Arsenic Mobility in a Changing Northern Climate: Implications for Geochemical Baselines and Long-Term Stability of Contaminants in Lake Systems

ARSENIC MOBILITY IN A CHANGING NORTHERN CLIMATE: IMPLICATIONS FOR GEOCHEMICAL BASELINES AND LONG-TERM STABILITY OF CONTAMINANTS IN LAKE SYSTEMS

Clare Bethune Miller

A thesis submitted to the Department of Geological Sciences & Geological Engineering

In conformity with the requirements for

the degree of Doctor of Philosophy

Queen’s University

Kingston, Ontario, Canada

January 2020

Climate change is influencing the biogeochemical dynamics of lake systems in northern Canada.

These changes may affect the loading and cycling of naturally occurring metal(loid)s and the long-term stability of mining-derived contaminants in sub-Arctic lakes. Arsenic (As) concentrations of lakes in the

Courageous Lake Greenstone Belt (CLGB), Northwest Territories, Canada, are elevated from the weathering of mineralized bedrock and/or the operation of historical gold mines (Tundra and Salmita mines). In this region, the cumulative effects of resource extraction and modern climate warming make it difficult to discern between anthropogenic impacts and baseline geochemistry. This study integrates As geochemistry, organic petrography, paleoclimate proxies (particle size, organic matter (OM) type and quantity), and radiometric dating (14C and 210Pb) to determine the influence of past and present climate warming on the long-term stability of As in lakes surrounding Tundra Mine. The findings of this study demonstrate baseline As concentrations in lake sediments ranging from 28 to 170 mg·kg−1 (median: 40 mg·kg−1; n = 102) and provide evidence that weathering of mineralized bedrock and terrigenous material provides an ongoing source of naturally derived As to some lakes of the CLGB. An increased accumulation of OM in the near-surface sediment, as a result of climate warming, influences redox dynamics and results in As release from minerals to pore waters via reductive dissolution of As-bearing minerals (i.e. scorodite and Fe-(oxy)hydroxides). Under these changing redox conditions, solid phase OM mediates the diffusion of dissolved As to overlying surface waters by providing a substrate for As sequestration and facilitating the precipitation of authigenic As-bearing minerals (i.e. framboidal pyrite,

As-sulphides, Fe-(oxy)hydroxides). However, the effect of these changes will differ between lakes as the long-term stability of As is influenced by the source and primary hosts of As in lake sediments.

Knowledge from this study will help predict future climate-driven changes to metal(loid) cycling in sub-

Arctic lakes and improve environmental monitoring and remediation strategies at northern metal mines.

The possibilities and benefits of weaving traditional knowledge with our research and teaching of geological sciences and engineering are also explored in this thesis, based on hands-on experience gained through this study.

iii

Co-Authorship

The three manuscript chapters of this thesis (Chapters 2, 3, and 4) are intended for journal publication. My co-supervisors, Michael Parsons (Geological Survey of Canada) and Heather Jamieson

(Queen’s University), and Jennifer Galloway (Principle Investigator, Geological Survey of Canada) are co-authors on each of these chapters. These authors provided scientific and editorial guidance in addition to their instrumental roles in project planning, and organization of fieldwork logistics.

Chapter 2 entitled “Lake-specific controls on the long-term stability of mining-related, legacy arsenic contamination and geochemical baselines in a changing northern environment, Tundra Mine,

Northwest Territories, Canada” was submitted to Applied Geochemistry on March 5th, 2019, peer- reviewed, submitted in revised form on June 28th, 2019, and accepted for publication on August 15th,

2019. In addition to the above, the co-authors on this manuscript are Graeme T. Swindles, University of

Leeds, and Nawaf A. Nasser, Carleton University for their contribution and assistance with statistical analysis.

Chapter 3 entitled “Influence of late-Holocene climate change on the solid-phase speciation and long-term stability of arsenic in sub-Arctic lake sediments” was submitted to Science of the Total

Environment on September 3rd, 2019, peer-reviewed and accepted with minor edits on November 7th,

2019. This manuscript will be submitted in revised form on December 20th, 2019. In addition to the above, co-authors include Omid H. Ardakani, Geological Survey of Canada, and Braden R.B. Gregory,

Carleton University. Omid Ardakani provided laboratory space in addition to invaluable guidance in the collection and interpretation of organic petrography data. Guidance in the age-depth modelling of 14C dates was provided by Braden Gregory.

Chapter 4 entitled “Solid phase organic matter control on arsenic mobility in mining-impacted sediment, Tundra Mine, Northwest Territories” will be submitted shortly after submission of this thesis.

In addition to the above, Omid H. Ardakani is also a co-author on this paper for his invaluable assistance with the collection and interpretation of fluorescence microscopy data. iv

As Chapter 5 entitled “Exploring different ways of knowing: Braiding scientific data and

Indigenous knowledge in the geological sciences – A graduate student’s perspective” is written as a critical reflection, it has no official co-authors. However, the foundations for this chapter were provided by the lessons and invaluable mentorship of Timothy Yearington Grey Thunderbird (Métis-

Algonquin Knowledge Keeper, Queen’s University), Jessica Perritt (Nuclear Waste Management

Organization), and Elisabeth and Attima Hadlari (Hadlari Consulting, Nunavut).

Acknowledgements

The knowledge generated in this thesis is greater than the sum of its parts and extends well beyond the pages of this dissertation due to the incredible guidance, mentorship, encouragement, and support I have received over the last 4 years.

First, to my supervisors, Heather Jamieson and Michael Parsons. Thank you both for this wonderful opportunity. Your passion for science, relentless optimism, and unwavering support is truly inspirational. From long days in the field to long nights at the pub, I have learned more from you both than I could have possibly imagined. To the Jamieson research group (Chris, Anezka, Brent, Sean, Kat,

Jon, Alex, Kirsten, et al.), working with you all has been an honour and highlight of my graduate school experience. From brainstorming project ideas and late nights in the lab to office birthday decorations and table hockey rivalries, the guidance and support from each of you has made both this thesis and me, as a scientist, stronger. To my co-authors and the CHARS field team (Nawaf, Brady, Andrew, Hendrik, Mike, and our fearless leader - Jen), your enthusiasm, even when it’s -40 °C and our project seems doomed, is contagious and uplifting. If you are ever going to do winter Arctic field work, don’t go without Hendrik

Falck. Without his knowledge and ingenuity this project truly would have been doomed.

Thank you to the people who have guided both my research and growth along the way. Agatha

Dobosz and Brian Joy, for your patience and problem solving to help me analyze “grunge”. Tony

Lanzirotti, Matt Newville, and Matt Ward, for your guidance and long days at the synchrotron. Omid

Ardakani, for guiding and struggling alongside me as we explored the complex world living in these lakes. Your enthusiasm for a new challenge is truly inspiring. Erin Palmer, I was taught from a young age that librarians have super-powers. Thank you for showing me how true that is. A special thank you to

Attima, Elisabeth, Francis, Tim Yearington, and Jessica Perritt for welcoming me into your world and opening both my mind and my eyes. Tchi-miigwetch, qujanaqquq.

I thank the full afternoons on Stone City patio when research plans went awry. I thank the mountains and weekends in Frontenac Park for preserving my sanity. But most importantly, I thank the vi friends who willingly abandoned reality to climb a mountain, cook a hash, or drink a beer (or two).

Although I am sometimes the red goat of the bunch, you can’t deny I have a word with ways.

It is difficult to articulate how important my family has been throughout this journey. To my co- author and collaborator in life, Jessica Renaud. Thank you for your unwavering support and patience during this crazy time in our lives. Thank you for being big when I was small. Finally, to our in-house editor in chief, Cheri Bethune, MD, MCISc, BSW, BScH Geology, BA Classics and Psychology. I believe it is now appropriate for you to add PhD Environmental Geochemistry to your list of distinctions.

Thank you to both you and Poppy Smurf for your contagious enthusiasm, unwavering support, and endless inspiration. But most importantly, thank you for a lifetime filled with stories, adventures, and wonderful worlds of imagination that truly taught us how to think.

To all of you, thank you for the lessons which I will carry with me always. This process has made me a more inquisitive scientist and researcher, but more importantly these experiences have changed the way I see and perceive our beautiful and dynamic world.

vii

Statement of Originality

As the author of this thesis, I confirm that this dissertation constitutes original findings and all the intellectual content is the product of my own work. All published and unpublished ideas and/or techniques from the contribution of others have been fully acknowledged.

Clare Miller

viii

Table of Contents

Abstract ...... ii Co-Authorship...... iv Acknowledgements ...... vi Statement of Originality ...... viii List of Figures ...... xiv List of Tables ...... xx List of Abbreviations ...... xxi Chapter 1 Introduction ...... 1 1.1 Research motivation and objectives ...... 1 1.2 Background ...... 6 1.2.1 Physiography and site location ...... 6 1.2.2 Geology of the Courageous Lake Greenstone Belt ...... 7 1.2.2.1 Regional geology ...... 7 1.2.2.2 Mineralization ...... 8 1.2.3 Historical mining and remediation activities ...... 11 1.2.4 Effects of climate warming on As mobility ...... 13 1.2.5 Sub-Arctic Holocene climate cycles ...... 16 1.3 Thesis organization ...... 17 1.4 References ...... 19 Chapter 2 Lake-specific controls on the long-term stability of mining-related, legacy arsenic contamination and geochemical baselines in a changing northern environment, Tundra Mine, Northwest Territories, Canada ...... 30 2.1 Abstract ...... 30 2.2 Introduction ...... 31 2.3 Study Location ...... 34 2.3.1 Physiography ...... 34 2.3.2 Geological setting ...... 36 2.4 Methods...... 36 2.4.1 Sampling locations and sample preparation ...... 36 2.4.2 Analyses ...... 38 2.4.2.1 Surface and porewater geochemistry ...... 38 2.4.2.2 Chronology and sedimentation rates ...... 39

2.4.2.3 Sediment geochemistry ...... 39 2.4.2.4 Characterization of As-bearing solid phases ...... 40 2.4.2.5 Sediment textural and organic characteristics ...... 41 2.4.2.6 Statistical analysis of controls on As distribution ...... 42 2.5 Results ...... 42 2.5.1 Surface Water ...... 42 2.5.1.1 Physiochemical characteristics and major dissolved species ...... 42 2.5.1.2 Metal(loid)s and total As speciation ...... 44 2.5.2 Chronology and sedimentation rates ...... 44 2.5.3 Geochemically distinct groupings ...... 44 2.5.4 Sediment geochemistry ...... 48 2.5.5 Porewater geochemistry ...... 49 2.5.6 Characterization of As-bearing solid phases ...... 52 2.5.7 Sediment textural and organic characteristics ...... 56 2.5.8 Statistical analysis of controls on As distribution ...... 56 2.6 Discussion ...... 57 2.6.1 Delineating mining impacts in lake sediment ...... 57 2.6.2 Sources of As in near-surface sediment ...... 59 2.6.3 Intra-lake variance in As mobility and implications of climate warming on the long-term stability of As ...... 64 2.7 Conclusions ...... 69 2.8 Acknowledgements ...... 70 2.9 References ...... 72 Chapter 3 Influence of late-Holocene climate change on the solid-phase speciation and long-term stability of arsenic in sub-Arctic lake sediments ...... 83 3.1 Abstract ...... 83 3.2 Introduction ...... 84 3.3 Study Area ...... 86 3.4 Methods and Materials ...... 87 3.4.1 Sampling locations and sample preparation ...... 87 3.4.2 Age-depth model and stratigraphy ...... 90 3.4.3 Sedimentary grain size ...... 91 3.4.4 Organic matter characterization ...... 91 3.4.5 Sediment and porewater geochemistry ...... 92 x

3.4.5.1 Sediment elemental analysis ...... 92 3.4.5.2 Porewater elemental analysis and dissolved As speciation ...... 93 3.4.6 Solid phase As speciation analysis ...... 93 3.4.6.1 SEM-based Automated Mineralogy and Electron Microprobe analyses ...... 94 3.4.6.2 Bulk X-ray Absorption Near-Edge Structure (XANES) ...... 94 3.4.7 Statistical analyses ...... 95 3.5 Results ...... 96 3.5.1 Core chronology ...... 96 3.5.2 Sedimentary grain size ...... 97 3.5.3 Organic matter characterization ...... 98 3.5.3.1 Sediment organic matter composition ...... 98 3.5.3.2 Petrographic classification of organic matter ...... 101 3.5.4 Sediment and porewater geochemistry ...... 102 3.5.4.1 Elemental concentrations in geochemical groups ...... 102 3.5.4.2 Dissolved As speciation ...... 106 3.5.5 Solid phase As speciation ...... 106 3.5.6 Statistical relationship between organic matter and As ...... 108 3.6 Discussion ...... 110 3.6.1 Climate controls on As speciation and mobility ...... 110 3.6.1.1 Holocene Thermal Maximum (ML-G5; ca. 5430 ± 110 to 4070 ± 130 cal. yrs BP) .... 110 3.6.1.2 Transitional climate (ML-G4; ca. 4070 ± 130 to 2490 ± 140 cal. yrs BP) ...... 116 3.6.1.3 Latest Holocene cooling (CL-G4 & CL-G3; ca. 2110 ± 80 to 400 ± 120 cal. BP) ...... 118 3.6.2 Implications for 21st century warming ...... 118 3.7 Conclusions ...... 119 3.8 Acknowledgements ...... 120 3.9 References ...... 122 Chapter 4 Solid phase organic matter control on arsenic mobility in mining-impacted sediment, Tundra Mine, Northwest Territories, Canada ...... 133 4.1 Introduction ...... 133 4.2 Study area ...... 136 4.2.1 Historical mining and remediation activities ...... 136 4.2.2 Sample locations ...... 137 4.3 Methods and Materials ...... 139 4.3.1 Sample collection and preparation ...... 139 xi

4.3.2 Sediment and porewater geochemistry ...... 140 4.3.3 Sediment organic and textural characterization ...... 140 4.3.4 Solid phase As speciation analysis ...... 141 4.3.4.1 SEM-based Automated Mineralogy and Electron Microprobe analyses ...... 142 4.3.4.2 Synchrotron-based µXRF and XRD ...... 142 4.3.4.3 Bulk X-ray Absorption Near-Edge Structure (XANES) ...... 143 4.4 Results ...... 143 4.4.1 Sediment and porewater geochemistry ...... 143 4.4.2 Sediment organic characterization ...... 145 4.4.3 Solid and aqueous As speciation ...... 146 4.4.3.1 SEM-based Automated Mineralogy and Electron Microprobe analyses ...... 146 4.4.3.2 Synchrotron-based µXRF and XRD ...... 147 4.4.3.3 HG-AFS and bulk XANES ...... 149 4.5 Discussion ...... 151 4.5.1 Sources of As and organic matter to lakes ...... 151 4.5.2 OMAs and As sequestration ...... 153 4.5.3 Arsenic sequestration with TOM ...... 155 4.5.4 Implications for long-term environmental monitoring...... 156 4.6 Limitations and Future Work ...... 157 4.7 Conclusions ...... 157 4.8 Acknowledgements ...... 158 4.9 References ...... 160 Chapter 5 Exploring different ways of knowing: Braiding scientific data and Indigenous knowledge in the geological sciences – A graduate student’s perspective ...... 169 5.1 The Journey ...... 169 5.1.1 Learning on-the-land ...... 169 5.1.2 In the community ...... 170 5.1.3 Back at home ...... 171 5.2 The Lessons and Stories ...... 174 5.2.1 Learning to Listen ...... 174 5.2.2 Language ...... 175 5.2.3 Timescales ...... 176 5.3 Finding Common Ground ...... 177 5.4 Barriers ...... 178 xii

5.5 Moving Forward Together ...... 179 5.5.1 The self ...... 179 5.5.2 Exploring different ways of knowing in the Department of GSGE ...... 180 5.6 References ...... 182 Chapter 6 Conclusions and future research ...... 183 6.1 Delineating mining impacts and geochemical baseline in the CLGB ...... 183 6.2 Long-term stability of legacy contaminants in a changing climate, Tundra Mine, NT ...... 184 6.3 Past sub-Arctic warming as an analogue to inform interpretation of modern-day geochemical trends… ...... 185 6.4 Role of solid phase organic matter on As mobility ...... 188 6.5 Implications for environmental monitoring of legacy and modern mine sites ...... 190 6.6 Combining solid-phase arsenic and organic matter speciation techniques in lake sediment studies...... 192 6.7 Recommendations for future research ...... 193 6.8 References ...... 196 Appendix A Supplementary data to Chapter 2 ...... 199 Appendix B Supplementary data to Chapter 3 ...... 219 Appendix C Supplementary data for Chapter 4 ...... 240 Appendix D Geochemical characterization of Tundra Mine tailings ...... 248 D.1 Introduction ...... 249 D.2 Methods...... 250 D.3 Results and Discussion ...... 251 D.3.1 Field Observations ...... 251 D.3.2 Elemental concentration and distribution of As-bearing phases ...... 251 D.3.3 Tailings organic matter characterization ...... 255 D.4 References ...... 259 Appendix E Field and Laboratory Photos ...... 260

xiii

List of Figures

Figure 1.1 Map showing (A) Location of Slave Geological Province (B) Bedrock geology of the Slave Geological Province and study location with a dashed line denoting boundary between Taiga Shield and Southern Arctic Ecozones (Agriculture and Agri-Food Canada, 2013); (C) Bedrock geology of the Courageous Lake Greenstone Belt and location of Tundra and Salmita mines...... 4 Figure 1.2 Stratigraphic succession of the Courageous-Mackay Lake region (modified from Ransom and Robb, 1986; Bleeker, et al. 1999; Adam, 2016)...... 9 Figure 1.3 Schematic outlining the features and facilities at the former Tundra Mine site (modified from Golder Associates Ltd., 2005 and Lorax Environmental, 2007)...... 13 Figure 1.4 Schematic diagram illustrating the post-depositional behavior of As and associated elements in lakes impacted by the weathering of arsenopyrite-bearing bedrock and deposition of mine wastes (tailings and waste rock); organic matter fractions (S1, S2, S3 as determined by Rock Eval pyrolysis) (modified from Van den Berghe et al., 2018 and Schuh, 2019)...... 16 Figure 2.1 A) Map showing simplified bedrock geology, sediment gravity core sampling locations, predominant mine drainage pathways originating from the former Tundra Mine site, and lake depths at sampling locations (in brackets) (geology modified from Folinsbee and Moore (1955) & Thompson and Kerswill (1994)); B) Bathymetry map of Hambone Lake with numbered locations of sediment grab samples (bathymetry map rendered from 50 water depth measurements)...... 35 Figure 2.2 Sediment grain size and solid-phase concentrations of As, elements enriched in ore and used in mineral processing (Ag, Au, Hg, Pb, Sb, W, Zn) and redox sensitive elements (S, Fe, and Mn) from sediment cores extracted from Powder Mag Lake and Bulldog Lake. Element concentration profiles are plotted on different scales and scales differ between lakes. Shaded intervals separate mining-impacted (MI, I) and background (BG) geochemical populations identified using CONISS (G1- G4), 210Pb dating, and multivariate element profiles...... 46 Figure 2.3 Sediment grain size and solid-phase concentrations of As, elements enriched in ore and used in mineral processing (Ag, Au, Hg, Pb, Sb, W, Zn) and redox sensitive elements (S, Fe, and Mn) from sediment cores extracted from Matthews Lake and Control Lake. Element concentration profiles are plotted on different scales and scales differ between lakes. Shaded intervals separate impacted (MI, I) and background (BG) geochemical populations identified using CONISS (G1- G5), 210Pb dating, and multivariate element profiles...... 47 Figure 2.4 Depth profiles of sediment and dissolved porewater As, Fe, Mn, and S concentrations, porewater As speciation, TOC, and OM fractions S1 and S2 in Powder Mag Lake, Bulldog Lake,

xiv

Matthews Lake, and Control Lake. Concentrations for each element are plotted on different scales and scales are different for each lake...... 51 Figure 2.5 Backscatter electron (BSE) images and mean As content (wt. %), determined by EPMA, of the predominant As-hosting solid phases identified in the near-surface sediment of lakes of the Tundra Mine region. a) Arsenopyrite (Bulldog Lake 0 to 1 cm); b) framboidal pyrite (Powder Mag 4 to 5 cm); c) As- sulphide (Powder Mag 3 to 4 cm); d) FeSx/FeO (Powder Mag 3 to 4 cm); e) Fe-(oxy)hydroxide (Matthews Lake 4 to 5 cm); f) Fe-bearing phyllosilicate with intercalated minerals along cleavage planes (HAM-1). Representative EDS spectra for each phase are presented in Appendix A...... 52 Figure 2.6 Relative contribution of each As-hosting solid phase to total As concentrations in the top 15 cm of sediment samples from Hambone, Powder Mag, Bulldog, Matthews and Control lakes. Sediment grab samples collected from Hambone Lake are plotted as distance from effluent discharge location and not relative to depth in the sediment column...... 55 Figure 2.7 Spearman’s Rank correlations of inorganic and organic geochemical parameters for the near- surface sediment of Bulldog Lake (B-MI) (n = 7), Hambone Lake (HL-MI) (n = 6), Matthews Lake (ML- MI) (n = 13) and Control Lake (CL-I) (n = 14) sediment cores. Organic fractions include TOC, S1, S2, and S3. Data shown as Spearman’s Rank (rs) correlation coefficients. Italicized rs values represent p ≤

0.001, non-italicized values represent p ≤ 0.05; only correlations with rs ≥ 0.6 are included...... 58 Figure 2.8 Schematic diagrams showing the predominant As sources, relative concentration of aqueous As species, and relative contribution of each As-hosting solid phase to total As concentrations in near- surface sediment samples from lakes of the Tundra Mine region. Arrows denote the predicted influence of increased OM on As mobility in each lake. Vertical scales are not meant to be representative of the actual bathymetric features of each lake...... 61 Figure 3.1 A) Map showing location of Tundra Mine (red star) in relation to the City of Yellowknife and other nearby paleolimnological records in the central Northwest Territories (all lake names unofficial): McMaster Lake (MacDonald et al., 1993), Queen’s Lake (Pienitz et al., 1999), Toronto Lake (MacDonald et al., 1993), Waite Lake and Danny’s Lake (Sulphur et al., 2016), and Carleton Lake (Upiter et al., 2014). The transition from boreal forest to tundra is approximated with darker shading (modified after Upiter et al., 2014 and Sulphur et al., 2016). B) Bedrock geology and sampling locations in the CLGB (Folinsbee and Moore, 1955; Thompson and Kerswill, 1994)...... 89 Figure 3.2 Bayesian age-depth model developed using BACON for (A) Matthews Lake and (B) Control Lake. The age-depth model was constructed using the Clam software package and incorporates 210Pb- derived CRS dates and AMS 14C dates (Blaauw, 2010; Reimer et al., 2013); the gray area represents the 95 % confidence interval for the age-depth model. The top leftmost panel shows that Markov Chain Monte Carlo runs were stable, middle plot denotes the prior (curve) and posterior (filled histogram) xv distributions for accumulation rate (yr·cm − 1), and the rightmost plot shows the prior (curve) and posterior (filled histogram) for the dependence of accumulation rate between sections. The main plot shows age distributions of calibrated 14C ages and the age-depth model (grey). Dark grey areas indicate more precisely dated sections of the chronology, while lighter grey areas indicate less chronologically secure sections...... 97 Figure 3.3 Downcore plots of sediment grain size, solid-phase and porewater concentrations of As, S, and Fe; organic matter fractions (S1, S2, S3, and TOC) and dissolved As speciation in the Matthews Lake sediment core. Note changes in scale. Porewater As concentrations are plotted on a log scale. Geochemical populations (ML-G1 to ML-G5) are based on stratigraphically constrained cluster analysis of log-transformed concentrations of 46 elements. Modeled ages incorporate 210Pb-derived CRS dates and AMS 14C dates; dates with a (*) have been calculated from present day (1950 AD) and are reported in AD...... 99 Figure 3.4 Downcore plots of sediment grain size, solid-phase and porewater concentrations of As, S, and Fe; organic matter fractions (S1, S2, S3, and TOC) and dissolved As speciation from the Control Lake sediment core. Note changes in scale. Porewater As concentrations are plotted on a log scale to allow for relative increases to be discerned. Geochemical populations (CL-G1 to CL-G4) are based on stratigraphically constrained cluster analysis of log-transformed concentrations of 46 elements. Modeled ages incorporate 210Pb-derived CRS dates and AMS 14C dates; dates with a (*) have been calculated from present day (1950 AD) and are reported in AD...... 100 Figure 3.5 Fluorescent-light photomicrographs under oil immersion and blue light excitation of organic matter preserved in Matthews Lake and Control Lake sediment. (a) Cuticle of land-plant leaves or stems (Cutinite); (b) Particulate macerals (e.g. sporinite (spore or pollen grain), alginate (planktonic or benthic), chlorophyllinite in diatom frustules, algae (Botryococcus) are dispersed in an AOM matrix; (c) Framboidal pyrite co-occurring with sporinite and AOM matrix; (d) Terrestrial plant epidermis or cell- wall (Subertinite) and red fluorescing chlorophyllinite; (e) Pyritized funginite; (f) Framboidal pyrite co- occurring with alginate; (g) Particulate macerals (e.g. algae (Botryococcus), red fluorescing chlorophyllinite, and alginate within AOM matrix; (h) Particulate macerals (e.g. sporinite (spore or pollen grain), chlorophyllinite in diatom frustules, and algae (Botryococcus) are dispersed in an AOM matrix; (i) Framboidal pyrite co-occurring with sporinite, algae (Botryococcus), and AOM...... 102 Figure 3.6 SEM-BSE images of As-bearing phases in lake sediment cores from Matthews Lake (ML) and Control Lake (CL); Arsenopyrite (ML-A: 44 to 45 cm; CL-A: 28 to 29cm); framboidal pyrite (ML-B: 44 to 45 cm; CL-B: 28 to 29 cm); detrital Fe-oxide (ML-C: 44 to 45 cm; CL-C: 44 to 45 cm); scorodite (ML- E: 26 to 27 cm);authigenic Fe-(oxy)hydroxides (ML-D: 44 to 45 cm; CL-D: 39 to 40 cm). Representative EDS spectra for each phase are presented in Appendix B...... 107 xvi

Figure 3.7 Relative distribution of As species and proportion of As-bearing solid phases with depth in the Matthews Lake and Control Lake sediment cores. Relative distribution was calculated by multiplying As concentration in samples by the fraction of As-species obtained by SEM-based automated mineralogy and LCF of As K-edge XANES data. Results of linear combination fitting of XANES spectra are presented in Appendix B...... 108 Figure 3.8 Summary stratigraphic figure comparing arsenic concentrations (porewater and sediment) and selected organic matter parameters (S2, S3) preserved in the Matthews Lake and Control Lake sediment cores(*) to regional paleoclimate reconstructions using the relative abundance of Picea pollen (1Sulphur et al., 2016), C/N ratios (2Griffith, 2013), diatom-inferred dissolved organic carbon (DOC) (3Pienitz et al., 1999), OC (lost on ignition) (4Upiter et al., 2014), and chironomid-inferred mean July average temperatures (MJAT) (4Upiter et al., 2014). Dashed lines define CONISS-based geochemical groupings and shading represents inferred periods of climate variability in the Courageous Lake Greenstone Belt. Dates with an * are calculated based on median ages between Matthews and Control lakes...... 115 Figure 4.1 Map showing A) Study location and approximate extents of modern day treeline (red line; Brandt, 2009) and B) Bedrock geology and sampling locations (sediment cores – blue; sediment grab samples – yellow) in the Courageous Lake Greenstone Belt (Folinsbee and Moore, 1955; Thompson and Kerswill, 1994). Surface water flow paths from the former Tundra Mine site (INAC, 2005)...... 138 Figure 4.2 Downcore plots of As, Fe, Mn, and S concentrations in sediment and porewater, and solid- phase organic matter composition. The horizontal dotted line represents the onset of mining activities in 1964 (Miller et al., 2019a)...... 144 Figure 4.3 Paired (1) fluorescence-light photomicrographs under oil immersion and (2) SEM-BSE images of OMAs (A, B, E, F) and TOM (C, D) preserved in near-surface sediment of Powder Mag, Bulldog, and Hambone lakes. Organic macerals include suberin (D1; green fluorescing), funginite (C1; green to yellow fluorescing inner layer), chlorophyllinite (A1, B1, E1, F1; pink to red fluorescing), oxidized OM (A1, B1, C1, D1, E1, F1; brown fluorescing), alginate (A1, B1. F1; green fluorescing), and AOM (A1, B1, E1, F1; green to yellow fluorescing)...... 146 Figure 4.4 SEM-BSE images and EDS element maps of (a) OMAs and (b) TOM preserved in the near- surface sediment of Bulldog and Powder Mag lakes...... 147 Figure 4.5 Micro-XRF element maps showing As-Kα (red), Fe-Kα (green), and Sr-Kα (blue) intensities overlain on SEM-BSE images of As-bearing TOM (a, b, c, d) and OMAs (e, f, g, h). Colour intensities are not comparable between maps and are not intended to represent exact concentrations. Cross hairs demonstrate location of EPMA analysis, note that symbol is larger than beam diameter...... 148 Figure 4.6 Micro-XRF element maps showing As-Kα (red), Fe-Kα (green), and Sr-Kα (blue) intensities of As-bearing OMAs (a) and TOM (b,). Colour intensities are not comparable between maps and are not xvii intended to represent exact concentrations. Integration of the µ-XRD patterns shows the presence of discrete grains of pyrite (Py) and orpiment (Orp) associated with OMAs which are comprised of a heterogenous mixture of detrital (not shown) and authigenic minerals (i.e. feroxyhyte (Feroxy)). Authigenic minerals associated with TOM are comprised of goethite (Gt) and mackinawite (Mkw). Cross hairs demonstrate location of µ-XRF analysis, note that symbol is larger than beam diameter...... 149 Figure 4.7 Relative distribution of As species in sediment (bulk XANES) and porewater (HG-AFS) with depth in the (a) Powder Mag Lake and (b) Bulldog Lake sediment cores and (c) with distance from the effluent discharge location in surface sediments of Hambone Lake...... 151

Figure 4.8 Schematic of (a) organo-mineral aggregates (A – authigenic FeSx/FeO grains; B – framboidal pyrite; C – detrital mineral; D – Organic macerals (i.e. sporinite, chlorophylinite, alginite); E – amorphous organic matter (AOM); F – authigenic Fe-(oxy)hydroxides; G – As-sulphides and As bound to AOM; H – oxidized AOM) and (b) Terrigenous organic matter (A – authigenic FeSx/FeO; B – authigenic Fe- (oxy)hydroxides; C – labile cellular structure; D – oxidized cellular structure; E – authigenic sulphides (pyrite, mackinawite))...... 154 Figure 5.1 Photos of research group participating in an on-the-land camp in Nunavut (A to F); A & B) Camp set-up at Elu Inlet, NU; C & D) Daily language lessons; E & F) Learning to cook over a heather fire and clean fish...... 172 Figure 5.2 Photos of research group participating in an on-the-land camp in Nunavut (G to J) and set-up for Department of Geological Sciences and Geological Engineering TEK workshop at Queen’s University (K & L); G) Elder working with a younger Inuk to set a fishing net; H) Rocks that mark hunting and trapping pathways; I) Learning about edible plants on one of our daily walks; J) Participating in Inuit games; K & L) Medicine wheel and tobacco placed at the centre of the workshop circle...... 173 Figure 6.1 Summary stratigraphic figure comparing arsenic concentrations (porewater and sediment) and selected aquatic (A) and terrigenous-derived (B) organic macerals preserved in the Matthews Lake sediment core to inferred periods of climate variability in the Courageous Lake Greenstone Belt based on regional paleoclimate reconstructions (Pienitz et al., 1999, Griffith, 2013, Upiter et al., 2014, Sulphur et al., 2016)...... 187 Figure 6.2 Schematic of the solid phase speciation and post-depositional mobility of As in lake sediment (black) and porewater (blue). Insets display the authigenic formation of (a) organo-mineral aggregates (A

– authigenic FeSx/FeO grains; B – framboidal pyrite; C – detrital mineral; D – Organic macerals (i.e. sporinite, chlorophylinite, alginite); E – amorphous organic matter (AOM); F – authigenic Fe- (oxy)hydroxides; G – As-sulphides and As bound to OM matrix; H – oxidized AOM) and (b) Terrigenous organic matter (A - authigenic FeSx/FeO; B - authigenic Fe-(oxy)hydroxides; C – cellular structure of

xviii

TOM; B – Fe-(oxy)hydroxides; C – oxidized cell wall; D – authigenic sulphides (pyrite, mackinawite)), which facilitate the sequestration of As under changing redox conditions...... 189

xix

List of Tables

Table 1.1 Hypothetical genesis of mineralization in the Courageous-Matthews Lake area (Ransom and Robb, 1986)...... 10 Table 2.1 Selected surface water physicochemical parameters, anions (dissolved, filtered), alkalinity (as

CaCO3), total (unfiltered) concentrations of selected major and trace elements, and inorganic As speciation from surface waters of lakes of the Tundra Mine region...... 43 Table 2.2 Sediment concentrations of selected elements in grab samples collected from Hambone Lake; the highest concentrations are bolded for each element. Sampling locations presented on Figure 2.1...... 49 Table 3.1 Median sediment and porewater (pw) concentrations of selected elements for defined geochemical groupings within Matthews Lake and Control Lake sediment cores (ranges in brackets). . 105 Table 3.2 Spearman’s correlation of As and selected elemental and organic parameters for defined geochemical groupings within Matthews Lake (ML-G3 to ML-G5) and Control Lake (CL-G3 and CL-

G4) sediment cores. Only rs that are significant at the p ≤ 0.05 level are shown...... 110

List of Abbreviations

AANDC Aboriginal Affairs and Northern Development Canada

AMS Accelerator mass spectrometry

AOM Amorphous organic matter

APS Advanced Photon Source

As Arsenic

Au Gold

BG Background

Bi Bismuth

BSE Backscatter electron

C Carbon

CARD Contaminants and Remediation Directorate

CCME Canadian Council of Ministers of the Environment

CIRNAC Crown-Indigenous Relations and Northern Affairs Canada

CLGB Courageous Lake Greenstone Belt

CONISS Constrained incremental sum of squares

CRS Constant rate of supply

Cs Cesium

DL Detection limit

DO Dissolved oxygen

DOC Dissolved organic carbon

EDS Energy-dispersive x-ray spectroscopy

EDTA Ethylenediaminetetraacetic acid

EPMA Electron probe microanalyzer

FAT Felsic ash tuff xxi

Fe Iron

FRE Freshwater reservoir effect

GSGE Geological Sciences and Geological Engineering

GSC Geological Survey of Canada

GSC-A Geological Survey of Canada – Atlantic Division

GSC-C Geological Survey of Canada – Calgary

HC Hydrocarbon

HDPE High-density polyethylene

Hg Mercury

HG-AFS Hydride generation-atomic fluorescence spectrometry

HI Hydrogen index

HTM Holocene Thermal Maximum

I Impacted

ICP-AES Inductively coupled plasma – atomic emission spectroscopy

ICP-MS Inductively coupled plasma – mass spectroscopy

INAC Indigenous and Northern Affairs Canada

ISQG Interim sediment quality guideline

IQ Inuit Quajimajatuqanit

LCF Linear combination fitting

LDL Lower detection limit

MI Mining impacted

MLA Mineral liberation analyser

MJAT Mean July average temperature

Mn Manganese

MPD Mean percent difference

xxii

N2 Nitrogen

NSERC Natural Sciences and Engineering Research Council of Canada

O2 Oxygen

OC Organic carbon

OI Oxygen index

OM Organic matter

OMAs Organo-mineral aggregates

Pb Lead

PC Pyrolysable carbon

PEARL Paleoecological Environmental Assessment Research Laboratory

PEL Probable effects level

RC Residual carbon

RI Redox interface rs Spearman’s correlation ranking

RSD Relative standard deviation

S Sulphur

SD Standard deviation

SEM Scanning electron microscopy

SGP Slave Geological Province

SPL Sparse phase liberation

SWI Sediment-water interface

TCA Tailings confinement area

TK Traditional knowledge

TEK Traditional ecological knowledge

TF Transfer function

xxiii

TOC Total organic carbon

TOM Terrigenous-derived organic matter

μ-XRF Micro x-ray fluorescence

μ-XRD Micro x-ray diffraction

VRS Visible reflectance spectroscopy

WDS Wavelength dispersive spectroscopy

χ2 Chi-squared

XANES X-ray absorption near-edge structure

Zmax Maximum water depth

Zn Zinc

xxiv

Chapter 1

Introduction

1.1 Research motivation and objectives

At northern latitudes, climate warming is influencing the biogeochemistry of lake systems.

Increasing ambient temperatures result in many interrelated changes to sub-Arctic environments including: the lengthening of both aquatic and terrestrial growing seasons, decreased duration of ice-cover on lakes, changes in hydrology and solute transport, degradation of permafrost, enhanced delivery of organic matter to surface water bodies, and accelerated weathering rates in catchments (Smol et al., 2005;

Minyuk et al., 2007; Axford et al., 2009; Rydberg et al., 2010). These changes may alter geochemical baselines in sediments and waters and influence the stability of contaminants related to historical mining activities in Canada’s North. Previous studies have demonstrated that climate-related changes influence the cycling of metals such as mercury (Hg) and lead (Pb) and can alter the baseline concentrations of these elements in the environment (Lindeberg et al., 2006; Minyuk et al., 2007; Outridge et al., 2007,

2011, 2017; Augustsson et al., 2010; Jiang et al., 2011; Cooke et al., 2012). However, the influence of climate change on the mobility of other elements, such as As, is not well established, leading to uncertainty in the determination of geochemical baselines and the long-term monitoring of mining- derived contaminants in mineralized regions.

Geochemical baselines and background values reflect variations in the concentrations of elements in surface water and lake sediments, providing guidance for environmental monitoring and the development of remediation objectives upon mine closure. In the context of this thesis, the differentiation between baseline and background is important. Geochemical baseline includes the cumulative influence of geogenic (or background) variation, local mineralization, anthropogenic impacts, and climate-related influences (Salminen and Gregorauskiene, 2000). Conversely, background or ‘pre-mining background’ is used to refer to concentration ranges in sediment that can be attributed solely to natural processes.

In mineralized regions of Canada’s North, As concentrations in lakes likely reflect a combination of geogenic input from the weathering of bedrock, enriched in As-bearing minerals, and anthropogenic input from both legacy and modern-day mining/remediation operations. In the Slave Craton, Northwest

Territories (NT), sub-Arctic Canada, orogenic gold (Au) deposits are hosted in arsenopyrite (FeAsS)- bearing quartz-carbonate veins occurring in deformed greenstone belts (Goldfarb et al., 1995; Groves et al., 1998). The enrichment of As-bearing sulphides in these mineralized regions may lead to naturally elevated As concentrations, exceeding national environmental quality guidelines, in sediment and surface waters of some lakes (CCME, 2001 a, b). Additionally, a long history of mining and processing of Au has resulted in increased As concentrations in soils, sediments, and surface waters surrounding some former mine sites in the NT (Wagemann et al., 1978; Groves et al., 1998; Bright et al., 1994, 1996; Golder

Associates Ltd., 2005; Andrade et al., 2010; Galloway et al., 2012, 2017; AECOM, 2015; Palmer et al.,

2015, 2019; Houben et al., 2016; Jamieson et al., 2017; Schuh et al., 2018; Van Den Berghe et al., 2018).

The Slave Geological Province (SGP) hosts 52 legacy metal mines (NOAMI, 2017) and remains prospective for future natural resource development (i.e. Seabridge Gold Inc., 2010, 2018; Tetra Tech

Wardrop, 2012; Mirza and Elliot, 2019; Terrax Minerals, 2019); therefore, distinguishing legacy mining impacts from climate-related changes in sediment geochemistry and As mobility will help to guide the sustainable development of Canada’s northern resources.

The historical operation of two high-grade (18.4 to 27.8 g·tonne-1), low-tonnage (285,000 oz Au)

Au-mines in the Courageous Lake Greenstone Belt (CLGB), Tundra Mine (1964 to 1968) and the Salmita

Mine (1983 to 1987), resulted in multi-element contamination of lakes downstream of the former mine sites (Figure 1.1; INAC, 2005; URS, 2005; SENES Consultants, 2006; Lorax Environmental, 2007). The absence of roasting at this site, unlike at the Giant and Con mines in Yellowknife, limited the radial distance of legacy impacts, providing the opportunity to study baseline geochemistry within the mineralized region of a Au-bearing greenstone belt (Jamieson, 2014; Galloway et al., 2015; Palmer et al.,

2015; Houben et al., 2016; Jamieson et al., 2017). The Tundra Mine site provides an ideal location to

study the confounding impacts of legacy Au-mining and current climate warming on metal(loid) mobility in both mining-impacted and natural environments.

Over the late-Holocene (~ ca. 7,000 yrs. BP), the study region experienced profound landscape changes in response to climate warming, including tree line movement of up to 150 km, that affected the physical and chemical properties of lake sediments. Through paleoclimate reconstruction, previous studies have documented biogeochemical changes in lakes proximal to Tundra Mine in response to this period of warming in the sub-Arctic (MacDonald et al., 1993; Pienitz et al., 1999; Upiter et al. 2014;

Sulphur et al., 2016). Therefore, lakes in this region provide records of past climate warming that can be used an analogue to help predict the impact of 21st century climate change on As mobility.

In the Canadian North, Indigenous knowledge systems including Traditional Knowledge (TK),

Traditional Ecological Knowledge (TEK), and Inuit Quajimajatuqanit (IQ) have the potential to identify and fill information gaps provided by the limitations of analytical science (i.e. detection limits, sample size, sampling resolution, cultural and social perspectives). This Indigenous knowledge can help us to understand these dynamic environmental systems and provide a holistic perspective regarding the effects of climate change on people living in northern communities and the ecosystems they rely on (Usher,

2000; Huntington et al., 2004; Alexander et al., 2011; Wolf et al., 2013). The value of integrating western science and Indigenous worldviews has been widely demonstrated by climate scientists; however, this mixed methods approach is still underused by other scientists, especially in studies of geology and resource extraction (Dowsley, 2009; Ford and Furgal, 2009; Gearheard et al., 2010; Green and

Raygorodetsky, 2010; Alexander et al., 2011). Through engagement with the North Slave Métis Alliance, the Yellowknives Dene First Nation, the Tłįchǫ Research and Training Institute, Hadlari Consulting Ltd., an Inuit-owned-and-operated business, Métis-Algonquin Knowledge Keepers at Queen’s University, and industry advisors of Indigenous knowledge, this thesis explores braiding different ways of knowing and meaningful community engagement in the fields of geological sciences, geological engineering, and natural resource development.

The main objective of this study is to combine geochemical speciation methods, organic and inorganic geochemistry, and paleolimnological reconstruction to discern the climate-related factors influencing the mobility of As in northern lake systems. Through addressing the following research objectives, the results of this study aim to improve environmental monitoring and remediation strategies for legacy, modern, and future metal mines in Canada’s North.

• Research objective # 1: Delineate and characterize the spatial extent of impacts related to

historical mining activities at the Tundra Mine, NT, to determine sources of contamination and

the effects of ongoing climate warming on the long-term stability of mining-related

contamination.

• Research objective # 2: Examine the influence of late-Holocene climate change on As loading

and speciation in lake sediments to generate new knowledge that can be used to improve

predictions of the impact of 21st century climate change on As mobility.

• Research objective # 3: Determine the influence of increasing solid phase organic matter (OM)

on redox dynamics and As mobility in mining-impacted sediment.

• Research objective # 4: Explore how TK, TEK, and IQ may enrich our understanding of earth

processes and expand our research and teaching perspectives in the geological sciences.

1.2 Background

1.2.1 Physiography and site location

Tundra and Salmita mines (64.042448 N, 111.169942 W) are located 80 km north of the present-day tree line within the Canadian sub-Arctic. The sub-Arctic region is defined by distinct climate

(i.e. hydrology, precipitation, temperature) characteristics and corresponds with the transition from boreal forest to tundra vegetation. The study site is located between the Taiga Shield, which denotes the northern edge of the modern-day boreal coniferous forest, and the Southern Arctic ecoregions (Figure 1.1;

Atkinson, 1981; Ecosystem Classification Group, 2008, 2012). The physiography of the region is characterized by many small lakes and the typical low relief and slightly rounded topography of the

Canadian tundra (Hatfield Consulting, 1982). Topographic highs are predominantly formed in areas of frost heave or frost broken outcrops (Hatfield Consulting, 1982). Retreat of the Laurentide Ice Sheet, between 10,000 and 9,000 yrs. BP, deposited glacial till and glaciofluvial sediment in topographic lows

(Dyke, 2004; Geological Survey of Canada, 2014). This area is in the transition zone between discontinuous and continuous permafrost; median active-layer thicknesses is seasonally and spatially variable, ranging in thickness from 37 to 150 cm (URS, 2005; Karunaratne, 2011; AECOM, 2015).

Regional climate is characterized by long, cold winters and short, warm summers. Temperatures in the region range from -31 to 10 C with an average annual temperature (1992 to 2011) between -12 and -9 C

(INAC 2005; SENES Consultants, 2006). Data from the nearest long-term weather station in Yellowknife

document a mean January temperature of -26.8 °C and mean July temperature of 16.8 °C from 1971 to

2000 (Environment Canada, 2010). Mean annual precipitation, recorded from 1992 to 2011, is approximately 260 mm and open water season ranges from 94 to 126 days (INAC, 2005; AANDC, 2013).

Lake ice in the region is on average 1.5 to 2 m thick with break-up occurring in early to mid-June and freeze-up beginning in mid- to late October (Golder Associates Ltd., 2017).

The Tundra Mine is approximately 240 km NE of Yellowknife, NT, Canada, south of

Courageous Lake and on the eastern shore of Matthews Lake (Figure 1.1). In the winter months, the site is accessible by a spur road off the Tibbitt to Contwoyto Winter Road. A landing strip near the mine site provides access during the summer. Located within the Treaty 8 Claim, the Akaitcho Territory, the

Wek’eeshii and Monwhi Gogha De Nittaee Areas of the Tłı̨ chǫ Land Claim Agreement, and the North

Slave Metis traditional lands, the Courageous Lake region is used for hunting and harvesting (AANDC,

2012). The Tłı̨ chǫ community of Wekweti, situated approximately 90 km to the west of the site, is the closest permanent community.

1.2.2 Geology of the Courageous Lake Greenstone Belt

1.2.2.1 Regional geology

The mineralized Archean greenstone belts of the Slave Craton have historically been productive mining districts and include the former Giant, Con, and Lupin orogenic gold mines (Geusebroek and

Duke, 2005; Bullen and Robb, 2006; Moir et al., 2006). These deposits are primarily hosted in the

Yellowknife Supergroup of the Central Slave Province (Figure 1.1). The Yellowknife Supergroup is a steeply dipping, north to northwest trending, homoclinal sequence consisting of meta-volcanic and metasedimentary rocks (~2.73 to 2.66 Ga) (McGlynn and Henderson, 1970; Henderson, 1985). The CLGB is one of 20 volcanic belts which make up the Yellowknife Supergroup and is defined by a gradual evolution from volcanism, commencing at 2.73 Ga (Kam Group and Banting Group), to sedimentary accumulation (Burwash Formation (2.67 Ga)) (Figure 1.2; McGlynn and Henderson, 1970; Henderson,

1975; Bleeker and Davis, 1999). In the Courageous Lake region, the CLGB unconformably overlies the

Central Slave Basement Complex (~3.325 Ga) consisting of diorite to tonalite gneiss overlain by thin, 7

discontinuous sequences of volcanic, clastic, and banded iron formations (Bleeker et al., 1999). The

CLGB is flanked to the west by a sodic granite pluton of the Courageous Lake Batholith and to the east by conformably overlying turbidite metasedimentary rocks of the Burwash Formation. Meta-gabbro and masses of felsic porphyry intrude strata of the Yellowknife Supergroup and Proterozoic diabase dykes cross-cut all Archean strata (Dillion-Leitch, 1981; Ransom and Robb, 1986).

1.2.2.2 Mineralization

Multiple Au showings have been discovered and delineated across the CLGB. In the Tundra

Mine region, Au-mineralization is primarily hosted in quartz-carbonate–bearing shear zones that occur at the contact between finely laminated Burwash Formation argillaceous metasediments and Banting Group volcanics, or in quartz stringers filling narrow shear zones, mainly within the metasediments (Ransom and

Robb, 1986). This mineralization has many characteristics compatible with the orogenic gold deposit model which may attributed to regional deformation events commencing at 2.60 to 2.59 Ga (Lambert and

Henderson, 1980; Henderson, 1981). Ransom and Robb (1986) postulate that mineralization of the CLGB is related to felsic intrusions, geothermal hot spots, and late stage silica emplacement occurring during synorogenic volcanism (Table 1.1).

Figure 1.2 Stratigraphic succession of the Courageous-Mackay Lake region (modified from Ransom and Robb, 1986; Bleeker, et al. 1999; Adam, 2016).

Mineralization within the largest auriferous structure (Salmita B-Vein) has been interpreted by

Ransom and Robb (1986) to be a result of episodic silica flooding along hydrothermal conduits formed from east-west trending fault structures and brecciation related to the Salmita Shear Zone (Table 1.1).

Gold also occurs in finely disseminated sulphides within sericitic altered upper felsic tuffaceous units of each volcanic cycle (Ransom and Robb, 1986; Stanley and Sequin, 1989). There are few studies examining the metallogeny of Au deposits in the CLGB; however, Ransom and Robb (1986) and Adam

(2016) suggest that these Au-rich felsic lithologies may be a result of pre-orogenic epithermal

mineralization. Gold mineralization associated with these deposits is predominately refractory in nature, hosted within acicular and rhombic arsenopyrite.

Table 1.1 Hypothetical genesis of mineralization in the Courageous-Matthews Lake area (modified from Ransom and Robb, 1986).

Stage Mineralizing Geologic Event Period Six -- Faulting, diabase intrusion

Five 4 Turbidite sedimentation with penecontemporaneous-to-post deformation. "Last- gasp" silica emplacement into weak, well- capped structures and shear zones (e.g. Matthews-vein) Four 3 b) "sinters" continue and/or new centres develop higher up the pile during or immediately post felsic tuff deposition (terminal volcanism) , e.g . T- Vein, Noranda/Getty a) local basaltic flows fill shallow basins and cap siliceous "sinters Three 2 Quiescence in extrusive activity. “Lagoonal" facies black shale development in shallow basin(s) . "Black Smokers" produce carbon (and sulphide) rich layer. Geothermal hot spots are active producing siliceous "sinter,"* (e.g. B-Vein) Continues into Stage 4 Two -- Development of local calc-alkali centres which produce quartz porphry stocks and release large volumes of rhyodacitic material, contemporaneously with continuous basaltic flows laterally One 1 Outpourings of basalts along a major rift. Minor periods of felsic ejecta and associated structurally controlled sulphide accumulations with weak base metal (Zn, Cu, Ag) mineralization * Siliceous “sinter” - intended to include any process of silicification or silica flooding associated with hydrothermal activity during the terminal stages of volcanism. This process possibly occurred as a semi-continuum in a combination of aerial, subaerial and subaqueous environments.

Since the closure of these mines in 1987, further gold occurrences (FAT deposit and the Walsh

Lake deposit) have been delineated by Adam (2016) and Seabridge Gold Inc. (2010) that differ in both host rock and mineralization style to the Tundra and Salmita deposits. Exploration of these deposits commenced in 2008, pre-feasibility studies for the construction of an open pit mine were released in

2012, and reserves are estimated at 48,963 tonnes (2.18 g·tonne-1; 6.5 million oz Au) (Seabridge Gold

Inc., 2010). Exploration drilling continued in 2018 and identified two new target zones with similar size 10

and grade to Walsh Lake (3.24 g·tonne-1) (Seabridge Gold Inc., 2018). Results from the exploration program concluded that, at 2018 gold prices, economic potential exists for higher-grade, free-milling deposits such as Walsh Lake and the refractory reserves in the FAT deposit (Seabridge Gold Inc., 2018).

1.2.3 Historical mining and remediation activities

Over two phases of mining and milling, approximately 285,000 oz Au were produced from the

Tundra (1964 to 1968) and Salmita (1983 to 1987) mines at grades ranging from 18.4 to 27.8 g·tonne-1

(Ransom and Robb, 1986; Silke, 2011). Ore was milled at the Tundra site at a rate of 125 to 150 tonnes per day with a 95 to 97 % gold recovery achieved through direct cyanidation (Brien, 1964; Lakefield

Research Centre, 1975; Giant Yellowknife Mines Ltd., 1980). In the early days of mining, tailings, waste waters, and associated processing chemicals were disposed sub-aqueously into an adjacent lake (Russel

Lake; Figure 1.1). Dams were constructed around the shoreline of the lake, using low permeability tailings at the core, to create the Tailing Confinement Area (TCA) during later phases of processing.

Subsequent to milling operations, the TCA was divided into Upper and Lower Ponds to allow for flooding and long-term storage (URS, 2005).

Following bankruptcy of the previous owner in 1999, maintenance and remediation of the Tundra

Mine site became the responsibility of the Contaminants and Remediation Directorate (CARD) of

Indigenous and Northern Affairs Canada (INAC). At the initiation of remediation activities in 2007, approximately 155,000 m3 of waste rock, 240,000 m3 of tailings solids, and 1.28 x 106 m3 of impacted water were present on-site (URS, 2005; Golder Associates Ltd., 2008). Remediation activities have included: removal of on-site buildings, dam and landfill repairs, geotechnical inspection of dams, tailings water management, waste rock and tailings kinetic testing, capping of waste rock and tailings, and water quality monitoring (Golder Associates Ltd., 2005, 2008; INAC, 2005, 2010, 2014; URS, 2005; SENES

Consultants, 2006, 2008, 2011; Lorax Environmental, 2007; AECOM, 2015). Weathering of exposed waste rock used in construction of mine facilities, in addition to seasonal overtopping and seepage of

TCA waters, have resulted in the contamination of downstream lakes (INAC, 2014). Waste rock was also

used to infill the northern-most shoreline of Bulldog Lake to maintain water levels (Department of

Fisheries and Oceans, 2004).

Rising water levels in the TCA warranted the implementation of an emergency water treatment program in 2009 (INAC, 2010). A permanent water treatment plant was constructed at the Tundra Mine site in 2014 to reduce concentrations of As, Zn, and Pb in tailings waters prior to disposal (AECOM,

2015). In this treatment process, ferric sulphate and lime were added to the overlying waters to precipitate metal hydroxides and remove As through co-precipitation and sorption processes. Elevated concentrations of Zn and Pb in tailings effluent were treated with sodium metabisulfite (WESA, 2013). The precipitate and other suspended solids were separated from the effluent water through dewatering in Geotubes© before testing and subsequent discharge of effluent to Hambone Lake. Due to the accumulation of seasonal melt water and precipitation in the TCA, an average of ~ 254,500 m3 of treated effluent were released to the downstream environment on an annual basis from 2009 to 2018, during active remediation activities (INAC, 2005, 2008, 2010; AANDC, 2012, 2013, 2014). In the final phases of remediation

(2017 to 2018) tailings were placed in a lined pit, capped and covered. The site is presently in an adaptive management and long-term monitoring phase following the completion of remediation activities in

August 2018 (AECOM, 2018).

Figure 1.3 Schematic outlining the features and facilities at the former Tundra Mine site (modified from Golder Associates Ltd., 2005 and Lorax Environmental, 2007).

1.2.4 Effects of climate warming on As mobility

The implications of climate warming on As mobility in lakes are described in detail in subsequent chapters of this thesis and are therefore only described briefly in this section.

Lake sediment may provide a sink or source of As to surface waters in both mining-impacted and unimpacted regions (Martin and Pedersen, 2002; Toevs et al., 2006; Andrade et al., 2010; Schuh et al.,

2018). In the sub-Arctic, lakes are particularly vulnerable to climate change (Schindler and Smol, 2006;

Griffiths et al., 2017; Woelders et al., 2018). Increases in temperature and duration of ice-free seasons, as a result of climate warming, can cause whole-scale changes to organic carbon cycles and influence the physical and chemical processes within lake systems (Frey and McClelland, 2009; Prowse et al., 2011;

Stern et al., 2012). It has been demonstrated that As speciation and mobility in lakes can be altered by limnologic changes, seasonal fluctuations, and increases in organic matter; however, the influence of regional climate change on the long-term stability of As-bearing minerals in lake environments is not well established (Martin and Pedersen, 2002; Palmer et al., 2019).

At high latitudes, aquatic and terrigenous biological production is closely related to air temperature and the duration of the ice-free seasons (Douglas and Smol, 1999; Jankovská and Komárek,

2000; Griffiths et al., 2017). Permafrost thawing, as a result of increasing ambient temperatures, may also contribute a source of organic carbon into sub-Arctic lakes (Guo et al., 2007; McGuire et al., 2009). In lacustrine systems, total organic carbon (TOC) concentrations affect dissolved oxygen levels and influence sediment redox dynamics which, in turn, influence the speciation, mobility, and cycling of metal(loid)s (Pienitz et al., 1999; Martin and Pedersen, 2002; Couture and Van Cappellen, 2011). In sediment, As mobility is highly susceptible to the changing redox conditions and concentrations of OM associated with seasonal fluctuations (Aggett and O’Brien, 1985; La Force et al., 2000; Martin and

Pedersen, 2002; Redman, 2002; Smedley and Kinniburgh, 2002; Keimowitz et al., 2005; Toevs et al.,

2006; Kirk et al., 2010; Wang and Jiao, 2014; Wang et al., 2018). It is therefore postulated that long-term, climate-driven biogeochemical and physicochemical changes will influence the release, transport, and fate of redox-sensitive metal(loid)s, such as As, in sub-Arctic lake systems.

The mobility of As in environmental media (i.e. soils, surface waters, sediment) is dependent on mineralogy, aqueous speciation, and biological processes. In lacustrine environments, post-depositional mobility of As is largely controlled by redox conditions and closely related to the cycling of iron (Fe), sulfur (S) and OM (Figure 1.4; Harvey et al., 2002; O’Day et al., 2004; Wilkin and Ford, 2006; Couture et al., 2010). Under oxidizing conditions, such as in the near-surface sediment, the precipitation of authigenic Fe/Mn- (oxy)hydroxides sequesters dissolved As through sorption and co-precipitation reactions (Dzombak and Morel, 1990; Waychunas et al., 1993; Dixit and Hering, 2003). As oxygen concentrations decrease during burial in the sediment, the reductive dissolution of Fe- and Mn-

(oxy)hydroxides releases As to sediment porewater. This As then diffuses upward where it may re- 14

precipitate at the sediment-water interface (SWI) through reactions with authigenic metal oxyhydroxides or may diffuse into overlying surface waters (Figure 1.4; Boudreau, 1999; Martin and Pedersen, 2002;

Campbell et al., 2008). Over time, this redox-driven cycling may result in substantial surface sediment enrichment of As. In anoxic environments, such as reducing sediments and those with high concentrations of OM, dissolved As may also be removed from solution through the precipitation of authigenic sulphides

(i.e. mackinawite (FeS), pyrite (FeS2), orpiment (As2S3) and realgar (AsS) (Figure 1.4; Bostick and

Fendorf, 2003; O'Day et al., 2004). Under these conditions, arsenite (As (III) as H3AsO3) is the predominant dissolved As species; however, thioarsenic species may also play an important role in As mobility (Planer-Friedrich et al., 2007). In general, the release of dissolved As into the water column depends on sediment redox conditions, which are, in turn, governed by the accumulation rate of OM. As

OM in sediment increases as a result of climate warming, understanding its role in As mobilization and transport in surface and subsurface environments is of utmost importance.

Organic compounds are redox reactive and play an important role in driving redox transformations in sediment; however, dissolved OM (e.g. OM < 0.22 to 0.45 μm) may also influence As mobility through competitive adsorption on mineral surfaces (Grafe et al., 2001; Redman et al., 2002), formation of Fe-OM aqueous complexes (Langner et al., 2012, 2014) and microbially-mediated precipitation of As-bearing minerals (Kirk et al., 2004). These interactions have been well documented in soils (Kalbitz and Wennrich, 1998; Bauer and Blodau, 2006; Dobran and Zagury, 2006), and sediments

(i.e. stream, wetland, and deltaic) (La Force et al., 2000; Langner et al., 2012, 2014; Wang et al., 2018).

However, the role of solid phase OM (OM > 0.22 to 0.45 μm) on element mobility in sediment is largely unknown (Sanei and Goodarzi, 2006; Langner et al., 2012). In this study, stepwise pyrolysis (Rock-Eval®

6), visible reflectance spectroscopy, and fluorescent light microscopy were used to determine the amount and type of OM in bulk lake sediments. In recent sediments, the S1 fraction of organic matter (determined by pyrolysis) is comprised of aquatic-derived OM (e.g. algal-derived lipids; amino acids, chlorophyll, and small volatile molecules) and S2 compounds (determined by pyrolysis) are derived from the biomacromolecule structure of algal cell walls and other aquatic biological matter, such as phytoplankton 15

and copepods (Sanei et al., 2005; Carrie et al., 2012). Terrigenous plant materials (i.e. conifer needles, roots, bark), in addition to humic and fulvic acids, comprise the S3 fraction (determined by pyrolysis;

Carrie et al., 2012; Albrecht et al., 2015). These sources of OM to lakes and their roles in As mobility are highlighted in Figure 1.4.

Figure 1.4 Schematic diagram illustrating the post-depositional behavior of As and associated elements in lakes impacted by the weathering of arsenopyrite-bearing bedrock and deposition of mine wastes (tailings and waste rock); organic matter fractions (S1, S2, S3 as determined by Rock Eval pyrolysis) (modified from Van den Berghe et al., 2018 and Schuh, 2019).

1.2.5 Sub-Arctic Holocene climate cycles

This section provides an overview of paleolimnological reconstructions and their use as analogues to improve predictions of the impact of 21st century climate change. Specifics regarding paleolimnological reconstructions of Holocene climate trends in the central NT of sub-Arctic Canada are discussed in detail in Chapter 3.

Lake sediments preserve archives of past ecological changes and are widely used to reconstruct biogeochemical responses to regional climate trends, especially in remote areas where long-term

environmental monitoring data are limited. Documenting historical changes in climate and their influence on ecology and biogeochemistry through the analyses of fossil pollen, macrofossils, solid phase geochemistry, sediment grain size, and both stable and radioactive isotopes in lake sediments has increasingly become a focal point of environmental research in Arctic and sub-Arctic regions (Smol and

Douglas, 2007; Burge et al., 2018). During the first half of the present interglacial period, ambient temperatures rose in the central NT at approximately 7,000 cal. yr BP and marked the onset of the

Holocene Thermal Maximum (HTM) (MacDonald, 1983, 1987; Ritchie, 1984; Slater, 1985; Moser and

MacDonald, 1990; MacDonald et al., 1993; Szeicz and MacDonald, 1995, 2001; Pienitz et al., 1999;

Huang et al., 2004; Upiter et al., 2014). Ambient air temperatures increased by 1 to 2 °C over the HTM, resulting in increased aquatic production and a shift in terrestrial vegetation from tundra to forest-tundra.

Studies examining the influence of historical climate cycles on the distribution of metal(loid)s in pre- industrial sediment of the sub-Arctic demonstrate that Hg, Pb, redox sensitive elements (Fe/Mn), and other trace metal(loid)s are affected by climate-related factors (i.e. primary production, sedimentation rates, influxes of allochthonous mineral matter) (Lindeberg et al., 2006; Augustsson et al., 2010; Cooke et al., 2012; Outridge et al., 2017). However, to our knowledge, no studies have examined the effects of past warming cycles on the speciation and mobility of As in sub-Arctic lacustrine systems.

1.3 Thesis organization

Including the Introduction, this thesis comprises six chapters. Chapters 2, 3, and 4 are written in manuscript format. Chapter 2 is published in Applied Geochemistry (Miller et al., 2019) and describes the influence of modern-day climate warming on the long-term stability of legacy mining contamination in lakes downstream of the former Tundra Mine, NT. Chapter 3 was accepted for publication in Science of the Total Environment on November 7th, 2019 and will be submitted following minor revisions on

December 20th, 2019. This chapter examines the influence of late-Holocene climate change on the loading and mobility of As in lake sediments as a possible analogue for current and projected warming in the central NT. Chapter 4, which is intended for publication in Chemical Geology, focuses on the role of solid

phase organic matter on the mobility of As under changing redox conditions. Chapter 5 is a critical reflection of my experiences learning about different knowledge systems which aims to inform future action and considers the implications of these experiences on my work as an academic and geoscientist.

Chapter 6 provides a summary of the main conclusions of this study and recommendations for future research.

1.4 References

1AANDC, 2012. Water Surveillance Network Program (SNP) Report for Type ‘A’ Water Licence MV2009L8-0008 (Tundra Mine Remediation) June 2012 Water Quality Results. http://registry.mvlwb.ca/Documents/MV2009L8-0008/MV2009L8-0008%20-%20AANDC- CARD%20-%20Tundra%20Mine%20Remediation%20- %20June%202012%20SNP%20Report%20-%20Jul30-12.pdf

AANDC, 2013. Water Balance Monitoring at the Tundra Mine Site: 2012 Update. http://registry.mvlwb.ca/Documents/MV2009L8-0008/MV2009L8-0008%20- %20AANDC%20%20CARD%2020Tundra%20Minesite%202012%20WL%20Annual%20Report% 20-%20Mar28-13.pdf

AANDC, 2014. Tundra Mine, NWT Remediation Environmental Monitoring Program: Water Quality Part L: 2013 and Historic Results. http://registry.mvlwb.ca/Documents/MV2009L8- 0008/MV2009L8-0008%20-%20AANDC-CARD%20- %202013%20Annual%20Report%20for%20Tundra%20Mine%20Site%20-%20Mar27-14.pdf

Adam, M.A., 2016. Gold Mineralization at the FAT Deposit , Courageous Lake Greenstone Belt, Northwest Territories, Canada (Master’s Thesis). University of Western Ontario. https://ir.lib.uwo.ca/etd/3640/

AECOM, 2015. Tundra Mine, NWT Remediation Environmental Monitoring Program: Water Quality Part L: 2014 and Historical Results. http://registry.mvlwb.ca/Documents/MV2009L8- 0008/MV2009L8-0008%20-%20AANDC-CARD%20- %20Tundra%20Mine%202014%20Annual%20WL%20Report%20-%20May5-15.pdf

AECOM, 2018. Tundra Mine Adaptive Management Plan. AECOM http://registry.mvlwb.ca/Documents/MV2016L8-0003/MV2016L8-0003%20-%20INAC- CARD%20-%20Tundra%20-%20Adaptive%20Management%20Plan%20-%20Mar8-18.pdf

Aggett, J., O’Brien, G.A., 1985. Detailed model for the mobility of arsenic in lacustrine sediments based on measurements in Lake Ohakuri. Environ. Sci. Technol. 19, 231–238. https://doi.org/10.1021/es00133a002

Agriculture and Agri-Food Canada, 2013. Terrestrial Ecoregions of Canada. https://open.canada.ca/data/en/dataset/ade80d26-61f5-439e-8966-73b352811fe6

Albrecht, R., Sebag, D., Verrecchia, E., 2015. Organic matter decomposition: bridging the gap between Rock–Eval pyrolysis and chemical characterization (CPMAS13C NMR). Biogeochemistry 122, 101–111. https://doi.org/10.1007/s10533-014-0033-8

Alexander, C., Bynum, N., Johnson, E., King, U., Mustonen, T., Neofotis, P., Oettlé, N., Rosenzweig, C., Sakakibara, C., Shadrin, V., Vicarelli, M., Waterhouse, J., Weeks, B., 2011. Linking Indigenous and Scientific Knowledge of Climate Change. Bioscience 61, 477–484. https://doi.org/10.1525/bio.2011.61.6.10

1 INAC, AANDC, CIRNAC are all acronyms for the Federal Department responsible for the Government of Canada's work in northern Canada. 19

Atkinson, K., 1981. Vegetation zonation in the Canadian subarctic. R. Geogr. Soc. 13, 13–17. https://doi.org/https://www.jstor.org/stable/20001647

Augustsson, A., Peltola, P., Bergbäck, B., Saarinen, T., Haltia-Hovi, E., 2010. Trace metal and geochemical variability during 5,500 years in the sediment of Lake Lehmilampi, Finland. J. Paleolimnol. 44, 1025–1038. https://doi.org/10.1007/s10933-010-9471-z

Axford, Y., Briner, J.P., Cooke, C.A., Francis, D.R., Michelutti, N., Miller, G.H., Smol, J.P., Thomas, E.K., Wilson, C.R., Wolfe, A.P., 2009. Recent changes in a remote Arctic lake are unique within the past 200,000 years. Proc. Natl. Acad. Sci. 106, 18443–18446. https://doi.org/10.1073/pnas.0907094106

Bauer, M., Blodau, C., 2006. Mobilization of arsenic by dissolved organic matter from iron oxides, soils and sediments. Sci. Total Environ. 354, 179–190. https://doi.org/10.1016/j.scitotenv.2005.01.027

Bleeker, W., Davis, W., 1999. The 1991-1996 NATMAP Slave Province Project: Introduction. Can. J. Earth Sci. 36, 1033–1042. https://doi.org/10.1139/cjes-36-7-1033

Bleeker, W., Ketchum, J.W.F., Jackson, V.A., Villeneuve, M.E., 1999. The Central Slave Basement Complex, Part I: its structural topology and autochthonous cover. Can. J. Earth Sci. 36, 1083–1109. https://doi.org/10.1139/e98-102

Bostick, B.C., Fendorf, S., 2003. Arsenite sorption on troilite (FeS) and pyrite (FeS2). Geochim. Cosmochim. Acta 67, 909–921. https://doi.org/10.1016/S0016-7037(02)01170-5

Boudreau, B.P., 1999. Metals and models: Diagenetic modelling in freshwater lacustrine sediments. J. Paleolimnol. 22, 227–251. https://doi.org/10.1023/A:1008144029134

Brien, F.B., 1964. Tundra Gold Mines - Milling Operations. West. Min. 34–38.

Bright, D.A., Coedy, B., Dushenko, W.T., Reimer, K.J., 1994. Arsenic transport in a watershed receiving gold mine effluent near Yellowknife, Northwest Territories, Canada. Sci. Total Environ. 155, 237– 252. https://doi.org/10.1016/0048-9697(94)90503-7

Bright, D.A., Dodd, M., Reimer, K.J., 1996. Arsenic in subarctic lakes influenced by gold mine effluent: The occurrence of organoarsenicals and ‘hidden’ arsenic. Sci. Total Environ. 180, 165–182. https://doi.org/https://doi.org/10.1016/0048-9697(95)04940-1

Bullen, W., Robb, M., 2006. Economic contribution of gold mining in the Yellowknife mining district; Chapter 4 in Gold in the Yellowknife Greenstone Belt, Northwest Territories: Results of the EXTECH III Multidisciplinary Research Project, (ed.) C.D. Anglin, H. Falck, D.F. Wright, and E.J. Ambrose; Geological Association of Canada, Mineral Deposits Division, Special Publication No. 3, pp. 38–49.

Burge, D.R.L., Edlund, M.B., Frisch, D., 2018. Paleolimnology and resurrection ecology: The future of reconstructing the past. Evol. Appl. 11, 42–59. https://doi.org/10.1111/eva.12556

Campbell, K.M., Root, R., O’Day, P.A., Hering, J.G., 2008. A gel probe equilibrium sampler for measuring arsenic porewater profiles and sorption gradients in sediments: II. Field application to Haiwee reservoir sediment. Environ. Sci. Technol. 42, 504–510. https://doi.org/10.1021/es071120a

Carrie, J., Sanei, H., Stern, G., 2012. Standardisation of Rock-Eval pyrolysis for the analysis of recent sediments and soils. Org. Geochem. 46, 38–53. https://doi.org/10.1016/j.orggeochem.2012.01.011

CCME, 2001a. Canadian Sediment Quality Guidelines for the Protection of Aquatic Life, Environmental Protection, CCME. https://ccme.ca/en/resources/canadian_environmental_quality_guidelines/

CCME, 2001b. CCME Water Quality Guidelines for the Protection of Aquatic Life, CCME. https://ccme.ca/en/resources/canadian_environmental_quality_guidelines/

Cooke, C.A., Wolfe, A.P., Michelutti, N., Balcom, P.H., Briner, J.P., 2012. A Holocene perspective on algal mercury scavenging to sediments of an Arctic lake. Environ. Sci. Technol. 46, 7135–7141. https://doi.org/10.1021/es3003124

Couture, R.M., Gobeil, C., Tessier, A., 2010. Arsenic, iron and sulfur co-diagenesis in lake sediments. Geochim. Cosmochim. Acta 74, 1238–1255. https://doi.org/10.1016/j.gca.2009.11.028

Couture, R.M., Van Cappellen, P., 2011. Reassessing the role of sulfur geochemistry on arsenic speciation in reducing environments. J. Hazard. Mater. 189, 647–652. https://doi.org/10.1016/j.jhazmat.2011.02.029

Department of Fisheries and Oceans, 2004. Matthews Lake and Area Fish Habitat Restoration Program: 2003 Site Baseline Assessment.

Dillion-Leitch, H.C., 1981. Volcanic stratigraphy, structure and metamorphism in the Courageous- Mackay Lake Greenstone Belt, Slave Province, Northwest Territories. University of Ottawa. https://doi.org/10.20381/ruor-13514

Dixit, S., Hering, J.G., 2003. Comparison of arsenic (V) and arsenic (III) sorption onto iron oxide minerals: Implications for arsenic mobility. Environ. Sci. Technol. 37, 4182–4189. https://doi.org/10.1021/es030309t

Dobran, S., Zagury, G.J., 2006. Arsenic speciation and mobilization in CCA-contaminated soils: influence of organic matter content. Sci. Total Environ. 364, 1–3.

Douglas, M. S. V., Smol, J. P. 1999 in The Diatoms: Applications for the Environmental and Earth Sciences, eds. Stoermer, E. F., Smol J. P. Cambridge Univ. Press, Cambridge, U.K. pp. 227–244.

Dowsley, M., 2009. Community clusters in wildlife and environmental management: Using TEK and community involvement to improve co-management in an era of rapid environmental change. Polar Res. 28, 43–59. https://doi.org/10.1111/j.1751-8369.2008.00093.x

Dyke, A.S., 2004. An outline of North American deglaciation with emphasis on central and northern Canada. Dev. Quat. Sci. 2, 373–424. https://doi.org/10.1016/S1571-0866(04)80209-4

Dzombak, D., Morel, F.M.M., 1990. Surface complexation modelling- hydrous ferric oxide. John Wiley, New York.

Ecosystem Classification Group. 2008. Ecological Regions of the Northwest Territories - Taiga Shield. Department of Environment and Natural Resources, Government of the Northwest Territories, 21

Yellowknife, NT. viii + 146 pp. + insert map.

Ecosystem Classification Group. 2012. Ecological Regions of the Northwest Territories - Southern Arctic. Department of Environment and Natural Resources, Government of the Northwest Territories, Yellowknife, NT. x + 170 pp. + insert map

Environment Canada, 2010. National Climate Data and Information Archive, Canadian Climate Normals or Averages 1971-2000, Yellowknife A, Northwest Territories; Environment Canada. http://www.climate.weatheroffice.gc.ca

Ford, J.D., Furgal, C., 2009. Foreword to the special issue: Climate change impacts, adaptation and vulnerability in the Arctic. Polar Res. 28, 1–9. https://doi.org/10.1111/j.1751-8369.2009.00103.x

Frey, K.E., McClelland, J.W., 2009. Impacts of permafrost degradation on arctic river biochemistry. J. Glaciol. 23, 169–182. https://doi.org/10.1002/hyp

Galloway, J.M., Sanei, H., Patterson, R.T., Mosstajiri, T., Hadlari, T., Falck, H., 2012. Total arsenic concentrations of lake sediments near the City of Yellowknife, Northwest Territories. Geological Survey of Canada Open File 7037, 1–47.

Galloway, J M; Palmer, M; Jamieson, H E; Patterson, R T; Nasser, N; Falck, H; Macumber, A L; Goldsmith, S A; Sanei, H; Normandeau, P; Hadlari, T; Roe, H M; Neville, L A; Lemay, D., 2015. Geochemistry of lakes across ecozones in the Northwest Territories and implications for the distribution of arsenic in the Yellowknife region. Part 1: Sediments. Geological Survey of Canada, Open File 7908, 1-49, https://doi.org/10.4095/296954

Galloway, J.M., Swindles, G.T., Jamieson, H.E., Palmer, M.J., Parsons, M.B., Sanei, H., Macumber, A.L., Patterson, R.T., Falck, H., 2017. Organic matter control on the distribution of arsenic in lake sediments impacted by ~ 65 years of gold ore processing in subarctic Canada. Sci. Total Environ. 622–623, 1668–1679. https://doi.org/10.1016/j.scitotenv.2017.10.048

Gearheard, S., Pocernich, M., Stewart, R., Sanguya, J., Huntington, H.P., 2010. Linking Inuit knowledge and meteorological station observations to understand changing wind patterns at Clyde River, Nunavut. Clim. Change 100, 267–294. https://doi.org/10.1007/s10584-009-9587-1

Geological Survey of Canada, 2014. Surfical Geology: Lac de Gras, Northwest Territories, NTS 76-D; Canadian Geosciences Map 183 (2nd Addition, preliminary, Surfical Data Model v. 20 conversion of Map 1870A) scale 1:125 000. https://doi.org/10.4095/295497

Geusebroek, P.A. & Duke, N.A., 2005. An Update on the Geology of the Lupin Gold Mine, Nunavut, Canada., Exploration Mining Geology, 13, pp.1–13.

Giant Yellowknife Mines Ltd., 1980. Salmita Project - A Brief Project Overview, 1–17. Obtained from NTGO Library.

Golder Associates Ltd., 2005. Terrestrial and Aquatic Field Surveys and Ecological Risk Assessment of the Tundra Gold Mine. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

Golder Associates Ltd., 2008. Phase II Remedial Action Plan: Tundra Mine Site. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

Golder Associates Ltd., 2017. Public Works and Government Services Canada Tundra Mine, NWT Remediation Environmental Monitoring Program: 2016 and Historical Water Quality Results.

Goldfarb, R.J., Berger, B.R., Klein, T.L., Pickthorn, W.J., Klein, D.P., 1995. Low-Sulfide Au Quartz Veins. In: E.A. du Bray (Ed.), Preliminary Compilation of Descriptive Geoenvironmental Mineral Deposit Models. U.S. Geol. Surv. Open-File Report 95–831, Denver CO, pp. 261–267.

Grafe, M., Eick, M.J., Grossl, P.R., 2001. Adsorption of arsenate (V) and arsenite (III) on goethite in the presence and absence of dissolved organic carbon. Soil Sci. Soc. Am. J. 65, 1680. https://doi.org/10.2136/sssaj2001.1680

Green, D., Raygorodetsky, G., 2010. Indigenous knowledge of a changing climate. Clim. Change 100, 239–242. https://doi.org/10.1007/s10584-010-9804-y

Griffiths, K., Michelutti, N., Sugar, M., Douglas, M.S.V., Smol, J.P., 2017. Ice-cover is the principal driver of ecological change in High Arctic lakes and ponds. PLoS One 12. https://doi.org/10.1371/journal.pone.0172989

Groves, D.I., Goldfarb, R.J., Gebre-Mariam, M., Hagemann, S.G., Robert, F., 1998. Orogenic gold deposits—a proposed classification in the context of their crustal distribution and relationship to other gold deposit types. Ore Geology Reviews 13, 7–27. https://doi.org/10.1016/S0169- 1368(97)00012-7

Guo, L., Ping, C.L., Macdonald, R.W., 2007. Mobilization pathways of organic carbon from permafrost to arctic rivers in a changing climate. Geophys. Res. Lett. 34, 1–5. https://doi.org/10.1029/2007GL030689

Harvey, C.F., Swartz, C.H., Badruzzaman, A.B.M., Keon-Blute, N., Yu, W., Ashraf Ali, M., Jay, J., Beckie, R., Niedan, V., Brabander, D., Oates, P.M., Ashfaque, K.N., Islam, S., Hemond, H.F., Feroze Ahmed, M., 2002. Arsenic Mobility and Groundwater Extraction in Bangladesh. Science. 298, 1602-06. https://doi.org/10.1126/science.1081789

Hatfield Consulting, 1982. Initial Environmental Evaluation for Giant Yellowknife Mines - Salmita Project. Unpublished report prepared for Giant Yellowknife Mines Ltd., Yellowknife, NT. pp. 1–24.

Henderson, J.B., 1975. Sedimentology of the Archean Yellowknife Supergroup at Yellowknife, District of Mackenzie; Bulletin 246. https://doi.org/https://doi.org/10.4095/103977

Henderson, J.B., 1981. Archean basin evolution in the Slave Province, Canada. Developments in Precambrian Geology, 4, 213–235. https://doi.org/10.1016/S0166-2635(08)70014-8

Henderson, J.B., 1985. Geology of the Yellowknife-Hearn Lake Area, District of Mackenzie: A Segment Across an Archean Basin (Geological Survey of Canada Memoir 414).

Houben, A.J., D'Onofrio, R., Kokelj, S.V., Blais, J.M., 2016. Factors affecting elevated arsenic and methyl mercury concentrations in small shield lakes surrounding gold mines near the Yellowknife, NT, (Canada) Region. PLoS One 11, e0150960. https://doi.org/ 10.1371/journal.pone.0150960.

Huang, C.C., MacDonald, G., Cwynar, L., 2004. Holocene landscape development and climatic change in the low arctic, Northwest Territories, Canada. Palaeogeogr. Palaeoclimatol. Palaeoecol. 205, 221– 234. http://dx.doi.org/10.1016/j.palaeo.2003.12.009

Huntington, H., Callaghan, T., Fox, S., Krupnik, I., 2004. Matching traditional and scientific and scientific observations to detect environmental change : A discussion on Arctic terrestrial ecosystems. Ambio Special Re. 18–23. https://doi.org/10.2307/25094583

INAC, 2005. Tundra Mine, NWT Environmental Monitoring Program: Water Quality Part D: 2005 Results. Prepared for: Contaminants and Remediation Directorate, Indian and Northern Affairs Canada. http://registry.mvlwb.ca/Documents/MV2005X0031/PartD-WaterQualityMon-Nov05.pdf

INAC, 2008. Tundra Mine, NWT Environmental Monitoring Program: Water Quality Part F: 2007 Results. Contaminants and Remediation Directorate, Indian and Northern Affairs Canada. http://registry.mvlwb.ca/Documents/MV2005L8-0014/2007- Part F - Water Quality Monitoring Program (Mar_08).pdf

INAC, 2010. Remediation Environmental Monitoring Program: Water Quality Part H: 2009 and Historic Results. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

INAC, 2014. Tundra Mine, NWT Remediation Environmental Monitoring Program: Water Quality Part L: 2013 and Historic Results.

Jamieson, H.E., 2014. The legacy of arsenic contamination from mining and processing refractory gold ore at Giant Mine, Yellowknife, Northwest Territories, Canada. Rev. Mineral. Geochemistry 79, 533–551. https://doi.org/10.2138/rmg.2014.79.12

Jamieson, H.E., Maitland, K.M., Oliver, J.T., Palmer, M.J., 2017. Regional distribution of arsenic in near- surface soils in the Yellowknife area. NWT Open File 2017-03.

Jankovská, V., Komárek, J., 2000. Indicative value of Pediastrum and other coccal green algae in palaeoecology. Folia Geobot. 35, 59–82. https://doi.org/10.1007/BF02803087

Jiang, S., Liu, X., Chen, Q., 2011. Distribution of total mercury and methylmercury in lake sediments in Arctic Ny-Ålesund. Chemosphere 83, 1108–1116. https://doi.org/10.1016/j.chemosphere.2011.01.031

Kalbitz, K., Wennrich, R., 1998. Mobilization of heavy metals and arsenic in polluted wetland soils and its dependence on dissolved organic matter. Sci. Total Environ. 209, 27–99.

Keimowitz, A.R., Zheng, Y., Chillrud, S.N., Mailloux, B., Jung, H.B., Stute, M., Simpson, H.J., 2005. Arsenic redistribution between sediments and water near a highly contaminated source. Environ. Sci. Technol. 39, 8606–8613. https://doi.org/10.1021/es050727t

Kirk, M.F., Holm, T.R., Park, J., Jin, Q., Sanford, R.A., Fouke, B.W., Bethke, C.M., 2004. Bacterial sulfate reduction limits natural arsenic contamination in groundwater. Geology 32, 953–956.

Kirk, M.F., Roden, E.E., Crossey, L.J., Brealey, A.J., Spilde, M.N., 2010. Experimental analysis of arsenic precipitation during microbial sulfate and iron reduction in model aquifer sediment reactors. Geochim. Cosmochim. Acta 74, 2538–2555. https://doi.org/10.1016/j.gca.2010.02.002

La Force, M.J., Hansel, C.M., Fendorf, S., 2000. Arsenic speciation, seasonal transformations and co- distribution with iron in a mine waste-influenced palustrine emergent wetland. Environ. Sci. Technol. 34, 3937–3943. https://doi.org/10.1021/es0010150

Lakefield Research Centre, 1975. Results of Metallurgical Testwork on Salmita Muck. Obtained from NTGO Library.

Lambert, M.B. and Henderson, J.B., 1980. A uranium-lead age of zircons from volcanics and metasediments of the Back River Volcanic Complex, eastern Slave Province, District of Mackenzie; in Current Research, Part C, Geological Survey of Canada, Paper 80-1 C, p. 239-242.

Langner, P., Mikutta, C., Kretzschmar, R., 2012. Arsenic sequestration by organic sulphur in peat. Nat. Geosci. 5, 66–73. https://doi.org/10.1038/ngeo1329

Langner, P., Mikutta, C., Kretzschmar, R., 2014. Oxidation of organosulfur-coordinated arsenic and realgar in peat: Implications for the fate of arsenic. Environ. Sci. Technol. 48, 2281–2289. https://doi.org/10.1021/es4049785

Lindeberg, C., Bindler, R., Renberg, I., Emteryd, O., Karlsson, E., Anderson, N.J., 2006. Natural fluctuations of mercury and lead in Greenland lake sediments. Environ. Sci. Technol. 40, 90–95. https://doi.org/10.1021/es051223y

Lorax Environmental, 2007. Tundra Mine Geochemical Assessment. Pp. 105. Internal Report.

NOAMI, 2017. NOAMI Inventory of Orphaned/Abandoned Mines. http://noami.org/choosemap_e.php

MacDonald, G.M., 1983. Holocene vegetation history ofthe upper Natla River Area, North- west Territories, Canada. Arct. Alp. Res. 15, 169–180. https://doi.org/ 10.1080/00040851.1983.12004341

MacDonald, G.M., 1987. Postglacial vegetation history of the Mackenzie River Basin. Quat. Res. 28, 245–262. https://doi.org/10.1016/0031-0182(89)90001-1

MacDonald, G.M., Edwards, T.W.D., Moser, K.A., Pienitz, R., Smol, J.P., 1993. Rapid response of treeline vegetation and lakes to past climate warming. Nature 361, 243–246. https://doi.org/10.1038/361243a0

Martin, A.J., Pedersen, T.F., 2002. Seasonal and interannual mobility of arsenic in a lake impacted by metal mining. Environ. Sci. Technol. 36, 1516–1523. https://doi.org/10.1021/es0108537

McGlynn, J., Henderson, J., 1970. Archean volcanism and sedimentation in the Slave Structural Province, in: Symposium on Basins and Geosynclines of the Canadian Shield. Geological Survey of Canada, pp. 31–44. https://doi.org/https://doi.org/10.4095/119840

McGuire, A.D., Anderson, L.G., Christensen, T.R., Dallimore, S., Guo, L., Hayes, D.J., Heimann, M., Lorenson, T.D., Macdonald, R.W., Roulet, N., 2009. Sensitivity of the carbon cycle in the Arctic to climate change. Ecol. Monogr. 79, 523–555. https://doi.org/10.1890/08-2025.1

Miller, C.B., Parsons, M.B., Jamieson, H.E., Swindles, G.T., Nasser, N.A., Galloway, J.M. 2019. Lake- specific controls on the long-term stability of mining-related, legacy arsenic contamination and geochemical baselines in a changing northern environment, Tundra Mine, Northwest Territories, Canada. Applied Geochemistry, 109, 104403. https://doi.org/10.1016/j.apgeochem.2019.104403

Mirza, A.M., Elliott, B., 2019. Aeromagnetic survey of the Itchen Lake and Point Lake area, Northwest Territories, parts of NTS 086H, 086A, and 076E; Northwest Territories Geological Survey, NWT Open Report 2019-003. https://www.nwtgeoscience.ca/news/new-geophysical-data-released-slave- province

Moir, I., Falck, H., Hauser, B., and Robb, M., 2006. The history of mining and its impact on the development of Yellowknife; Chapter 2 in Gold in the Yellowknife Greenstone Belt, Northwest Territories: Results of the EXTECH III Multidisciplinary Research Project, (ed.) C.D. Anglin, H. Falck, D.F. Wright, and E.J. Ambrose; Geological Association of Canada, Mineral Deposits Division, Special Publication No. 3, pp. 11–28.

Moser, K., MacDonald, G., 1990. Holocene vegetation change at treeline north of Yellowknife, Northwest Territories, Canada. Quaternary Research, 34(2), 227–239. https://doi.org/10.1016/0033- 5894(90)90033-H

Minyuk, P.S., Brigham-Grette, J., Melles, M., Borkhodoev, V.Y., Glushkova, O.Y., 2007. Inorganic geochemistry of El’gygytgyn Lake sediments (northeastern Russia) as an indicator of paleoclimatic change for the last 250 kyr. J. Paleolimnol. 37, 123–133. https://doi.org/10.1007/s10933-006-9027-4

O’Day, P.A, Vlassopoulos, D., Root, R., Rivera, N., 2004. The influence of sulfur and iron on dissolved arsenic concentrations in the shallow subsurface under changing redox conditions. Proc. Natl. Acad. Sci. U. S. A. 101, 13703–13708. https://doi.org/10.1073/pnas.0402775101

Outridge, P., Sanei, H., Goodarzi, F., 2007. Evidence for Control of Mercury Accumulation Rates in Canadian High Arctic Lake Sediments by Variations of Aquatic Primary Productivity. Environ. Sci. Technol. 41, 5259–5265. https://doi.org/10.1021/es070408x

Outridge, P.M., Sanei, H., Stern, G.A., Goodsite, M., Hamilton, P.B., Carrie, J., Goodarzi, F., Macdonald, R.W., 2011. Response to comment on climate change and mercury accumulation in Canadian high and sub-Arctic lakes. Environ. Sci. Technol. 45, 6705–6706. https://doi.org/10.1007/s10533-010- 9543-1.(8)

Outridge, P.M., Sanei, H., Courtney Mustaphi, C.J., Gajewski, K., 2017. Holocene climate change influences on trace metal and organic matter geochemistry in the sediments of an Arctic lake over 7,000 years. Appl. Geochemistry 78, 35–48. https://doi.org/10.1016/j.apgeochem.2016.11.018

Palmer, M.J., Galloway, J.M., Jamieson, H.E., Patterson, R.T., Falck, H., Kokelj, S. V, 2015. The concentration of arsenic in lake waters of the Yellowknife area. NWT Open File 2015-06 29. https://doi.org/10.13140/RG.2.1.2582.5041

Palmer, M.J., Chételat, J., Richardson, M., Jamieson, H.E., Galloway, J.M., 2019. Seasonal variation of arsenic and antimony in surface waters of small subarctic lakes impacted by legacy mining pollution near Yellowknife, NT, Canada. Sci. Total Environ. 684, 326–339. https://doi.org/10.1016/j.scitotenv.2019.05.258

Pienitz, R., Smol, J.P., MacDonald, G.M., 1999. Paleolimnological reconstruction of Holocene climatic trends from two boreal treeline lakes, Northwest Territories, Canada. Arctic, Antarct. Alp. Res. 31, 82. https://doi.org/10.2307/1552625

Planer-Friedrich, B., London J., McCleskey R.B., Nordstrom D.K., Wallschläger D., 2007. Thioarsenates in geothermal waters of Yellowstone National Park: Determination, preservation, and geochemical importance. Environ Sci Technol. 41, 5245–5251. 26

Prowse, T., Alfredsen, K., Beltaos, S., Bonsal, B., Duguay, C., Korhola, A., McNamara, J., Pienitz, R., Vincent, W.F., Vuglinsky, V., Weyhenmeyer, G.A., 2011. Past and future changes in arctic lake and river ice. Ambio. 40, 53–62. https://doi.org/10.1007/s13280-011-0216-7

Ransom, A., Robb, M., 1986. The Salmita Gold Deposit. Gold West. Shield 38, 285–305.

Redman, A.D., 2002. Natural organic matter affects arsenic speciation and sorption onto hematite 2889– 2896. https://doi.org/10.1021/es0112801

Ritchie, J.C., 1984. A Holocene pollen record of boreal forest history from the Travaillant Lake area, Lower Mackenzie River Basin. Can. J. Bot. 62, 1385–1392. https://doi.org/ 10.1139/b84-188

Rydberg, J., Klaminder, J., Rosén, P., Bindler, R., 2010. Climate driven release of carbon and mercury from permafrost mires increases mercury loading to sub-arctic lakes. Sci. Total Environ. 408, 4778– 4783. https://doi.org/10.1016/j.scitotenv.2010.06.056

Salminen, R., Gregorauskiene, V., 2000. Considerations regarding the definition of geochemical baseline of elements in the surficial materials in areas of differing basic geology. Applied Geology. v. 15, p. 647-653. https://doi.org/10.1016/S0883-2927(99)00077-3

Sanei, H., Goodarzi, F., 2006. Relationship between organic matter and mercury in recent lake sediment: The physical–geochemical aspects. Appl. Geochemistry 21, 1900-1912. http://10.1016/j.apgeochem.2006.08.015

Schindler, D.W., Smol, J.P., 2006. Cumulative effects of climate warming and other human activities on freshwaters of Arctic and subarctic North America. Ambio. 35, 160–68. https://doi.org/10.1579/0044-7447(2006)35[160:CEOCWA]2.0.CO;2

Schuh, C.E., Jamieson, H.E., Palmer, M.J., Martin, A.J., 2018. Solid-phase speciation and post- depositional mobility of arsenic in lake sediments impacted by ore roasting at legacy gold mines in the Yellowknife area, Northwest Territories, Canada. Appl. Geochemistry 91, 208–220. https://doi.org/10.1016/j.apgeochem.2017.10.025

Schuh, C.E., 2019. Geochemical and mineralogical characterization of arsenic in lake sediments impacted by legacy gold mining in the Yellowknife area. PhD Thesis. Queen’s University. https://qspace.library.queensu.ca/handle/1974/26133

Seabridge Gold Inc., 2010. Courageous Lake Project Description in Support of a Class “ A ” Land Use Permit (Former Permit MV2003C0050), Seabridge Gold Inc. http://registry.mvlwb.ca/Documents/MV2010C0046/MV2010C0046 - Project Description - Oct06- 10.pdf

Seabridge Gold Inc., 2018. Seabridge discovers new gold zones at Courageous Lake. News Release: May, 9, Seabridge Gold Inc. http://seabridgegold.net/pdf/NR/NMay9-18.pdf

SENES Consultants, 2006. Tundra Mine Site Ecological Risk Assessment of Water Management. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

SENES Consultants, 2008. Remediation Completion Report for Tundra Mine Phase I 2007 Restoration Program Tundra Mine , NT. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

SENES Consultants Ltd., 2011. Development of Tundra Mine construction monitoring, long-term monitoring, and status of the environment programs. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

Silke, R., 2011. Tundra Mine Heritage Project. http://oocities.org/baja/outback/6406/Tundra.html.tmp

Slater, D.S., 1985. Pollen analysis of postglacial sediments from Eildun Lake, District of Mackenzie, NWT, Canada. Can. J. Earth Sci. 22, 663–674. https://doi.org/10.1139/e85-073

Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochemistry 17, 517–568. https://doi.org/10.1016/S0883-2927(02)00018-5

Smol, J.P., Wolfe, A.P., Birks, H.J.B., Douglas, M.S. V., Jones, V.J., Korhola, A., Pienitz, R., Ruhland, K., Sorvari, S., Antoniades, D., Brooks, S.J., Fallu, M.-A., Hughes, M., Keatley, B.E., Laing, T.E., Michelutti, N., Nazarova, L., Nyman, M., Paterson, A.M., Perren, B., Quinlan, R., Rautio, M., Saulnier-Talbot, E., Siitonen, S., Solovieva, N., Weckstrom, J., 2005. Climate-driven regime shifts in the biological communities of arctic lakes. Proc. Natl. Acad. Sci. 102, 4397–4402. https://doi.org/10.1073/pnas.0500245102

Smol, J.P., Douglas, M.S. V, 2007. From controversy to consensus: making the case for recent climate using lake sediments. Front. Ecol. Environ. 5, 466–474. https://doi.org/10.1890/060162

Stanley, M.C., Sequin, J.M., 1989. Geological Exploration Report of the Courageous-MacKay Lake Volcanic Belt. Obtained from: NTGO Library.

Stern, G.A., Macdonald, R.W., Outridge, P.M., Wilson, S., Chételat, J., Cole, A., Hintelmann, H., Loseto, L.L., Steffen, A., Wang, F., Zdanowicz, C., 2012. How does climate change influence arctic mercury? Sci. Total Environ. 414, 22–42. https://doi.org/10.1016/j.scitotenv.2011.10.039

Sulphur, K.C., Goldsmith, S.A., Galloway, J.M., Macumber, A., Griffith, F., Swindles, G.T., Patterson, R.T., Falck, H., Clark, I.D., 2016. Holocene fire regimes and treeline migration rates in sub-arctic Canada. Glob. Planet. Change 145, 42–56. https://doi.org/10.1016/j.gloplacha.2016.08.003

Szeicz, J.M., MacDonald, G.M., 1995. Recent white spruce dynamics at the subarctic alpine treeline of North-Western Canada. J. Ecol. 83, 873–875. https://doi.org/ 10.2307/2261424

Szeicz, J.M., MacDonald, G.M., 2001. Montane climate and vegetation dynamics in east- ernmost Beringia during the Late Quaternary. Quat. Sci. Rev. 20, 247–257. https://doi.org/10.1016/S0277- 3791(00)00119-0

Terrax Minerals, 2019. In Control of Major Canadian Gold District Yellowknife City Gold Project Prioritized Targets. Pg. 2, Terrax Minerals. https://www.terraxminerals.com/site/assets/files/2458/txr_2pager_15jan2019.pdf

Tetra Tech Wardrop, 2012. Courageous Lake Prefeasibility Study, Tetra Tech Wardrop. https://secure.kaiserresearch.com/i/jk/tr16/TRSEA20120905.pdf

Toevs, G.R., Morra, M.J., Polizzotto, M.L., Strawn, D.G., Bostick, B.C., Fendorf, S., 2006. Metal(loid) diagenesis in mine-impacted sediments of lake Coeur d’Alene, Idaho. Environ. Sci. Technol. 40, 2537–2543. https://doi.org/10.1021/es051781c

Upiter, L.M., Vermaire, J.C., Patterson, R.T., Crann, C.A., Galloway, J.M., Macumber, A.L., Neville, L.A., Swindles, G.T., Falck, H., Roe, H.M., Pisaric, M.F.J., 2014. Middle to late Holocene chironomid-inferred July temperatures for the central Northwest Territories, Canada. J. Paleolimnol. 52, 11–26. https://doi.org/10.1007/s10933-014-9775-5

URS, 2005. Geochemical Assessment of Acid Rock Drainage and Metal Leaching Potential of Tailings and Waste Rock, Tundra Mine, NWT. Internal Report.

Usher, P.J., 2000. Traditional Ecological Knowledge in Environmental Assessment and Management. Arctic, 52, 183–193. https://doi.org/10.14430/arctic849

Van Den Berghe, M.D., Jamieson, H.E., Palmer, M.J., 2018. Arsenic mobility and characterization in lakes impacted by gold ore roasting, Yellowknife, NWT, Canada. Environ. Pollut. 234, 630–641. https://doi.org/10.1016/j.envpol.2017.11.062

Wagemann, R., Snow, N.B., Rosenberg, D.M., Lutz, A., 1978. Arsenic in sediments, water and aquatic biota from lakes in the vicinity of Yellowknife, Northwest Territories, Canada. Arch. Environ. Contam. Toxicol. 7, 169–191. https://doi.org/10.1007/BF02332047

Wang, Y., Jiao, J.J., 2014. Multivariate statistical analyses on the enrichment of arsenic with different oxidation states in the Quaternary sediments of the Pearl River Delta, China. J. Geochemical Explor. 138, 72–80. https://doi.org/10.1016/j.gexplo.2013.12.012

Wang, Y., Le Pape, P., Morin, G., Asta, M.P., King, G., Bártová, B., Suvorova, E., Frutschi, M., Ikogou, M., Pham, V.H.C., Vo, P. Le, Herman, F., Charlet, L., Bernier-Latmani, R., 2018. Arsenic Speciation in Mekong Delta Sediments Depends on Their Depositional Environment. Environ. Sci. Technol. 52, 3431–3439. https://doi.org/10.1021/acs.est.7b05177

Waychunas, G.A., Rea, B.A., Fuller, C., Davis, J.A., 1993. Surface chemistry of ferrihydrite: Part 1.EXAFS studies on geometry of coprecipitated and adsorbed arsenate. Geochim. Cosmochim. Acta 57, 2251–2269. https://doi.org/10.1016/0016-7037(93)90567-G

WESA, 2013. Tundra Phase II Remediation Activities: Water Treatment Plant Operation (Monitoring Plan: MV2009L8-0008).

Wilkin, R.T., Ford, R.G., 2006. Arsenic solid-phase partitioning in reducing sediments of a contaminated wetland. Chem. Geol. 228, 156–174. https://doi.org/10.1016/j.chemgeo.2005.11.022

Woelders, L., Lenaerts, J.T.M., Hagemans, K., Akkerman, K., Van Hoof, T.B., Hoek, W.Z., 2018. Recent climate warming drives ecological change in a remote high-Arctic lake. Sci. Rep. 8, 1–8. https://doi.org/10.1038/s41598-018-25148-7

Wolf, J., Allice, I., Bell, T., 2013. Values, climate change, and implications for adaptation: Evidence from two communities in Labrador, Canada. Glob. Environ. Chang. 23, 548–562. https://doi.org/10.1016/j.gloenvcha.2012.11.007

Chapter 2

Lake-specific controls on the long-term stability of mining-related, legacy

arsenic contamination and geochemical baselines in a changing northern

environment, Tundra Mine, Northwest Territories, Canada

2.1 Abstract

Climate change is influencing the biogeochemistry of northern lake ecosystems. These changes may affect the mobility of naturally occurring metal(loid)s and long-term stability of anthropogenic contaminants. Arsenic (As) concentrations in lake sediments in the Courageous Lake Greenstone Belt,

Northwest Territories, Canada, are elevated from the operation of two high-grade, low-tonnage historical gold mines (Tundra Mine and Salmita Mine) and the weathering of mineralized bedrock. In sensitive sub-

Arctic environments, it is not currently known how the cumulative effects of resource extraction and climate warming will impact geochemical baselines and the long-term stability of legacy contaminants. In this study, measurements of As concentration and speciation in waters and sediments are combined with multivariate analyses of climate proxies (sediment particle size and organic matter composition) from five lakes downstream of the former Tundra Mine site. Data from lake sediment cores were divided into geochemically distinct populations using a combination of radiometric dating and constrained incremental sum-of-squares cluster analysis to define geochemical baselines, examine the lake-specific controls on As distribution, and determine climate-related factors that may influence the long-term stability of As.

Median As concentrations in near-surface impacted sediments (median: 110 mg·kg-1; range: 31 to 1,010 mg·kg-1; n = 22) and pre-mining sediment (median: 40 mg·kg-1; range: 28 to 170 mg·kg-1; n = 102) exceed the Canadian Council of the Ministers of the Environment Probable Effects Level of 17 mg·kg−1.

Near the Tundra Mine, the long-term stability of As in the near-surface sediment is influenced by the source of As (direct disposal and weathering of waste rock, tailings overtopping and seepage, discharge of treated tailings effluent, weathering and airborne deposition of tailings and waste rock, and natural

weathering of mineralized bedrock), lithology of the sediment, and composition of sediment organic matter. This study demonstrates that in lakes impacted by weathering of waste rock and mineralized bedrock, As in sediments is primarily hosted by Fe-(oxy)hydroxides and may be more susceptible to remobilization with climate warming relative to those lakes impacted by direct discharge of mine wastes where As-bearing sulphides are the most abundant As host. Continued climate warming is expected to increase the natural loading of metal(loid)s and organic matter to lake sediments; however, the effects of these changes on the long-term stability of legacy contaminants will vary between lakes.

2.2 Introduction

Geochemical baselines and background values provide guidance for monitoring mining impacts in the environment and are used in the development of remediation objectives upon mine closure.

Presently, mine site remediation projects in Canada commonly rely on regional geochemical baseline values or guidelines based on national average metal(loid) concentrations in environmental media

(CCME, 2001a, b; Environment Canada, 2012). As metal(loid) concentrations in lakes reflect local variations of elements in the environment, naturally occurring concentrations in water and sediment may exceed national averages in mineralized areas. In Canada’s North, low and highly variable sedimentation rates in lakes also contribute to local variations in the distribution and concentration of metal(loid)s in sediments (Salminen and Tarvainen, 1997; Macumber et al., 2011; Crann et al., 2015). At northern latitudes, climate change is profoundly affecting the seasonality, biological productivity, and hydrology of lakes and has been shown to impact the mobility of certain metal(loid)s, such as mercury (Hg), and influence geochemical baselines (Macdonald et al., 2005; Outridge et al., 2007, 2011, 2017; McGuire et al., 2009; Jiang et al., 2011; Sanei et al., 2012). The thawing of permafrost, as a result of climate warming, has also increased the flux of organic carbon and loading of metal(loid)s (i.e. Hg) into sub-

Arctic surface waters (Guo et al., 2007; McGuire et al., 2009; Rydberg et al., 2010). However, the influence of climate change on the mobility of other elements, such as As, is not well established, leading

to uncertainty in the determination of geochemical baselines and the long-term monitoring of mining- derived contaminants in mineralized regions.

In northern Canada, Archean orogenic gold (Au) deposits are often hosted in metamorphosed mafic to ultramafic volcanic sequences with associated sedimentary rocks known as greenstone belts

(Goldfarb et al., 1995; Groves et al., 1998). The Au-mineralization in these deposits commonly occurs in quartz veins containing carbonates, sericite, pyrite and arsenopyrite (FeAsS). Due to the occurrence of

As-bearing sulphides in association with these deposits, mining and processing of Au may result in increased As concentrations in sediments and surface waters surrounding both modern and historical mine sites in the Northwest Territories (Wagemann et al., 1978; Groves et al., 1998; Bright et al., 1994, 1996;

Golder Associates Ltd., 2005; Andrade et al., 2010; Galloway et al., 2012, 2017; AECOM, 2015; Palmer et al., 2015; Houben et al, 2016; Jamieson et al., 2017; Van Den Berghe et al., 2018). Numerous studies have examined mechanisms influencing As mobility in mining-impacted lake systems to better understand implications for human and environmental health and to establish regional baselines (e.g.

Mudroch et al., 1989; Azcue et al., 1994; Bright et al., 1994; Thornton, 1996; Martin and Pedersen,

2002). These studies demonstrate that controls on As speciation and mobility differ locally within lakes and between studied lakes due to factors such as water depth and morphology, seasonal thermal stratification, water residence time, primary productivity, sediment organic and inorganic geochemistry, and catchment bedrock lithology (Martin and Pedersen, 2002; Hollibaugh et al., 2005; Wang and

Mulligan, 2006; Andrade et al., 2010; Hasegawa et al., 2010; Campbell and Nordstrom, 2014; Galloway et al., 2017; Schuh et al., 2018, 2019; Van Den Berghe et al., 2018; Barrett et al., 2019; Palmer et al.,

2019). As a result of lake-specific controls, the dispersion and long-term stability of mining-related As contamination in lacustrine environments may also vary within a scale of a single mining district.

Detailed characterization of lake sediment can help to accurately monitor the environmental impacts of mining activities and evaluate the efficacy of mine site remediation projects.

Two high-grade, low-tonnage historical gold mines have operated in the Courageous Lake

Greenstone Belt (CLGB) of the Northwest Territories (NT): Tundra Mine (1964 to 1968) and Salmita 32

Mine (1983 to 1987) (Figure 2.1). During both phases of mining, gold was recovered through mercury amalgamation and cyanidation at the Tundra Mine mill (Silke, 2009). Gold in the region is free-milling and associated with arsenopyrite from hydrothermal mineralization (Ransom and Robb, 1986). As a result, historical mining and milling activities led to the accumulation of As-bearing waste rock, tailings, and impacted waters on site (Figure 2.1; URS, 2005; Golder Associates Ltd., 2008). The Tundra Mine was in care and maintenance mode from 1999 to 2007 under the ownership of the Contaminants and

Remediation Directorate (CARD) of Crown-Indigenous Relations and Northern Affairs Canada

(CIRNAC). Remediation activities occurred from 2007 to 2018 including: removal of on-site buildings, dam and landfill repairs; geotechnical inspection of dams; tailings water treatment and disposal; waste rock and tailings kinetic testing; capping and lining of waste rock and tailings; and water quality monitoring (Golder Associates Ltd., 2005, 2008, 2016, 2017; INAC, 2005, 2008, 2010; URS, 2005;

SENES Consultants, 2006, 2008, 2011; Lorax Environmental, 2007; AANDC, 2012, 2013, 2014;

AECOM, 2015). Waste rock at the former mine site contains 0.07% to 6.5% sulphides, with arsenopyrite, pyrite (FeS2), and pyrrhotite (Fe1-xS) as the primary sulphide minerals (URS, 2005). In tailings, As concentrations range from 70 to > 10,000 mg·kg-1 (median = 1, 630 mg kg-1; n = 35) with arsenopyrite accounting for 60 to 98 % of total As (Miller et al., unpublished data). Analysis of surface waters and groundwaters and sediment grab samples during environmental monitoring shows that weathering of waste rock and overtopping and seepage of waters from the tailings confinement area (TCA) have resulted in contamination of lakes up to 4 km downstream of the former mine site (INAC, 2005; URS,

2005, SENES Consultants, 2006; Lorax Environmental, 2007). Lakes of the Tundra Mine region may contain elevated As from both hydrothermal mineralization and legacy gold mining activities. However, no baseline data were collected prior to exploration and the initial development of Tundra Mine in 1951, thus, delineating mining impacts and monitoring the long-term effectiveness of remediation remains a challenge (Moore, 1978; Hatfield Consulting, 1982). Following the completion of remediation activities in August 2018, the site is presently in an adaptive management and long-term monitoring phase

(AECOM, 2018). The CLGB is located within the Tłı̨ chǫ Land Claim Agreement and the North Slave 33

Metis traditional lands and is also highly prospective for future mineral development (Seabridge, 2010;

Tetra Tech-Wardrop, 2012; Government of the Northwest Territories, Executive and Indigenous Affairs,

2007).

Through combined geochemical, mineralogical, and limnological techniques, the present study aims to characterize legacy mining impacts in lakes downstream of the former Tundra gold mine and determine the lake-specific controls on the mobility of As. The main objective of this study is to assess the implications of continued climate warming on the long-term stability of mining-derived As contamination in sub-Arctic lakes to help inform long-term monitoring activities and support sustainable development of mineral resources in northern Canada.

2.3 Study Location

2.3.1 Physiography

Tundra Mine (64.042448 N, 111.169942 W) is located approximately 240 km NE of

Yellowknife on the eastern shore of Matthews Lake (Figure 2.1). In winter months the site is accessible by a spur road off the Tibbitt to Contwoyto Winter Road. A landing strip near the mine site provides access during the summer. Located within the boreal forest-open tundra zone 80 km north of the treeline, the landscape of the region is characterized by low relief and slightly rounded topography of the Canadian tundra (Hatfield Consulting, 1982). Temperatures in the region range from -31 to 10 C with an average annual temperature of -12 to -9 C (SENES Consultants, 2006). Open water season ranges from 94 to 126 days with break-up in late June and freeze-up in September (INAC, 2005).

Figure 2.1 A) Map showing simplified bedrock geology, sediment gravity core sampling locations, predominant mine drainage pathways originating from the former Tundra Mine site, and lake depths at sampling locations (in brackets) (geology modified from Folinsbee and Moore (1955) & Thompson and Kerswill (1994)); B) Bathymetry map of Hambone Lake with numbered locations of sediment grab samples (bathymetry map rendered from 50 water depth measurements).

2.3.2 Geological setting

The Tundra Mine deposits and other regional gold showings of the CLGB are associated with the northwest-trending Yellowknife Supergroup of the Central Slave Geological Province (Henderson, 1970;

McGlynn and Henderson, 1970; Dillion-Leitch, 1981; Ransom and Robb, 1986). In the Courageous Lake region, the CLGB unconformably overlies thin, discontinuous sequences of volcanic, clastic, and banded iron formations of the Central Slave Basement Cover Group (~2.39 Ga) and diorite to tonalite gneiss of the Central Slave Basement Complex (~3.325 Ga) (Bleeker et al., 1999). The CLGB is flanked to the east by conformably overlying turbiditic metasedimentary rocks (Figure 2.1; Folinsbee and Moore, 1950,

1955; Moore, 1951). Multiple gold showings have been discovered throughout the CLGB. In the

Tundra/Salmita Mine region, Au-mineralization is primarily free-milling and hosted in quartz-carbonate– bearing shear zones that occur at the contact between finely laminated argillaceous metasediments of the

Burwash Formation and Banting Group volcanics, or in quartz stringers filling narrow shear zones within the metasediments (Ransom and Robb, 1986). In mineralized bedrock of the CLGB, arsenopyrite is found as stringers within quartz veins, in zones of disseminated sulphides within felsic to intermediate tuffaceous rocks, in association with carbonatization and silicification of mafic volcanic rocks, and in trace amounts within slate (Ransom and Robb, 1986) (Figure 2.1).

2.4 Methods

2.4.1 Sampling locations and sample preparation

Two primary surface water flow paths, originating from the former Tundra Mine site, have been identified by environmental monitoring activities (Figure 2.1) (INAC, 2005, 2008, 2010; SENES

Consultants Ltd., 2006, 2008, 2011; AANDC, 2012, 2013, 2014). Two lakes from each flow path (I:

Hambone Lake, Powder Mag Lake; Pathway II: Bulldog Lake, Matthews Lake) were sampled for the present study (Figure 2.1). One additional sampling site, Control Lake, is not connected to the hydrological flow paths originating at the mine site and has been used as a background (unimpacted) reference for ongoing monitoring activities (Golder Associates Ltd., 2005; Figure 2.1).

In March 2016, sediment cores, porewaters, and surface water samples were collected through the ice from Powder Mag, Bulldog, Matthews, and Control lakes. In July 2016, near-surface sediment samples were collected from Hambone Lake (Figure 2.1). Prior to collection of sediment cores, surface water samples were collected from the middle of the water column using a Kemmerer water sampler. At each site, water quality variables (dissolved oxygen (DO), temperature, and conductivity) were measured in-situ using a YSI sonde (Professional Plus with a Quattro Probe Cable) and pH was measured with a hand-held pH meter (HACH H138 MiniLab ISFET). Surface water samples were placed in a 1-L High-

Density Polyethylene (HDPE) bottle, triple-rinsed with sample water, and transported to a clean indoor location for filtering and preservation within 12 hours of collection. Samples for dissolved cations and anions were filtered (< 0.45 µm) into 60 mL HDPE bottles using an all-plastic 50 mL syringe (Norm-

Ject® Sterile Luer-Lock Syringe) and Sterivex® capsule filters. No evidence of Fe oxidation was observed in the samples in the time between collection and filtration. Cation samples were acidified to pH

< 2 using ultrapure nitric acid. Samples for total As (unfiltered) and inorganic As speciation were collected in amber-coloured 60 mL HDPE bottles (Nalgene® 2106-0002). Samples for As speciation were acidified with 300 µL concentrated acetic acid and preserved with 600 µL of 0.25 M EDTA; total

As samples were preserved with 2% HCl. Duplicate surface water samples were collected at one of the four lakes. Travel blanks, acid blanks, and sample blanks were included for analysis. All samples were then refrigerated at 4C and stored in a dark location prior to shipment.

Due to the remote location of the Tundra Mine site, transport of sediment cores from the sampling location had the potential to disturb the sediment and result in the oxidation of redox-sensitive elements, such as As. Sediment cores were therefore collected in the winter to allow for the transport of nitrogen (N2) to the site and subsampling of sediment cores on the frozen lake surface. Sediment cores were collected using a gravity corer and transparent polycarbonate core tubes (7.5 cm x 60 cm). To account for the influence of sediment focusing (Blais and Kalff, 1995) and to avoid sediment disturbance caused by seasonal grounding of lake ice, all sediment cores were collected from the deepest location in

each lake (Zmax) which was determined based on previous bathymetric surveys (Golder Associates Ltd.,

2005). Immediately following collection, sediment cores were vertically extruded and subsampled at 1- cm resolution in a high purity (99.998 %) N2-filled glove bag. Sediment was subsampled into N2 flushed

50 mL polypropylene centrifuge tubes (Corning® Falcon®) and refrigerated pending shipment to the

Geological Survey of Canada – Atlantic Division (GSC-A) in Dartmouth, NS, for porewater extraction.

Surface sediment grab samples were collected in July 2016 from Hambone Lake using an Ekman dredge sampler. Samples were collected along a traverse from the TCA effluent discharge location to the outlet into Powder Mag Lake (Figure 2.1). As the near-surface sediment in Hambone Lake likely freezes during winter months, the collected sediment samples were frozen and sent to Queen’s University, Kingston,

ON, for subsampling and preparation.

To separate porewater from the sediment, samples were centrifuged (Thermo Scientific Sorvall

Legend™ XF Centrifuge) at 4,400 RPM for 30 minutes (centrifugal force of 4221 Gs) at GSC-A.

Porewater was then pipetted from the sample tubes in a high purity (99.998%) N2-filled glove bag.

Samples were prepared following the protocol described above for filtered cations, filtered anions, total

As, and total inorganic As speciation. For each sample pH was measured but due to low water content, Eh could not be measured. Following porewater extraction, sediment samples were refrigerated at 4C and stored in a dark location prior to shipment to Queen’s University for preparation and geochemical analysis.

2.4.2 Analyses

2.4.2.1 Surface and porewater geochemistry

Surface water samples were analyzed for anions, alkalinity, dissolved organic carbon (DOC), and a 64-element suite of metal(loid)s at the Inorganic Geochemical Research Laboratory of the GSC.

Analyses of major elements were performed by Inductively Coupled Plasma – Atomic Emission

Spectroscopy (ICP-AES) using a Perkin-Elmer 3000 DV. Trace elements were analyzed using ICP-MS with a Thermo Corporation X-7 Series II. Anion concentrations were measured using a Dionex DX-600

ion chromatograph using an AS-18 column and gradient elution. Dissolved organic carbon was measured on a Shimadzu TOC-5000 analyzer following removal of inorganic carbon using phosphoric acid.

Alkalinity measurements were completed using a PC-Titrate system. For each sample set, analyses were completed on laboratory duplicates, blanks, and certified standards (SLRS-5). For laboratory duplicates, relative standard deviation (RSD) ranges from 0.08% to 1.40 % and mean percent difference (MPD) ranges from 0.12% to 1.98 % for As. For field duplicates RSD and MPD are higher for unfiltered samples

(2.44 % and 3.45 %, respectively) than filtered samples (2.05 % and 2.90 %). Arsenic was below detection limit (0.1 µg·L-1) in all field blanks (n = 4). Inorganic As speciation was analyzed by hydride generation-atomic fluorescence spectrometry (HG-AFS; model PSA 10.055 Millennium Excalibur) at the

Université de Montréal. Dissolved As speciation by HG-AFS is comprised of three constituents: inorganic As (III) and As (V), and a residual fraction (AsR = AsTotal – (As (V) + As (III))). The residual fraction includes As species that are not detected using HF-AFS, including thioarsenates in sulfidic waters and non-reducible organoarsenic compounds (Planer-Friedrich and Wallschläger, 2009; PS Analytical,

2018). Standard solutions and certified inter-calibration samples were used to assess accuracy and precision. Trace amounts of As (0.12, 0.16 µg·L-1) were detected in two blanks analyzed for total As, however, As was not detected in the other (n = 6) duplicate blanks.

2.4.2.2 Chronology and sedimentation rates

Near-surface sediment was dated using gamma spectrometry for the four cores collected. Samples for gamma dating were prepared following Schelske et al. (1994). Analysis was completed at the

Paleoecological Environmental Assessment Research Laboratory (PEARL, Queen’s University) using an

Ortec High Purity Germanium Gamma Spectrometer. Total 137Cs, 214Bi, 210Pb, and 214Pb results were processed using ScienTissiME 2.1.2 software.

2.4.2.3 Sediment geochemistry

Freeze dried sediment sub-samples (n = 211) were submitted to Acme Analytical Laboratories

(Bureau Veritas), Vancouver, BC, for geochemical analyses. A modified aqua regia digestion protocol

(1:1:1 HCl:HNO3:H2O at 95 °C for one hour) was used prior to elemental analysis via inductively coupled 39

plasma – mass spectroscopy (ICP-MS) (AQ250-EXT package (53 elements)). Near-total (4-acid) digestions were not used as they involve high-temperature fuming with hydrofluoric acid that may volatilize As, antimony (Sb), Hg, and sulphur (S), which are key elements of interest in this study

(Parsons et al., 2012, 2019). Sediments were pulverized to approximately 200 mesh size (74 µm) prior to digestion.

One field duplicate was submitted for every 10 samples to assess sampling accuracy and precision. In duplicate samples, RSD ranges from 0.21% to 7.22% and MPD ranges from 0.29% to

10.21% for As. Certified reference materials (STSD-3 (Lynch, 1990)) and laboratory standard reference materials (STD DS10, STD OREAS45EA) were included with each sample set to assess analytical accuracy and precision. For STSD-3 mean measured As concentration was 24.34 ± 0.81 mg·kg-1 (n = 8) vs. an expected concentration of 22 mg·kg−1 for As following aqua regia digestion (RSD 3.3%). Mean measured As concentration for STD DS10 was 42.3 ± 0.77 mg·kg-1 (n = 8) vs. an expected concentration of 46.2 mg·kg−1 As following aqua regia digestion.

2.4.2.4 Characterization of As-bearing solid phases

The distribution of As-bearing solids in lake sediment was characterized using Scanning Electron

Microscopy (SEM) coupled with Automated Mineralogy and integrated Energy-Dispersive X-ray

Spectroscopy (EDS) analyses at Queen's University, Kingston, ON. Through a combination of backscatter electron (BSE) image analysis and EDS, SEM-based Automated Mineralogy was used to identify and quantify relative abundance of common As-bearing minerals (Gu, 2003; Fandrich et al., 2007;

Buckwalter-Davis, 2013; Van Den Berghe, 2018). Sparse phase liberation (SPL) analysis mode with a user-defined BSE greyscale range of 120 to 255 was used to selectively identify As-hosting phases, therefore does not provide bulk mineralogy information (Fandrich et al., 2007). Solid phases thought to contain trace concentrations of As were analyzed using a JEOL JXA-8230 electron probe microanalyzer

(EPMA) operating in wavelength dispersive spectroscopy (WDS) mode. The relative contribution of each

As-hosting phase to bulk sediment As concentrations was calculated using the average As concentration

of each phase and the fraction of each As-bearing species obtained by SEM-based automated mineralogy following the methods of Van Den Berghe et al. (2018).

2.4.2.5 Sediment textural and organic characteristics

Sedimentary grain size was determined using a Beckman Coulter LS 13320 laser diffraction particle size analyzer fitted with a Universal Liquid Module over a measurement range between 0.4 and

2000 µm. Followings methods outlined in van Hengstum et al. (2007), Donato et al. (2009), and Murray

(2002) sediments were treated with 10 % HCl to remove authigenic carbonate particles and 35 % H2O2 to remove organic matter (OM). Results were processed with the software package GRADISTAT v8.0

(Blott and Pye, 2001) to calculate grain size statistical parameters. Garnet15 (mean diameter 15 ± 2 μm), an accuracy standard supplied by Beckman Coulter, was run once per month. To assess instrumental precision, an in-house mud sample (Cushendun Mud; mean diameter = 20.5 ± 0.76 μm) was run at the start of every session.

Stepwise pyrolysis (Rock-Eval ® 6) was used to characterize the type and source of OM in the sediment samples (e.g. Sanei et al., 2005; Carrie et al., 2012). Sediment sub-samples were analyzed to determine the proportion of reactive OM (defined as pyrolysable carbon). Pyrolysable carbon is comprised of: (i) the S1 fraction; composed of small, readily degradable, volatile molecules (e.g. biolipids); (ii) the S2 fraction; composed in part of larger molecules of hydrocarbons derived from kerogen (e.g. algal cell wall detritus); and, (iii) the S3 fraction; derived from oxygen-containing organic molecules (Sanei et al., 2005; Sanei and Goodarzi, 2006). A standard reference material (Internal 9107 shale standard, Geological Survey of Canada – Calgary (GSC-C); Ardakani et al., 2016) was run every 5th sample and demonstrates < 1% RSD for Total Organic Carbon (TOC), < 3% RSD for S1 and S2 fractions, and 11% RSD for S3. The lower precision for S3 in bulk samples is expected due to poor peak integration and distinction between S3 OM and S3 carbonates (siderite in shale standard) (Ardakani et al.,

2016). A duplicate sample was run every 10th sample; the majority of duplicate samples demonstrate an

RSD of < 5% for all parameters and an MPD of < 15%.

2.4.2.6 Statistical analysis of controls on As distribution

Elements with concentrations below instrumental detection limits in 25% or more of the samples

(B, Ge, Hf, In, Pd, Pt, Re, Ta, Te) and those that demonstrated little down-core variability (Ga, K, Na, Th,

Tl) were removed prior to statistical analyses. One half of the lower detection limit (LDL) was used for element concentrations below the LDL for retained data (Reimann et al., 2008). No element concentrations exceeded upper detection limits.

Stratigraphically constrained incremental sum of squares (CONISS) cluster analysis compared against a broken stick model using the rioja package in R (R Core Team, 2014; Juggins, 2017) was used to isolate distinct geochemical populations and delineate impacted (MI, I) from background (BG) sediment. Both Anderson Darling and Shapiro-Wilk tests demonstrate a highly non-normal distribution of

As (p < 0.0005 and p < 0.0005, respectively) in sediment samples; therefore, non-parametric tests were chosen for statistical analysis. For cluster analysis, data were log-transformed and converted to standardized values (Davis, 2011).

2.5 Results

2.5.1 Surface Water

2.5.1.1 Physiochemical characteristics and major dissolved species

Surface waters in all four lakes sampled were oxic (range: 3.60 to 12.8 mg·L-1) with circumneutral pH (range: 6.58 to 7.01). Dissolved oxygen concentrations during winter sampling in 2016 were lowest in Powder Mag Lake and lower than reported during the previous summer in Bulldog (range:

8.0 to 16 mg·L-1) and Powder Mag lakes (range: 3.0 to 18 mg·L-1) (AANDC, 2014; Golder Associates

Ltd., 2016). Specific conductance was highest in lakes directly impacted by discharge and disposal associated with mining activities (Powder Mag Lake and Bulldog Lake) and decreased with distance from the TCA. Dissolved organic carbon was highest in Powder Mag Lake, historical monitoring also demonstrates high DOC (16 mg·L-1) concentrations in Hambone Lake (INAC, 2005). Concentrations of major dissolved species are consistent with measured trends in specific conductance and reflect decreased

-2 loading of metals and SO4 with increasing distance from the TCA (Table 2.1; Appendix A). These observed trends are similar to those reported by annual water quality monitoring at Tundra Mine that show decreasing concentrations of dissolved species with increasing distance from point sources of contamination.

Table 2.1 Selected surface water physicochemical parameters, anions (dissolved, filtered), alkalinity (as CaCO3), total (unfiltered) concentrations of selected major and trace elements, and inorganic As speciation from surface waters of lakes of the Tundra Mine region. Lake Name Units CCMEA Powder Mag Bulldog Matthews Control

Date of Sampling -- -- 11-Mar-16 10-Mar-16 13-Mar-16 12-Mar-16

Physical Parameters (field measurements)

Sample Depth m -- 1.0 3.5 5.7 1.8

Temperature  C -- 0.20 3.40 2.90 2.50

Specific Conductance µS·cm-1 -- 833 86.2 49.3 30.7

pH S.U. 6.5 – 9.0 7.01 6.75 6.90 6.58

DO mg·L-1 -- 3.60 7.80 10.4 12.8

Anions, Alkalinity and DOC

-1 SO4 mg·L -- 257 7.87 4.17 1.29

-1 Alkalinity (as CaCO3) mg·L -- 113 23.5 13.9 9.60

DOC mg·L-1 -- 16.6 6.90 4.80 7.50

Major and Trace Elements

Fe mg·L-1 0.30 0.39 0.06 0.02 0.02

AsB µg·L-1 5.0 19.4 3.01 0.95 0.81

Ba µg·L-1 -- 47.1 5.30 3.10 4.00

Sr µg·L-1 -- 238 15.6 13.0 9.60

Zn µg·L-1 7.0 2.40 1.20 0.70 3.20

Arsenic Speciation

As (III)B µg·L-1 -- 0.80 0.17 < 0.09 0.15

As (V)B µg·L-1 -- 15.4 2.57 0.56 0.11

B -1 AsR µg·L -- 3.17 0.28 0.35 0.55

A CCME Water Quality Guidelines for the protection of aquatic life (CCME, 2001b), bold and italicized values exceed guidelines; B Concentration reported based on HG-AFS analysis

2.5.1.2 Metal(loid)s and total As speciation

Concentrations in filtered (< 0.45 µm) and unfiltered surface water samples are similar for all elements with the exception of Al, Mn, Pb and Fe, which demonstrated higher concentrations in unfiltered samples. Analysis of surface waters in Powder Mag Lake demonstrates increased total metal(loid) concentrations with respect to other lakes sampled (Table 2.1; Appendix A). At the time of sampling, As and Fe concentrations exceeded CCME Water Quality Guidelines in Powder Mag Lake. No other elements were measured at concentrations exceeding these guidelines. Total As concentrations are highest in surface waters impacted by tailings effluent and lowest in surface waters farthest from the former mine site (Table 2.1). The highest percentage of AsR is observed in Control Lake (68 %) (Table

2.1).

2.5.2 Chronology and sedimentation rates

Sediment ages and sedimentation rates were calculated using the constant rate of supply model

(Appleby, 2001; Appendix A). Based on the 210Pb decay profile for each lake, disturbance or mixing (i.e. bioturbation, occurrence of slump deposits, sampling error) between sampling intervals was determined to be minimal. Average sedimentation rates in near-surface sediment vary between lakes, ranging from 52 to

20 yr·cm-1. The fastest sedimentation rates are observed in the near-surface sediment of Powder Mag

Lake (2.0 yr·cm-1) and Control Lake (1.75 yr·cm-1); the lowest sedimentation rates in near-surface sediment are observed in the largest lake, Matthews Lake (130 yr·cm-1). The lower sedimentation rates observed in Matthews Lake are similar to those reported by Crann et al. (2015) for similar sized lakes in tundra settings.

2.5.3 Geochemically distinct groupings

As northern lacustrine systems are typically characterized by low sedimentation rates, defining the precise onset of anthropogenic activities in the sediment cores is not possible at the sampling resolution used in this study (1 cm) (Gasiorowski, 2008; Crann et al., 2015; Macumber, 2015). To delineate impacted (MI, I) sediment from background (BG), individual sediment cores were divided into distinct geochemical groupings using a combination of cluster analysis, multivariate element profiles, and 44

radiometric dating (Figure 2.2, 2.3). These combined methods allowed for the identification of changes in sediment geochemistry attributed to the onset of mining activities in each of the sediment cores. In lakes proximal to the mine site (Powder Mag Lake and Bulldog Lake) mining impacts are delineated on Figure

2.2 by the groupings P-MI and B-MI, respectively. Increases observed in As concentration in sediment pre-dating mining activities in these two lakes are attributed to post-depositional As mobility across the redox interface (RI) and are denoted by P-RI and B-RI (Figure 2.2). In lakes more distal to the former mine site impacts are delineated by the groupings ML-MI and CL-I in the Matthews Lake and Control

Lake sediment cores, respectively (Figure 2.3). In all lakes, geochemical groupings in background sediment are apparent from cluster analysis (Figure 2.2, 2.3). These groupings (BG) indicate changes in sediment geochemistry that reflect natural fluctuations in geochemical baselines.

Figure 2.2 Sediment grain size and solid-phase concentrations of As, elements enriched in ore and used in mineral processing (Ag, Au, Hg, Pb, Sb, W, Zn) and redox sensitive elements (S, Fe, and Mn) from sediment cores extracted from Powder Mag Lake and Bulldog Lake. Element concentration profiles are plotted on different scales and scales differ between lakes. Shaded intervals separate mining-impacted (MI, I) and background (BG) geochemical populations identified using CONISS (G1- G4), 210Pb dating, and multivariate element profiles. 46

Figure 2.3 Sediment grain size and solid-phase concentrations of As, elements enriched in ore and used in mineral processing (Ag, Au, Hg, Pb, Sb, W, Zn) and redox sensitive elements (S, Fe, and Mn) from sediment cores extracted from Matthews Lake and Control Lake. Element concentration profiles are plotted on different scales and scales differ between lakes. Shaded intervals separate impacted (MI, I) and background (BG) geochemical populations identified using CONISS (G1- G5), 210Pb dating, and multivariate element profiles.

2.5.4 Sediment geochemistry

Arsenic concentrations range from 31 to 1,010 mg·kg-1 in near-surface, impacted sediment (MI &

I) and 28 to 170 mg·kg-1 in background sediment (BG) (Figure 2.4; Appendix A). In all lakes, the highest solid phase As concentrations (Assed) are evident in mining-impacted sediment, with maximum values observed in Bulldog Lake (median: 570 mg·kg-1; range: 200 to 1,010 mg·kg -1; n = 7) (Figure 2.2, 2.3;

Appendix A). The lowest As concentrations in mining-impacted sediment are observed in Matthews Lake

(range: 31 to 90 mg·kg-1; median: 48 mg·kg-1; n = 13). Control Lake is currently used as a background reference for environmental monitoring activities and is the farthest lake from the former Tundra Mine site; As concentrations in the near-surface sediment of Control Lake range from 110 to 220 mg·kg-1

(Figure 2.4). The highest median background As concentration is observed in Control Lake (CL-BG) (120 mg·kg-1) (Figure 2.4). In all lakes studied, increased As concentrations in shallow sediment are associated with increases in the concentration of other elements enriched in the ore body or used in processing and refining (Zn, Pb, Ag, Au, Hg, Sb, W, and Sr) (Figure 2.2, 2.3).

In lakes proximal to the former Tundra Mine site, peaks in As concentration occur just below the sediment-water interface (SWI) (Figure 2.2, 2.3). Sediment As concentrations in the shallow sediment of

Powder Mag and Bulldog lakes demonstrate profiles similar to S. Sediment concentrations of S in shallow, mining-impacted sediment demonstrate maximum concentrations at 2 to 3 cm and are 1 to 2 orders of magnitude higher in Bulldog Lake (median: 0.58 %; range: 0.35 to 1.02 %; n = 7) and Powder

Mag Lake (median: 1.28 %; range: 1.06 to 1.29 %; n = 3) compared to lakes farther from the former mine site (Figure 2.2). Conversely, in more distal lakes, peaks in As concentration occur at the SWI and have profiles similar to Fe and Mn (Figure 2.3). In both Control Lake and Matthews Lake, concentrations of

As, Fe, and Mn have maxima at the SWI and gradually decrease in the near-surface sediment.

In Hambone Lake, sediment As concentrations are highest in the two samples collected from the northeastern end of the lake (Figure 2.1). Concentrations of other elements enriched in tailings and used in mineral processing (Ag, Au, Hg, Pb, W, Zn) also follow this trend (Table 2.2).

Table 2.2 Sediment concentrations of selected elements in grab samples collected from Hambone Lake; the highest concentrations are bolded for each element. Sampling locations presented on Figure 2.1. Sample ID Units HAM-1 HAM-2 HAM-3 HAM-4 HAM-5 HAM-6 Distance from discharge m 540 515 310 230 125 250 Water Depth m 0.70 0.90 1.80 1.90 1.70 1.00 As mg·kg-1 622 576 304 195 302 79.6 Ag mg·kg-1 469 497 247 138 373 69.0 Au µg·kg-1 3,750 3,730 2,400 210 2,381 380 Cu mg·kg-1 85.9 96.5 72.3 57.5 83.0 29.5 Hg µg·kg-1 104 87.0 86.0 50.0 84.0 18.0 Mn mg·kg-1 122 133 105 110 112 96 Ni mg·kg-1 97.3 123 85.1 73.2 83.7 53.1 Pb mg·kg-1 8.12 8.56 5.03 3.42 5.81 2.74 S % 1.20 1.80 0.99 0.97 1.02 0.21 Sb mg·kg-1 0.90 1.25 1.02 0.38 0.7 0.34 W mg·kg-1 4.60 4.30 2.30 1.30 2.40 0.80 Zn mg·kg-1 143 165 96.0 80.5 108 75.7

2.5.5 Porewater geochemistry

Arsenic porewater concentrations in near-surface sediment are highly variable, ranging from 0.75 to 383 µg L-1 (median: 13.5 µg·L-1; n = 22; Figure 2.4). Dissolved As concentrations in the mining- impacted sediment of Powder Mag Lake (range: 29 to 383 µg·L-1) are an order of magnitude higher than those observed in other lakes sampled.

The concentrations and distributions of dissolved Fe, Mn, and S in porewaters vary between lakes studied as does the relationship between these elements and porewater As concentrations (Figure 2.4).

The distribution of dissolved As in porewaters in Bulldog and Control Lakes is directly related to the redox behavior of Mn, Fe and S. Conversely, in Matthews Lake, the concentrations of all three elements

(Fe, Mn, S) increase at depths in porewaters just below the SWI and no variations in dissolved As are observed in the shallow sediment. Low porewater volumes in deeper sediment of Matthews Lake (below

20 cm) precluded elemental analysis. Profiles of dissolved Fe, Mn, and S are notably different in Powder

Mag Lake. No surface enrichment of these elements is observed; however, increases in Fe, Mn, S, and Zn concentrations occur with depth in the sediment column (n = 30) (Figure 2.4).

The pH of porewaters from both mining-impacted (range: 5.3 to 7.5; median: 5.7; n = 21) and background sediment (range: 4.8 to 6.7; median: 5.7; n = 85) range from slightly acidic to circumneutral.

The highest pH values were observed in Powder Mag (range: 5.9 to 7.5) and are attributed to the annual

disposal of alkaline treated tailings effluent along this drainage pathway (pH range: 7.54 to 8.61; Golder

Associates Ltd., 2016). Hambone Lake, located upstream of Powder Mag Lake and also influenced by treated effluent, had high pH values in the water column during past environmental monitoring (range:

6.4 to 8.3; AANDC, 2014; Golder Associates Ltd., 2016). The lowest porewater pH was observed in

Matthews Lake sediments (range: 4.8 to 5.8).

As shown in Figures 2.2 and 2.3, no consistent trend in inorganic As speciation is apparent in sediment porewaters of the lakes studied near Tundra Mine. In Powder Mag, Bulldog and Matthews lakes

As (V) is the predominant aqueous species in porewaters. Conversely, As (III) is the most abundant inorganic species of As in porewater samples analyzed from the Control Lake sediment core. The difference between total As and the sum of As (III) and As (V) is substantial in porewaters of Powder

Mag, Matthews, and Control lakes, suggesting that there may be other As species present (e.g. thioarsenates, organoarsenic species) in addition to inorganic As (Figure 2.4). Further analyses would be required to identify these species.

Figure 2.4 Depth profiles of sediment and dissolved porewater As, Fe, Mn, and S concentrations, porewater As speciation, TOC, and OM fractions S1 and S2 in Powder Mag Lake, Bulldog Lake, Matthews Lake, and Control Lake. Concentrations for each element are plotted on different scales and scales are different for each lake.

2.5.6 Characterization of As-bearing solid phases

Six As-hosting solid phases were identified by combined SEM-based automated mineralogy and

EPMA analysis in the near-surface sediments of Hambone, Powder Mag, Bulldog, Matthews, and Control lakes. The abundance of these minerals varies between lakes and with depth in the sediment as shown by the horizontal bar graphs on Figure 2.6 (Appendix A).

Figure 2.5 Backscatter electron (BSE) images and mean As content (wt. %), determined by EPMA, of the predominant As-hosting solid phases identified in the near-surface sediment of lakes of the Tundra Mine region. a) Arsenopyrite (Bulldog Lake 0 to 1 cm); b) framboidal pyrite (Powder Mag 4 to 5 cm); c) As-sulphide (Powder Mag 3 to 4 cm); d) FeSx/FeO (Powder Mag 3 to 4 cm); e) Fe-(oxy)hydroxide (Matthews Lake 4 to 5 cm); f) Fe-bearing phyllosilicate with intercalated minerals along cleavage planes (HAM-1). Representative EDS spectra for each phase are presented in Appendix A.

Arsenopyrite was identified in the near-surface sediment of lakes proximal to the former mine site

(Figure 2.5). Arsenopyrite grains range in size from 1 to 25 µm, with the largest and most abundant grains present at 5 to 6 cm depth in the Bulldog Lake sediment core. In the near-surface sediment of Bulldog

Lake, FeAsS accounts for up to 40 % of total As. Trace arsenopyrite was also identified in Powder Mag

Lake (20 % of total As) and Hambone Lake (2 % of total As), but no arsenopyrite was observed in the shallow sediment of Matthews Lake or Control Lake.

Authigenic Fe-oxyhydroxides are more prevalent than detrital Fe-oxides in near-surface sediments of all lakes studied. These two phases were differentiated based on grain morphology, using

BSE images; detrital particles have sharper edges and more well-defined grain boundaries than authigenic particles which tend to intergrow with surrounding grains (Figure 2.5). EPMA demonstrates that these phases host an average of 0.50 wt. % As. similar proportions of As (mean = 0.50 wt. %, and 0.60 wt. %, respectively; combined mean = 0.50 wt. %). In the near-surface sediments of Matthews and Control lakes,

As is predominantly hosted by Fe-(oxy)hydroxides, accounting for 98 to 100 % of total As. In the near- surface sediments of Bulldog, Hambone, and Powder Mag lakes, authigenic Fe-(oxy)hydroxides were identified, but account for a smaller percentage of total As (5 to 50 % of total As).

The main host of As near the redox interface of Bulldog Lake and Powder Mag Lake is a fine grained aggregate of poorly crystalline Fe-(oxy)hydroxide and iron monosulphide (FeSx/FeO). SEM-based automated mineralogy and EPMA reveal that this phase is heterogeneous in nature with a wide range of morphologies and chemical compositions. Sulphur content of this phase ranges from 0.25 wt. % to 27 wt.

% (n = 37). Arsenic associated with this phase ranges from < D.L. to 0.50 wt. % (mean = 0.20 wt. %; 20 particles; 27 spots). In the mining-impacted sediment of lakes proximal to the former mine site, this phase accounts for 20 to 65 % of total As (Figure 2.6).

In lakes close to Tundra Mine, a discrete As-sulphide phase was identified by SEM-based automated mineralogy. This phase is distinguishable from FeAsS as the Fe content is markedly lower, as determined through EDS spectra (Appendix A). The small grain size precludes the use of EPMA to definitively identify this phase as realgar (As4S4) or orpiment (As2S3). This phase is most abundant in the mining-impacted sediment of Powder Mag Lake, accounting for up to 50 % of the total As (Figure 2.5,

2.6).

Arsenic-bearing framboidal pyrite was identified in all sediments studied. Electron microprobe analysis of 26 particles (27 spots) demonstrates an average of 0.20 wt. % As. In lakes of the Tundra mine region, this phase accounts for 0.10 to 25 % of total As in the near-surface sediment (Figure 2.6).

Iron-bearing phyllosilicates (illite, biotite, and chlorite) with intercalated mineralization along cleavage planes were also detected through SEM-based automated mineralogy in all samples (Figure 2.5). Electron microprobe analysis demonstrates that in lakes proximal to the former mine site, this phase contains elevated S and As. Sulfur content of this phase ranges from 0.1 to 4.2 wt. % (mean = 1.6 wt. %). The presence of trace S associated with phase suggests that traces of sulphide minerals may be forming between the phyllosilicate sheets, leading to the sequestration of As through the formation of discrete precipitates and/or adsorption at the layer edges. EDS analysis of selected mineral grains suggest that the composition of this discrete mineral phase is FeS; however, a more comprehensive investigation would be required to draw conclusions regarding the average chemical composition of this phase (Appendix A). In the near-surface sediment of Powder Mag, Bulldog, and Hambone lakes the As content of this phase ranges from < D.L. to 0.20 wt. % (mean = 0.10 wt. %; n = 46), accounting for 1 to 5 % of total As (Figure

2.6).

Figure 2.6 Relative contribution of each As-hosting solid phase to total As concentrations in the top 15 cm of sediment samples from Hambone, Powder Mag, Bulldog, Matthews and Control lakes. Sediment grab samples collected from Hambone Lake are plotted as distance from effluent discharge location and not relative to depth in the sediment column.

2.5.7 Sediment textural and organic characteristics

All lake sediment samples are dominated by silt (< 63 µm) sized particles (median: 71%; range:

52% to 78%; n = 160) (Figure 2.2, 2.3). In all lakes, with the exception of Matthews Lake, an increase in particle size is observed near the sediment surface (Figure 2.2, 2.3). Particle size in the sediments of

Powder Mag Lake and Bulldog Lake is highly variable throughout the sediment column, demonstrating short intervals of increased sand content (Figure 2.2, 2.3). Total organic carbon content of the sediment ranges from 7.17 to 21.5 % (median = 11.6 %; n = 173) (Figure 2.4). In shallow, impacted sediment, TOC concentrations are highest in Control Lake (17.2 wt. %), Hambone Lake (15.5 wt %) and Powder Mag

Lake (21.5 wt. %). The majority of OM in all sediment samples is S2 (median = 3.17 wt. %; range 1.80 to

5.44 wt. %; n = 172), however, concentrations are different between lakes studied (Appendix A).

Increases in OM concentrations are observed at the SWI of most lakes sampled, with the exception of

Matthews Lake, which has lower sediment OM in near-surface sediment (Figure 2.4).

2.5.8 Statistical analysis of controls on As distribution

Spearman’s Rank Correlation analysis shows that the associations between As and other elements, sedimentary particle size, and OM are different in the near-surface sediment of each of the study lakes (Figure 2.7). Sample size for the near-surface sediment (P–MI) of Powder Mag Lake was not large enough (n < 4) to perform this analysis. In Hambone Lake, As is highly positively (rs ≥ 0.70) and significantly (p ≤ 0.05; n = 6) correlated to other elements enriched in the ore or used in processing at

Tundra Mine (Au, Sb, W) (Figure 2.7). Although also proximal to the former Tundra Mine site, these relationships are not observed in the near-surface sediment of Bulldog Lake (Figure 2.7). Positive (rs ≥

0.80) and significant (p ≤ 0.05) correlations between As and redox-sensitive elements (Fe, Mn, S) are observed in the near-surface sediments of Matthews Lake (n = 13) and Control Lake (n = 14).

Correlations in concentrations of As and S1 OM are evident in the mining-impacted sediment of

Matthews Lake and Hambone Lake but the relationship between As and S1 OM is non-significant (p ≥

0.05) in near-surface sediment of Control Lake (n = 14) and Bulldog Lake (n = 7) (Figure 2.7). The relationships between TOC and OM fractions (S1, S2, S3) are different in the near-surface sediment of all 56

lakes studied (Figure 2.7). The association (rs) between variables is stronger when lakes are assessed as individual populations. This increase in statistical significance may be indicative of differing natural processes contributing to the mobility of As in each of these systems and highlights the importance of examining distinct geochemical populations to determine relevant associations. Significant correlations between As and grain size exist in certain geochemical populations in Powder Mag, Matthews and

Control lakes; however, no consistent relationship was observed.

2.6 Discussion

2.6.1 Delineating mining impacts in lake sediment

Dating of modern sediments suggests that increases in solid phase As concentrations in the shallow sediment of each lake (Figure 2.2, 2.3) correspond approximately to the initiation of exploration and mine development activities in 1946 (Silke, 2009). The sediment geochemistry of lakes near the former Tundra and Salmita mines has been altered through disposal of waste rock and its use in mine site construction, tailings overtopping and seepage, deposition of airborne dust, and/or disposal of mine water and treated tailings effluent. However, mining impacts extend to different depths in the sediment of lakes in the Tundra Mine region (Figure 2.2, 2.3). Cluster analysis and radiometric dating shows distinct changes in the loading of elements enriched in the mined ore and waste rock in the upper 4 to 14 cm of sediment cores from all lakes studied (Figure 2.2, 2.3).

Figure 2.7 Spearman’s Rank correlations of inorganic and organic geochemical parameters for the near-surface sediment of Bulldog Lake (B-MI) (n = 7), Hambone Lake (HL-MI) (n = 6), Matthews Lake (ML-MI) (n = 13) and Control Lake (CL-I) (n = 14) sediment cores. Organic fractions include TOC, S1, S2, and S3. Data shown as Spearman’s Rank (rs) correlation coefficients. Italicized rs values represent p ≤ 0.001, non-italicized values represent p ≤ 0.05; only correlations with rs ≥ 0.6 are included.

2.6.2 Sources of As in near-surface sediment

Increased concentrations of As in surficial sediments of lakes of the former Tundra and Salmita mines region are attributed to the increased natural input of arsenopyrite and secondary As-bearing hosts released during weathering of bedrock and surficial material and to anthropogenic activities associated with mining and mineral processing. Five sources of As to lakes of the Tundra Mine region have been identified including: direct disposal and weathering of waste rock, tailings overtopping and seepage, discharge of treated tailings effluent, weathering and airborne deposition of tailings and waste rock, and natural weathering of mineralized bedrock (Figure 2.8).

Hambone Lake

Seepage and overtopping of tailings effluent (winter historical average (2007 to

2010): 2,830 µg·L-1 As), and the disposal of treated tailings effluent (average (2009 to 2013): 79 µg·L-1

As) have provided the primary sources of As to the sediment and surface water of Hambone Lake.

Arsenopyrite-bearing waste rock used in construction of the original tailings dam and mine facilities and windblown tailings are the most likely sources of trace arsenopyrite observed in the near-surface sediment of Hambone Lake (Figure 2.8). A positive (rs ≥ 0.8) and significant (p ≤ 0.05) relationship between sediment As concentrations and W, Au (Figure 2.7) and Zn (rs = 0.9; p ≤ 0.05; n = 6) also suggest that waste rock and tailings provide the primary sources of As to this lake. In mineralized veins of the Tundra

Mine region, there is a strong correlation between As and Sb in Au-bearing arsenopyrite (Figure 2.7;

Adam, 2016). A significant (rs ≥ 0.8; p ≤ 0.05) relationship between solid-phase As and Sb concentrations is evident in Hambone Lake, suggesting that arsenopyrite provides a primary source of As to this lake.

However, the abundance of As-bearing authigenic sulphides indicate that diagenetic remobilization and active sequestration of As is occurring in the near-surface sediment (Figure 2.6, 2.8). Concentrations of

As, Zn, Ag, and Cu are highest in sediment at the northeastern end of the lake, farthest from the treated effluent discharge location (Figure 2.1; Table 2.2). The increased concentration of elements, which have historically exceeded CCME guidelines in effluent discharge, suggest that this area may provide an ongoing source of downstream contamination (Table 2.2; AECOM, 2015). Relatively oxidized effluent 59

water may have oxidized authigenic As-bearing sulphides, or an increase in pH may have decreased the sorption of arsenate on Fe-(oxy)hydroxides due to competitive sorption (Jeong et al., 2010; Campbell and

Nordstrom, 2014; Desbarats et al., 2015; Golder Associates Ltd., 2017). In addition, the overtopping and seepage of alkaline tailings waters (pH = 8.22) and disposal of treated effluent (pH = 8.4 to 8.5) into

Hambone Lake has increased the pH of surface waters in this lake (average pH = 7.35 to 7.92; AANDC,

2014; AECOM, 2015). As the post-depositional mobility of As is governed by redox conditions and pH of sediment porewaters, any change in geochemistry resulting from tailings seepage and discharge of treated tailings effluent may influence the long-term stability of legacy contaminants (Smedley and

Kinniburgh, 2002).

Figure 2.8 Schematic diagrams showing the predominant As sources, relative concentration of aqueous As species, and relative contribution of each As-hosting solid phase to total As concentrations in near-surface sediment samples from lakes of the Tundra Mine region. Arrows denote the predicted influence of increased OM on As mobility in each lake. Vertical scales are not meant to be representative of the actual bathymetric features of each lake.

Powder Mag Lake

Surface water and porewater As concentrations in Powder Mag Lake are an order of magnitude higher than the other lakes studied, reflecting the transport of dissolved and particulate-bound As species from the TCA through Hambone Lake to the downstream environment (Table 2.1; Figure 2.1, 2.2). The oxidation and dissolution of secondary As-minerals due to the disposal of treated tailings effluent may also have contributed to increased dissolved As concentrations in Powder Mag Lake. Trace arsenopyrite occurs in sediment deposited after the onset of mining activities, but is much less abundant than observed in Bulldog Lake sediments. Within Powder Mag Lake, the main As-hosts are authigenic Fe-

(oxy)hydroxides and Fe-sulphides (Figure 2.6, 2.8). The relative absence of arsenopyrite, abundance of secondary As minerals, and prevalence of oxidized aqueous As species in the near-surface sediment of

Powder Mag Lake (Figure 2.4, 2.6) are attributed to tailings seepage and discharge of treated effluent into this drainage pathway. Oxidation of arsenopyrite in the tailings necessitated treatment with ferric sulphate to precipitate As-bearing Fe-(oxy)hydroxides during remediation activities (WESA Technologies Inc.,

2014). This precipitate was separated by dewatering using Geotubes ®; however, the formation of suspended or colloidal As-bearing Fe-(oxy)hydroxides in the effluent discharge has contributed to the transport of As along this drainage pathway. Anoxic wetlands have been demonstrated to be an effective trap of As (Langner et al., 2011); therefore, the peatland between Hambone and Powder Mag Lake may also provide and long-term sink and/or source of mining-derived As and influence its mobility to the downstream environment. In sediment pre-dating mining activities, As concentrations exceed CCME

Probable Effects Level (PEL) of 17 mg·kg-1 and interim sediment quality guideline (ISQG) of 5.9 mg kg-1

(range: 42 to 70 mg·kg-1; median = 60 mg·kg−1; n = 12); however, fall within the global average range for greywacke and slate which surround Powder Mag Lake (0.5 to 143 mg·kg−1; Figure 2.1; Boyle and

Jonasson, 1973; Smedley and Kinniburgh, 2002).

Bulldog Lake

The greatest abundance of arsenopyrite identified in sediment of the Tundra Mine region is in

Bulldog Lake, accounting for 1 to 40 % of total As in these shallow sediments (Figure 2.6). The largest 62

and most abundant grains are present at 5 to 6 cm depth in the Bulldog Lake sediment column, and this depth coincides temporally with the onset of surface exploration activities in 1946 at the former Tundra

Mine. The presence of arsenopyrite grains, ranging in size from 1 to 25 µm, suggests the disposal and/or weathering of As-bearing waste rock is the primary source of As to this lake. Bulldog Lake is adjacent to the former mill pad (Figure 2.1), which was constructed using waste rock (Hatfield Consulting, 1982;

Lorax Environmental, 2007). The higher sedimentation rates observed at the surface of Bulldog Lake

(0.16 cm·yr-1) are attributed to the increased activity and associated terrain disturbance on-site during recent remediation activities. In the early days of mining, water pumped from the underground workings was discharged into Bulldog Lake (Hatfield Consulting, 1982) and may have provided an additional source of dissolved As, which has since become incorporated into the lake sediment in secondary As- bearing minerals. Dissolution and surface oxidation of arsenopyrite grains is apparent in the shallow

Bulldog Lake sediment, providing a source of dissolved As to the porewaters (Figure 2.4). A significant

(rs ≥ 0.8; p ≤ 0.05) relationship between solid-phase As and Sb concentrations is also evident in the mining-impacted sediment of Bulldog Lake, suggesting that arsenopyrite is a source of both of these elements. Low concentrations of As in pre-mining sediment (median = 34 mg·kg−1; n = 17) suggest that the adjacent volcanic lithologies and mineralized shear vein have had little influence on As loading and mobility in this lake (Figure 2.1).

Matthews Lake and Control Lake

Arsenic concentrations in the near-surface sediment of both Matthews and Control lakes exceed both the CCME PEL and ISQG (CCME, 2001a). No arsenopyrite was identified in the near-surface sediments of these lakes. Arsenic concentrations in the shallow sediments of both lakes are similar to those observed in sediment pre-dating mining activities. Elevated As concentration at the SWI in both of these lakes is attributed primarily to reductive dissolution of naturally-occurring, As-bearing Fe-

(oxy)hydroxides during burial, and re-precipitation near the SWI (Figure 2.3; Martin and Pedersen, 2002).

Increasing concentrations of As, Au, Sb, and W in the sediment cores of Matthews and Control lakes correspond to the onset of advanced exploration activities around 1946, indicating that runoff and dust 63

from the former waste rock and ore piles from both Salmita and Tundra mine sites may have provided sources of mining-related contamination to these lakes (Hatfield Consulting, 1982). Higher sedimentation rates (1.75 yr·cm-1) and an increase in grain size in sediments of Control Lake coincide temporally with runway construction and increased development activity at the Tundra Mine site at ca. 1946 (Figure 2.3).

An increase in grain size is not observed in the other lake sediment cores and therefore may be the result of erosion from the nearby airstrip and roadways (Figure 2.1) or natural changes within the catchment of

Control Lake (Figure 2.3). Grain size of particles in the near-surface sediment of Control Lake ranges from 27 to 173 µm (arithmetic mean), suggesting that some of this coarser grained material may be derived from the mine site, mine roads, and/or the airstrip and transported short distances by aerial suspension (Livingstone and Warren, 1996; Petavratzi et al., 2005).

In both of these lakes, authigenic As-bearing Fe-(oxy)hydroxide is the most abundant As host suggesting that the weathering and oxidation of As-bearing iron sulphides in mine wastes or mineralized bedrock may contribute dissolved Fe, Mn and As to these lakes (Figure 2.6, 2.8; Corkhill and Vaughan,

2009). Matthews Lake is primarily bound by felsic volcanics (rhyolite and dacite) whereas Control Lake is bound by metasedimentary (greywacke and slate) lithologies; global average As content for these bedrock types ranges from 3.2 to 13.4 mg·kg-1 and 0.5 to 143 mg·kg−1, respectively (Smedley and

Kinniburgh, 2002). Higher than global crustal average concentrations of As in pre-mining sediment of

Matthews Lake may be attributed to As associated with Au-mineralization along the eastern shore of this lake or to the weathering of other bedrock types and derived surficial materials in the region. The presence of trace arsenopyrite in pre-mining sediment of both Control and Matthews Lakes also confirms the presence of a geogenic source of As and suggests local-scale variability in geochemical baselines in sediments of lakes in the CLGB.

2.6.3 Intra-lake variance in As mobility and implications of climate warming on the long-term stability of As

The source of As, distance from the mine site, sedimentation rates, and sediment geochemistry

(organic and inorganic) control the depth and magnitude of impacts in lake sediments of the Tundra Mine 64

region. These and other parameters also influence the speciation and post-depositional mobility of As in sediment. These factors vary for each lake studied, demonstrating that the mechanisms influencing the distribution of As can be highly variable, even on the scale of a single mining district. As a result, the effects of past and future climate change on the long-term stability of mining-related As contamination is expected to differ for each lake.

Post-depositional mobility of As in most lacustrine environments is largely governed by dissolution of Fe-(oxy)hydroxides and the synchronous co-precipitation with poorly crystalline Fe-

(oxy)hydroxides and sulphides (Moore et al., 1988; Rickard and Luther, 1997; Nickson et al., 2000;

O’Day et al., 2004). An influx of reactive, labile organic carbon (S1 and S2) to lake sediments, expected as a result of climate warming (Smol et al., 2005; Frey and McClelland, 2009; Prowse et al., 2011; Stern et al., 2012), may influence these redox-mediated mechanisms and thus the long-term stability of As in lake systems. The source of As and type of organic material are likely to be important in determining the magnitude of As mobility associated with organic loading (Nickson et al., 2000; Redman et al., 2002;

Bauer and Blodau, 2006; Sharma et al., 2010; Galloway et al., 2017). Due to the association of As with metal oxides and sulphides in near-surface sediment, changes to redox conditions may significantly influence the distribution and mobility of As in lakes of the Tundra Mine region (Figure 2.8; Martin and

Pedersen, 2002; Couture et al., 2010).

The highest concentrations of As in surface waters and sediments are observed in lakes proximal to the mine site and impacted by the direct disposal of mine wastes (Hambone, Powder Mag, and Bulldog lakes); however, As concentrations and solid-phase speciation vary between these three lakes. The highest sediment As concentrations are observed in sediments of Bulldog Lake, whereas surface water and mean porewater concentrations in Powder Mag Lake and Hambone Lake (surface water average (summer

2016): 98.7 µg·L-1) are higher than in Bulldog Lake (Table 2.1; Figure 2.2; AANDC, 2014). Powder Mag

Lake and Bulldog Lake both demonstrate multiple peaks in porewater As concentrations in the near- surface sediment. One peak occurs at the SWI with subsequent increases just below this boundary, suggesting remobilization of As is occurring in both of these lakes (Figure 2.2). However, increased 65

porewater concentrations in the near-surface sediment of Powder Mag Lake suggest that As is being actively remobilized in sediment to a larger degree than observed in Bulldog Lake.

Total surface water concentration of As in Powder Mag Lake is lower than total As concentrations in treated effluent (average As (2009-2014): 75.3 µg·L-1) and surface waters of Hambone

Lake (average As: 35 to 40 µg·L-1), suggesting that As is being removed from overlying surface water and settling into the bottom sediments (AANDC, 2013; AECOM, 2015). Additionally, the low As concentrations in surface water and sediment of lakes further downstream of Powder Mag Lake indicate that As is being sequestered in the sediment of this lake (AECOM, 2015). Colloidal and suspended As- bearing Fe-(oxy)hydroxides are deposited to the surface sediment of Powder Mag Lake where they are reductively dissolved; As is then sequestered through the precipitation of authigenic sulphides (Figure

2.8). The rapid onset of reducing conditions in the sediment column of Powder Mag Lake likely results from the accumulation of labile OM derived from the peatland connecting Powder Mag to Hambone Lake

(Figure 2.1, 2.2). In general, reducing conditions are thought to increase the mobility of As due to the reductive dissolution of Fe-(oxy)hydroxides and the transformation of As (V) to As (III) (Smedley and

Kinniburgh, 2002; Martin and Pedersen, 2002; Dixit and Hering, 2003; Andrade et al., 2010). However, a positive relationship between labile OM (S1) and sediment As concentrations in modern sediment impacted by mining activities in sub-Arctic Canada was recently observed by Galloway et al. (2017), who suggested that increased labile OM may provide a substrate for microbial growth that in turn mediates the precipitation of As-bearing secondary minerals (i.e. framboidal pyrite, As-sulphides).

Organic matter concentrations in surface water (DOC) and near-surface sediment (TOC, S1, S2) of Powder Mag Lake are higher than other lakes studied (Table 2.1, Appendix A). This is attributed to the influence of seasonal effluent discharge to Hambone Lake (mean = 254,500 m3·yr-1; median: 180,000 m3·yr-1; n = 7) that increased the flow of water through the peatland area into Powder Mag Lake and mobilized organic material (Figure 2.1) (INAC, 2005, 2008, 2010; AANDC, 2012, 2013, 2014). The abundance of As-bearing sulphides in the near-surface sediment of Powder Mag Lake suggest that sediments enriched in S, Fe and natural OM may be efficient geochemical traps for As and support the 66

mechanisms of As attenuation observed by Drahota et al. (2009) in organic-rich groundwaters, and

Langner et al. (2011) in wetlands. Burton et al. (2013) demonstrated that microbial sulfate reduction can result in changes in As speciation and highlighted the important role of thioarsenate species on the mobility of As in mining-impacted environments. The presence of these complexes may account for the abundance of AsR species in the porewaters of Powder Mag Lake (Figure 2.4) and contribute to the mobility of As; however, further porewater speciation analyses would be required to test this hypothesis

(Couture et al., 2013).

A positive (rs) relationship between S1 and sediment As concentrations occurs in two of the five lakes studied (Hambone Lake (rs = 0.88; p ≤ 0.05; n = 6) and Matthews Lake (rs = 0.68; p ≤ 0.05; n = 13)

(Figure 2.7); however, no relationship between Assed and labile OM is observed in Bulldog Lake or

Control Lake. Increases in reactive OM input to lakes may lead to higher rates of sulfide production and promote the onset of reducing conditions by increasing sediment oxygen demand (Wilkin and Barnes,

1997; Toevs et al., 2006). The influx of OM facilitates the microbially mediated reduction of both Fe (III)

-2 and SO4 , leading to the formation of As-bearing sulphide minerals in the near-surface sediment (Wilkin and Barnes, 1997; Kirk et al., 2004; Lowers et al., 2007; Thamdrup, 2000). As a result, the influx of OM may facilitate the formation of microbially mediated As-bearing sulphide minerals in the near-surface sediment (Nordstrom and Archer, 2003). In lakes proximal to the former mine site, As is primarily hosted in sulphides or authigenic agglomerations of authigenic Fe-(oxy)hydroxide and iron monosulphide, possibly amorphous mackinawite (FeS1-x). It has been demonstrated that amorphous mackinawite may provide an effective mechanism of As attenuation under changing redox conditions in lake sediment

(Vega et al., 2017; An et al., 2017). Increased OM loading to these lakes may, therefore, facilitate the precipitation of As-bearing sulfides and increase the stability of As in sediment; however, the relationships between As, S, and OM are not consistent (Figure 2.7). Previous studies show that OM influences the speciation and mobility of As in peatlands, lake sediment, and soils, through various mechanisms, including competition with As for sorption on mineral surfaces (Redman et al., 2002; Bauer and Blodau, 2006; Wang and Mulligan, 2006; Stuckey et al., 2016; Galloway et al., 2017; Wang et al., 67

2018). However, the interaction of OM with elements is influenced by its characteristics and composition

(Grafe et al., 2002; Ritter et al., 2006; Langner et al., 2011; Burton et al., 2013). In the near-surface sediment of lakes proximal to the former mine site, the relationship between TOC and labile OM fractions

(S1, S2, S3) is different (Figure 2.7). This reflects the natural heterogeneity of sediment OM and suggests that the varied sources of both As and OM influence the speciation and mobility of As in mining- impacted lake sediments of the Tundra Mine region.

In lakes more distal to the former mine site, weathering and dust transport of exposed waste rock used for construction of mine facilities may have influenced As concentrations in the near-surface sediment; however, the predominant source of As to these lakes is likely natural weathering of As-rich bedrock (Figure 2.8). In the near-surface sediment of Matthews and Control lakes, median As concentrations are lower than lakes closer to Tundra Mine and As in the near-surface sediment is primarily hosted by Fe-(oxy)hydroxides (Figure 2.6, 2.8). In sediment cores from both Matthews Lake and Control Lake, a significant relationship between sediment As, Fe, and Mn indicates the important role of Fe-(oxy)hydroxides in the sequestration of As at the SWI (Figure 2.7). Increased OM input to these lakes may result in enhanced diffusion of As from sediment to surface waters by promoting reducing conditions and thus drive reductive dissolution of As-bearing Fe-(oxy)hydroxides (Nickson et al., 2000;

Harvey et al., 2002; Islam et al., 2004). However, the presence of As-bearing framboids in the near- surface sediment of both Matthews Lake and Control Lake suggests that the precipitation of authigenic pyrite may also sequester As released from reductive dissolution of As-bearing Fe-(oxy)hydroxides

(Figure 2.8). A significant correlation between labile OM (S1), S (rs = 0.91; p ≤ 0.001; n = 13) and As (rs

= 0.67; p ≤ 0.05; n = 13) in near-surface sediments of Matthews Lake may represent this mechanism of

As attenuation (Figure 2.7; Galloway et al., 2017).

The results of this study support Galloway et al. (2017), who suggested that in lakes impacted by

As from legacy gold mining in sub-Arctic Canada, OM may facilitate the precipitation of As-bearing sulfides and increase the stability of As in sediment. In contrast, this study highlights that in lakes where

As is hosted mainly by Fe-(oxy)hydroxides, increasing OM may enhance the diffusion of As from sediment to surface water.

2.7 Conclusions

The sediment geochemistry of lakes near a former gold mine site in sub-Arctic Canada has been altered due to the influence of historical mining, mineral processing, and subsequent terrain disturbance associated with remediation. Five main sources have contributed As to lakes around the former Tundra

Mine: disposal of waste rock, tailings overtopping and seepage, weathering and airborne deposition of waste rock and tailings, discharge of treated tailings effluent, and natural weathering of mineralized bedrock. Impacts were observed in each of the lakes studied through analysis of sediment, surface water, and porewaters. The depth of impacts in the sediment column and primary controls on the post- depositional mobility of mining-derived As contamination are highly variable between lakes.

The findings of this study suggest that benthic grab samples, commonly collected during environmental site assessments, may not provide an accurate assessment of mining impacts as changes in geochemistry occur on a sub-centimeter scale and deposition rates vary between lakes. As a result, interpretation of sediment geochemistry and mining impact based only on grab samples with uncertain depth control complicates the accurate determination of geochemical baselines. Sediment gravity cores can more precisely capture the uppermost sediments so that they may be sub-sampled at higher resolution (e.g. 0.5 to 1.0 cm). This study highlights that sampling depth control is essential for the interpretation of sediment geochemistry and differentiation of mining impact in lacustrine systems. This is especially important in sub-Arctic lakes where sedimentation rates are low.

The distribution of As in these lakes is influenced by lake-specific variables including: the source of As, Fe and S concentrations in sediment, depth of the lake, sedimentation rate, and OM type. This study also demonstrates variability in pre-mining background As concentrations in sediment cores and provides evidence for the influence of bedrock weathering on the natural loading of As to lakes in this region.

Labile OM (S1 as determined by Rock Eval pyrolysis) demonstrated significant positive relationships to sedimentary As and S concentrations in two of the five lakes studied. The relationship between S1, As, and S, in addition to the abundance of As-bearing framboidal pyrite in mining-impacted sediment, suggests that S1 may mediate the authigenic precipitation of As-bearing sulphides. Differing relationships between near-surface sediment As concentrations and OM in lakes of the Tundra Mine region may be explained by the natural heterogeneity of sediment OM. The findings of this study suggest that increasing OM loading to lakes expected with continued climate warming will have varying effects on As mobility and fate in mining-impacted lakes and that these effects depend on the main hosts of As in lake sediments. In lakes where As is hosted in reduced sulphide form, increasing OM input may increase the stability of As in sediment, whereas in lakes where the primary host of As is Fe-(oxy)hydroxides, increasing OM to these lakes may enhance the diffusion of As from sediment to surface waters. These results have implications for long-term monitoring of lakes impacted by legacy contamination.

Knowledge from this study can aid in the development of robust geochemical baselines and be used to improve environmental monitoring and remediation strategies at future northern metal mines.

2.8 Acknowledgements

This project is jointly funded by Polar Knowledge Canada (Project # 1519-149, awarded to J.M.

Galloway (GSC Calgary) and T.R. Patterson (Carleton University)), the Environmental Geoscience

Program, Natural Resources Canada (Metal Mining Project, M. B. Parsons; Northern Baselines Activity;

J.M. Galloway), a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery

Grant (H. E. Jamieson), and NSERC Northern Research Supplement (Jamieson). Funding for the project has also been contributed by funds awarded to C.B. Miller by the NSERC Create Mine of Knowledge and the Northern Scientific Training Program and substantial in-kind support from CIRNAC. The authors would like to extend thanks to Brian Cummings, Christopher Grooms, Agatha Dobosz, and Brian Joy of

Queen's University for their assistance and invaluable guidance. A special thank you to Braden Gregory

(Carleton University), Andrew Macumber (Carleton University), and Hendrik Falck (Northwest

Territories Geological Survey) for facilitating logistics and providing relentless optimism. A warm thank you to CIRNAC, Delta Engineering, and Nahanni Construction for their hospitality and support for our northern field work initiatives. This is Contribution Number 20180420 of the Lands and Minerals Sector,

Natural Resources Canada.

2.9 References

2AANDC, 2012. Water Surveillance Network Program (SNP) Report for Type ‘A’ Water Licence MV2009L8-0008 (Tundra Mine Remediation) June 2012 Water Quality Results. http://registry.mvlwb.ca/Documents/MV2009L8-0008/MV2009L8-0008%20-%20AANDC- CARD%20-%20Tundra%20Mine%20Remediation%20- %20June%202012%20SNP%20Report%20-%20Jul30-12.pdf

Adam, M.A., 2016. Gold Mineralisation at the FAT Deposit , Courageous Lake Greenstone Belt, Northwest Territories, Canada (Masters Thesis). University of Western Ontario. https://ir.lib.uwo.ca/cgi/viewcontent.cgi?article=5245&context=etd

AECOM, 2018. Tundra Mine Adaptive Management Plan. http://registry.mvlwb.ca/Documents/MV2016L8-0003/MV2016L8-0003 – INAC CARD – Adaptive Management Plan – Revision from May 24-18 - Oct26-18.pdf

An, X.L., Huang, F.G., Ren, H.T., Wang, Y.F., Chen, Y., Liu, Z.M., Zhang, H.W., Han, X., 2017. Oxidative dissolution of amorphous FeS and speciation of secondary Fe minerals: Effects of pH and As(III) concentration. Chem. Geol. 462, 44–54. https://doi.org/10.1016/j.chemgeo.2017.04.025

Appleby, P.G., 2001. Chronostratigraphic techniques in recent sediments, in: Last, W.M., Smol, J.P. (Eds.), Tracking Environmental Change Using Lake Sediments Volume 1: Basin Analysis, Coring, and Chronological Techniques. Kluwer, Dordrecht, pp. 171–203.

2 INAC, AANDC, CIRNAC are all acronyms for the Federal Department responsible for the Government of Canada's work in northern Canada. 72

Ardakani, O.H., Sanei, H., Snowdon, L.R., Outridge, P.M., Obermajer, M., Stewart, R., Vandenberg, R., Boyce, K., 2016. The accepted values for the internal Geological Survey of Canada (GSC) 9107 Rock-Eval 6 standard (Upper Cretaceous Second White Speckled Shale, Colorado Group), western Canada Open File 8043. https://doi.org/10.4095/298729

Azcue, J.M., Mudroch, A., Rosa, F., Hall, G., 1994. Effects of abandoned gold mine tailings on the arsenic concentrations in water and sediments of Jack of Clubs Lake, B.C. Environ. Technol. (United Kingdom) 15, 669–678. https://doi.org/10.1080/09593339409385472

Barrett, P.M., Hull, E.A., Burkart, K., Hargrave, O., Mclean, J., Taylor, V.F., Jackson, B.P., Gawel, J.E., Neumann, R.B., 2019. Contrasting arsenic cycling in strongly and weakly stratified contaminated lakes : Evidence for temperature control on sediment – water arsenic fluxes. https://doi.org/10.1002/lno.11119

Blais, J.M., Kalff, J., 1995. The influence of lake morphometry on sediment focusing. Limnol. Oceanogr. 40, 582–588. https://doi.org/10.4319/lo.1995.40.3.0582

Blott, S.J., Pye, K., 2001. Gradistat: A grain size distribution and statistics package for the analysis of unconsolidated sediments. Earth Surf. Process. Landforms 26, 1237–1248. https://doi.org/10.1002/esp.261

Boyle, R.W., Jonasson, I.R., 1973. The geochemistry of As and its use as an indicator element in geochemical prospecting. J. Geochem. Explor. 2, 251–296. http://doi.org/0.1016/0375- 6742(73)90003-4

Buckwalter-Davis, M., 2013. Automated Mineral Analysis of Mine Waste. Queen's Univeristy (MSc. Thesis). https://qspace.library.queensu.ca/bitstream/handle/1974/8200/Buckwalter- Davis_Martha_J_201308_MSc.pdf?sequence=1&isAllowed=y Burton, E.D., Johnston, S.G., Planer-Friedrich, B., 2013. Coupling of arsenic mobility to sulfur transformations during microbial sulfate reduction in the presence and absence of humic acid. Chem. Geol. 343, 12–24. https://doi.org/10.1016/j.chemgeo.2013.02.005

Campbell, K.M., Nordstrom, D.K., 2014. Arsenic speciation and sorption in natural environments. Rev. Mineral. Geochemistry 79, 185–216. https://doi.org/http://dx.doi.org/10.2138/rmg.2014.79.3

CCME, 2001a. Canadian Sediment Quality Guidelines for the Protection of Aquatic Life, Environmental Protection. https://ccme.ca/en/resources/canadian_environmental_quality_guidelines/

CCME, 2001b. CCME Water Quality Guidelines for the Protection of Aquatic Life. https://ccme.ca/en/resources/canadian_environmental_quality_guidelines/

Corkhill, C.L., Vaughan, D.J., 2009. Arsenopyrite oxidation - A review. Appl. Geochemistry 24, 2342– 2361. https://doi.org/10.1016/j.apgeochem.2009.09.008

Couture, R.M., Gobeil, C., Tessier, A., 2010. Arsenic, iron and sulfur co-diagenesis in lake sediments. Geochim. Cosmochim. Acta 74, 1238–1255. https://doi.org/10.1016/j.gca.2009.11.028

Couture, R.M., Rose, J., Kumar, N., Mitchell, K., Wallschläger, D., Van Cappellen, P., 2013. Sorption of arsenite, arsenate, and thioarsenates to iron oxides and iron sulfides: A kinetic and spectroscopic investigation. Environ. Sci. Technol. 47, 5652–5659. https://doi.org/10.1021/es3049724

Crann, C.A., Patterson, R.T., Macumber, A.L., Galloway, J.M., Roe, H.M., Blaauw, M., Swindles, G.T., Falck, H., 2015. Sediment accumulation rates in subarctic lakes: Insights into age-depth modeling from 22 dated lake records from the Northwest Territories, Canada. Quat. Geochronol. 27, 131–144. https://doi.org/10.1016/j.quageo.2015.02.001

Davis, J., 2011. Statistics and data analysis in geology 3rd Edition. John Wiley & Sons, Ltd.

Desbarats, A.J., Parsons, M.B., Percival, J.B., 2015. Arsenic mobility in mildly alkaline drainage from an orogenic lode gold deposit, Bralorne mine, British Columbia. Appl. Geochemistry 57, 45–54. https://doi.org/10.1016/j.apgeochem.2014.11.015

Dixit, S., Hering, J.G., 2003. Comparison of arsenic (V) and arsenic (III) sorption onto iron oxide minerals: implications for arsenic mobility. Environ. Sci. Technol. 37, 4182–4189. https://doi.org/10.1021/es030309t

Donato, S. V., Reinhardt, E.G., Boyce, J.I., Pilarczyk, J.E., Jupp, B.P., 2009. Particle-size distribution of inferred tsunami deposits in Sur Lagoon, Sultanate of Oman. Mar. Geol. 257, 54–64. https://doi.org/10.1016/j.margeo.2008.10.012

Drahota, P., Rohovec, J., Filippi, M., Mihaljevic, M., Rychlovsky, P., Cerveny, V., Pertold, Z., 2009. Mineralogical and geochemical controls of arsenic speciation and mobility under different redox conditions in soil, sediment and water at the Mokrsko-West gold deposit, Czech Republic. Sci. Total Environ. 407:3372–3384. http://doi.org/ 10.1016/j.scitotenv.2009.01.009

Environment Canada, 2012. Metal Mining Technical Guidance for EEM. http://ec.gc.ca/esee- eem/AEC7C481-D66F-4B9B-BA08-A5DC960CDE5E/COM-1434---Tec-Guide-for-Metal-Mining- Env-Effects-Monitoring_En_02[1].pdf

Fandrich, R., Gu, Y., Burrows, D., Moeller, K., 2007. Modern SEM-based mineral liberation analysis. Int. J. Miner. Process. 84, 310–320. https://doi.org/10.1016/j.minpro.2006.07.018

Folinsbee, R.E., Moore, J.C.G., 1950. Matthews Lake, Northwest Territories, Paper 50-2. https://doi.org/https://doi.org/10.4095/108448

Folinsbee, R.E., Moore, J.C.G., 1955. Matthews Lake, District of Mackenzie, Northwest Territories. Geological Survey of Canada, "A" Series Map 1024A. https://doi.org/10.4095/107491

Frey, K.E., McClelland, J.W., 2009. Impacts of permafrost degradation on arctic river biochemistry. J. Glaciol. 23, 169–182. https://doi.org/10.1002/hyp

Galloway, J.M., Swindles, G.T., Jamieson, H.E., Palmer, M., Parsons, M.B., Sanei, H., Macumber, A.L., Timothy Patterson, R., Falck, H., 2017. Organic matter control on the distribution of arsenic in lake sediments impacted by ~ 65 years of gold ore processing in subarctic Canada. Sci. Total Environ. 622–623, 1668–1679. https://doi.org/10.1016/j.scitotenv.2017.10.048

Gasiorowski, M., 2008. Deposition rate of lake sediments under different alternative stable states. Geochronometria 32, 29–35. https://doi.org/10.2478/v10003-008-0020-y

Golder Associates, L., 2005. Terrestrial and Aquatic Field Surveys and Ecological Risk Assessment of the Tundra Gold Mine. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

Golder Associates Ltd., 2008. Phase II Remedial Action Plan: Tundra Mine Site. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

Golder Associates Ltd., 2016. Public Works and Government Services Canada Tundra Mine, NWT Remediation Environmental Monitoring Program: Water Quality: 2015 and Historical Results.

Golder Associates Ltd., 2017. Public Works and Government Services Canada Tundra Mine, NWT Remediation Environmental Monitoring Program: 2016 and Historical Water Quality Results.

Goldfarb, R.J., Berger, B.R., Klein, T.L., Pickthorn, W.J., Klein, D.P., 1995. Low-sulfide Au quartz veins. In: E.A. du Bray (Ed.), Preliminary Compilation of Descriptive Geoenvironmental Mineral Deposit Models. U.S. Geol. Surv. Open-File Report 95–831, Denver CO, pp. 261–267 (Chapter 34).

Government of the Northwest Territories, Executive and Indigenous Affairs. 2007. Interim Land Withdrawal - Akaitcho Dene First Nations. https://www.eia.gov.nt.ca/sites/eia/files/adfn_interim_land_withdrawal_map_crownland.pdf

Grafe, M., Eick, M.J., Grossl, P.R., Saunders, A.M., 2002. Adsorption of arsenate and arsenite on ferrihydrite in the presence and absence of dissolved organic carbon. J. Environ. Qual. 31, 1115. https://doi.org/10.2134/jeq2002.1115

Gu, Y., 2003. Automated scanning electron microscope based mineral liberation analysis an introduction to JKMRC/FEI Mineral Liberation Analyser. J. Miner. Mater. Charact. Eng. 02, 33–41. https://doi.org/10.4236/jmmce.2003.21003

Harvey C. F., Swartz C. H., Badruzaman A. B. M., Keon-Blute N., Yu W., Ali M. A., Jay J., Bechie R., Niedan V., Brabander D., Oates M. F., Ashfaque K. N., Islam S., Hemond H. F., and Ahmed M. F. 2002. Arsenic mobility and groundwater extraction in Bangladesh. Science. 298, 1602–1606.

Hasegawa, H., Rahman, M.A., Kitahara, K., Itaya, Y., Maki, T., Ueda, K., 2010. Seasonal changes of arsenic speciation in lake waters in relation to eutrophication. Sci. Total Environ. 408, 1684–1690. https://doi.org/10.1016/j.scitotenv.2009.11.062

Hatfield Consulting, 1982. Initial Environmental Evaluation for Giant Yellowknife Mines - Salmita Project. Unpublished report prepared for Giant Yellowknife Mines Ltd., Yellowknife, NT. pp. 1 - 24.

Henderson, J., 1970. Stratigraphy of the Yellowknife Supergroup, Yellowknife Bay - Prosperous Lake area, District of MacKenzie. https://doi.org/https://doi.org/10.4095/100647

Hollibaugh, J.T., Carini, S., Gürleyük, H., Jellison, R., Joye, S.B., LeCleir, G., Meile, C., Vasquez, L., Wallschläger, D., 2005. Arsenic speciation in Mono Lake, California: Response to seasonal stratification and anoxia. Geochim. Cosmochim. Acta 69, 1925–1937. https://doi.org/10.1016/j.gca.2004.10.011

INAC, 2010. Tundra Mine , NWT Remediation Environmental Monitoring Program: Water Quality Part H: 2009 and Historic Results. http://registry.mvlwb.ca/Documents/MV2009L8-0008/MV2009L8- 0008%20-%202010%20Annual%20Report%20-%20Mar31-11.pdf

Islam F. S., Gault A. G., Boothman C., Polya D. A., Charnock J. M., Chatterjee D., and Lloyd J. 2004. Role of metal-reducing bacteria in arsenic release from Bengal Delta sediments. Nature. 430, 68–71.

Jamieson, H.E., Maitland, K.M., Oliver, J.T., Palmer, M.J., 2017. Regional distribution of arsenic in near- surface soils in the Yellowknife area. NWT Open File 2017-03.

Jeong, H.Y., Han, Y.S., Park, S.W., Hayes, K.F., 2010. Aerobic oxidation of mackinawite (FeS) and its environmental implication for arsenic mobilization. Geochim. Cosmochim. Acta 74, 3182–3198. https://doi.org/10.1016/j.gca.2010.03.012

Juggins, S., 2017. Analysis of Quaternary Science Data, Package ‘ rioja ' (http://cran.r- project.org/package= rioja)

Kirk, M.F., Holm, T.R., Park, J., Jin, Q., Sanford, R.A., Fouke, B.W., Bethke, C.M., 2004. Bacterial sulfate reduction limits natural arsenic contamination in groundwater. Geology 32, 953–956.

Langner, P., Mikutta, C., Kretzschmar, R., 2011. Arsenic sequestration by organic sulphur in peat. Nat. Geosci. 5, 66–73. https://doi.org/10.1038/ngeo1329

Livingstone, I., Warren, A., 1996. Aeolian Geomorphology: An Introduction. Harlow: Longman, pp.1-63.

Lorax Environmental, 2007. Tundra Mine Geochemical Assessment. Pp. 1 – 105. Internal report.

Lowers, H.A., Breit, G.N., Foster, A.L., Whitney, J., Yount, J., Uddin, Md N., Muneem, Ad A., 2007. Arsenic incorporation into authigenic pyrite, Bengal Basin sediment, Bangladesh. Geochim. Cosmochim. Acto 71 (11), 2699 – 2717. https://doi.org/10.1016/j.gca.2007.03.022

Lynch, J., 1990. Provisional elemental values for eight new lake sediment and stream sediment reference materials LKSD-1, LKSD-2, LKSD-3, LKSD-4, STSD-1, STSD-2, STSD-3 and STSD-4. Geostand. Newsl. 14, 153–167. https://doi.org/10.1111/j.1751-908X.1990.tb00070.x

Macdonald, R., Harner, T., Fyfe, J., 2005. Recent climate change in the Arctic and its impact on contaminant pathways and interpretation of temporal trend data. Sci Total Env. 342, 5–86. https://doi.org/10.1016/j.scitotenv.2004.12.059

Macumber, A.L., Patterson, R.T., Neville, L.A., Falck, H., 2011. A sledge microtome for high resolution subsampling of freeze cores. J. Paleolimnol. 45, 307–310. https://doi.org/10.1007/s10933-010-9487- 4

Macumber, A.L., 2015. Grain-size analysis and Arcellinida as tools for inferring paleoclimatological variability : a case study from Holocene Canadian Subarctic lake sediments. PhD. Thesis. Carleton University. https://curve.carleton.ca/system/files/etd/622aaee1-0a90-4077-bdf7- 838298d14c0e/etd_pdf/083ece3a2a5ee991a0e74ef0fa67858e/macumber- grainsizeanalysisandarcellinidaastoolsfor.pdf

Martin, A.J., Pedersen, T.F., 2002. Seasonal and interannual mobility of arsenic in a lake impacted by metal mining. Environ. Sci. Technol. 36, 1516–1523. https://doi.org/10.1021/es0108537

Moore, J.C.G., 1951. Courageous Lake; Geological Survey of Canada Paper 51-14. https://doi.org/https://doi.org/10.4095/108378

Moore, J.W., 1978. Biological and water quality surveys at potential mines in the Northwest Territories. III. Salmita Gold Property, Mathews Lake. Environment Canada, Environmental Protection Service, Manuscr. Rep. NW-78-7: 24 p. + Appendices. Obtained from DFO Library, Yellowknife, NT.

Moore, J.N., Ficklin, W.H., Johns, C., 1988. Partitioning of arsenic and metals in reducing sulfidic sediments. Environ. Sci. Technol. 22, 432–437. https://doi.org/10.1021/es00169a011

Mudroch, A., Joshi, S.R., Sutherland, D., Mudroch, P., Dickson, K.M., 1989. Geochemistry of sediments in the back bay and Yellowknife Bay of the Great Slave Lake. Environ. Geol. Water Sci. 14, 35–42. https://doi.org/10.1007/BF01740583

Murray, M., 2002. Is laser particle size determination possible for carbonate rich lake sediments? J. Paleolimnol. 27, 173–183

Nickson, R.T., Mcarthur, J.M., Ravenscroft, P., Burgess, W.G., Ahmed, K.M., 2000. Mechanism of arsenic release to groundwater, Bangladesh and West Bengal. Appl. Geochemistry 15, 403–413. https://doi.org/10.1016/S0883-2927(99)00086-4

Nordstrom, D. K. and Archer, D. G. (2003) in Arsenic in Goundwater, eds. Welch, A. H. & Stollenwerk, K. G. (Kluwer, Boston), pp. 1–25

O’Day, P.A., Vlassopoulos, D., Root, R., Rivera, N., 2004. The influence of sulfur and iron on dissolved arsenic concentrations in the shallow subsurface under changing redox conditions. Proc. Natl. Acad. Sci. U. S. A. 101, 13703–13708. https://doi.org/10.1073/pnas.0402775101

Outridge, P., Sanei, H., Goodarzi, F., 2007. Evidence for control of mercury accumulation rates in Canadian high arctic lake sediments by variations of aquatic primary productivity. Environ. Sci. Technol. 41, 5259–5265. https://doi.org/10.1021/es070408x

Outridge, P.M., Sanei, H., Stern, G.A., Goodsite, M., Hamilton, P.B., Carrie, J., Goodarzi, F., Macdonald, R.W., 2011. Response to comment on climate change and mercury accumulation in Canadian high and subarctic lakes. Environ. Sci. Technol. 45, 6705–6706. https://doi.org/10.1007/s10533-010- 9543-1.(8)

Parsons, M.B., LeBlanc, K.W.G., Hall, G.E.M., Sangster, A.L., Vaive, J., E, Pelchat, P., 2012. Environmental geochemistry of tailings, sediments and surface waters collected from 14 historical gold mining districts in Nova Scotia. Geological Survey of Canada, Open File 7150. https://doi.org/10.4095/291923

Parsons, M.B., Jamieson, H.E., Miller, C.B., Galloway, J.M., 2019. Choosing an appropriate digestion protocol for environmental risk assessments of mineralized earth materials; Geological Survey of Canada, Scientific Presentation 97, 1 poster. https://doi.org/10.4095/314499

Petavratzi, E., Kingman, S., Lowndes, I., 2005. Particulates from mining operations: A review of sources, effects and regulations. Miner. Eng. 18, 1183–1199. https://doi.org/10.1016/j.mineng.2005.06.017

Planer-Friedrich, B., Wallschläger, D., 2009. A critical investigation of hydride generation-based arsenic speciation in sulfidic waters. Environ. Sci. Technol. 43, 5007–5013. https://doi.org/10.1021/es900111z

Prowse, T., Alfredsen, K., Beltaos, S., Bonsal, B., Duguay, C., Korhola, A., McNamara, J., Pienitz, R., Vincent, W.F., Vuglinsky, V., Weyhenmeyer, G.A., 2011. Past and future changes in arctic lake and river ice. Ambio 40, 53–62. https://doi.org/10.1007/s13280-011-0216-7

PS Analytical. 2018. Available online : https://www.psanalytical.com/

R Core Team. 2014. R: A Language and Environment for Statistical Computing. http://www.r-project.org

Ransom, A., Robb, M., 1986. The Salmita gold deposit, Courageous Lake, Northwest Territories. In: L.A. Clark and D.R. Francis (Editors), Gold in the Western Shield. Can. Inst. Min. Met., 38: 285-305

Redman, A.D., Macalady, D.L., Ahmann, D., 2002. Natural organic matter affects Arsenic speciation and sorption onto hematite. Environ. Sci. Technol. 36, 2889–2896. https://doi.org/10.1021/es0112801

Reimann, C., Filzmoser, P., Garrett, R., Dutter, R., 2008. Statistical Data Analysis Explained: Applied Environmental Statistics with R. John Wiley & Sons, Ltd.

Rickard, D., Luther, G.W., 1997. Kinetics of pyrite formation by the H2S oxidation of iron (II) monosulfide in aqueous solutions between 25 and 125 degrees C: The mechanism. Geochim. Cosmochim. Acta 61, 135–147. https://doi.org/doi:10.1016/S0016-7037(96)00321-3

Ritter, K., Aiken, G.R., Ranville, J.F., Bauer, M., Macalady, D.L., 2006. Evidence for the aquatic binding of arsenate by natural organic matter-suspended Fe (III). Environ. Sci. Technol. 40, 5380–5387. https://doi.org/10.1021/es0519334

Salminen, R., Tarvainen, T., 1997. The problem of defining geochemical baselines. A case study of selected elements and geological materials in Finland. J. Geochemical Explor. 60, 91–98. https://doi.org/10.1016/S0375-6742(97)00028-9

Sanei, H., Stasiuk, L.D., Goodarzi, F., 2005. Petrological changes occurring in organic matter from recent lacustrine sediments during thermal alteration by Rock-Eval pyrolysis. Org. Geochem. 36, 1190 – 1203. http://doi.org/10.1016/j.orggeochem.2005.02.009

Sanei, H., Goodarzi, F. 2006. Relationship between organic matter and mercury in recent lake sediment: The physical–geochemical aspects. Appl. Geochemistry 21, 1900-1912. http://10.1016/j.apgeochem.2006.08.015

Sanei, H., Outridge, P.M., Dallimore, A., Hamilton, P.B., 2012. Mercury-organic matter relationships in pre-pollution sediments of thermokarst lakes from the Mackenzie River Delta, Canada: The role of depositional environment. Biogeochemistry 107, 149–164. https://doi.org/10.1007/s10533-010- 9543-1

Schelske, C.L., Peplow, A., Brenner, M., Spencer, C.N., 1994. Low-background gamma counting: applications for 210Pb dating of sediments. J. Paleolimnol. 10, 115–128. https://doi.org/10.1007/BF00682508

Schuh, C.E., Jamieson, H.E., Palmer, M.J., Martin, A.J., Blais, J.M., 2019. Controls governing the spatial distribution of sediment arsenic concentrations and solid-phase speciation in a lake impacted by legacy mining pollution. Sci. Total Environ. 654, 563–575. https://doi.org/10.1016/j.scitotenv.2018.11.065

Seabridge Gold Inc., 2010. Courageous Lake Project Description in Support of a Class “ A ” Land Use Permit (Former Permit MV2003C0050). http://registry.mvlwb.ca/Documents/MV2010C0046/MV2010C0046%20- %20Project%20Description%20-%20Oct06-10.pdf

SENES Consultants Ltd., 2006. Tundra Mine Site Ecological Risk Assessment of Water Management. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

SENES Consultants Ltd., 2008. Remediation completion report for Tundra Mine Phase I (2007) Restoration Program Tundra Mine, NT. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

SENES Consultants Ltd., 2011. Development of Tundra Mine construction monitoring, long-term monitoring, and status of the environment programs. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

Sharma, P., Rolle, M., Kocar, B.D., Fendorf, S., Kapppler, A., 2010. Influence of Natural Organic Matter on As Transport and Retention. Environ. Sci. Technol. 45, 546–553. https://doi.org/10.2134/jeq2003.1393

Silke, R., 2009. The Operational History of Mines in the Northwest Territories, Canada, pp. 487 – 500. http://miningnorth.com/_rsc/site-content/library/NWT_Mines_History_RSilke2009.pdf

Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochemistry 17, 517–568. https://doi.org/10.1016/S0883-2927(02)00018-5

Stuckey, J.W., Schaefer, M. V., Kocar, B.D., Benner, S.G., Fendorf, S., (2016). Arsenic release metabolically limited to permanently water-saturated soil in Mekong Delta. Nat. Geosci. 9, 70–76. https://doi.org/10.1038/ngeo2589

Tetra Tech Wardrop, 2012. Courageous Lake Prefeasibility Study. https://secure.kaiserresearch.com/i/jk/tr16/TRSEA20120905.pdf

Thamdrup B. 2000. Bacterial Manganese and Iron Reduction in Aquatic Sediments. In: Schink B. (eds) Advances in Microbial Ecology. Advances in Microbial Ecology, vol 16. Springer, Boston, MA

Thompson, P.H., Kerswill, J.A., 1994. Preliminary Geology of the Winter Lake - Lac de Gras Area, District of MacKenzie, Northwest Territories, Geological Survey of Canada, Open File 2740 (revised), scale 1:250000. https://geoscan.nrcan.gc.ca/starweb/geoscan/servlet.starweb?path=geoscan/downloade.web&search1 =R=194794

Thornton, I., 1996 Sources and pathways of arsenic in the geochemical environment: health implications, in Environmental Geochemistry and Health (eds J.D. Fuge and R. McCall), Geological Society Special Publication No 113, pp. 153–61.

URS, 2005. Geochemical Assessment of Acid Rock Drainage and Metal Leaching Potential of Tailings and Waste Rock, Tundra Mine, NWT. Internal Report.

van Hengstum, P.J., Reinhardt, E.G., Boyce, J.I., Clark, C., 2007. Changing sedimentation patterns due to historical land-use change in Frenchman’s Bay, Pickering, Canada: Evidence from high-resolution textural analysis. J. Paleolimnol. 37, 603–618. https://doi.org/10.1007/s10933-006-9057-y

Vega, A.S., Planer-Friedrich, B., Pastén, P.A., 2017. Arsenite and arsenate immobilization by preformed and concurrently formed disordered mackinawite (FeS). Chem. Geol. 475, 62–75. https://doi.org/10.1016/j.chemgeo.2017.10.032

WESA Technologies Inc., 2014. Water Treatment Monitoring and Discharge Plan 2014 (Monitoring Plan: MV2009L8-0008)

Wang, S., Mulligan, C.N., 2006. Effect of natural organic matter on arsenic release from soils and sediments into groundwater. Environ. Geochem. Health 28, 197–214. https://doi.org/10.1007/s10653-005-9032-y

Wilkin, R.T., Barnes, H.L. 1997. Formation processes of framboidal pyrite. Geochim. Cosmochim. Acta 61, 323-339. https://doi.org/10.1016/S0016-7037(96)00320-1

Chapter 3

Influence of late-Holocene climate change on the solid-phase speciation and

long-term stability of arsenic in sub-Arctic lake sediments

3.1 Abstract

Sediment cores were collected from two lakes in the Courageous Lake Greenstone Belt (CLGB), central Northwest Territories, Canada, to examine the influence of late-Holocene warming on the transport and fate of arsenic (As) in sub-Arctic lakes. In both lakes, allochthonous As-bearing minerals

(i.e. arsenopyrite and scorodite) were identified in sediment deposited during times of both regional warming and cooling, suggesting that weathering of bedrock and derived surficial materials provides a continual source of As to lakes of the CLGB. However, maximum porewater As (84 µg·L-1 and 15 µg·L-1) and reactive organic matter (OM; aquatic and terrestrial-derived) concentrations in each lake are coincident with known periods of regional climate warming. It is inferred that increased biological production in surface waters and influx of terrigenous OM led to the release of sedimentary As to porewater through reductive dissolution of As-bearing Fe-(oxy)hydroxides and scorodite during episodes of regional warming. Elevated sedimentary As concentrations (median: 36 mg·kg-1; range: 29 to 49 mg·kg-1) are observed in sediment coeval with the Holocene Thermal Maximum (ca. 5430 ± 110 to 4070

± 130 cal. yrs. BP); at these depths, authigenic As-bearing framboidal pyrite is the primary host of As in sediment and the influence of organic matter on the precipitation of As-bearing framboidal pyrite is apparent petrographically. These findings suggest that increased biological productivity and weathering of terrestrial OM associated with climate warming influences redox cycles in the near-surface sediment and enhances the mobility of As in northern lakes. Knowledge generated from this study is relevant for predicting future climate change-driven variations in metal(loid) cycling in aquatic systems and can be used to interpret trends in long-term environmental monitoring data at historical, modern, and future metal mines in northern environments.

3.2 Introduction

Global climate warming is disproportionately affecting high northern latitudes, and in particular, northwestern Canada (ACIA, 2004). Temperatures are predicted to continue rising in the northern hemisphere at more than double the global rate, and to increase by 2 to 6 C by the late 21st century

(Zhang et al., 2019). Rising temperatures are expected to affect the biogeochemistry of lake systems and impact the loading and cycling of metal(loid)s (Pienitz et al., 1999; Macdonald et al., 2005; Smol et al.,

2005; Axford et al., 2009; Rydberg et al., 2010; Galloway et al., 2017; Outridge et al., 2017). Across

Canada, natural ranges of metal concentrations in surface waters and sediment are used to guide environmental monitoring activities and the development of remediation objectives at active, abandoned, and proposed mine sites (CCME, 2001a, b). In mineralized regions where concentrations of naturally occurring metal(loids) are commonly above national guidelines, understanding the transport and fate of elements and drivers of chemical change(s) is especially relevant. Improved knowledge of climate-related processes that influence the mobility of metal(loid)s will help guide the development of geochemical baselines, inform environmental assessments and monitoring activities, and facilitate sustainable development of mineral resources in northern Canada.

The Slave Geological Province (SGP), Northwest Territories (NT), Canada, hosts numerous legacy metal mines and remains prospective for future development (Seabridge Gold Inc., 2010, 2018;

Tetra Tech Wardrop, 2012; Terrax Minerals, 2019). In the SGP, gold (Au)-mineralization occurs in orogenic Au deposits in quartz veins, which are commonly also enriched with arsenopyrite (FeAsS)

(Goldfarb et al., 1995; Groves et al., 1998). Due to the association of arsenic (As)-bearing sulphides with these deposits, As concentrations in sediment and surface waters of some lakes are naturally elevated above Canadian guidelines (CCME, 2001 a,b; Galloway et al., 2012, 2015, 2017). Additionally, mining and processing of Au has resulted in increased As concentrations in environmental media surrounding modern and historical mine sites (DeSisto et al., 2011; Craw and Bowell, 2014; Desbarats et al., 2015).

Lakes play an important role in the storage and mobilization of As and other elements (La Force et al.,

2000; Nordstrom, 2002; Smedley and Kinniburgh, 2002; Andrade et al. 2010; Schuh et al., 2018; Miller et al., 2019; Palmer et al., 2019). Differences in the type and quality of organic matter (OM) are important in metal speciation and transport in environmental systems (Pickhardt et al., 2002; Jiang et al., 2011;

Carrie et al., 2012; Sanei et al., 2012; Galloway et al., 2017; Outridge et al., 2017). Recent studies have suggested that OM, expected to increase with climate warming and associated effects such as permafrost thaw (Guo et al., 2007; Vonk et al., 2013, 2015) and increased productivity in lakes (Prowse et al., 2006;

2011; Smol and Douglas, 2007), may be particularly important for controlling the mobility and fate of As in both natural and mining-impacted northern aquatic ecosystems (MacDonald et al., 2005; Galloway et al., 2017).

In near-surface sediments, As mobility is primarily driven by sorption and co-precipitation reactions with authigenic Fe/Mn-(oxy)hydroxides and sulphides (Smedley and Kinniburgh, 2002; Bowell et al., 2014). An increased flux of OM to near-surface sediments may affect these and other mechanisms of As sequestration by driving changes in redox conditions and influencing the stability of As-bearing minerals (Martin and Pedersen, 2002; Couture et al., 2010), enhancing competitive adsorption (Grafe et al., 2001, 2002), facilitating microbially-mediated precipitation of As-bearing minerals (Bostick and

Fendorf, 2003; Kirk et al., 2004, 2010; Wang et al., 2018), and driving reduction of aqueous As species

(Redman et al., 2002; Bauer and Blodau, 2006; Du Laing et al., 2009). Galloway et al. (2017) and Miller et al. (2019) demonstrated the influence of reactive OM on the cycling and sequestration of As in mining- impacted lakes. However, differentiating between the cumulative effects of resource development and climate change in modern lakes remains a challenge. Redistribution and sediment-surface enrichment of

As due to the alteration of the organic geochemistry of aquatic sediments by enhanced organic carbon fluxes may occur in both mining-impacted and natural systems (Macdonald et al., 2005). An improved understanding of the effects of climate change on lake biogeochemistry is needed to more accurately predict the impacts of future climate variations on As mobility in northern ecosystems.

Lake sediments provide an archive that can be used to examine the effects of past climate variability on the biogeochemistry of lacustrine systems and serve as an analogue for current and projected warming. The central NT is an ideal location to examine past analogues of warming as it experienced profound landscape change, including tree line movement of up to 150 km, over the late-

Holocene (~ ca. 7,000 yrs. BP) that affected the physical and chemical properties of lake sediments

(Sulphur et al., 2016). This region, therefore, offers a unique opportunity to examine the influence of various types of OM on As concentrations and speciation. Using paleolimnological and geochemical methods applied to lake sediment cores, this study examines the influence of late-Holocene (ca. 5430 ±

110 to 400 ± 120 cal. BP) climate change on As loading and speciation to generate new knowledge that can be used to improve predictions of the impact of 21st century climate change on As mobility.

3.3 Study Area

Matthews Lake (64.06014 N, 111.23566 W) and Control Lake (64.07771 N, 111.13493 W; unofficial name) are located approximately 240 km NE of Yellowknife, NT (Figure 3.1). These lakes are

430 m above sea level and within the Lockhart River drainage basin, which has a catchment area of approximately 26,600 km2 and discharges to Great Slave Lake (Topographical Survey of Canada, 1925).

Matthews Lake has a surface area of 10.7 km2 and a maximum depth of 12 m; Control Lake has a surface area of 2.2 km2 and a maximum depth of 4.0 m (Figure 3.1; Golder Associates Ltd., 2005). Located 80 km north of the present-day tree line, between the tundra shield high and low Arctic ecoregions, the physiography of the region is characterized by many small lakes and the typical low relief topography of the Canadian tundra (Ecosystem Classification Group, 2008; 2012). Organic deposits and peat occur in depressions and along creek valley and drainage channel bottoms (Seabridge Gold Inc., 2010). The area is in the transition zone between discontinuous and continuous permafrost; median active-layer thicknesses is seasonally and spatially variable, ranging in thickness from 37 to 150 cm (URS, 2005; Karunaratne,

2011; AECOM, 2015). Modern climate is characterized as relatively cold and dry with long winters and short summers; average annual temperature ranges from -12 to -9 C (SENES Consultants Ltd., 2006).

Upiter et al. (2014) reconstructed a mean July air temperature of 12.8 °C (ca. 1942 to 2011) at Carleton

Lake (unofficial name), 55 km NE of Tundra Mine (Figure 3.1). Lake ice in the region is on average 1.5 to 2 m thick with break-up occurring in early to mid-June and freeze-up beginning in mid- to late October

(Golder Associates Ltd. 2017).

The study site is located within the Courageous Lake Greenstone Belt (CLGB), which is associated with the northwest-trending Yellowknife Supergroup of the Central Slave Province (McGlynn and Henderson, 1970; Henderson, 1970, 1985). Regionally, arsenopyrite is the predominant As-bearing mineral in bedrock and is primarily associated with Au mineralization within hydrothermal quartz veins at the contact between Archean volcanics and argillaceous sediments (Figure 3.1). Arsenopyrite is also found as a trace component within altered mafic volcanics and slate (Ransom and Robb, 1986). Matthews

Lake is bound by felsic and mafic volcanics of the CLGB, while bedrock underlying Control Lake consists of granitoid and metasedimentary lithologies (Figure 3.1). Locally, glacial till ranges from 2 to

10 m in thickness; bedrock striations indicate an eastern to southeastern orientation of glacial advance

(Geological Survey of Canada, 2014).

3.4 Methods and Materials

3.4.1 Sampling locations and sample preparation

Sediment cores were collected in March 2016 from the frozen lake surface using a Glew gravity corer and transparent polycarbonate core tubes (7.5 cm (inner diameter) x 60 cm) (Glew et al., 2001).

Samples were taken from the deepest locations in Matthews (10.4 m) and Control (3.7 m) lakes, based on previous bathymetric mapping by Golder Associates Ltd. (2005) and Hatfield Consulting (1982) (Figure

3.1). Sediments from each core were extruded at 1-cm resolution in a high purity (99.998 %) nitrogen

(N2)-filled glove bag immediately following collection at the coring site.

Samples were shipped to the Geological Survey of Canada – Atlantic Division (GSC-A) in

Dartmouth, Nova Scotia, for porewater extraction. Subsequent elemental and As speciation analyses of all water samples were completed at the GSC’s Inorganic Geochemical Research Laboratory in Ottawa and

the Université de Montréal, respectively (for further details, see Miller et al., 2019). At Queen’s

University, aliquots of sediment were freeze-dried at 1.0 Pa and -50 °C for grain size determinations (0.5 to 1 g), organic geochemical (0.2 g) and inorganic geochemical analyses (0.5 g), and radiometric dating.

A separate aliquot was dried in a high purity N2-filled glove bag to preserve solid-phase As speciation for scanning electron microscope (SEM)-based automated mineralogy, electron probe microanalysis (EPMA) and bulk X-ray Absorption Near-Edge Structure (XANES) analysis (Huang and Ilgen, 2006).

Figure 3.1 A) Map showing location of Tundra Mine (red star) in relation to the City of Yellowknife and other nearby paleolimnological records in the central Northwest Territories (all lake names unofficial): McMaster Lake (MacDonald et al., 1993), Queen’s Lake (Pienitz et al., 1999), Toronto Lake (MacDonald et al., 1993), Waite Lake and Danny’s Lake (Sulphur et al., 2016), and Carleton Lake (Upiter et al., 2014). The transition from boreal forest to tundra is approximated with darker shading (modified after Upiter et al., 2014 and Sulphur et al., 2016). B) Bedrock geology and sampling locations in the CLGB (Folinsbee and Moore, 1955; Thompson and Kerswill, 1994). 89

3.4.2 Age-depth model and stratigraphy

Near-surface sediments from both cores were dated using gamma spectrometry at the

Paleoecological Environmental Assessment Research Laboratory (PEARL, Queen’s University) using an

Ortec High Purity Germanium Gamma Spectrometer (Cooke et al., 2012; Appendix B). Sediment ages were calculated using the constant rate of supply model (Appleby, 2001). Bulk sediment (Matthews Lake n = 6; Control Lake n = 7), humic acid (Matthews Lake n = 3; Control Lake n = 2) and humin fractions

(Matthews Lake n = 3; Control Lake n = 2) were dated using accelerator mass spectrometry (AMS) 14C at the André E. Lalonde AMS Laboratory in Ottawa (Appendix B). Radiocarbon dates were calibrated using the IntCal13 terrestrial radiocarbon calibration curve (Reimer et al., 2013). Age-depth relationships for both cores were modelled using BACON software (version 2.2; Blaauw, 2010; Blaauw and Christen,

2013). Radiocarbon dating is based on the assumption that the carbon in sediment forms in equilibrium with atmospheric 14C (Philippsen, 2013). However, numerous sources of 14C-depleted carbon may accumulate in lake systems (i.e. water rock-interactions with carbonate lithologies, weathering of old terrigenous organic material, thawing of permafrost) and result in anomalously old radiocarbon ages of sediment core samples from lakes. This phenomenon is referred to as the freshwater reservoir effect

(FRE). To estimate the FRE, preliminary age-depth models were generated with only radiocarbon results.

As the sediment-water interface was captured by the cores (visually determined in the field), the preliminary age-depth models were extrapolated to 0 cm depth and the difference between the modelled age at 0 cm and the age of the sediment at the surface (1950 AD; -66 years BP) was assumed to represent a FRE, which is known to affect lakes in the SGP (e.g. Dalton et al., 2018). The age-depth models suggest a FRE of ca. 110 yrs. at Control Lake and an FRE of ca. 230 yrs. at Matthews Lake (see Appendix B for

14C-only age models). These age offsets are similar to the 430-year FRE reported in Dalton et al. (2018).

After adjusting uncalibrated radiocarbon results for the FRE, age-depth models based on both 210Pb and corrected and calibrated 14C dates was generated in BACON using the IntCal13 curve (Figure 3.2; Reimer et al., 2013).

3.4.3 Sedimentary grain size

Sedimentary grain size was determined at Carleton University using a Beckman Coulter LS

13320 laser diffraction particle size analyzer fitted with a Universal Liquid Module over a measurement range between 0.4 and 2000 µm (for further details, see Miller et al., 2019). Results were used in conjunction with the software package GRADISTAT v.8.0 (Blott and Pye, 2001) to calculate grain size statistical parameters. Garnet15 (mean diameter 15 ± 2 μm), an accuracy standard supplied by Beckman

Coulter, was run once per month. To assess instrumental precision, an in-house mud sample (Cushendun

Mud; mean diameter = 20.5 ± 0.76 μm) was run at the start of every session.

3.4.4 Organic matter characterization

Visible reflectance spectroscopy (VRS) was used to determine chlorophyll-a concentrations in lake sediment (Matthews n = 52; Control n = 48). A FOSS NIRSystems Model 6500 series Rapid

Content Analyzer (Tidestone Technologies Inc.) was used to collect VRS of sediment over the range of

400 to 2,500 nm. The preparation of sediment samples, calculation and interpretation of VRS results, and use of reflectance spectroscopy for inferring lake trophic status and evaluation of past primary production is detailed by Das et al. (2005), Michelutti et al. (2005, 2010), and Wolfe et al. (2006).

Programmed pyrolysis (Rock-Eval® 6) was used to determine the amount and type of OM in bulk lake sediments (Matthews Lake n = 52; Control Lake n = 48; Lafargue et al., 1998). This technique can be used to identify different fractions of OM in lake sediments and is described in detail by Carrie et al. (2012), Albrecht et al. (2015), Galloway et al. (2017), and Outridge et al. (2017). In lake sediments, the S1 fraction is predominantly derived from autochthonous OM (e.g. algal-derived lipids; Carrie et al.,

2012) and S2 compounds are derived from the biomacromolecule structure of algal cell walls and other aquatic biological matter (Sanei et al., 2005; Carrie et al., 2012). The S3 portion of organic matter is comprised of terrigenous plant materials in addition to humic and fulvic acids (Carrie et al., 2012;

Albrecht et al., 2015). An influx of terrigenous organic matter is generally characterized by high oxygen index (OI) values while algal-derived organic fractions demonstrate higher hydrogen index (HI) values

(Appendix B; Carrie et al., 2012; Sebag et al., 2013; Outridge et al., 2017). A standard reference material 91

(Internal 9107 shale standard, Geological Survey of Canada – Calgary (GSC-C), Ardakani et al., 2016) was run every 5th sample and demonstrated < 3 % relative standard deviation (RSD) for S1 and S2 and 11

% RSD for S3. The lower precision for S3, and in turn oxygen index (OI), in bulk samples, is expected due to poor peak integration and lack of distinction between S3 derived from OM, carbonates, and siderite in reference samples (Ardakani et al., 2016). A duplicate sample was run every 10th sample; the majority of duplicates demonstrate RSD of < 5 % for all parameters and a mean percent difference (MPD) of < 15

%. Further details regarding sample preparation and methodologies are provided in Appendix B.

Organic petrography was used to optically and qualitatively examine OM at GSC-C. Polished epoxy mounts for eleven samples (Matthews Lake: n = 7; Control Lake: n = 4) were prepared following methods outlined by Reyes et al. (2006). Incident white light and fluorescent light microscopy were conducted using a Zeiss Axioimager II microscope system (50× magnification) equipped with the Diskus-

Fossil system. Fluorescence microscopy was conducted using ultraviolet G 365 nm excitation with a 420 nm filter. Organic matter was classified based on the classification of macerals in coal (ICCP, 1971; Sanei et al., 2005; Reyes et al., 2006).

3.4.5 Sediment and porewater geochemistry

3.4.5.1 Sediment elemental analysis

Elemental analysis of bulk sediment (n = 211) was completed at Acme Analytical Laboratories

(Bureau Veritas), Vancouver, via inductively coupled plasma-mass spectrometry (ICP-MS) (AQ250-EXT package (53 elements)). A modified aqua regia protocol (1:1:1 HCl:HNO3:H2O at 95 °C for one hour) was used to digest the sediments prior to analysis. Near-total (4-acid) digestions were not used for this study, as some key elements of interest (i.e. As, antimony (Sb), mercury (Hg), and sulphur (S)) may be lost through volatilization during high-temperature fuming with hydrofluoric acid (Parsons et al., 2012,

2019).

One field duplicate was submitted for every 10 samples to assess sampling accuracy and precision of commercial ICP-MS analysis. RSD ranged from 0.21 % to 7.22 % and MPD ranged from

0.29 % to 10.21 % for As in duplicate samples. Certified reference materials (STSD-3; Lynch, 1990) and 92

laboratory standard reference materials (STD DS10, STD OREAS45EA) were also included with each sample set to assess analytical accuracy and precision. For STSD-3 the mean measured As concentration was 24.34 ± 0.81 mg·kg-1 (n = 8) vs. an expected concentration of 22 mg·kg−1 for As following aqua regia digestion (RSD 3.3 %). Mean measured As concentration for STD DS10 was 42.3 ± 0.77 mg·kg-1 (n

= 8) vs. an expected concentration of 46.2 mg·kg−1 As following aqua regia digestion.

3.4.5.2 Porewater elemental analysis and dissolved As speciation

Porewater samples were analyzed for a 64-element suite of metal(loid)s at the GSC’s Inorganic

Geochemical Research Laboratory in Ottawa. Analyses of major elements were performed by Inductively

Coupled Plasma-Atomic Emission Spectroscopy (ICP-AES) using a Perkin-Elmer 3000 DV. Trace elements were analyzed using ICP-MS with a Thermo Corporation X-7 Series II. For each sample set, analyses were also completed on field and laboratory duplicates, blanks, and certified standards (SLRS-5 and TMRAIN). For laboratory duplicates, RSD ranged from 0.08 % to 1.40 % and MPD ranged from

0.12 % to 1.98 % for As. Mean percent difference for standard reference materials SLRS-5 and TMRAIN was 4.77 % (n = 2) and 1.18 % (n = 2), respectively. Arsenic was below the detection limit (0.1 µg·L-1) in all field blanks (n = 4). Inorganic As speciation was determined by hydride generation-atomic fluorescence spectrometry (HG-AFS; model PSA 10.055 Millennium Excalibur) at the Université de

Montréal. Standard solutions and certified inter-calibration samples were used to assess accuracy and precision. Trace amounts of As (0.12, 0.16 µg·L-1) were detected in two blanks analyzed for total As, however, As was not detected in the other (n = 6) duplicate blanks. Arsenic speciation by HG-AFS measures arsenite (As (III)) and arsenate (As (V)) species in solution (Planer-Friedrich and Wallschläger,

2009). The difference (AsR) between total As and ∑ As (III) and As (V) may represent organo-arsenic compounds or thiolated As species (Figure 3.3, 3.4; Hollibaugh et al., 2005; Hasegawa et al., 2010;

Campbell and Nordstrom, 2014; Nearing et al., 2014).

3.4.6 Solid phase As speciation analysis

Thirteen N2-dried lake sediment samples were selected for detailed speciation analysis based on bulk elemental composition and position within each core. Polished sections (35-50 μm thickness) were 93

prepared at Vancouver Petrographics. Preparation of these samples followed procedures outlined by Van

Den Berghe et al. (2018) and Schuh et al. (2018). From these samples, intervals with the highest sediment and porewater As concentrations (n = 8) were selected for bulk XANES analysis at the Advanced Photon

Source (APS) at Argonne National Laboratory (Lemont, IL).

3.4.6.1 SEM-based Automated Mineralogy and Electron Microprobe analyses

The relative abundance of common As-bearing minerals in lake sediment was characterized using a combination of backscatter electron (BSE) image analysis and integrated Energy-Dispersive X-ray

Spectroscopy (EDS) (Gu, 2003; Fandrich et al., 2007; Buckwalter-Davis, 2013; Van Den Berghe, 2018).

Sparse phase liberation (SPL) mode with a user-defined BSE greyscale range of 120 to 255 was used to selectively identify As-hosting phases, and therefore does not provide bulk mineralogy information

(Fandrich et al., 2007). Automated mineralogy analysis was completed using a FEI Quanta 650 FEG

ESEM operating under high vacuum following procedures outlined by Van Den Berghe (2018). Elements must form approximately 3 % of the solid phase being analyzed to produce a discernible peak in the EDS spectrum (Pirrie and Rollinson, 2011). Therefore, the As concentrations of solid phases thought to contain trace concentrations of As were analyzed using a JEOL JXA-8230 electron microprobe operating in wavelength dispersive spectroscopy (WDS) mode. The relative contribution of each As-hosting phase to bulk sediment concentrations was calculated using the average As concentration of each phase determined through EPMA analysis and the fraction of each As-bearing species obtained by SEM-based automated mineralogy following the methods of Van Den Berghe et al. (2018).

3.4.6.2 Bulk X-ray Absorption Near-Edge Structure (XANES)

The relative abundance and distribution of As oxidation states present in the lake sediment samples was determined using synchrotron-based bulk XANES analysis at Sector 20-BM of the APS.

Preparation of samples for these analyses followed procedures described by Van Den Berghe et al.

(2018). Many sediment samples had low bulk As concentrations (< 100 mg·kg-1); therefore, a focused 7 mm incident beam was used to maximize fluorescence. All XANES analyses were performed in a cryostat chamber (12 to 55 K) to limit beam-induced changes in As oxidation state. XANES analyses 94

were conducted by scanning across the pre-edge, adsorption, and post-edge regions (11,717 to 12,167 eV). Standard materials were selected to cover the principal oxidation states exhibited by As in lacustrine systems (–1 to +5; Appendix B). Processing of bulk As K-edge XANES data was carried out using

ATHENA (Demeter 0.9.25) software. The raw XANES spectra were normalized, aligned, and merged prior to statistical analysis (Ravel and Newville, 2005). Linear combination fitting (LCF) was performed from 30 eV below the adsorption edge up to 100 eV over the As Kα-edge position for the normalized derivative spectra. Reported percentages of each model compound are absorption-corrected amounts

(post-hoc normalization to 100 %; Appendix B). Goodness-of-fit was assessed based on chi-squared (χ2) values and R-factors. LCF-best fit was also evaluated from the difference between the experimental data and fitted data plotted as a difference curve as well as through visual confirmation of identified phases through SEM-based automated mineralogy. The relative distribution of As-oxidation states was calculated by multiplying bulk As concentration by the fraction of As-species obtained through LCF of As K-edge

XANES data.

3.4.7 Statistical analyses

Zones of distinct geochemical composition were determined for each sediment core using element geochemistry of bulk sediments in the two cores subjected to stratigraphically constrained incremental sum of squares (CONISS) cluster analysis compared against a broken stick model using the rioja package in R (Figure 3.3, 3.4; Grimm, 1987; R Core Team, 2014; Juggins, 2017). Elements with concentrations below instrumental detection limits in 25 % or more of the samples and those that demonstrated little down-core variability were removed prior to statistical analyses. One half of the lower detection limit (LDL) was used for element concentrations below the LDL for retained data (Reimann et al., 2008). No element concentrations exceeded upper detection limits. The 46-element suites of data were log-transformed and converted to standardized values to calculate CONISS clusters (Davis, 2011). Five geochemical zones (ML-G1 through ML-G5) were delineated in the Matthews Lake core and 4 zones were delineated in the Control Lake core (CL-G1 through CL-G4; Figure 3.3, 3.4). The relationships between metal(loid) concentrations and OM fractions in the clusters that pre-date mining activity (ML-G3 95

to ML-G5; CL-G3 and CL-G4) were the focus of statistical analyses for this paper to isolate climate influence on the geochemistry. Anderson Darling (p < 0.05, AD = 0.95) and Shapiro–Wilk (p < 0.0005,

W = 0.72) tests demonstrate a non-normal distribution of As in both Matthews Lake (whole dataset, n =

52) and Control Lake (whole dataset, n = 48), therefore non-parametric tests were used for statistical analyses. Spearman's rank correlation analysis was used to explore the relationship between solid-phase

As concentration in bulk sediment and other geochemical and environmental variables in the stratigraphic clusters.

3.5 Results

3.5.1 Core chronology

Humin and humic OM fractions were extracted from the Control Lake and Matthews Lake sediment cores and dated using AMS 14C to determine if mobile OM accounts for the offset in age-depth models and apparent increase in sedimentation rates (Figure 3.2). The humic acid fractions are consistently younger than the humin fraction and humin dates are within 5 % RPD of the AMS 14C age determinations of bulk OM (Appendix B). This result indicates that there is a source of old carbon in the bulk OM. That source could be carbonates derived from the landscape (mineralized zones, surficial material) or OM that experienced a long residence time on the landscape prior to deposition in the lake sediments (Crann et al., 2015). As a result of the inherent uncertainty of composite age-depth modelling, the dates for this study are treated as general ranges and, wherever possible, supported by secondary lines of evidence.

An age-depth model based on seven 210Pb and twelve 14C dates indicates that the 52-cm-long

Matthew Lake core spans approximately the last 5,400 years (Figure 3.3; Appendix B). A distinct change in sedimentation rate from ca. 0.1 mm·yr-1 to 0.6 mm·yr-1 occurs at 12 cm depth in the core (ca. 640 – 410 cal. yrs. BP; Figure 3.2).

The 48-cm long Control Lake core spans the last ca. 2,110 cal. yrs. BP based on fourteen 210Pb dates and nine 14C dates (Figure 3.4; Appendix B). Similar to the age model for Matthews Lake, an abrupt

shift in sedimentation rate from ca. 0.3 to ca.1 mm·yr-1 occurs at the base of the 210Pb profile between 12 and 18 cm (ca. 1080 – 670 cal. yrs BP to ca. 150 – 65 cal. yrs BP; Figure 3.2).

Figure 3.2 Bayesian age-depth model developed using BACON for (A) Matthews Lake and (B) Control Lake. The age-depth model was constructed using the Clam software package and incorporates 210Pb-derived CRS dates and AMS 14C dates (Blaauw, 2010; Reimer et al., 2013); the gray area represents the 95 % confidence interval for the age-depth model. The top leftmost panel shows that Markov Chain Monte Carlo runs were stable, middle plot denotes the prior (curve) and posterior (filled histogram) distributions for accumulation rate (yr·cm − 1), and the rightmost plot shows the prior (curve) and posterior (filled histogram) for the dependence of accumulation rate between sections. The main plot shows age distributions of calibrated 14C ages and the age-depth model (grey). Dark grey areas indicate more precisely dated sections of the chronology, while lighter grey areas indicate less chronologically secure sections.

3.5.2 Sedimentary grain size

Lake sediments preserved in the Matthews Lake core are composed predominately of silt (< 63

µm; mean = 73.4 %, SD = 15.4 %), with some sand (mean = 11.1 %, SD = 6.5 %) and clay (mean = 12.2

%, SD = 1.8 %) (Appendix B). Grain size distribution was generally consistent downcore, however,

decreases in silt content and corresponding increases in sand content occur in intervals ML-G1 and ML-

G4 (Figure 3.3).

The Control Lake core also contains sediments that are predominantly silt-sized (mean = 71.2 %,

SD = 7.4 %), with some clay (mean = 16.0 %, SD = 4.0 %) and sand (mean = 12.9 %, SD = 10.3 %)

(Appendix B). An increase in particle size is observed in the surface sediment (C-G1; median grain size =

55.3 µm) compared to deeper sediment (CL-G3; median grain size = 19.7 µm) (Figure 3.4).

3.5.3 Organic matter characterization

3.5.3.1 Sediment organic matter composition

Total organic carbon content of the sediment from both cores ranges from 9.6 to 17.2 wt. %

(Matthews Lake: median = 11.4 wt. %; range: 9.6 to 12.6 wt. %, n = 52; Control Lake: median = 13.3 wt.

%; range: 11.2 to 17.2 wt. %; n = 48) (Figure 3.3, 3.4). The highest TOC concentrations in the Matthews

Lake sediment core occur from 18 to 19 cm and correspond with maximum residual carbon (RC) values

(Figure 3.3). In the Control Lake sediment core, the highest concentration of TOC is observed in the near- surface sediment (0 to 5 cm); however, the highest S2 concentrations are observed at 43 to 48 cm (Figure

3.4). The majority of OM in both lakes is composed of S2 OM (Matthew Lake: median = 3.02 wt. %; range 2.29 to 3.53 wt. %; Control Lake: median = 3.37 wt. %; range 2.71 to 4.46 wt. %). In the Matthews

Lake core, an increasing trend in S2 and pyrolysable carbon (PC) is observed between 25 and 48 cm

(Figure 3.3). This is accompanied by an increase in chlorophyll-a from 48 to 52 cm (Figure 3.3). Trends in chlorophyll-a concentrations are not consistent with those of Rock-Eval parameters in Control Lake

(Figure 3.4).

For all bulk sediment samples from both cores, Rock-Eval pyrograms demonstrate a large S1 peak and a bimodal S2 peak. In recent sediments, a bimodal S2 peak was commonly observed in the range of 300 to 650°C. The first peak shoulder is referred to as S2a and is attributed to fresh biological matter, the second peak is known as S2b and is related to kerogen-like OM. Both S2 fractions are part of labile or reactive OM in the sediments (Carrie et al., 2012).

Figure 3.3 Downcore plots of sediment grain size, solid-phase and porewater concentrations of As, S, and Fe; organic matter fractions (S1, S2, S3, and TOC) and dissolved As speciation in the Matthews Lake sediment core. Note changes in scale. Porewater As concentrations are plotted on a log scale. Geochemical populations (ML-G1 to ML-G5) are based on stratigraphically constrained cluster analysis of log-transformed concentrations of 46 elements. Modeled ages incorporate 210Pb-derived CRS dates and AMS 14C dates; dates with a (*) have been calculated from present day (1950 AD) and are reported in AD. 99

Figure 3.4 Downcore plots of sediment grain size, solid-phase and porewater concentrations of As, S, and Fe; organic matter fractions (S1, S2, S3, and TOC) and dissolved As speciation from the Control Lake sediment core. Note changes in scale. Porewater As concentrations are plotted on a log scale to allow for relative increases to be discerned. Geochemical populations (CL-G1 to CL-G4) are based on stratigraphically constrained cluster analysis of log-transformed concentrations of 46 elements. Modeled ages incorporate 210Pb-derived CRS dates and AMS 14C dates; dates with a (*) have been calculated from present day (1950 AD) and are reported in AD. 100

3.5.3.2 Petrographic classification of organic matter

Organic matter in sediment of the Tundra Mine region is derived from biological sources of both aquatic and terrestrial origin based on pyrolysis results and organic petrography (Sanei et al., 2005; Reyes et al., 2006). The reactive, fluorescing component of OM in the sediment is comprised of easily degradable algal-derived lipids, various pigments, as well as plant-derived particles (Figure 3.5). Yellow to yellow-greenish spores and green fluorescing algae (predominately Botryococcus) are abundant in the sediment of Matthews Lake and Control Lake; however, the majority of preserved reactive OM in both of these lakes appears to be alginate and amorphous organic matter (AOM) co-occurring with authigenic minerals (Figure 3.5 b, f, g, h, i). Chlorophyllinite is observed within diatom frustules and authigenic mineral matrices (Figure 3.5 b, h). In both lakes, terrigenous-derived OM consists of trace fibrous plant material (Figure 3.5 a, c). Terrestrial-derived OM is more abundant in the Control Lake sediments than in those of the Matthews Lake core. Fungal spores (sporinite) are notably abundant (Figure 3.5 b, c, d, e).

101

Figure 3.5 Fluorescent-light photomicrographs under oil immersion and blue light excitation of organic matter preserved in Matthews Lake and Control Lake sediment. (a) Cuticle of land-plant leaves or stems (Cutinite); (b) Particulate macerals (e.g. sporinite (spore or pollen grain), alginate (planktonic or benthic), chlorophyllinite in diatom frustules, algae (Botryococcus) are dispersed in an AOM matrix; (c) Framboidal pyrite co-occurring with sporinite and AOM matrix; (d) Terrestrial plant epidermis or cell-wall (Subertinite) and red fluorescing chlorophyllinite; (e) Pyritized funginite; (f) Framboidal pyrite co-occurring with alginate; (g) Particulate macerals (e.g. algae (Botryococcus), red fluorescing chlorophyllinite, and alginate within AOM matrix; (h) Particulate macerals (e.g. sporinite (spore or pollen grain), chlorophyllinite in diatom frustules, and algae (Botryococcus) are dispersed in an AOM matrix; (i) Framboidal pyrite co-occurring with sporinite, algae (Botryococcus), and AOM.

3.5.4 Sediment and porewater geochemistry

3.5.4.1 Elemental concentrations in geochemical groups

Five geochemical zones are delineated in the Matthews Lake core (ML-G1 to ML-G5) and four in the Control Lake core (CL-G1 to CL-G4) (Figure 3.3, 3.4).

Matthews Lake

ML-G1 (ca. 1920 ± 30 AD to 2016 ± 70 AD; 0 to 6 cm; n = 6) and ML-G2 (ca. 780 ± 140 cal. BP to 1920 ± 30 AD; 7 to 13 cm; n = 7) have been impacted by historical mining activities and are not the

102

focus of this study (see Miller et al. (2019) for details). Increases in As concentration in sediment pre- dating mining activities (ML-G2) may be attributed to post-depositional mobility of As. These groupings are characterized by increased concentrations of As, Au, Hg, Fe, Mn, and W (Table 3.1).

The concentrations of these elements decrease with depth in the core. The identification of distinct geochemical groupings in the top 0 to 5 cm in the Matthews Lake sediment is, in part, attributed to post-depositional redox cycling and surface enrichment, particularly of redox-sensitive elements (As,

Fe, Mn). Increased concentrations of dissolved Fe, Mn, and S are observed just below the sediment-water interface (SWI) and no variations in dissolved As are observed in the shallow sediment. Low porewater volumes in deeper sediment of Matthews Lake (below 20 cm) precluded analysis of elements other than

As.

Geochemical zones in background sediment of the Matthews Lake core are characterized by relative changes in the distribution of several elements (Table 3.1; Appendix B). Increased concentrations of S, Sr, and Zn relative to over- and under-lying sediment are observed in ML-G3 (ca. 2490 ± 140 to 780

± 140 cal. yrs BP; 13 to 20 cm; n = 7) while lower concentrations of Au, Zn, Hg, and Cu, relative to ML-

G1 to ML-G3, characterize ML-G4 (ca. 4070 ± 130 to 2490 ± 140 cal. yrs BP; 20 to 37 cm; n = 17; Table

3.1). Dissolved As concentrations are also lower in ML-G4; however, begin to increase in sediments deeper than 34 cm. ML-G5 is predominantly characterized by increased concentrations of As and S. A concurrent decrease in sediment Fe and slight increases in Co are also observed. Maximum porewater As concentrations in Matthews Lake (84 µg·L-1) occur at approximately 40 cm depth (Figure 3.3).

Control Lake

CL-G1 (ca. 1980 ± 30 to 2016 ± 60 AD; 0 to 5 cm; n = 5) and CL-G2 (ca. 400 ± 120 cal BP to

1980 ± 30 AD; 5 to 14 cm; n = 9) may have been impacted by metal-rich dusts derived from the mine site, mine roads, and/or the nearby airstrip (see Miller et al. (2019) for details). As observed in Matthews

Lake, post-depositional mobility of As may account for increased As concentrations in sediment pre- dating mining activities. These sediments contain elevated concentrations of As, Au, Ca, Hg, Pb, Sb, Fe and Mn (Table 3.1). Dissolved As, Fe and Mn profiles display sharp peaks at the SWI (CL-G1) and 103

decrease to background concentrations within the top 5 cm of the sediment column; maximum dissolved

S concentrations (0.85 mg·L-1) occur below the SWI at 7 cm (CL-G2). Concentrations of dissolved Fe decrease to below detection at 6 cm, corresponding with maximum dissolved S concentrations.

The two distinct geochemical groups (CL-G3 (ca. 1770 ± 60 to 400 ± 120 cal. BP; 14 to 36 cm; n

= 22) and CL-G4 (ca. 2110 ± 80 to 1770 ± 60 cal. BP; 36 to 48 cm; n = 12)) in background sediment of the Control Lake core are characterized by relative changes in the concentrations of several elements

(Table 3.1; Appendix B). CL-G3 is characterized by the highest median concentration of As in the core.

Concentrations of Co, Zn, and Fe are elevated relative to CL-G2 and CL-G4. Dissolved As concentrations increase from 28 to 38 cm. Porewater S concentrations are variable, ranging from 0.08 to 0.58 mg·L-1.

CL-G4 has lower As, Mn, and Fe sediment concentrations than CL-G3, but higher concentrations of S and Ag. Increased concentrations of S correspond with slight increases in sediment As concentrations

(Figure 3.3, 3.4). Maximum porewater As concentrations (15 µg·L-1) occur at approximately 40 cm depth.

104

Table 3.1 Median sediment and porewater (pw) concentrations of selected elements for defined geochemical groupings within Matthews Lake and Control Lake sediment cores (ranges in brackets).

As Aspw Au Ag Ca Cu Co Fe Hg Mn Pb S Sb Sr W Zn mg·kg-1 µg·L-1 µg·kg-1 µg·kg-1 wt. % mg·kg-1 mg·kg-1 wt. % µg·kg-1 mg·kg-1 mg·kg-1 wt. % mg·kg-1 mg·kg-1 mg·kg-1 mg·kg-1

Matthews Lake sediment core

ML-G1 48.0 3.85 29.6 210 0.41 81.1 11.9 2.09 53.0 324 6.07 0.29 0.18 14.0 1.20 128 & (30.8- (2.31- (7.60- (188- (0.37- (74.9- (11.3- (1.87- (32.0- (295- (4.28- (0.24- (0.11- (12.1- (0.90- (117- ML-G2 89.5) 4.81) 230) 241) 0.47) 94.8) 13.6) 2.68) 79.0) 554) 10.2) 0.34) 0.51) 14.7) 3.60) 167) (n = 13)

34.0 2.50 10.2 230 0.43 86.3 12.2 2.15 30.0 359 4.84 0.33 0.14 15.0 1.00 189 ML-G3 (32.0- (1.96- (7.10- (197- (0.41- (78.7- (11.9- (1.96- (19.0- (344- (4.49- (0.29- (0.12- (14.6- (1.00- (181- (n = 7) 37.8) 3.67) 12.2) 243) 0.45) 91.4) 12.5) 2.47) 47.0) 405) 5.83) 0.34) 0.17) 15.4) 1.20) 232)

32.0 2.84 5.60 189 0.41 70.7 12.2 2.17 25.0 353 4.05 0.31 0.11 13.5 1.10 145 ML-G4 (29.6- (1.99- (2.90- (157- (0.39- (65.8- (11.5- (1.87- (16.0- (324- (3.69- (0.28- (0.09- (12.9- (1.00- (133- (n = 17) 36.6) 6.97) 11.4) 209) 0.42) 75.5) 13.0) 2.30) 40.0) 378) 4.65) 0.33) 0.13) 14.1) 2.00) 170)

36.2 48.5 3.20 162 0.40 68.2 13.6 1.83 22.0 337 3.55 0.43 0.1 13.1 0.8 124 ML-G5 (31.4- (23.9- (2.60- (148- (0.39- (66.1- (12.1- (1.65- (11.0- (318- (3.34- (0.32- (0.08- (12.4- (0.7- (117- (n = 15) 48.8) 84.4) 5.20) 186) 0.42) 77.5) 14.7) 2.09) 30.0) 363) 4.01) 0.55) 0.12) 14.8) 1.2) 141)

Control Lake sediment core CL-G1 103 1.15 8.00 129 0.23 61.8 13.4 4.83 53.0 173 8.01 0.37 0.21 16.2 1.05 83.6 & (86.1- (0.64- (4.70- (94.0- (0.21- (42.9- (11.8- (4.18- (21.0- (161- (5.61- (0.34- (0.09- (14.4- (0.90- (61.5- CL-G2 224) 5.66) 14.1) 167) 2.39) 72.4) 17.7) 7.31) 76.0) 241) 9.94) 0.39) 0.31) 46.2) 1.20) 110) (n = 14)

124 2.57 4.70 151 0.20 71.6 16.1 5.22 26.5 178 5.18 0.32 0.08 14.2 1.05 93.9 CL-G3 (56.1- (2.44- (1.20- (147- (0.19- (58.7- (14.4- (2.82- (18.0- (134- (4.20- (0.31- (0.04- (12.2- (0.70- (62.6- (n = 22) 167) 6.45) 8.20) 182) 0.32) 78.8) 37.4) 6.90) 36.0) 200) 9.06) 0.36) 0.12) 16.2) 1.20) 106)

89.1 6.88 2.70 171 0.21 69.3 16.0 4.10 26.0 144 4.52 0.36 0.07 13.2 0.95 82.0 CL-G4 (56.1- (2.44- (1.20- (147- (0.19- (58.7- (14.4- (2.82- (18.0- (134- (4.20- (0.31- (0.04- (12.2- (0.70- (62.6- (n = 12) 130) 14.8) 5.10) 195) 0.23) 75.0) 25.1) 5.80) 42.0) 153) 4.97) 0.38) 0.09) 14.2) 1.50) 92.9)

105

3.5.4.2 Dissolved As speciation

In the Matthews Lake core, As (V) is the predominant inorganic species in porewater; conversely,

As (III) is most prevalent in the porewater of the Control Lake core In both the Matthews Lake and

Control Lake sediment cores, decreasing proportions of As (V) are observed with depth and correspond with increases in As (III) and AsR, respectively. In the Control Lake core, the lowest As (III) percentages occur at 5 to 6 cm and correspond to maximum concentrations of dissolved S (0.81 mg·L-1). Residual As accounts for 5 to 79 % (median = 47 %; n = 32) of total As in interstitial waters of the Matthews Lake sediment core and demonstrates a general trend of increasing relative proportions with depth. In the

Control lake core, AsR accounts for a lower proportion of total dissolved As (median = 25 %; range: 0.6 to

68 %; n = 47) and is more abundant in shallow sediment (0 to 25 cm) than deeper in the sediment core

(Figure 3.3, 3.4).

3.5.5 Solid phase As speciation

The main As-hosting solid phases identified by SEM-based automated mineralogy in the sediments of Matthews and Control lakes include: arsenopyrite, framboidal pyrite, and authigenic Fe-

(oxy)hydroxides (Figure 3.6; Appendix B). In pre-mining sediment of the Matthews Lake core, discrete grains of scorodite (FeAsO4·2H2O) are also observed (Figure 3.6, 3.7). The abundance of these minerals varies with depth in the sediment in both lakes (Figure 3.7). Grain morphology can be used to distinguish between authigenic Fe-(oxy)hydroxides and detrital Fe-oxides (Schuh et al., 2018; Miller et al., 2019;

Figure 3.6). Electron microprobe analysis of secondary As-bearing minerals demonstrates that framboidal pyrite (n = 28) and authigenic Fe-(oxy) hydroxides (n = 35) host an average of 0.50 and 0.30 wt. % As, respectively. Scorodite particles range in size from 1 to 5 µm and contain an average of 31.0 wt. % As (n

= 7). EPMA scans on a funginite indicate the presence of trace As associated with these reworked particles of OM (Figure 3.5 e; Appendix B).

106

Figure 3.6 SEM-BSE images of As-bearing phases in lake sediment cores from Matthews Lake (ML) and Control Lake (CL); Arsenopyrite (ML-A: 44 to 45 cm; CL-A: 28 to 29cm); framboidal pyrite (ML-B: 44 to 45 cm; CL-B: 28 to 29 cm); detrital Fe-oxide (ML-C: 44 to 45 cm; CL-C: 44 to 45 cm); scorodite (ML-E: 26 to 27 cm);authigenic Fe- (oxy)hydroxides (ML-D: 44 to 45 cm; CL-D: 39 to 40 cm). Representative EDS spectra for each phase are presented in Appendix B.

Pre-mining As concentrations in the sediments from Matthews Lake increase from 41 to 48 cm, reaching maximum concentrations (49 mg·kg-1) at 47 to 48 cm (Figure 3.3). Over this interval, automated mineralogy results demonstrate that As is primarily hosted by framboidal pyrite (75 to 80 wt. % of total

As) and As-bearing Fe-oxyhydroxides (15 to 20 wt. % of total As), while arsenopyrite and scorodite comprise a trace amount of total As in the sediment (2 wt. % and 1 wt. %, respectively) (Figure 3.7;

Appendix B). Reduced arsenic species (As (-I), As (III)) are more abundant over this interval relative to higher in the sediment column and are likely attributed to the observed increase in framboidal pyrite

(Figure 3.7).

107

High sedimentary As concentrations in basal sediments of the Control Lake core are hosted by

As-bearing Fe-(oxy)hydroxides (80 wt. % of total As) and framboidal pyrite (20 wt. % of total As).

Elevated As in mid-core sediments is also associated mainly with Fe-(oxy)hydroxides and framboidal pyrite (98.5 wt. % and 1.50 wt. %, respectively), with trace arsenopyrite accounting for < 1 wt. % of total

As. Over intervals with increasing abundance of framboidal pyrite and arsenopyrite, the presence of reduced As species ((As (-I), As (III)) is observed (Figure 3.7). No arsenopyrite was observed with the increase in sediment As concentrations from 39 to 40 cm in Control Lake.

3.5.6 Statistical relationship between organic matter and As

Spearman Rank correlation shows that OM fractions S1, S2, S3, PC, RC, TOC, OI, and inferred- chlorophyll-a have significant (p ≤ 0.05) relationships with total sedimentary As concentration in the

Matthews Lake and Control Lake sediment cores; however, these relationships vary with geochemical

108

zones in each core (Table 3.2). The relationship between different OM fractions also varies with depth in each sediment core (Appendix B).

Increases in dissolved and solid phase As are observed in clusters ML-G1, ML-G3 (porewater only) and ML-G5 (Figure 3.3) in the Matthews Lake sediment core. The pre-mining sediments represented by clusters ML-G3 to ML-G5 are the focus of statistical analyses for this paper. In cluster

ML-G4, As concentrations are most strongly correlated with Fe; however, correlations with OM fractions

(OI, and chlorophyll-a) are also observed (Table 3.2). Within the ML-G5 geochemical population the strongest correlation exists between As and S concentrations (rs= 0.88; p ≤ 0.001; n = 15). Arsenic is also positively and significantly correlated with S2, PC, and TOC (Table 3.2). Over ML-G5, sediment S concentrations are significantly related to HI (Table 3.2).

In the Control Lake sediment core, As, Fe, and OI are positively correlated to each other; however, a significant negative relationship between As and S2 is observed (Table 3.2). Increases in dissolved and solid phase As occur in clusters CL-G1, CL-G3, and CL-G4; however, pre-mining sediment represented by clusters CL-G3 and CL-G4 are the focus of this paper. Within CL-G3 and CL-

G4, the strongest correlation exists between sediment As and Fe concentrations (Table 3.2). Within CL-

G3 both As and Fe are significantly correlated to S3 and OI. Within CL-G4, a strong and significant relationship exists between As, chlorophyll-a and OI while significant negative relationships are observed between As and other OM variables (S1, S2, TOC, PC, HI) (Table 3.2). No relationship between chlorophyll-a and other OM fractions are observed in CL-G3 or CL-G4 (Appendix B).

109

Table 3.2 Spearman’s correlation of As and selected elemental and organic parameters for defined geochemical groupings within Matthews Lake (ML-G3 to ML-G5) and Control Lake (CL-G3 and CL-G4) sediment cores. Only rs that are significant at the p ≤ 0.05 level are shown.

Grouping Age range (cal. yrs BP) Variable rs p Matthews Lake sediment core ML-G3 (n = 7) 2490 ± 140 to 780 ± 140 Au 0.89 0.05 Fe 0.84 0.001 OI 0.70 0.05 ML-G4 (n = 17) 4070 ± 130 to 2490 ± 140 Chlorophyll-a 0.70 0.05 HI - 0.74 0.05 S 0.88 0.001 S2 0.65 0.05 ML-G5 (n = 15) 5430 ± 110 to 4070 ± 130 PC 0.62 0.05 TOC 0.60 0.05 Control Lake sediment core Fe 0.97 0.001 Mn 0.85 0.001 OI 0.84 0.001 CL-G3 (n = 22) 1770 ± 60 to 400 ± 120 S3 0.65 0.05 RC 0.60 0.05 S 0.55 0.05 HI - 0.67 0.05 Fe 1.0 0.001 Chlorophyll-a 0.86 0.001 OI 0.74 0.05 CL-G4 (n = 12) 2110 ± 80 to 1770 ± 60 TOC - 0.62 0.05 HI - 0.71 0.05 PC - 0.80 0.05 S2 - 0.80 0.05

3.6 Discussion

3.6.1 Climate controls on As speciation and mobility

3.6.1.1 Holocene Thermal Maximum (ML-G5; ca. 5430 ± 110 to 4070 ± 130 cal. yrs BP)

Pre-mining As concentrations in sediment and porewaters of the Matthews Lake core are highest in sediments dating in age from approximately 5,400 to 3,800 cal. yrs BP (median ages; 52 to 34 cm;

Figure 3.8). In the Matthews Lake sediment core, higher concentrations of S, S1, S2, TOC, and chlorophyll-a are also observed from approximately 5,400 to 3,800 cal. yrs BP (ML-G5) relative to shallower in the core, suggesting an increased production of OM (Figure 3.3; Outridge, et al., 2007, 2017;

110

Michelutti et al., 2010, Carrie et al., 2012). A gradual increase in HI is observed with depth in the

Matthews Lake core, reaching a maximum at approximately ca. 5,400 cal. yrs BP (52 cm; Figure 3.3).

This increase suggests an interval of relatively high primary production (Sebag et al., 2013). However, a negative correlation between inferred chlorophyll-a and S2 in ML-G5 (rs = -0.55; p ≤ 0.05; n = 15) suggests that terrigenous OM is also an important source of S2 to this lake (Sanei et al., 2005; Carrie et al., 2012). Over the Holocene Thermal Maximum (HTM), melting permafrost and the northern migration of the treeline provided abundant sources of terrestrial OM to the sediment (i.e. conifer needles, leaves, humic acid) that may contribute to the S2 signal (Sulphur et al., 2016; Macumber et al., 2018). The presence of a bimodal distribution in the S2 signal of the pyrolysis data suggests that OM in this part of the core is derived from both aquatic (first peak, S2a) and terrigenous sources (second peak, S2b; Sanei et al., 2005). Organic petrology of sediments from this interval in the Matthews Lake core reveals green- fluorescing algae, chlorophyllinite within diatom frustules and preserved within the authigenic mineral matrix, abundant algal biomass (alginate) and partially oxidized terrestrial organic material (e.g. suberinite, cutinite, funginite). These specific varieties of OM reflect abundant aquatic production and terrigenous OM input at this time. Increased abundance of terrestrial plant-derived OM, notably suberinite, is observed over this interval and suggests that OM input to this lake system is influenced by changes in overland hydrology (Figure 3.5 d).

During the time period represented by sediments in interval ML-G5, warm early Holocene temperatures occurred across the northern hemisphere associated with the HTM (MacDonald et al., 1993;

Edwards et al., 1996; Pienitz et al., 1999; Andersen et al. 2004; Kaufman et al., 2004; Upiter et al., 2014;

Macumber, 2015, 2018; Sulphur et al., 2016). In the NT, the HTM resulted in a 1 to 2 °C increase in air temperature. A chironomid-based temperature reconstruction from nearby Carleton Lake suggests that average July temperatures of 12.6 °C prevailed between ca. 6,000 cal. yr BP to ca. 4,000 cal. yr BP

(Upiter et al., 2014; Figure 3.8). This warming event resulted in the northward advance of the tree line by at least 150 km relative to its present position between 4,600 to 4,000 cal. yrs BP (Figure 3.8; Upiter et al., 2014; Sulphur et al., 2016). The tree line reached at least as far north as Toronto, Waterloo, Queen's 111

and McMaster lakes (Figure 3.1 A; MacDonald et al., 1993; Pienitz et al., 1999, Upiter et al., 2014;

Sulphur et al., 2016; all informal names) by ca. 5,500 to 5,000 cal. yrs BP (Moser and MacDonald, 1990).

This northward migration of the tree line resulted in changes to organic carbon input to these lake systems with a peak in TOC recorded at ca. 4,250 cal. yrs BP in Carleton Lake (Upiter et al., 2014). Warmer air temperatures in sub-Arctic Canada during the Holocene resulted in decreased ice cover, longer growing seasons, and increased biological productivity in lake systems (Smol, 1988; Douglas and Smol 1999;

Griffiths et al., 2017).

The mineralogy of sediments was examined in the Matthews Lake core where there was an increase in porewater and solid-phase As concentrations and S1 and S2 at 42 cm (ca. 4590 ± 160 cal. yrs

BP) and 45 cm (ca. 4790 ± 110 cal. yrs BP) (Figure 3.3). At these depths, authigenic As-bearing framboidal pyrite is the primary host of As in sediment (80 wt. % of total As) while arsenopyrite and scorodite, a common weathering product of arsenopyrite, account for trace amounts of total As. The presence of discrete scorodite grains (0.5 wt. % of total As; Figure 3.6, 3.7) suggests that weathering of mineralized bedrock provided an ongoing background source of As to this lake. Although they were not detected in analysis, other weathering products of As-enriched bedrock (i.e. As-bearing Fe-oxyhydroxides and dissolved As) likely would have contributed to the enrichment of As in the near-surface sediment at this time. However, possible post-depositional dissolution of minerals precludes accurate identification of these original mineral assemblages.

Increased porewater As concentrations and the lack of correlation between As and Fe in ML-G5 also suggest that the onset of reducing conditions during early diagenesis resulted in the reductive dissolution of As-bearing solid phases. Scorodite is only thermodynamically stable under oxidizing conditions (Majzlan et al. 2014); therefore, reductive dissolution of these mineral grains would liberate As to porewater in the near-surface sediment. The positive relationship between As and S (rs = 0.88; p ≤

0.001, n = 15), combined with a predominance of authigenic As hosts over this horizon, supports the interpretation of remobilization and re-precipitation in As-bearing framboidal pyrite during early diagenesis (Nesbitt et al., 1995; Walker et al., 2006; Corkhill and Vaughan, 2009). Increased OM content 112

in the water column and near-surface sediment may have also resulted in increased rates of sulfate reduction and subsequent precipitation of Fe-sulfides (Berner et al. 1985). Authigenic Fe-sulphide minerals have been shown to sequester As in sediment through sorption or co-precipitation mechanisms

(Farquhar et al., 2002; Bostick and Fendorf, 2003; Lowers et al., 2007; Le Pape et al., 2017; Wang et al.,

2018). In low-temperature aqueous environments, increases in reactive OM may also provide localized reducing micro-environments that facilitate the formation of more abundant FeS2 (Wilkin and Barnes,

1997; MacLean et al., 2008). The association between labile, reactive OM and framboidal pyrite is evident in fluorescent light microscopy (Figure 3.5). This relationship is most apparent over the HTM warming interval archived in sediments of ca. 5020 ± 30 cal. yrs BP (46 to 47 cm) in the Matthews Lake core. The hypothesis that OM may facilitate sulphide precipitation, and, consequently, As sequestration, in lake sediments is further supported by a significant (p ≤ 0.05) relationship between HI and S.

Over the ML-G5 interval, As is significantly correlated (rs ≥ 0.60; p ≤ 0.05; n = 15) with both S2 and TOC which suggests that OM plays a direct role in the sequestration of As in sediment. The mechanism of As sequestration could include direct sorption to OM, precipitation of authigenic As- bearing sulphides, or a combination of both (Redman et al. 2002; Biswas et al., 2019). The presence of oxygen-bound As (V) in reducing sediments has been observed in previous bulk As XAS studies and may suggest that As is bound to oxygen-containing functional groups of OM (Lowers et al., 2007; Stuckley et al., 2016; Wang et al., 2018; Biswas et al., 2019).

The geochemical and organic petrological evidence for the role of labile OM in the sequestration of As in sediment in this study supports findings reported by Galloway et al. (2017). In their study of modern sediments of the Yellowknife region impacted by mining and mineral processing, it was reported that reactive OM (S1) acts as a substrate for microbial growth and mediation of As-sulphide precipitation.

The S1 fraction of OM is generally altered through humification and loss of functional groups during early diagenesis, therefore, the relationship between S1 and As was not expected to be as strong in the current study examining older, buried sediments (Sanei et al., 2005; Carrie et al., 2012). The inconsistent

113

relationship between As, S and S2 over the HTM warming interval warrants further investigation of the role of reactive OM on the binding and sequestration of As in lake sediment.

114

Figure 3.8 Summary stratigraphic figure comparing arsenic concentrations (porewater and sediment) and selected organic matter parameters (S2, S3) preserved in the Matthews Lake and Control Lake sediment cores(*) to regional paleoclimate reconstructions using the relative abundance of Picea pollen (1Sulphur et al., 2016), C/N ratios (2Griffith, 2013), diatom-inferred dissolved organic carbon (DOC) (3Pienitz et al., 1999), OC (lost on ignition) (4Upiter et al., 2014), and chironomid-inferred mean July average temperatures (MJAT) (4Upiter et al., 2014). Dashed lines define CONISS-based geochemical groupings and shading represents inferred periods of climate variability in the Courageous Lake Greenstone Belt. Dates with an * are calculated based on median ages between Matthews and Control lakes. 115

3.6.1.2 Transitional climate (ML-G4; ca. 4070 ± 130 to 2490 ± 140 cal. yrs BP)

Increasing scorodite abundance is observed from older to younger sediment in interval ML-G4 of the Matthews Lake sediment core, concurrent with a decrease in S2 concentrations. Sand content increases in the core at this time in ML-G4, and TOC declines relative to zone ML-G5 (Figure 3.3).

Sediment As concentrations demonstrate little variability over this interval, but a progressive decrease in porewater As concentrations in younger sediments is observed (Figure 3.8). The timing of these changes at ca. 4,000 cal. yrs BP coincides with the onset of cooler and wetter conditions and consequently, tree line retreat (Moser and MacDonald, 1990; MacDonald et al., 1993; Pienitz et al., 1999; Huang et al.,

2004; Upiter et al., 2014; Sulphur et al., 2016).

A decrease in microscopic charcoal and gradual decline of the fine silt component in a sediment core analyzed from Danny's Lake and a reduced rate of pollen production in Waite Lake indicate progressive regional climate cooling in the central NT over this period (Macumber, 2015; Sulphur et al.,

2016). Chironomid-inferred mean July air temperatures from Carleton Lake record temperatures of approximately 11.5 to 12.5°C from 2,500 to 1,500 cal. yrs BP, only slightly lower than those recorded over the HTM (Upiter et al., 2014). Relatively wet Neoglacial climate conditions could have resulted in enhanced erosion of mineralized bedrock, contributing more scorodite and sand to Matthews Lake. The abundance of scorodite grains suggests bottom waters were oxygenated, thus increasing the preservation of scorodite in near-surface sediments at this time (Figure 3.7, 3.8). Total organic carbon is lower in ML-

G4 than in preceding intervals, suggesting less primary production at this time. Lower OM production coupled with less stable and/or shorter intervals of thermal stratification during cool climate conditions, would have promoted oxygenated bottom water conditions in the lake.

In ML-G4, As is most strongly correlated with Fe and no relationship between As and S is observed, consistent with mineralogical data that identified scorodite and Fe-(oxy)hydroxides as the predominant hosts of As in these sediments. The predominance of allochthonous As minerals suggests that weathering of mineralized bedrock was the main source of As in Matthews Lake sediment during this time. If bottom water conditions during this interval were predominantly oxidizing, detrital arsenopyrite 116

would likely have been oxidized in situ following deposition, although no textural evidence of this process was observed. A positive correlation between As, OI and Fe (Table 3.2; rs = 0.83, p ≤ 0.001; n =

17) over ML-G4 suggests that As may also be associated with terrigenous plant materials in the catchment of Matthews Lake and introduced to the lake through weathering processes (Carrie et al.,

2012). A slight increase in S2 is observed in Matthews Lake sediments at approximately 2,500 cal. yrs BP

(ML-G3; 13 to 20 cm), suggesting that a brief warming event punctuated late-Holocene Neoglacial conditions in the central NT. However, there is no change in As concentrations in the sediments or porewaters of Matthews Lake during this episode, possibly due to its brevity (Figure 3.3, 3.8).

The development of relatively cool and wet climate conditions associated with the Neoglacial also affected the sediment and porewater geochemistry of Control Lake. Arsenic concentrations in porewaters extracted from sediments of the Control Lake core reach maximum concentrations at ca.

2,100 – 1,700 cal. yrs BP (36 to 48 cm; CL-G4) and decrease in younger sediments. Concurrent with an increase in porewater As concentrations, S2 and TOC reach maximum concentrations at this time and begin to decrease in younger sediment (Figure 3.4, 3.8). Petrographic evidence of aquatic (alginate and diatoms) and terrestrially-derived OM (funginite and sporinite) provide evidence of both autochthonous and allochthonous sources of reactive OM to Control Lake during this progressive cooling event (Figure

3.5). A significant correlation between OI and As suggests detrital terrigenous loading of As to Control

Lake over this interval (Table 3.2; Figure 3.4, 3.8).

Arsenic is associated primarily with Fe-(oxy)hydroxides (80 %) and framboidal pyrite (20 %) in

CL-G4 (Figure 3.7). An increase in porewater As concentrations and a significant (p ≤ 0.05, n = 12) negative correlation between sediment As concentrations and S2 (rs = -0.80), and TOC (rs = -0.62) suggests that sufficient reactive organic material was present in lake sediment during these transitional climate conditions to drive the reductive dissolution of As-bearing phases and mobilize As, as observed in

Matthews Lake over the HTM. The influence of reactive OM on the localized precipitation of As-bearing framboidal pyrite in lake sediments is apparent petrographically in the Control Lake sediment core

(Figure 3.5 f). Funginite macerals observed in CL-G4 have undergone pyritization, with Fe-oxides and 117

FeS nucleating outward from the cell wall; EPMA analysis shows that these phases contain trace amounts of As (Figure 3.5; Appendix B). Despite the depth in the sediment core, bulk XANES analysis demonstrates the presence of both oxidized and reduced As species in zone CL-G4 (Figure 3.7). The abundance of As (V) species may reflect the affinity of As (V) for Fe-(oxy)hydroxides over the pH range of the Control Lake porewaters (median = 5.5; range = 5.2 to 5.9; n = 7) (Dixit and Hering, 2003).

3.6.1.3 Latest Holocene cooling (CL-G4 & CL-G3; ca. 2110 ± 80 to 400 ± 120 cal. BP)

Maximum solid-phase sediment As concentrations are observed in sediments deposited between ca. 1,700 to ca. 500 cal. yr BP (21 to 29 cm) in Control Lake (CL-G3) (Figure 3.4, 3.8). Over this stratigraphic interval, the Control Lake sediment core exhibits decreasing S2 and a corresponding increase in S3 and OI, indicating that allochthonous OM delivery to Control Lake was important during this time

(Figure 3.4). Chironomid-inferred air temperatures from nearby Carleton Lake suggest that the lowest temperatures over the last ca. 6,000 years occurred between approximately ca. 1,550 and ca. 900 cal. yr.

BP in the central NT (Upiter et al., 2014). Episodic increases in sand content are observed in the Control

Lake sediment core over this period (Figure 3.4) and also recorded in Carleton Lake (Figure 3.1) from ca.

1,700 to ca. 700 cal. yrs BP (Upiter et al., 2014). Reduced OM content and increased input of sand to lake sediment likely reflect increased erosion and/or a reduction in lake productivity (Upiter et al., 2014). Low porewater As concentrations and a significant correlation between As and S3 (rs = 0.65; p ≤ 0.05), suggest that during cooler and wetter climate conditions, weathering of mineralized bedrock may still contribute

As to lake sediment but at the same time, less autochthonous OM production and/or increased input of terrigenous OM may reduce the mobility of As in sediments.

3.6.2 Implications for 21st century warming

Using dated sediments and comparisons to regional late-Holocene climate records, this study demonstrates that past climate has influenced As mobility and preservation in lake sediments. Climate- related influences on As transport and fate must therefore be considered to predict the response of sediment geochemistry to 21st c. warming. Climate change has already affected organic carbon cycling in high northern latitudes through a variety of processes, including permafrost thaw (Frey and McClelland, 118

2009; Vonk et al., 2013; 2015), vegetation change (Lantz et al., 2012), longer ice-free seasons (Douglas and Smol, 1999), and increased water temperature (Smol and Douglas, 2007). Increasing concentrations of S1, S2, and TOC are observed in the near-surface sediment of Matthews and Control lakes, and this is interpreted to be a result of 21st century warming (Miller et al., 2019). In mining-impacted northern lakes, continued climate warming is expected to lead to changes in the organic geochemistry of lake sediments and this will affect the cycling and long-term stability of As and other elements of potential concern

(Galloway et al., 2017; Miller et al., 2019). Using past episodes of climate warming, such as the HTM, as an analogue for on-going and future climate warming, the results of this study demonstrate that increases in the abundance of reactive OM during climate warming will impact geochemical baselines in aquatic systems.

In both Matthews and Control lakes, the deposition of allochthonous As-bearing minerals in times of both warming and cooling suggest that weathering of bedrock and derived surficial materials is a continual source of As to these lakes. However, increased aquatic production and vegetation reorganization associated with the HTM resulted in increased porewater As concentrations in both lakes.

This response suggests that 21st c. climate warming and resulting increased biological productivity will influence redox conditions in bottom waters and near-surface sediment (Barrett et al., 2019).

Development of longer intervals of reducing conditions will promote reductive dissolution of As-bearing minerals such as Fe-(oxy)hydroxides and scorodite, which may result in increases in As concentrations in near-surface sediments (MacDonald et al., 2005). However, increases in reactive OM may also facilitate the precipitation of As-bearing framboidal pyrite that can sequester As. In mineralized or mining- contaminated regions, warming climate conditions can result in increased surface water and sediment As concentrations in lakes.

3.7 Conclusions

Detailed geochemical and sedimentological analyses of sediment and porewaters preserved in a ca. 5,400 cal BP-year-old sediment core from Matthews Lake and a ca. 2,100 cal. yr BP-year-old

119

sediment core from Control Lake located in the central Canadian sub-Arctic were used to study the influence of late-Holocene climate change on the loading and mobility of As in lake sediments.

Weathering of mineralized bedrock and terrigenous material provides an ongoing source of As to

Matthews and Control lakes; however, increased aquatic production and reorganization of terrestrial vegetation, associated with the HTM, resulted in reductive dissolution and re-precipitation of As in the sediment of both lakes. Stepwise pyrolysis and petrographic analyses provide evidence for increased aquatic (i.e. alginate and diatoms) and terrestrially derived OM (i.e. funginite and sporinite) in the lake sediment during this late-Holocene warming event. Through increased autochthonous OM production and allochthonous OM delivery, warmer climate conditions caused an increase in the mobility of As, resulting from As release from minerals to pore waters via reductive dissolution followed by sequestration in authigenic As-bearing framboidal pyrite. The findings of this study suggest that increased biological production, resulting from 21st c. climate warming, will promote reducing conditions in near-surface sediment and bottom waters and, thereby, enhance the mobility of As in lakes.

Progressive cooling and the onset of Neoglacial conditions in the central NT correspond with a decrease in reactive OM and a concurrent decrease in porewater As concentrations in younger sediments of both lakes. An increased abundance of scorodite and As-bearing Fe-(oxy)hydroxides with the onset of cooler regional climate conditions suggests that a decrease in OM production resulted in more oxidizing conditions in bottom waters and near-surface sediment, limiting the reductive dissolution of As-bearing minerals and the mobility of As in sediment. Cooler conditions resulted in more stable As minerals, and increased sequestration in sediment.

3.8 Acknowledgements

This project was jointly funded by Polar Knowledge Canada (Project# 1519-149, awarded to J.M.

Galloway (GSC Calgary) and T.R. Patterson (Carleton University)), the Environmental Geoscience

Program, Natural Resources Canada (Metal Mining Project, M.B. Parsons; Northern Baselines Activity,

J.M. Galloway), a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery

120

Grant (H.E. Jamieson; RGPIN/03736-2016), and a NSERC Northern Research Supplement (H.E.

Jamieson; RGPNS/305500-2016). Financial support was also contributed through student awards to C.B.

Miller from the NSERC Create Mine of Knowledge and the Northern Scientific Training programs. Field programs were conducted under Aurora Research Institute Licence No. 15858 (J.M. Galloway). We are grateful for substantial in-kind support from Crown-Indigenous Relations Northern Affairs Canada

(Murray Somers and Joel Gowman) during the field program, including fixed-wing transport and field accommodations at Tundra Mine in 2016. Synchrotron-based analyses were performed at the Advanced

Photon Source, Argonne National Laboratory, Sector 20-BM, which is supported by a partnership between the X-ray Science Division and Canadian Light Source. The authors would like to thank Brian

Cummings (Paleoecological Environmental Assessment and Research Lab (PEARL), Queen’s

University), Christopher Grooms (PEARL), Agatha Dobosz (Queen’s Facility for Isotope Research

(QFIR)), Brian Joy (QFIR), and Matt Newville of the APS for their assistance and invaluable guidance. A special thank you to Andrew Macumber (Carleton University), Hendrik Falck (Northwest Territories

Geological Survey), Nawaf Nasser (Carleton University), and the staff of Delta Engineering and Nahanni

Construction for their support during our northern field work. JMG contributed to this manuscript with support from a Marie Skłodowska-Curie Fellowship under the European Union’s Horizon 511 2020

(Grant agreement # 754513) and the Aarhus University Research Foundation at the Aarhus Institute of

Advanced Studies, Aarhus University, Denmark. This is Contribution Number 20190286 of the Lands and Minerals Sector, Natural Resources Canada.

121

3.9 References

ACIA, 2004. Impacts of a Warming Arctic: Arctic Climate Impact Assessment. ACIA Overview report. Cambridge University Press. 140 pp.

AECOM, 2015. Tundra Mine, NWT Remediation environmental monitoring program: water quality Part L: 2014 and historical results, AECOM. http://mvlwb.ca/Registry.aspx?a=MV2009L8-0008

Andersen, C., Koç, N., Moros, M., 2004. A highly unstable Holocene climate in the subpolar North Atlantic: Evidence from diatoms. Quat. Sci. Rev. 23, 2155–2166. https://doi.org/10.1016/j.quascirev.2004.08.004

Barrett, P.M., Hull, E.A., Burkart, K., Hargrave, O., McLean, J., Taylor, V.F., Jackson, B.P., Gawel, J.E., Neumann, R.B., 2019. Contrasting arsenic cycling in strongly and weakly stratified contaminated lakes: Evidence for temperature control on sediment–water arsenic fluxes. Limnol. Oceanogr. 64, 1333–1346. https://doi.org/10.1002/lno.11119

Berner, R.A., De Leeuw, J.W., Spiro, B., Murchison, D.G., Eglinton, G., 1985. Sulphate reduction, organic matter decomposition and pyrite formation. Philos. Trans. R. Soc. London. Ser. A, Math. Phys. Sci. 315, 25–38. https://doi.org/https://doi.org/10.1098/rsta.1985.0027

Biswas, A., Besold, J., Sjostedt, C., Gustafsson, J.P., Scheinost, A.C., Planer-Friedrich, B., 2019. Complexation of arsenite, arsenate, and monothioarsenate with oxygen-containing functional groups of natural organic matter: An XAS Study. Environ. Sci. Technol. acs.est.9b03020. https://doi.org/10.1021/acs.est.9b03020 122

Blaauw, M., 2010. Methods and code for “classical” age-modelling of radiocarbon sequences. Quat. Geochronol. 5, 512–518. https://doi.org/10.1016/j.quageo.2010.01.002

Blaauw, M., Christen, J.A., 2013. Bacon manual – v2.2. Tutorial 1–11. http://chrono.qub.ac.uk/blaauw/manualBacon_2.2.pdf

Bostick, B.J., Fendorf, S., 2003. Arsenite sorption on troilite (FeS) and pyrite (FeS2). Geochim. Cosmochim. Acta 67, 909–921. https://doi.org/10.1016/S0016-7037(02)01170-5

Bowell, R.J., Alpers, C.N., Jamieson, H.E., Nordstrom, D.K., Majzlan, J., 2014. The Environmental Geochemistry of Arsenic – An Overview. Rev. Mineral. Geochemistry 79, 1–16. https://doi.org/10.2138/rmg.2014.79.1

Buckwalter-Davis, M., 2013. Automated Mineral Analysis of Mine Waste. MSc. Thesis. Queen's University. https://qspace.library.queensu.ca/bitstream/handle/1974/8200/Buckwalter- Davis_Martha_J_201308_MSc.pdf?sequence=1&isAllowed=y

Campbell, K.M., Nordstrom, D.K., 2014. Arsenic speciation and sorption in natural environments. Rev. Mineral. Geochemistry 79, 185–216. https://doi.org/http://dx.doi.org/10.2138/rmg.2014.79.3

CCME, 2001a. Canadian Sediment Quality Guidelines for the Protection of Aquatic Life, Environmental Protection, CCME. https://ccme.ca/en/resources/canadian_environmental_quality_guidelines/

CCME, 2001b. CCME Water Quality Guidelines for the Protection of Aquatic Life, CCME. https://ccme.ca/en/resources/canadian_environmental_quality_guidelines/

Corkhill, C.L., Vaughan, D.J., 2009. Arsenopyrite oxidation - A review. Appl. Geochemistry 24, 2342– 2361. https://doi.org/10.1016/j.apgeochem.2009.09.008

Craw, D., Bowell, R.J., 2014. The Characterization of Arsenic in Mine Waste. Rev. Mineral. Geochemistry 79, 473–505. https://doi.org/https://doi.org/10.2138/rmg.2014.79.10

Couture, R.M., Gobeil, C., Tessier, A., 2010. Arsenic, iron and sulfur co-diagenesis in lake sediments. Geochim. Cosmochim. Acta 74, 1238–1255. https://doi.org/10.1016/j.gca.2009.11.028

123

Dalton, A.S., Patterson, T., Roe, H., AL., M., Swindles, G., Galloway, J., Vermaire, J., Crann, C., Falck, H., 2018. Late-Holocene climatic variability in Subarctic Canada : insights from a high-resolution lake record from the central Northwest Territories. PLoS One 13, 1–21. https://doi.org/10.1371/journal.pone.0199872

Das B, Vinebrooke RD, Sanchez-Azofeifa A, Rivard B, Wolfe AP., 2005. Inferring lake sediment chlorophyll concentrations with reflectance spectroscopy: a novel approach to reconstructing trophic status changes in alpine lakes. Can J Fish Aquat Sci 62:1067–1078. http://doi:10.1139/f05-016

Davis, J., 2011. Statistics and data analysis in geology 3rd Edition. John Wiley & Sons, Ltd.

DeSisto, S. L., Jamieson, H. E., & Parsons, M. B. (2011). Influence of hardpan layers on arsenic mobility in historical gold mine tailings. Applied Geochemistry, 26(12), 2004–2018. http://doi.org/10.1016/j.apgeochem.2011.06.030

Douglas, M.S. V., Smol, J.P., 1999. Freshwater diatoms as indicators of environmental change in the High Arctic, in: Stoermer, E.F., Smol, J.P. (Eds.), The Diatoms. Cambridge University Press, Cambridge, pp. 227–244. https://doi.org/10.1017/CBO9780511613005.011

Du Laing, G., Chapagain, S.K., Dewispelaere, M., Meers, E., Kazama, F., Tack, F.M.G., Rinklebe, J., Verloo, M.G., 2009. Presence and mobility of arsenic in estuarine wetland soils of the Scheldt estuary (Belgium). J. Environ. Monit. 11, 873–881. http://doi.org/10.1039/B815875D

Ecosystem Classification Group. 2008. Ecological Regions of the Northwest Territories - Taiga Shield. Department of Environment and Natural Resources, Government of the Northwest Territories, Yellowknife, NT. viii + 146 pp. + insert map.

Edwards, T.W.D., Wolfe, B.B., Macdonald, G.M., 1996. Influence of changing atmospheric circulation on precipitation δ18O-temperature relations in Canada during the Holocene. Quat. Res. 46, 211– 218. https://doi.org/10.1006/qres.1996.0061

Fandrich, R., Gu, Y., Burrows, D., Moeller, K., 2007. Modern SEM-based mineral liberation analysis. Int. J. Miner. Process. 84, 310–320. https://doi.org/10.1016/j.minpro.2006.07.018

Farquhar, M. L., Charnock, J. M., Livens, F. R., Vaughan, D. J. 2002, Mechanisms of arsenic uptake from aqueous solution by interaction with goethite, lepidocrocite, mackinawite, and pyrite: An X-ray absorption spectroscopy study. Environ. Sci. Technol. 36, 1757–1762. https://doi.org/10.1021/es010216g

124

Folinsbee, R.E., Moore, J.C.G., 1955. Matthews Lake, District of Mackenzie, Northwest Territories. Geological Survey of Canada, "A" Series Map 1024A. https://doi.org/10.4095/107491

Frey, K.E., McClelland, J.W., 2009. Impacts of permafrost degradation on arctic river biochemistry. J. Glaciol. 23, 169–182. https://doi.org/10.1002/hyp

Geological Survey of Canada, 2014. Surficial geology, Lac de Gras, Northwest Territories, NTS 76-D; Geological Survey of Canad, Canadian Geoscience Map 184 (2nd edition, preliminary, Surficial Data Model v. 2.0 conversion of Map 1870A), scale 1:125,000. doi: 10.4905/295497

Glew, J.R., Smol, J.P., Last, W.M., 2001. Sediment core collection and extrusion. pp. 73-105. In: Last, W.M. and Smol, J.P. [Editors]. Tracking environmental change using lake sediments. Vol 1: Basin Analysis, Coring, and Chronological Techniques. Kluwer Academic Publishers, Dordrecht.

Golder Associates, Ltd., 2005. Terrestrial and aquatic field surveys and ecological risk assessment of the Tundra Gold Mine, Golder Associates Ltd. http://mvlwb.ca/Registry.aspx?a=MV2009L8-0008

Golder Associates, Ltd., 2017. 2016 Aquatic Effects Monitoring Program Annual Report. Snap Lake Mine, Type A Water Licence MV2011L2-0004, Golder Associates Ltd. http://registry.mvlwb.ca/Documents/MV2011L2-0004/MV2011L2-0004%20- %20De%20Beers%20Snap%20Lake%20-%202016%20AEMP%20Annual%20Report%20- %20May1-17.pdf

125

Griffith, F., 2013. Holocene and recent paleoclimate investigations using carbon and nitrogen isotopes from bulk sediment of two subarctic lakes, central Northwest Territories. University of Ottawa, Ontario. (MSc Thesis). http://dx.doi.org/10.20381/ruor-3393

Grimm, E.C., 1987. CONISS: A FORTRAN 77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares. Comput. Geosci. 13, 13–35

Hatfield Consulting, 1982. Initial Environmental Evaluation for Giant Yellowknife Mines - Salmita Project. Unpublished report prepared for Giant Yellowknife Mines Ltd., Yellowknife, NT. pp. 1 - 24.

Henderson, J., 1970. Stratigraphy of the Yellowknife Supergroup, Yellowknife Bay - Prosperous Lake area, District of MacKenzie: Geological Survey of Canada, Paper 70-26, p. 562 - 566.

Henderson, J.B., 1985. Geology of the Yellowknife-Hearne Lake Area, District of Mackenzie: A Segment Across an Archean Basin; Memoir 414, 135 p.

Huang, C.C., MacDonald, G., Cwynar, L., 2004. Holocene landscape development and climatic change in the low arctic, Northwest Territories, Canada. Palaeogeogr. Palaeoclimatol. Palaeoecol. 205, 221e234. http://dx.doi.org/10.1016/j.palaeo.2003.12.009

Huang, J.H., Ilgen, G., 2006. Factors affecting arsenic speciation in environmental samples: Sample drying and storage. Int. J. Environ. Anal. Chem. 86, 347–358. https://doi.org/10.1080/03067310500227878

126

Juggins, S., 2017. Analysis of quaternary science data, Package ‘ rioja .’ http://cran.r.project.org/package= rioja.

Kaufman, D.S., Ager, T.A., Anderson, N.J., Anderson, P.M., Andrews, J.T., Bartlein, P.J., Brubaker, L.B., Coats, L.L., Cwynar, L.C., Duvall, M.L., Dyke, A.S., Edwards, M.E., Eisner, W.R., Gajewski, K., Geirsdóttir, A., Hu, F.S., Jennings, A.E., Kaplan, M.R., Kerwin, M.W., Lozhkin, A. V., MacDonald, G.M., Miller, G.H., Mock, C.J., Oswald, W.W., Otto-Bliesner, B.L., Porinchu, D.F., Rühland, K., Smol, J.P., Steig, E.J., Wolfe, B.B., 2004. Holocene thermal maximum in the western Arctic (0-180°W). Quat. Sci. Rev. 23, 529–560. https://doi.org/10.1016/j.quascirev.2003.09.007

Kirk, M.F., Holm, T.R., Park, J., Jin, Q., Sanford, R.A., Fouke, B.W., Bethke, C.M., 2004. Bacterial sulfate reduction limits natural arsenic contamination in groundwater. Geology 32, 953–956.

Lafargue, E., Marquis, F., Pillot, D., 1998. Rock-Eval 6 Applications in hydrocarbon exploration, production, and soil contamination studies. Rev. l’Institut Français du Pétrole 53, 421–437. https://doi.org/10.2516/ogst:1998036

Lantz T.C., Marsh P., Kokelj S.V., 2012. Recent shrub proliferation in the Mackenzie Delta uplands and microclimatic implications. Ecosystems, doi:10.1007/s10021-012- 9595-2

Le Pape, P., Blanchard, M., Brest, J., Boulliard, J.C., Ikogou, M., Stetten, L., Wang, S., Landrot, G., Morin, G., 2017. Arsenic incorporation in pyrite at ambient temperature at both tetrahedral S-I and octahedral Fe-II Sites: Evidence from EXAFS-DFT analysis. Environ. Sci. Technol. 51, 150–158. https://doi.org/10.1021/acs.est.6b03502

Lowers, H.A., Breit, G.N., Foster, A.L., Whitney, J., Yount, J., Uddin, Md N., Muneem, Ad A., 2007. Arsenic incorporation into authigenic pyrite, Bengal Basin sediment, Bangladesh. Geochim. Cosmochim. Acta 71 (11), 2699–2717. https://doi.org/10.1016/j.gca.2007.03.022

Lynch, J., 1990. Provisional elemental values for eight new lake sediment and stream sediment reference materials LKSD-1, LKSD-2, LKSD-3, LKSD-4, STSD-1, STSD-2, STSD-3, and STSD-4. Geostand. Newsl. 14, 153–167. https://doi.org/10.1111/j.1751-908X.1990.tb00070.x

127

MacDonald, R.W., Harner, T., Fyfe, J., 2005. Recent climate change in the Arctic and its impact on contaminant pathways and interpretation of temporal trend data. Sci Total Env. 342, 5–86. https://doi.org/S0048-9697(04)00838-1 [pii]\n10.1016/j.scitotenv.2004.12.059

MacLean, L.C., Tyliszczak, T., Gilbert, P.U., Zhou, D., Pray, T.J., Onstott, T.C., Southam, G., 2008. A high-resolution chemical and structural study of framboidal pyrite formed within a low-temperature bacterial biofilm. http://doi.org/10.1111/j.1472-4669.2008.00174.x

Macumber, A.L., 2015. Grain-size analysis and Arcellinida as tools for inferring paleoclimatological variability : a case study from Holocene Canadian subarctic lake sediments. Carleton University. Doctoral dissertation. https://curve.carleton.ca/system/files/etd/622aaee1-0a90-4077-bdf7- 838298d14c0e/etd_pdf/083ece3a2a5ee991a0e74ef0fa67858e/macumber- grainsizeanalysisandarcellinidaastoolsfor.pdf

Macumber, A.L., Patterson, R.T., Galloway, J.M., Falck, H., Swindles, G.T., 2018. Reconstruction of Holocene hydroclimatic variability in subarctic treeline lakes using lake sediment grain-size end- members. Holocene 28, 845–857. https://doi.org/10.1177/0959683617752836

Majzlan J, Drahota P, Filippi M., 2014. Parageneses and crystal chemistry of arsenic minerals. Rev Mineral Geochem 79:17-184.

Martin, A.J., Pedersen, T.F., 2002. Seasonal and interannual mobility of arsenic in a lake impacted by metal mining. Environ. Sci. Technol. 36, 1516–1523. https://doi.org/10.1021/es0108537

Michelutti, N., Wolfe, A. P., Vinebrooke, R.D., Rivard, B., Briner, J.P., 2005. Recent primary production increases in arctic lakes, Geophys. Res. Lett., 32, L19715, http://doi.org/10.1029/2005GL023693

Michelutti, N., Blais, J.M., Cumming, B.F., Paterson, A.M., Rühland, K., Wolfe, A.P., Smol, J.P., 2010. Do spectrally inferred determinations of chlorophyll a reflect trends in lake trophic status? J. Paleolimnol. 43, 205–217. https://doi.org/10.1007/s10933-009-9325-8

Moser, K., MacDonald, G., 1990. Holocene vegetation change at treeline north of Yellowknife, Northwest Territories, Canada. Quaternary Research, 34(2), 227-239. https://doi.org/10.1016/0033- 5894(90)90033-H

Nearing, M.M., Koch, I., Reimer, K.J., 2014. Complementary arsenic speciation methods: A review. Spectrochim. Acta Part B 99, 150–162.

128

Nesbitt, H.W., Muir, I.J., Prarr, A.R., 1995. Oxidation of arsenopyrite by air and air-saturated, distilled water, and implications for mechanism of oxidation. Geochim. Cosmochim. Acta 59, 1773–1786. https://doi.org/10.1016/0016-7037(95)00081-A

Nordstrom, D.K., 2002. Worldwide Occurrences of Arsenic in Ground Water. Science 296, 2143–2145. https://doi.org/10.1126/science.1072375

Outridge, P., Sanei, H., Goodarzi, F., 2007. Evidence for control of mercury accumulation rates in Canadian High Arctic lake sediments by variations of aquatic primary productivity. Enviromental Sci. Technlogy 41, 5259–5265. https://doi.org/10.1021/es070408x

Philippsen, B., 2013. The freshwater reservoir effect in radiocarbon dating. Herit. Sci. 1, 1–19. https://doi.org/10.1186/2050-7445-1-24

Pickhardt, P.C., Folt, C.L., Chen, C.Y., Klaue, B., Blum, J.D., 2002. Algal blooms reduce the uptake of toxic methylmercury in freshwater food webs. Proc. Natl. Acad. Sci. U. S. A. 99, 4419–23. https://doi.org/10.1073/pnas.072531099

Pirrie, D., Rollinson, G.K., 2011. Unlocking the applications of automated mineral analysis. Geol. Today 27, 226–235. https://doi.org/10.1111/j.1365-2451.2011.00818.x

Prowse, T.D., Wrona, F.J., Reist, J.D., Gibson, J.J., Hobbie, J.E., Lévesque, L.M.J., Vincent, W.F., 2006. Climate change effects on hydroecology of arctic freshwater ecosystems. Ambio 35, 347–58.

129

Prowse, T., Alfredsen, K., Beltaos, S., Bonsal, B., Duguay, C., Korhola, A., McNamara, J., Pienitz, R., Vincent, W.F., Vuglinsky, V., Weyhenmeyer, G.A., 2011. Past and future changes in arctic lake and river ice. Ambio 40, 53–62. https://doi.org/10.1007/s13280-011-0216-7

R Core Team. 2014. R: A Language and Environment for Statistical Computing. http://www.r-project.org

Ransom, A., Robb, M., 1986. The Salmita Gold Deposit. In: Gold in the Western Shield. Edited by L.A. Clarke and D.R. Francisis. The Canadian Insitute of Mining and Metallurgy, Special Volume 38, pp. 285–305.

Ravel, B., Newville, M., 2005. Athena, Artemis, Hephaestus: Data analysis for X-ray absorption spectroscopy using Ifeffit. J. Synchrot. Radiat. 12, 537–541.

Redman, A.D., Macalady, D.L., Ahmann, D., 2002. Natural organic matter affects arsenic speciation and sorption onto hematite. Environ. Sci. Technol. 36, 2889–2896. https://doi.org/10.1021/es0112801

Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Buck, C.E., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Haflidason, H., Hajdas, I., Hatté, C., Heaton, T.J., Hoffmann, D.L., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M., Reimer, R.W., Richards, D.A., Scott, E.M., Southon, J.R., Turney, C.S.M., van der Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration curves, 0-50,000 years cal BP. Radiocarbon 55: 1869-1887

Reimann, C., Filzmoser, P., Garrett, R., Dutter, R., 2008. Statistical Data Analysis Explained: Applied Environmental Statistics with R. John Wiley & Sons, Ltd.

Reyes, J., Goodarzi, F., Sanei, H., Stasiuk, L.D., Duncan, W., 2006. Petrographic and geochemical characteristics of organic matter associated with stream sediments in Trail area British Columbia, Canada. Int. J. Coal Geol. 65, 146–157. https://doi.org/10.1016/j.coal.2005.04.016

Sanei, H., Stasiuk, L.D., Goodarzi, F., 2005. Petrological changes occurring in organic matter from recent lacustrine sediments during thermal alteration by Rock-Eval pyrolysis. Org. Geochem. 36, 1190– 1203. https://doi.org/10.1016/j.orggeochem.2005.02.009

130

Seabridge Gold Inc., 2018. Seabridge discovers new gold zones at Courageous Lake. News Release: May, 9, Seabridge Gold Inc. http://seabridgegold.net/pdf/NR/NMay9-18.pdf

Sebag, D., Debret, M., M’voubou, M., Obame, R.M., Ngomanda, A., Oslisly, R., Bentaleb, I., Disnar, J.- R., Giresse, P., 2013. Coupled Rock-Eval pyrolysis and spectrophotometry for lacustrine sedimentary dynamics: Application for West Central African rainforests (Kamalété and Nguène lakes, Gabon). The Holocene 23, 1173–1183. https://doi.org/10.1177/0959683613483622

SENES Consultants Ltd., 2006. Tundra Mine Site Ecological Risk Assessment of Water Management, SENES Consultants Ltd. http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochemistry 17, 517–568. https://doi.org/10.1016/S0883-2927(02)00018-5

Smol, J. P., 1988. Paleoclimate proxy data from freshwater Arctic diatoms. Verhandlungen Internationale Vereinigung fur theoretische und angewandte Limnologie, 23: 837–844.

Smol, J.P., Douglas, M.S. V., 2007. From controversy to consensus: making the case for recent climate change using lake sediments. Front. Ecol. Environ. 5, 466–474. https://doi.org/10.1890/060162

Stuckey, J.W., Schaefer, M. V., Kocar, B.D., Benner, S.G., Fendorf, S., 2016. Arsenic release metabolically limited to permanently water-saturated soil in Mekong Delta. Nat. Geosci. 9, 70–76. https://doi.org/10.1038/ngeo2589

Tetra Tech Wardrop, 2012. Courageous Lake Prefeasibility Study, Tetra Tech Wardrop. https://secure.kaiserresearch.com/i/jk/tr16/TRSEA20120905.pdf

131

Topographical Survey of Canada, 1925. Map of Lockhart River Basin, Great Slave Lake District, Northwest Territories: including the head waters of Thelon, Back and Coppermine Rivers. Scale 1:380,160. Yale University, Beinecke Rare Book and Manuscript Library, Call Number: 743 L79 1925.

Upiter, L.M., Vermaire, J.C., Patterson, R.T., Crann, C.A., Galloway, J.M., Macumber, A.L., Neville, L.A., Swindles, G.T., Falck, H., Roe, H.M., Pisaric, M.F.J., 2014. Middle to late-Holocene chironomid-inferred July temperatures for the central Northwest Territories, Canada. J. Paleolimnol. 52, 11–26. https://doi.org/10.1007/s10933-014-9775-5

URS, 2005. Geochemical Assessment of Acid Rock Drainage and Metal Leaching Potential of Tailings and Waste Rock, Tundra Mine, NWT, URS. http://mvlwb.ca/Registry.aspx?a=MV2009L8-0008

Vonk, J.E., Mann, P.J., Davydov, S., Davydova, A., Spencer, R.G.M., Schade, J., Sobczak, W.V., Zimov, N., Zimov, S., Bulygina, E., Eglinton, T.I., Holmes, R.M., 2013. High biolability of ancient permafrost carbon upon thaw. Geophys. Res. Lett. 40, 2689–2693

Vonk, J.E., Tank, S.E., Bowden, W.B., Laurion, I., Vincent, W.F., Alekseychik, P., Amyot, M., Billet, M.F., Canário, J., Cory, R.M., Deshpande, B.N., Helbig, M., Jammet, M., Karlsson, J., Larouche, J., MacMillan, G., Rautio, M., Walter Anthony, K.M., Wickland, K.P., 2015. Reviews and Syntheses: Effects of permafrost thaw on arctic aquatic ecosystems. Biogeosciences Discuss. 12, 10719–10815. https://doi.org/10.5194/bgd-12-10719-2015

Walker, F.P., Schreiber, M.E., Rimstidt, J.D., 2006. Kinetics of arsenopyrite oxidative dissolution by oxygen. Geochim. Cosmochim. Acta 70, 1668–1676. https://doi.org/10.1016/j.gca.2005.12.010

Wang, Y., Le Pape, P., Morin, G., Asta, M.P., King, G., Bártová, B., Suvorova, E., Frutschi, M., Ikogou, M., Pham, V.H.C., Vo, P. Le, Herman, F., Charlet, L., Bernier-Latmani, R., 2018. Arsenic speciation in Mekong Delta sediments depends on their depositional environment. Environ. Sci. Technol. 52, 3431–3439. https://doi.org/10.1021/acs.est.7b05177

Wilkin, R.T., Barnes, H.L., 1997. Formation processes of framboidal pyrite. Geochim. Cosmochim. Acta 61, 323–339. https://doi.org/10.1016/S0016-7037(96)00320-1

Wolfe, AP., Vinebrooke, RD., Michelutti, N., Rivard, B., Das, B. 2006. Experimental calibration of lake- sediment spectral reflectance to chlorophyll a concentrations: methodology and paleolimnological validation. J Paleolimnol 36:91–100. doi: 10.1007/s10933-006-0006-6

Zhang, X., Flato, G., Kirchmeier-Young, M., Vincent, L., Wan, H., Wang, X., Rong, R., Fyfe, J., Li, G., Kharin, V.V. (2019): Changes in temperature and precipitation across Canada; Chapter 4 in Bush, E. and Lemmen, D.S. (Eds.) Canada’s Changing Climate Report. Government of Canada, Ottawa, Ontario, pp 112-193.

132

Chapter 4

Solid phase organic matter control on arsenic mobility in mining-impacted

sediment, Tundra Mine, Northwest Territories, Canada

4.1 Introduction

In Canada’s north, modern climate warming has resulted in both increasing temperatures and duration of ice-free seasons, promoting increased biological productivity and transport of organic matter

(OM) to lake systems (Frey and McClelland, 2009; Prowse et al., 2011; Stern et al., 2012; Griffiths et al.,

2017). The accumulation of OM influences the biogeochemistry and redox dynamics of near-surface sediments in northern lakes (Toevs et al., 2006; McGuire et al., 2009; Williamson et al., 2015; Rantala et al., 2016; Spence et al., 2016). These climate-driven changes may, in turn, influence the mobility of redox-sensitive elements, such as arsenic (As) (Macdonald et al., 2005; Wang and Mulligan, 2006). Lakes in proximity to mining activities may have elevated concentrations of As in aquatic sediments due to the input of tailings, waste rock, mine effluent, airborne emissions, and/or runoff from mine waste and naturally mineralized bedrock (DeSisto et al., 2011; Craw and Bowell, 2014; Miller et al., 2019a).

Changing redox conditions, due to seasonal variations and resultant fluctuations in OM, are a predominant mechanism of As release from sediments to lake waters (Martin and Pedersen, 2002; Bauer and Blodau, 2006; Andrade et al., 2010; Couture and Cappellen, 2011; Anawar et al., 2013; Barrett et al.,

2019; Miller et al., 2019b; Palmer et al., 2019). The influence of changing redox conditions on the geochemical cycling of As has been well documented; however, the role of solid phase OM (OM > 0.22 to 0.45 µm) in this process is not as well understood (Langner et al., 2012; Anawar et al., 2013; Biswas et al., 2019), particularly in lake environments. As climate warming continues in Canada’s north, an improved understanding of the effects of increasing OM on the long-term stability of As is needed to more accurately predict the impacts of future climate variations on As mobility in northern lake systems.

133

In lacustrine environments, the post-depositional mobility of As is governed by the solid phase speciation of As, which depends primarily on the redox conditions of the sediment and associated porewaters (Smedley and Kinniburgh, 2002). Under oxic conditions in near-surface lake sediments, As is sequestered through adsorption and co-precipitation processes with authigenic and detrital iron (Fe)- and manganese (Mn)-(oxy)hydroxides at the sediment-water interface (SWI) (Dixit and Hering, 2003).

Development of reducing conditions, through burial or increased abundance of labile OM, allows for the sequestration of dissolved As through sorption and co-precipitation reactions involving authigenic sulphides (Farquhar and Livens, 2002; Bostick et al., 2005; Lowers et al., 2007; Le Pape et al., 2017;

Schuh et al., 2019). However, the onset of reducing conditions in sediments is progressive and results in an oxic-to-anoxic transition zone where decreasing concentrations of available oxygen (O2) result in the release of As to porewater through both the oxidation of As-bearing sulfides and reductive dissolution of

Fe- and Mn-(oxy)hydroxides (Wolthers et al., 2005; Bennett et al., 2012; An et al., 2017). Under changing redox conditions, the formation of organo-mineral aggregates (OMAs) and/or iron monosulphides (FeS) (i.e. mackinawite) may promote the sequestration of trace metal(loid)s (i.e. As, Cu,

Cd, Ni, Pb, Zn) through co-precipitation with or sorption to these authigenic phases; however, the effectiveness of these mechanisms in natural environments is difficult to predict (Moon and Peacock,

2012; Chen et al., 2014; Kleber et al., 2015; An et al., 2017; Vega et al., 2017; Du et al., 2018; Qu et al.,

2019).

In the near-surface sediment, dissolved oxygen concentrations of sediment pore waters are controlled largely by labile OM; therefore, redox conditions may be influenced by increases in labile OM due to climate warming (Meyers and Ishiwatari, 1993). The potential effects of increased OM on As mobility are diverse and largely depend on the speciation of As in sediment and porewater (Biswas et al.,

2019; Miller et al., 2019a). Many authors have examined the influence of OM on the stability of As- bearing minerals (Redman, 2002; Bostick and Fendorf, 2003; Dixit and Hering, 2003; Islam et al., 2004;

Wang and Mulligan, 2006; Wang et al., 2018) and As sorption to mineral surfaces in near-surface lake sediments (Davis, 1982; Grafe et al., 2001, 2002; Redman, 2002; Bauer and Blodau, 2006; Sharma et al., 134

2010; Guo et al., 2011; Stuckey et al., 2016). Recent studies highlight the importance of solid phase OM as an alternative sorbent of As in anoxic and oxic sediment (Langner et al., 2011; Wang et al., 2018;

Biswas et al., 2019). The role of solid phase OM in the sequestration of As may be of particular importance under transitional redox conditions where Fe reduction and S oxidation occur simultaneously and As is more susceptible to remobilization.

Solid phase OM in sediment is a heterogeneous mixture of organic compounds derived from both aquatic biological matter and terrigenous plant materials (Tissot and Welte, 1978; Meyers and Ishiwatari,

1993; Reyes et al., 2006). In lakes, OM comprises a diversity of compounds with varying structural and functional properties that influence their reactivity (Gu et al., 1994; Chen et al., 2002, 2004; Fellman et al., 2010). The affinity of these functional groups for some metal ions may strongly affect the mobility of these metals in aquatic systems (reviewed in Mostofa et al., 2009). A large percentage of organic carbon

(OC) in sediment is associated with authigenic Fe-oxides which form in-situ at sediment redox boundaries through the oxidation of porewater Fe (II) in the presence of OC, resulting in the formation of stable Fe-OC aggregates (Lalonde et al., 2012; Barber et al., 2017). At the oxic/anoxic boundary, interactions between OM, Fe, and sulphur (S) may result in the co-localized precipitation of authigenic minerals (i.e. Fe-oxides and sulphides) with OMAs or terrigenous-derived OM (TOM) (Wilkin and

Barnes, 1997; Maclean et al., 2008; Couture and Cappellen, 2011). As the biogeochemical cycling of As in lake sediments is closely associated with the cycling of Fe, S, and OM, these dynamics will, in turn, influence the mobility of As in the near-surface sediment (Couture et al., 2010). Arsenic binding to OM may occur via ternary complexes through the formation of a metal-cation bridge (i.e. Fe (III), Al (III), Ca

(II)) (Ritter et al., 2006; Sharma et al., 2010; Hoffmann et al., 2013), thiol-bonding to sulphydryl groups of NOM (Langner et al., 2011; Wang et al., 2018), or complexation with the oxygen-containing functional groups of OM (Tessier et al., 1996; Biswas et al., 2019). Co-precipitation of Fe and OM may also enhance the sorption of trace metal(loid)s to authigenic Fe minerals (i.e. Zn, Cd, Pb) (Tessier et al., 1996).

These studies suggest that OM may provide an alternative substrate for As immobilization and influence the behavior of As in near-surface sediments. 135

In Canada’s north, As concentrations in lakes surrounding former gold mine sites are commonly elevated due to historical mining and mineral processing activities (Wagemann et al., 1978; Bright et al.,

1994, 1996; Groves et al., 1998; Andrade et al., 2010; Galloway et al., 2012, 2017; Palmer et al., 2015;

Jamieson et al., 2017; Schuh et al., 2018, 2019; Miller et al., 2019a). As temperatures continue to rise and disproportionately affect high northern latitudes (ACIA, 2004), an improved knowledge of how increasing OM will influence the mobility of legacy contaminants in lakes is required to help inform environmental assessments and monitoring activities in northern Canada. The main objective of this study is to examine the association between As and solid phase OM to determine the influence of increasing

OM on the long-term stability of mining-derived As contamination in sub-Arctic lakes.

4.2 Study area

4.2.1 Historical mining and remediation activities

Tundra and Salmita mines are approximately 240 km NE of Yellowknife, NT, Canada, on the eastern shore of Matthews Lake (Figure 4.1). Over two phases of mining and milling, 1964 to 1968 and

1983 to 1987, approximately 285,000 oz of Au were produced at grades ranging from 18.4 to 27.8 g·t-1

(Ransom and Robb, 1986; Silke, 2001). During these two periods of mining, all ore was processed at the

Tundra Mine facilities and wastes were deposited sub-aqueously into an adjacent lake. During the second period of mining, tailings-core dams were built to reinforce this lake and create a tailings confinement area (Figure 4.1). The Tundra Mine site has been in care and maintenance mode since 1999 under the stewardship of the Contaminants and Remediation Directorate (CARD) of Crown-Indigenous Relations and Northern Affairs Canada (CIRNAC). At the initiation of remediation activities in 2007, approximately 155,000 m3 of waste rock, 240,000 m3 of tailings, and 1,280,000 m3 of impacted water were present on-site (URS, 2005; Golder Associates Ltd., 2008). In lakes downstream of the tailings impoundment, environmental monitoring has demonstrated a wide range of As concentrations in surface water (0.80 to 105 µg·L-1) and near-surface sediment (24 to 1,010 mg·kg-1) (Staples, 2014; Miller et al.,

2019a). Arsenic concentrations in the near-surface sediment of all 5 lakes sampled by Miller et al. (2019a)

136

exceed the CCME Interim Sediment Quality Guideline (ISQG; 5.9 mg kg−1) (CCME, 2001a, 2001b). The disposal of waste rock, tailings overtopping and seepage, weathering and airborne deposition of waste rock and tailings, discharge of treated tailings effluent, and natural weathering of mineralized bedrock have all contributed to the elevated As concentrations in lakes around the former Tundra Mine (Miller et al., 2019a). The site is currently in an adaptive management and long-term monitoring phase, following the completion of remediation activities in August 2018 (AECOM, 2018).

4.2.2 Sample locations

The study area is located in the central portion of the Slave Geological Province of the Canadian

Shield and is 80 km north of the present-day tree line (Figure 4.1). The area is part of the transition zone between discontinuous and continuous permafrost with numerous small lakes and organic deposits occurring in low lying areas (Seabridge Gold Inc., 2010; Karunaratne, 2011; AECOM, 2015). The modern climate is characterized as relatively cold and dry with long winters and short summers. Small lakes, such as those investigated in this study, may be ice-covered from mid-October to late June (Palmer et al., 2019). Sediment cores were collected from Powder Mag (64.05114  N, 111.15042  W) and

Bulldog (64.03758  N, 111.18337  W) lake in March 2016. Sediment grab samples were collected from

Hambone Lake (64.047033  N, 111.154958  W) in July 2016 (Figure 4.1). Mean sedimentation rates in the near-surface sediment of Powder Mag (52 yr·cm-1) and Bulldog (40 yr·cm-1) lakes are higher than average for small northern lakes (70 yr·cm-1) and attributed to the onset of mining activities at Tundra and

Salmita mines (Crann et al., 2015; Miller et al., 2019a).

137

Figure 4.1 Map showing A) Study location and approximate extents of modern day treeline (red line; Brandt, 2009) and B) Bedrock geology and sampling locations (sediment cores – blue; sediment grab samples – yellow) in the Courageous Lake Greenstone Belt (Folinsbee and Moore, 1955; Thompson and Kerswill, 1994). Surface water flow paths from the former Tundra Mine site (INAC, 2005).

138

4.3 Methods and Materials

4.3.1 Sample collection and preparation

Sediment cores were collected from the frozen surface of Powder Mag and Bulldog lakes using a gravity corer with polycarbonate core tubes (7.5 × 60 cm). All sediment cores were collected from the deepest location in each lake to limit possible disturbance of the sediment caused by sediment focusing or ice scouring in the winter months (Blais and Kalff, 1995). Immediately following collection, cores were vertically extruded and subsampled at 1-cm resolution in a high purity (99.998 %) N2-filled glove bag.

Sediment was subsampled into N2-flushed 50 mL polypropylene centrifuge tubes (Corning® Falcon®) and refrigerated pending shipment to the Geological Survey of Canada – Atlantic (GSC-A) in Dartmouth,

NS, for porewater extraction. Near-surface (0 to 15 cm) sediment grab samples were collected from

Hambone Lake using an Ekman dredge sampler. These sediment samples were frozen and sent to Queen's

University, Kingston, ON, for subsampling and preparation. Aliquots of sediment were freeze-dried at 1.0

Pa and -50 °C for organic geochemical (0.2 g) and inorganic geochemical analyses (0.5 g). A separate aliquot was dried in a high purity N2-filled glove bag to preserve solid-phase As speciation for scanning electron microscope (SEM)-based automated mineralogy, electron microprobe analysis (EPMA), synchrotron-based bulk X-ray Absorption Near-Edge Structure (XANES), micro X-ray fluorescence (μ-

XRF), and micro X-ray diffraction (μ-XRD) (Huang and Ilgen, 2006).

Porewater was extracted from the sediment core samples by centrifugation (Thermo Scientific

Sorvall Legend™ XF Centrifuge) at 4,400 RPM for 30 min (centrifugal force of 4221 Gs) at GSC-A.

Samples were prepared and handled following the protocol detailed in Miller et al. (2019a) for filtered cations, filtered anions, total As, and total inorganic As speciation. For each sample, pH was measured with a hand-held pH meter (HACH H138 MiniLab ISFET) but due to low water content, Eh could not be measured.

139

4.3.2 Sediment and porewater geochemistry

Metal(loid) concentrations in porewaters were measured at the Inorganic Geochemical Research

Laboratory of the Geological Survey of Canada in Ottawa. Analyses of major elements were performed by Inductively Coupled Plasma – Atomic Emission Spectroscopy (ICP-AES) using a Perkin-Elmer 3000

DV. Trace elements were analyzed using Inductively Coupled Plasma – Mass Spectroscopy (ICP-MS) with a Thermo Corporation X-7 Series II. Standard solutions and certified inter-calibration samples were used to assess accuracy and precision; QA/QC results are presented in Miller et al. (2019a). The relative proportions of dissolved As (V), As (III), and a residual fraction (AsR) in water samples were determined by hydride generation–atomic fluorescence spectrometry (HG-AFS; model PSA 10.055 Millennium

Excalibur) at the Université de Montréal. The residual fraction likely includes As species that are not detected using HF-AFS, including thioarsenates in sulfidic waters and non-reducible organoarsenic compounds (Planer-Friedrich and Wallschläger, 2009; PS Analytical, 2018).

Element concentrations in sediment samples (n = 211) were determined by ICP-MS (1F/AQ250 package) following digestion by a modified aqua regia treatment (1:1:1 HCl:HNO3:H2O at 95 °C) at

Acme Analytical Laboratories (Bureau Veritas), Vancouver, BC. Partial digestion with aqua regia was used as key elements of interest (i.e. As, antimony (Sb), mercury (Hg), and S)) may be lost through volatilization during high-temperature fuming with hydrofluoric acid in more aggressive, near-total digestions (Parsons et al., 2012, 2019). Field duplicates, certified reference materials (STSD-3; Lynch,

1990) and laboratory standard reference materials (STD DS10, STD OREAS45EA) were included with each sample set to assess analytical accuracy and precision (see Miller et al. 2019a for details).

4.3.3 Sediment organic and textural characterization

The amount and type of OM in lake sediments (Hambone Lake n = 6; Powder Mag Lake n = 30;

Bulldog Lake n = 37) was determined by programmed pyrolysis at the Geological Survey of Canada

– Calgary (GSC-C; Lafargue et al., 1998). Further details regarding sample preparation, methodologies, and precision measurements are provided in Appendix B.

140

In recent sediments, the S1 fraction organic matter is comprised of aquatic-derived OM (e.g. algal-derived lipids; amino acids, chlorophyll, and small volatile molecules) and S2 compounds are derived from the biomacromolecule structure of algal cell walls and other aquatic biological matter, such as phytoplankton and copepods (Sanei et al., 2005; Carrie et al., 2012). Terrigenous plant materials (i.e. conifer needles, roots, bark), in addition to humic and fulvic acids, comprise the S3 fraction (Carrie et al.,

2012; Albrecht et al., 2015). Studies by Carrie et al. (2012), Albrecht et al. (2015), Galloway et al. (2017),

Outridge et al. (2017), and Miller et al. (2019b) detail the application of Rock-Eval pyrolysis in the study of recent lake sediments.

The distribution and origin of solid phase OM (n = 5) was assessed through qualitative petrographic analysis following methods outlined by Reyes et al. (2006). Incident white light and fluorescent light microscopy were conducted using a Zeiss Axioimager II microscope system (50× magnification) equipped with the Diskus-Fossil system. Fluorescence microscopy was conducted using ultraviolet G 365 nm excitation with a 420 nm filter. Organic matter was classified based on the classification of macerals in International Committee for Coal Petrology (1971), Sanei et al. (2005) and

Reyes et al. (2006).

Comparison of optical properties (fluorescence spectrometry) between samples which had previously been examined using SEM and EPMA analysis to those not exposed to an electron beam suggests that prior SEM and EPMA analysis seems to have resulted in thermal alteration of the OM.

Thermal alteration of OM in lithified rocks by ion milling was demonstrated by Sanei and Ardakani

(2016) but, to the best of our knowledge, the effects of prolonged electron beam exposure on OM in unconsolidated sediment has not been investigated. As a result, SEM/EPMA analysis and organic petrography were conducted on different sub-samples from the same sediment interval.

4.3.4 Solid phase As speciation analysis

Eleven N2-dried lake sediment samples (Hambone Lake n = 3; Powder Mag Lake n = 4; Bulldog

Lake n = 4) were selected for detailed speciation analysis based on bulk elemental composition and position within each core. Polished sections (35 to 50 μm thickness) were prepared at Vancouver 141

Petrographics. To preserve soluble phases, preparation of these samples used limited water and heat

(Schuh et al. 2018, Van Den Berghe et al. 2018).

4.3.4.1 SEM-based Automated Mineralogy and Electron Microprobe analyses

The relative distribution of As-hosting solid phases in selected sediment samples was completed through automated mineralogy, using a FEI™ Quanta 650 Field Emission Gun Environmental SEM, and

Mineral Liberation Analyzer software. This integrated technique quantifies mineral phases through a combination of backscatter electron (BSE) image analysis and Energy-dispersive X-ray Spectroscopy

(Buckwalter-Davis, 2013). Operating conditions, the mineral reference library used for phase classification, and the calculations used to quantify the relative contribution of each As-hosting phase are described in Miller et al. (2019a). Fine grained aggregates of poorly crystalline, As-bearing Fe-sulphides and Fe-(oxy)hydroxides (FeSx/FeO), which co-occur with OM (Figure 4.3), were targeted for analysis using a JEOL JXA-8230 electron microprobe operating in wavelength dispersive spectroscopy mode. Due to the small size of the grains and the somewhat uneven topography of the polished surface which can cause X-ray scattering and inaccurate peak intensity (Rönnhelt et al. 1987), the EPMA results are considered semi-quantitative.

4.3.4.2 Synchrotron-based µXRF and XRD

Arsenic-bearing authigenic grains associated with solid phase OM were identified through combined SEM-based automated mineralogy, EPMA, and fluorescent light microscopy analysis. These grains were subsequently targeted for synchrotron-based μ-XRF and μ-XRD analysis using a monochromatic beam at undulator beamline 13-ID-E at the APS, Argonne National Laboratory. X-ray fluorescence maps and XRD patterns were collected using an incident beam energy of 18.0 keV and a 2

μm beam diameter. Two-dimensional, continuous µ-XRF/XRD mapping was completed at frame rates of

50 to 100 ms per pixel. Specific points identified on μXRF maps were subsequently selected as targets for longer duration μXRD measurements (frame rate: 10,000 ms per pixel). The fluorescent radiation was measured using a Hitachi 4-element Vortex ME4 silicon drift diode positioned at 90 ° to the incident beam and connected to a Xspress 3 digital X-ray multi-channel analyzer system. X-ray diffraction was 142

measured using a Perkin Elmer XRD1621 digital flat panel detector placed 400 mm from the sample and operating in transmission mode. Post-hoc processing and analysis of µ-XRF maps was performed using

Larch software (Version 0.9.46; Newville, 2019). Dioptas software (Prescher and Prakapenka, 2015) was used to calibrate and integrate XRD data; phase identification was performed in X’Pert HighScore Plus.

4.3.4.3 Bulk X-ray Absorption Near-Edge Structure (XANES)

To determine the relative distribution of As oxidation states in the sediments, As K-edge XANES data were collected at Sector 20-BM of the Advanced Photon Source, Argonne National Laboratory

(Chicago, IL). Preparation of samples for these analyses (Hambone Lake n = 3; Powder Mag Lake n = 3;

Bulldog Lake n = 3) followed procedures described by Van Den Berghe et al. (2018). Standard materials were selected to cover the principal oxidation states exhibited by As in lacustrine systems (-1 to +5;

Appendix C). Operational conditions and post-hoc data processing are detailed in Miller et al. (2019b).

The percentages of each model compound reported in this paper are absorption-corrected amounts (post- hoc normalization to 100 %; Miller et al., 2019b).

4.4 Results

4.4.1 Sediment and porewater geochemistry

Sediment As concentrations range from 80 to 1,010 mg·kg−1 in the three lakes sampled near

Tundra Mine. The highest solid-phase As concentrations are observed in Bulldog Lake (median = 570 mg·kg−1; range = 200 to 1,010 mg·kg−1; n = 7) with maximum values occurring from 2 to 3 cm depth in the sediment core. Conversely, maximum dissolved As concentrations are observed in Powder Mag Lake

(maximum: 383 μg·L−1) and are an order of magnitude higher than Bulldog Lake. The concentration and down-core distribution of redox-sensitive elements (Fe and S) varies between the two sediment cores

(Figure 4.2). As sediment cores were not collected from Hambone Lake, the distribution of metal(loid)s in the near surface sediment was not examined; however, spatial variance in the near-surface sediment concentrations of As (79.6 to 622 mg·kg−1), S, and Fe is observed. Sulphur concentrations in Hambone

143

Lake sediment range from 0.99 to 1.90 wt. % (median = 1.0 wt. %; n = 6) and Fe concentrations range from 0.84 to 1.60 % (median = 1.30 %; n = 6).

Porewater As concentrations in Powder Mag and Bulldog lakes range from 21 to 383 μg·L−1 in the near-surface sediment (Figure 4.2). In these two lakes, maximum dissolved As concentration in sediment occur at or just below the sediment-water interface (SWI; Figure 4.2). In Powder Mag Lake, maximum porewater As concentrations occur at the SWI while in Bulldog Lake, peak concentrations are observed at 2.5 cm depth (Figure 4.2).

The pH of porewaters in the near-surface sediment is circumneutral, ranging from 5.7 to 7.5

(median = 5.8; n = 11). The highest pH values are observed in Powder Mag Lake (range = 6.3 to 7.5). No porewater geochemistry is available for Hambone lake; however, past environmental monitoring recorded a surface water pH range from 6.4 to 8.3 (AANDC, 2013; Golder Associates Ltd., 2016).

Figure 4.2 Downcore plots of As, Fe, Mn, and S concentrations in sediment and porewater, and solid-phase organic matter fractions (S1, S2, and TOC). The horizontal dotted line represents the onset of mining activities in 1964 (Miller et al., 2019a).

144

4.4.2 Sediment organic characterization

Total organic carbon content in the near-surface sediment of all lakes studied ranges from 8.7 to

21.5 wt. % (median = 11.4 wt. %; n = 25), with maximum concentrations observed at the SWI in Powder

Mag Lake (Figure 4.2). In both Powder Mag and Bulldog lake sediment cores, an increase in OM is observed at the SWI. The majority of OM is composed of S2 OM (median = 2.4 wt. %; range = 1.2 to 5.4 wt. %; n = 25). In Hambone Lake, the highest TOC (15.5 wt. %) and S2 OM (4.96 wt. %) occur farthest from the eluent discharge location (HAM-2; Figure 4.1).

Classification and identification of sediment OM, based on optical properties (fluorescence and reflectance) and morphology, indicate that OM in the near-surface sediment of lakes in the Tundra Mine region is derived from both aquatic and terrigenous sources (Sanei et al., 2005; Reyes et al., 2006). Woody fragments and resin, suberin (i.e. roots and bark; Figure 4.3 D1), cutinite (i.e. leaves and stems), funginite

(Figure 4.3 C1), and spores originating from terrestrial vegetation are common in the near-surface sediment.

Aquatic-derived OM is abundant in the sediment and comprised primarily of alginate (benthic and planktonic; Figure 4.3 A1, B1, E1, F1). Chlorophyllinite is derived from chlorophyll pigments and can be either aquatic or terrestrial in origin (Figure 4.3 A1, B1, E1, F1). Organo-mineral aggregates (comprised of amorphous organic matter (AOM), particulate macerals (e.g. alginate, chlorophyllinite), oxidized AOM

(Figure 4.3 A1, B1, E1, F1), and mineral matter) are abundant in all sediment samples studied (Figure 4.3

A1, B1, E1, F1). Oxidation of AOM and TOM is commonly observed and results in a spectral shift towards lower energy fluorescence wavelengths (i.e. redshift; Figure 4.3). Through reflected light microscopy, the precipitation of sulphide and Fe-(oxide) minerals are observed associated with oxidized terrestrial and aquatic-derived OM (Figure 4.3). Unaltered (i.e. unoxidized) terrigenous and aquatic OM is identifiable in the sediment as it does not display a spectral shift and preserves its characteristic fluorescence (Figure 4.3).

145

Figure 4.3 Paired (1) fluorescence-light photomicrographs under oil immersion and (2) SEM-BSE images of OMAs (A, B, E, F) and TOM (C, D) preserved in near-surface sediment of Powder Mag, Bulldog, and Hambone lakes. Organic macerals include suberin (D1; green fluorescing), funginite (C1; green to yellow fluorescing inner layer), chlorophyllinite (A1, B1, E1, F1; pink to red fluorescing), oxidized OM (A1, B1, C1, D1, E1, F1; brown fluorescing), alginate (A1, B1. F1; green fluorescing), and AOM (A1, B1, E1, F1; green to yellow fluorescing).

4.4.3 Solid and aqueous As speciation

4.4.3.1 SEM-based Automated Mineralogy and Electron Microprobe analyses

Fine grained aggregates of crystalline Fe-sulphides and Fe-(oxy)hydroxides (FeSx/FeO) are abundant in the near-surface sediment of all lakes studied and observed in association with both OMAs and

146

TOM. SEM-based EDS mapping illustrates the precipitation of both Fe and S within these authigenic minerals (Figure 4.4). Arsenic concentration associated with these authigenic minerals ranges from

Figure 4.4 SEM-BSE images and EDS element maps of (a) OMAs and (b) TOM preserved in the near-surface sediment of Bulldog and Powder Mag lakes.

4.4.3.2 Synchrotron-based µXRF and XRD

Micro-XRF element maps demonstrate that As is associated with OMAs and TOM (Figure 4.5). In

TOM grains, As is generally observed with authigenic FeSx/FeO minerals and dispersed throughout the grain (Figure 4.5a-d). In contrast, As in OMAs is not associated with authigenic FeSx/FeO minerals and is instead observed either as micron-sized grains within the organic material or dispersed throughout the AOM

(Figure 4.5e-h). Micro-XRD results suggest these grains may be As-bearing sulphide minerals, such as orpiment (As2S3); however, analysis did not provide identifiable diffraction patterns for many of these As- rich regions (Figure 4.6).

147

Micro-XRF results confirm both the heterogeneous nature of OMAs and TOM, as well as the association of As with both of these authigenic phases (Figure 4.5). Organo-mineral aggregates are comprised of both fine-grained detrital grains (i.e. silicates, phyllosilicates, quartz, hematite) and mineral phases which may be authigenic in origin (i.e. ferrihydrite, feroxyhyte, goethite, maghemite, mackinawite, pyrite; Figure 4.6). Terrigenous organic matter grains consist of mixtures of goethite, ferrihydrite, maghemite, lepidocrocite, pyrite, and mackinawite (Figure 4.6).

Figure 4.5 Micro-XRF element maps showing As-Kα (red), Fe-Kα (green), and Sr-Kα (blue) intensities overlain on SEM-BSE images of As-bearing TOM (a, b, c, d) and OMAs (e, f, g, h). Colour intensities are not comparable between maps and are not intended to represent exact concentrations. Cross hairs demonstrate location of EPMA analysis, note that symbol is larger than beam diameter.

148

Figure 4.6 Micro-XRF element maps showing As-Kα (red), Fe-Kα (green), and Sr-Kα (blue) intensities of As- bearing OMAs (a) and TOM (b,). Colour intensities are not comparable between maps and are not intended to represent exact concentrations. Integration of the µ-XRD patterns shows the presence of discrete grains of pyrite (Py) and orpiment (Orp) associated with OMAs which are comprised of a heterogenous mixture of detrital (not shown) and authigenic minerals (i.e. feroxyhyte (Feroxy)). Authigenic minerals associated with TOM are comprised of goethite (Gt) and mackinawite (Mkw). Cross hairs demonstrate location of µ-XRF analysis, note that symbol is larger than beam diameter.

4.4.3.3 HG-AFS and bulk XANES

In near-surface (0 to 10 cm) sediment porewaters of Powder Mag and Bulldog lakes, As (V) is the predominant inorganic aqueous As species (Figure 4.7). However, at the SWI in Bulldog Lake, As (III) is

149

the dominant species. Maximum dissolved As concentrations in both lakes occur where As (V) proportions are highest and proportions of AsR are lowest (Figure 4.7). In Bulldog Lake, a trend of increasing proportions of As (V) with depth is observed, while below a depth of 1 cm, the relative proportion of As

(III) is consistently around 10 %. In Powder Mag Lake, relative proportions of both As (III) and AsR increase with depth from the SWI.

The As K-edge XANES spectra demonstrate that reduced solid phase As species (As (-I), As (III)) predominate in the near-surface sediment of all lakes studied, accounting for an average of 80 % (n = 9) total As (Figure 4.7). In Powder Mag Lake, relative proportions of O-bound As (III) and As (V) species

(11871.4 eV and 11875.0 eV, respectively) are higher than S-coordinated As (-I) or As (III) (11868.7 and

11869.5 eV, respectively). At the SWI, O-bound As (III) and As (V) account for 26 % and 47 % of total

As, respectively. An increase in S-bound As species is observed at 3.5 cm depth and corresponds to maximum As concentrations (271 mg kg-1) in the near-surface sediment. Conversely, in Bulldog Lake, S- coordinated As species account for approximately 40 % of solid-phase As species in the upper 3 cm of sediment. A trend of decreasing S-coordinated As species and increasing O-bound As species is observed with depth in Bulldog Lake. In Hambone Lake, ratios of O- and S-bound As species are similar in sediment close to the treated tailings effluent discharge location; however, the relative proportions of S-bound species increase with distance toward the peatland at the NE end of the lake (Figure 4.1, 4.5).

150

Figure 4.7 Relative distribution of As species in porewater (HG-AFS) and sediment (bulk XANES) with depth in the (a) Powder Mag Lake and (b) Bulldog Lake sediment cores and (c) with distance from the effluent discharge location in surface sediments of Hambone Lake.

4.5 Discussion

4.5.1 Sources of As and organic matter to lakes

The sediment and porewater geochemistry of Powder Mag, Bulldog, and Hambone lakes have been altered due to the influence of historical mining practices and subsequent remediation activities. Miller et al. (2019a) identified five main sources of As to lakes downstream of the former Tundra gold mine site: disposal of waste rock, seepage through and over-topping of the tailings dams, weathering and airborne deposition of waste rock and tailings, discharge of treated tailings effluent, and natural weathering of mineralized bedrock. The oxidation of mining-derived sulphides (e.g. arsenopyrite) and authigenic As- bearing framboidal pyrite is the predominant source of increased porewater As concentrations in the near-

151

surface sediment of Bulldog Lake (Miller et al. 2019a). Conversely, in Hambone and Powder Mag lakes, the seepage and overflow of tailings waters and disposal of treated effluent contributed As, in dissolved or suspended form, to the near-surface sediment (Miller et al., 2019a). The abundance of secondary As minerals and prevalence of oxidized As species in the near-surface porewaters and sediment of Powder

Mag Lake (Figure 4.3, 4.5) are attributed to the dissolution of As-bearing Fe-(oxy)hydroxides deposited in the near-surface sediment of this lake and re-precipitated as authigenic As-sulphides.

Arsenic may also be transferred to the lake sediment in association with autochthonous and allochthonous OM. Dissolved As is bioavailable and may be accumulated by aquatic organisms living in the water column, such as phytoplankton and zooplankton, and has been observed to concentrate in these organisms at orders of magnitude higher than in ambient waters (Eisler, 1988; Hellweger et al., 2003; Lopez et al., 2017). Increased labile OM, due to modern day climate warming, is observed both petrographically

(alginate and diatoms) and geochemically (S1 and S2) in the near-surface sediment of Bulldog and Powder

Mag lakes suggesting that the detritus of these organisms provide a source of As to near-surface sediment

(Figure 4.2; Caumette et al., 2012). Warming air temperatures in sub-Arctic Canada are anticipated to result in longer growing seasons and increased biological productivity in lake systems (Smol, 1988; Michelutti et al., 2005; Griffiths et al., 2017). As a result, increasing aquatic production may enhance the loading of As to the near-surface sediment of mining-impacted lakes. In surface waters, As may also form both aqueous and colloidal complexes with natural OM and dissolved or suspended Fe (III), respectively, and settle in the near-surface sediment (Ritter et al., 2006; Sundman et al., 2014). High dissolved organic carbon (DOC) and total Fe in both Powder Mag (DOC = 16.6 mg·L-1; Fe = 0.39 mg·L-1) and Hambone (DOC = 16 mg·L-

1 (average 1990 to 2004); Fe = 0.55 mg·L-1 (average 1990 to 2004), 0.28 mg·L -1 (2015 average)) surface waters suggests that the formation of complexes between arsenate, suspended Fe-(oxy)hydroxide colloids in solution, and natural OM may play a significant role in the transport of both As and OM to the bottom sediments (INAC, 2005; Ritter et al., 2006; Golder Associates Ltd., 2016; Miller et al., 2019a).

Additionally, petrographic evidence of terrestrially-derived OM (i.e. cutinite, funginite, and sporinite) and

152

presence of detrital As-bearing minerals provides evidence that weathering also contributes both OM and

As to these lake systems (Anawar et al., 2013; Miller et al. 2019b).

4.5.2 OMAs and As sequestration

Organo-mineral aggregates, comprised of AOM, particulate macerals (e.g. sporinite (spore or pollen grains), alginate (planktonic or benthic), diatom frustules, and/or algae), and both authigenic and detrital mineral matter are common in the near-surface sediment of the lakes studied (Figure 4.3). In these aggregates, mixtures of detrital minerals (i.e. silicates, quartz, pyrrhotite), fine-grained authigenic mineral phases (i.e. ferrihydrite, goethite, maghemite, mackinawite, pyrite) occur co-localized with AOM (Figure

4.6). Arsenic is associated with these aggregates as: (1) as discrete clusters, (2) evenly dispersed within the

AOM matrix, and (3) sorbed to and/or co-precipitated with authigenic FeSx/FeO portions of the grain.

Within these OMAs, As most commonly occurs as discrete clusters (Figure 4.5). The As-rich clusters are too small for EPMA analysis; however, µ-XRD suggests As is associated with small grains of orpiment (Figure 4.6). The presence of orpiment, particularly in Bulldog Lake, is supported by bulk XANES which indicates that S-coordinated As (III) accounts for approximately 60 % of solid-phase As in the near- surface sediment. A shift in fluorescence colour (from green to red wavelengths) is observed in the AOM associated with these aggregates. This fluorescence shift is caused by the oxidation of OM as it acts as an electron donor to drive sulphide mineralization, suggesting that these discrete As grains are forming in-situ

(Figure 4.3; Davis et al., 1990). The aggregated nature of OMAs protects portions of the AOM from microbial degradation and oxidation (Figure 4.3). The role of reactive OM on the precipitation of authigenic

As-bearing minerals has been well established in both laboratory (Kirk et al., 2004, 2010) and field-based sediment studies (Langner et al., 2011; Galloway et al., 2017; Wang et al., 2018; Miller et al., 2019b).

Within OMAs, coupled µ-XRF and fluorescence microscopy analysis demonstrates that As is also dispersed throughout the AOM (Figure 4.5). These regions of concentrated As did not diffract, suggesting that As may complexed with OM through the formation of ternary complexes with Fe ions and/or bonding with hydroxyl groups (Caumette et al., 2012; Biswas et al., 2019). Miller et al. (2019a) attributed the presence of O-bound As in the near-surface sediment of Bulldog, Hambone, and Powder Mag lakes to 153

sorption to and/or co-precipitation with Fe-(oxy)hydroxides; however, this study demonstrates that the association of As with OMAs may account for the abundance of O-bound As species, particularly below the SWI where Fe-(oxy)hydroxides would begin to reductively dissolve (Figure 4.7).

Less commonly, As is observed co-precipitated with or sorbed to authigenic Fe-(oxy)hydroxide and Fe-sulphide minerals within the OMAs (Figure 4.5). Regions of authigenic mineralization with higher proportions of As are most commonly comprised of a mixture of ferrihydrite and goethite (Figure 4.6). At mildly acidic to circumneutral pH, As (III) and As (V) have similar affinities for Fe-(oxy)hydroxides, suggesting that As is associated with these phases through sorption, which may explain the presence of both

As valence states in the near-surface sediment of all three lakes studied (Figure 4.7; Dixit and Herring,

2003). Mackinawite mineralization is observed within these OMAs; however, As does not appear to co- precipitate or sorb to this phase but instead precipitates as fine-grained As-sulphides. These observations are in agreement with the lab-based studies conducted by Wolthers et al. (2005) who observed the formation of poorly crystalline As2S3 precipitates at the surface of FeS minerals at neutral pH.

154

4.5.3 Arsenic sequestration with TOM

Arsenic-bearing TOM is observed in the near-surface sediment of all lakes studied. Ranging in size from 20 to 200 µm, these organic macerals are primarily derived from terrestrial vegetation (i.e. roots, leaves, spores) (Figure 4.5). The modern landscape of the Tundra region consists of shrubland and small conifers with organic deposits and peat in low-lying areas; therefore, terrestrial weathering may contribute

TOM to the near-surface sediment of these lakes (Ecosystem Classification Group, 2008; 2012; Figure 4.1).

Within the organic structure of these macerals, precipitation of authigenic minerals is observed consisting of mixtures of goethite, ferrihydrite, maghemite, lepidocrocite, pyrite, and mackinawite (Figure 4.6;

Appendix C). The heterogenous nature of these grains, comprised of both sulfide and oxide precipitates may allude to their formation and stability under changing redox conditions. The direct chelation or co- precipitation of macromolecular OC-Fe structures has recently been found to play a significant role in the stabilization of OM under dynamic redox conditions. These studies suggest that vascular plant-derived aromatic and pyrogenic compounds may be preferentially preserved under changing redox conditions

(Lalonde et al., 2012; Riedel et al., 2013; Chen et al., 2015).

Combined µ-XRD and EPMA analysis demonstrates that authigenic Fe-(oxy)hydroxide and/or Fe- sulphide minerals associated with TOM commonly exceed 1 wt. % As (Figure 4.5). Micro-XRF analysis illustrates that As is generally concentrated at the centre of these grains and decreases outwards; however, within a TOM grain, no consistent patterns were observed in either the distribution of authigenic minerals or the association of As with these phases. Recent research investigating the immobilization of metal(loid)s under dynamic redox conditions suggests that the timing of authigenic mineral (i.e. FeSx) precipitation influences the effectiveness of contaminant removal (Vega et al., 2017; Du et al., 2018). For example, scavenging of dissolved As is more effective in systems where As co-precipitates with FeS in comparison to in settings where FeS forms in the sediment prior to introduction of the contaminant (Vega et al., 2017).

The high As concentrations observed in Fe-sulphide minerals associated with TOM in lakes near Tundra

Mine suggest that these sulphides precipitated authigenically in the mining-impacted lake sediments, sequestering dissolved As during their formation. Additionally, the co-precipitation of OM with Fe- 155

(oxy)hydroxides can alter their reactivity towards metal sorption at mid to low pH (Du et al., 2018). The results of the current study suggest that while the influx of reactive TOM is likely to result in As release from minerals to pore waters by enhancing reductive dissolution, it may also mediate the flux of dissolved

As to overlying surface waters by facilitating the precipitation of reactive authigenic Fe-(oxy)hydroxide and FeSx minerals in the sediments under changing redox conditions.

4.5.4 Implications for long-term environmental monitoring

The results of this study help to improve knowledge about the long-term fate of As in lake systems when increased autochthonous OM production and allochthonous OM delivery, as a result of continued climate warming, promotes a shift from oxidizing to reducing conditions in the near-surface sediment of lakes. Accumulation of labile OM will increase the sediment oxygen demand, driving the progressive onset of reducing conditions in the near-surface sediment and influence the stability of redox-sensitive elements

(i.e. Fe, S, As) and minerals (i.e. Fe-oxides and sulphides). These changing dynamics are expected to release

As to pore waters and overlying surface waters via reductive dissolution of Fe-(oxy)hydroxides or the oxidation of sulphides. However, this study demonstrates that solid phase OM may also promote the sequestration of As under transitional redox conditions by facilitating the formation of fine-grained As- sulphides, the precipitation of authigenic As-bearing FeSx/FeO minerals, and the direct sorption of As to

AOM functional groups. As a result, the diffusion of As into porewaters may also be abated by the presence of solid phase OM in near-surface sediment. However, the net effect of these competing processes is difficult to predict as the present-day mineral hosts for As in sediments vary between lakes which will, in turn, influence their long-term stability under changing redox conditions.

Based on the results of this study, it is expected that increased concentrations of aquatic and terrestrial-derived labile OM will drive the redistribution of As in near-surface sediment and result in surface-enrichment of sediment As concentrations. The natural enrichment of As in the near-surface sediment may convolute the interpretation of long-term monitoring data at former mine sites and make discerning between anthropogenic impacts and the influence of current warming trends in the sub-Arctic difficult. 156

4.6 Limitations and Future Work

Sorption and co-precipitation mechanisms are highly dependent on the biogeochemical conditions of the near-surface sediment. As highlighted by Miller et al. (2019a), the biogeochemistry of Hambone,

Powder Mag, and Bulldog lake sediments is markedly different from lake to lake. These differences (i.e. metal(loid) concentrations, pH, TOC, OM composition, etc.) will influence the effectiveness of solid phase

OM in the sequestration of As. The limited sample size of this study does not allow for broad predictions about how other sub-Arctic lakes might also behave. This study demonstrates that solid phase OM plays an important role in the cycling of As in near-surface sediment and therefore warrants further study regarding the influence of seasonality, intra- and inter-lake variances, in addition to the complexation mechanisms between As (III), As (V) and solid phase OM.

4.7 Conclusions

In this study, the association between As and solid phase OM was examined to determine the influence of modern climate warming on the long-term stability of mining-derived As contamination in lakes of the Tundra Mine region, NT. Detailed geochemical and organic petrological analyses suggest that increased aquatic production in lakes and weathering of terrestrial vegetation near a former gold mine site has impacted the mobility of legacy As contamination. While increased labile OM will change redox conditions in lake sediments and release As to porewaters via reduction dissolution of As-bearing Fe oxyhydroxides (Martin and Pedersen, 2002; Miller et al., 2019b), this study demonstrates that solid phase

OM may also provide an additional substrate for As sequestration and promote the precipitation of As- bearing minerals (both oxides and sulphides) under dynamic redox conditions.

Increased aquatic-derived OM, as a result of climate warming, was observed in the near surface sediment of all lakes studied. Clumping and/or bind together of mineral matter and aquatic-derived OM results in the formation of OMAs, comprised of: AOM, particulate macerals (e.g. sporinite (spore or pollen grain), alginate (planktonic or benthic), diatom frustules, and/or algae), and both authigenic and detrital mineral matter, in the near surface sediment. Arsenic is associated with these phases, primarily in the form

157

of authigenic, fine-grained, poorly crystalline, As-sulphides (i.e. orpiment). Within these OMAs, As is also observed associated with AOM and, to a lesser extent, with authigenic FeSx/FeO precipitates (i.e. pyrite, goethite, ferrihydrite, mackinawite).

Modern climate warming has also resulted in enhanced terrestrial weathering, providing an additional source of labile OM to the near surface sediment of these mining-impacted lakes. Terrigenous- derived OM may facilitate the precipitation of reactive authigenic FeSx/FeO minerals, (i.e. lepidocrocite, ferrihydrite, goethite, mackinawite, pyrite) providing a substrate for As sequestration. The presence of these

As-bearing substrates at and below the SWI suggests the presence of TOM plays a role in stabilizing redox sensitive authigenic minerals (i.e. sulphides and Fe-(oxy)hydroxides), and resultantly As, under dynamic redox conditions.

This study improves knowledge about As mobility in lake sediments and may inform long-term monitoring of other trace metal(loid)s (i.e. Zn, Cd, Pb) as redox conditions change due to climate warming.

Based on the results of this study, it is expected that increased concentrations of aquatic and terrestrial- derived labile OM will result in the enrichment of As in the near-surface sediment, making it more difficult to discern between anthropogenic impacts and the influence of current warming trends in the sub-Arctic.

4.8 Acknowledgements

This project was jointly funded by Polar Knowledge Canada (Project# 1519-149, awarded to JMG and TR Patterson (Carleton University)), the Environmental Geoscience Program, Natural Resources

Canada (Metal Mining Project, MBP; Northern Baselines Activity, JMG), a Natural Sciences and

Engineering Research Council of Canada (NSERC) Discovery Grant (HEJ; RGPIN/03736-2016), a

NSERC Northern Research Supplement (HEJ; RGPNS/305500-2016), the NSERC Create Mine of

Knowledge (CBM, Principle Investigator: Marc Amyot, Université de Montréal) and the Northern

Scientific Training programs (CBM, Project # 306001). Field programs were conducted under Aurora

Research Institute Licence No. 15858 (J.M.G). Synchrotron-based analysis were performed at

GeoSoilEnviroCARS (GSECARS; The University of Chicago, Sector 13) and Sector 20 at the APS,

158

Argonne National Laboratory. We are grateful to Nawaf Nasser, Braden Gregory, and Crown-Indigenous

Relations Northern Affairs Canada (Murray Somers and Joel Gowman) for their in-kind support and field assistance. The authors would like to thank Agatha Dobosz (Queen’s Facility for Isotope Research (QFIR)),

Brian Joy (QFIR), Antonio Lanzirotti (GSECARS), Matthew Newville (APS) for their assistance and invaluable guidance in data acquisition and processing. We are grateful for the guidance of Alexandra

Bailey (SRK Consulting) and Anežka Borbina-Radková (Queen’s University) in the processing our synchrotron data.

159

4.9 References

3AANDC, 2013. Water Balance Monitoring at the Tundra Mine Site: 2012 Update. ACIA, 2004. Impacts of a Warming Arctic: Arctic Climate Impact Assessment. ACIA Overview report. Cambridge University Press. 140 pp. AECOM, 2015. Tundra Mine, NWT Remediation Environmental Monitoring Program: Water Quality Part L: 2014 and Historical Results AECOM, 2018. Tundra Mine Adaptive Management Plan. AECOM Available from: http://registry.mvlwb.ca/Documents/MV2016L8-0003/MV2016L8-0003%20-%20INAC- CARD%20-%20Tundra%20-%20Adaptive%20Management%20Plan%20-%20Mar8-18.pdf Albrecht, R., Sebag, D., Verrecchia, E., 2015. Organic matter decomposition: bridging the gap between Rock–Eval pyrolysis and chemical characterization (CPMAS13C NMR). Biogeochemistry 122, 101–111. https://doi.org/10.1007/s10533-014-0033-8 An, X., Huang, F., Ren, H., Wang, Y., Chen, Y., Liu, Z., Zhang, H., Han, X., 2017. Oxidative dissolution of amorphous FeS and speciation of secondary Fe minerals : Effects of pH and As (III) concentration. Chem. Geol. 462, 44–54. https://doi.org/10.1016/j.chemgeo.2017.04.025 Anawar, H., Tareq, S.M., Ahmed, G., 2013. Is organic matter a source or redox driver or both for arsenic release in groundwater? Phys. Chem. Earth 60, 49–56. https://doi.org/https://doi.org/10.1016/j.pce.2013.04.009 Andrade, C.F., Jamieson, H.E., Kyser, T.K., Praharaj, T., Fortin, D., 2010. Biogeochemical redox cycling of arsenic in mine-impacted lake sediments and co-existing pore waters near Giant Mine, Yellowknife Bay, Canada. Appl. Geochemistry 25, 199–211. https://doi.org/10.1016/j.apgeochem.2009.11.005 Barber, A., Brandes, J., Leri, A., Lalonde, K., Balind, K., Wirick, S., Wang, J., Gélinas, Y., 2017. Preservation of organic matter in marine sediments by inner-sphere interactions with reactive iron. Sci. Rep. 7, 1–10. https://doi.org/10.1038/s41598-017-00494-0 Bauer, M., Blodau, C., 2006. Mobilization of arsenic by dissolved organic matter from iron oxides, soils and sediments. Sci. Total Environ. 354, 179–190. https://doi.org/10.1016/j.scitotenv.2005.01.027 Bennett, W.W., Teasdale, P.R., Panther, J.G., Welsh, D.T., Zhao, H., Jolley, D.F., 2012. Investigating arsenic speciation and mobilization in sediments with DGT and DET: A mesocosm evaluation of oxic-anoxic transitions. Environ. Sci. Technol. 46, 3981–3989. https://doi.org/10.1021/es204484k Biswas, A., Besold, J., Sjostedt, C., Gustafsson, J.P., Scheinost, A.C., Planer-Friedrich, B., 2019. Complexation of arsenite, arsenate, and monothioarsenate with oxygen-containing functional groups of natural organic matter: An XAS Study. Environ. Sci. Technol. acs.est.9b03020. https://doi.org/10.1021/acs.est.9b03020 Blais, J.M., Kalff, J., 1995. The influence of lake morphometry on sediment focusing. Limnol. Oceanogr. 40, 582–588. https://doi.org/10.4319/lo.1995.40.3.0582

3 INAC, AANDC, CIRNAC are all acronyms for the Federal Department responsible for the Government of Canada's work in northern Canada. 160

Bostick, B.C., Fendorf, S., 2003. Arsenite sorption on troilite (FeS) and pyrite (FeS2). Geochim. Cosmochim. Acta 67, 909–921. https://doi.org/0016-7037/03 Bostick, B.C., Fendorf, S., Brown Jr., G.E., 2005. In situ analysis of thioarsenite complexes in neutral to alkaline arsenic sulphide solutions. Mineral. Mag. 69, 781–795. https://doi.org/10.1180/0026461056950288 Brandt, J.P., 2009. The extent of the North American boreal zone. Environ. Rev. 17, 101–161. https://doi.org/10.1139/A09-004 Bright, D.A., Coedy, B., Dushenko, W.T., Reimer, K.J., 1994. Arsenic transport in a watershed receiving gold mine effluent near Yellowknife, Northwest Territories, Canada. Sci. Total Environ. 155, 237– 252. https://doi.org/10.1016/0048-9697(94)90503-7 Bright, D.A., Dodd, M., Reimer, K.J., 1996. Arsenic in subarctic lakes influenced by gold mine effluent: the occurrence of organoarsenicals and ‘hidden’ arsenic. Sci. Total Environ. 180, 165–182. https://doi.org/10.1016/0048-9697(95)04940-1 Buckwalter-Davis, M., 2013. Automated Mineral Analysis of Mine Waste. MSc. Thesis. Queen's University. https://qspace.library.queensu.ca/bitstream/handle/1974/8200/Buckwalter- Davis_Martha_J_201308_MSc.pdf?sequence=1&isAllowed=y

Caumette, G., Koch, I., Moriarty, M., Reimer, K.J., 2012. Arsenic distribution and speciation in Daphnia pulex. Sci. Total Environ. 432, 243–250. https://doi.org/10.1016/j.scitotenv.2012.05.050 Carrie, J., Sanei, H., Stern, G., 2012. Standardisation of Rock-Eval pyrolysis for the analysis of recent sediments and soils. Org. Geochem. 46, 38–53. https://doi.org/10.1016/j.orggeochem.2012.01.011 CCME, 2001a. Canadian Sediment Quality Guidelines for the Protection of Aquatic Life, Environmental Protection. CCME, 2001b. Canadian Water Quality Guidelines for the Protection of Aquatic Life Water Quality Index 1.0 Technical Report. Chen, C., Dynes, J.J., Wang, J., Sparks, D.L., 2014. Properties of Fe-Organic Matter Associations via Coprecipitation versus Adsorption. Environ.Sci. Technol. 48, 13751−13759. https://doi.org/10.1021/es503669u Chen, C., Kukkadapu, R., Sparks, D.L., 2015. Influence of Coprecipitated Organic Matter on Fe2+- catalyzed transformations of ferrihydrite: Implications for carbon dynamics. Enviromental Sci. Technlogy 49, 10927−10936. https://doi.org/10.1021/acs.est.5b02448 Chen, J., Gu, B., Leboeuf, E.J., Pan, H., Dai, S., 2002. Spectroscopic characterization of the structural and functional properties of natural organic matter fractions. Chemosphere 48, 59–68 Chen, J., Wan, G., Zhang, D.D., Zhang, F., Huang, R., 2004. Environmental records of lacustrine sediments in different time scales: Sediment grain size as an example. Sci. China 47, 954–960. https://doi.org/10.1360/03yd0160 Couture, R.-M., Gobeil, C., Tessier, A., 2010. Arsenic, iron and sulfur co-diagenesis in lake sediments. Geochim. Cosmochim. Acta 74, 1238–1255. https://doi.org/10.1016/j.gca.2009.11.028 Couture, R.-M., Van Cappellen, P., 2011. Reassessing the role of sulfur geochemistry on arsenic speciation in reducing environments. J. Hazard. Mater. 189, 647–652. https://doi.org/10.1016/j.jhazmat.2011.02.029

161

Craw, D., Bowell, R.J., 2014. The Characterization of Arsenic in Mine Waste. Rev. Mineral. Geochemistry 79, 473–505. https://doi.org/https://doi.org/10.2138/rmg.2014.79.10 Davis A., Rathbone R. F., Lin R. and Quick J. C., 1990. Observations concerning the nature of maceral fluoresence alteration with time. In Advances in Organic Geochemistry 1989 (Edited by Durand B. and Brhar F.). Org. Geochem. 16, 897-906. Pergamon Press, Oxford Davis, J.A., 1982. Adsorption of natural dissolved organic matter at the oxide/water interface. Geochim. Cosmochim. Acta 46, 2381–2393. Desisto, S.L., Jamieson, H.E., Parsons, M.B., 2011. Influence of hardpan layers on arsenic mobility in historical gold mine tailings. Appl. Geochemistry 26, 2004–2018. https://doi.org/10.1016/j.apgeochem.2011.06.030 Dixit, S., Hering, J.G., 2003. Comparison of arsenic (V) and arsenic (III) sorption onto iron oxide minerals: implications for arsenic mobility. Environ. Sci. Technol. 37, 4182–4189. https://doi.org/10.1021/es030309t Du, H., Peacock, C.L., Chen, W., Huang, Q., 2018. Binding of Cd by ferrihydrite organo-mineral composites : Implications for Cd mobility and fate in natural and contaminated environments. Chemosphere 207, 404–412. https://doi.org/10.1016/j.chemosphere.2018.05.092 Ecosystem Classification Group. 2008. Ecological Regions of the Northwest Territories - Taiga Shield. Department of Environment and Natural Resources, Government of the Northwest Territories, Yellowknife, NT. viii + 146 pp. + insert map Ecosystem Classification Group. 2012. Ecological Regions of the Northwest Territories - Southern Arctic. Department of Environment and Natural Resources, Government of the Northwest Territories, Yellowknife, NT. x + 170 pp. + insert map Eisler, R., 1988. Arsenic hazards to fish, wildlife, and invertebrates: a synoptic review. Contam. Hazard Rev. 85, 1–65. Farquhar, M.L., Livens, F.R., 2002. Mechanisms of Arsenic Uptake from Aqueous Solution by Interaction with Goethite, Lepidocrocite, Mackinawite, and Pyrite : An X-ray Absorption Spectroscopy Study. Environ. Sci. Technol. 1757–1762. https://doi.org/10.1021/es010216g Fellman, J.B., Hood, E., Spencer, R.G.M., 2010. Fluorescence spectroscopy opens new windows into dissolved organic matter dynamics in freshwater ecosystems: A review. Limnol. Oceanogr. 55, 2452–2462. https://doi.org/10.4319/lo.2010.55.6.2452 Folinsbee, R.E., Moore, J.C.G., 1955. Matthews Lake, District of Mackenzie, Northwest Territories. Geological Survey of Canada, "A" Series Map 1024A. https://doi.org/10.4095/107491 Frey, K.E., McClelland, J.W., 2009. Impacts of permafrost degradation on arctic river biochemistry. J. Glaciol. 23, 169–182. https://doi.org/10.1002/hyp Galloway, J.M., Sanei, H., Patterson, R.T., Mosstajiri, T., Hadlari, T., Falck, H., 2012. Total arsenic concentrations of lake sediments near the City of Yellowknife, Northwest Territories. Geological Survey of Canada Open File 7037, 1–47

162

Golder Associates Ltd., 2016. Public Works and Government Services Canada Tundra Mine, NWT Remediation Environmental Monitoring Program: Water Quality: 2015 and Historical Results Grafe, M., Eick, M.J., Grossl, P.R., 2001. Adsorption of arsenate (V) and arsenite (III) on goethite in the presence and absence of dissolved organic carbon. Soil Sci. Soc. Am. J. 65, 1680. https://doi.org/10.2136/sssaj2001.1680 Grafe, M., Eick, M.J., Grossl, P.R., Saunders, A.M., 2002. Adsorption of arsenate and arsenite on ferrihydrite in the presence and absence of dissolved organic carbon. J. Environ. Qual. 31, 1115– 1123. https://doi.org/10.2136/sssaj2001.1680 Griffiths, K., Michelutti, N., Sugar, M., Douglas, M.S.V., Smol, J.P., 2017. Ice-cover is the principal driver of ecological change in High Arctic lakes and ponds. PLoS One 12. https://doi.org/10.1371/journal.pone.0172989 Groves, D.I., Goldfarb, R.J., Gebre-Mariam, M., Hagemann, S.G., Robert, F., 1998. Orogenic gold deposits—a proposed classification in the context of their crustal distribution and relationship to other gold deposit types. Ore Geology Reviews 13, 7–27. https://doi.org/10.1016/S0169- 1368(97)00012-7 Gu, B., Schmitt, J., Chen, Z., Liang, L., Mccarthy, J.F., 1994. Adsorption and desorption of natural organic matter on iron oxides: mechanisms and models. Enviromental Sci. Technlogy 28, 38–46. https://doi.org/10.1021/es00050a007 Guo, H., Zhang, B., Zhang, Y., 2011. Control of organic and iron colloids on arsenic partitioand transport in high arsenic groundwaters in the Hetao basin, Inner Mongolia. Appl. Geochemistry 26, 360–370. https://doi.org/10.1016/j.apgeochem.2010.12.009 Hellweger, F.L., Farley, K.J., Lall, U., Di Toro, D.M., 2003. Greedy algae reduce arsenate. Limnol. Oceanogr. 48, 2275–2288. https://doi.org/10.4319/lo.2003.48.6.2275 Hoffmann, M., Mikutta, C., Kretzschmar, R., 2013. Arsenite binding to natural organic matter: Spectroscopic evidence for ligand exchange and ternary complex formation. Environ. Sci. Technol. 47, 12165–12173. https://doi.org/10.1021/es4023317 Huang, J.H., Ilgen, G., 2006. Factors affecting arsenic speciation in environmental samples: Sample drying and storage. Int. J. Environ. Anal. Chem. 86, 347–358. https://doi.org/10.1080/03067310500227878 INAC, 2005. Tundra Mine, NWT Environmental Monitoring Program: Water Quality Part D: 2005 Results. Prepared for: Contaminants and Remediation Directorate. Indian and Northern Affairs Canada. http://registry.mvlwb.ca/Documents/MV2005X0031/PartD-WaterQualityMon-Nov05.pdf. International Committee for Coal Petrology (ICCP), 1971. International Handbook of Coal Petrography, 2nd ed. (1963); suppl. To 2nd ed.: Centre Nat. de Rech. Sci., Paris

163

Islam, F.S., Gault, A.G., Boothman, C., Polya, D.A., Chamok, J.M., Chatterjee, D., Lloyd, J.R., 2004. Role of metal-reducing bacteria in arsenic release from Bengal delta sediments. Nature 430, 68–71. https://doi.org/10.1038/nature02638 Jamieson, H.E., Maitland, K.M., Oliver, J.T., Palmer, M.J., 2017. Regional distribution of arsenic in near- surface soils in the Yellowknife area. NWT Open File 2017-03.

Karunaratne, K. C. 2011. A field examination of climate-permafrost relations in continuous and discontinuous permafrost of the Slave geological province. Doctoral dissertation. Carleton University, Ottawa, Canada. https://curve.carleton.ca/a9c9335d-e1d8-4ffe-a772-4701ab6c2f89 Keil, R.G., Mayer, L.M., 2013. Mineral Matrices and Organic Matter. Treatise Geochemistry Second Ed. 12, 337–359. https://doi.org/10.1016/B978-0-08-095975-7.01024-X Kirk, M.F., Holm, T.R., Park, J., Jin, Q., Sanford, R.A., Fouke, B.W., Bethke, C.M., 2004. Bacterial sulfate reduction limits natural arsenic contamination in groundwater. Geology 32, 953–956. https://doi.org/10.1130/G20842.1 Kirk, M.F., Roden, E.E., Crossey, L.J., Brealey, A.J., Spilde, M.N., 2010. Experimental analysis of arsenic precipitation during microbial sulfate and iron reduction in model aquifer sediment reactors. Geochim. Cosmochim. Acta 74, 2538–2555. https://doi.org/10.1016/j.gca.2010.02.002 Kleber, M., Eusterhues, K., Keiluweit, M., Mikutta, C., Mikutta, R., Nico, P.S., 2015. Mineral – Organic Associations: Formation, Properties, and Relevance in Soil Environments, Advances in Agronomy. Elsevier Ltd. https://doi.org/10.1016/bs.agron.2014.10.005 Lafargue, E., Marquis, F., Pillot, D., 1998. Rock-Eval 6 Applications in hydrocarbon exploration, production, and soil contamination studies. Rev. l’Institut Français du Pétrole 53, 421–437. https://doi.org/10.2516/ogst:1998036 Lalonde, K., Mucci, A., Ouellet, A., Gélinas, Y., 2012. Preservation of organic matter in sediments promoted by iron. Nature 483, 198–200. https://doi.org/10.1038/nature10855 Langner, P., Mikutta, C., Kretzschmar, R., 2011. Arsenic sequestration by organic sulphur in peat. Nat. Geosci. 5, 66–73. https://doi.org/10.1038/ngeo1329 Le Pape, P., Blanchard, M., Brest, J., Boulliard, J.C., Ikogou, M., Stetten, L., Wang, S., Landrot, G., Morin, G., 2017. Arsenic Incorporation in Pyrite at Ambient Temperature at Both Tetrahedral S- Iand Octahedral FeIISites: Evidence from EXAFS-DFT Analysis. Environ. Sci. Technol. 51, 150– 158. https://doi.org/10.1021/acs.est.6b03502 Lopez, A.R., Funk, D.H., Buchwalter, D.B., 2017. Arsenic (V) bioconcentration kinetics in freshwater macroinvertebrates and periphyton is influenced by pH. Environ. Pollut. 224, 82–88. https://doi.org/10.1016/j.envpol.2016.12.066 Lowers, H.A., Breit, G.N., Foster, A.L., Whitney, J., Yount, J., Uddin, M.N., Muneem, A.A., 2007. Arsenic incorporation into authigenic pyrite, Bengal Basin sediment, Bangladesh. Geochim. Cosmochim. Acta 71, 2699–2717. https://doi.org/10.1016/j.gca.2007.03.022 Lynch, J., 1990. Provisional Elemental Values for Eight New Lake Sediment and Stream Sediment Reference STSD-3 and STSD-4”. Geostand. Newsl. 14, 153–167. Macdonald, R.W., Harner, T., Fyfe, J., 2005. Recent climate change in the Arctic and its impact on contaminant pathways and interpretation of temporal trend data. Sci Total Env. 342, 5–86. https://doi.org/S0048-9697(04)00838-1 [pii]\n10.1016/j.scitotenv.2004.12.059

164

Maclean, L.C., Tyliszcak, T., Gilbert, P.U.P., Zhou, D., Pray, T., Onstott, T., Southam, G., 2008. A high- resolution chemical and structural study of framboidal pyrite formed within a low-temperature bacterial biofilm. Geobiology 6, 471–480. https://doi.org/10.1111/j.1472-4669.2008.00174.x Martin, A.J., Pedersen, T.F., 2002. Seasonal and interannual mobility of arsenic in a lake impacted by metal mining. Environ. Sci. Technol. 36, 1516–1523. https://doi.org/10.1021/es0108537 McGuire, A.D., Anderson, L.G., Christensen, T.R., Dallimore, S., Guo, L., Hayes, D.J., Heimann, M., Lorenson, T.D., Macdonald, R.W., Roulet, N., 2009. Sensitivity of the carbon cycle in the Arctic to climate change. Ecol. Monogr. 79, 523–555. https://doi.org/10.1890/08-2025.1 Meyers, P.A., Ishiwatari, R., 1993. Lacustrine organic geochemistry - an overview of indicators of organic matter sources and diagenesis in lake sediments. Org. Geochem. 20, 867–900. https://doi.org/10.1016/0146-6380(93)90100-P Michelutti, N., Wolfe, A.P., Vinebrooke, R.D., Rivard, B., Briner, J.P., 2005. Recent primary production increases in arctic lakes. Geophys. Res. Lett. 32, 1–4. https://doi.org/10.1029/2005GL023693 Miller, C.B., Parsons, M.B., Jamieson, H.E., Swindles, G.T., Nasser, N.A., Galloway, J.M., 2019a. Lake- specific controls on the long-term stability of mining-related, legacy arsenic contamination and geochemical baselines in a changing northern environment, Tundra Mine, Northwest Territories, Canada. Appl. Geochemistry 109, 104403. https://doi.org/10.1016/j.apgeochem.2019.104403 Miller, C.B., Parsons, M.B., Jamieson, H.E., Ardakani, O.H., Gregory, B.R.B., Galloway, J.M., 2019b. Influence of late-Holocene climate change on the solid-phase speciation and long-term stability of arsenic in sub-Arctic lake sediments. In press, Science of the Total Environment. Moon, E.M., Peacock, C.L., 2012. Adsorption of Cu (II) to ferrihydrite and ferrihydrite – bacteria composites : Importance of the carboxyl group for Cu mobility in natural environments. Geochim. Cosmochim. Acta 92, 203–219. https://doi.org/10.1016/j.gca.2012.06.012 Mostofa, K.M.., Liu, C., Feng, X., Yoshioka, T., Vione, D., Pan, X., Wu, F., 2009. Complexation of Dissolved Organic Matter with Trace Metal Ions in Natural Waters in Photobiogeochemistry of Organic Matter, Biogeochemistry. https://doi.org/10.1007/978-3-642-32223-5 Newville, M. 2019. Larch: Data Analysis Tools for X-ray Spectroscopy. Version 0.9.46. https://xraypy.github.io/xraylarch/ Outridge, P.M., Sanei, H., Courtney Mustaphi, C.J., Gajewski, K., 2017. Holocene climate change influences on trace metal and organic matter geochemistry in the sediments of an Arctic lake over 7,000 years. Appl. Geochemistry 78, 35–48. https://doi.org/10.1016/j.apgeochem.2016.11.018 Palmer, M.J., Galloway, J.M., Jamieson, H.E., Patterson, R.T., Falck, H., Kokelj, S. V, 2015. The concentration of arsenic in lake waters of the Yellowknife area. NWT Open File 2015-06 29. https://doi.org/10.13140/RG.2.1.2582.5041 Palmer, M.J., Chételat, J., Richardson, M., Jamieson, H.E., Galloway, J.M., 2019. Seasonal variation of arsenic and antimony in surface waters of small subarctic lakes impacted by legacy mining pollution near Yellowknife, NT, Canada. Sci. Total Environ. 684, 326–339. https://doi.org/10.1016/j.scitotenv.2019.05.258 Parsons, M.B., LeBlanc, K.W.G., Hall, G.E.M., Sangster, A.L., Vaive, J., E, Pelchat, P., 2012. Environmental geochemistry of tailings, sediments and surface waters collected from 14 historical gold mining districts in Nova Scotia. Geological Survey of Canada, Open File Report 7150. https://doi.org/10.4095/291923

165

Parsons, M.B., Jamieson, H.E., Miller, C.B., and Galloway, J.M., 2019. Choosing an appropriate digestion protocol for environmental risk assessments of mineralized earth materials; Geological Survey of Canada, Scientific Presentation 97, 1 poster. https://doi.org/10.4095/314499 Pirrie, D., Rollinson, G.K., 2011. Unlocking the applications of automated mineral analysis. Geol. Today 27, 226–235. https://doi.org/10.1111/j.1365-2451.2011.00818.x Planer-Friedrich, B., Wallschläger, D., 2009. A critical investigation of hydride generation-based arsenic speciation in sulfidic waters. Environ. Sci. Technol. 43, 5007–5013. https://doi.org/10.1021/es900111z Prescher, C., Prakapenka, V.B., 2015. DIOPTAS : a program for reduction of two- dimensional X-ray diffraction data and data exploration. High Press. Res. 7959. https://doi.org/10.1080/08957959.2015.1059835 Prowse, T., Alfredsen, K., Beltaos, S., Bonsal, B., Duguay, C., Korhola, A., McNamara, J., Pienitz, R., Vincent, W.F., Vuglinsky, V., Weyhenmeyer, G.A., 2011. Past and future changes in arctic lake and river ice. Ambio 40, 53–62. https://doi.org/10.1007/s13280-011-0216-7 PS Analytical, 2018. Available from: https://www.psanalytical.com/ Qu, C., Chen, W., Hu, X., Cai, P., Chen, C., Yu, X., 2019. Heavy metal behaviour at mineral-organo interfaces : Mechanisms, modelling and influence factors. Environ. Int. 131. https://doi.org/10.1016/j.envint.2019.104995 Ransom, A., Robb, M., 1986. The Salmita gold deposit, Courageous Lake, Northwest Territories. In: L.A. Clark and D.R. Francis (Editors), Gold in the Western Shield. Can. Inst. Min. Met., 38: 285-305

Rantala, M. V., Nevalainen, L., Rautio, M., Galkin, A., Luoto, T.P., 2016. Sources and controls of organic carbon in lakes across the subarctic treeline. Biogeochemistry 129, 235–253. https://doi.org/10.1007/s10533-016-0229-1 Redman, A.D., 2002. Natural Organic Matter Affects Arsenic Speciation and Sorption onto Hematite. Environ. Sci. Technol. 36, 2889–2896. https://doi.org/10.1021/es0112801 Reyes, J., Goodarzi, F., Sanei, H., Stasiuk, L.D., Duncan, W., 2006. Petrographic and geochemical characteristics of organic matter associated with stream sediments in Trail area British Columbia, Canada. Int. J. Coal Geol. 65, 146–157. https://doi.org/10.1016/j.coal.2005.04.016 Riedel, T., Zak, D., Biester, H., Dittmar, T., 2013. Iron traps terrestrially derived dissolved organic matter at redox interfaces. Proc. Natl. Acad. Sci. 110, 10101–10105. https://doi.org/10.1073/pnas.1221487110 Ritter, K., Aiken, G.R., Ranville, J.F., Bauer, M., Macalady, D.L., 2006. Evidence for the aquatic binding of arsenate by natural organic matter-suspended Fe(III). Environ. Sci. Technol. 40, 5380–5387. https://doi.org/10.1021/es0519334 Rönnhult, T., Brox, B., Fritze, G., 1987. The influence of surface topography on the x‐ray intensity in electron microprobe analysis (EDS/WDS). Scanning 9, 81–87. https://doi.org/10.1002/sca.4950090205 Sanei, H., Stasiuk, L.D., Goodarzi, F., 2005. Petrological changes occurring in organic matter from recent lacustrine sediments during thermal alteration by Rock-Eval pyrolysis. Org. Geochem. 36, 1190– 1203. https://doi.org/10.1016/j.orggeochem.2005.02.009

166

Schuh, C.E., Jamieson, H.E., Palmer, M.J., Martin, A.J., 2018. Applied Geochemistry Solid-phase speciation and post-depositional mobility of arsenic in lake sediments impacted by ore roasting at legacy gold mines in the Yellowknife area, Northwest Territories, Canada. Appl. Geochemistry 91, 208–220. https://doi.org/10.1016/j.apgeochem.2017.10.025 Schuh, C.E., Jamieson, H.E., Palmer, M.J., Martin, A.J., Blais, J.M., 2019. Controls governing the spatial distribution of sediment arsenic concentrations and solid-phase speciation in a lake impacted by legacy mining pollution. Sci. Total Environ. 654, 563–575. https://doi.org/10.1016/j.scitotenv.2018.11.065 Seabridge Gold Inc., 2010. Courageous Lake Project Description in Support of a Class “ A ” Land Use Permit (Former Permit MV2003C0050). Sharma, P., Rolle, M., Kocar, B.D., Fendorf, S., Kapppler, A., 2010. Influence of Natural Organic Matter on As Transport and Retention. Environ. Sci. Technol. 45, 546–553. https://doi.org/10.2134/jeq2003.1393 Silke, R., 2001. Tundra Mine Heritage Project 1–15. http://www.geocities.ws/silkeryan/Tundra.html Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochemistry 17, 517–568. https://doi.org/10.1016/S0883-2927(02)00018-5 Smol, J. P., 1988. Paleoclimate proxy data from freshwater Arctic diatoms. Verhandlungen Internationale Vereinigung fur theoretische und angewandte Limnologie, 23: 837–844. Smol, J.P., Wolfe, A.P., Birks, H.J.B., Douglas, M.S. V., Jones, V.J., Korhola, A., Pienitz, R., Ruhland, K., Sorvari, S., Antoniades, D., Brooks, S.J., Fallu, M.-A., Hughes, M., Keatley, B.E., Laing, T.E., Michelutti, N., Nazarova, L., Nyman, M., Paterson, A.M., Perren, B., Quinlan, R., Rautio, M., Saulnier-Talbot, E., Siitonen, S., Solovieva, N., Weckstrom, J., 2005. Climate-driven regime shifts in the biological communities of arctic lakes. Proc. Natl. Acad. Sci. 102, 4397–4402. https://doi.org/10.1073/pnas.0500245102 Spence, G.H., Le Heron, D.P., Fairchild, I., 2016. Sedimentological perspectives on climatic, atmospheric and environmental change in the Neoproterozoic Era. Sedimentology 253–306. https://doi.org/10.1111/sed.12261 Staples, R., 2014. Tundra Mine, NWT Remediation Environmental Monitoring Program: Water Quality Part L: 2013 and Historic Results Stern, G.A., Macdonald, R.W., Outridge, P.M., Wilson, S., Chételat, J., Cole, A., Hintelmann, H., Loseto, L.L., Steffen, A., Wang, F., Zdanowicz, C., 2012. How does climate change influence arctic mercury? Sci. Total Environ. 414, 22–42. https://doi.org/10.1016/j.scitotenv.2011.10.039 Stuckey, J.W., Schaefer, M. V., Kocar, B.D., Benner, S.G., Fendorf, S., 2016. Arsenic release metabolically limited to permanently water-saturated soil in Mekong Delta. Nat. Geosci. 9, 70–76. https://doi.org/10.1038/ngeo2589 Sundman, A., Karlsson, T., Sjoberg, S., Persson, P., 2014. Complexation and precipitation reactions in the ternary As (V)–Fe (III)–OM (organic matter) system. Geochim. Cosmochim. Acta 145, 297–314. https://doi.org/10.1016/j.gca.2014.09.036 Tessier, A., Fortin, D., Belzile, N., DeVitre, R.R., Leppard, G.G., 1996. Metal sorption to diagenetic iron and manganese oxyhydroxides and associated organic matter: Narrowing the gap between field and laboratory measurements. Geochim. Cosmochim. Acta 60, 387–404. https://doi.org/10.1016/0016- 7037(95)00413-0

167

Tetra Tech Wardrop, 2012. Courageous Lake Prefeasibility Study. Tetra Tech Wardrop Available from: https://secure.kaiserresearch.com/i/jk/tr16/TRSEA20120905.pdf Tissot B.P., Welte D.H. (1978) Early Transformation of Organic Matter: The Diagenetic Pathway from Organisms to Geochemical Fossils and Kerogen. In: Petroleum Formation and Occurrence. Springer, Berlin, Heidelberg. Toevs, G.R., Morra, M.J., Polizzotto, M.L., Strawn, D.G., Bostick, B.C., Fendorf, S., 2006. Metal(loid) diagenesis in mine-impacted sediments of lake Coeur d’Alene, Idaho. Environ. Sci. Technol. 40, 2537–2543. https://doi.org/10.1021/es051781c Thompson, P.H., Kerswill, J.A., 1994. Preliminary Geology of the Winter Lake - Lac de Gras Area, District of MacKenzie, Northwest Territories, Geological Survey of Canada, Open File 2740 (revised), scale 1:250000. URS, 2005. Geochemical Assessment of Acid Rock Drainage and Metal Leaching Potential of Tailings and Waste Rock, Tundra Mine, NWT. Van Den Berghe, M.D., Jamieson, H.E., Palmer, M.J., 2018. Arsenic mobility and characterization in lakes impacted by gold ore roasting, Yellowknife, NWT, Canada. Environ. Pollut. 234, 630–641. https://doi.org/10.1016/j.envpol.2017.11.062 Vega, A.S., Planer-Friedrich, B., Pastén, P.A., 2017. Arsenite and arsenate immobilization by preformed and concurrently formed disordered mackinawite (FeS). Chem. Geol. 475, 62–75. https://doi.org/10.1016/j.chemgeo.2017.10.032 Wagemann, R., Snow, N.B., Rosenberg, D.M., Lutz, A., 1978. Arsenic in sediments, water and aquatic biota from lakes in the vicinity of Yellowknife, Northwest Territories, Canada. Arch. Environ. Contam. Toxicol. 7, 169–191. https://doi.org/10.1007/BF02332047 Wang, S., Mulligan, C.N., 2006. Effect of natural organic matter on arsenic release from soils and sediments into groundwater. Environ. Geochem. Health 28, 197–214. https://doi.org/10.1007/s10653-005-9032-y Wang, Y., Le Pape, P., Morin, G., Asta, M., King, G., Bartova, B., Suvorova, E., Frutschi, M., Ikogou, M., Hoai, V., Pham, C., Le Vo, P., Herman, F., Charlet, L., Bernier-Latmani, R., 2018. Arsenic speciation in Mekong Delta sediments depends on their depositional environment. Environ. Sci. Technol. 52, 3431-3439. https://doi.org/10.1021/acs.est.7b05177 Wilkin, R.T., Barnes, H.L., 1997. Formation processes of framboidal pyrite. Geochim. Cosmochim. Acta 61, 323–339. https://doi.org/10.1016/S0016-7037(96)00320-1 Williamson, C.E., Overholt, E.P., Pilla, R.M., Leach, T.H., Brentrup, J.A., Knoll, L.B., Mette, E.M., Moeller, R.E., 2015. Ecological consequences of long- term browning in lakes. Nat. Publ. Gr. 1–10. https://doi.org/10.1038/srep18666 Wolthers, M., Charlet, L., Van Der Weijden, C.H., Van Der Linde, P.R., Rickard, D., 2005. Arsenic mobility in the ambient sulfidic environment: Sorption of arsenic (V) and arsenic (III) onto disordered mackinawite. Geochim. Cosmochim. Acta 69, 3483–3492. https://doi.org/10.1016/j.gca.2005.03.003

168

Chapter 5

Exploring different ways of knowing: Braiding scientific data and Indigenous

knowledge in the geological sciences – A graduate student’s perspective

“Where do you begin telling someone their world is not the only one” – Lee Maracle

5.1 The Journey

As scientists and engineers, the objective of our work is to better understand dynamic earth systems to ultimately improve our knowledge and benefit communities. A critical, and too often overlooked, component in the understanding of remote and complex environments lies in the knowledge and memory of the people who intimately rely on this land to support their way of life. This knowledge, accumulated by a community over generations and based on keen observation, is known as traditional knowledge (TK), traditional ecological knowledge (TEK), and Inuit Quajimajatuqanit (IQ). These knowledge systems provide crucial information which can be used to fill the gaps provided by the limitations of analytical science (e.g. budgets, scope of work, sample sizes, deadlines, positivist framework) and improve our understanding of the World we all share. Specifically, as geologists, we estimate the value of land (i.e. resource estimates) and calculate potential risk factors based on samples and metrics. The purpose of this journey was to explore different types of knowledge which can’t be added to our spreadsheets.

5.1.1 Learning on-the-land

To develop a broader understanding of past, present, and potential future environmental changes in northern environments and the implications on both natural systems and the people of the Canadian

North, our collaborative research group (Carleton University, Queen’s University, Geological Survey of

Canada) participated in an on-the-land camp (August 2017) in Elu Inlet, Nunavut (NU). Led by Inuit elders, this week-long camp provided the opportunity for our team of analytical scientists to better

169

understand the culture and people of this area of research interest and resultantly, develop a more holistic understanding of the implications of our research to the peoples of Canada’s North. Most importantly, this experience provided a unique opportunity to see the world through a different lens.

Through language lessons, daily excursions, and storytelling over meals of fresh Arctic char, we were privileged to experience a brief but intimate insight into the beautiful and complex culture of the

Inuit people. Our days began with basic language lessons of one of many Inuit dialects, Nattilingmiutut.

As we struggled through an introduction to grammar, structure, and pronunciation, we learned of this culture’s core values, their perspective and relationship with their surroundings, and strong desire to both preserve and share their culture. Walking along hunting routes used by the community for generations, we learned of the culture’s resilience and intimate connection with this ever-changing northern landscape. On these walks, we followed paths marked by rock arrows which guided hunters to the next fish cache and onto rich hunting grounds (Figure 5.2 H). As we walked over meadows and past lakes, we stopped regularly to gather berries and roots, fish for Arctic char, and share stories over a heather fire (Figure 5.1

E, F; Figure 5.2 I). The time of day was less important in dictating our daily schedule than the lessons to be learned from that particular place. Through personal stories, we learned of the Inuit’s humble spirit, ceaseless optimism, and deep-rooted connection with family and the land. These stories also provided insight into an emotional and frustrating history of failed engagement with southern researchers and an explanation for the current distrust between northern communities and scientists (Cochran et al., 2008).

During this time on the land, we stepped away from our spreadsheets, test tubes, and positivist approach to perceiving, and slowly began to see the world around us in a very different way.

5.1.2 In the community

The Geoscience Tools for Environmental Assessment Project, of which this dissertation is a part, was designed as a collaborative research project that aimed to integrate knowledge from First Nation,

Métis, and Inuit communities; government; industry; and academia (Polar Knowledge Canada Project

#1516-149). The communication of analytical research findings in a plain-language format at community- based stakeholder workshops has been an important focus of this project from its inception. Through 170

these workshops, our research group aimed to foster effective collaboration between researchers,

Indigenous elders, mining companies, and local community members. In addition to these workshops, this project also focused on engagement with the North Slave Métis Alliance, Yellowknives Dene First

Nation, Tłįchǫ Research and Training Institute, and Hadlari Consulting Ltd., an Inuit-owned-and-operated business. The objective of this engagement was to generate a collection of knowledge that incorporates both western-based geoscience and human contextual information to develop a deeper understanding of climate change in northern Canada. The collection and curation of TEK/IQ was outside the scope of my doctoral research; however, the opportunity to participate in one-on-one meetings between lead researchers and community representatives and observe the development of this project has provided me, as an early career researcher, with some of the crucial skills required for developing pertinent and appropriate projects in direct collaboration with the Indigenous peoples of the North. An overview of the approach to this project and summary of key findings is provided by Galloway and Patterson (2018).

5.1.3 Back at home

Following my experiences on-the-land and participating in community-based workshops, sharing these lessons and stories seemed like a very natural next step. I reflected on these experiences and worked to introduce the next generation of geological scientists and engineers to the concepts of “different ways of knowing” through several undergraduate lectures. Building on the keen interest of both the undergraduate students and faculty, I worked with the Department of Geological Sciences and Geological

Engineering (GSGE), Knowledge Holders at Queen’s University, and a Senior Advisor for Indigenous

Knowledge from industry to develop a workshop to introduce and explore concepts in meaningful

Indigenous community engagement in the fields of geological sciences and geological engineering.

Through this workshop, faculty and students were introduced to TK, examined methods of effective engagement and collaboration, and explored the next steps of braiding Indigenous and scientific knowledge in the Department of GSGE at Queen’s University (Figure 5.2 K, L).

171

Figure 5.1 Photos of research group participating in an on-the-land camp in Nunavut (A to F); A & B) Camp set-up at Elu Inlet, NU; C & D) Daily language lessons; E & F) Learning to cook over a heather fire and clean fish.

172

Figure 5.2 Photos of research group participating in an on-the-land camp in Nunavut (G to J) and set-up for Department of Geological Sciences and Geological Engineering TEK workshop at Queen’s University (K & L); G) Elder working with a younger Inuk to set a fishing net; H) Rocks that mark hunting and trapping pathways; I) Learning about edible plants on one of our daily walks; J) Participating in Inuit games; K & L) Medicine wheel and tobacco placed at the centre of the workshop circle.

173

5.2 The Lessons and Stories

Land (nuna), lake (tahiq), rock (ujarak), person (inuk) … endless similarities exist between analytical geoscience and TK; however, a stark contrast remains between the interpretation and application of the knowledge derived from the relationships between each of these entities. To begin braiding the knowledge from both of these worlds, we must first start to not only understand one another but also to acknowledge that our way of knowing is not the only one. This section highlights and reflects on just a few of the many lessons these experiences have provided to me and hopefully will provide a foundation for better understanding to build on for the future. It is very important to note that these teachings come from Inuit, Algonquin, Anishinaabe, and Haudenosaunee backgrounds and a lesson learned from one experience may not necessarily reflect the perspective of other Indigenous groups.

5.2.1 Learning to Listen

A lesson I have learned time and time again over these past three years (and one I still struggle with) is learning to listen. Not ask questions, not provide critique, debate or discourse … to just listen.

Traditional knowledge is developed through constant observation of the environment and requires taking in information through all the senses. As speech is not one of these senses, if it used too frequently or at inappropriate times, it may distract the senses from acquiring pertinent information.

The story about how the jackrabbit got his long ears was passed on to me by a Métis-

Algonquin Knowledge Keeper at the very start of this journey. This story centres around the unhappiness the jackrabbit causes with his careless answers to the animals' questions and shares lessons about the importance of listening and not just hearing. The purpose of sharing this story was to articulate the history of the relationship between Indigenous groups and researchers in which we, as the researchers, have made a habit of consulting communities but in the end drawing conclusions based on what we think the answers are or should be.

In many Indigenous cultures, a talking stick is used during group meetings to ensure a code of conduct of respect is followed. In both the Algonquin and Ojibwe cultures, an eagle feather is used and

174

the person holding the feather is given the space to speak while others listen quietly and respectfully. This practice ensures that all voices are shared equally, that the insight and wisdom of quieter personalities are heard, and guards against assertive personalities having advantage through the dominance of their voice.

Researchers could benefit from borrowing a feather from this cap…

5.2.2 Language

“Language is the dwelling place of ideas that do not exist anywhere else. It is a prism through which to

see the world.”- Robin Wall Kimmerer

Before you can understand a culture and its people, you must understand its language, as the foundations of language provide the purest insight to a culture’s world perspective. The structure of a language provides a basis for understanding a people’s values and relationship to the world around them.

This concept is illustrated by the Nattilingmiutut dialect of the Inuit language. Nattilingmiutut is structured in a way which does not detach the speaker from the object or place to which they are referring.

All observations include the observer as they are simply a living being amongst other living beings (i.e. animals, plants, etc.). This structure creates an innate sense of belonging that would be impossible to feel or know if the observer is detached from what is being observed. For example, in Inuit, there is no word for sister in a detached sense. The Inuit word would change according to the gender of the speaker and the relationship with the speaker (younger or older). From this perspective, there is no “I”, but there is a “we” or an “us”. A language is so much more than a means of daily communication, the words and structure of a language also communicate the fundamental values of a culture. When translated, as Indigenous languages commonly are for research and academic’s sake, many of these nuances are lost and with them go the ability to meaningfully engage and braid two vastly different ways of knowing.

The deep connection of people to their family and environment is also beautifully illustrated in the way in which communities introduce themselves at a gathering or meeting. In this approach, all members introduce themselves first as people. For example, I am Clare Miller, I am a partner, daughter, sister, aunt, explorer, athlete, and a geochemist. This approach dissolves any hierarchy which comes from first stating your position or title (i.e. Dr., CEO, President) and again creates a sense of humble belonging 175

and interconnectedness between people. This approach also instills a deep sense of responsibility to preserve those relationships and engrains the mutualistic connection between people and the land.

5.2.3 Timescales

Traditional knowledge is a cumulative body of knowledge, practice, and belief, handed down through generations, over thousands of years, through lessons and oral traditions (Alexander et al., 2011).

This legacy of knowledge, acquired through adaptation to the surrounding environment, is used to inform decision making for the well-being of generations to come. These decisions are made based on the approach that actions made today should benefit the children many generations into the future. This time scale is vastly different than the approach taken when planning a research project or resource development (i.e. the life cycle of a mine) and creates an immediate disconnect between us and the people of the communities in which we are proposing to work. This disconnect is exemplified in the approach to mining and remediation at Giant Mine, Yellowknife, NT, where current plans include storing contaminants underground in perpetuity. In the minds of the local community, the solutions proposed by western consultants are too short-sighted and communication of the perpetual risk relies not only on our ability to read spreadsheets and reports, but also on the assumption that there will be a government and budget maintaining the site for eternity. As a result, the Yellowknives Dene First Nation have created a modern-day story, the Guardians of Eternity, which they hope will communicate and preserve knowledge through a means that has and will continue to endure technological and societal changes (Benoit et al.,

2015).

Similarly, communication of knowledge through oral tradition and the sharing of stories occurs at very different timescales compared to giving a lecture or scheduling a meeting. To have the privilege of learning from this body of knowledge requires first building relationships and trust within the community which takes more time and different skills than setting up a purchase order or attending a conference.

Providing the resources and skills for our students, faculty, and employees to begin building these meaningful relationships will be an essential component to making these first steps.

176

5.3 Finding Common Ground

None of the learnings outlined above are novel, and the mutual benefits and challenges in braiding of western science and TK systems have been well studied (Berkes 1993, 1999; Ingold 2000).

However, these experiences have shown me that while a chasm exists between the worlds of geoscience and TEK, the foundation of “learning from the land” provides an essential common ground and a perfect starting point to begin understanding one another. This section aims to outline some of the common ground between these knowledge systems and provide possible first steps for us, as scientists and engineers, to begin bringing these two worlds together and continuing this journey side by side.

The most crucial part of this journey is that each knowledge system is equally respected, one does not replace the other, and one voice does not drown out the other. We must acknowledge that any attempt to join scientific and Indigenous knowledge systems may reflect the history of power relationships between

Indigenous groups and scientists (Simpson, 2004). The Two-Row Wampum, a treaty between the

Haudenosaunee and early Dutch settlers, provides insightful guidance and perspective to begin this journey. In this agreement, both groups travel down the same river together, but each in their respective boats. Neither group was to try to steer the other’s vessel or interfere in the internal affairs of the other. A pathway towards meaningful collaboration only begins with a willingness to respect the methods, voices, and perspectives of both world views. One of the most valuable lessons I have learned along this journey is the importance of allowing each collaborating group the space and resources required to define their own methods, objectives, and priorities. As Queen’s University is situated on Haudenosaunee territory, starting this journey by honouring the Two-Row Wampum treaty and passing this knowledge on to the next generation of scientists and engineers seems fitting.

As geological scientists, we look to environmental systems (i.e. rocks, water, trees) to tell us about the evolution of the Earth and have the unique ability to perceive larger time scales than many western scientists. However, by seeing the environment as atoms and molecules to be studied, we are missing vital information to help us fully understand these complex natural systems and the impacts of

177

our research or resource development on the well-being of the land and the people of the area. The braiding of western geoscience and TEK can has the potential to contribute invaluable information and generate insight into climate change, land-use change, changes in seasonality, and spatial extent of contamination (Alexander et al., 2011; North Slave Métis Alliance community members et al., 2017;

Galloway and Patterson, 2018). However, and more importantly, braiding these two knowledge systems provides a new perspective and ways of learning for the geological scientists and engineers responsible for the development of a mine, tailings pond, or dam. By spending time on the land and learning the language of a local culture, we, as geological scientists and engineers, may begin to see the interactions between these natural systems through different eyes and develop a more holistic understanding of the intricate balance which cannot be quantified through equilibrium constants or mass-balance equations.

5.4 Barriers

This journey will not be without challenges, many of which will be unique to specific projects or situations. This section outlines just a few of the barriers and challenges which I have encountered along this short journey with the aim of providing some reflection and insight as we continue along this path.

Geology, as a discipline, prides itself on the rigor and validity of peer-reviewed scientific studies, statistical analysis, and stringent monitoring of independent and dependent variables which can be quantitatively assessed. These metrics are a construct of western science and do not exist within the TEK realm which is instead based on generations of knowledge acquired through adaptation and survival in a particular environment. The perceived absence of quantifiable quality control measures in oral histories or teachings creates a major barrier for many scientists and engineers who define “truth” and “fact” on statistical significance and numerical repeatability. From my perspective as a scientist, the survival of a community is a stronger test of reliability and validity than duplicates and blanks will ever be. To effectively braid these ways of knowing, we, as scientists and engineers, need to acknowledge both ways of knowing as equally valid.

178

As an engaged, enthusiastic, eager student, my learning style has been praised in western academia. However, to many Indigenous cultures, these traits and qualities are perceived as aggressive and potentially even rude. This created a hurdle for me as I struggled to adapt to learning in a new cultural setting and fully grasp a different way of knowing. Through time and much needed guidance, I have come to realize different ways of acquiring knowledge are equally valuable. As scientists and engineers, we have become very comfortable and accustomed to a positivist approach to learning in which logic and reason provide the foundation for knowing. In this system, knowledge derived from intuition or introspection is perceived as less valid or approached with skepticism. However, as much as we hate to admit it, science and research are inherently biased by the unconscious choices of the researcher conducting the study. From the decision of which samples are collected to the interpretation of data, the results of a scientific study are influenced by the researcher’s motivations, experiences, and world perspectives. This is not a bad thing, but it is imperative that we acknowledge that our way of “knowing” is not flawless.

5.5 Moving Forward Together

5.5.1 The self

Braiding different ways of knowing, for me, will be a life-long journey which, through this experience, I have realized has always been a part of my upbringing. However small the steps may have been, I was fortunate enough to be raised in a family where “conventional learning” wasn’t the only way.

In my family learning through stories, creative play, and being in nature were as valuable if not more-so than the algebra homework we brought home that day. Reflecting on that now, these life lessons may have provided me with the opportunity to be open to new ways of being and seeing.

Writing these experiences in a formal dissertation, in itself, does not do these lessons justice.

However, as academics communicate knowledge through published documents, this provides a starting place for an important conversation. To those who have passed these lessons on to me: I promise to take these lessons beyond the pages of this thesis. I will continue to listen first and do second, to build

179

relationships and trust in my community, to be mindful of the quiet voices, and continue to learn and always make space for new perspectives.

“Don’t be afraid to be a beginner” - Robin Wall Kimmerer

5.5.2 Exploring different ways of knowing in the Department of GSGE

Through these experiences, I have learned that filling the knowledge gaps left by analytical science is more than adding to a dataset, rather, to truly have a meaningful impact as scientists and engineers we must learn to embrace different ways of knowing and perceiving the world.

Reflecting on these experiences, the objective of narrowing the chasm and braiding these knowledge systems sometimes seems insurmountable; however, these few experiences have also shown me that there is hope. From my perspective, that hope lies in our willingness to begin challenging ourselves to open our minds to different ways of knowing instead of fearing the unknown. These things are beginning at Queen’s and in the Department of Geological Sciences and Engineering. Within our department, workshops and guest lectures have brought an awareness to this subject and sparked an interest in the next generation of geological scientists and engineers. Our challenge is to continue building on them. A TEK perspective brings with it a sense of belonging and as a result, a sense of responsibility.

With the land as our common denominator, we can build on that responsibility and continue working towards a better tomorrow. Some suggestions for starting points:

· Use of Treaty maps of the Kingston area in Field Methods · Follow-up workshop for faculty and students · Budgeting for graduate students to engage with community groups in area of study · Invitation for a Knowledge Keeper to teach on-the-land in the Kingston area · Continuing to have guest lectures to speak on specialized areas · Moving away from solely lecture based teaching styles · Providing resources and encouragement for faulty to begin adapting courses or teaching styles

180

This journey does not end with the final line of this chapter. Instead, perhaps the last line represents a perfect starting point for me to really begin.

So, to you, as the academic reader: Remember your world is not the only one.

181

5.6 References

Berkes, F., 1993. Traditional Ecological Knowledge in Perspective, in: Harris, C.E., Pritchard, M.S. (Eds.), Traditional Ecological Knowledge: Concepts and Cases. International Program International Program on Traditional Ecological Knowledge and International Development and Research Centre, pp. 1–9

Berkes, F., 1999. Sacred Ecology: Traditional Ecological Knowledge and Resource Management, Second Edi. ed. Taylor & Francis

Benoit, F., Saxberg, K., Harpelle, G., Harpelle, R., 2015. Guardians of Eternity. ShebaFilms. http://www.guardiansofeternity.ca/film-team/

Cochran, P.A.L., Marshall, C.A., Garcia-Downing, C., Kendall, E., Cook, D., McCubbin, L., Gover, R.M.S. 2008. Indigenous ways of knowing: Implications for participatory research and community. Am J Public Health. 98, 22–27. https://doi.org/10.2105/AJPH.2006.093641

Galloway, J.M., Patterson, R.T. 2018. Introduction to traditional knowledge studies in support of geoscience tools for assessment of metal mining in northern Canada. Polar Knowledge: Aqhaliat 2018, Polar Knowledge Canada, 92–98. https://doi.org/10.35298/pkc.2018.12

Ingold, T. 2000. The perception of the environment: essays on livelihood, dwelling and skill. Routledge, London, UK. http://dx.doi.org/10.4324/9780203466025

Kimmer, R., 2015. Braiding Sweetgrass: Indigenous Wisdom, Scientific Knowledge and the Teachings of Plants. Milkweed Editions.

North Slave Métis Alliance community members, Shiga, S., Evans, P., King, D., and Keats, B., 2017. Continual change and gradual warming: A summary of the North Slave Métis Alliance’s recorded cultural knowledge on climate and environmental change. Report prepared for Geological Survey of Canada Geoscience Tools for Environmental Assessment of Metal Mining (compiled by Jennifer M. Galloway & R. Timothy Patterson); Polar Knowledge: Aqhaliat 2018, Polar Knowledge Canada.

Simpson, L., 2004. Anticolonial Strategies for the Recovery and Maintenance of Indigenous Knowledge. The American Indian Quarterly. 28, 373-384. https://doi.org/10.1353/aiq.2004.0107.

182

Chapter 6

Conclusions and future research

6.1 Delineating mining impacts and geochemical baseline in the CLGB

The results of this study show that the sediment geochemistry of five lakes within 5 km of the former Tundra Mine has been altered due to the influence of historical mining and subsequent terrain disturbance associated with remediation activities (Miller et al., 2019a). However, the depth of impacts and baseline sediment geochemistry are highly variable between lakes. Radiometric dating of four sediment cores demonstrates that in these lakes, changes in As, S and Fe geochemistry occur on a sub- centimetre scale. As a result, the depth control provided by sediment gravity cores (in contrast to grab samples) is essential to evaluate post-depositional mobility of contaminants and accurately determine geochemical baselines in these low sedimentation rate environments.

Environmental baseline data were not collected prior to the exploration and initial development of

Tundra Mine in 1951. Therefore, delineating mining impacts and establishing pre-mining background conditions in the Tundra Mine area is important for the long-term monitoring of legacy contamination and guiding future resource development within the CLGB. This study demonstrates that weathering leads to naturally elevated sediment As concentrations (range: 28 to 170 mg·kg−1) in some lakes of the CLGB, pre-dating mining activities. No correlation between bedrock type and elevated baseline As concentrations could be discerned; however, this may also be due to the lack of detailed bedrock geology mapping in this region and/or the limited sample size of this study. Locally, glacial till ranges from 2 to

10 m in thickness and may also provide a source of As to these lakes. In mineralized regions near gold deposits in the Yellowknife Greenstone Belt, 400 km south-east of Tundra Mine, tills contain As concentrations up to 1,560 mg·kg−1 although typical As concentrations are between 5 and 30 mg·kg−1

(Kerr, 2006).

183

Control Lake is used by Crown-Indigenous Relations and Northern Affairs Canada as a background reference for ongoing environmental monitoring activities at Tundra Mine and was selected in this study to evaluate the impacts of modern-day climate warming on geochemical baselines and metal(loid) mobility in unimpacted lakes (Golder Associates Ltd., 2005). Further investigation determined that the near-surface sediment of Control Lake has been impacted by dust generated by increased activity at the Tundra Mine site during historical mining, construction of roads and an airstrip, and recent remediation activities. The predicted impact of climate warming on geochemical baselines is discussed in Sections 6.3, 6.4, and 6.5.

6.2 Long-term stability of legacy contaminants in a changing climate, Tundra Mine, NT

In sub-Arctic regions, increases in temperature and duration of ice-free seasons, as a result of climate warming, will promote an increase in biological production and organic matter (OM) transport to aquatic environments (Frey and McClelland, 2009; Prowse et al., 2011; Stern et al., 2012; Griffiths et al.,

2017). In near-surface sediments, dissolved oxygen concentrations of sediment pore waters are largely controlled by the availability of labile OM. As a result, modern climate warming may drive the progressive onset of reducing conditions in the near-surface sediment of lakes and influence the stability of redox-sensitive elements (i.e. Fe, S, As) and minerals (i.e. Fe-oxides and sulphides). The impacts of increased OM influx on the post-depositional cycling of As in lake sediments in the Tundra Mine region is illustrated in Figure 6.1. Increased concentrations of labile OM, as a result of 21st century warming, are observed in the near-surface sediment of lakes in the Tundra Mine region (Miller et al., 2019a). This study reveals that the source of As to these lakes influences the solid-phase speciation of As in lake sediments, which in turn dictates the fate and mobility of legacy As contamination under changing climate conditions. In lakes proximal to the former mine site, As, introduced by the direct disposal of mine wastes, is hosted primarily by sulphides in the near-surface sediment. Because sulphide minerals are inherently stable under reducing conditions, increasing the flux of labile OM to the sediment may increase the stability of legacy contamination in these lakes. In sub-Arctic lakes impacted by the disposal of

184

sulphide-bearing waste rock or tailings, climate warming may therefore increase the stability of As in contaminated lake sediments over time. Conversely, in lakes farther from the former Tundra Mine, As was largely introduced to the near-surface sediment through aerial deposition of dusts, weathering of waste rock and mineralized bedrock, or suspended sediments transported by streams. In these lake sediments, As is primarily hosted by Fe-(oxy)hydroxides which are susceptible to reductive dissolution as oxygen concentrations in the near-surface sediment decrease. As a result, continued climate warming may lead to increased surface water As concentrations in these distal lakes in the future. However, the presence of trace As-bearing framboidal pyrite in these distal lakes suggests that increased loading of labile OM may also moderate the flux of As to the overlying surface waters sub-Arctic lakes through adsorption or co-precipitation.

6.3 Past sub-Arctic warming as an analogue to inform interpretation of modern-day geochemical trends

Through the high-resolution (1-cm) geochemical and sedimentological analyses of two sediment cores, this study uses paleoclimate reconstruction to better understand As cycling in modern environments and inform decision-making at legacy mine sites in sub-Arctic Canada. Paleoenvironmental analysis of lake deposits are useful for assessing natural changes in environmental systems pre-dating the influence of anthropogenic impacts. This technique, therefore, may allow us to differentiate between the cumulative effects of resource development and climate change in modern lakes.

Over the late Holocene (~ ca. 7,000 yrs. BP), the central Arctic experienced profound landscape change in response to the Holocene Thermal Maximum. This global warming event affected the physical and chemical properties of lakes, which are preserved in the sediment record and provide insights to the impacts of current warming trends on sub-Arctic lakes (Pienitz et al., 1999; Griffith, 2013; Upiter et al.,

2014; Sulphur et al., 2016). In sediment cores retrieved from Matthews and Control lakes, increased concentrations of As and S are observed in both sediment and porewaters deposited well before the onset of mining activities. These elevated concentrations coincide with maximum concentrations of labile OM

185

and occur over periods of well-documented warming events in the sub-Arctic (Figure 3.8; Figure 6.1).

This study demonstrates that increased biological production (both aquatic and terrigenous), as a result of

Holocene warming in the Tundra Mine region, enhanced both the loading and mobility of naturally derived As in lakes. The increased flux of labile OM to sediment approximately 5,000 yrs BP promoted the onset of reducing conditions in the near-surface sediment. As a result, As-bearing Fe-(oxy)hydroxide minerals deposited in the lake sediment from the weathering of mineralized bedrock underwent reductive dissolution and increased dissolved As concentrations in porewaters. However, the presence of labile OM over this time also mediates As release by facilitating the precipitation of As-bearing authigenic sulphides

(e.g. framboidal pyrite; Figure 6.1). This study suggests that climate warming creates what Schuh (2019) describes as a continuous “redox loop” that drives the release of As to porewaters through reductive dissolution and subsequently promotes its sequestration through the precipitation of authigenic sulphides.

This mechanism would tentatively lead to progressive increases in near-surface sediment As concentrations in mineralized regions of the sub-Arctic in response to climate warming.

186

Figure 6.1 Summary stratigraphic figure comparing arsenic concentrations (porewater and sediment) and selected aquatic (A) and terrigenous-derived (B) organic macerals preserved in the Matthews Lake sediment core to inferred periods of climate variability in the Courageous Lake Greenstone Belt based on regional paleoclimate reconstructions (Pienitz et al., 1999, Griffith, 2013, Upiter et al., 2014, Sulphur et al., 2016).

These findings suggest that modern warming climate conditions may result in elevated As concentrations in near-surface lake sediments and surface waters; however, limitations with this analogue approach complicate the interpretation of metal(loid) mobility in modern day systems. Most notably, the tendency for As to migrate both upwards and downwards within the sediment column post-deposition makes the interpretation of paleorecords inherently difficult. The burial and diagenetic processes which occur in lake sediment influence redox conditions and, therefore, the speciation and mobility of elements such as As, Fe, and S (Andrade et al., 2010; Couture et al., 2010; Miller et al., 2019b). As a result, the As species and concentration profiles preserved deeper in the lake record may not provide an exact analogue to mechanisms influencing As mobility in the near-surface sediment. By examining trends over a longer

187

duration (i.e. centuries) as opposed to decadal changes in the sediment record, this study attempts to provide insight into long-term trends of As speciation and mobility in response to climate warming.

While this approach provides insight into the mechanisms influencing the long-term stability of As, a notable gap exists in understanding the shorter duration changes which may help inform present day decision making at northern mine sites. Proposed approaches to addressing this knowledge gap are discussed in Section 6.7.

6.4 Role of solid phase organic matter on As mobility

Chapters 2 and 3 of this thesis demonstrate that increased labile OM leads to a progressive onset of reducing conditions in the near-surface lake sediment and highlight the effect of changing redox conditions on the stability of As-bearing minerals. Building on these findings, Chapter 4 highlights the important role of solid phase OM in mediating the flux of As under dynamic redox conditions in the near- surface sediment. This chapter builds on knowledge in Chapter 3 to address gaps identified in Chapter 2 and focuses specifically on the role of aquatic- and terrigenous-derived organic matter on the mobility of

As in three mining-impacted, sub-Arctic lakes.

The influence of increasing solid phase organic matter on the predominant mechanism influencing As-mobility in the near-surface sediment is summarized in Figure 6.2. In oxidized near- surface sediments, As is sequestered through the precipitation of metal oxyhydroxides, particularly Fe- oxyhydroxide. The onset of reducing conditions, through burial or increased abundance of labile OM, results in the release of As to porewater through reductive dissolution and subsequent sequestration through the precipitation of As-bearing sulphides (Bostick and Fendorf, 2003; O'Day et al., 2004; Lowers et al., 2007). However, the onset of reducing conditions in the near-surface sediment is progressive and results in a transitional zone of depleted O2 concentrations where As may be released to porewater through both the oxidation of As-bearing sulfides and reductive dissolution of Fe-(oxy)hydroxides.

Through combined microscale geochemical and organic matter speciation analysis, this study demonstrates that under dynamic redox conditions increased algal and terrestrial-derived OM will

188

mediate the potential flux of As to overlying surface water by facilitating the precipitation of authigenic sulphides, enhancing the stability of authigenic oxides, and allowing for the direct sequestration of As to

OM.

Figure 6.2 Schematic of the solid phase speciation and post-depositional mobility of As in lake sediment (black) and porewater (blue). Insets display the authigenic formation of (a) organo-mineral aggregates (A – authigenic FeSx/FeO grains; B – framboidal pyrite; C – detrital mineral; D – Organic macerals (i.e. sporinite, chlorophylinite, alginite); E – amorphous organic matter (AOM); F – authigenic Fe-(oxy)hydroxides; G – As-sulphides and As bound to OM matrix; H – oxidized AOM) and (b) Terrigenous organic matter (A - authigenic FeSx/FeO; B - authigenic Fe-(oxy)hydroxides; C – cellular structure of TOM; B – Fe-(oxy)hydroxides; C – oxidized cell wall; D – authigenic sulphides (pyrite, mackinawite)), which facilitate the sequestration of As under changing redox conditions.

Increased biological productivity within a lake’s watershed results in an increased accumulation of the remains of microorganisms, algae, phyto- and zoo-plankton and degraded plant materials in the near-surface sediment. The accumulation of aquatic derived OM in the lake sediment promotes the formation of OMAs, which are comprised of amorphous organic matter (AOM), particulate macerals (e.g.

189

sporinite (spore or pollen grain), alginate (planktonic or benthic), diatom frustules, and/or algae), and both authigenic and detrital mineral matter which are abundant in the near-surface sediment (Figure 6.2).

These authigenic grains provide micro-environments which allow for the sequestration of As predominantly through the formation of authigenic, fine-grained, poorly crystalline, As-sulphides (i.e. orpiment). Arsenic is also observed associated with amorphous OM in these grains, providing preliminary evidence that solid-phase organic matter may directly sequester As. Tentatively, AOM may attenuate As through bonding with hydroxyl groups and formation of ternary complexes via metal-cation bridges (i.e.

Fe (III), Al (III), Ca (II)) (Ritter et al., 2006; Sharma et al., 2010; Hoffmann et al., 2013). Terrigenous- derived OM, introduced to the sediment through weathering, also plays an important role in attenuating the mobility of As in the near-surface sediment of mining-impacted, sub-Arctic lakes. The observation of these As-bearing authigenic grains at and below the sediment-water interface (SWI) suggesting that terrigenous OM effectively stabilizes redox sensitive authigenic minerals (i.e. sulphides and Fe-

(oxy)hydroxides) and provides a reactive substrate for As sequestration under dynamic redox conditions.

The findings of this study suggest that while increased delivery of labile OM to sediment may enhance the mobility of As in the near-surface sediment as conditions shift from oxidizing to reducing, solid phase organic matter may also play a vital role in abating the flux of As to overlying waters by providing an additional substrate for sequestration.

6.5 Implications for environmental monitoring of legacy and modern mine sites

The findings of this study can be used to help guide interpretation of long-term environmental monitoring data for the nearby, and recently remediated, Tundra Mine and are also broadly applicable to environmental assessment and monitoring at other northern mine sites. This study highlights the anticipated consequences of modern-day climate warming on the loading and mobility of metal(loid)s in both natural and mining-impacted environments as well as the potential unintended impacts of remediation activities on contaminant mobility.

190

In mineralized regions of Canada’s north, continued climate warming is expected to influence geochemical baselines and the mobility of metal(loid)s in both mining-impacted and non-impacted lakes.

Based on studies of past warming events, this study demonstrates that weathering of mineralized bedrock will increase loading of metal(loid)s to lake sediment. Additionally, warming conditions in sub-Arctic regions cause enhanced aquatic production and increased weathering of terrigenous material resulting in the accumulation of detrital organic matter in the near-surface sediment. These changing dynamics will promote the progressive onset of reducing conditions, potentially resulting in the remobilization of metal(loid)s which are scavenged by redox-sensitive authigenic minerals (i.e. As, Cu, Cd, Pb, Fe, Mn, Zn) and lead to increased concentrations of these metal(loid)s at both the SWI and in overlying surface waters. The natural enrichment of metal(loid)s in surface waters and near-surface sediment may complicate the interpretation of long-term monitoring data and make it difficult to discern between anthropogenic impacts and the influence of current warming trends in the Arctic. This study demonstrates that the collection of long sediment cores (> 30 cm length) is imperative for interpreting long-term trends in geochemical baselines and controls on metal(loid) mobility in sub-Arctic lakes.

The conclusions and predictions of long-term trends in this study are largely based on the assumption that modern climate warming will continue to increase the natural loading of OM and lead to the progressive onset of reducing conditions in the near-surface sediment. These assumptions are based on the observation of increasing labile OM in the near-surface sediment as a result of 21st century climate warming and are supported by observed trends elsewhere in the Arctic (e.g. Fahnestock et al., 2019). It is important to note that there are several other climate related factors which may influence these predictions. Most pertinent of these factors is the possibility of increased oxygenation of surface waters due to longer ice-free seasons and increased flow of oxygenated surface waters to these lakes. Increasing oxygen concentration in the surface water of these lakes could result in opposing trends to those presented in this study. Under this scenario, diffusion of O2 into the near-surface sediment would increase the stability of As in lakes where Fe-(oxy)hydroxides are the primary hosts. Conversely, in lakes where sulphides are the primary host of As, oxidation would result in the release of As to surface waters. 191

However, the evidence of changing redox conditions due to an increased flux of labile OM over past warming events, provided in this study, suggests that changes in biological production is a more prominent driver of As cycling in northern lakes. This assumption is also widely supported by research on past and present climate warming events which demonstrates that ecological changes in Arctic lakes are predominantly driven by increases in ambient temperature and duration of ice-cover (i.e. Smol, 1988;

Douglas et al., 1999; Sulphur et al., 2016; Griffiths et al., 2017).

6.6 Combining solid-phase arsenic and organic matter speciation techniques in lake sediment studies

Although there is very little published data on the nature or origin of solid phase OM in lake sediments or about its role in element mobility in these settings, there is a growing consensus that solid phase OM plays an important role in the mobility of elements in other low-temperature, environmental systems (i.e. peats, deltas; Langner et al., 2011; Wang et al., 2018; Biswas et al., 2019). This study combined organic petrography techniques (i.e. Rock Eval, fluorescence microscopy) with geochemical speciation methods (i.e. SEM-based automated mineralogy, EPMA, synchrotron-based µXRF/XRD) to characterize the role of OM in controlling As mobility in mining-impacted lake sediments. The former techniques are traditionally used by the petroleum geoscientists for exploration and resource evaluation, while the latter techniques are commonly used for characterization of rocks, mineral deposits, and mine waste. While combining these techniques revealed novel observations, a number of hurdles were encountered which limited the scope of this multidisciplinary approach and provided insight to guide future studies.

Originally, we planned to combine organic petrography and geochemical speciation methods to study individual grains and provide visual evidence of the relationship between solid phase OM and the dispersion of As in mining-impacted sediment. I aimed to select authigenic As-bearing grains based on

SEM and EPMA analysis of thin sections for subsequent organic petrological and µXRF/XRD analysis.

However, upon inspection of selected grains it was determined that previous SEM and EPMA analysis on

192

these samples may have resulted in thermal maturation of the OM (Sanei and Ardakani, 2016).

Quantifying the extent of maturation may provide guidance for future studies examining OM in recent sediment (discussed in Section 6.7).

Each of these analytical methods interacts with the sample in a different way, requiring different physical set-up and sample preparation procedures. For example, fluorescence microscopy uses light to excite and fluoresce the specimen of interest resulting in the subsequent emission of characteristic wavelengths from the sample surface. To limit the passage of incident light through the sample and maximize emission, samples are prepared as 1-cm impregnated epoxy pucks. Conversely, synchrotron- based µXRD is a penetrative analysis in which X-rays pass through a sample to produce a characteristic spectrum for each grain encountered within the beam path. For this technique, thin section samples 35 to

50 μm thick produce the optimal signal-to-noise ratio for analyses. Attempts were made to prepare samples in a method which would optimize the results for both techniques. Suggestions to improve and inform these techniques are outlined in Section 6.7.

6.7 Recommendations for future research

Although the research objectives outlined in the Introduction to this thesis were largely achieved, several important questions could not be fully addressed in the scope of this thesis. General recommendations for future work based on knowledge generated in this study are discussed below:

Effects of modern-day climate warming on geochemical baselines:

• This study demonstrates that Control Lake has been impacted by dust generated during mining,

construction and/or remediation activities at Tundra Mine. Future work should examine baseline

concentrations and the influence of modern warming on As mobility in lakes further from point

sources of contamination in the CLGB.

• This study demonstrates significant inter-lake variability in geochemical baselines in sediments

from lakes in the CLGB. Future studies examining the impacts of climate warming on natural

193

lakes should include the sampling of several lakes to capture this natural variability, and ideally,

include multiple cores from individual lakes.

• Limited lithogeochemistry is available for the bedrock or overburden of the CLGB. A study of

baselines should therefore include the sampling and analysis (mineralogy and bulk geochemistry)

of bedrock, tills and soils to fully understand the natural sources of As and other metal(loid)s to

these lakes.

Paleoclimate reconstruction to study changes in metal(loid) mobility and baseline:

• Through the integration of geochemical and paleolimnological techniques, this study examined

the natural fluctuations in As concentrations from the late Holocene to present in sub-Arctic lakes

and correlated these fluxes to changes in regional climate. However, as discussed in Section 6.3,

the post-depositional mobility of redox-sensitive elements may convolute the interpretation of

downcore trends. Examining other paleoclimate indicators (e.g. changes in Arcellinidan or diatom

communities) may provide more insight as to the regional climate trends and near-surface redox

conditions at the time of maximum metal(loid) concentrations (cf. Nasser et al., 2016).

• In sub-Arctic regions, low sedimentation rates mean that 1-cm sampling resolution may represent

decades of sediment accumulation in some lakes. The use of freeze coring techniques allows for

sampling on a sub-mm scale therefore allows higher resolution studies in these regions

(Macumber et al., 2011). Freeze cores were collected alongside our gravity cores in March 2016.

Analysis of paleoclimate trends preserved in the Control Lake freeze core was conducted by

Gregory (2019) and may provide an opportunity to compare these two methods side-by-side.

Role of solid-phase OM on As mobility

• The combination of geochemical speciation (µ-XRF/XRD, XANES, SEM-based automated

mineralogy) and organic petrological techniques (fluorescence microscopy) in Chapter 3 provides

strong evidence that solid phase organic matter plays an important role in the sequestration of As

under changing redox conditions. The addition of sequential extractions, extended X-ray

194

absorption fine structure, gas chromatography-mass spectrometry, and/or nanoscale secondary

ion mass spectrometry may provide clearer insight as to the interactions and binding mechanisms

between As, OM and associated authigenic minerals (i.e. Biwas et al., 2019).

• Quantifying the influence of SEM and synchrotron-based analysis on the thermal maturation of

organic matter would allow for an improved methodology to be developed in order to combine

these techniques more effectively in future studies.

• Experimentation to optimize sample preparation which would permit the analysis of an individual

grain using each of these techniques sequentially would limit the assumptions required when

combining these techniques in the current study.

• Use of alternative sample preparation methods to preserve the cellular structure of OM in the

sediment to allow further investigation of the binding of As to solid-phase OM (i.e. Omelon et al.,

2013).

195

6.8 References

Andrade, C.F., Jamieson, H.E., Kyser, T.K., Praharaj, T., Fortin, D., 2010. Biogeochemical redox cycling of arsenic in mine-impacted lake sediments and co-existing pore waters near Giant Mine, Yellowknife Bay, Canada. Appl. Geochemistry. 25, 199–211. https://doi.org/10.1016/j.apgeochem.2009.11.005 Biswas, A., Besold, J., Sjostedt, C., Gustafsson, J.P., Scheinost, A.C., Planer-Friedrich, B., 2019. Complexation of arsenite, arsenate, and monothioarsenate with oxygen-containing functional groups of natural organic matter: An XAS Study. Environ. Sci. Technol. acs.est.9b03020. https://doi.org/10.1021/acs.est.9b03020 Bostick, B.J., Fendorf, S., 2003. Arsenite sorption on troilite (FeS) and pyrite (FeS2). Geochim. Cosmochim. Acta 67, 909–921. https://doi.org/10.1016/S0016-7037(02)01170-5 Couture, R.M., Gobeil, C., Tessier, A., 2010. Arsenic, iron and sulfur co-diagenesis in lake sediments. Geochim. Cosmochim. Acta 74, 1238–1255. https://doi.org/10.1016/j.gca.2009.11.028 Douglas, M.S. V., Smol, J.P., 1999. Freshwater diatoms as indicators of environmental change in the High Arctic, in: Stoermer, E.F., Smol, J.P. (Eds.), The Diatoms. Cambridge University Press, Cambridge, pp. 227–244. https://doi.org/10.1017/CBO9780511613005.011 Fahnestock, M.F., Bryce, J.G., McCalley, C.K., Montesdeoca, M., Bai, S., Li, Y., Driscoll, C.T., Crill, P.M., Rich, V.I., Varner, R.K., 2019. Mercury reallocation in thawing subarctic peatlands. Geochemical Perspect. Lett. 33–38. https://doi.org/10.7185/geochemlet.19 Frey, K.E., McClelland, J.W., 2009. Impacts of permafrost degradation on arctic river biochemistry. J. Glaciol. 23, 169–182. https://doi.org/10.1002/hyp Golder Associates, Ltd., 2005. Terrestrial and aquatic field surveys and ecological risk assessment of the Tundra Gold Mine, Golder Associates Ltd. http://mvlwb.ca/Registry.aspx?a=MV2009L8-0008 Gregory, B.R.B. 2019. Understanding the impact of millennial to sub-decadal climate and limnological change on the stability of arsenic in lacustrine sediments. Carleton University, Ontario, Canada (PhD Thesis). https://curve.carleton.ca/system/files/etd/6f75fc6f-304b-45ca-914f- 15eaa623224a/etd_pdf/059ec3620c91087887bbecf117128d0b/gregory- understandingtheimpactofmillennialtosubdecadal.pdf#page=123 Griffith, F., 2013. Holocene and recent paleoclimate investigations using carbon and nitrogen isotopes from bulk sediment of two subarctic lakes, central Northwest Territories. University of Ottawa, Ontario. (MSc Thesis). http://dx.doi.org/10.20381/ruor-3393 Griffiths, K., Michelutti, N., Sugar, M., Douglas, M.S.V., Smol, J.P., 2017. Ice-cover is the principal driver of ecological change in High Arctic lakes and ponds. PLoS One 12. https://doi.org/10.1371/journal.pone.0172989 Hoffmann, M., Mikutta, C., Kretzschmar, R., 2013. Arsenite binding to natural organic matter: Spectroscopic evidence for ligand exchange and ternary complex formation. Environ. Sci. Technol. 47, 12165–12173. https://doi.org/10.1021/es4023317 Kerr, D.E., 2006. Chapter 20, Quaternary geology and exploration geochemistry. In: Anglin, C.D., Falck, H., Wright, D.F., Ambrose, E.J. (Eds.), GAC Special Publication No.3, Gold in the Yellowknife Greenstone Belt, Northwest Territories: Results of the Extech III Mul- tidisciplinary Research Project, pp. 301–324

196

Langner, P., Mikutta, C., Kretzschmar, R., 2011. Arsenic sequestration by organic sulphur in peat. Nat. Geosci. 5, 66–73. https://doi.org/10.1038/ngeo1329 Lowers, H.A., Breit, G.N., Foster, A.L., Whitney, J., Yount, J., Uddin, Md N., Muneem, Ad A., 2007. Arsenic incorporation into authigenic pyrite, Bengal Basin sediment, Bangladesh. Geochim. Cosmochim. Acta 71 (11), 2699–2717. https://doi.org/10.1016/j.gca.2007.03.022 Macumber, A.L., Patterson, R.T., Neville, L.A., Falck, H., 2011. A sledge microtome for high resolution subsampling of freeze cores. J. Paleolimnol. 45, 307–310. https://doi.org/10.1007/s10933-010-9487- 4 Miller, C.B., Parsons, M.B., Jamieson, H.E., Swindles, G.T., Nasser, N.A., Galloway, J.M. 2019a. Lake- specific controls on the long-term stability of mining-related, legacy arsenic contamination and geochemical baselines in a changing northern environment, Tundra Mine, Northwest Territories, Canada. Applied Geochemistry, 109, 104403. http://doi.org/10.1016/j.apgeochem.2019.104403 Miller, C.B., Parsons, M.B., Jamieson, H.E., Ardakani, O.H., Gregory, B.R.B., Galloway, J.M., 2019b. Influence of late-Holocene climate change on the solid-phase speciation and long-term stability of arsenic in sub-Arctic lake sediments. In press, Science of the Total Environment. Nasser, N.A., Patterson, R.T., Roe, H.M., Galloway, J.M., Falck, H., Palmer, M.J., Spence, C., Sanei, H., Macumber, A.L., Neville, L.A., 2016. Lacustrine Arcellinina (Testate Amoebae) as bioindicators of arsenic contamination. Microb. Ecol. 72, 130–149. https://doi.org/10.1007/s00248-016-0752-6 O’Day, P. a, Vlassopoulos, D., Root, R., Rivera, N., 2004. The influence of sulfur and iron on dissolved arsenic concentrations in the shallow subsurface under changing redox conditions. Proc. Natl. Acad. Sci. U. S. A. 101, 13703–13708. https://doi.org/10.1073/pnas.0402775101 Omelon, C.R., Brady, A.L., Slater, G.F., Laval, B., Lim, D.S.S., Southam, G., 2013. Microstructure variability in freshwater microbialites, Pavilion Lake, Canada. Palaeogeogr. Palaeoclimatol. Palaeoecol. 392, 62–70. https://doi.org/10.1016/j.palaeo.2013.08.017 Pienitz, R., Smol, J.P., MacDonald, G.M., 1999. Paleolimnological reconstruction of Holocene climatic trends from two boreal treeline lakes, Northwest Territories, Canada. Arctic, Antarct. Alp. Res. 31, 82. https://doi.org/10.2307/1552625 Prowse, T., Alfredsen, K., Beltaos, S., Bonsal, B., Duguay, C., Korhola, A., McNamara, J., Pienitz, R., Vincent, W.F., Vuglinsky, V., Weyhenmeyer, G.A., 2011. Past and future changes in arctic lake and river ice. Ambio 40, 53–62. https://doi.org/10.1007/s13280-011-0216-7 Ritter, K., Aiken, G.R., Ranville, J.F., Bauer, M., Macalady, D.L., 2006. Evidence for the aquatic binding of arsenate by natural organic matter-suspended Fe (III). Environ. Sci. Technol. 40, 5380–5387. https://doi.org/10.1021/es0519334 Sanei, H., Ardakani, O.H., 2016. Alteration of organic matter by ion milling. Int. J. Coal Geol. 163, 123– 131. https://doi.org/10.1016/j.coal.2016.06.021 Schuh, C.E., Jamieson, H.E., Palmer, M.J., Martin, A.J., Blais, J.M., 2019. Controls governing the spatial distribution of sediment arsenic concentrations and solid-phase speciation in a lake impacted by legacy mining pollution. Sci. Total Environ. 654, 563–575. https://doi.org/10.1016/j.scitotenv.2018.11.065 Sharma, P., Rolle, M., Kocar, B.D., Fendorf, S., Kapppler, A., 2010. Influence of natural organic matter on As transport and retention. Environ. Sci. Technol. 45, 546–553. https://doi.org/10.2134/jeq2003.1393

197

Smol, J. P., 1988. Paleoclimate proxy data from freshwater Arctic diatoms. Verhandlungen Internationale Vereinigung fur theoretische und angewandte Limnologie, 23: 837–844. Stern, G.A., Macdonald, R.W., Outridge, P.M., Wilson, S., Chételat, J., Cole, A., Hintelmann, H., Loseto, L.L., Steffen, A., Wang, F., Zdanowicz, C., 2012. How does climate change influence arctic mercury? Sci. Total Environ. 414, 22–42. https://doi.org/10.1016/j.scitotenv.2011.10.039 Sulphur, K.C., Goldsmith, S.A., Galloway, J.M., Macumber, A., Griffith, F., Swindles, G.T., Patterson, R.T., Falck, H., Clark, I.D., 2016. Holocene fire regimes and treeline migration rates in sub-arctic Canada. Glob. Planet. Change 145, 42–56. https://doi.org/10.1016/j.gloplacha.2016.08.003 Upiter, L.M., Vermaire, J.C., Patterson, R.T., Crann, C.A., Galloway, J.M., Macumber, A.L., Neville, L.A., Swindles, G.T., Falck, H., Roe, H.M., Pisaric, M.F.J., 2014. Middle to late-Holocene chironomid-inferred July temperatures for the central Northwest Territories, Canada. J. Paleolimnol. 52, 11–26. https://doi.org/10.1007/s10933-014-9775-5 Wang, Y., Le Pape, P., Morin, G., Asta, M.P., King, G., Bártová, B., Suvorova, E., Frutschi, M., Ikogou, M., Pham, V.H.C., Vo, P. Le, Herman, F., Charlet, L., Bernier-Latmani, R., 2018. Arsenic speciation in Mekong Delta sediments depends on their depositional environment. Environ. Sci. Technol. 52, 3431–3439. https://doi.org/10.1021/acs.est.7b05177

198

Appendix A Supplementary data to Chapter 2

The data presented in this appendix reflect the supplementary material that appears online in association with the publication of Chapter 2 in Applied Geochemistry.

The Excel files listed provide more comprehensive data and associated QA/QC: · Miller_BulkGeochemistry_Sediment · Miller_AqueousGeochemistry · Miller_RockEval · Miller_GrainSizeAnalysis · Miller_EPMA_AppendixA · Miller_SpearmansCorrelations_AppendixA

199

Table A-1 Concentrations of selected anions (filtered, <0.45 µm) and major and trace elements (unfiltered) in surface waters of lakes of the Tundra Mine region (additional data provided in Table 2.1).

Units CCMEA Powder Mag Bulldog Matthews Control

Date of Sampling -- -- 11-Mar-16 10-Mar-16 13-Mar-16 12-Mar-16

Anions

Cl mg·L-1 120 5.40 0.40 0.76 0.33

Major Elements

Ca mg·L-1 -- 104 8.94 5.64 1.92

Mg mg·L-1 -- 11.6 1.82 1.05 1.03

Na mg·L-1 -- 35.6 0.91 0.69 1.04

Trace Elements

Sb µg·L-1 -- 0.33 0.02 0.02 < 0.01

W µg·L-1 -- 0.18 < 0.02 < 0.02 < 0.02

A CCME Water Quality Guidelines for the protection of aquatic life (CCME, 2001b), bold and italicized values exceed guidelines; B Concentration reported based on HG-AFS analysis

200

Table A-2 Concentrations of As, Ag, Au, Hg, Pb, Sb, W, Zn, S, Fe, Mn and organic matter fractions S1 & S2 in the Powder Mag Lake (POW) sediment core (64.05114° N, 111.15042° W).

Sample Depth As Ag Au Hg Pb Sb W Zn S Fe Mn S1 S2 Units cm mg·kg-1 µg·kg-1 µg·kg-1 µg·kg-1 mg·kg-1 mg·kg-1 mg·kg-1 mg·kg-1 % % mg·kg-1 mg HC/gA mg HC/gA Detection Limit -- 0.1 2.0 0.20 2.0 0.01 0.02 0.1 0.1 0.02 0.01 1.0 0.01 0.01 POW 0-1 0.5 ISB IS IS IS IS IS IS IS IS IS IS 37.0 65.1 POW 1-2 1.5 210 168 1,010 35.0 4.75 1.75 4.40 90.7 1.29 1.71 102 28.2 48.6 POW 2-3 2.5 195 183 882 64.0 4.47 1.40 2.90 106 1.06 1.75 97.0 19.0 35.3 POW 3-4 3.5 271 122 467 64.0 3.90 0.92 1.60 159 1.28 2.29 108 10.1 26.6 POW 4-5 4.5 260 105 50.2 38.0 3.54 0.28 0.80 125 1.20 2.19 112 7.57 23.7 POW 5-6 5.5 191 133 10.6 41.0 3.89 0.11 1.50 171 0.98 1.86 106 8.54 28.5 POW 6-7 6.5 163 149 9.60 27.0 4.35 0.08 0.70 129 0.84 1.73 110 10.7 35.6 POW 7-8 7.5 114 138 8.80 39.0 3.92 0.07 0.80 119 0.65 1.50 102 10.5 35.0 POW 8-9 8.5 77.2 103 4.80 27.0 3.44 0.04 1.20 86.5 0.46 1.33 96.0 9.26 30.3 POW 9-10 9.5 69.5 129 8.80 25.0 3.68 0.06 0.50 93.6 0.50 1.28 95.0 10.3 35.9 POW 10-11 10.5 87.4 151 16.6 45.0 4.49 0.08 0.60 127 0.61 1.45 111 13.3 43.3 POW 11-12 11.5 89.7 162 24.1 30.0 4.52 0.09 0.70 108 0.62 1.52 120 13.3 46.0 POW 12-13 12.5 89.0 161 23.7 51.0 4.29 0.08 0.70 109 0.65 1.57 116 12.2 40.3 POW 13-14 13.5 81.2 125 14.7 43.0 3.41 0.06 1.00 94.6 0.56 1.86 109 12.7 38.9 POW 14-15 14.5 79.7 154 19.6 40.0 4.33 0.07 1.00 119 0.58 1.72 112 13.4 44.2 POW 15-16 15.5 66.0 163 11.3 36.0 4.22 0.06 0.70 92.3 0.60 1.54 114 13.8 45.6 POW 16-17 16.5 66.6 144 14.6 29.0 3.85 0.07 0.70 84.8 0.62 1.53 122 13.0 42.6 POW 17-18 17.5 58.0 131 8.40 36.0 3.80 0.04 1.30 87.7 0.60 1.64 123 13.6 47.5 POW 18-19 18.5 66.0 114 6.20 48.0 3.32 0.05 0.80 94.1 0.63 1.68 126 14.4 50.1 POW 19-20 19.5 70.7 99.0 3.60 56.0 3.16 0.04 0.70 97.3 0.67 1.84 134 15.3 54.7 POW 20-21 20.5 68.9 106 4.90 68.0 3.50 0.05 0.80 118 0.67 1.95 150 16.7 62.2 POW 21-22 21.5 52.6 92.0 4.20 43.0 2.98 0.03 0.60 91.5 0.51 1.78 148 13.4 47.3 POW 22-23 22.5 59.7 103 2.50 48.0 3.21 0.04 0.70 104 0.63 1.91 158 13.2 52.6 POW 23-24 23.5 56.8 77.0 2.40 53.0 2.87 0.02 0.60 103 0.60 1.88 149 12.1 45.1 201

POW 24-25 24.5 54.6 89.0 1.70 45.0 2.75 0.03 0.80 112 0.59 2.06 149 13.8 51.4 POW 25-26 25.5 52.5 92.0 2.50 56.0 2.71 0.03 0.80 116 0.58 2.15 155 12.7 48.9 POW 26-27 26.5 42.4 89.0 1.10 44.0 2.48 0.03 1.20 102 0.50 1.84 147 12.0 45.8 POW 27-28 27.5 44.5 81.0 3.70 48.0 2.37 <0.02 0.80 95.4 0.52 1.80 125 14.2 52.8 POW 28-29 28.5 58.8 125 <0.20 43.0 2.55 <0.02 0.60 97.6 0.71 1.80 128 17.8 65.6 POW 29-30 29.5 52.6 113 2.90 38.0 2.65 0.03 0.60 86.6 0.72 1.65 122 17.2 63.9 A mg HC/g – milligrams of hydrocarbon per gram of sediment; B IS – insufficient sample.

Table A-3 Concentrations of As, Ag, Au, Hg, Pb, Sb, W, Zn, S, Fe, Mn and organic matter fractions S1 & S2 from the Bulldog Lake (BUL) sediment core (64.03785° N, 111.18337° W).

Sample Depth As Ag Au Hg Pb Sb W Zn S Fe Mn S1 S2 Units cm mg·kg-1 µg·kg-1 µg·kg-1 µg·kg-1 mg·kg-1 mg·kg-1 mg·kg-1 mg·kg-1 % % mg·kg-1 mg HC/gA mg HC/gA Detection Limit -- 0.1 2.0 0.20 2.0 0.01 0.02 0.1 0.1 0.02 0.01 1.0 0.01 0.01 BUL 0-1 0.5 590 746 4,580 212 34.0 0.60 80.1 140 0.46 4.09 993 12.3 35.9 BUL 1-2 1.5 351 902 7,350 253 45.0 0.75 85.9 182 0.83 2.94 609 10.8 29.3 BUL 2-3 2.5 1,010 710 3,120 246 39.3 0.8 53.3 167 1.02 2.86 462 11.1 30.6 BUL 3-4 3.5 805 493 1,640 190 25.4 0.87 35.3 144 0.72 2.47 449 12.5 33.8 BUL 4-5 4.5 566 393 859 149 15.9 0.75 22.4 133 0.58 2.15 442 12.0 35.1 BUL 5-6 5.5 305 355 644 132 12.7 0.52 15.1 119 0.44 1.92 412 13.2 35.7 BUL 6-7 6.5 198 325 375 119 10.4 0.27 8.60 112 0.35 1.80 406 12.5 36.4 BUL 7-8 7.5 114 290 184 114 5.85 0.18 4.40 115 0.30 1.51 381 12.6 35.9 BUL 8-9 8.5 103 283 136 89.0 5.44 0.17 3.60 109 0.27 1.47 343 11.2 32.4 BUL 9-10 9.5 90.4 266 98.7 76.0 5.15 0.17 3.10 117 0.27 1.54 346 10.8 32.8 BUL 10-11 10.5 87.9 278 101 81.0 5.38 0.18 3.20 124 0.27 1.58 363 10.6 34.7 BUL 11-12 11.5 83.5 242 118 87.0 5.02 0.18 3.40 116 0.28 1.63 357 11.7 34.7 BUL 12-13 12.5 74.5 240 108 77.0 4.95 0.14 2.90 114 0.26 1.70 317 10.5 31.2 BUL 13-14 13.5 71.2 228 95.8 71.0 4.84 0.14 2.70 116 0.23 1.80 306 9.80 28.1 BUL 14-15 14.5 62.7 221 91.4 56.0 4.56 0.15 2.50 118 0.23 1.71 279 9.42 27.4 BUL 15-16 15.5 55.8 216 60.9 61.0 4.51 0.12 2.00 122 0.21 1.69 286 8.54 25.9

202

BUL 16-17 16.5 57.6 221 77.4 68.0 4.80 0.16 2.00 121 0.22 1.69 300 9.10 26.6 BUL 17-18 17.5 49.7 222 49.2 66.0 4.54 0.13 2.20 119 0.21 1.70 254 8.85 25.4 BUL 18-19 18.5 46.5 194 38.1 58.0 4.07 0.11 1.50 111 0.19 1.69 253 7.39 21.7 BUL 19-20 19.5 41.4 295 25.5 66.0 4.11 0.11 1.40 117 0.14 1.67 247 7.82 23.1 BUL 20-21 20.5 34.2 163 17.8 53.0 3.55 0.08 1.00 94.9 0.15 1.57 207 7.67 23.4 BUL 21-22 21.5 34.4 175 7.80 69.0 3.75 0.09 0.90 105 0.15 1.55 216 7.78 23.3 BUL 22-23 22.5 34.3 173 5.80 53.0 3.71 0.08 0.80 105 0.18 1.53 212 8.41 27.0 BUL 23-24 23.5 36.2 190 5.10 79.0 3.92 0.08 1.20 106 0.19 1.5 221 8.60 26.9 BUL 24-25 24.5 33.1 171 9.40 59.0 3.65 0.08 0.70 101 0.19 1.42 212 9.39 28.5 BUL 25-26 25.5 33.0 191 10.1 53.0 4.31 0.10 0.80 107 0.19 1.35 196 10.1 29.4 BUL 26-27 26.5 36.4 194 13.8 60.0 4.21 0.11 1.00 118 0.22 1.47 218 10.4 29.9 BUL 27-28 27.5 34.5 189 12.8 70.0 3.80 0.09 1.10 109 0.22 1.46 219 10.5 30.9 BUL 28-29 28.5 33.1 177 16.5 54.0 3.43 0.09 0.80 96 0.21 1.39 209 9.78 30.0 BUL 29-30 29.5 33.0 183 17.9 72.0 3.73 0.08 1.00 96 0.20 1.36 201 10.6 30.6 BUL 30-31 30.5 36.4 153 16.3 51.0 3.48 0.07 1.10 101 0.20 1.40 200 10.5 30.6 BUL 31-32 31.5 34.5 159 15.9 66.0 3.76 0.07 1.10 107 0.23 1.50 197 11.0 31.2 BUL 32-33 32.5 33.1 181 11.9 62.0 3.69 0.08 0.80 105 0.23 1.45 216 10.6 30.7 BUL 33-34 33.5 35.6 181 6.80 66.0 3.98 0.10 0.90 129 0.23 1.38 213 11.6 34.1 BUL 34-35 34.5 34.2 154 6.50 51.0 3.31 0.08 0.90 108 0.20 1.37 195 9.68 28.6 BUL 35-36 35.5 36.8 131 5.10 47.0 3.19 0.06 0.80 97.5 0.19 1.41 188 8.99 26.7 BUL 36-37 36.5 37.5 136 4.60 44.0 4.70 0.06 0.60 92.4 0.19 1.31 175 9.09 27.3 A mg HC/g – milligrams of hydrocarbon per gram of sediment.

203

Table A-4 Concentrations of As, Ag, Au, Hg, Pb, Sb, W, Zn, S, Fe, Mn and organic matter fractions S1 & S2 from the Matthews Lake (MAT) sediment core (64.06014° N, 111.23566° W).

Sample Depth As Ag Au Hg Pb Sb W Zn S Fe Mn S1 S2 Units cm mg·kg-1 µg·kg-1 µg·kg-1 µg·kg-1 mg·kg-1 mg·kg-1 mg·kg-1 mg·kg-1 % % mg·kg-1 mg HC/gA mg HC/gA Detection Limit -- 0.1 2.0 0.2 2.0 0.01 0.02 0.1 0.1 0.02 0.01 1.0 0.01 0.01 MAT 0-1 0.5 89.5 213 230 79.0 10.2 0.51 3.60 124 0.34 2.68 554 14.5 27.6 MAT 1-2 1.5 66.8 204 179 74.0 9.53 0.50 2.70 126 0.32 2.35 445 16.9 30.7 MAT 2-3 2.5 48.0 192 91.0 72.0 8.27 0.38 1.30 118 0.29 2.10 350 16.1 30.3 MAT 3-4 3.5 47.0 207 68.8 53.0 7.89 0.28 1.50 120 0.29 2.02 319 15.2 30.5 MAT 4-5 4.5 63.0 188 51.7 73.0 7.21 0.25 1.40 117 0.29 2.19 357 13.8 29.5 MAT 5-6 5.5 50.5 212 37.0 74.0 7.19 0.22 1.30 125 0.29 2.08 348 13.4 30.6 MAT 6-7 6.5 47.9 201 29.6 72.0 6.07 0.17 1.20 128 0.29 2.09 324 12.9 31.0 MAT 7-8 7.5 49.3 221 23.0 43.0 5.23 0.18 1.00 133 0.29 2.10 317 13.2 32.2 MAT 8-9 8.5 52.7 241 26.5 47.0 5.54 0.18 1.20 144 0.28 2.27 339 13.5 34.6 MAT 9-10 9.5 40.1 235 20.7 32.0 5.11 0.17 1.20 142 0.26 2.09 310 14.6 36.0 MAT 10-11 10.5 38.3 237 27.9 52.0 5.06 0.14 1.10 135 0.27 2.07 308 14.6 36.3 MAT 11-12 11.5 31.2 210 10.2 38.0 4.28 0.11 0.90 137 0.25 1.89 295 14.8 36.3 MAT 12-13 12.5 30.8 208 7.60 33.0 4.48 0.11 1.00 167 0.24 1.87 296 14.5 37.2 MAT 13-14 13.5 36.4 235 10.2 40.0 5.29 0.16 1.00 223 0.29 2.15 359 15.4 37.5 MAT 14-15 14.5 33.7 241 8.40 47.0 4.97 0.15 1.10 232 0.31 2.02 357 16.2 38.6 MAT 15-16 15.5 35.7 230 11.9 30.0 5.83 0.17 1.00 202 0.30 1.96 344 15.1 37.1 MAT 16-17 16.5 32.0 243 7.10 23.0 4.84 0.14 1.00 186 0.33 2.11 357 14.6 35.8 MAT 17-18 17.5 32.6 204 7.20 33.0 4.67 0.13 1.00 185 0.33 2.21 361 15.0 35.7 MAT 18-19 18.5 34.0 210 10.7 26.0 4.49 0.12 1.20 189 0.34 2.38 387 14.2 35.6 MAT 19-20 19.5 37.8 197 12.2 19.0 4.54 0.12 1.20 181 0.33 2.47 405 13.9 33.2 MAT 20-21 20.5 34.6 198 9.10 26.0 4.34 0.12 1.20 170 0.30 2.24 348 14.8 36.1 MAT 21-22 21.5 31.3 194 8.90 25.0 4.22 0.12 2.00 160 0.31 2.25 378 15.4 36.6 MAT 22-23 22.5 34.3 203 11.4 22.0 4.22 0.12 1.00 159 0.30 2.23 360 14.8 35.9 MAT 23-24 23.5 36.6 193 8.50 34.0 4.33 0.12 1.10 159 0.31 2.30 373 14.5 34.7

204

MAT 24-25 24.5 35.4 180 9.20 27.0 4.13 0.11 1.00 149 0.31 2.20 350 14.4 36.0 MAT 25-26 25.5 32.3 183 5.50 22.0 4.05 0.11 1.10 145 0.32 2.28 363 15.6 38.5 MAT 26-27 26.5 31.7 201 5.70 16.0 3.93 0.12 1.10 144 0.32 2.17 365 15.4 38.0 MAT 27-28 27.5 32.0 189 9.20 25.0 3.99 0.12 1.00 140 0.31 2.12 341 14.8 36.2 MAT 28-29 28.5 32.1 209 5.80 16.0 4.65 0.11 1.10 145 0.32 2.22 369 14.0 36.0 MAT 29-30 29.5 32.5 200 5.60 26.0 4.08 0.12 1.10 146 0.32 2.20 354 15.4 38.0 MAT 30-31 30.5 32.2 176 3.10 30.0 4.04 0.10 1.20 156 0.32 2.15 353 16.2 39.1 MAT 31-32 31.5 29.8 164 3.60 40.0 3.74 0.10 1.00 136 0.28 1.97 324 15.9 39.4 MAT 32-33 32.5 29.6 157 3.70 22.0 3.69 0.10 1.00 135 0.30 1.97 344 15.3 38.8 MAT 33-34 33.5 29.6 165 4.40 30.0 3.82 0.10 1.00 133 0.31 1.94 344 15.4 36.6 MAT 34-35 34.5 30.7 166 4.20 20.0 3.75 0.09 1.00 134 0.31 1.91 337 15.8 38.7 MAT 35-36 35.5 30.5 189 2.90 17.0 4.27 0.13 1.10 152 0.33 1.94 353 17.6 41.7 MAT 36-37 36.5 30.1 162 4.10 21.0 3.74 0.11 1.10 140 0.31 1.87 340 15.6 36.7 MAT 37-38 37.5 29.2 157 3.70 14.0 3.33 0.10 0.80 121 0.30 1.73 320 16.7 39.6 MAT 38-39 38.5 31.4 165 2.70 15.0 3.93 0.10 1.00 129 0.32 1.85 343 16.6 41.1 MAT 39-40 39.5 33.8 186 3.70 22.0 3.61 0.10 1.20 141 0.37 2.09 363 17.2 40.1 MAT 40-41 40.5 33.5 153 2.70 30.0 3.38 0.11 0.90 126 0.37 1.86 327 16.6 40.6 MAT 41-42 41.5 41.4 168 3.00 22.0 3.46 0.09 0.70 118 0.45 1.86 330 16.4 42.5 MAT 42-43 42.5 40.4 171 3.30 17.0 3.76 0.11 0.90 129 0.43 1.87 350 16.7 40.8 MAT 43-44 43.5 37.0 162 3.00 26.0 3.70 0.12 1.00 129 0.42 1.83 344 16.0 40.1 MAT 44-45 44.5 45.9 157 4.00 13.0 3.51 0.10 0.90 122 0.49 1.85 341 14.3 34.5 MAT 45-46 45.5 43.5 165 2.60 22.0 3.34 0.09 1.00 120 0.49 1.83 337 16.1 38.2 MAT 46-47 46.5 48.8 162 3.00 11.0 3.45 0.09 0.80 119 0.55 1.90 344 16.1 39.5 MAT 47-48 47.5 36.2 162 5.20 25.0 3.53 0.08 0.80 120 0.46 1.76 318 15.9 40.1 MAT 48-49 48.5 34.6 150 3.20 23.0 3.55 0.11 0.70 124 0.46 1.71 324 14.5 27.6 MAT 49-50 49.5 34.5 154 4.30 15.0 4.01 0.10 0.80 133 0.42 1.69 344 16.9 30.7 MAT 50-51 50.5 33.3 148 3.10 26.0 3.79 0.08 0.80 117 0.42 1.65 334 16.1 30.3 MAT 51-52 51.5 38.9 164 3.70 15.0 3.89 0.10 0.80 130 0.46 1.67 331 15.2 30.5 A mg HC/g – milligrams of hydrocarbon per gram of sediment.

205

Table A-5 Concentrations of As, Ag, Au, Hg, Pb, Sb, W, Zn, S, Fe, Mn and organic matter fractions S1 & S2 from the Control Lake (CON) sediment core (64.07771° N, 111.13493° W).

Sample Depth As Ag Au Hg Pb Sb W Zn S Fe Mn S1 S2 Units cm mg·kg-1 µg·kg-1 µg·kg-1 µg·kg-1 mg·kg-1 mg·kg-1 mg·kg-1 mg·kg-1 % % mg·kg-1 mg HC/gA mg HC/gA Detection Limit -- 0.1 2.0 0.2 2.0 0.01 0.02 0.1 0.1 0.02 0.01 1.0 0.01 0.01 CON 0-1 0.5 224 94.0 14.1 65.0 9.07 0.37 0.90 67.5 0.37 7.31 241 28.2 46.7 CON 1-2 1.5 218 167 11.7 58.0 7.90 0.37 1.00 61.5 0.37 6.88 205 27.6 42.5 CON 2-3 2.5 151 110 8.70 71.0 8.11 0.38 1.10 63.1 0.38 5.92 198 24.2 44.0 CON 3-4 3.5 109 125 10.5 62.0 9.56 0.39 1.20 78.6 0.39 5.11 184 26.2 48.6 CON 4-5 4.5 91.1 117 8.30 76.0 9.94 0.38 1.20 83.6 0.38 4.82 179 23.2 46.7 CON 5-6 5.5 86.1 129 7.70 57.0 9.79 0.37 1.10 88.4 0.37 4.36 162 20.2 44.7 CON 6-7 6.5 89.3 128 9.00 60.0 8.12 0.36 1.00 79.9 0.36 4.22 161 19.0 42.7 CON 7-8 7.5 104 120 5.60 43.0 8.39 0.37 1.10 90.3 0.37 4.80 164 17.8 44.3 CON 8-9 8.5 102 154 4.80 44.0 6.72 0.39 1.20 110 0.39 4.95 172 18.6 42.8 CON 9-10 9.5 89.3 140 6.30 44.0 5.90 0.34 1.20 83.5 0.34 4.18 168 17.4 42.3 CON 10-11 10.5 107 133 7.00 49.0 5.88 0.36 1.00 75.1 0.36 4.84 169 17.5 40.1 CON 11-12 11.5 94.8 140 12.0 32.0 5.93 0.36 0.90 85.1 0.36 4.42 170 17.9 42.4 CON 12-13 12.5 95.4 140 4.70 41.0 5.73 0.36 0.90 83.9 0.36 4.64 173 16.9 39.5 CON 13-14 13.5 110 128 6.70 21.0 5.61 0.34 1.00 83.6 0.34 5.10 179 15.9 38.9 CON 14-15 14.5 129 153 7.20 22.0 5.65 0.38 1.20 106 0.38 5.79 189 17.5 43.0 CON 15-16 15.5 152 141 7.60 21.0 5.85 0.36 1.20 93.5 0.36 6.16 189 15.0 38.5 CON 16-17 16.5 147 135 5.20 11.0 9.06 0.34 1.20 101 0.34 5.87 180 15.3 37.8 CON 17-18 17.5 137 144 5.20 34.0 5.60 0.32 1.10 94.6 0.32 5.49 182 14.7 39.6 CON 18-19 18.5 116 163 8.10 36.0 6.30 0.34 1.20 95.5 0.34 4.84 185 15.2 37.5 CON 19-20 19.5 73.6 131 5.60 30.0 4.98 0.29 1.10 70.9 0.29 3.32 146 16.7 40.7 CON 20-21 20.5 155 161 5.10 26.0 5.38 0.34 1.00 85.0 0.34 6.04 190 14.6 37.5 CON 21-22 21.5 121 150 4.50 34.0 5.31 0.32 1.00 80.9 0.32 5.13 174 16.4 42.7 CON 22-23 22.5 127 158 4.20 32.0 5.14 0.30 1.00 96.2 0.30 5.30 171 15.9 39.0 CON 23-24 23.5 134 157 8.20 32.0 5.27 0.32 1.10 84.9 0.32 5.78 181 15.2 39.3

206

CON 24-25 24.5 153 182 4.10 33.0 5.56 0.32 1.20 103 0.32 6.43 196 15.4 40.7 CON 25-26 25.5 152 164 4.00 33.0 5.58 0.31 1.20 106 0.31 6.15 200 14.6 39.5 CON 26-27 26.5 108 134 3.20 27.0 4.97 0.30 1.00 94.4 0.30 4.75 176 15.6 36.9 CON 27-28 27.5 137 154 4.90 30.0 5.18 0.33 1.10 104 0.33 5.59 190 14.4 36.5 CON 28-29 28.5 167 152 4.10 25.0 5.01 0.34 1.20 95.2 0.34 6.90 183 14.5 35.7 CON 29-30 29.5 114 129 3.90 21.0 4.82 0.30 1.00 79.7 0.30 4.99 161 13.6 32.7 CON 30-31 30.5 109 146 4.90 22.0 5.15 0.30 1.00 92.7 0.30 4.74 161 13.8 33.9 CON 31-32 31.5 109 137 3.90 23.0 4.89 0.30 1.00 77.2 0.30 4.83 154 13.3 35.4 CON 32-33 32.5 104 171 4.20 23.0 5.17 0.33 1.00 91.7 0.33 4.68 169 15.4 39.8 CON 33-34 33.5 89.1 157 5.00 31.0 5.16 0.32 0.90 94.2 0.32 4.01 155 15.5 40.6 CON 34-35 34.5 76.4 141 3.60 26.0 4.85 0.31 0.90 69.1 0.31 3.59 146 15.0 38.7 CON 35-36 35.5 86.0 142 3.80 18.0 5.11 0.30 0.90 78.0 0.30 3.85 140 14.5 36.3 CON 36-37 36.5 87.4 149 2.30 18.0 4.61 0.34 0.90 77.4 0.34 4.01 140 15.5 40.0 CON 37-38 37.5 95.0 183 326* 22.0 4.50 0.36 1.00 81.5 0.36 4.41 134 15.3 40.3 CON 38-39 38.5 105 167 2.30 23.0 4.81 0.36 1.00 79.4 0.36 4.72 146 15.8 40.5 CON 39-40 39.5 130 195 3.20 25.0 4.97 0.38 1.30 92.9 0.38 5.80 153 14.7 39.3 CON 40-41 40.5 119 170 2.60 22.0 4.42 0.36 1.00 84.6 0.36 5.43 140 14.9 41.1 CON 41-42 41.5 110 177 2.80 31.0 4.54 0.37 1.00 83.7 0.37 5.07 148 14.5 43.5 CON 42-43 42.5 90.8 178 2.70 20.0 4.70 0.37 1.50 82.9 0.37 4.19 151 17.7 50.3 CON 43-44 43.5 68.5 154 1.70 29.0 4.30 0.38 0.80 74.5 0.38 3.29 136 17.1 48.6 CON 44-45 44.5 72.6 183 26.4 27.0 4.63 0.36 0.80 84.5 0.36 3.31 139 16.2 47.4 CON 45-46 45.5 63.8 172 5.10 35.0 4.45 0.36 0.70 62.6 0.36 2.96 141 17.9 53.7 CON 46-47 46.5 56.1 147 1.20 42.0 4.20 0.31 0.80 67.2 0.31 2.82 146 17.2 53.5 CON 47-48 47.5 81.4 156 3.00 27.0 4.28 0.34 0.80 82.4 0.34 3.99 147 16.4 46.6

* Outlier removed from figures in publication; A mg HC/g – milligrams of hydrocarbon per gram of sediment.

207

Table A-6 Concentrations of organic matter fractions S1 & S2 from Hambone Lake sediment grab samples.

Sample Units Detection Limits HAM-1 HAM-2 HAM-3 HAM-4 HAM-5 HAM-6 Distance from discharge* m -- 540 515 310 230 125 250 Sediment sample depth cm -- 0-15 0-15 0-15 0-15 0-15 0-15 S1 mg HC/gA 0.01 16.7 19.6 11.9 11.4 13.6 -- S2 mg HC/gA 0.01 50.8 52.7 45.4 53.4 59.7 -- * Location shown on Figure 2.1; A mg HC/g – milligrams of hydrocarbon per gram of sediment.

208

Table A-7 Results of 210Pb dating for sediment core retrieved from Powder Mag Lake.

Sediment Depth Unsupported Activity Supported Activity Year (CRS)A Error cm Bq·kg-1 Bq·kg-1 -- Years 0 303 -- 2016 0.7 0.5 303 0 2014 1.0 1.5 536 13.4 2006 1.7 2.5 398 20.9 1986 2.5 3.5 104 11.7 1938 7.4 4.5 1.41 15.2 -- -- 5.5 0 16.6 -- -- 6.5 0 19.5 -- -- 9.5 0 18.5 -- -- 20.5 0 19.5 -- -- 24.5 0 15.8 -- -- 28.5 0 15.3 -- -- A CRS – Constant rate of supply

Table A-8 Results of 210Pb dating for sediment core retrieved from Bulldog Lake.

Sediment Depth Unsupported Activity Supported Activity Year (CRS)A Error cm Bq·kg-1 Bq·kg-1 -- Years 0 1098 -- 2016 0.2 0.5 1098 52.8 2010 1.0 1.5 651 39.3 1994 1.0 2.5 382 40.0 1977 2.0 3.5 219 35.3 1962 3.0 4.5 128 29.3 1947 5.0 5.5 72.7 28.2 1933 7.5 6.5 62.2 27.3 1918 11 9.5 5.93 24.6 -- -- 10.5 16.0 23.6 -- -- 14.5 0.1 27.4 -- -- 20.5 0 24.2 -- -- 36.5 0 24.5 -- -- A CRS – Constant rate of supply

209

Table A-9 Results of 210Pb dating for sediment core retrieved from Matthews Lake.

Sediment Depth Unsupported Activity Supported Activity Year (CRS)A Error cm Bq·kg-1 Bq·kg-1 -- Years 0 501 -- 2016 2.0 0.5 501 15.9 2014 2.0 1.5 425 14.6 2006 2.0 2.5 248 19.9 1996 2.0 3.5 191 24.9 1988 3.0 4.5 73.5 24.8 1981 4.0 5.5 116 11.5 1976 4.0 7.5 65.4 18.1 1959 7.0 9.5 41.4 19.2 1942 11 10.5 46.5 22.2 1930 16 14.5 0.24 21.4 -- -- 20.5 0 27.8 -- -- 24.5 0 23.9 -- -- A CRS – Constant rate of supply

Table A-10 Results of 210Pb dating for sediment core retrieved from Control Lake.

Sediment Depth Unsupported Activity Supported Activity Year (CRS)A Error cm Bq·kg-1 Bq·kg-1 -- Years 0 936 -- 2016 1.0 0.5 936 21.6 2015 1.0 1.5 981 8.20 2010 1.0 2.5 902 21.5 2004 1.0 3.5 731 14.8 1997 1.5 4.5 622 15.6 1987 2.0 5.5 289 16.3 1978 3.0 6.5 155 22.0 1972 4.0 7.5 117 23.0 1967 4.0 9.5 79.0 22.8 1958 6.0 12.5 39.3 22.1 1946 8.0 14.5 29.8 21.2 1939 10 20.5 33.5 19.9 1895 24 24.5 0 21.4 -- -- 28.5 0 19.1 -- -- 32.5 0 24.6 -- -- 36.5 0 23.3 -- -- A CRS – Constant rate of supply

210

Table A-11 Results of automated mineralogy and mass distribution calculations for Powder Mag Lake sediments.

As-Hosting Solid Phase Number of Particles Total Area of Particles Density As Mass Distribution Units -- µm2 g·cm-3 wt.% % Powder Mag Lake (0 – 1 cm) [ID: POW 0-1] Fe/Mn-(oxy)hydroxide 353 5540 4.28 0.50* 5.27 Framboidal pyrite 122 5180 5.01 0.20* 2.30 Arsenopyrite 4 150 6.07 46.0 18.6 As-Sulphide 99 430 3.56 70.0 48.1 Phyllosilicate 817 25340 3.09 0.10* 3.48 A FeSx/FeO 1583 53850 4.65 0.20* 22.2 Powder Mag Lake (1 – 2 cm) [ID: POW 1-2] Fe/Mn-(oxy)hydroxide 890 45407 4.28 0.50* 20.4 Framboidal pyrite 109 6334 5.01 0.20* 1.33 Arsenopyrite 12 13 6.07 46.0 0.77 As-Sulphide 38 190 3.56 70.0 9.91 Phyllosilicate 3440 77994 3.09 0.10* 5.06 A FeSx/FeO 8340 320904 4.65 0.20* 62.5 Powder Mag Lake (3 – 4 cm) [ID: POW 3-4] Fe/Mn-(oxy)hydroxide 1591 88393 4.28 0.50* 18.5 Framboidal pyrite 294 28623 5.01 0.20* 2.81 Arsenopyrite 21 64 6.07 46.0 1.76 As-Sulphide 72 618 3.56 70.0 15.1 Phyllosilicate 6705 135292 3.09 0.10* 4.10 A FeSx/FeO 17531 633473 4.65 0.20* 57.7 1 1 * Determined using electron microprobe analysis (EPMA); A 휌 = 휌 + 휌 (FeSxFeO) 2 푝푦푟𝑖푡푒 2 푓푒푟푟𝑖ℎ푦푑푟𝑖푡푒

Table A-12 Results of automated mineralogy and mass distribution calculations for Bulldog Lake sediments.

As-Hosting Solid Phase Number of Particles Total Area of Particles Density As Mass Distribution Units -- µm2 g·cm-3 wt.% % Bulldog Lake (0 – 1 cm) [ID: BUL 0-1] Fe/Mn-(oxy)hydroxide 2540 131464 4.28 0.50* 53.0 Framboidal pyrite 7 956 5.01 0.20* 0.17 Arsenopyrite 13 27 6.07 46.0 1.42 As-Sulphide 1 7 3.56 70.0 0.34 Phyllosilicate 1130 13278 3.09 0.10* 0.77 A FeSx/FeO 9825 253236 4.65 0.20* 44.3 Bulldog Lake (1 – 2 cm) [ID: BUL 1-2] Fe/Mn-(oxy)hydroxide 2509 55430 4.28 0.50* 13.6 Framboidal pyrite 66 6121 5.01 0.20* 0.66

211

Arsenopyrite 34 388 6.07 46.0 12.4 As-Sulphide 74 413 3.56 70.0 11.8 Phyllosilicate 3288 80277 3.09 0.10* 2.84 A FeSx/FeO 13090 552333 4.65 0.20* 58.7 Bulldog Lake (2 – 3 cm) [ID: BUL 2-3] Fe/Mn-(oxy)hydroxide 3743 81555 4.28 0.50* 7.61 Framboidal pyrite 28 5778 5.01 0.20* 0.23 Arsenopyrite 167 1417 6.07 46.0 17.2 As-Sulphide 688 5385 3.56 70.0 50.9 Phyllosilicate 4662 102849 3.09 0.10* 1.38 A FeSx/FeO 13407 559664 4.65 0.20* 22.7 Bulldog Lake (5 – 6 cm) [ID: BUL 5-6] Fe/Mn-(oxy)hydroxide 1429 31745 4.28 0.50* 15.8 Framboidal pyrite 34 1966 5.01 0.20* 0.45 Arsenopyrite 12 600 6.07 46.0 39.0 As-Sulphide 81 459 3.56 70.0 26.7 Phyllosilicate 731 10864 3.09 0.10* 0.78 A FeSx/FeO 5644 79494 4.65 0.20* 17.2 1 1 * Determined using electron microprobe analysis (EPMA); A 휌 = 휌 + 휌 (FeSxFeO) 2 푝푦푟𝑖푡푒 2 푓푒푟푟𝑖ℎ푦푑푟𝑖푡푒

Table A-13 Results of automated mineralogy and mass distribution calculations for Hambone Lake sediments.

As-Hosting Solid Phase Number of Particles Total Area of Particles Density As Mass Distribution Units -- µm2 g·cm-3 wt.% % Hambone Lake (1) [ID: HAM-1] Fe/Mn-(oxy)hydroxide 1131 38381 4.28 0.50* 8.26 Framboidal pyrite 1728 84733 5.01 0.20* 8.53 Arsenopyrite 12 35 6.07 46.0 0.99 As-Sulphide 283 1109 3.56 70.0 27.8 Phyllosilicate 2795 141350 3.09 0.10* 4.39 A FeSx/FeO 9013 535716 4.65 0.20* 50.0 Hambone Lake (2) [ID: HAM-2] Fe/Mn-(oxy)hydroxide 684 21960 4.28 0.50* 5.38 Framboidal pyrite 1358 61859 5.01 0.20* 7.08 Arsenopyrite 10 57 6.07 46.0 1.83 As-Sulphide 93 589 3.56 70.0 16.8 Phyllosilicate 2795 146958 3.09 0.10* 5.20 A FeSx/FeO 8662 599466 4.65 0.20* 63.7 Hambone Lake (5) [ID: HAM-5] Fe/Mn-(oxy)hydroxide 534 17706 4.28 0.50* 9.2 Framboidal pyrite 2977 100567 5.01 0.20* 24.3 212

Arsenopyrite 4 11 6.07 46.0 0.73 As-Sulphide 44 433 3.56 70.0 26.1 Phyllosilicate 1291 49876 3.09 0.10* 3.73 A FeSx/FeO 9681 160200 4.65 0.20* 36.0 1 1 * Determined using electron microprobe analysis (EPMA); A 휌 = 휌 + 휌 (FeSxFeO) 2 푝푦푟𝑖푡푒 2 푓푒푟푟𝑖ℎ푦푑푟𝑖푡푒

Table A-14 Results of automated mineralogy and mass distribution calculations for Matthews Lake sediments.

Table A-15 Results of automated mineralogy and mass distribution calculations for Control Lake sediments.

Figure A-1: Box and whisker plot of sedimentary As concentrations from all lakes sampled. Sediment cores are divided into distinct geochemical clusters: impacted (filled boxes) and background (hollow boxes), as defined by CONISS analysis. Note that samples were collected with a sediment gravity corer for all lakes except Hambone Lake which was sampled with an Ekman Grab sampler. On these plots, the ends of the boxes represent the upper and lower quartiles, the median value is marked by a horizontal line inside the box, the whiskers extend to the to the farthest points that are not outliers (i.e., 1.5 times the interquartile range), and data that exceed this range are plotted as a circle. The dotted line denotes the CCME Sediment Guideline for the protection of aquatic life (CCME, 2001a).

214

Figure A-2 EDS spectra to supplement Figure 2.5.

215

216

217

218

Appendix B Supplementary data to Chapter 3

The data presented in this appendix reflect the supplementary material that will appear online in association with the publication of Chapter 3 in Science of the Total Environment.

The Excel files listed provide more comprehensive data and associated QA/QC:

· Miller_BulkGeochemistry_Sediment

· Miller_AqueousGeochemistry

· Miller_RockEval

· Miller_InferredChlorophyllA

· Miller_GrainSizeAnalysis

· Miller_EPMA_AppendixB

· Miller_SpearmansCorrelations_AppendixB

· Miller_XANES_AppendixB

219

Methods B-1 Detailed methodology for programmed pyrolysis (Rock-Eval® 6; Vinci Technologies, France).

Sediment sub-samples (50 mg; Matthews Lake n = 52; Control Lake n = 48) were first subjected to pyrolysis under an inert atmosphere (N2). During pyrolysis, the quantity of free/volatile hydrocarbons present in the sample (S1, mg HC·g-1) and the amount of hydrocarbon released by the thermal cracking of OM (S2, mg HC·g-1) were detected with a Flame Ionization

-1 -1 Detector. Simultaneously, S3CO2 (mg CO2·g ) and S3CO (mg CO·g ), derived from thermal cracking of oxygen-bearing compounds, were measured by infrared spectroscopy (Lafargue et al.,

1998; Carrie et al., 2012). Following the pyrolysis stage, the sample was transferred to an oxidation

-1 -1 oven and all remaining OM (S4CO2, mg CO2·g ; S4CO, mg CO·g ) and residual carbon (RC, %) was measured. Total organic carbon (TOC; wt %), was determined from the sum of all carbon released during pyrolysis and oxidation (TOC = pyrolysable organic carbon (PC %) + residual organic carbon (RC%)). The hydrogen index (HI = S2 × 100/TOC, mg HC/g TOC) and oxygen

-1 index (OI = S3 × 100/TOC, mg CO2·g TOC) were calculated.

220

Table B-1 Results of 210Pb dating on Matthews Lake sediment core.

Table B-2 Results of 210Pb dating on Control Lake sediment core.

221

Table B-3 Unprocessed 210Pb (Gamma) and 14C (AMS) (bulk, humic, and humin) dates from Matthews and Control lakes.

Sample ID Method Depth (cm) Age (yrs*/cal BP ± 1c) Material dated cal BPA Matthews Lake UOC‐5503 AMS 2.5 317 ± 22 Bulk lake sediment 460‐347 (74.8%)

UOC-8809 AMS 2.5 366 ± 21 Humin 499–426 (58.0%) UOC-8814 AMS 2.5 400 ± 27 Humic 514–429 (80.1%) UOC‐5504 AMS 15.5 2406 ± 22 Bulk lake sediment 2491‐2351 (91.9%)

UOC-8810 AMS 15.5 2407 ± 32 Humin 2499–2348 (80.9%) UOC-8815 AMS 15.5 2299 ± 25 Humic 2355–2306 (86.1%) UOC‐5505 AMS 25.5 2982 ± 22 Bulk lake sediment 3218‐3074 (95.4%) UOC‐5506 AMS 34.5 3755 ± 23 Bulk lake sediment 4160‐4076 (76.1%)

UOC‐5507 AMS 44.5 4455 ± 24 Bulk lake sediment 5280‐5163 (47.2%)

UOC-8811 AMS 44.5 4536 ± 30 Humin 5187–5057 (50.1%)

UOC-8816 AMS 44.5 4588 ± 43 Humic 5332–5258 (35.7%) UOC-2128 AMS 51.5 4872 ± 24 Bulk lake sediment -- MAT:0-1 Gamma 0.5 2014 ± 2* Bulk lake sediment -- MAT:1-2 Gamma 1.5 2006 ± 2* Bulk lake sediment -- MAT:2-3 Gamma 2.5 1996 ± 2* Bulk lake sediment -- MAT:3-4 Gamma 3.5 1988 ± 3* Bulk lake sediment -- MAT:4-5 Gamma 4.5 1981 ± 4* Bulk lake sediment -- MAT:5-6 Gamma 5.5 1976 ± 4* Bulk lake sediment -- MAT:7-8 Gamma 7.5 1959 ± 7* Bulk lake sediment -- MAT:9-10 Gamma 9.5 1942 ± 11* Bulk lake sediment -- MAT:10-11 Gamma 10.5 1930 ± 16* Bulk lake sediment -- MAT:14-15 Gamma 14.5 -- Bulk lake sediment -- MAT:20-21 Gamma 20.5 -- Bulk lake sediment -- MAT:24-25 Gamma 24.5 -- Bulk lake sediment -- Control Lake UOC‐5489 AMS 4.5 Modern Bulk lake sediment 1955‐1956* (95.4%) UOC‐5490 AMS 12.5 1136 ± 22 Bulk lake sediment 1087‐968 (90.2%)

UOC-8812 AMS 12.5 1212 ± 27 Humin 1187–1061(83.7%) UOC-8817 AMS 12.5 1318 ± 35 Humic 1300–1181(95.4%) UOC‐5491 AMS 19.5 1350 ± 23 Bulk lake sediment 1308‐1260 (93.9%) UOC‐5492 AMS 25.5 1133 ± 22 Bulk lake sediment 1085‐965 (92.3%) UOC‐5493 AMS 31.5 1761 ± 22 Bulk lake sediment 1730‐1605 (95.4%)

UOC-8813 AMS 31.5 1837 ± 23 Humin 1827–1710(93.9%) UOC-8818 AMS 31.5 NHA Humic NHA UOC‐5494 AMS 39.5 1958 ± 22 Bulk lake sediment 1952‐1865 (91.6%) UOC-2129 AMS 47.5 2223 ± 25 Bulk lake sediment --

222

CON:0-1 Gamma 0.5 2015 ± 1* Bulk lake sediment -- CON:1-2 Gamma 1.5 2010 ± 1* Bulk lake sediment -- CON:2-3 Gamma 2.5 2004 ± 1* Bulk lake sediment -- CON:3-4 Gamma 3.5 1997 ± 1.5* Bulk lake sediment -- CON:4-5 Gamma 4.5 1987 ± 2.0* Bulk lake sediment -- CON:5-6 Gamma 5.5 1978 ± 3* Bulk lake sediment -- CON:6-7 Gamma 6.5 1972 ± 4* Bulk lake sediment -- CON:7-8 Gamma 7.5 1967 ± 4* Bulk lake sediment -- CON:9-10 Gamma 9.5 1958 ± 6* Bulk lake sediment -- CON:12-13 Gamma 12.5 1946 ± 8* Bulk lake sediment -- CON:14-15 Gamma 14.5 1939 ± 10* Bulk lake sediment -- CON:20-21 Gamma 20.5 1895 ± 24* Bulk lake sediment -- CON:24-25 Gamma 24.5 -- Bulk lake sediment -- CON:28-29 Gamma 28.5 -- Bulk lake sediment -- CON:32-33 Gamma 32.5 -- Bulk lake sediment -- CON:36-37 Gamma 36.5 -- Bulk lake sediment -- NHA – No humic acid produced; * Date reported in year cal. AD; A Calibrated results are given as a range with an associated probability.

Table B-4 Arsenic speciation results based on linear combination fitting of XANES spectra.

Depth As(-I)-S As(III)-Sa As(III)-Oa As(V)-Oa Fitted Sumb R valuec Reduced χ2 Reference Arsenopyrite Orpiment Hematite-As(III) HFO-As(V) ------cm % % % % % -- -- Matthews Lake 0 - 1 -- 3.00 30.0 67.0 107.8 0.028 0.013 4 - 5 -- -- 27.0 73.0 114.3 0.050 0.030 44 - 45 0.30 23.5 24.6 51.8 113.2 0.073 0.034 Control Lake 1 - 2 -- 1.00 35.0 64.0 101.8 0.004 0.002 3 - 4 -- -- 37.0 63.0 110.5 0.034 0.017 8 - 9 -- 8.00 43.0 49.0 104.2 0.015 0.006 28 - 29 -- 8.00 46.0 46.0 103.6 0.008 0.003 39 - 40 10.0 2.00 46.0 41.0 103.5 0.020 0.007 a Component sums were normalized to 100%; b Fitted sum of all references before normalization to a sum of 100%; c Mean square misfit between the data and the fit.

223

Table B-5 Concentrations of As, S, Fe, organic matter fractions S1, S2, S3 and TOC, grain size distribution, and porewater pH and As concentrations from the Matthews Lake (MAT) sediment core (64.06014° N, 111.23566° W).

Sample Depth As S Fe S1 S2 S3 TOC Chlr-A Sand Silt Clay pH As Units cm mg·kg-1 % % mg HC/gA mg HC/gA mg HC/gA wt. % mg·g-1 % % % s.u. µg·L-1 D.L. -- 0.1 0.02 0.01 0.01 0.01 0.01 0.2 0.01 ------0.09 MAT 0-1 0.5 89.5 0.34 2.68 14.5 27.6 19.2 11.5 0.103 11.8 74.9 13.3 5.5 2.31 MAT 1-2 1.5 66.8 0.32 2.35 16.9 30.7 18.5 11.4 0.088 8.22 81.2 10.6 5.3 3.16 MAT 2-3 2.5 48.0 0.29 2.10 16.1 30.3 17.1 10.7 0.068 14.7 74.0 11.3 5.3 3.94 MAT 3-4 3.5 47.0 0.29 2.02 15.2 30.5 16.8 10.6 0.055 13.3 76.4 10.3 5.5 4.10 MAT 4-5 4.5 63.0 0.29 2.19 13.8 29.5 15.2 9.60 0.049 33.1 59.5 7.43 5.3 4.32 MAT 5-6 5.5 50.5 0.29 2.08 13.4 30.6 17.1 10.8 0.045 15.5 73.0 11.5 5.4 4.81 MAT 6-7 6.5 47.9 0.29 2.09 12.9 31.0 16.4 10.5 0.040 12.0 76.4 11.6 5.3 4.66 MAT 7-8 7.5 49.3 0.29 2.10 13.2 32.2 15.8 10.3 0.038 7.59 80.3 12.1 5.3 3.82 MAT 8-9 8.5 52.7 0.28 2.27 13.5 34.6 15.8 10.1 0.036 12.3 77.5 10.2 5.7 3.56 MAT 9-10 9.5 40.1 0.26 2.09 14.6 36.0 15.7 10.2 0.033 13.0 75.8 11.2 5.7 3.85 MAT 10-11 10.5 38.3 0.27 2.07 14.6 36.3 15.7 10.3 0.036 11.6 78.1 10.2 5.8 3.04 MAT 11-12 11.5 31.2 0.25 1.89 14.8 36.3 15.8 10.5 0.037 14.4 74.8 10.8 5.7 3.19 MAT 12-13 12.5 30.8 0.24 1.87 14.5 37.2 16.2 10.9 0.034 ------5.8 3.93 MAT 13-14 13.5 36.4 0.29 2.15 15.4 37.5 17.0 11.5 0.037 11.0 77.3 11.7 5.7 3.19 MAT 14-15 14.5 33.7 0.31 2.02 16.2 38.6 17.2 11.6 0.042 13.2 75.7 11.0 5.6 3.38 MAT 15-16 15.5 35.7 0.30 1.96 15.1 37.1 17.3 11.7 0.039 10.0 78.3 11.7 5.6 3.67 MAT 16-17 16.5 32.0 0.33 2.11 14.6 35.8 17.6 11.8 0.037 9.11 78.0 12.9 5.7 1.96 MAT 17-18 17.5 32.6 0.33 2.21 15.0 35.7 18.0 12.0 0.034 8.66 78.4 12.9 5.7 2.35 MAT 18-19 18.5 34.0 0.34 2.38 14.2 35.6 19.2 12.6 0.031 10.1 76.8 13.0 5.6 2.24 MAT 19-20 19.5 37.8 0.33 2.47 13.9 33.2 18.2 12.0 0.029 12.6 76.1 11.3 5.6 2.50 MAT 20-21 20.5 34.6 0.30 2.24 14.8 36.1 17.6 11.7 0.035 9.27 76.1 14.7 5.4 1.99 MAT 21-22 21.5 31.3 0.31 2.25 15.4 36.6 17.8 11.7 0.032 ------5.6 2.21 MAT 22-23 22.5 34.3 0.30 2.23 14.8 35.9 17.4 11.5 0.030 6.32 80.4 13.3 5.4 2.07 MAT 23-24 23.5 36.6 0.31 2.30 14.5 34.7 16.4 10.9 0.031 6.93 79.1 14.0 5.4 2.36

224

MAT 24-25 24.5 35.4 0.31 2.20 14.4 36.0 17.2 11.5 0.033 40.1 52.8 7.06 5.5 2.11 MAT 25-26 25.5 32.3 0.32 2.28 15.6 38.5 17.6 11.7 0.029 10.2 76.6 13.2 5.5 2.84 MAT 26-27 26.5 31.7 0.32 2.17 15.4 38.0 16.8 11.3 0.029 12.4 75.7 11.9 5.6 2.69 MAT 27-28 27.5 32.0 0.31 2.12 14.8 36.2 16.4 11.1 0.031 13.0 76.4 10.6 5.5 3.08 MAT 28-29 28.5 32.1 0.32 2.22 14.0 36.0 16.5 11.4 0.033 21.7 68.7 9.58 5.2 3.31 MAT 29-30 29.5 32.5 0.32 2.20 15.4 38.0 17.2 11.8 0.029 10.8 76.7 12.5 5.2 3.51 MAT 30-31 30.5 32.2 0.32 2.15 16.2 39.1 17.2 11.8 0.027 7.44 79.2 13.3 5.4 5.54 MAT 31-32 31.5 29.8 0.28 1.97 15.9 39.4 16.6 11.4 0.026 14.5 74.5 11.0 5.2 2.07 MAT 32-33 32.5 29.6 0.30 1.97 15.3 38.8 16.5 11.2 0.029 12.1 75.5 12.4 5.2 4.09 MAT 33-34 33.5 29.6 0.31 1.94 15.4 36.6 16.5 11.5 0.026 12.7 74.7 12.5 5.2 6.97 MAT 34-35 34.5 30.7 0.31 1.91 15.8 38.7 16.7 11.7 0.023 7.51 79.4 13.1 -- -- MAT 35-36 35.5 30.5 0.33 1.94 17.6 41.7 16.7 11.7 0.024 16.6 72.0 11.3 5.2 6.69 MAT 36-37 36.5 30.1 0.31 1.87 15.6 36.7 15.8 11.3 0.023 4.61 80.7 14.7 -- -- MAT 37-38 37.5 29.2 0.30 1.73 16.7 39.6 14.9 10.7 0.024 2.69 80.6 16.8 5.0 27.1 MAT 38-39 38.5 31.4 0.32 1.85 16.6 41.1 15.9 11.4 0.025 10.9 79.0 10.1 5.1 25.1 MAT 39-40 39.5 33.8 0.37 2.09 17.2 40.1 17.6 12.4 0.027 5.64 79.1 15.3 -- -- MAT 40-41 40.5 33.5 0.37 1.86 16.6 40.6 15.4 10.9 0.027 8.78 77.8 13.5 5.0 51.5 MAT 41-42 41.5 41.4 0.45 1.86 16.4 42.5 16.1 11.5 0.026 12.9 75.4 11.6 5.0 59.2 MAT 42-43 42.5 40.4 0.43 1.87 16.7 40.8 16.6 11.9 0.025 8.86 77.4 13.7 4.9 23.9 MAT 43-44 43.5 37.0 0.42 1.83 16.0 40.1 15.9 11.5 0.022 10.2 76.5 13.4 4.8 -- MAT 44-45 44.5 45.9 0.49 1.85 14.3 34.5 16.0 11.5 0.025 5.23 80.8 13.9 5.2 43.0 MAT 45-46 45.5 43.5 0.49 1.83 16.1 38.2 16.4 11.9 0.021 9.45 76.6 13.9 4.9 47.5 MAT 46-47 46.5 48.8 0.55 1.90 16.1 39.5 16.1 11.6 0.022 10.9 75.8 13.4 5.1 84.4 MAT 47-48 47.5 36.2 0.46 1.76 15.9 40.1 15.4 11.2 0.022 8.15 79.2 12.6 5.1 47.0 MAT 48-49 48.5 34.6 0.46 1.71 14.5 27.6 13.2 9.7 0.046 7.90 78.5 13.6 5.2 49.6 MAT 49-50 49.5 34.5 0.42 1.69 16.9 30.7 14.7 10.8 0.043 8.46 78.8 12.7 5.1 54.3 MAT 50-51 50.5 33.3 0.42 1.65 16.1 30.3 14.8 10.8 0.038 4.48 81.0 14.6 5.1 57.3 MAT 51-52 51.5 38.9 0.46 1.67 15.2 30.5 15.1 11.1 0.034 12.0 76.5 11.5 5.0 -- A mg HC/g – milligrams of hydrocarbon per gram of sediment.

225

Table B-6 Concentrations of As, S, Fe, organic matter fractions S1, S2, S3 and TOC, grain size distribution, and porewater pH and As concentration from the Control Lake (CON) sediment core (64.07771° N, 111.13493° W).

Sample Depth As S Fe S1 S2 S3 TOC Chlr-A Sand Silt Clay pH As mg mg mg Units cm mg·kg-1 % % wt. % mg·g-1 % % % s.u. µg·L-1 HC/gA HC/gA HC/gA Detection Limit -- 0.1 0.02 0.01 0.01 0.01 0.01 0.2 0.01 ------0.09 CON 0-1 0.5 224 0.37 7.31 28.2 46.7 29.4 17.2 0.103 34.9 54.3 10.8 6.0 5.66 CON 1-2 1.5 218 0.37 6.88 27.6 42.5 27.8 16.5 0.088 19.9 66.7 13.3 6.0 3.34 CON 2-3 2.5 151 0.38 5.92 24.2 44.0 24.3 15.5 0.068 19.0 68.0 12.9 5.7 2.32 CON 3-4 3.5 109 0.39 5.11 26.2 48.6 24.3 16.3 0.055 47.7 44.7 7.63 5.8 1.19 CON 4-5 4.5 91.1 0.38 4.82 23.2 46.7 21.9 15.3 0.049 26.3 59.4 14.3 5.6 0.75 CON 5-6 5.5 86.1 0.37 4.36 20.2 44.7 19.8 14.4 0.045 26.0 62.4 11.6 5.6 0.96 CON 6-7 6.5 89.3 0.36 4.22 19.0 42.7 18.7 13.4 0.040 28.3 59.2 12.5 5.6 1.31 CON 7-8 7.5 104 0.37 4.80 17.8 44.3 19.7 14.1 0.038 36.1 54.1 9.81 5.6 0.83 CON 8-9 8.5 102 0.39 4.95 18.6 42.8 19.3 13.7 0.036 12.7 70.8 16.5 5.6 0.64 CON 9-10 9.5 89.3 0.34 4.18 17.4 42.3 18.7 13.4 0.033 44.5 48.3 7.12 5.6 0.88 CON 10-11 10.5 107 0.36 4.84 17.5 40.1 17.8 12.8 0.036 0.00 75.2 24.8 5.6 1.93 CON 11-12 11.5 94.8 0.36 4.42 17.9 42.4 18.9 13.5 0.037 7.23 72.3 20.5 5.6 0.84 CON 12-13 12.5 95.4 0.36 4.64 16.9 39.5 17.7 12.7 0.034 2.96 74.0 23.1 5.7 1.22 CON 13-14 13.5 110 0.34 5.10 15.9 38.9 18.4 12.9 0.037 5.93 72.3 21.8 5.3 1.11 CON 14-15 14.5 129 0.38 5.79 17.5 43.0 20.5 14.2 0.042 2.07 75.4 22.5 5.6 1.13 CON 15-16 15.5 152 0.36 6.16 15.0 38.5 18.9 13.0 0.039 6.31 73.0 20.7 5.7 1.34 CON 16-17 16.5 147 0.34 5.87 15.3 37.8 18.5 12.7 0.037 1.05 76.3 22.6 5.4 0.96 CON 17-18 17.5 137 0.32 5.49 14.7 39.6 18.7 13.0 0.034 3.20 76.5 20.3 5.4 1.39 CON 18-19 18.5 116 0.34 4.84 15.2 37.5 17.1 12.2 0.031 3.57 77.4 19.0 5.6 1.09 CON 19-20 19.5 73.6 0.29 3.32 16.7 40.7 16.6 12.5 0.029 6.61 77.2 16.2 5.6 0.89 CON 20-21 20.5 155 0.34 6.04 14.6 37.5 18.6 12.7 0.035 10.3 72.5 17.2 5.5 1.23 CON 21-22 21.5 121 0.32 5.13 16.4 42.7 19.3 13.8 0.032 10.5 73.2 16.3 5.7 2.23 CON 22-23 22.5 127 0.30 5.30 15.9 39.0 18.5 13.1 0.030 3.69 76.1 20.2 5.6 2.52

226

CON 23-24 23.5 134 0.32 5.78 15.2 39.3 19.3 13.3 0.031 7.04 75.4 17.5 5.6 2.64 CON 24-25 24.5 153 0.32 6.43 15.4 40.7 21.6 14.4 0.033 5.02 76.1 18.9 5.8 2.38 CON 25-26 25.5 152 0.31 6.15 14.6 39.5 21.3 13.9 0.029 7.99 74.7 17.3 5.5 1.93 CON 26-27 26.5 108 0.30 4.75 15.6 36.9 19.1 13.1 0.029 10.3 73.2 16.5 5.5 2.62 CON 27-28 27.5 137 0.33 5.59 14.4 36.5 19.0 12.8 0.031 5.78 75.3 18.9 5.3 5.42 CON 28-29 28.5 167 0.34 6.90 14.5 35.7 19.2 12.8 0.033 7.76 73.5 18.8 5.3 3.29 CON 29-30 29.5 114 0.30 4.99 13.6 32.7 16.3 11.2 0.029 12.7 71.1 16.2 5.6 3.78 CON 30-31 30.5 109 0.30 4.74 13.8 33.9 16.5 11.4 0.027 16.1 70.6 13.2 5.6 5.03 CON 31-32 31.5 109 0.30 4.83 13.3 35.4 16.6 11.7 0.026 ------4.11 CON 32-33 32.5 104 0.33 4.68 15.4 39.8 17.7 12.8 0.029 14.4 70.6 15.0 5.7 5.81 CON 33-34 33.5 89.1 0.32 4.01 15.5 40.6 17.6 12.8 0.026 15.6 70.3 14.1 5.5 2.96 CON 34-35 34.5 76.4 0.31 3.59 15.0 38.7 15.5 11.7 0.023 17.9 70.8 11.3 5.7 6.45 CON 35-36 35.5 86.0 0.30 3.85 14.5 36.3 16.3 11.9 0.024 ------5.5 5.98 CON 36-37 36.5 87.4 0.34 4.01 15.5 40.0 17.4 12.8 0.023 15.2 71.6 13.2 5.9 7.92 CON 37-38 37.5 95.0 0.36 4.41 15.3 40.3 17.8 13.1 0.024 13.6 72.8 13.7 5.7 5.76 CON 38-39 38.5 105 0.36 4.72 15.8 40.5 18.1 13.1 0.025 9.07 74.9 16.0 5.5 2.44 CON 39-40 39.5 130 0.38 5.80 14.7 39.3 19.2 13.3 0.027 9.82 72.7 17.5 5.5 13.8 CON 40-41 40.5 119 0.36 5.43 14.9 41.1 19.3 13.6 0.027 0.00 0.0 0.0 5.3 14.8 CON 41-42 41.5 110 0.37 5.07 14.5 43.5 19.5 14.1 0.026 10.0 74.0 16.0 5.2 10.4 CON 42-43 42.5 90.8 0.37 4.19 17.7 50.3 20.9 15.5 0.025 7.40 77.3 15.3 -- 5.85 CON 43-44 43.5 68.5 0.38 3.29 17.1 48.6 18.7 14.4 0.022 7.98 78.0 14.0 -- 4.39 CON 44-45 44.5 72.6 0.36 3.31 16.2 47.4 18.1 13.9 0.025 3.87 80.1 16.0 -- 5.21 CON 45-46 45.5 63.8 0.36 2.96 17.9 53.7 20.8 15.7 0.021 14.9 72.6 12.5 -- 8.62 CON 46-47 46.5 56.1 0.31 2.82 17.2 53.5 22.7 16.7 0.022 13.6 74.8 11.6 -- 3.38 CON 47-48 47.5 81.4 0.34 3.99 16.4 46.6 21.3 15.2 0.022 8.83 77.0 14.2 -- 9.36 A mg HC/g – milligrams of hydrocarbon per gram of sediment.

227

Table B-7 Spearman’s correlation matrix of selected elemental and organic parameters for Matthews Lake sediment from 14 – 19 cm depth (ML-G3). Bold values indicate p ≤ 0.05; bold and italicized indicate p ≤ 0.001.

As S Fe S1 S2 S3 TOC PC RC HI OI Chl-a Sand Silt Clay As 1.00 -0.33 0.32 0.18 -0.50 -0.04 -0.21 -0.21 -0.04 -0.47 0.15 -0.21 0.57 -0.36 -0.43 S 1.00 0.63 0.59 0.96 0.93 0.93 0.93 0.93 -0.51 0.82 -0.74 -0.26 -0.07 0.59 Fe 1.00 0.57 0.54 0.71 0.64 0.64 0.71 -0.62 0.75 -0.93 0.07 -0.21 0.25 S1 1.00 0.61 0.75 0.79 0.79 0.75 -0.93 0.77 -0.57 -0.11 0.04 0.46 S2 1.00 0.86 0.93 0.93 0.86 -0.44 0.73 -0.64 -0.43 0.11 0.71 S3 1.00 0.96 0.96 1.00 -0.75 0.97 -0.86 -0.25 0.00 0.57 TOC 1.00 1.00 0.96 -0.69 0.92 -0.79 -0.43 0.18 0.71 Chl-a 1.00 0.96 -0.69 0.92 -0.79 -0.43 0.18 0.71 PC 1.00 -0.75 0.97 -0.86 -0.25 0.00 0.57 RC 1.00 -0.81 0.64 -0.07 0.11 -0.33 HI 1.00 -0.90 -0.24 0.06 0.51 OI 1.00 0.18 0.00 -0.43 Sand 1.00 -0.93 -0.79 Silt 1.00 0.61 Clay 1.00

Table B-8: Spearman’s correlation matrix of selected elemental and organic parameters for Matthews Lake sediment from 20 – 37 cm depth (ML-G4). Bold values indicate p ≤ 0.05; bold and italicized indicate p ≤ 0.001.

As S Fe S1 S2 S3 TOC PC RC HI OI Chl-a Sand Silt Clay As 1.00 0.84 0.09 -0.32 -0.43 0.48 0.17 -0.22 0.50 -0.74 0.70 0.70 -0.21 0.18 0.19 S 1.00 0.04 -0.51 -0.63 0.47 0.05 -0.49 0.57 -0.89 0.83 0.81 -0.10 0.00 0.12 Fe 1.00 0.49 0.44 0.06 0.44 0.46 0.13 0.28 -0.33 -0.16 0.20 -0.09 -0.16 S1 1.00 0.88 0.20 0.71 0.94 0.12 0.68 -0.66 -0.72 0.05 0.08 0.04 S2 1.00 0.04 0.62 0.95 0.01 0.85 -0.81 -0.78 0.03 0.05 0.04 S3 1.00 0.70 0.26 0.88 -0.37 0.49 0.24 -0.09 0.03 0.14

228

TOC 1.00 0.78 0.73 0.22 -0.18 -0.22 -0.01 0.04 0.13 Chl-a 1.00 0.20 0.71 -0.66 -0.67 -0.02 0.09 0.13 PC 1.00 -0.39 0.41 0.35 0.02 -0.10 0.08 RC 1.00 -0.95 -0.85 0.07 0.03 -0.06 HI 1.00 0.78 -0.15 0.02 0.12 OI 1.00 0.18 -0.28 -0.15 Sand 1.00 -0.94 -0.93 Silt 1.00 0.79 Clay 1.00

Table B-9 Spearman’s correlation matrix of selected elemental and organic parameters for Matthews Lake sediment from 38 – 52 cm depth (ML-G5). Bold values indicate p ≤ 0.05; bold and italicized indicate p ≤ 0.001.

As S Fe S1 S2 S3 TOC PC RC HI OI Chl-a Sand Silt Clay As 1.00 0.88 0.33 0.55 0.65 0.50 0.59 0.62 0.41 0.29 -0.16 -0.44 0.43 -0.56 -0.15 S 1.00 0.02 0.24 0.50 0.19 0.31 0.38 0.08 0.59 -0.46 -0.28 0.24 -0.36 -0.12 Fe 1.00 0.57 0.49 0.83 0.75 0.59 0.81 -0.71 0.80 -0.39 0.25 -0.30 0.08 S1 1.00 0.73 0.72 0.80 0.87 0.66 -0.12 0.26 -0.36 0.24 -0.33 0.08 S2 1.00 0.82 0.91 0.96 0.75 0.14 0.17 -0.53 0.20 -0.21 0.14 S3 1.00 0.97 0.88 0.98 -0.38 0.63 -0.54 0.33 -0.31 0.10 TOC 1.00 0.95 0.94 -0.21 0.47 -0.54 0.34 -0.33 0.08 Chl-a 1.00 0.81 0.00 0.28 -0.54 0.25 -0.29 0.15 PC 1.00 -0.46 0.66 -0.49 0.36 -0.28 0.03 RC 1.00 -0.88 0.07 -0.06 0.04 -0.06 HI 1.00 -0.26 0.00 0.01 0.28 OI 1.00 -0.21 0.23 -0.05 Sand 1.00 -0.86 -0.80 Silt 1.00 0.52 Clay 1.00

229

Table B-10 Spearman’s correlation matrix of selected elemental and organic parameters for Control Lake sediment core from 15 to 36 cm depth (CL-G3). Bold values indicate p ≤ 0.05; bold and italicized indicate p ≤ 0.001.

As S Fe S1 S2 S3 TOC PC RC HI OI Chl-a Sand Silt Clay As 1.00 0.55 0.97 -0.35 -0.15 0.65 0.36 -0.08 0.59 -0.67 0.84 -0.20 -0.48 0.27 0.66 S 0.56 -0.02 0.06 0.32 0.15 0.01 0.24 -0.04 0.24 -0.23 -0.41 0.06 0.58 Fe 1.00 -0.25 -0.08 0.72 0.46 0.02 0.66 -0.68 0.83 -0.16 -0.51 0.27 0.70 S1 1.00 0.75 0.20 0.49 0.79 0.28 0.45 -0.49 0.14 -0.25 0.32 0.11 S2 1.00 0.37 0.63 0.96 0.36 0.51 -0.40 0.38 -0.20 0.27 0.08 S3 1.00 0.89 0.50 0.99 -0.54 0.59 -0.02 -0.39 0.29 0.47 TOC 1.00 0.75 0.91 -0.25 0.27 0.22 -0.34 0.26 0.37 Chl-a 1.00 0.51 0.34 -0.26 0.39 -0.24 0.30 0.14 PC 1.00 -0.54 0.56 -0.01 -0.40 0.32 0.46 RC 1.00 -0.93 0.26 0.14 -0.04 -0.27 HI 1.00 -0.29 -0.27 0.11 0.42 OI 1.00 0.31 -0.27 -0.34 Sand 1.00 -0.85 -0.94 Silt 1.00 0.66 Clay 1.00

Table B-11 Spearman’s correlation matrix of selected elemental and organic parameters for Control Lake sediment core from 37 to 48 cm depth (CL-G4). Bold values indicate p 0.05; bold and italicized indicate p ≤ 0.001.

As S Fe S1 S2 S3 TOC PC RC HI OI Chl-a Sand Silt Clay

As 1.00 0.47 1.00 -0.79 -0.78 -0.32 -0.62 -0.78 -0.44 -0.71 0.73 0.86 -0.08 -0.33 0.78 S 1.00 0.47 -0.22 -0.14 -0.12 -0.09 -0.21 -0.12 0.00 -0.01 0.21 -0.41 0.14 0.61 Fe 1.00 -0.79 -0.78 -0.32 -0.62 -0.78 -0.44 -0.71 0.73 0.86 -0.08 -0.33 0.78 S1 1.00 0.88 0.53 0.76 0.87 0.59 0.67 -0.62 -0.60 -0.12 0.39 -0.59 S2 1.00 0.63 0.89 0.98 0.72 0.76 -0.69 -0.58 -0.15 0.47 -0.52 S3 1.00 0.90 0.72 0.98 0.04 0.08 -0.03 -0.07 0.28 -0.23 230

TOC 1.00 0.93 0.95 0.45 -0.35 -0.38 -0.11 0.40 -0.40 Chl-a 1.00 0.80 0.65 -0.60 -0.53 -0.13 0.48 -0.55 PC 1.00 0.17 -0.07 -0.18 -0.13 0.38 -0.28 RC 1.00 -0.94 -0.76 -0.11 0.32 -0.42 HI 1.00 0.82 0.21 -0.43 0.41 OI 1.00 0.10 -0.39 0.52 Sand 1.00 -0.89 -0.59 Silt 1.00 0.22 Clay 1.00

Table B-13 EMPA analysis of scorodite grains.

A Grain ID As2O5 SiO2 SO3 FeO H2O Total Units wt. % wt. % wt. % wt. % wt. % wt. % MAT27 453 51.79 0.95 1.24 35.01 15.60 104.59 MAT27 12815 52.52 2.39 1.21 34.94 15.60 106.66 MAT27 18377 45.69 1.21 1.37 32.82 15.60 96.69 MAT27 19679 45.66 1.65 1.22 31.34 15.60 95.47 MAT37 5219 51.63 0.97 1.34 33.52 15.60 103.06 MAT37 6298 39.11 1.14 1.17 24.98 15.60 82.00 MAT37 7195 44.06 9.57 1.15 28.95 15.60 99.33

A Constant 15.60 wt % value chosen based on stoichiometric values for scorodite; H2O is added to allow for post-hoc matrix corrections.

231

Table B-14 Results of automated mineralogy and mass distribution calculations for Matthews Lake.

As-Hosting Solid Phase Number of Particles Total Area of Particles Density As Mass Distribution Units -- µm2 g·cm-3 wt.% % Matthews Lake (0 – 1 cm) [ID: MAT 0-1] Fe/Mn-(oxy)hydroxide 10062 177853 4.28 0.30* 99.2 Framboidal pyrite 32 723 5.01 0.50* 0.80 Arsenopyrite 0 0 6.07 46 0 As-Sulphide 0 0 3.56 70 0 Scorodite 0 0 3.27 31 0 Matthews Lake (4 – 5 cm) [ID: MAT 4-5] Fe/Mn-(oxy)hydroxide 1808 82640 4.28 0.30* 98.8 Framboidal pyrite 46 1174 5.01 0.50* 1.20 Arsenopyrite 0 0 6.07 46 0 As-Sulphide 0 0 3.56 70 0 Scorodite 0 0 3.27 31 0 Matthews Lake (26 – 27 cm) [ID: MAT 26-27] Fe/Mn-(oxy)hydroxide 1289 17107 4.28 0.30* 8.30 Framboidal pyrite 37 1977 5.01 0.50* 1.90 Arsenopyrite 0 0 6.07 46 0 As-Sulphide 0 0 3.56 70 0 Scorodite 206 2345 3.27 31 89.8 Matthews Lake (36 – 37 cm) [ID: MAT 36-37] Fe/Mn-(oxy)hydroxide 1204 20798 4.28 0.30* 54.2 Framboidal pyrite 31 635 5.01 0.50* 3.20 Arsenopyrite 0 0 6.07 46 0.00 As-Sulphide 0 0 3.56 70 0.00 Scorodite 24 207 3.27 31 42.6 Matthews Lake (41 – 42 cm) [ID: MAT 41-42] Fe/Mn-(oxy)hydroxide 1478 8190 4.28 0.30* 17.4 Framboidal pyrite 749 18152 5.01 0.50* 75.3 Arsenopyrite 0 0 6.07 46 0.0 As-Sulphide 0 0 3.56 70 0.0 Scorodite 5 44 3.27 31 7.30 Matthews Lake (44 – 45 cm) [ID: MAT 44-45] Fe/Mn-(oxy)hydroxide 1842 18282 4.28 0.30* 20.0 Framboidal pyrite 1123 36147 5.01 0.50* 77.5 Arsenopyrite 2 8 6.07 46 2.0 As-Sulphide 0 0 3.56 70 0 Scorodite 1 6 3.27 31 0.50 * Determined using electron microprobe analysis (EPMA)

232

Table B-15 Results of automated mineralogy and mass distribution calculations for Control Lake.

As-Hosting Solid Phase Number of Particles Total Area of Particles Density As Mass Distribution Units -- µm2 g·cm-3 wt.% % Control Lake (0 – 1 cm) [ID: CON 0-1] Fe/Mn-(oxy)hydroxide 21529 1269196 4.28 0.30* 99.8 Framboidal pyrite 80 1603 5.01 0.50* 0.20 Arsenopyrite 0 0 6.07 46 0 As-Sulphide 0 0 3.56 70 0 Scorodite 0 0 3.27 31 0 Control Lake (2 – 3 cm) [ID: CON 2-3] Fe/Mn-(oxy)hydroxide 21585 1450646 4.28 0.30* 99.9 Framboidal pyrite 34 111 5.01 0.50* 0.01 Arsenopyrite 0 0 6.07 46 0 As-Sulphide 0 0 3.56 70 0 Scorodite 0 0 3.27 31 0 Control Lake (3 – 4 cm) [ID: CON 3-4] Fe/Mn-(oxy)hydroxide 24305 874422 4.28 0.30* 99.8 Framboidal pyrite 32 714 5.01 0.50* 0.2 Arsenopyrite 0 0 6.07 46 0 As-Sulphide 0 0 3.56 70 0 Scorodite 0 0 3.27 31 0 Control Lake (8 – 9 cm) [ID: CON 8-9] Fe/Mn-(oxy)hydroxide 17619 607543 4.28 0.30* 98.4 Framboidal pyrite 554 5171 5.01 0.50* 1.6 Arsenopyrite 0 0 6.07 46 0 As-Sulphide 0 0 3.56 70 0 Scorodite 0 0 3.27 31 0 Control Lake (28 – 29 cm) [ID: CON 28-29] Fe/Mn-(oxy)hydroxide 49289 3106630 4.28 0.30* 98.5 Framboidal pyrite 3024 23735 5.01 0.50* 1.46 Arsenopyrite 3 4 6.07 46 0.03 As-Sulphide 0 0 3.56 70 0 Scorodite 0 0 3.27 31 0 Control Lake (39 – 40 cm) [ID: CON 39-40] Fe/Mn-(oxy)hydroxide 6123 149552 4.28 0.30* 79.9 Framboidal pyrite 1482 19405 5.01 0.50* 20.1 Arsenopyrite 0 0 6.07 46 0 As-Sulphide 0 0 3.56 70 0 Scorodite 0 0 3.27 31 0 * Determined using electron microprobe analysis (EPMA)

233

Table B-16 Model compounds for post-hoc Linear Combination Fitting XANES analysis.

Formula As Species Source Edge As Position wt. %* Arsenopyrite FeAsS As(-I)-S Deloro, ON, Canada 11,867.0 46.01

Realgar As4S4 As(II)-S Hunan, China 11,867.5 70.03

Orpiment As2S3 As(III)-S Moldawa, Hungary 11,867.7 60.90

A Hematite-As(III) 0.026As·Fe2O3 Fe2O3-As(III) A 11869.7 1.20*

A HFO-As(V) 0.0012As·Fe(OH)3 Fe(OH)3- A, B 11873.2 0.08* As(V) * Determined by ICP-MS; A Adsorption procedures by Liu et al. (2006); B Synthesized following Schwertmann & Cornell (2000) and Liu et al. (2006)

234

Figure B-1 14C-only Bayesian age-depth model developed using Bacon for Matthews Lake and Control Lake.

235

Figure B-2 EDS spectra to supplement Figure 3.4.

236

237

238

Figure B-3 EPMA As K and L-edge wavelength scan of pyritized funginite from Control Lake sediment core (39 – 40 cm).

1.6 As La1,2 Channel 1/TAP 1.4 V = 15 kV I = 100 nA 1.2 step = 10 mm

1 tdwell = 5 s

0.8

0.6

0.4

0.2 counts per second per nanoamp

0 101 102 103 104 105 106 107 L [mm]

239

Appendix C Supplementary data for Chapter 4

The data presented in this appendix reflect the supplementary material that will be submitted in association with Chapter 4 for publication in a peer-reviewed journal following thesis submission.

The Excel files listed provide more comprehensive data and associated QA/QC:

· Miller_BulkGeochemistry_Sediment

· Miller_AqueousGeochemistry

· Miller_RockEval

· Miller_EPMA_AppendixC

· Miller_XANES_AppendixC

· Miller_µXRD_AppendixC

240

Table C-1 Concentrations of As, S, Fe, Mn and organic matter fractions (S1, S2, S3, TOC) in the Powder Mag Lake (POW) sediment core (64.05114° N, 111.15042° W).

Sample Depth As S Fe Mn S1 S2 S3 TOC mg mg mg Units cm mg·kg-1 wt. % wt. % mg·kg-1 wt. % HC/gA HC/gA HC/gA Detection Limit -- 0.1 0.02 0.01 1.0 0.01 0.01 0.01 -- POW 0-1 0.5 ISB IS IS IS 37.0 65.1 44.0 21.5 POW 1-2 1.5 210 1.29 1.71 102 28.2 48.6 29.3 15.6 POW 2-3 2.5 195 1.06 1.75 97.0 19.0 35.3 20.1 11.3 POW 3-4 3.5 271 1.28 2.29 108 10.1 26.6 14.7 8.67 POW 4-5 4.5 260 1.20 2.19 112 7.57 23.7 11.4 7.46 POW 5-6 5.5 191 0.98 1.86 106 8.54 28.5 12.7 8.69 POW 6-7 6.5 163 0.84 1.73 110 10.7 35.6 14.3 10.3 POW 7-8 7.5 114 0.65 1.50 102 10.5 35.0 13.8 9.97 POW 8-9 8.5 77.2 0.46 1.33 96.0 9.26 30.3 11.7 8.58 POW 9-10 9.5 69.5 0.50 1.28 95.0 10.3 35.9 13.3 9.88 POW 10-11 10.5 87.4 0.61 1.45 111 13.3 43.3 16.5 12.1 POW 11-12 11.5 89.7 0.62 1.52 120 13.3 46.0 17.8 13.0 POW 12-13 12.5 89.0 0.65 1.57 116 12.2 40.3 16.1 11.6 POW 13-14 13.5 81.2 0.56 1.86 109 12.7 38.9 16.0 11.4 POW 14-15 14.5 79.7 0.58 1.72 112 13.4 44.2 17.5 12.7 POW 15-16 15.5 66.0 0.60 1.54 114 13.8 45.6 17.5 12.9 POW 16-17 16.5 66.6 0.62 1.53 122 13.0 42.6 16.3 12.0 POW 17-18 17.5 58.0 0.60 1.64 123 13.6 47.5 18.1 13.4 POW 18-19 18.5 66.0 0.63 1.68 126 14.4 50.1 21.3 15.1 POW 19-20 19.5 70.7 0.67 1.84 134 15.3 54.7 23.8 16.8 POW 20-21 20.5 68.9 0.67 1.95 150 16.7 62.2 27.2 19.1 POW 21-22 21.5 52.6 0.51 1.78 148 13.4 47.3 20.7 14.7 POW 22-23 22.5 59.7 0.63 1.91 158 13.2 52.6 25.7 17.5 POW 23-24 23.5 56.8 0.60 1.88 149 12.1 45.1 21.5 15.0 POW 24-25 24.5 54.6 0.59 2.06 149 13.8 51.4 25.0 17.2 POW 25-26 25.5 52.5 0.58 2.15 155 12.7 48.9 24.0 16.3 POW 26-27 26.5 42.4 0.50 1.84 147 12.0 45.8 21.9 15.1 POW 27-28 27.5 44.5 0.52 1.80 125 14.2 52.8 22.7 16.3 POW 28-29 28.5 58.8 0.71 1.80 128 17.8 65.6 27.1 19.4 POW 29-30 29.5 52.6 0.72 1.65 122 17.2 63.9 25.4 18.6 A mg HC/g – milligrams of hydrocarbon per gram of sediment; B IS – insufficient sample.

241

Table C-2 Concentrations of As, S, Fe, Mn and organic matter fractions (S1, S2, S3, TOC) from the Bulldog Lake (BUL) sediment core (64.03785° N, 111.18337° W).

Sample Depth As S Fe Mn S1 S2 S3 TOC mg mg mg Units cm mg·kg-1 wt. % wt. % mg·kg-1 wt. % HC/gA HC/gA HC/gA Detection Limit -- 0.1 0.02 0.01 1.0 0.01 0.01 0.01 -- BUL 0-1 0.5 590 0.46 4.09 993 12.3 35.9 23.8 13.3 BUL 1-2 1.5 351 0.83 2.94 609 10.8 29.3 16.1 10.2 BUL 2-3 2.5 1,010 1.02 2.86 462 11.1 30.6 17.1 11.0 BUL 3-4 3.5 805 0.72 2.47 449 12.5 33.8 17.2 11.6 BUL 4-5 4.5 566 0.58 2.15 442 12.0 35.1 17.0 11.8 BUL 5-6 5.5 305 0.44 1.92 412 13.2 35.7 16.7 11.8 BUL 6-7 6.5 198 0.35 1.80 406 12.5 36.4 16.3 11.6 BUL 7-8 7.5 114 0.30 1.51 381 12.6 35.9 15.7 11.4 BUL 8-9 8.5 103 0.27 1.47 343 11.2 32.4 15.0 10.5 BUL 9-10 9.5 90.4 0.27 1.54 346 10.8 32.8 14.8 10.5 BUL 10-11 10.5 87.9 0.27 1.58 363 10.6 34.7 15.5 11.1 BUL 11-12 11.5 83.5 0.28 1.63 357 11.7 34.7 16.0 11.4 BUL 12-13 12.5 74.5 0.26 1.70 317 10.5 31.2 14.7 10.3 BUL 13-14 13.5 71.2 0.23 1.80 306 9.80 28.1 13.7 9.34 BUL 14-15 14.5 62.7 0.23 1.71 279 9.42 27.4 13.2 9.06 BUL 15-16 15.5 55.8 0.21 1.69 286 8.54 25.9 11.9 8.48 BUL 16-17 16.5 57.6 0.22 1.69 300 9.10 26.6 12.5 8.70 BUL 17-18 17.5 49.7 0.21 1.70 254 8.85 25.4 12.2 8.36 BUL 18-19 18.5 46.5 0.19 1.69 253 7.39 21.7 10.4 7.17 BUL 19-20 19.5 41.4 0.14 1.67 247 7.82 23.1 10.5 7.36 BUL 20-21 20.5 34.2 0.15 1.57 207 7.67 23.4 10.6 7.44 BUL 21-22 21.5 34.4 0.15 1.55 216 7.78 23.3 10.5 7.36 BUL 22-23 22.5 34.3 0.18 1.53 212 8.41 27.0 11.8 8.43 BUL 23-24 23.5 36.2 0.19 1.5 221 8.60 26.9 11.6 8.36 BUL 24-25 24.5 33.1 0.19 1.42 212 9.39 28.5 12.6 8.97 BUL 25-26 25.5 33.0 0.19 1.35 196 10.1 29.4 12.8 9.19 BUL 26-27 26.5 36.4 0.22 1.47 218 10.4 29.9 13.2 9.46 BUL 27-28 27.5 34.5 0.22 1.46 219 10.5 30.9 13.5 9.68 BUL 28-29 28.5 33.1 0.21 1.39 209 9.78 30.0 12.9 9.23 BUL 29-30 29.5 33.0 0.20 1.36 201 10.6 30.6 13.3 9.56 BUL 30-31 30.5 36.4 0.20 1.40 200 10.5 30.6 13.1 9.49 BUL 31-32 31.5 34.5 0.23 1.50 197 11.0 31.2 13.6 9.70 BUL 32-33 32.5 33.1 0.23 1.45 216 10.6 30.7 13.3 9.53 BUL 33-34 33.5 35.6 0.23 1.38 213 11.6 34.1 14.7 10.6 BUL 34-35 34.5 34.2 0.20 1.37 195 9.68 28.6 12.4 8.87 242

BUL 35-36 35.5 36.8 0.19 1.41 188 8.99 26.7 11.5 8.26 BUL 36-37 36.5 37.5 0.19 1.31 175 9.09 27.3 11.5 8.35 A mg HC/g – milligrams of hydrocarbon per gram of sediment.

Table C-3 Concentrations of As, S, Fe, Mn and organic matter fractions (S1, S2, S3, TOC) from Hambone Lake sediment grab samples.

Detection Sample Units HAM1 HAM2 HAM3 HAM4 HAM5 HAM6 Limit Distance from m -- 540 515 310 230 125 250 discharge* Sediment sample depth cm -- 0-15 0-15 0-15 0-15 0-15 0-15 As mg·kg-1 0.01 622 576 304 195 303 80 S wt. % 0.02 1.2 1.8 0.99 0.97 1.02 0.21 Fe wt. % 0.01 1.59 1.80 1.32 1.34 1.20 0.84 S1 mg HC/gA 0.01 16.7 19.6 11.9 11.4 13.6 5.48 S2 mg HC/gA 0.01 50.8 52.7 45.4 53.4 59.7 17.3 S3 mg HC/gA 0.01 21.4 25.8 16.2 19.0 19.9 6.49 TOC wt. % -- 14.5 16.0 11.8 14.0 15.1 4.73 * Location shown on Figure 4.1; A mg HC/g – milligrams of hydrocarbon per gram of sediment.

243

Table C-4 Arsenic speciation (HG-AFS) and concentrations of dissolved As, S, Fe, Mn in porewaters extracted from the Powder Mag Lake (POW) sediment core (64.05114° N, 111.15042° W).

As A A B C C D Sample Depth S Fe Mn Astot As (V) AsR (III)C Units cm mg·L-1 mg·L-1 mg·kg-1 µg·L-1 µg·L-1 µg·L-1 µg·L-1 Detection Limit -- 0.05 0.005 0.1 0.09 0.09 0.09 -- POW 0-1 0.5 71.3 2.18 372 383 80.0 284 18.6 POW 1-2 1.5 66.1 0.90 460 304 53.5 189 60.9 POW 2-3 2.5 72.6 0.30 463 58.9 7.30 41.1 10.4 POW 3-4 3.5 73.2 1.04 507 28.8 7.50 13.4 8.00 POW 4-5 4.5 77.5 0.38 381 89.0 30.3 49.7 9.00 POW 5-6 5.5 83.2 0.09 345 57.3 14.4 37.9 5.00 POW 6-7 6.5 88.7 0.03 335 16.0 2.60 10.0 3.50 POW 7-8 7.5 94.0 1.21 413 10.4 2.10 6.20 2.10 POW 8-9 8.5 99.4 0.04 267 8.70 2.50 4.40 1.90 POW 9-10 9.5 106 0.02 323 4.20 1.00 1.70 1.50 POW 10-11 10.5 111 1.65 478 2.30 0.40 0.70 1.20 POW 11-12 11.5 112 3.76 544 2.40 0.40 1.00 0.90 POW 12-13 12.5 114 0.06 524 1.80 0.40 0.40 0.90 POW 13-14 13.5 117 1.42 518 1.80 0.30 0.30 1.20 POW 14-15 14.5 118 0.87 558 1.30 0.20 0.20 0.90 POW 15-16 15.5 119 0.86 525 1.30 0.20 0.20 0.90 POW 16-17 16.5 116 20.3 819 1.50 0.30 0.40 0.80 POW 17-18 17.5 117 1.63 585 0.80 0.30 0.10 0.40 POW 18-19 18.5 116 8.16 723 0.50 0.20 0.20 0.10 POW 19-20 19.5 116 1.87 639 2.40 0.90 0.30 1.20 POW 20-21 20.5 113 0.17 566 1.80 1.70 0.10 0.00 POW 21-22 21.5 109 30.8 915 0.60 0.30 0.10 0.20 POW 22-23 22.5 102 6.27 818 3.40 1.60 0.30 1.50 POW 23-24 23.5 99.4 13.8 825 1.60 0.40 0.20 0.90 POW 24-25 24.5 94.0 2.04 655 5.20 3.10 0.30 1.70 POW 25-26 25.5 92.1 24.8 843 2.20 0.90 0.40 0.90 POW 26-27 26.5 87.8 18.4 810 1.40 0.70 0.20 0.50 POW 27-28 27.5 83.6 32.9 823 2.70 0.90 0.30 1.50 POW 28-29 28.5 78.3 26.7 767 4.70 0.80 3.00 0.80 POW 29-30 29.5 71.9 5.09 629 3.20 2.10 0.30 0.90 A ICP-OES; B ICP-MS; C HG-AFS; D Calculated from HG-AFS results

244

Table C-5 Arsenic speciation (HG-AFS) and concentrations of dissolved As, S, Fe, Mn in porewaters extracted from the Bulldog Lake (BUL) sediment core (64.03785° N, 111.18337° W).

A A B C C C D Sample Depth S Fe Mn Astot As (III) As (V) AsR Units cm mg·L-1 mg·L-1 mg·kg-1 µg·L-1 µg·L-1 µg·L-1 µg·L-1 Detection Limit -- 0.05 0.005 0.1 0.09 0.09 0.09 -- BUL 0-1 0.5 0.22 5.02 2756 74.3 32.84 27.9 13.5 BUL 1-2 1.5 1.77 0.22 1646 67.7 11.35 30.7 25.6 BUL 2-3 2.5 2.87 1.82 1473 82.7 12.42 67.7 2.61 BUL 3-4 3.5 5.52 0.02 1309 21.4 3.77 9.92 7.73 BUL 4-5 4.5 10.8 0.02 932 33.2 2.46 20.0 10.8 BUL 5-6 5.5 10.0 0.11 797 22.2 2.06 17.1 3.04 BUL 6-7 6.5 6.54 0.18 34.8 37.6 3.58 28.8 5.25 BUL 7-8 7.5 2.32 0.12 610 30.5 2.68 24.6 3.18 BUL 8-9 8.5 4.34 0.16 28.0 24.3 2.06 20.5 1.72 BUL 9-10 9.5 1.75 0.13 213 20.5 1.79 13.2 5.55 BUL 10-11 10.5 5.19 0.18 78.8 20.4 1.99 17.6 0.75 BUL 11-12 11.5 1.96 0.14 223 16.8 2.24 14.2 0.30 BUL 12-13 12.5 2.50 0.13 66.7 16.8 1.30 11.6 3.84 BUL 13-14 13.5 2.40 0.15 25.5 17.2 1.50 14.4 1.27 BUL 14-15 14.5 1.45 0.08 378 9.42 1.08 6.25 2.09 BUL 15-16 15.5 3.37 0.13 53.7 12.1 1.23 10.1 0.80 BUL 16-17 16.5 2.01 0.04 65.2 11.8 1.93 8.71 1.18 BUL 17-18 17.5 3.86 0.02 24.5 12.2 1.12 8.33 2.72 BUL 18-19 18.5 2.03 0.02 66.7 10.4 1.15 6.45 2.75 BUL 19-20 19.5 1.26 0.10 18.1 11.6 3.73 6.85 1.06 BUL 20-21 20.5 0.40 0.01 198 12.2 4.92 5.24 2.07 BUL 21-22 21.5 0.59 0.07 27.5 13.2 5.18 5.67 2.34 BUL 22-23 22.5 0.45 0.03 46.9 16.1 11.14 2.27 2.69 BUL 23-24 23.5 1.52 0.02 43.8 12.8 2.62 7.68 2.53 BUL 24-25 24.5 1.03 0.02 74.4 12.0 3.21 4.08 4.73 BUL 25-26 25.5 0.55 0.03 175 17.3 10.80 2.25 4.27 BUL 26-27 26.5 0.80 0.01 139 16.9 3.09 11.3 2.52 BUL 27-28 27.5 0.80 0.01 67.4 13.2 5.55 3.71 3.89 BUL 28-29 28.5 0.63 0.01 187 15.4 8.77 2.21 4.40 BUL 29-30 29.5 0.22 ------BUL 31-32 31.5 1.77 0.03 50.4 16.0 3.28 8.95 3.73 BUL 32-33 32.5 2.87 0.09 36.4 15.2 6.08 7.53 1.55 BUL 33-34 33.5 5.52 0.04 17.4 19.2 11.76 6.33 1.06 BUL 34-35 34.5 10.6 0.01 265 13.2 13.07 <0.09 -- BUL 35-36 35.5 10.0 0.02 63.0 19.9 10.46 3.94 5.45 BUL 36-37 36.5 6.54 < 0.005 230 14.4 11.63 0.32 2.43 A ICP-OES; B ICP-MS; C HG-AFS; D Calculated from HG-AFS results

245

Table C-6 Solid-phase arsenic speciation in lake sediments based on linear combination fitting of XANES spectra.

Depth As(-I)-S As(III)-Sa As(III)-Oa As(V)-Oa Fitted Sumb R valuec Reduced χ2 Reference Arsenopyrite Orpiment Hematite-As(III) HFO-As(V) ------cm % % % % % -- -- Powder Mag Lake 1 - 2 -- 28.0 25.0 47.0 101.6 0.00281 0.00092 3 - 4 16.0 41.0 24.0 19.0 101.1 0.00234 0.00068 6 - 7 -- 37.0 29.0 34.0 100.9 0.00372 0.00115 Bulldog Lake 0 - 1 -- 5.00 40.0 55.0 100.6 0.00119 0.00043 1 - 2 22.0 55.0 14.0 9.00 101.0 0.00346 0.00102 2 - 3 14.0 60.0 17.0 10.0 101.0 0.00545 0.00170 5 - 6 -- 44.0 26.0 30.0 101.3 0.00347 0.00107 Hambone Lake HAM1 56.0 10.0 26.0 8.00 100.0 0.01035 0.00236 HAM2 63.0 -- 28.0 9.00 98.2 0.00936 0.00193 HAM5 28.0 25.0 21.0 25.0 100.5 0.00681 0.00170 a Component sums were normalized to 100%; b Fitted sum of all references before normalization to a sum of 100%; c Mean square misfit between the data and the fit.

Table C-7 Model compounds for post-hoc Linear Combination Fitting XANES analysis.

Formula As Species Source Edge As Position wt. %* Arsenopyrite FeAsS As(-I)-S Deloro, ON, Canada 11,867.0 46.01

Realgar As4S4 As(II)-S Hunan, China 11,867.5 70.03

Orpiment As2S3 As(III)-S Moldawa, Hungary 11,867.7 60.90

A Hematite-As(III) 0.026As·Fe2O3 Fe2O3-As(III) A 11869.7 1.21*

246

Figure C-1 Additional µXRF maps with corresponding µXRD spectra from Bulldog (a, b, c) and Hambone (d, e) lakes; (a) Qtz – quartz, Lep – lepidocrocite, Mgh – maghemite; (b) Gt – goethite, Mkw – mackinawite; (c) Tnt – titanite, Fh – ferrihydrite, Gt – goethite; (d) Gt – goethite; (e) Mkw – mackiniwite, Mgh – maghemite.

247

Appendix D Geochemical characterization of Tundra Mine tailings

The Excel files listed provide more comprehensive data and associated QA/QC:

· Miller_BulkGeochemistry_Tailings

· Miller_EPMA_AppendixD

· Miller_XANES_AppendixD

248

D.1 Introduction

During mining, tailings were disposed of and stored subaqueously in a tailings confinement area

(TCA) on the Tundra Mine site. However, following mine closure (1987) water levels were not closely monitored leading to seasonal overtopping of TCA waters and seepage through/under the tailings dams

(INAC, 2014). In July 2016, we collected two cores of subaerially exposed tailings from the TCA, prior to remediation, to characterize one of the primary sources of downstream As contamination (Figure D-1).

The main objective of this work was to carry out mineralogical analysis of secondary precipitates in the tailings to supplement data from previous geochemical studies conducted to examine the mobility and risk of As in the TCA (Lorax Environmental, 2007). In the final phases of remediation (2017 to 2018), tailings waters were treated and subsequently disposed into Hambone Lake and tailings were placed in a lined pit, capped and covered (AECOM, 2018).

Figure D-1 Location of Tundra Mine and tailings core samples.

249

D.2 Methods

Two samples of tailings were collected using aluminum push cores from the TCA beach (Figure

D-1). Prior to collection, a pit was dug in the tailings and horizons with distinct colours and textures were described (Figure D-2). Cores were sealed on both ends and frozen prior to shipment to Queen’s

University for subsampling and analysis. Tailings cores were defrosted in a nitrogen (N2)-filled glove bag then cut lengthwise using a table saw and carefully separated using a ceramic blade. Cores were subsampled in the N2-filled glove bag based on visual observation of distinct colour changes and dried

3 (Figure D-2). Tailings subsamples were placed in N2-flushed 50 cm polypropylene centrifuge tubes for storage prior to analysis.

Tailings sub-samples (n = 35, not including duplicates or reference material) were submitted to

Acme Analytical Laboratories (Bureau Veritas), Vancouver, for geochemical analyses. A modified aqua regia digestion protocol (1:1:1 HCl:HNO3:H2O at 95 °C for one hour) was used prior to elemental analysis via inductively coupled plasma-mass spectrometry (ICP-MS 1F/AQ250 package (53 elements)).

Tailings subsamples were selected for detailed solid-phase speciation analysis based on bulk elemental composition and position within each core. Polished sections (35 to 50 μm thickness) were prepared at Vancouver Petrographics for SEM-based automated mineralogy (n = 4; operating conditions detailed in Chapter 2), electron microprobe (EPMA) analysis (n = 4; operating conditions described in

Chapter 3), and bulk XANES (n = 5) analysis at the Advanced Photon Source (APS) at Argonne National

Laboratory (Lemont, IL; sample preparation and operating conditions detailed in Chapter 3).

Qualitative analysis of organic matter in tailings was completed through petrographic analysis following methods outlined in Chapter 3 and by Reyes et al. (2006). Incident white light and fluorescent light microscopy were conducted using a Zeiss Axioimager II microscope system (50× magnification) equipped with the Diskus-Fossil system. Fluorescence microscopy was conducted using ultraviolet G 365 nm excitation with a 420 nm filter.

250

D.3 Results and Discussion

D.3.1 Field Observations

Depth of tailings was variable between sample locations. Tailings core TC-02 reached bedrock refusal at 17 cm while tailings extended well below the 30-cm length of tailings core TC-01. Tailings at both sampling locations were dry at the surface and became progressively wetter with depth, reaching saturation at approximately 15 cm. Alternating banding of orange and grey tailings (1 to 2 cm in thickness) is apparent from 0 to 10 cm in both cores collected (Figure D-2).

Figure D.2 Subsampling of Tundra Mine tailings prior to remediation; A) Overview of tailings confinement area B) Alternating zones of oxidation in upper tailings (T1 to T5 designate field observations of distinct tailings horizons); C) Tailings core ready for subsampling.

D.3.2 Elemental concentration and distribution of As-bearing phases

In the Tundra Mine tailings, solid phase As concentrations range from 62 to 10,000 mg·kg- 1 (n =

35; median = 1,600 mg·kg-1). Scanning electron microscope-based automated mineralogy analyses show that secondary As-bearing phases account for up to 45 % of total As suggesting that As was mobile

251

following deposition in the tailings. The post-depositional remobilization of As and the seasonal stability of secondary phases may have played a in the mobility and seepage of As to the downstream environment

(Figure D-3). Arsenopyrite accounts for up to 95 % of total As; however, the presence of mixed As oxidation states (As (-I), As (III) and As (V)) at all depths suggests that in-situ oxidation of this phase likely accounts for a large percentage of these authigenic secondary phases (Figure D-4). Surface enrichment of As, Fe, and Mn was observed in the visibly oxidized upper tailings corresponding with minor increases in both authigenic Fe-(oxy)hydroxides and the presence of mixed As oxidation states

(Figure D-3, D-4). Minor increases in S are also apparent just below the surface of the tailings and correspond to increased abundance of As-bearing framboidal pyrite. The increased concentrations of As and other trace metal(oid)s (i.e. Ag, Au, Hg, Zn, Cr, Cu, Sb, Pb and W) in the near-surface tailings suggests that accumulation of authigenic Fe and Mn oxides and framboidal pyrite contributed to the scavenging of trace metals at the tailings-water interface (Table D-1, D-2). At the depth of the water table

(~ 13 to 15 cm) bulk As concentrations increase slightly due to the accumulation of As-sulphides suggesting the onset of reducing conditions in the saturated tails (Figure D-2, D-3; Table D-3). The susceptibility of these authigenic phases to changes in redox conditions may have increased the loading of metals to tailings porewater and surface waters seasonally or with variations in water cover levels.

Secondary Ca-Fe arsenate minerals account for 0.50 to 15 % of total As (Figure D-3; Table D-3).

The presence of relict carbonate minerals in the gangue (calcite: 2.9 to 4.5%) and utilization of lime for cyanide leaching promoted the formation of these authigenic minerals in the TCA (Brien, 1964; Giant

Yellowknife Mines Ltd., 1980; Lorax Environmental, 2007). Their size, morphology, and chemical composition is highly variable; however, molar ratios of Ca/As suggests a composition trending toward yukonite (Paktunc et al., 2004; Walker et al., 2009). The variable composition of this phase and absence of porewater chemistry makes it difficult to predict its solubility; however, its presence as both discrete grains and as rims on arsenopyrite suggests that weathering of arsenopyrite in the tailings is liberating As to solution and promoting the formation of this secondary phase.

252

Figure D-3 Arsenic concentration profiles in tailings core TC-01. Pie charts show the relative contribution of each As-hosting solid phase, by mass, to total As concentrations in selected sediment intervals. Filled circles denote samples for which SEM-AM was completed.

Seepage and overtopping of water from the TCA were two primary pathways of downstream contamination from the former Tundra Mine site (INAC, 2005; URS, 2005; SENES Consultants, 2006;

Lorax Environmental, 2007). Dissolved As or suspended As-bearing particles in tailings surface water or porewater both contributed to this contamination. The mineral form and stability of As-bearing particles in the tailings therefore influenced the degree of contamination derived from the tailings. The abundance of secondary phases suggests that As was mobile in the tailings pore waters; however, the surface

253

enrichment of As suggests that sequestration of As on authigenic minerals may have mediated the flux to the overlying waters. In general, SEM-based automated mineralogy accounts for a larger percentage of arsenopyrite than LCF best-fit XANES results. The surface oxidation of arsenopyrite is observed through

SEM analysis; however, the formation of intermediate As oxidation products at the surface of the grain may not be accurately accounted for through SEM-based automated mineralogy and account for the discrepancy between results. In addition, the small size and microcrystalline/amorphous nature of other authigenic As-bearing phases may not be accurately accounted for through SEM-based automated mineralogy contribute to the discrepancy.

Figure D-4 Selected XANES spectra for four subsamples from tailings core TC01. The XANES spectrum of the bulk sample suggests the presence of As in more than one oxidation state.

254

D.3.3 Tailings organic matter characterization

Fluorescent light microscopy of tailings samples reveals that trace OM is present in the near- surface tailings. Classification and identification of OM, based on optical properties (fluorescence and reflectance) and morphology, indicate that OM is derived from both aquatic (i.e. diatoms, alginate) and terrigenous sources (i.e. sporinite) (Figure D-5; Sanei et al., 2005; Reyes et al., 2006). The identification of trace OM in the tailings cores is supported by TOC concentrations ranging from 3 to 29 mg/L (2003 to

2006) in TCA surface waters (Lorax, 2007). The presence of labile OM in the near-surface tailings also supports the observation of abundant framboidal pyrite and suggests that increased organic OM transport to the TCA interfacial sediments has influenced redox conditions and the cycling of metal(loid)s in the tailings. These findings suggest that increases of both autochthonous and allochthonous sources of OM, as a result of climate warming, not only influence metal(loid) cycling in lakes but may also influence the long-term stability of tailings at current and abandoned mine sites in the sub-Arctic.

Figure D-5 Fluorescent-light photomicrographs under oil immersion and blue light excitation of organic matter preserved in the near-surface tailings of the TCA; A) Organo-mineral complex with aliginite; B) Sporinite.

255

Table D-1 Solid phase concentrations of As, Fe, Mn, S, Ca and S from tailings cores extracted from the tailings containment area (TC-01).

Depth (mid) As Fe Mn S Ca Sb Sr cm mg·kg-1 wt. % mg·kg-1 wt. % wt. % mg·kg-1 mg·kg-1 0.75 1,630 1.42 274 0.12 0.29 0.53 9.30 2.0 1,630 1.50 253 0.15 0.29 0.80 9.20 3.75 4,390 2.50 478 0.26 0.52 2.16 14.2 4.75 6,350 3.16 591 0.20 0.56 2.6 18.5 6.25 6,540 3.43 602 0.50 0.75 3.21 16.3 8.0 7,320 3.59 457 0.41 0.36 2.04 15.1 8.5 5,070 3.21 348 0.54 0.23 1.62 12.3 10.5 2,260 1.31 107 0.47 0.33 0.51 13.8 11.5 1,440 1.13 93.0 0.4 0.31 0.25 12.6 12.5 1,850 1.38 149 0.42 0.36 0.63 14.9 13.5 1,470 1.38 161 0.47 0.38 0.56 16.0 14.5 677 1.25 137 0.58 0.42 0.33 18.6 15.5 530 1.26 127 0.74 0.46 0.26 21.2 16.5 131 1.19 112 0.99 0.53 0.09 28.3 17.5 63.0 1.20 119 1.10 0.55 0.05 29.4 18.5 323 1.23 125 0.90 0.51 0.14 25.0 19.5 175 1.25 123 1.09 0.54 0.10 28.3 20.5 646 1.48 144 1.18 0.52 0.31 28.6 21.5 353 1.35 144 1.16 0.51 0.22 29.7 22.5 69.5 1.29 114 1.23 0.51 0.06 31.9

Table D-2 Solid phase concentrations of As, Fe, Mn, S, Ca and S from tailings cores extracted from the tailings containment area (TC-02).

Depth As Fe Mn S Ca Sb Sr cm mg·kg-1 wt. % mg·kg-1 wt. % wt. % mg·kg-1 mg·kg-1 0.5 4,410 2.41 744 0.55 0.79 1.47 21.1 1.5 7,000 3.17 539 0.40 0.51 1.62 17.3 2.5 4,560 2.8 466 0.40 0.44 1.83 13.8 3.5 8,340 3.61 628 0.40 0.63 2.63 19.5 5.25 6,320 3.55 624 0.29 0.59 4.04 18.1 6.25 7,640 3.62 577 0.56 0.53 2.92 14.8 7.5 > 10,000 4.11 439 0.31 0.38 1.81 18.6 9 4,840 3.25 372 0.52 0.23 1.70 12.1 10.5 3,550 1.67 172 0.44 0.37 0.96 16.3 11.5 1,920 1.17 126 0.46 0.44 0.46 18.2 12.5 1,430 1.13 136 0.49 0.47 0.25 19.1

256

13.5 773 1.06 143 0.55 0.49 0.14 21.5 14.5 648 1.07 140 0.58 0.53 0.12 21.1 15.5 311 1.03 134 0.58 0.53 0.08 25.3 16.5 345 1.05 137 0.61 0.55 0.09 27.5

Table D-3 Results of automated mineralogy and mass distribution calculations for the TC-01 core.

As-Hosting Solid Phase Number of Particles Total Area of Particles Density As Mass Distribution Units -- µm2 g·cm-3 wt.% % TC-01 (0 – 1.5 cm) [ID TC-01_1] Arsenopyrite 989 13754 6.07 46 91.4 As-Sulphide 11 146 3.56 70 0.86 Framboidal pyrite 479 8924 5.01 0.5* 0.25 Fe-Oxides 769 14864 4.28 3.0* 4.15 Ca-Fe-Arsenate 205 2090 2.74 25 3.35 TC-01 (8.5 - 10 cm) [ID TC-01_10] Arsenopyrite 17643 217161 6.07 46 96.6 As-Sulphide 55 1047 3.56 70 0.60 Framboidal pyrite 5078 138220 5.01 0.5* 0.90 Fe-Oxides 2200 43732 4.28 3.0* 1.40 Ca-Fe-Arsenate 127 2733 2.74 25 0.50 TC-01 (12 - 13 cm) [ID TC-01_13] Arsenopyrite 17643 217161 6.07 46 96.6 As-Sulphide 55 1047 3.56 70 0.60 Framboidal pyrite 5078 138220 5.01 0.5* 0.90 Fe-Oxides 2200 43732 4.28 3.0* 1.40 Ca-Fe-Arsenate 127 2733 2.74 25 0.50 TC-01 (20 – 21 cm) [ID TC-01_13] Arsenopyrite 5290 55441 6.07 46 59.7 As-Sulphide 623 16569 3.56 70 23.6 Framboidal pyrite 2340 65583 5.01 0.5* 1.0 Fe-Oxides 2247 76430 4.28 3.0* 5.8 Ca-Fe-Arsenate 685 21961 2.74 25 9.9 * Determined using electron microprobe analysis (EPMA)

257

Table D-4 Results of automated mineralogy and mass distribution calculations for TC-02 core.

As-Hosting Solid Phase Number of Particles Total Area of Particles Density As Mass Distribution Units -- µm2 g·cm-3 wt.% % TC-02 (7.5 - 8.5 cm) [ID TC-02_8] Arsenopyrite 3508 36282 6.07 46 77.8 As-Sulphide 0 0 3.56 70 0 Framboidal pyrite 1312 29689 5.01 0.5* 0.90 Fe-Oxides 3012 18191 4.28 3.0* 2.80 Ca-Fe-Arsenate 1294 20528 2.74 25 18.5 * Determined using electron microprobe analysis (EPMA)

258

D.4 References

AECOM, 2018. Tundra Mine Adaptive Management Plan. Available online: http://registry.mvlwb.ca/Documents/MV2016L8-0003/MV2016L8- 0003%20%E2%80%93%20INAC%20CARD%20%E2%80%93%20Adaptive%20Management%20 Plan%20%E2%80%93%20Revision%20from%20May%2024-18%20%20-%20Oct26-18.pdf

Brien, F.B., 1964. Tundra Gold Mines - Milling Operations. West. Min. 34–38. Obtained from NTGO Library.

Giant Yellowknife Mines Ltd., 1980. Salmita Project - A Brief Project Overview. Obtained from NTGO Library.

4INAC, 2005. Tundra Mine, NWT Environmental Monitoring Program: Water Quality Part D: 2005 Results. Prepared for: Contaminants and Remediation Directorate, Indian and Northern Affairs Canada. Available online: http://registry.mvlwb.ca/Documents/MV2005X0031/PartD- WaterQualityMon-Nov05.pdf

INAC, 2014. Tundra Mine, NWT Remediation Environmental Monitoring Program: Water Quality Part L: 2013 and Historic Results

Lorax Environmental, 2007. Tundra Mine Geochemical Assessment. 105 p. Internal report.

Paktunc, D.P., Foster, A.F., Heald, S.H., Laflamme, G.L., 2004. Speciation and characterization of arsenic in gold ores and cyanidation tailings using X-ray absorption spectroscopy. Geochim. Cosmochim. Acta 68, 969–983. https://doi.org/10.1016/j.gca.2003.07.013

Sanei, H., Stasiuk, L.D., Goodarzi, F., 2005. Petrological changes occurring in organic matter from recent lacustrine sediments during thermal alteration by Rock-Eval pyrolysis. Org. Geochem. 36, 1190– 1203. https://doi.org/10.1016/j.orggeochem.2005.02.009

SENES Consultants Ltd., 2006. Tundra Mine Site Ecological Risk Assessment of Water Management. Available online: http://www.mvlwb.ca/Registry.aspx?a=MV2009L8-0008

URS, 2005. Geochemical Assessment of Acid Rock Drainage and Metal Leaching Potential of Tailings and Waste Rock, Tundra Mine, NWT. Internal Report.

Walker, S.R., Parsons, M.B., Jamieson, H.E., Lanzirotti, A., 2009. Arsenic mineralogy of near-surface tailings and soils: Influences on arsenic mobility and bioaccessibility in the nova scotia gold mining districts. Can. Mineral. 47, 533–556. https://doi.org/10.3749/canmin.47.3.533

4 INAC, AANDC, CIRNAC are all acronyms for the Federal Department responsible for the Government of Canada's work in northern Canada. 259

Appendix E Field and Laboratory Photos

Figure E-1 Photos from field work at Tundra Mine in March 2016; A) Driving on the Tibbitt to Contwoyto winter road; B) Transporting field gear and nitrogen canister to sampling sites; C) Augering holes in the ice for sampling; D) Measuring depth of the ice and surface water; E-G) Recovering and assessing a sediment gravity core sample.

260

Figure E-2 Photos from field work at Tundra Mine in March 2016; H) Sampling set up for water sampling, freeze coring, sediment gravity coring, and extrusion of sediment cores on lake surface; I) Measuring depth of ice and surface water; J-K) Set-up for extrusion of sediment gravity cores in N2-flled glove bag; L-M) Collecting surface water parameters and samples; N) Field laboratory set-up for filtering and preserving surface water samples.

261

Figure E-3 Sediment gravity core photos from Tundra Mine field work, March 2016; O) Powder Mag Lake; P) Bulldog Lake; Q) a) Matthews Lake sediment core, b) with SWI preserved; R) a) Control Lake sediment core, b) with SWI preserved.

262

Figure E-4 S-T) Collection of Ekman grab samples from Hambone Lake, July 2016; U) Sediment samples in centrifuge to separate porewaters; V) Filtering and preservation of porewater in N2-filled glove bag; W) Sediment samples prior to (a) and following (b) centrifuging; X) Drying sediment samples in N2-filled glove bag.

263