cm depth increments. No significant differences among

treatments for any other peat depths.

125

Once below the water table level of the Low WT treatment, a significant influence of the

vascular plant functional group was apparent (p < 0.01) and this effect was dependent upon the

water table treatment, with a significant interaction between water table and plant functional

group (p = 0.02; Figure 4-5). Both the sedge only and Ericaceae only vegetation treatments with

high water tables and the Low WT Ericaceae had significantly lower kmeth values than the High

WT control vegetation treatment (Figure 4-5). No significant differences in kmeth among the

treatments were observed below 40 cm.

4.4.2 Hg(II) and MeHg Partitioning

The partitioning of MeHg and inorganic Hg between the peat and pore water was significantly

affected by the variations in water table position and shifts in plant functional group (Figures 4-6

and C-4-3). At a depth of 20 cm below the peat surface, the water table treatment effect was

significant in both 2013 and 2014 (p < 0.01), with lower KD values in the Low WT monoliths

(Figures 4-6c and C-4-3c). Lower KD values in the Low WT treatment mesocosms (Figure 4-6a)

suggest that enhanced partitioning of Hg into the mobile pore water phase occurred with

lowered, more variable water table position. No significant water table or plant functional group

effects on the KD values were observed at a depth of 40 cm in 2013 (Figure 4-6b). Inorganic Hg

partitioning at both 20 and 40 cm were not significantly affected by either the water table or the

plant functional group treatments in 2013 (Figure C-4-3a, b). The vascular plant functional

groups present in the mesocosms significantly affected solid phase-pore water partitioning of

both MeHg and inorganic Hg at a depth of 40 cm in 2014 (p = 0.02 for MeHg, Figure 4-6d; p <

0.01 for THg, Figure C-4-3d).

126

Figure 4-6 Soil-water partition coefficients (logKD) for MeHg at 20 and 40 cm below the peat


2014 (n = 4 per treatment). Letters denote statistically similar groups.

4.5 Discussion

Both the imposed water table treatments as well as the shifts in the vascular plant functional

groups significantly affected the accumulation of THg and MeHg within the peat profile. The

THg profile of the peat installed in the mesocosms prior to any experimental manipulations of

water table and vascular vegetation communities was similar to that observed by Rydberg et al.

(2010) in a Swedish mire and likely reflects the historical record of past Hg atmospheric

deposition (Biester et al. 2007). The observed significant influence of water table position on

peat MeHg and THg concentrations and the shift in peak concentrations to approach the zone of

lowered mean water table fluctuation under the experimental conditions may suggest that

lowered, fluctuating water tables may, over time, enhance MeHg and THg accumulation in the

127

solid phase peat. Significant differences in MeHg concentrations and %MeHg in the upper 30

cm of peat as a result of the plant functional group treatments were also observed. This depth

coincides with the location of the majority of the shallowly-rooted Ericaceae shrub roots in the

Ericaceae only and control vegetation mesocosms. Beyond the shallow depths, sedge root

density typically increased in the PEATcosm monoliths (Romanowicz et al. 2015).

Multiple mechanisms may be responsible for the observed increased accumulation of MeHg

within the zone of water table fluctuation, which coincides with the Ericaceae shrub rooting

zone. One such mechanism is the balance between the competing processes of Hg methylation

and MeHg demethylation. The strong positive relationship between %MeHg and kmeth and

overall lack of a significant relationship between %MeHg and kdemeth may collectively suggest

that Hg methylation was a stronger control on net MeHg accumulation in the peat monoliths than

the opposing process of MeHg demethylation (Mitchell and Gilmour 2008; Lehnherr et al.

2012b). Enhanced Hg methylation has been observed to occur in peat subject to water table

fluctuations, due to the regeneration of the terminal electron acceptors, such as sulfate, required

for Hg methylating microbial communities (Coleman Wasik et al. 2015). The observed

significant influence of water table position and variability on %MeHg may be an indication that

Hg methylation may be affected by the changes in redox conditions within the zone of water

table fluctuation.

However, potential Hg methylation rate constants were not significantly influenced by water

table position and variability at the depth of the lowered water table treatments. Within the peat

depths near the lowered water table position in the Low WT mesocosms, kmeth values were not

significantly different between the Low and High WT treatments. This is surprising given that

the mean water table position for the Low WT treatments was approximately 40 cm below the

peat surface during the peat collection and isotopic incubation. The lack of difference in kmeth at

the lowered water table position may suggest that the fluctuating water table position may be

suitable for Hg methylation to occur and this process is not suppressed by periodic oxic

conditions. In the PEATcosm mesocosms, Haynes et al. (2017a) determined that the pore water

sulfate concentrations were significantly influenced by the water table position with elevated

pore water sulfate concentrations in the Low WT mesocosms. These pore water sulfate

concentrations were typically highest in the 40 cm depth pore waters of the Low WT bins, which

corresponds with the zone of water table fluctuation and the peak peat MeHg concentrations.

128

Therefore, increased regeneration and availability of sulfate under lowered and variable water

tables may prime the production of MeHg in the solid phase peat and result in similar

methylation potential as is observed under flooded conditions.

Increased availability of labile carbon under Low WT conditions may also contribute to the

enhanced MeHg concentrations in the solid phase peat in the zone of water table fluctuation and

the similarity in kmeth between the Low and High WT treatments. Available labile forms of

carbon, including acetate, act as an energy source for methylating microbial communities

including iron- and sulfate-reducing bacteria and methanogens (Mitchell et al. 2008; Yang et al.

2016b). King et al. (2000) demonstrated that acetate-consuming sulfate-reducing bacteria

actively methylate Hg. The strong correlation between pore water acetate concentrations at 40

cm depth and the mean solid phase peat MeHg concentrations in the corresponding depths across

the six treatment combinations suggest that an available source of labile carbon for methylating

microbial communities may promote methylation in these monoliths. This effect has similarly

been observed in association with added sulfate which stimulates microbial Hg methylation

(Mitchell et al. 2008), and combined with sulfate and other terminal electron acceptor

regeneration due to fluctuating water levels in the PEATcosm mesocosms, MeHg production

may be augmented over time.

The vascular plant functional groups may also influence the balance contributing to net MeHg

production in the peat. Within the rooting zone of the Ericaceae shrubs, oxygen may be taken up

by the roots and associated mycorrhizal fungi of these plants thereby limiting its availability to

heterotrophs (Romanowicz et al. 2015). The presence of Ericaceae shrubs has been observed to

suppress heterotrophic peat decomposition and Hg mobility in pore water due to oxygen

availability (Haynes et al., 2017a). Limited availability of labile carbon sources due to

heterotrophic suppression of decomposition with the presence of Ericaceae shrubs may

contribute to the lower observed MeHg concentrations as a result of reduced Hg methylation.

The consumption of oxygen by heterotrophic bacteria may also suppress terminal electron

acceptor regeneration, thereby limiting methylation potential. Therefore, the removal of

Ericaceae shrubs and their mycorrhizal fungi may result in elevated MeHg under both high water

table and dry conditions. The establishment of sedges may also influence net MeHg production

in the peat. The shuttling of oxygen by the aerenchymous tissues of the sedges may influence

Hg methylation in the peat due to more oxic conditions within the rooting zone (Crow and

129

Wieder 2005). The presence of both sedge and Ericaceae vascular vegetation in the control

treatments may balance the availability of oxygen and, in conjunction with the peat water table

position, provide an optimal redox state for methylation to occur in the solid phase. This may

account for the elevated potential Hg methylation rate constants in the control vegetation

treatments observed just below the water table.

Given that collectively the observed kmeth across the treatments cannot solely account for the

shifts in peak MeHg and THg concentrations in the peat profile, partitioning of inorganic Hg and

MeHg from the solid phase into the mobile pore waters may be a stronger control. The

mechanism by which accumulation at the observed depths may occur likely includes

translocation of Hg within the peat profile (Canario et al. 2008; Franzen et al. 2004).

Mobilization of Hg from decomposition of peat soil and subsequent re-sorption in the underlying

mineral layers has previously been observed to account for natural internal enrichment within the

soil profile (Franzen et al. 2004). The similarity between the THg and MeHg peat profiles over

the course of the experiment provides support for this mechanism as it may suggest that a

common process is influencing both inorganic Hg and MeHg. The observed changes in the peat

profiles over time are not likely a function of changes in peat bulk density as a result of

decomposition, resulting in elevated inorganic Hg and MeHg concentrations at certain depths.

The peat profile inorganic Hg and MeHg concentrations are similar to the inorganic Hg and

MeHg stock profiles, which account for changes in bulk density (see Figure C-4-1). The trends

in KD coefficients reflect the observed patterns in pore water MeHg and THg concentrations

(reported in Haynes et al., 2017a), which may suggest that changes in pore water inorganic Hg

and MeHg, as opposed to those in the solid phase peat, may be driving the observed partitioning.

Leaching of Hg in association with dissolved organic matter from the solid phase peat in the Low

WT mesocosms was primarily attributed to enhanced decomposition in the aerobic upper depths

(Haynes et al., 2017a), which may also account for the enhanced partitioning of both inorganic

Hg and MeHg in the shallow peat depths under Low WT conditions (Figures 4-6 and C-4-3).

The deepening of soil oxygenation with water table drawdown has been observed to stimulate

microbial decomposition in peat (Bragazza et al. 2016) resulting in enhanced concentrations of

dissolved organic carbon liberated in peatland pore waters (Liu et al. 2016), including labile

carbon sources such as acetate (Yang et al. 2016b). The positive correlation between pore water

acetate concentrations and MeHg concentrations may also indicate that, as a result of aerobic

130

decomposition, Hg may be liberated from the solid phase in association with small, mobile

organic carbon molecules. Leaching of inorganic Hg and MeHg from the peat into the pore

waters in association with dissolved organic carbon has previously been observed to occur in the

Low WT treatment mesocosms (Haynes et al. 2017a). Translocation of Hg within the peat

profile to the depths associated with the lowered water table position may be facilitated by

dissolved organic carbon movement. This mobile inorganic Hg and MeHg may subsequently be

transported by fluctuating water levels within the peat profile and re-sorbed to the solid phase in

the depths coinciding with the greatest variation in water table position; potentially binding to the

reduced sulfur groups in the organic matter (Skyllberg et al. 2000). This mechanism may play an

important role in contributing to the shifting peak peat THg and MeHg concentrations toward the

position of the lowered water table.

The presence of different plant functional types have also been observed to affect dissolved

organic matter in peat pore waters, with vascular plants such as graminoids and Ericaceae shrubs

responsible for the destabilization of organic matter and the facilitation of carbon losses

(Robroek et al. 2016). Aeration of the peat in the rooting zone via oxygen leakage from the

aerenchymous tissues of the sedge vegetation (Greenup et al. 2000; Crow and Wieder 2005;

Waddington et al. 2015) and the lack of competition for available oxygen by Ericaceae shrubs

(Romanowicz et al. 2015) may contribute to peat degradation deeper within the peat profile and

further augment Hg partitioning to pore water (Figures 4-6d and C-4-3d). In addition to

affecting peat decomposition, the removal of different plant functional types has also been

observed to result in a shift of the microbial community composition (Robroek et al. 2015) and

therefore influence such processes as methanogenesis (Hodgkins et al. 2014). Removal of

sedges, and the subsequent loss of a labile source of carbon from root exudation, may limit the

amount of labile carbon available for microbial decomposer communities; thereby affecting peat

decomposition (Robroek et al. 2016) and potential MeHg and inorganic Hg translocation within

the peat profile.

4.5.1 Conclusions and Implications

This study demonstrates that inorganic Hg and MeHg partitioning, translocation and re-sorption

to the solid phase may contribute to the shifts in the depth of peak concentrations as a result of

changes in hydrology and plant community. Inorganic Hg and MeHg will likely be partitioned

131

into pore waters as a result of aerobic peat decomposition above the lowered water table level

and mobilized within the peat in association with dissolved organic carbon. Some of this Hg

may be translocated and re-sorbed to the solid phase peat within the zone of the lowered water

tables. Alterations in peatland water table positions may also affect potential Hg methylation

rate constants, with comparable kmeth values under both Low and High WT conditions in the peat

depths subjected to water table fluctuations. The comparable MeHg production under aerated

conditions to that observed in anoxic peat around the lowered water table depth may be

facilitated by residual pockets of anoxia and the regeneration of terminal electron acceptors such

as sulfate. Root oxygen loss from sedges may also contribute to terminal electron acceptor

regeneration. The availability of a labile carbon source for methylating communities, such as

acetate, liberated from enhanced peat decomposition under aerobic conditions and via sedge root

exudation may also be a contributing factor leading to greater MeHg production. Climate-

induced shifts in water table position and vascular plant functional groups have the potential to

enhance the transport and export of MeHg from peatlands to downstream aquatic ecosystems

(Haynes et al., 2017a), which may harm vulnerable wildlife and human populations which rely

upon aquatic life for sustenance (Mergler et al. 2007; Scheuhammer et al. 2007). Additionally, if

a changing climate favors the establishment of sedges and the removal of Ericaceae shrubs,

inorganic Hg may be delivered to the peat from the atmosphere by the sedge aerenchyma

(Haynes et al., 2017b) resulting in elevated peat THg concentrations, which may be available for

methylation. Combined changes in both water table levels and vascular plant communities have

the potential to significantly affect the cycling of inorganic Hg and MeHg, both speciation and

mobility, within the solid phase peat and the partitioning of Hg species into peatland pore waters.

Further research is required to examine the role of Hg translocation within the profile of peatland

systems, as this mechanism of Hg accumulation is understudied in the literature.

132


We would like to acknowledge the laboratory assistance of P. Huang, K. Ng, R. Co and B.

Perron as well as the field assistance of K. Ng. We thank L. Jamie Lamit for ambient peat

collection. Funding was provided through a Natural Sciences and Engineering Research Council

of Canada (NSERC) Alexander Graham Bell Canada Graduate Scholarship (CGS-Doctoral) to

K.M.H and a NSERC Discovery Grant to C.P.J.M. The PEATcosm experiment was funded by

the USDA Forest Service Northern Research Station Climate Change Program and the National

Science Foundation (DEB-1146149).

4.7 References

Armstrong, W., S. H. F. W. Justin, P. M. Beckett, and S. Lythe (1991), Root adaptation to soil

waterlogging. Aquatic Botany, 39, 57–73.

Biester, H., R. Bindler, A. Martinez-Cortizas, and D. R. Engstrom (2007), Modeling the past

atmospheric deposition of mercury using natural archives. Environmental Science and

Technology, 41, 4851-4860.

Bragazza, L., A. Buttler, B. J. M. Robroek, R. Albrecht, C. Zaccone, V. E. J. Jassey, and C.

Signarbieux (2016), Persistent high temperature and low precipitation reduce peat carbon

accumulation. Global Change Biology, 22, 4114-4123.



and Applied Ecology, 10, 330-339.

Canario, J., L. Poissant, N. O’Driscoll, J. Ridal, T. Delongchamp, M. Pilote, P. Constant, J.

Blais, and D. Lean (2008), Mercury partitioning in surface sediments of the Upper St.

Lawrence River (Canada): Evidence of the importance of the sulphur chemistry. Water,

Air and Soil Pollution, 187, 219-231.




133

Chen, J., S. O. Pehkonen, and C. –J. Lin (2003), Degradation of monomethylmercury chloride by

hydroxyl radicals in simulated natural waters. Water Research, 37, 2496-2504.





doi:10.1002/2015JG002993.

Crow, S. E. and R. K. Wieder (2005), Sources of CO2 emission from a northern peatland: root

respiration, exudation, and decomposition. Ecology, 86, 1825-1834.




Franzen, C., R. Kilian, and H. Biester (2004), Natural mercury enrichment in a minerogenic fen

– evaluation of sources and processes. Journal of Environmental Monitoring, 6, 466-472.




11810-11820.

Greenup, A., L. M. Bradford, A. N. McNamara, P. P. Ineson, and J. A. Lee (2000), The role of

Eriophorum vaginatum in CH4 flux from an ombrotrophic peatland. Plant and Soil, 227,

265-272.



Groisman, P. Y., R. W. Knight, and T. R. Karl (2012), Changes in intense precipitation over the

central United States. Journal of Hydrometeorology, 13, 47-66.

134

Hammerschmidt, C. R., and W. F. Fitzgerald (2010), Iron-mediated photochemical

decomposition of methylmercury in an arctic Alaskan lake. Environmental Science and

Technology, 44, 6138-6143.

Heyes, A., T. R. Moore, J. W. M. Rudd, and J. J. Dugoua (2000), Methyl mercury in pristine and

impounded boreal peatlands, Experimental Lakes Area, Ontario. Canadian Journal of

Fisheries and Aquatic Sciences, 57, 2211-2222.

Haynes, K. M., E. S. Kane, L. Potvin, E. A. Lilleskov, R. K. Kolka and C. P. J. Mitchell (2017a),

Mobility and transport of mercury and methylmercury in peat as a function of changes in

water table regime and plant functional groups. Global Biogeochemical Cycles, 31,

DOI:10.1002/2016GB005471.

Haynes, K. M., E. S. Kane, L. Potvin, E. A. Lilleskov, R. K. Kolka and C. P. J. Mitchell (2017b),

Gaseous mercury fluxes in peatlands and the potential influence of climate change.

Atmospheric Environment, 154, 247-259.

Hintelmann, H., R. D. Douglas, and J. Y. Villeneuve (1995), Measurement of mercury

methylation in sediments by using enriched stable mercury isotopes combined with

methylmercury determination by gas chromatography-inductively coupled plasma

mass spectrometry. Journal of Analytical Atomic Spectrometry, 10, 619-624.

Hintelmann, H., and R. D. Evans (1997), Application of stable isotopes in environmental tracer

studies-Measurement of monomethylmercury (CH3Hg+) by isotope dilution ICP-MS and

detection of species transformation, Fresenius Journal of Analytical Chemistry, 358, 378-

385.

Hintelmann H, and N. Ogrinc (2003), Determination of stable mercury isotopes by ICP/MS and

their application in environmental studies. In Biogeochemistry of Environmentally

Important Trace Elements, Cai Y, Braids CO. (eds). ACS Symp. Ser. Vol. 835, American

Chemical Society: Washington, DC; 321-338.

Hodgkins, S.B., M. M. Tfaily, C. K. McCalley, T. A. Logan, P. M. Crill, S. R. Saleska, V. I.

Rich, and J. P. Chanton (2014), Changes in peat chemistry associated with permafrost

135

thaw increase greenhouse gas production. Proceedings of the National Academy of

Sciences, 111, 5819-5824.



763-766.

Janssen, E., D. J. Wuebbles, K. E. Kunkel, S. C. Olsen, and A. Goodman (2014), Observational-

and model-based trends and projections of extreme precipitation over the contiguous

United States. Earth’s Future, 2, 99–113.

Kim, M. –K. and K. –D. Zoh (2013), Effects of natural water constituents on the photo-

decomposition of methylmercury and the role of hydroxyl radical. Science of the Total


King, J. K., J. E. Kostka, M. E. Frischer, and F. M. Saunders (2000), Sulfate-reducing bacteria

methylate mercury at variable rates in pure culture and in marine sediments. Applied and

Environmental Microbiology, 66, 2430–2437.



Journal, 65, 897-905.

Kolka, R. K., C. P. J. Mitchell, J. D. Jeremiason, N. A. Hines, D. F. Grigal, D. R. Engstrom, J. K.

Coleman Wasik, E. A. Nater, E. B. Swain, B. A. Monson, J. A. Fleck, B. Johnson, J. E.

Almendinger, B. A. Branfireun, P. L. Brezonik, and J. B. Cotner (2011), Mercury cycling

in peatland watersheds. Chapter 11. In: Kolka, R. K., S. D. Sebestyen, E. S. Verry, K. N.

Brooks, eds. Peatland biogeochemistry and watershed hydrology at the Marcell

Experimental Forest. Boca Raton, FL: CRC Press: 349-370.

Lehnherr, I., and V. L. St. Louis (2009), Importance of ultraviolet radiation in the

photodemethylation of methylmercury in freshwater ecosystems. Environmental Science


136

Lehnherr, I., V. L. St. Louis, C. A. Emmerton, J. D. Barker, and J. L. Kirk (2012a),

Methylmercury cycling in high Arctic wetland ponds: Sources and sinks. Environmental

Science and Technology, 46, 10514-10522.

Lehnherr, I., V.L. St. Louis, and J.L. Kirk (2012b), Methylmercury cycling in high Arctic

wetland ponds: Controls on sedimentary production. Environmental Science and

Technology, 46, 10523-10531.

Lehnherr, I. (2014), Methylmercury biogeochemistry: a review with special reference to Arctic

aquatic ecosystems. Environmental Reviews, 22, 229-243.



global implications – A synthesis. Biogeosciences, 5, 1475-1491.

Liu, L., H. Chen, Q. Zhu, G. Yang, E. Zhu, J. Hu, C. Peng, L. Jiang, W. Zhan, T. Ma, Y. He and

D. Zhu (2016), Responses of peat carbon at different depths to simulated warming and

oxidizing. Science of the Total Environment, 548-549, 429-440.

Marvin-DiPasquale, M. C., and R. S. Oremland (1998), Bacterial Methylmercury Degradation in

Florida Everglades Peat Sediment. Environmental Science and Technology, 32, 2556-

2563.

Marvin-DiPasquale, M., J. Agee, C. McGowan, R. S. Oremland, M. Thomas, D. Krabbenhoft,

and C. C. Gilmour (2000), Methyl-mercury degradation pathways: A comparison among

three mercury-impacted ecosystems. Environmental Science and Technology, 34, 4908-

4916.

Matilainen, T., M. Verta, M. Niemi, and A. Uusirauva (1991), Specific rates of net

methylmercury production in lake sediments. Water Air and Soil Pollution, 56, 595-605.




137

Mitchell, C. P. J., B. A. Branfireun, and R. K. Kolka (2008), Assessing sulfate and carbon



Mitchell, C. P. J., and C. C. Gilmour (2008), Methylmercury production in a Chesapeake Bay

salt marsh, Journal of Geophysical Research, 113, G00C04.



Oremland, R. S., C. W. Culbertson, and M. R. Winfrey (1991), Methylmercury decomposition

in sediments and bacterial cultures: Involvement of methanogens and sulfate reducers

in oxidative demethylation. Applied and Environmental Microbiology, 57, 130-137.




and Soil, 387, 277-294.



Robroek, B. J. M., V. E. J. Jassey, M. A. R. Kox, R. L. Berendsen, R. T. E. Mills, L. Cécillon, J.

Puissant, M. Meima-Franke, P. A. H. M. Bakker and P. L. E. Bodelier (2015), Peatland

vascular plant functional types affect methane dynamics by altering microbial community

structure. Journal of Ecology, 103, 925–934.

Robroek, B. J. M., R. J. H. Albrecht, S. Hamard, A. Pulgarin, L. Bragazza, A. Buttler, and V. E.

J. Jassey (2016), Peatland vascular plant functional types affect dissolved organic matter

chemistry. Plant and Soil, 407, 135-143.

Romanowicz, K. J., E. S. Kane, L. R. Potvin, A. L. Daniels, R. K. Kolka, and E. A. Lilleskov

(2015), Understanding drivers of peatland extracellular enzyme activity in the

PEATcosm experiment: mixed evidence for enzymic latch hypothesis. Plant and Soil,

397, 371-386.

138

Rydberg, J., J. Karlsson, R. Nyman, I. Wanhatalo, K. Näthe, and R. Bindler (2010), Importance

of vegetation type for mercury sequestration in the northern Swedish mire,

Rödmossamyran. Geochimica et Cosmochimica Acta, 74, 7116-7126.



12-18.

Sellers, P., C. A. Kelly, and J. W. M. Rudd (2001), Fluxes of Methylmercury to the Water

Column of a Drainage Lake: The Relative Importance of Internal and External Sources.

Limnology and Oceanography, 46, 623-631.






Canadian Journal of Fisheries and Aquatic Sciences, 51, 1065–1076.

Strack, M., J. M. Waddington, L. Rochefort, and E.-S. Tuittila (2006), Response of vegetation






Tjerngren, I., T. Karlsson, E. Bjorn, and U. Skyllberg (2012), Potential Hg methylation and

MeHg demethylation rates related to the nutrient status of different boreal wetlands.

Biogeochemistry, 108, 335-350.

Ullrich, S. M., T. W. Tanton, and S. A. Abdrashitova (2001), Mercury in the aquatic

environment: A review of factors affecting methylation. Critical Reviews in


139

U.S. EPA, Method 1630: Methyl Mercury in Water by Distillation, Aqueous Ethylation, Purge

and Trap, and Cold Vapor Atomic Fluorescence Spectrometry, U.S. Environmental

Protection Agency: Washington, D.C., 1998.








Ecology, 81, 3464-3478.



9, 141–151.



3589-3600.


Heredia-Middleton, C. Coates, and E. Kakouros (2009), Experimental removal of

wetland emergent vegetation leads to decreased methylmercury production in surface

sediment. Journal of Geophysical Research, 114, G00C05.

Windham-Myers, L., M. Marvin-DiPasquale, C. A. Stricker, J. L. Agee, L. H. Kieu, and E.

Kakouros (2014), Mercury cycling in agricultural and managed wetlands of California,

USA: Experimental evidence of vegetation-driven changes in sediment biogeochemistry

and methylmercury production. Science of the Total Environment, 484, 300-307.

140

Yang, Z., W. Fang, X. Lu, G.-P. Sheng, D. E. Graham, L. Liang, S. D. Wullschleger, and B. Gu

(2016a), Warming increases methylmercury production in an Arctic soil. Environmental

Pollution, 214, 204-209.

Yang, Z., S. D. Wullschleger, L. Liang, D. E. Graham, and B. Gu (2016b), Effects of warming

on the degradation and production of low-molecular-weight labile organic carbon in an

Arctic tundra soil. Soil Biology and Biochemistry, 95, 202-211.

Yu, L., S. Zhong, L. Pei, X. Bian, and W. E. Heilman (2016), Contribution of large-scale

circulation anomalies to changes in extreme precipitation frequency in the United States.

Environmental Research Letters, 11, 044003.

141

Chapter 5 Contrasting Mercury Isotopic Compositions of Two Sub-boreal

Peatlands

5

5.1 Abstract

Peatland ecosystems store large amounts of mercury (Hg) in organic peat soils, which are

vulnerable to hydrological changes and shifts in vascular plant communities likely to result from

climate change. Stable isotope fractionation of Hg aids in identifying sources and the processes

controlling Hg cycling, but this approach has not been widely applied in peatland ecosystems.

To compare different peatland systems and to examine the potential influence of climate change

impacts on peat Hg cycling, surface (0-10 cm) peat was collected for Hg isotopic analysis from a

peatland in northern Minnesota and a plot-scale climate change experiment (PEATcosm) in

northern Michigan where water table positions and vascular plant functional groups were

manipulated. Based on Δ201

Hg mass balance mixing models, gaseous Hg(0) deposition accounts

for 57 ± 9% of the peat Hg concentrations at the northern Minnesota site, whereas Hg(II)

deposition dominated (74 ± 12%) the Hg stores at the northern Michigan site, potentially as a

result of significant differences in seasonal snow accumulation. Higher Hg(II) contributions are

observed in peat where higher water tables are maintained through greater precipitation and

snowmelt inputs. A strong relationship is observed between mass dependent fractionation

(MDF) and total gaseous Hg fluxes measured at the PEATcosm experiment, but not at the

northern Minnesota site, potentially due to the presence of a tree canopy, which is not present in

the PEATcosm site and which likely affects boundary layer stability. Odd isotope mass

independent fractionation (MIF) of peat at the northern Minnesota site is indicative of

photochemical reduction. In contrast, the PEATcosm peat exhibits significant positive Δ201

Hg

and Δ200

Hg anomalies with only a minimal or no corresponding shift in Δ199

Hg and Δ204

Hg,

respectively, which does not align with any previously identified MIF mechanisms. This study

demonstrates that Hg cycling in peatlands varies by site and isotopic analyses may aid in

identifying how cycling mechanisms may be altered with a changing climate. Further research

142

in peatland ecosystems is required to determine the processes responsible for MDF in peat soil,

the isotopic anomalies observed in the PEATcosm peatland and the influence of snow Hg

accumulation processes in peat.

5.2 Introduction

Peatland ecosystems cover approximately 3% of the global landscape and are predominantly

located in subarctic, boreal and temperate regions globally (Gorham 1991). Despite their small

global land area, these ecosystems store between 10% (Gorham 1991) to 30% (Turunen et al.

2002) of the global soil carbon pool. With their unique biogeochemical and hydrological

characteristics these ecosystems play an important role in global climate; acting as sinks of

atmospheric carbon dioxide (CO2) throughout the Holocene epoch via accumulation of immense

quantities of organic matter (Gorham 1991). Accumulation of organic matter in thick peat soils

occurs in peatlands due to a net imbalance between plant net primary production and microbial

decomposition (Clymo 1984; Moore and Basiliko 2006). Hydrological processes that maintain

peatland water tables are significant controls on the delicate balance between carbon

accumulation and carbon losses via the microbial mineralization of organic matter (Whittington

and Price 2006; Waddington et al. 2015).

Climate change threatens to alter the hydrological budget in peatlands in the boreal and sub-

boreal regions (Bridgham et al. 1995) with both changes in precipitation patterns and greater

evapotranspiration expected in association with increasing temperatures (Kunkel et al. 2003;

Groisman et al. 2005; Thomson et al. 2005). Changes in hydrology may further result in a shift

in plant community composition towards vascular-dominated vegetation including graminoids

and Ericaceae shrubs, as compared to Sphagnum moss-dominated systems (Weltzin et al. 2003;

Strack et al. 2006; Dieleman et al. 2015), with potential feedbacks on peatland carbon cycling

(Bridgham et al. 1995). Therefore, the carbon storage capacity of peatlands may be greatly

reduced and ecosystem feedbacks may result in considerable re-release of sequestered carbon to

the atmosphere (Trettin et al. 2006).

In addition to storing immense stocks of carbon, peatland ecosystems also act as strong sinks of

inorganic mercury (Hg), sequestering more Hg than associated upland systems (Kolka et al.

2001; Grigal 2003). Atmospheric deposition is the principal process delivering Hg to terrestrial

ecosystems from local, regional and global sources (Driscoll et al. 2007). Gaseous elemental

143

mercury, Hg(0), may be directly deposited to terrestrial surfaces as dry deposition (Driscoll et al.

2013). Vascular vegetation has also been observed to influence the exchange of gaseous Hg with

the atmosphere via uptake by stomata and accumulation in leaf tissues as well as by sorption to

leaf surfaces (Marsik et al. 2005; Laacouri et al. 2013; Haynes et al. 2017). This accumulated

Hg in vegetation may subsequently be contributed to the soil in litter (Demers et al. 2013).

Atmospheric Hg(0) may also be oxidized to Hg(II) and deposited either in association with

precipitation or directly to soils (Driscoll et al. 2013). Once deposited, peatlands provide an

optimal setting, in terms of hydrology and redox potential, for the process of Hg methylation to

occur (St. Louis et al. 1994; Tjerngren et al. 2012), which is mediated by numerous anaerobic

microbes (Gilmour et al. 2013). Mercury in peatlands may also be reduced to Hg(0) and re-

emitted to the atmosphere (Kyllönen et al. 2012; Fritsche et al. 2014). Solar radiation, soil and

air temperatures, soil moisture and redox potential all influence the dynamic exchange between

the atmosphere and background soils (Edwards et al. 2001; Gustin and Stamenkovic 2005;

Moore and Carpi 2005; Moore and Castro 2012). Processes by which Hg may be reduced

include photochemical reduction (Bergquist and Blum 2007), microbial reduction (Kritee et al.

2007) and natural organic matter reduction (Zheng and Hintelmann 2010), as evidenced by

studies of Hg isotope fractionation. However, the relative influences of deposition and reduction

processes, as well as the controls governing these processes in peatland ecosystems, are not well

understood.

Mercury stable isotope fractionation can be utilized to elucidate the processes controlling the

cycling of Hg through the environment and to attribute sources of Hg at a particular site (Yin et

al. 2010). Mercury has seven stable isotopes: 196

Hg, 198

Hg, 199

Hg, 200

Hg, 201

Hg, 202

Hg and 204

Hg

with a mass difference of up to 4%. Both mass-dependent (MDF) and mass-independent (MIF)

isotope fractionation has been observed during transformations and reactions (Bergquist and

Blum 2009). Mass-dependent fractionation is most often represented by δ202

Hg, while MIF is

most commonly observed for odd-mass-number isotopes (MIFodd, Δ199

Hg and Δ201

Hg)

(Bergquist and Blum 2009). Even mass number MIF (MIFeven, Δ200

Hg and Δ204

Hg) is a less

common isotopic anomaly, but has been observed in atmospheric Hg samples likely due to gas

phase atmospheric processes (Gratz et al. 2010; Sherman et al. 2012; Cai and Chen 2016) and is

reflected in large positive Δ200

Hg precipitation Hg(II) isotopic signatures (Chen et al. 2012).

However, MIFeven has not been similarly observed in terrestrial environments.

144

Recent studies have begun to investigate natural abundance stable Hg isotopic signatures in

peatland ecosystems to distinguish Hg sources to these landscapes and to determine the

mechanisms influencing Hg cycling within peatlands, including deposition and re-emission

pathways (Jiskra et al. 2015; Enrico et al. 2016). Jiskra et al. (2015) concluded that the majority

of Hg found in a boreal forest peat soil was the result of litter deposition, with some contribution

from wet deposition via precipitation. Similarly, Enrico et al. (2016) determined in a French

Pyrenees bog that predominantly gaseous elemental Hg, with negative Δ199

Hg and near zero

Δ200

Hg, accumulated via dry deposition in the living Sphagnum moss and recently accumulated

peat. In terms of the controls on soil-air exchange dynamics in peatlands, Jiskra et al. (2015)

attributed significant Hg re-emission from the peat soil to non-photochemical abiotic reduction

of Hg facilitated by natural organic matter reduction, while Enrico et al. (2016) observed

evidence of photochemical re-emission. These two studies take an important first step in

identifying Hg sources and cycling processes in peatland ecosystems using Hg isotope analysis.

Given the vulnerability of peatland ecosystems to climate change, it is important to better

understand the processes governing Hg deposition, reduction and re-emission to the atmosphere

from these systems, particularly across different sites.

The purpose of this study was to compare the Hg isotopic signatures of two sub-boreal

peatlands. Surface peat samples were collected for isotopic analysis from the PEATcosm

(Peatland Experiment at the Houghton Mesocosm Facility) experimental peat monoliths located

in Houghton, Michigan both prior to (2011) and during (2014) the experimental manipulations.

PEATcosm is an experiment wherein water table levels and vascular vegetation communities

were manipulated to simulate potential climate change impacts. Surface peat samples were also

collected from the S1 bog of the Marcell Experimental Forest in north-central Minnesota. The

main objectives of this study were 1) to compare the Hg isotopic signatures of surface peat from

both sites, which may differ because of different amounts of snow during the winter months and

different vascular vegetation canopy structures, 2) to investigate the processes governing peat Hg

cycling, particularly as it relates to the total gaseous Hg fluxes measured at these sites in a

previous study (Haynes et al. 2017) and 3) to discern any treatments effects of the PEATcosm

experimental water table and plant functional group treatments on the Hg isotopic signature of

surface peat.

145

5.3 Methods

5.3.1 S1 Peatland Site Description

The S1 peatland is an 8.1 ha ombrotrophic, spruce-Sphagnum bog located within the Marcell

Experimental Forest (MEF) in north-central Minnesota (47.51907° N, 93.45966° W). Typical

annual precipitation for the MEF is approximately 780 mm, with approximately one-third of this

occurring as snow. Mean temperatures at the MEF range from -15°C in January to 19°C in July

(Sebestyen et al. 2011). The overstory is dominated by two tree species: Picea mariana (Mill.)

B.S.P and Larix laricina (Du Roi) K. Koch. The understory consists primarily of Ericaceae

shrubs including Chamaedaphne calyculata, Kalmia polifolia, Vaccinium angustifolium Aiton,

Vaccinium oxycoccos, and Rhododendron groenlandicum as well as sedges Eriophorum spp. and

the lily Maianthemum trifolium (L.) Sloboda. The dominant bryophyte present on hummocks is

Sphagnum magellanicum, while hollows are mainly colonized by S. angustifolium (C.E.O.

Jensen ex Russow).

Surface peat (0-10 cm) from this site was collected and analyzed for isotopic signatures to serve

as an in situ peat contrast to the experimental conditions of the PEATcosm mesocosms. The S1

peatland also receives considerably less snow accumulation than the PEATcosm site located in

the Upper Peninsula of Michigan.

The S1 peatland has now become the site of the SPRUCE (Spruce and Peatland Responses

Under Climatic and Environmental Change) experiment. The SPRUCE study is a climate

change field experiment that involves a ten year manipulation of both soil and air temperature

(up to 9°C above ambient temperatures in a regression design), and atmospheric carbon dioxide

(CO2; elevated to approx. 800 ppm) concentrations within open-top, approximately 12 m

diameter enclosures along three transects within the S1 bog (Hanson et al. 2017; see Figure D-5-

1a for photo of a SPRUCE plot). This study describes the isotopic signatures of surface peat

samples collected in May 2014 prior to the initiation of any manipulations.

5.3.2 PEATcosm Site Description

The PEATcosm Mesocosm Facility is located in Houghton, Michigan, USA (47.11469° N,

88.54787° W) at the United States Department of Agriculture (USDA) Forest Service Northern

Research Station - Forestry Sciences Laboratory. Typical annual precipitation at this site is

146

approximately 870 mm, with approximately 50% of this precipitation in the form of snow. As

compared to the approximately 275 mm of snow water equivalent received at the S1 peatland,

the PEATcosm site receives more than 1.5 times the amount of snow water equivalent, with an

annual average of 435 mm. Mean temperatures in this area range from -13°C in January to 24°C

in July (Potvin et al. 2015). This facility housed twenty-four intact 1 m3 (1 m x 1 m x 1 m) peat

monoliths that were harvested from an ombrotrophic peatland in Meadowlands, Minnesota, USA

in May 2010 and placed in individual Teflon-coated stainless steel mesocosm bins. The bins

were open on top, exposing the peat to ambient climate conditions and inserted into a climate-

controlled tunnel allowing belowground access to one face of each of the bins. The mesocosms

were insulated on the sides which facilitated the maintenance of a natural vertical temperature

gradient to simulate that of a natural peatland (see Figure D-5-1b for photos of the Facility).

Potvin et al. (2015) provides a comprehensive and detailed explanation of the peat harvest and

experimental set-up. The PEATcosm experiment concluded in July 2015 with the destructive

harvest of each of the monoliths.

The PEATcosm experiment was designed to assess the potential effects of anticipated shifts due

to global climate change in vascular plant communities and water table regimes on numerous

aspects of peatland ecosystem functioning, including Hg cycling. This study was a full-factorial

experimental design with two water table (WT) prescriptions crossed with three vascular plant

functional group treatments in a randomized complete block design. There were four replicate

mesocosms per treatment combination for a total of 24 experimental units. The water table

treatments were based on long-term (approximately 50 years) data from the Marcell

Experimental Forest in north-central Minnesota (47.51907° N, 93.45966° W), located north of

the peat monolith harvest site. The two target water table treatment profiles were modelled after

1) typical variability, average water table position (referred to as ‘High WT’) and 2) comparably

high variability, low water table position (referred to as ‘Low WT’). The mean difference

between the High and Low WT positions was approximately 20 cm throughout the experiment.

These target water table profiles were maintained via a combination of artificial precipitation

additions, rain-exclusion covers and regulated outflow during the spring months from

approximately the acrotelm-catotelm boundary from each of the bins. The three vascular plant

functional group treatments simulated anticipated climate change-induced community

composition (Chapin et al. 1996; Weltzin et al. 2000; Strack et al. 2006). The treatments

147

included 1) all Ericaceae removed (referred to as ‘sedge only’ treatment), 2) all sedge removed

(‘Ericaceae only’ treatment) and 3) both sedge and Ericaceae present (‘unmanipulated control’

treatment). The plant functional group treatments were initiated in June 2011 and maintained

weekly by the clipping of stems of the excluded species. The dominant sedge species present

was Carex oligosperma Michx., while the dominant ericaceous shrubs included Chamaedaphne

calyculata (L.) Moench., Kalmia polifolia Wangenh., and Vaccinium oxycoccos L.. The mosses

Sphagnum rubellum Wilson, S. magellanicum Brid., S. fuscum (Schimp.) Klinggr and

Polytrichum strictum Brid. comprised the dominant bryophyte species in all of the 24

mesocosms. To a lesser extent, P. commune Hedw., Eriophorum vaginatum L., Andromeda

polifolia L. var. glaucophylla (Link) DC., Rhododendron groenlandicum (Oeder) Kron and Judd,

and Drosera rotundifolia L. were also present.

5.3.3 Peat Sampling and THg Analysis

In June 2014 two peat cores were collected from hollow microforms within each enclosure plot

in the S1 bog. Surface peat (0-10 cm only for this study) was collected using a serrated knife,

which was rinsed with deionized water between samples. Peat from the two cores was combined

in a large sealable plastic bag and homogenized by hand. Samples were taken from the bag

using clean, gloved hands. The samples were placed on ice and frozen following collection.

At the PEATcosm study, peat analyzed for isotopes was collected from each of the 24 mesocosm

bins in 2011 prior to the initiation of the experimental treatments and in 2014 during the

experimental manipulations. Cores were collected using a stainless steel corer from which the

peat was extruded. The corer was rinsed with deionized water between each mesocosm. Peat

samples were immediately frozen following sub-sectioning. As one objective of this study is to

relate trends in total gaseous Hg (TGM) fluxes to Hg isotopic signatures, only the surface 0-10

cm samples were analyzed for isotopes. Due to low bulk density and relatively low THg

concentrations (mean ± standard deviation [THg] = 31 ± 7 ng g-1

) of these surface peat samples

as well as the limited amount of material available within the overall PEATcosm framework, the

four replicate peat samples for each of the six crossed water table and plant functional group

treatments were pooled to ensure enough Hg for isotopic analytical detection. To identify the

potential influence of the PEATcosm treatments on the Hg isotopic signatures of the peat, only

three of the 2011 pre-treatment samples from both water table treatments and two of the three

148

plant functional groups (Low WT sedge only, Low WT Ericaceae only and High WT sedge only)

were analyzed for isotopes. Comparing the pre-treatment to the treatment peat Hg signatures

also facilitated the assessment of the potential impact of the pre-treatment climate in the harvest

site of Meadowlands, MN to the experimental site climate in Houghton, MI.

All peat samples were freeze-dried and homogenized prior to digestion in hot nitric acid.

Digestates were diluted and bromine monochloride (BrCl) was added to oxidize all Hg in the

samples overnight prior to analysis. Total Hg analysis was performed using a Tekran 2600 Total

Mercury Analyzer by cold vapor atomic fluorescence spectroscopy (CVAFS) according to US

EPA Method 1631. Recovery of a THg spike was 104 ± 1 % (mean ± standard deviation, n=4),

replication of duplicates was 1.2 ± 0.1 % (n=4) and the detection limit, calculated as three

standard deviations of matrix blanks, was 0.01 ng g-1

(n=4). Recovery of standard reference

material (MESS-3) following digestion was 99 ± 3% (n=4).

5.3.4 Total Gaseous Mercury Flux Measurements

Ambient TGM fluxes were measured using Teflon dynamic flux chambers (DFCs) placed on the

peat surface and connected to a Tekran 2537A Gaseous Mercury Analyzer at both the

PEATcosm and S1 peatland sites. Full details of the measurement procedures and equipment are

provided in Haynes et al. (2017). Briefly, subsets of the total number of experimental treatments

in each experiment were monitored for TGM fluxes over individual 24-hour measurement

periods. In order to discern potential treatment effects, similar ambient conditions were required.

At PEATcosm, TGM fluxes were monitored on 8 of the 24 mesocosm bins in July 2014 at the

peak of the growing season. These measurements were performed approximately one month

prior to the peat sampling. Duplicate mesocosms of the following experimental treatment

pairings were monitored for TGM fluxes: High WT – control vegetation, Low WT – control

vegetation, Low WT – Ericaceae only, and Low WT – sedge only. These treatments span the

range of vascular plant functional group treatments with the Low WT prescription representing

the greatest anticipated hydrological changes in peatlands due to climate change. The High WT

– control vegetation treatment acted as the unmanipulated, control scenario.

At the S1 peatland twelve replicate TGM flux measurements were conducted in May-June 2014.

These fluxes were measured within the footprints of six experimental enclosures installed for the

SPRUCE experiment. Due to the considerable distance between plots and equipment

149

availability, two replicate locations within each experimental plot were measured simultaneously

during 24-hour measurement periods. These duplicate flux measurements within each plot were

averaged to relate TGM flux to the peat isotopic signatures.

5.3.5 Sample Preparation and Mercury Isotope Analysis

Total Hg in the peat samples was trapped in a 0.2% potassium permanganate (KMnO4) (w/w) in

5% sulfuric acid (H2SO4) (v/v) trapping solution using a thermal combustion method with a

modified Teledyne Leeman Labs Hydra IIc Direct Mercury Analyzer (Zheng et al. 2016). Once

trapped, an aliquot of the trapping solution was neutralized with hydroxylamine hydrochloride

(NH2OH∙HCl) and analyzed by CVAFS to determine the Hg concentration in each trap. Mean

trap recovery for all peat samples was 95 ± 14% (mean ± standard deviation). Method blanks

had an average concentration of 0.005 ng g-1

(range 0.0044 to 0.0049 ng g-1

, n=2), which

represented on average less than 0.5% of the concentration of peat sample traps. Standard

reference materials CCRMP TILL-1 (Ontario soils), NIST 1944 (New York/New Jersey

waterway sediment) and NIST 1646a (estuarine sediment) as well as the procedural standard

NIST SRM 3133 were also combusted in order to ensure the performance of the trapping method

as well as the accuracy and precision of isotopic measurements. Trap recovery of CCRMP

TILL-1 was 86 ± 5% (n=9), NIST 1944 was 107 ± 24% (n=2), NIST 1646a was 80% (n=1) and

the procedural standard NIST SRM 3133 was 91 ± 3% (n=7).

Mercury isotopic composition was determined by cold vapor multi-collector inductively coupled

plasma mass spectrometry (CV-MC-ICP-MS, Neptune Plus, Thermo Finnigan). Trap solutions

were neutralized with NH2OH∙HCl to reduce the KMnO4, and then diluted to 1 ng g-1

using a

pre-neutralized 0.2% KMnO4 solution. The diluted trap solutions were introduced into the

instrument using stannous chloride reduction and Hg(0) vapor separation. Instrumental mass

bias was corrected using an internal Tl standard (NIST 997), introduced as a desolvated aerosol.

Standard-sample-standard bracketing was conducted with NIST 3133 Hg standard. The

bracketing standard was prepared using the same matrix solution as samples and the

concentration was matched to samples such that the difference in signal intensities was within

10%. Isobaric interference from 204

Pb was corrected by measuring 206

Pb, but was consistently

negligible. On-peak zero corrections were applied to all Hg and Pb masses. Mercury isotope

compositions are reported using δ notation defined by the following equation:

150

𝛿𝑥𝑥𝑥𝐻𝑔 (‰) = [( 𝐻𝑔/ 𝐻𝑔198𝑥𝑥𝑥 )

𝑠𝑎𝑚𝑝𝑙𝑒

( 𝐻𝑔/ 𝐻𝑔198𝑥𝑥𝑥 )𝑁𝐼𝑆𝑇3133

− 1] × 1000 (5.1)

where x is the mass number of each Hg isotope from 199

Hg to 204

Hg. We used δ202

Hg to assess

MDF. Mass independent fractionation is defined as the difference between measured δxxx

Hg

values and the theoretical values determined based on kinetic mass dependent fractionation law

and is reported as Δxxx

Hg (Blum and Bergquist, 2007):

Δxx𝑥𝐻𝑔 = 𝛿xx𝑥𝐻𝑔 – 𝛽 × 𝛿202𝐻𝑔 (5.2)

where x is the mass number of each Hg isotope 199, 200, 201 and 204. β is a scaling constant to

estimate the theoretical kinetic MDF, which is 0.2520, 0.5024, 0.7520 and 1.4930 for 199

Hg,

200Hg,

201Hg and

204Hg, respectively. Each sample was measured twice and an in-house

secondary Hg standard (JTBaker) was measured at least 4 times in each analytical session to

monitor the performance of the instrument. The analytical uncertainties are reported as 2SD of

the JTBaker standard. The 2SD (n=13) for JTBaker was 0.05‰ for δ202

Hg, 0.04‰ for Δ199

Hg,

0.04‰ for Δ201

Hg, 0.02‰ for Δ200

Hg and 0.07‰ for Δ204

Hg (Table D-5-1).

5.3.6 Data Analyses


2014) with α = 0.05. All isotope data for all peat samples analyzed are provided in Table D-5-2.

Regression analyses using Pearson correlation coefficients were performed on the relationships

between δ202

Hg and Δ199

Hg, Δ199

Hg and Δ201

Hg, Δ200

Hg and Δ201

Hg, and Δ200

Hg and Δ204

Hg for

both the S1 and PEATcosm peat individually. For both peatland sites 24-hr TGM flux

measurements from a previous study (Haynes et al. 2017) were related to peat MDF signatures

(δ202

Hg). The relationships between the surface 0-10 cm peat THg concentrations and isotopic

values (Δ199

Hg, Δ200

Hg and Δ201

Hg) were also examined for both peatlands. As the surface

PEATcosm peat samples had to be pooled from all of the four replicate mesocosms for each of

the six treatment combinations in order to achieve enough Hg for isotopic detection, only a

qualitative assessment of the combined water table and plant functional group treatment effects

on isotopic signatures was performed.

The fraction of Hg(II) and Hg(0) contributions to the peat of both the S1 and PEATcosm sites

were estimated using Δ199

Hg, Δ201

Hg and Δ200

Hg values in a binary mixing model, with Hg(II)

151

and Hg(0) as the two end members, similar to the approach used by Enrico et al. (2016). The

isotopic values of the end members were selected based on the mean (± standard deviation) of

literature values and the appropriate post-depositional offsets were applied similar to the methods

of Zheng et al. (2016). The Δ199

Hg, Δ200

Hg and Δ201

Hg values for the Hg(0) and Hg(II) end

members are presented in Table D-5-3. To estimate the uncertainty associated with the

calculated Hg(II) and Hg(0) contributions, a Monte Carlo simulation was conducted with 1000

iterations, based on the uncertainties (standard deviation) in the selected end members, for each

isotopic value (Δ199

Hg, Δ200

Hg and Δ201

Hg) of each peat sample. The number of iterations was

determined to be sufficient to provide a stable solution with an uncertainty of approximately 1%

of the mean and has been applied in previous studies using Monte Carlo simulations (e.g. Zheng

et al. 2016).

5.4 Results and Discussion

5.4.1 S1 Peat Isotopic Composition

The Hg isotopic signatures of surface peat (top 0-10 cm) collected from the S1 bog had

significantly negative odd isotope anomalies (Δ199

Hg = -0.13 ± 0.10‰, Δ201

Hg = -0.15 ± 0.07‰;

mean ± standard deviation), positive Δ200

Hg anomalies (+0.07 ± 0.03‰) and near zero Δ204

Hg

values (+0.01 ± 0.02‰; see Table D-5-2). These values are similar to those recorded in living

Sphagnum moss (Δ199

Hg = -0.11 ± 0.09‰, Δ200

Hg = +0.03 ± 0.02‰, 1σ) and in recently

accumulated peat (Δ199

Hg = -0.22 ± 0.06‰, Δ200

Hg = 0.00 ± 0.04‰, 1σ) collected in a French

Pyrenees peat bog (Enrico et al. 2016). In contrast to that peatland site, wherein the majority

(~79%) of the Hg present was sourced from Hg(0) deposition, the S1 peat isotopic values

suggest that Hg(0) contributions from the atmosphere are approximately equal to Hg(II)

contributions to surface peat Hg concentrations. The binary mixing models suggest that Hg(0)

contributions range from 46 ± 9% based on the conservative Δ200

Hg values, up to 57 ± 9% based

on Δ201

Hg data (Table 5-1 for Δ201

Hg model, Tables D-5-4 and D-5-5 for Δ199

Hg and Δ200

Hg

models, respectively). One limitation to the model results may be the values selected to

represent the Hg(0) and Hg(II) end members. The isotopic signatures of Hg(0) and Hg(II) at

each of the two peatland sites were not measured, but rather synthesized from literature values

collected at multiple disparate sites. Therefore, given the range of values observed for Hg(0) and

Hg(II) in other studies, the mean of values reported in the literature may not be representative of

152

the isotopic signatures contributing to this peatland. Potential shifts in the isotopic values due to

post-depositional processes were accounted for by adding the associated offsets to the mean

literature values as has been performed in studies in forested environments (Zheng et al. 2016).

However, despite this caveat we believe that the model results indicating nearly equal Hg(0) and

Hg(II) contributions, with increasing Hg(0) contributions in some samples, is valid given the

trends in the isotopic data.

Table 5-1 Contributions of Hg(II) (fHg(II)) and Hg(0) (fHg(0)) to peat isotopic signatures for the

PEATcosm 2014 treatment peat and S1 bog 2014 peat. Results are the mean of 1000 iterations

of the isotopic mixing model with Monte Carlo simulation based on the Δ201

Hg isotopic values.

Δ201

Hg fHg(II) fHg(0)= 1 – fHg(II)

Mean Standard Deviation Mean Standard Deviation

PEATcosm Experimental Peat - 2014

Low WT – Sedge 0.64 0.12 0.36 0.12

Low WT – Ericaceae 0.59 0.12 0.41 0.12

Low WT – Control 0.69 0.13 0.31 0.13

High WT – Sedge 0.80 0.15 0.20 0.15

High WT – Ericaceae 0.82 0.16 0.18 0.16

High WT - Control 0.91 0.17 0.09 0.17

S1 Peat – 2014

Plot 4 0.34 0.08 0.66 0.08

Plot 6 0.31 0.08 0.69 0.08

Plot 10 0.48 0.10 0.52 0.10

Plot 13 0.52 0.10 0.48 0.10

Plot 17 0.50 0.10 0.50 0.10

Plot 19 0.42 0.09 0.58 0.09

Significant negative MDF anomalies, represented by the δ202

Hg values, were also observed in the

peat, ranging from -1.78 to -1.36‰. These values are higher than the mean δ202

Hg values of -

2.4‰ for Sphagnum moss observed by Enrico et al. (2016). Negative shifts in MDF have been

observed in pine needles, heather leaves and lichens in peat bogs, with δ202

Hg values ranging

from -2.3 to -2.8‰ (Enrico et al. 2016). Jiskra et al. (2015) also measured negative MDF values

153

(-2.59 ± 0.25‰) in boreal litter samples. Previous research in forest and bog ecosystems has

attributed this negative shift in δ202

Hg to MDF during stomatal uptake of Hg into the foliage of

trees and vascular vegetation and delivered to the soil via litterfall (Demers et al. 2013; Enrico et

al. 2016; Zheng et al. 2016). Some contribution of litter deposition from the vascular vegetation

to the S1 peat surface may occur but litterfall is not likely to be a dominant pathway for Hg

deposition to the peat. The higher δ202

Hg values recorded in the S1 peat as compared to those

observed in forest ecosystems may suggest an alternative mechanism is responsible for this

MDF. As Sphagnum mosses do not possess stomata it may be speculated that direct deposition

of Hg(0) via sorption to the moss surface may be the dominant source of Hg(0) in the surface

peat. Although MDF anomalies during direct Hg(0) deposition and oxidation in soil are

unconstrained, it has been suggested that a positive shift in MDF may occur due to preferential

deposition of heavier isotopes (Zheng et al. 2016). Therefore, the S1 δ202

Hg values likely

encompass both direct adsorption of Hg(0) to the Sphagnum peat (positive MDF) as well as

likely minor contributions from litter deposition of vascular vegetation, which accumulates Hg

via stomatal uptake (negative MDF). However, this remains a hypothesis as the mechanisms

governing MDF in Sphagnum peat soils with vascular vegetation cover have not been

experimentally determined in the literature.

The relationships between the isotopic signatures and the THg concentrations in the peat are also

indicative of Hg(0) contributions at the S1 site. A strong, significant, negative correlation was

observed between S1 peat THg concentrations and Δ201

Hg (r = - 0.94, p = 0.005; Figure 5-1b).

A similar trend was observed between THg concentrations and Δ199

Hg (r = - 0.90, p = 0.01),

with more negative MIFodd related to greater THg concentrations. There was no similar

relationship between the peat THg concentrations and Δ200

Hg (p = 0.24; Figure 5-1a) suggesting

that the S1 surface peat THg concentrations are not strongly reflective of Hg(II) isotopic

signatures.

154


Hg, Δ201

Hg and the surface peat THg concentrations (in ng

g-1

) for the S1 peatland and PEATcosm sites. Note different scale for Δ201

Hg for S1 bog and

PEATcosm data.

Collectively this may provide support for the increased contributions from Hg(0), rather than

Hg(II) deposition to the surface peat of the S1 peatland wherein higher THg concentrations were

observed. When investigating isotopic data in relation to peat TGM fluxes at the S1 site, there

was no relationship with MDF (p = 0.36; Figure 5-2a). Similarly, the relationship between TGM

fluxes from the peat and Δ201

Hg (p = 0.11) was not significant, while the relationship with

Δ199

Hg (r = - 0.79, p = 0.06) was marginally significant.

155

Figure 5-2 Relationship between δ202

Hg values (in ‰) and the mean daily TGM fluxes (in ng m-

2 d

-1) for the a) S1 and b) PEATcosm peatlands. Note different δ

202Hg and mean daily TGM flux

scales for both plots.

Total gaseous Hg exchange between soils and the atmosphere is a dynamic process influenced by

such factors as soil and air temperature, soil moisture, solar radiation and atmospheric Hg

concentrations over short timeframes (Edwards et al. 2001; Gustin and Stamenkovic 2005;

Moore and Carpi 2005; Moore and Castro 2012). Exchange of Hg(0) between standing vascular

vegetation via both stomatal and non-stomatal pathways including surface sorption may also

occur on short timescales affecting the magnitude and direction of TGM fluxes (Rea et al. 2002;

Stamenkovic and Gustin 2009). Gaseous Hg(0) is more likely to be quickly re-emitted to the

156

atmosphere following deposition due to its minimal solubility in water, but may also be

accumulated in leaf tissue through stomata where it may be oxidized to soluble Hg(II) (Hanson

et al. 1995; Rea et al. 2000). As litterfall contributions to peat are limited by slow rates of net

primary production of aboveground biomass (Frolking et al. 2001), gaseous Hg(0) exchange with

vascular vegetation may not be evident in the solid phase peat isotopic composition. Therefore,

as the isotopic signature of the peat is integrated over a longer time period, short term, dynamic

exchange in TGM fluxes dominated by Hg(0) may not be reflected in the isotopic data.

5.4.2 S1 Peat MIF

The mechanism for MIFodd in the surface 0-10 cm of peat from the S1 bog is likely

photochemical reduction. There was a strong, significantly positive linear relationship (r = 0.95,

p = 0.004) between Δ199

Hg and Δ201

Hg for the S1 peat with a slope of 1.35 ± 0.22 (Figure 5-3)

and no relationship in MIFeven (Figure D-5-2). The slope of the MIFodd relationship for the S1

peat was higher than the slope of 1.0 observed in aqueous solutions as a result of photochemical

reduction in the controlled experiments conducted by Bergquist and Blum (2007). This

relationship is also steeper for the S1 peat as compared to those of other peat soils (Jiskra et al.

2015; Enrico et al 2016) and those of other soils and plants (Demers et al. 2013; Zheng et al.

2016), which observed Δ199

Hg – Δ201

Hg slopes close to 1.0. A curved relationship between

Δ199

Hg and δ202

Hg was observed in the S1 peat (r = 0.87, p < 0.05 for second-order polynomial

fit), which is indicative of the influence of photochemical reduction (Bergquist and Blum 2007).

The potential source of this significant MIFodd may be a combination of both atmospheric

processes before Hg is deposited to the peat and mechanisms acting on the cycling of Hg in the

peat itself. Photo-oxidation of Hg(0) in the gas phase has previously been observed to result in a

slope of 1.6 to 1.9 for the relationship between Δ199

Hg and Δ201

Hg (Sun et al. 2016). This may

be occurring in the atmosphere before the Hg is deposited to the peat and is therefore reflected in

the peat isotopic signature. In addition to photo-oxidation of Hg(0) prior to deposition, abiotic

non-photochemical reduction of Hg(II) in the presence of organic matter, which produces a slope

of 1.5 to 1.6 (Zheng and Hintelmann 2010), may also contribute to the higher than expected

slope than would be observed due to photochemical reduction alone. It is reasonable that the

organic matter-rich peat may influence the reduction of Hg(II). Without direct experimental

evidence under controlled conditions however, the mechanisms governing MIFodd in addition to

photochemical reduction cannot be conclusively stated in this study.

157

Figure 5-3 Relationship between a) Δ199

Hg and Δ201

Hg and b) Δ199

Hg and δ202

Hg (all expressed

in ‰) for the surface 0-10 cm peat from both the S1 and PEATcosm peatland sites.

5.4.3 PEATcosm Peat Isotopic Signature

In contrast to the peat from the S1 bog, the control PEATcosm peat collected in 2011 prior to

experimental manipulations had near zero Δ199

Hg values (0.01 ± 0.02‰) and variable Δ201

Hg

values ranging from -0.04 to +0.23‰. Significant positive Δ200

Hg anomalies were also observed

in these samples (0.20 ± 0.14‰; Table D-5-2). The peat collected in 2014 from the PEATcosm

mesocosms, subjected to water table manipulations and altered plant communities, had higher

positive odd-isotope anomalies (Δ199

Hg = 0.10 ± 0.04‰, Δ201

Hg = 0.11 ± 0.10; mean ± standard

deviation) than both the pre-treatment peat and that from the S1 site. Similar to the S1 peat, the

PEATcosm peat had positive Δ200

Hg values (0.07 ± 0.02‰) with near-zero Δ204

Hg anomalies

(0.01 ± 0.03‰). Positive odd-mass isotopes have been previously observed in peat (Shi et al.

2011). Shi et al. (2011) observed enriched odd-mass isotopes in peat collected from a bog in

northern China, which was attributed to increasing anthropogenic Hg pollution with

industrialization. Similar to the S1 peat, negative MDF in the range of -1.47 to -1.04‰ was

observed in the PEATcosm peat. This again may be the result of direct deposition of Hg to the

peat via surface sorption as well as some deposition of vascular leaf litter in which Hg is

accumulated via stomatal uptake, which indirectly incorporates Hg into the peat (Zheng et al.

2016).

With MIFodd anomalies similar to that of Hg(II) in atmospheric samples (Gratz et al. 2010;

Sherman et al. 2012; Demers et al. 2013), the PEATcosm isotopic signatures suggest that the Hg

in the peat was predominantly the result of Hg(II) deposition. The mixing model using odd-

158

mass-number isotopes suggests that between 68-82% (74 ± 5%, mean ± standard deviation,

based on Δ199

Hg; Table D-5-4) to 59-91% (74 ± 12%, based on Δ201

Hg; Table 5-1) of the Hg in

the surface 10 cm of the PEATcosm peat was deposited as Hg(II). The fraction of Hg(II)

deposition is considerably lower when using the Δ200

Hg data, with only 53 ± 9% of Hg(II)

contributing to the peat Hg concentrations (Table D-5-5). The reason for this discrepancy in

mixing model results using the different isotopic values is unclear. Considerably more Hg(II)

deposition was reflected in the surface PEATcosm peat as compared to that collected from the

S1 bog based on the MIFodd model results. One potential factor contributing to this difference in

Hg(II) deposition may be the amount of snow accumulation during the winter months at these

two sites. Approximately 70% more snow falls at the PEATcosm site in Houghton, Michigan

(snow-water equivalent (SWE) of 435 mm per year) versus the Marcell Experimental Forest in

northern Minnesota (SWE of 260 mm per year) (Potvin et al. 2015; Sebestyen et al. 2011). Due

to the overstory tree canopy of spruce and tamarack within the S1 bog, the amount of snow

accumulation on the peatland is likely further diminished by canopy interception (Hedmond and

Pomeroy 1998; Sebestyen et al. 2011). Chen et al. (2012) observed predominantly positive

values of both Δ199

Hg (-0.29 to +0.73) and Δ200

Hg (+0.31 to +1.24) in non-Arctic snow collected

immediately after deposition. It was suggested that photochemically-driven Hg(0) oxidation

facilitated by snow crystals in the tropopause may account for the positive anomalies in Δ200

Hg,

while photoreduction on snow crystals may result in odd isotope MIF (Chen et al. 2012).

Lalonde et al. (2001) also observed photo-oxidation of Hg(0) in water and snow prior to

deposition. Ambient temperature may also influence the magnitude of Δ200

Hg anomalies, as

Chen et al. (2012) observed higher Δ200

Hg values in the winter months. Therefore, the positive

Δ200

Hg, Δ199

Hg and Δ201

Hg isotopic anomalies in the PEATcosm peat may be the result of

considerably greater snow accumulation that occurs at this site during the winter months as

compared to that received by the S1 peatland. Differences in snow accumulation between the

pre-treatment site of Meadowlands, MN (located southeast of the Marcell Experimental Forest)

and the experimental site of Houghton, MI may also account for the observed shift in the peat

isotopic signature between the pre-treatment (2011) and experimental (2014) peat.

Dominant contributions of oxidized Hg(II) to the PEATcosm peat was also supported by the

relationships among Δ201

Hg, Δ200

Hg and THg concentrations. Significant negative relationships

were observed not only between Δ201

Hg and peat THg concentrations (r = - 0.90, p = 0.02)

159

similar to the S1 peat, but also Δ200

Hg and surface peat THg (r = - 0.83, p = 0.02; Figure 5-1).

These relationships may suggest that the PEATcosm mesocosms each received similar amounts

of Hg(II), which supports the idea that snow accumulation may significantly influence the

isotopic signature of the surface peat. Smaller Δ201

Hg and Δ200

Hg anomalies with higher peat

THg concentrations may suggest the influence of enhanced Hg(0) contributions across the

treatments. Differences in Hg(0) deposition across the treatments may be reflected in a dilution

of the Hg(II) deposition signal. These influences were also reflected in the strong relationship

between MDF and TGM fluxes measured from the PEATcosm mesocosms (Figure 5-2b). The

mesocosms in which strong depositional fluxes were observed yielded the highest peat MDF

signatures, supporting the idea of direct Hg deposition via sorption to the surface peat. The

strong influence of Hg(II) deposition at the PEATcosm site as compared to the S1 peatland may

contribute to the observed differences in the relationship between fluxes and MDF at the two

sites (Figure 5-2). Previous research in forested systems suggested that Hg(II) may be more

readily available to be reduced upon deposition and subsequently re-emitted quickly to the

atmosphere (Graydon et al. 2009; Graydon et al. 2012). The control vegetation mesocosms

under both water table prescriptions which had the highest modelled Hg(II) contributions also

had the highest TGM emissions. Therefore, re-volatilization of Hg(II) deposited to the

PEATcosm peat surface may also occur quickly in peatlands and contribute to enhanced

emissions. Finally, the difference in the structure of the overstory at the two sites may also affect

the deposition patterns due to the stability of the atmospheric boundary layer above the peat

surface (Kyllönen et al. 2012; Mazur et al. 2014). The spruce and tamarack overstory tree

canopy above the vascular shrub cover at the S1 peatland may facilitate enhanced Hg(0)

deposition due to increased boundary layer stability as compared to the short Ericaceae shrub

cover and lack of overstory tree cover in the PEATcosm mesocosms. Further study under

controlled conditions is required to identify the role these factors may play in controlling Hg

cycling and the subsequent isotopic fractionation pattern.

5.4.4 PEATcosm Treatment Effects – Water Table and Plant Functional Groups

Effects of the PEATcosm water table and plant functional group treatments were reflected in the

isotopic signatures of the surface peat. The highest MIFodd anomalies and therefore the highest

modelled contributions of Hg(II) were observed in the High WT treatment peat, which received

160

the most precipitation to maintain the shallow water table positions (Figure 5-4). During the

spring months following the melting of the snowpack, runoff from the mesocosms was drained

to re-establish the treatment water table positions. Considerably more snowmelt runoff was

released from the Low WT treatments whereas less snowmelt was drained from the High WT

bins to maintain the relatively high water table position. Therefore, this distinction in Hg(II)

contributions between the water table treatments provides further support for the role of snow

and snowmelt water as the source of Hg(II) in the PEATcosm peat.

Figure 5-4 Δ199

Hg and Δ201

Hg values (in ‰) among the six crossed water table and plant

functional group treatments of the PEATcosm peat monoliths. Analyses were on pooled samples

from replicate treatments, thus error bars represent analytical precision, not sample replication

variability.

Under saturated conditions diffusion of Hg through soil pores is suppressed (Bahlmann et al.

2004; Gustin and Stamenkovic 2005; Song and Van Heyst 2005) and therefore gaseous Hg(0)

deposition into peat pores may be reduced. However, under lowered water tables, greater

gaseous Hg(0) may enter aerated peat pores where it may be oxidized and retained in the peat,

resulting in higher peat THg concentrations (Figure 5-1). Greater Hg(II) deposition in the High

161

WT treatments may be susceptible to reduction and re-emission from the peat (Graydon et al.

2009; Graydon et al. 2012). Mercury sorbed to the surfaces of leaves, particularly those of

vascular shrubs in the Ericaceae only and Control vegetation treatments, has been observed to

influence TGM fluxes (Haynes et al. 2017). Photo-reduction of carboxylic-bound Hg(II), which

is typically the dominant species on leaf surfaces, has been demonstrated to produce positive

MIF in the residual Hg(II) pool (Bergquist and Blum 2007; Zheng and Hintelmann 2009; Zheng

and Hintelmann 2010), which may contribute to the observed positive isotopic anomalies.

Smaller positive Δ199

Hg and Δ201

Hg anomalies for the Low WT treatments may also be due to

enhanced stomatal conductance of Hg(0) by vascular vegetation under drier conditions due to

increased productivity of the PEATcosm treatment aboveground biomass (Potvin et al. 2015),

which has been observed in forest soils under wetter conditions (Zheng et al. 2016).

5.4.5 PEATcosm Peat MIF

Unlike the clear relationship between Δ199

Hg and Δ201

Hg for the S1 peat, there was no similar

relationship for the experimental PEATcosm peat (p = 0.17; Figure 5-3). There was also no

significant relationship between Δ199

Hg and δ202

Hg (Figure 5-3). This trend of a significant

Δ201

Hg anomaly with no corresponding Δ199

Hg shift or MDF does not resemble any previously

constrained process such as abiotic non-photochemical Hg(II) reduction in the presence of

organic matter (Zheng and Hintelmann 2010), aqueous-phase photochemical reduction of Hg(II)

and photochemical degradation of methylmercury (Bergquist and Blum 2007), all of which may

be expected to play an important role in peatland Hg cycling (Jiskra et al. 2015). Interestingly,

there was a significant positive relationship between Δ201

Hg and Δ200

Hg for the PEATcosm peat

(r = 0.93, p = 0.01; Figure 5-5), while no similar relationship was observed for the S1 peat (p =

0.32).

162


Hg and Δ201


peat from both the S1 and PEATcosm peatland sites.

A positive relationship between Δ201

Hg and Δ200

Hg is also evident when collectively looking at

Hg(II) in precipitation collected by Gratz et al. (2010), Sherman et al. (2012) and Demers et al.

(2013), with the same slope of 0.18 (Figure D-5-3). This provides support to the binary mixing

model results demonstrating that the Hg in the surface PEATcosm peat is predominantly Hg(II),

with minimal Hg(0) contributions. However, no MIFeven relationship between Δ200

Hg and

Δ204

Hg was observed (Figure D-5-2), which is typically associated with atmospheric gas phase

processes (Gratz et al. 2010; Chen et al. 2012; Cai and Chen 2016) and has been documented in

forest soils (Zheng et al. 2016). To the best of our knowledge, this is the first documented case

of significant Δ201

Hg MIF with only a slight response in Δ199

Hg as well as a significant Δ200

Hg

anomaly without a corresponding shift in Δ204

Hg. As deposition to the PEATcosm peat was

occurring predominantly in the form of Hg(II), it is likely that the process producing this

fractionation pattern is associated with Hg(II). However, the mechanisms responsible for such

trends in the PEATcosm peat samples are not clear and require further investigation to assess the

processes that fractionate Hg isotopes in this manner.

5.4.6 Conclusions and Implications

This study demonstrates the contrasting Hg isotopic signatures in peat from two sub-boreal

peatlands. The observed peat odd- and even-mass isotopic anomalies reflect different Hg

sources as well as different processes, which act to influence Hg cycling in these peatlands.

163

Significant disparities in the fraction of Hg(0) and Hg(II) contributions were observed between

the S1 peatland in northern Minnesota and the PEATcosm peat in the Upper Peninsula of

Michigan. Differences in snow accumulation and subsequent snowmelt runoff between these

two sites may account for the observed peat isotopic compositions, with significant positive

Δ201

Hg and Δ200

Hg in the PEATcosm peat which receives nearly ~70% more snow (as SWE)

annually. Interception of snow by the spruce and tamarack cover in the S1 peatland as compared

to the lack of tree canopy at PEATcosm may further exacerbate the observed isotopic trends.

The influence of snow in contributing Hg to soils including peat warrants further study,

particularly for landscapes wherein snow accounts for a significant portion of the annual

hydrological budget. As climate change will likely affect snow accumulation in these regions

(Garris et al. 2015), understanding the contribution of snow Hg in controlling peat Hg

accumulation is crucial.

The strong relationship between the peat MDF and the TGM fluxes in the PEATcosm

mesocosms, and the lack of correlation for the S1 peat, may also be influenced by canopy cover

and the stability of the near-surface boundary layer. Deposition of predominantly Hg(II) in the

PEATcosm mesocosms and subsequent reduction, potentially on the leaf surfaces of the vascular

vegetation, may influence TGM fluxes from the peat. Whereas photochemical reduction appears

to be an important controlling influence on Hg cycling in the S1 peat, the mechanism responsible

for the MIFodd and MIFeven anomalies in the PEATcosm peat is unclear. The isotopic patterns as

a result of atmospheric processes require further study including the examination of non-arctic

snow. As the large stocks of Hg stored in peatlands (Kolka et al. 2001; Grigal 2003) are

vulnerable to the effects of climate change, understanding the mechanisms controlling Hg(0) and

Hg(II) contributions to peat, the mechanisms influencing Hg exchange with the peat, including

oxidation-reduction processes and the influence of vegetation, and therefore overall peat Hg

accumulation is warranted.


We would like to acknowledge the laboratory assistance of L. Zimmermann and M. Lee as well

as the field assistance of K. Ng and S. Rao. We thank L. Jamie Lamit for ambient peat collection

at the PEATcosm experiment. We would like to thank P. Hanson and W.R. Nettles of Oak

164

Ridge National Laboratory and the U.S. Forest Service for access to the SPRUCE plots in the S1

peatland of the Marcell Experimental Forest. Funding was provided through a Natural Sciences

and Engineering Research Council of Canada (NSERC) Alexander Graham Bell Canada

Graduate Scholarship (CGS-Doctoral) to K.M.H and a NSERC Discovery Grant (Fund #355866)

to C.P.J.M. The PEATcosm experiment was funded by the USDA Forest Service Northern

Research Station Climate Change Program and the National Science Foundation (DEB-

1146149).

5.6 References



Bergquist, B. A. and J. D. Blum (2007), Mass-dependent and –independent fractionation of Hg

isotopes by photoreduction in aquatic systems. Science, 318, 417–420.

Bergquist, B. A. and J. D. Blum (2009), The odds and evens of mercury isotopes: Applications

of mass-dependent and mass-independent isotope fractionation. Elements, 5, 353–357.

Blum, J. D. and B. A. Bergquist (2007), Reporting of variations in the natural isotopic

composition of mercury. Analytical and Bioanalytical Chemistry, 388, 353–359.

Bridgham, S. D., C. A. Johnston, J. Pastor, and K. Updegraff (1995), Potential feedbacks of

northern wetlands on climate change. BioScience, 45, 262–274.

Cai, H. and J. Chen (2016), Mass-independent fractionation of even mercury isotopes. Science

Bulletin, 61, 116–124.




165

Chen, J., H. Hintelmann, X. Feng, and B. Dimock (2012), Unusual fractionation of both odd and

even mercury isotopes in precipitation from Peterborough, ON, Canada. Geochimica et

Cosmochimica Acta, 90, 33–46.

Clymo, R. S. (1984), The limits to peat bog growth. Philosophical Transactions of the Royal

Society of London Series B, 303, 605-654.

Demers, J. D., J. D. Blum, and D. R. Zak (2013), Mercury isotopes in a forested ecosystem:

Implications for air-surface exchange dynamics and the global mercury cycle. Global

Biogeochemical Cycles, 27, 222–238.







Driscoll, C. T., R. P. Mason, H. M. Chan, D. J. Jacob, and N. Pirrone (2013), Mercury as a

global pollutant: sources, pathways, and effects. Environmental Science and Technology,

47, 4967–4983.


Hubble, E. K. Wong, L. Halfpenny-Mitchell, and M. S. Gustin (2001) Sources of

variability in mercury flux measurements. Journal of Geophysical Research –

Atmospheres, 106, 5421-5435.

Enrico, M., G. L. Roux, N. Marusczak, L. –E. Heimbürger, A. Claustres, X. Fu, R. Sun, and J. E.

Sonke (2016), Atmospheric mercury transfer to peat bogs dominated by gaseous

elemental mercury dry deposition. Environmental Science and Technology, 50, 2405–

2412.

166




Frolking, S., N. T. Roulet, T. R. Moore, P. J. H. Richard, M. Lavoie, and S. D. Muller (2001),

Modeling northern peatland decomposition and peat accumulation. Ecosystems, 4, 479-

498.

Garris, H. W., R. J. Mitchell, L. H. Fraser, and L. R. Barrett (2015), Forecasting climate change

impacts on the distribution of wetland habitat in the Midwestern United states. Global

Change Biology, 21, 766-776.




11810-11820.



Gratz, L. E., G. J. Keeler, J. D. Blum, and L. S. Sherman (2010), Isotopic composition and

fractionation of mercury in great lakes precipitation and ambient air. Environmental

Science and Technology, 44, 7764–7770.

Graydon, J. A., V. L. St. Louis, H. Hintelmann, S. E. Lindberg, K. A. Sandilands, J. W. M.

Rudd, C. A. Kelly, M. T. Tate, D. P. Krabbenhoft, and I. Lehnherr (2009), Investigation

of uptake and retention of atmospheric Hg(II) by boreal forest plants using stable Hg

isotopes. Environmental Science and Technology, 43, 4960–4966.

Graydon, J. A., V. L. St. Louis, S. E. Lindberg, K. A. Sandilands, J. W. M. Rudd, C. A. Kelly, R.

Harris, M. T. Tate, D. P. Krabbenhoft, C. A. Emmerton, H. Asmath, and M. Richardson

(2012), The role of terrestrial vegetation in atmospheric Hg deposition: Pools and fluxes

of spike and ambient Hg from the METAALICUS experiment. Global Biogeochemical

Cycles, 26, GB1022.

167





1326-1350.


emissions from soils. Biogeochemistry, 76, 215–232.

Hanson, P. J., S. E. Lindberg, T. Tabberer, J. G. Owens, and K. –H. Kim (1995), Foliar exchange

of mercury vapor: evidence for a compensation point. Water Air and Soil Pollution, 80,

373–382.

Hanson, P. J., J. S. Riggs, W. R. Nettles, J. R. Phillips, M. B. Krassovski, L. A. Hook, L. Gu, A.

D. Richardson, D. M. Aubrecht, D. M. Ricciuto, J. M. Warren, and C. Barbier (2017),

Attaining whole-ecosystem warming using air and deep soil heating methods with an

elevated CO2 atmosphere. Biogeosciences, 14, 861-883, DOI: 10.5194/bg-14-861-2017.

Haynes, K. M., E. S. Kane, L. Potvin, E. A. Lilleskov, R. K. Kolka, and C. P. J. Mitchell (2017),

Gaseous mercury fluxes in peatlands and the potential influence of climate change.

Atmospheric Environment, 154, 247-259.

Hedmond, N. R. and J. W. Pomeroy (1998), Measurements and modelling of snow interception

in the boreal forest. Hydrological Processes, 12, 1611-1625.

Jiskra, M., J. G. Wiederhold, U. Skyllberg, R. –M. Kronberg, I. Hajdas, and R. Kretzschmar

(2015), Mercury deposition and re-emission pathways in boreal forest soils investigated

with Hg isotope signatures. Environmental Science and Technology, 49, 7188–7196.



Journal, 65, 897-905.

168

Kritee, K., J. D. Blum, M. W. Johnson, B. A. Bergquist, and T. Barkay (2007), Mercury stable

isotope fractionation during reduction of Hg(II) to Hg(0) by mercury resistant

microorganisms. Environmental Science and Technology, 41, 1889–1895.



Letters, 30(17), 1900.



1182.




Lalonde, J. D., M. Amyot, A. M. L. Kraepiel, and F. M. M. Morel (2001), Photooxidation of

Hg(0) in artificial and natural waters. Environmental Science and Technology, 35, 1367-

1372.




Mazur, M., C. P. J. Mitchell, C. S. Eckley, S. L. Eggert, R. K. Kolka, S. D. Sebestyen, and E. B.

Swain (2014), Gaseous mercury fluxes from forest soils in response to forest harvesting

intensity: A field manipulation experiment. Science of the Total Environment, 496, 678-

687.

Moore, T. and N. Basiliko (2006), Decomposition in boreal peatlands. In R. K. Wieder and D. H.

Vitt, eds. Boreal peatland ecosystems. Springer-Verlag, Berlin, Heidelberg, Germany,

125-143.

Moore, C. and A. Carpi (2005), Mechanisms of the emission of mercury from soil: role of UV

radiation. Journal of Geophysical Research, 110, D24302.

169

Moore, C. W. and M. S. Castro (2012), Investigation of factors affecting gaseous mercury

concentrations in soils. Science of the Total Environment, 419, 136−143.




and Soil, 387, 277-294.







foliage over time in two northern mixed hardwood forests. Water Air and Soil Pollution,

133, 49-67.

Sebestyen, S. D., C. Dorrance, D. M. Olson, E. S. Verry, R. K. Kolka, A. E. Elling, et al. (2011),

Long-term monitoring sites and trends at the Marcell Experimental Forest. In: Kolka, R.

K., S. D. Sebestyen, E. S. Verry, K. N. Brooks, editors. Peatland biogeochemistry and

watershed hydrology at the Marcell Experimental Forest. Boca Raton: CRC Press, 15–

71.

Sherman, L. S., J. D. Blum, G. J. Keeler, J. D. Demers, and J. T. Dvonch (2012), Investigation of

local mercury deposition from a coal-fired power plant using mercury isotopes.

Environmental Science and Technology, 46, 382–390.

Shi, W. F., X. B. Feng, G. Zhang, L. L. Ming, R. S. Yin, Z. Q. Zhao, and J. Wang (2011), High-

precision measurement of mercury isotope ratios of atmospheric deposition over the past

150 years recorded in a peat core taken from Hongyuan, Sichuan Province, China.

Chinese Science Bulletin, 56, 877−882.

170








Strack, M., J. M. Waddington, L. Rochefort, and E. –S. Tuittila (2006), Response of vegetation



Sun, G., J. Sommar, X. Feng, C. –J. Lin, M. Ge, W. Wang, R. Yin, X. Fu, and L. Shang (2016),

Mass-dependent and -independent fractionation of mercury isotope during gas-phase

oxidation of elemental mercury vapor by atomic Cl and Br. Environmental Science and








Trettin, C. C., R. Laiho, K. Minkkinen, and J. Laine (2006), Influence of climate change factors

on carbon dynamics in northern forested peatlands. Canadian Journal of Soil Science, 86,

269-280.

Turunen, J., E. Tomppo, K. Tolonen, and A. Reinikainen (2002), Estimating carbon

accumulation rates of undrained mires in Finland - application to boreal and subarctic

regions. The Holocene, 12, 69-80.

171






Weltzin, J. F., J. Pastor, C. Harth, S. D. Bridgham, K. Updegraff and C. T. Chapin (2000),


Ecology, 81, 3464-3478.



9, 141-151.



3589-3600.

Yin, R., X. Feng, and W. Shi (2010), Application of the stable-isotope system to the study of

sources and fate of Hg in the environment: A review. Applied Geochemistry, 25, 1467-

1477.

Zheng, W. and H. Hintelmann (2009), Mercury isotope fractionation during photoreduction in

natural water is controlled by its Hg/DOC ratio. Geochimica et Cosmochimica Acta, 73,

6704−6715.

Zheng, W. and H. Hintelmann (2010), Nuclear field shift effect in isotope fractionation of

mercury during abiotic reduction in the absence of light. Journal of Physical Chemistry

A, 114, 4238–4245.

Zheng, W., D. Obrist, D. Weis, and B. A. Bergquist (2016), Mercury isotope compositions

across North American forests. Global Biogeochemical Cycles, 30, 1475-1492.

172

Chapter 6 Summary and Synthesis

6

6.1 Summary

The overarching objective of this research was to identify the potential impacts of climatic

change on Hg cycling in peatlands using mesocosm- and field-scale approaches. In order to

address the objectives and questions set forth at the beginning of this thesis, this research used

two novel experiments designed to explore different anticipated climate change effects in sub-

boreal peatland ecosystems. Sampling was conducted primarily at the mesocosm-scale at the

PEATcosm experiment, which investigated the effects of altered water table positions and

shifting vascular plant communities, anticipated to occur with climate change. Field-scale work

was also conducted in the SPRUCE (S1) peatland of the Marcell Experimental Forest, which

underwent a deep peat warming experiment beginning in June 2014 and is now the site of the

whole ecosystem warming and elevated atmospheric carbon dioxide experiment. Total Hg and

MeHg concentrations in pore water and snowmelt runoff as well as in the solid phase peat were

monitored in the PEATcosm peat monoliths over several years. Experimental research

examining gaseous Hg fluxes in natural environments is lacking in the literature, likely due to the

numerous challenges associated with field flux measurements. In this research, great care was

taken to ensure consistent meteorological conditions during sampling events and to account for

variations in such factors as plant community coverage and hydrology. By mitigating these

challenges, climate change impacts of hydrology and vascular plant communities at the

PEATcosm experiment and peat warming at depth at the SPRUCE experiment on total gaseous

Hg fluxes could be assessed. Natural abundance Hg isotopic analyses were applied to

investigate relative contributions of Hg(0) and Hg(II) deposition to surface peat of the two

peatlands as well as to gain insight into the mechanisms potentially controlling peat Hg cycling.

The effects of the PEATcosm climate change manipulations on Hg cycling processes in the

surface peat were also explored using isotopic analyses. The following subsections are meant

173

not only to summarize conclusions from individual thesis chapters, but also to progressively

synthesize across chapter findings within each section.

6.1.1 Hg and MeHg Mobility in Peat Pore Water and Runoff

Peatland ecosystems are known to be significant sources of MeHg to downstream aquatic

ecosystems. Lowered, fluctuating water tables and the removal of Ericaceae shrub vegetation

resulted in enhanced Hg and MeHg concentrations in the peat pore waters (Chapter 2). This

treatment yielded the highest concentrations of Hg and MeHg in snowmelt runoff exported from

the peat monoliths. A significant positive correlation was observed between Hg and dissolved

organic carbon (DOC) concentrations in both the peat pore waters and snowmelt runoff. This

suggests that one factor driving the enhanced leaching of Hg and MeHg from the solid peat into

the mobile phase with simulated climate change effects is increased peat decomposition.

Significant positive correlations between the amount of mass lost from cellulose decomposition

assays, representing potential peat decomposition, and both THg and MeHg concentrations in

pore waters are in agreement with peat decomposition influencing the mobilization and

accumulation of inorganic Hg and MeHg in pore waters. In addition to the influence of water

table position, plant functional groups affected Hg and MeHg mobility with increased

concentrations when Ericaceae shrubs were removed under both water table treatments.

Although not measured, it is likely that root exudation of oxygen and labile organic carbon by

sedges contribute to augmented pore water Hg and MeHg accumulation by enhancing peat

decomposition as well as potentially priming the process of methylation. Oxygenation within the

sedge rooting zone due to root oxygen loss may contribute to both aerobic peat mineralization

and regeneration of alternative electron acceptors. Within the PEATcosm pore waters, sulfate

concentrations were enhanced for both the lowered water table treatments and the sedge-

dominated peat monoliths, providing support for these conditions to regenerate terminal electron

acceptors and enhance Hg methylation.

174

6.1.2 Solid Phase Hg and MeHg – Translocation and Net MeHg Production

Over the course of the PEATcosm experiment, inorganic Hg and MeHg partitioning from the

peat to the pore waters was significantly influenced by the water table and plant functional group

treatments (Chapter 4). Greater partitioning into the aqueous phase was observed with lowered

and more variable water table positions in the upper depths of peat affected by increased peat

aeration. Changes in vascular vegetation cover also impacted inorganic Hg and MeHg

partitioning to pore water below the rooting zone, with enhanced mobilization in pore waters

when Ericaceae shrubs were removed. These results are in good agreement with the Chapter 2

findings in the peat pore waters.

In Chapter 4, significant changes in the peat THg and MeHg depth profiles were observed

throughout the course of the experiment, with the peak in THg and MeHg concentrations

occurring approximately 30-40 cm beneath the peat coinciding with the depth of the lowered

water tables in 2014. Within the rooting zone, MeHg was also significantly elevated when

Ericaceae shrubs were removed. Surprisingly, no significant differences in methylation or

demethylation were observed among either the water table or plant functional group treatments.

Collectively, this lends support to peat decomposition under aerobic conditions, subsequent

leaching of DOC and partitioning of both inorganic Hg and MeHg from the solid phase peat into

the aqueous phase as the processes instrumental in controlling the accumulation in pore waters.

Mercury partitioned into the pore waters may be transported to the depths near the lowered tables

and re-sorbed to the solid phase, potentially due to the affinity of Hg for organic matter (Figure

6-1). Translocation of inorganic Hg and MeHg within the peat profile appears to be a stronger

influence in governing the observed PEATcosm experimental peat profiles than net MeHg

production.

175

Figure 6-1 Influence of water table lowering on the partitioning of Hg and MeHg from the solid

phase into the aqueous pore water phase, translocation down the peat profile and re-sorption to

the solid phase within the zone of lowered water table fluctuation.

The fact that methylation potentials were similar between the two water table treatments at a

depth of 35-40 cm below the peat surface, where the lowered water tables were maintained at the

time of incubation, suggests that methylation may be stimulated despite increased oxygenation of

the peat. Methylmercury production is known to occur under anoxic conditions suitable for

sulfate- and iron-reduction, which accounts for the observed methylation potentials under

saturated conditions. Under lowered water table conditions, however, Hg methylation may be

comparable to the saturated conditions as a result of terminal electron acceptor regeneration

priming the process. The elevated sulfate concentrations observed in the pore waters of the

lowered water table treatments in Chapter 2, particularly at a depth of 40 cm below the peat

surface suggest that terminal electron acceptor regeneration may be a contributing factor to this

finding of similar methylation under drier conditions. Available terminal electron acceptors, as

well as a supply of labile carbon from both peat mineralization and sedge root exudation, may

collectively act to stimulate Hg methylation within the zone of lowered, fluctuating water tables

similar to that produced by the anoxia alone under saturated conditions. Despite the anticipated

loss of sustained anoxic conditions within the surface peat depths with a changing climate, the

hydrological conditions and subsequent shifts in plant communities may augment the mobility of

inorganic Hg and MeHg, both within the peat profile as well as partitioning into the peat pore

waters available for export from the systems during high-flow events such as spring snowmelt.

176

6.1.3 Vascular Plant Communities – Interactions with Atmospheric Hg Deposition

The dominant vascular plant communities present in the peatland may also significantly affect

the pathway by which Hg is deposited to the peat. In Chapter 3, the strongest trends of total

gaseous Hg (TGM) deposition in the PEATcosm peat monoliths were observed when sedges

were present. In comparison, Ericaceae-dominated peat resulted in predominantly emissive

fluxes, the strength of which was correlated with the amount of Ericaceae leaf coverage. As the

sedges had significantly lower surface-sorbed and tissue Hg concentrations than the Ericaceae

shrubs, TGM may be shuttled to the peat in the rooting zone, incidentally with oxygen, via the

aerenchymous tissues of the sedges. Shuttling of TGM to the roots of the sedges may contribute

to the elevated Hg concentrations in the solid phase at these depths observed in Chapter 4.

Strong TGM deposition in the sedge-dominated peat is evident in the Hg isotope data (Chapter

5). The highest mass-dependent fractionation (MDF) among the PEATcosm treatments was

indeed in the surface peat with sedge cover.

In contrast to the sedges, Ericaceae shrub leaves had higher surface as well as leaf tissue Hg

concentrations (Chapter 3). The character of the shrub leaves may promote surface sorption.

Stomatal uptake of gaseous Hg likely contributes to elevated tissue Hg concentrations. The

different pathways, both stomatal and non-stomatal, by which gaseous Hg is deposited to the

Ericaceae shrub cover, are reflected in the peat isotopic signatures measured in Chapter 5. More

negative MDF in the Ericaceae only and unmanipulated vegetation treatments of the PEATcosm

experiment suggest that stomatal uptake of gaseous Hg by the shrubs, and subsequent deposition

of litter of the leaves during senescence, is reflected in the surface peat. Reduction mechanisms

may contribute to the observed MDF anomalies, but further controlled study of such processes is

required to identify those active in the peat soils. The considerable TGM emissions observed

with the Ericaceae-dominated peat monoliths (Chapter 3) was evident in the peat isotopic data

(Chapter 5). Positive mass-independent fractionation (MIF) anomalies in the peat as well as the

observed relationship between TGM fluxes and MDF may be indicative of photo-reduction of

Hg(II), which is typically the dominant form of Hg found on leaf surfaces. The agreement

between the observed TGM fluxes among the PEATcosm plant community treatments in

Chapter 3 and the Hg isotopic signatures in Chapter 5 suggests that shifting plant functional

177

groups in a changing climate will alter the mechanisms and pathways by which gaseous Hg is

deposited to peatland systems.

6.1.4 Spatial Variability in Hg Isotopic Signatures

In addition to vegetation influences on Hg deposition, the geographical location and canopy

structure of peatlands influence the sources and cycling of Hg (Chapter 5). Dramatic differences

in peat isotopic signatures between the northern Minnesota (S1 peatland) and northern Michigan

(PEATcosm) peatland sites highlight considerable differences in Hg(0) and Hg(II) deposition.

The isotopic evidence suggests that significant differences in the amount of snow water

equivalent received by the two sites during spring snowmelt period accounts for the isotopic

disparities between the two peatlands. The increasing contributions of Hg(II) from snow at the

PEATcosm site was confirmed by the differences in MIFodd values between the water table

treatments; with higher Δ199

Hg and Δ201

Hg values in the wetter peat monoliths. The presence of

a tree canopy at the S1 (SPRUCE) peatland increased the contributions of Hg(0) to the peat,

possibly due to boundary layer stability, as compared to the PEATcosm site. The previously

undocumented MIFodd and MIFeven isotopic anomalies observed in the PEATcosm peat samples

require further research to assess the sources and processes that may fractionate Hg in this

manner.

6.1.5 Impacts of Climate Change on Peatland Hg Cycling

The two experimental sites, PEATcosm and SPRUCE, were established to simulate anticipated

climate change impacts on peatlands. In this thesis, the effects of water table variations, different

vascular plant functional group assemblages and peat warming at depth on Hg cycling were

investigated. The vascular plant functional groups examined in the PEATcosm experiment

interact with atmospheric Hg in different ways and thereby affect the pathways by which Hg is

deposited to peatlands. Shuttling of Hg to peat soils via aerenchymous tissues may act to lessen

atmospheric Hg burdens. However, enhanced deposition to the peat may increase the amount of

Hg available to be methylated. On the other hand, Ericaceae shrubs accumulate Hg in the leaf

tissues and are subsequently deposited to peat as litter upon senescence. Sorption of Hg to

178

Ericaceae leaf surfaces has the potential to be a pathway of significant Hg re-emission to the

atmosphere. Therefore, as the direction in which vascular plant communities will shift with

climate change (dominance of sedges vs. Ericaceae shrubs) is still unclear, the manner by which

gaseous Hg exchange between peatlands and the atmosphere will progress cannot yet be

conclusively stated. The SPRUCE peat warming at depth may be an indication that increasing

peat temperatures enhance gaseous Hg fluxes from the peat to the atmosphere. This effect may

be amplified when air temperatures are also increased as part of the SPRUCE experiment, which

will result in a more even warming of the peat. Partitioning of inorganic Hg and MeHg from the

solid phase to the mobile pore waters increased as a result of enhanced aerobic decomposition

with lowered water tables and the removal of Ericaceae shrub cover. Lowered and more variable

water table positions, and to a lesser extent the removal of Ericaceae shrubs, has the potential to

increase loading of Hg and MeHg to downstream aquatic ecosystems. The considerable

influence of snow meltwater in contributing to the peat Hg stocks was an unexpected result of

this research. In regions where snow accumulation is expected to increase with climate change,

greater deposition of Hg(II) to vegetation surfaces may delay and reduce the incorporation of Hg

into the soil pool and potentially result in enhanced gaseous Hg emissions to the atmosphere, as

Hg(II) on ground vegetation is susceptible to reduction and re-emission. The differences in

Hg(0) and Hg(II) contributions due to peatland canopy structure as well as deposition via snow

may significantly impact Hg cycling in these systems.

6.2 Limitations and Future Research Directions

Conducting peatland research at the mesocosm-scale and at the field plot-scale allows for the

direct and largely controlled manipulation of the factors under study. This approach effectively

removes the influences of potential confounding factors such as groundwater contributions,

which cannot be easily accounted for at larger scales in a natural system. These studies were

conducted in peatlands located at the southern extent of the boreal ecosystem. Given that this

research examined the direct influences of hydrology, plant communities and deep soil warming

on Hg cycling, the observations made in these sites are transferable to peatlands located in

temperate, northern boreal and arctic ecosystems. This research demonstrates how peatland Hg

cycling may be mechanistically impacted by the imposed climate change manipulations of

179

lowered water tables, different vascular plant community assemblages and increasing peat

temperatures. Depending upon how climate change may influence these factors within a certain

region, the potential effects on Hg cycling could be interpreted accordingly. An important next

step in this research could involve the investigation of these factors on Hg cycling across

latitudinal gradients. The observed significance in snow accumulation in contributing to peat Hg

stocks, for example, lends itself to further analysis. A review of the current literature examining

Hg isotope analyses across a gradient in snow accumulation could be conducted to further

examine the mechanistic trends observed in this thesis.

An exciting international collaborative approach, termed PeatDataHub, has been recently

initiated to address global-scale research questions in peatland science (Young et al. 2016). By

sharing data sets collected from long-term peatland monitoring sites as well as short-term

measurements collected around the world, peatland researchers can synthesize observations

across a range of scales and topics. The aggregation of peatland data would, for example,

advance the understanding of changes in hydrology and ecology anticipated with climate change

on the global peatland response in terms of carbon sink-source functions and feedback

mechanisms. Investigation of other potential issues of climate change including the cycling and

mobility of Hg in peatlands would benefit from global collaborative efforts such as the

PeatDataHub. Currently in the literature, few studies examine the impacts of climate change on

Hg cycling processes in peatlands. These studies are typically restricted to boreal landscapes in

North America, such as this research, and the arctic and boreal regions of Fennoscandia (e.g.

Rydberg et al. 2010; Fritsche et al. 2014). Expansion of peatland Hg research in relation to

global climate change would be facilitated by a collaborative global network, including the

potential for synthesizing results across latitudinal and other gradients. In addition to the impacts

of a changing climate, the effects of anthropogenic land use change, such as drainage of

peatlands and harvesting of peat, on the cycling of both Hg and carbon could also be an avenue

for extensive research facilitated by such a network. One major opportunity presented by the

PeatDataHub network is the potential to influence and inform policy and peatland management

strategy practices (Young et al. 2016). Collaborative global research could incorporate not only

peatland hydrology and carbon storage capabilities, but also strategies to mitigate potential

effects on Hg cycling, export from peatlands, and availability for uptake into vulnerable aquatic

food chains.

180

Planned future research with the PEATcosm data involves a collaborative, interdisciplinary

effort to examine the hgcAB gene pair in the peat using the metagenomic and metatranscriptomic

data on bacteria and archaea analyzed in association with the Joint Genome Institute (JGI)

project. This data will facilitate the assessment of the effects of the water table and plant

functional group treatments on Hg methylation gene presence and activity. These results could

then be related to the trends in methylation and demethylation rate potentials (Chapter 4) to

determine whether the genetic potential for methylation is either supportive of or decoupled from

kmeth and kdemeth.

The next steps in examining the potential effects of climate change on Hg cycling in peatlands

should involve experimental manipulation of water tables and vascular plant functional groups at

the peatland-scale. Long-term peatland water table manipulations, such as the poor fen sites at

the Seney National Wildlife Refuge in the Upper Peninsula of Michigan examining carbon

cycling (Chimner et al. 2016), would be ideal sites to examine the subsequent impacts on Hg

cycling. Free air carbon enrichment studies, such as the SPRUCE experiment now underway

with whole ecosystem warming, will provide insight into how plant communities may shift and

peatland water tables may react naturally in response to warming temperatures and elevated

atmospheric carbon dioxide concentrations. A targeted approach could then be taken to aid in

modelling the subsequent effects on Hg cycling.

Extensive further research should be carried out using natural abundance stable Hg isotope

fractionation patterns. Given the observed significance of geographic location on Hg cycling in

the two study sites in this thesis, further isotopic work is required to examine the potential spatial

variation in peatlands globally. Controlled experiments of Hg cycling mechanisms, including

deposition, reduction and atmospheric cycling mechanisms, would aid in identifying the isotopic

patterns observed in natural samples: peat, pore waters, vegetation, bulk gaseous samples and

precipitation including snow. Understanding the processes controlling Hg cycling will greatly

enhance the ability to model how peatland Hg cycling may be affected by the anticipated

changes in global climate.

181

6.3 References

Chimner, R. A., T. G. Pypker, J. A. Hribljan, P. A. Moore, and J. M. Waddington (2016), Multi-

decadal changes in water table levels alter peatland carbon cycling. Ecosystems,

DOI:10.1007/s10021-016-0092-x.




Rydberg, J., J. Klaminder, P. Rosén, and R. Bindler (2010), Climate driven release of carbon and

mercury from permafrost mires increases mercury loading to sub-arctic lakes. Science of

the Total Environment, 408, 4778-4783.

Young, D. M., P. J. Morris, and J. Holden (2016), Upscaling peatland science through

collaborative big data. Eos, 97, https://doi.org/10.1029/2016EO06125.

182

Appendix A. Supplementary Information for Chapter 2

Text A-2-1

All data were tested for normality (Shapiro-Wilk W test) and heteroscedasticity and were

successfully log-transformed or square root-transformed to achieve normality when parametric

assumptions were not met. The data from the midsummer pore water collections (August 2013

and July 2014) were removed from statistical analyses as only a small percentage (50% for

August 2013, 65% for July 2014) of the samples could be obtained. Due to low midsummer

water table levels insufficient water was available for extraction from the peat monoliths,

particularly at the 20 and 40 cm depths for the low WT treatment mesocosms, which inhibited a

direct statistical comparison only during this time. As the purpose of this study was to assess the

treatment effects on Hg mobilization in pore water which could subsequently be transported out

of the peatland during high-flow events, we feel that the inclusion of these partial midsummer

sampling events would in any case not address our objectives, because very little water is

exported from peatlands in streamflow during mid-summer low water periods. For all of the

other collection events, more than 80% of possible pore water samples were collected. Missing

samples, due to the lack of available sample for analysis during these collections, were treated as

‘not available’ for statistical analyses.

The main influences of water table position and vascular plant functional group on pore water

THg concentrations, MeHg concentrations and the percentage of THg as MeHg (%MeHg) over

the three sampling depths were assessed using repeated measures analyses of variance

(ANOVA), with water table, plant functional group and sampling depth as the main factors,

taking into consideration the repeated sampling events of the pore waters over time. No

significant interactive effects of water table position and vascular plant functional group were

observed for pore water Hg concentrations. Therefore, one-way ANOVAs with post-hoc Tukey

tests were performed to discern significant differences in THg and MeHg concentrations, and

%MeHg between the six crossed treatment combinations. Water table and plant community

effects as well as any interactive effect between these factors on THg and MeHg concentrations,

and %MeHg in snowmelt runoff were analyzed using three-way repeated measures ANOVAs to

183

account for any potential differences in the main factors between the two sampling years. One-

way ANOVAs with post-hoc Tukey tests were performed to assess significant differences in

THg concentrations, MeHg concentrations and %MeHg in snowmelt between the six crossed

treatment combinations given no interactive effects of the main experimental factors.

Correlation matrices were created using Pearson correlation coefficients in order to assess any

significant relationships between pore water Hg concentrations and dissolved organic carbon

(DOC), total phenolics concentrations as well as Hg and DOC concentrations in snowmelt

runoff.

184

Figure A-2-1 Photos of mesocosm bins.

Sedge Only

Ericaceae Only

Unmanipulated Control

185

Figure A-2-2 Runoff total Hg and MeHg loads across the six crossed water table and vascular

plant functional group treatments during the 2014 spring snowmelt period (n=4 per treatment). a)

Total Hg and b) MeHg. Letters denote statistically similar groups based on transformed data.

186

Figure A-2-3 Mean pore water MeHg concentrations (on log scale) during the five sampling


represents the mean pore water concentration of all three sampling depths across the four

replicate mesocosms (n=10 - 12).

187

Figure A-2-4 Mean pore water THg concentrations (on log scale) during the five sampling


represents the mean pore water concentrations of all three sampling depths across the four

replicate mesocosms (n=10 - 12).

188

Figure A-2-5 Pore water Hg-DOC relationships. a) THg (log scale) in relation to DOC (square

root transformed) concentration, and b) MeHg (square root transformed) in relation to DOC

(square root transformed) concentration. The data includes all three pore water depths from each

of the four replicates per each of the six treatments. All five sampling events are included.

189

Figure A-2-6 Relationships between shallow pore water (20 and 40 cm below the peat surface)

Hg and % mass loss of cellulose decomposition assays in the top 40 cm of the peat. a) THg

concentrations, b) MeHg concentrations, c) %MeHg. Decomposition assays were harvested in

2014 and represent potential peat decomposition among the treatments. From all five sampling

events, pore water Hg data collected at 20 and 40 cm depths was averaged by mesocosm. n=18-

20 per treatment.

190

Figure A-2-7 Locally-weighted scatterplot smoothing (LOWESS) relationships between shallow

pore water (20 and 40 cm below the peat surface) Hg and % mass loss of cellulose

decomposition assays in the top 40 cm of the peat. a) THg concentrations, b) MeHg

concentrations, c) %MeHg. Decomposition assays were harvested in 2014 and represent

potential peat decomposition among the treatments. From all five sampling events, pore water

Hg data collected at 20 and 40 cm depths was averaged by mesocosm. n=18-20 per treatment.

191

Figure A-2-8 Pore water Hg-total phenolics relationships. a) THg (log transformed) in relation

to total phenolics (square root transformed) concentrations, and b) MeHg (square root

transformed) in relation to total phenolics (square root transformed) concentrations. The data

includes all three pore water sampling depths from each of the four replicates per each of the six

treatments. All five sampling events are included.

192

Figure A-2-9 Sulfate concentrations among the six crossed WT and plant functional group

treatments including all three sampling depths for the 2013 and 2014 sampling events. Letters

denote statistically similar groups.

193

Table A-2-1 Volume of water drained (mean ± standard deviation, in L) from the mesocosm

bins during spring snowmelt 2014 to achieve the water table treatment positions (n=4 per

treatment).

Mean ± Standard Deviation

Volume of Water Drained (L)

High WT Ericaceae 68 ± 31

High WT Sedge 89 ± 35

High WT Control 59 ± 34

Low WT Ericaceae 119 ± 25

Low WT Sedge 165 ± 7

Low WT Control 108 ± 16

194

Appendix B. Supplementary Information for Chapter 3

Figure B-3-1 a) Relationship between mean inlet TGM concentrations (ng m-3

) and

corresponding 20-min TGM fluxes measured during daylight hours across the PEATcosm

experimental treatments. b) Boxplots of mean inlet Hg concentrations across the four PEATcosm

treatments. Letters denote statistically similar values.

195

Figure B-3-2 Piecewise regressions of PEATcosm hourly TGM fluxes (in ng m-2

h-1

) and

surface soil temperatures (in °C) for the a) High WT Control, b) Low WT Control, c) Low WT

Sedge, and d) Low WT Ericaceae treatments. Temperature thresholds are noted for each

treatment by the vertical dashed lines. The equations of the relationships before and after each

threshold temperature are specified.

196

Figure B-3-3 a), c) and e) Boxplots of mean inlet TGM concentrations measured during daylight

hours across the SPRUCE plots for each of the three sampling events. Letters denote statistically

similar values. b), d) and f) Relationship between mean inlet TGM concentrations (ng m-3

) and

corresponding 20-min TGM fluxes (daylight only) across the SPRUCE plots during each of the

three samplings.

197

Appendix C. Supplementary Information for Chapter 4

Figure C-4-1 Peat MeHg and THg stock (in ng cm-3

) profiles for the six combined water table

and vascular plant functional group treatments throughout the PEATcosm experiment from 2011

to 2014 and associated mean June to October water table positions for the High and Low WT

prescriptions. For clarity, error bars for each 10 cm depth increment have been omitted.

198

Figure C-4-2 Mean kdemeth (d-1

) in the upper 15 cm of the peat profile and 35-50 cm below the

peat surface for the a) High WT and b) Low WT treatments. Mean water table levels for the

month prior to sampling are denoted by the dashed lines. No statistically significant differences

among the depths for each of the High and Low WT data.

199

Figure C-4-3 Soil-water partition coefficients (logKD) for Hg(II) at 20 and 40 cm below the peat


2014 (n = 4 per treatment). Letters denote statistically similar groups.

200

Appendix D. Supplementary Information for Chapter 5

Figure D-5-1a S1 peatland of the Marcell Experimental Forest in north-central Minnesota

showing one SPRUCE plot footprint.

201

Figure D-5-1b PEATcosm Mesocosm Facility, Houghton, Michigan, with the 24 peat monoliths

(top) and underlain by a climate-controlled tunnel (bottom).

202


Hg and Δ204

Hg (all expressed in ‰) for the surface 0-10

cm peat from both the S1 and PEATcosm peatland sites.

203


Hg and Δ201


peat from both the S1 and PEATcosm peatland sites as compared to reported values of Hg(II)

deposition in the literature (Gratz et al. 2010; Sherman et al. 2012; Demers et al. 2013).

204

Table D-5-1 Mercury isotope signatures of reference materials and standards.

Sample ID n Recovery SD

δ204Hg

(‰) 2σ

δ202Hg

(‰) 2σ

δ201Hg

(‰) 2σ

δ200Hg

(‰) 2σ

δ199Hg

(‰) 2σ

Δ204Hg

(‰) 2σ

Δ201Hg

(‰) 2σ

Δ200Hg

(‰) 2σ

Δ199Hg

(‰) 2σ

NIST SRM 1944 2 107% 24% -0.58 0.06 -0.39 0.05 -0.30 0.06 -0.20 0.03 -0.08 0.04 0.01 0.07 -0.01 0.04 0.00 0.02 0.02 0.04

NIST SRM 1646a 1 80% N/A -1.40 0.18 -0.97 0.08 -0.68 0.14 -0.44 0.04 -0.10 0.04 0.04 0.07 0.05 0.08 0.05 0.02 0.14 0.04

CCRMP TILL-1 9 86% 5% -1.69 0.06 -1.13 0.05 -0.97 0.06 -0.55 0.03 -0.38 0.04 -0.01 0.07 -0.12 0.04 0.01 0.02 -0.09 0.04

NIST SRM 3133 7 91% 3% -0.05 0.06 -0.01 0.05 0.00 0.06 -0.01 0.03 0.00 0.04 -0.04 0.07 0.01 0.06 0.00 0.02 0.00 0.04

JTBaker 22 -0.87 0.06 -0.58 0.05 -0.43 0.06 -0.28 0.03 -0.12 0.04 -0.01 0.07 0.00 0.04 0.01 0.02 0.03 0.04

Note: 2σ is the higher of either 2 standard deviation (2SD) of the JTBaker standard or 2 standard error (2SE) of the mean.

205

Table D-5-2 Mercury isotope signatures of all peat samples. Each sample was analyzed twice for isotopes. 2σ is the higher of either 2

standard deviation (2SD) of the JTBaker standard or 2 standard error (2SE) of the two replicate results of each sample.

Sample ID

THg

(ng g-1)

δ204Hg

(‰) 2σ

δ202Hg

(‰) 2σ

δ201Hg

(‰) 2σ

δ200Hg

(‰) 2σ

δ199Hg

(‰) 2σ

Δ204Hg

(‰) 2σ

Δ201Hg

(‰) 2σ

Δ200Hg

(‰) 2σ

Δ199Hg

(‰) 2σ


Low WT – Sedge 39 -1.64 0.06 -1.07 0.05 -0.78 0.07 -0.49 0.07 -0.18 0.08 -0.04 0.07 0.03 0.04 0.05 0.05 0.09 0.07

Low WT – Ericaceae 38 -1.89 0.06 -1.27 0.05 -0.96 0.06 -0.60 0.05 -0.27 0.04 0.01 0.07 -0.01 0.04 0.04 0.04 0.05 0.04

Low WT – Control 30 -2.17 0.06 -1.47 0.05 -1.03 0.06 -0.67 0.04 -0.29 0.04 0.02 0.07 0.07 0.04 0.07 0.02 0.08 0.04

High WT – Sedge 32 -1.54 0.09 -1.04 0.05 -0.63 0.06 -0.44 0.03 -0.09 0.04 0.01 0.07 0.16 0.04 0.08 0.03 0.17 0.04

High WT – Ericaceae 24 -1.64 0.06 -1.12 0.05 -0.66 0.06 -0.49 0.03 -0.17 0.04 0.03 0.07 0.18 0.04 0.08 0.02 0.11 0.04

High WT - Control 24 -1.93 0.06 -1.33 0.05 -0.75 0.06 -0.58 0.03 -0.22 0.04 0.05 0.07 0.25 0.04 0.08 0.03 0.11 0.04

PEATcosm Pre-Treatment Peat – 2011

Low WT – Sedge 52 -1.88 0.20 -1.23 0.26 -0.95 0.19 -0.52 0.16 -0.31 0.18 -0.05 0.18 -0.03 0.04 0.09 0.03 -0.01 0.11

Low WT – Ericaceae 55 -2.11 0.11 -1.34 0.10 -0.78 0.06 -0.31 0.09 -0.30 0.04 -0.10 0.07 0.23 0.04 0.36 0.04 0.03 0.04

High WT – Sedge 54 -1.42 0.06 -1.01 0.05 -0.80 0.06 -0.35 0.07 -0.24 0.04 0.09 0.07 -0.04 0.05 0.15 0.07 0.01 0.04

S1 Peat – 2014

Plot 4 69 -2.60 0.06 -1.74 0.05 -1.53 0.06 -0.83 0.04 -0.72 0.04 -0.01 0.07 -0.22 0.04 0.04 0.03 -0.28 0.04

Plot 6 76 -2.61 0.08 -1.78 0.05 -1.58 0.07 -0.83 0.03 -0.66 0.04 0.04 0.07 -0.24 0.05 0.07 0.02 -0.21 0.04

Plot 10 42 -2.48 0.09 -1.66 0.05 -1.35 0.06 -0.80 0.03 -0.49 0.04 0.00 0.07 -0.10 0.04 0.04 0.03 -0.07 0.04

Plot 13 38 -2.02 0.06 -1.36 0.05 -1.09 0.06 -0.58 0.03 -0.36 0.04 0.01 0.07 -0.07 0.04 0.10 0.02 -0.02 0.04

Plot 17 29 -2.34 0.06 -1.55 0.05 -1.26 0.06 -0.68 0.03 -0.44 0.04 -0.02 0.07 -0.09 0.04 0.10 0.02 -0.05 0.04

Plot 19 63 -2.58 0.12 -1.74 0.11 -1.46 0.12 -0.82 0.09 -0.58 0.08 0.01 0.07 -0.15 0.04 0.05 0.03 -0.15 0.05

206

Table D-5-3 Mean and standard deviation values of Δ199

Hg, Δ200

Hg and Δ201

Hg (all expressed in

‰) for Hg(II) and Hg(0) used in the binary mixing model with Monte Carlo simulation.

Hg(II) Hg(0)


Δ199

Hg (‰) 0.36 0.28 -0.55 0.07

Δ200

Hg (‰) 0.26 0.25 -0.05 0.05

Δ201

Hg (‰) 0.35 0.26 -0.50 0.07

207




Hg isotopic values.

Δ199

Hg fHg(II) fHg(0)= 1 – fHg(II)



Low WT – Sedge 0.73 0.13 0.27 0.13

Low WT – Ericaceae 0.68 0.13 0.32 0.13

Low WT – Control 0.72 0.13 0.28 0.13

High WT – Sedge 0.82 0.15 0.18 0.15


High WT - Control 0.75 0.14 0.25 0.14

S1 Peat – 2014

Plot 4 0.30 0.07 0.70 0.07

Plot 6 0.38 0.08 0.62 0.08

Plot 10 0.54 0.10 0.46 0.10

Plot 13 0.60 0.11 0.40 0.11

Plot 17 0.57 0.11 0.43 0.11

Plot 19 0.45 0.09 0.55 0.09

208




Hg isotopic values.

Δ200

Hg fHg(II) fHg(0) = 1 – fHg(II)



Low WT – Sedge 0.44 0.39 0.56 0.39

Low WT – Ericaceae 0.41 0.36 0.59 0.36

Low WT – Control 0.53 0.49 0.47 0.49

High WT – Sedge 0.62 0.61 0.38 0.61


High WT - Control 0.62 0.61 0.38 0.61

S1 Peat – 2014

Plot 4 0.41 0.36 0.59 0.36

Plot 6 0.54 0.50 0.46 0.50

Plot 10 0.40 0.35 0.60 0.35

Plot 13 0.71 0.72 0.29 0.72

Plot 17 0.70 0.70 0.30 0.70

Plot 19 0.46 0.42 0.54 0.42

Date post:	14-Apr-2020
Category:	Documents
Upload:	others
View:	4 times
Download:	0 times

Climate Change Impacts on Mercury Cycling in …...ii Climate Change Impacts on Mercury Cycling in...

Documents