EPA910-R-14-001A | January 2014
EPA
United States
Environmental Protection
Agency
An Assessment of Potential Mining Impacts
on Salmon Ecosystems of Bristol Bay, Alaska
Volume 1 - Main Report
Region 10, Seattle, WA
www.epa.gov/bristolbay
-------
EPA910-R-14-001A
January 2014
AN ASSESSMENT OF POTENTIAL MINING IMPACTS ON
SALMON ECOSYSTEMS OF BRISTOL BAY, ALASKA
VOLUME 1—MAIN REPORT
U.S. Environmental Protection Agency
Region 10
Seattle, WA
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
Preferred citation: USEPA (U.S. Environmental Protection Agency). 2014. An Assessment of Potential
Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska. Region 10, Seattle, WA. EPA 910-R-14-001.
-------
List of Appendices viii
List of Tables ix
List of Figures xiv
List of Boxes xviii
Acronyms and Abbreviations xx
Preface xv
Authors, Contributors, and Reviewers xxvi
Photo Credits xxix
Acknowledgements xxxi
Executive Summary ES-1
Chapter 1. Introduction 1-1
1.1 Assessment Approach 1-2
1.2 Uses of the Assessment 1-6
Chapter 2. Overview of Assessment 2-1
2.1 Structure 2-1
2.1.1 Data Used in the Assessment 2-2
2.1.2 Types of Evidence and Inference 2-3
2.2 Scope 2-5
2.2.1 Topical Scope 2-5
2.2.2 Geographic Scales 2-8
Chapters. Region 3-1
3.1 Physiographic Divisions 3-1
3.2 Hydrologic Landscapes 3-13
3.3 Groundwater Exchange and Flow Stability 3-13
3.4 Quantity and Diversity of Aquatic Habitats 3-16
3.4.1 Stream Reach Characterization: Attributes 3-18
3.4.2 Stream Reach Characterization: Results 3-25
3.5 Water Quality 3-27
3.5.1 Water Chemistry 3-27
3.5.2 Water Temperature 3-30
3.6 Seismicity 3-30
3.7 Existing Development 3-33
3.8 Climate Change 3-34
3.8.1 Climate Change Projections for the Bristol Bay Region 3-36
3.8.2 Potential Climate Change Effects 3-41
Chapter 4. Type of Development 4-1
4.1 Mineral Deposits and Mining in the Bristol Bay Watershed 4-1
4.2 Porphyry Copper Deposits and Mining Processes 4-3
4.2.1 Genesis of Porphyry Copper Deposits 4-3
4.2.2 Chemistry and Associated Risks of Porphyry Copper Deposits 4-3
4.2.3 Overview of the Mining Process 4-6
Bristol Bay Assessment ••• January 2014
-------
4.2.4 Timeframes 4-20
Chapter 5. Endpoints 5-1
5.1 Overview of Assessment Endpoints 5-1
5.2 Endpoint 1: Salmon and Other Fishes 5-3
5.2.1 Species and Life Histories 5-9
5.2.2 Distribution and Abundance 5-13
5.2.3 Economic Implications 5-26
5.2.4 Biological Complexity and the Portfolio Effect 5-27
5.2.5 Salmon and Marine-Derived Nutrients 5-29
5.2.6 Bristol Bay Fisheries in the Global Context 5-30
5.3 Endpoint 2: Wildlife 5-31
5.3.1 Life Histories, Distributions, and Abundances of Species 5-32
5.3.2 Recreational and Subsistence Activities 5-35
5.4 Endpoint 3: Alaska Natives 5-35
5.4.1 Alaska Native Populations 5-36
5.4.2 Subsistence and Alaska Native Cultures 5-36
Chapter 6. Mine Scenarios 6-1
6.1 Basic Elements of the Mine Scenarios 6-1
6.1.1 Location 6-8
6.1.2 Mining Processes 6-8
6.1.3 Transportation Corridor 6-16
6.2 Specific Mine Scenarios 6-20
6.2.1 Mine Scenario Footprints 6-20
6.2.2 Water Balance 6-23
6.3 Closure and Post-Closure Site Management 6-27
6.3.1 Mine Pit 6-32
6.3.2 Tailings Storage Facilities 6-32
6.3.3 Waste Rock 6-33
6.3.4 Water Management 6-33
6.3.5 Premature Closure 6-35
6.4 Conceptual Models 6-36
6.4.1 Sources Evaluated 6-36
6.4.2 Stressors Evaluated 6-36
6.4.3 Endpoints Evaluated 6-42
6.4.4 Conceptual Model Diagrams 6-42
Chapter 7. Mine Footprint 7-1
7.1 Abundance and Distribution of Fishes in the Mine Scenario Watersheds 7-2
7.1.1 Fish Distribution 7-2
7.1.2 Spawning Salmon Abundance 7-12
7.1.3 Juvenile Salmon and Other Salmonid Abundance 7-15
7.2 Habitat Modification 7-16
7.2.1 Stream Segment Characteristics in the Mine Scenario Watersheds 7-16
7.2.2 Exposure: Habitat Lost to the Mine Scenario Footprints 7-19
7.2.3 Exposure-Response: Implications of Stream and Wetland Loss for Fish 7-28
7.2.4 Risk Characterization 7-33
7.2.5 Uncertainties 7-34
7.3 Streamflow Modification 7-35
7.3.1 Exposure: Streamflow 7-35
7.3.2 Exposure-Response: Streamflow 7-52
7.3.3 Risk Characterization 7-59
7.3.4 Uncertainties and Assumptions 7-60
7.4 Summary of Footprint Effects 7-61
Bristol Bay Assessment • January 2014
-------
Chapter 8. Water Collection, Treatment, and Discharge 8-1
8.1 Water Discharge Sources 8-1
8.1.1 Routine Operations 8-3
8.1.2 Wastewater Treatment Plant Failure 8-15
8.1.3 Spillway Release 8-17
8.1.4 Post-Closure Wastewater Sources 8-17
8.1.5 Probability of Contaminant Releases 8-19
8.2 Chemical Contaminants 8-19
8.2.1 Exposure 8-19
8.2.2 Exposure-Response 8-22
8.2.3 Risk Characterization 8-33
8.2.4 Additional Mitigation of Leachates 8-54
8.2.5 Uncertainties 8-54
8.3 Temperature 8-57
8.3.1 Exposure 8-58
8.3.2 Exposure-Response 8-59
8.3.3 Risk Characterization 8-60
8.3.4 Uncertainties 8-61
Chapter 9. Tailings Dam Failure 9-1
9.1 Tailings Dam Failures 9-2
9.1.1 Causes 9-2
9.1.2 Probabilities 9-7
9.1.3 Uncertainties 9-12
9.2 Material Properties 9-12
9.2.1 Tailings Dam Rockfill 9-12
9.2.2 Tailings Solids and Liquids 9-13
9.3 Modeling a Tailings Dam Failure 9-14
9.3.1 Hydrologic Characteristics 9-16
9.3.2 Sediment Transport and Deposition 9-16
9.3.3 Uncertainties 9-19
9.4 Scour, Sediment Deposition, and Turbidity 9-20
9.4.1 Exposure through Sediment Transport and Deposition 9-22
9.4.2 Exposure-Response 9-23
9.4.3 Risk Characterization 9-24
9.4.4 Uncertainties 9-27
9.5 Post-Tailings Spill Water Quality 9-28
9.5.1 Suspended Tailings Particles 9-28
9.5.2 Tailings Constituents 9-30
9.5.3 Weighing Lines of Evidence 9-43
9.6 Summary of Risks 9-45
9.6.1 Tailings Spill 9-45
9.6.2 Remediation of a Tailings Spill 9-45
Chapter 10. Transportation Corridor 10-1
10.1 Introduction 10-1
10.2 Fish Habitats and Populations along the Transportation Corridor 10-7
10.3 Potential Risks to Fish Habitats and Populations 10-15
10.3.1 Filling and Alteration of Wetlands, Ponds, and Small Lakes 10-20
10.3.2 Stream Crossings 10-21
10.3.3 Chemical Contaminants 10-31
10.3.4 Fine Sediment 10-35
10.3.5 Dust 10-38
10.3.6 Invasive Species 10-40
10.4 Overall Risk Characterization for the Transportation Corridor 10-43
10.5 Uncertainties 10-44
Bristol Bay Assessment January 2014
-------
Chapter 11. Pipeline Failures 11-1
11.1 Causes and Probabilities of Pipeline Failures 11-5
11.2 Potential Receiving Waters 11-7
11.3 Concentrate Pipeline Failure Scenarios 11-8
11.3.1 Sources 11-8
11.3.2 Exposure 11-9
11.3.3 Exposure-Response 11-11
11.3.4 Risk Characterization 11-12
11.3.5 Uncertainties 11-18
11.4 Return Water Pipeline Failure Scenarios 11-20
11.5 Diesel Pipeline Failure Scenarios 11-20
11.5.1 Sources 11-21
11.5.2 Exposure 11-22
11.5.3 Exposure-Response 11-24
11.5.4 Risk Characterization 11-28
11.5.5 Uncertainties 11-31
Chapter 12. Fish-Mediated Effects 12-1
12.1 Effects on Wildlife 12-1
12.2 Effects on Alaska Natives 12-6
12.2.1 Subsistence Use 12-8
12.2.2 Perception of Food Security 12-12
12.2.3 Economic Impacts 12-12
12.2.4 Social, Cultural, and Spiritual Impacts 12-14
12.2.5 Mitigation and Adaptation 12-16
12.3 Uncertainties 12-17
Chapter 13. Cumulative Risks of Multiple Mines 13-1
13.1 Cumulative and Induced Impacts 13-1
13.1.1 Definition 13-1
13.1.2 Vulnerability of Salmonids to Cumulative Impacts 13-2
13.1.3 Nature and Extent of Past, Present, and Future Impacts 13-6
13.2 Cumulative Impacts from Multiple Mines 13-6
13.2.1 Pebble South/PEB 13-8
13.2.2 Big Chunk South 13-10
13.2.3 Big Chunk North 13-22
13.2.4 Groundhog 13-23
13.2.5 AUDN/lliamna 13-24
13.2.6 Humble 13-26
13.2.7 Potential Impacts of Multiple Mines 13-27
13.3 Cumulative Impacts from Induced Development 13-31
13.4 Potential Effects on Assessment Endpoints 13-32
13.4.1 Fishes 13-32
13.4.2 Wildlife and Alaska Native Culture 13-33
13.5 Summary 13-33
Chapter 14. Integrated Risk Characterization 14-1
14.1 Overall Risk to Salmon and Other Fishes 14-1
14.1.1 Routine Operation 14-1
14.1.2 Accidents and Failures 14-4
14.2 Overall Loss of Wetlands, Ponds, and Lakes 14-12
14.3 Overall Fish-Mediated Risk to Wildlife 14-12
14.4 Overall Fish-Mediated Risk to Alaska Native Cultures 14-13
14.5 Summary of Uncertainties and Limitations in the Assessment 14-14
14.6 Summary of Uncertainties in Mine Design and Operation 14-17
14.7 Summary of Risks in the Mine Scenarios 14-17
Bristol Bay Assessment • January 2014
-------
14.8 Summary of Cumulative Risks of Multiple Mines 14-18
Chapter 15. References 15-1
15.1 References by Chapter 15-1
15.1.1 Executive Summary 15-1
15.1.2 Chapter 1—Introduction 15-2
15.1.3 Chapter 2—Overview of Assessment 15-2
15.1.4 Chapter 3—Region 15-3
15.1.5 Chapter 4—Type of Development 15-11
15.1.6 Chapter5—Endpoints 15-14
15.1.7 Chapter 6—Mine Scenarios 15-23
15.1.8 Chapter 7—Mine Footprint 15-26
15.1.9 Chapter 8—Water Collection, Treatment, and Discharge 15-35
15.1.10 Chapter 9—Tailings Dam Failure 15-42
15.1.11 Chapter 10—Transportation Corridor 15-50
15.1.12 Chapter 11—Pipeline Failures 15-59
15.1.13 Chapter 12-Fish-Mediated Effects 15-64
15.1.14 Chapter 13—Cumulative Risks of Multiple Mines 15-66
15.1.15 Chapter 14—Integrated Risk Characterization 15-73
15.2 CIS Base Map Citations 15-74
Bristol Bay Assessment ,, January 2014
-------
List of Appendices
Volume 2: Appendices A-D
Appendix A. Fishery Resources of the Bristol Bay Region
Appendix B. Non-Salmon Freshwater Fishes of the Nushagak and Kvichak River Drainages
Appendix C. Wildlife Resources of the Nushagak and Kvichak River Watersheds, Alaska
Appendix D. Ecological Knowledge and Cultures of the Nushagak and Kvichak Watersheds, Alaska
Volume 3: Appendices E-J
Appendix E. Bristol Bay Wild Salmon Ecosystem: Baseline Levels of Economic Activity and Values
Appendix F. Biological Characterization: Bristol Bay Marine Estuarine Processes, Fish, and Marine Mammal
Assemblages
Appendix G. Foreseeable Environmental Impact of Potential Road and Pipeline Development on Water
Quality and Freshwater Fishery Resources of Bristol Bay, Alaska
Appendix H. Geologic and Environmental Characteristics of Porphyry Copper Deposits with Emphasis on
Potential Future Development in the Bristol Bay Watershed, Alaska
Appendix I. Conventional Water Quality Mitigation Practices for Mine Design, Construction, Operation, and
Closure
Appendix J. Compensatory Mitigation and Large-Scale Hardrock Mining in the Bristol Bay Watershed
Bristol Bay Assessment ,,, January 2014
-------
List of Tables
Table 2-1 Geographic scales considered in the assessment 2-9
Table 3-1 Physiographic divisions (Wahrhaftig 1965) of the Nushagak and Kvichak River
watersheds 3-3
Table 3-2 Distribution of hydrologic landscapes in the Nushagak and Kvichak River watersheds 3-14
Table 3-3 Proportion of stream channel length within the Nushagak and Kvichak River
watersheds classified according to stream size (based on mean annual streamflow in
m3/s), channel gradient (%), and floodplain potential (based on %flatland in lowland) 3-30
Table 3-4 Examples of earthquakes in Alaska 3-31
Table 3-5 Average annual and seasonal air temperature for historical and projected periods
across the Bristol Bay watershed and the Nushagak and Kvichak River watersheds 3-37
Table 3-6 Average annual and seasonal precipitation for historical and projected periods across
the Bristol Bay watershed and the Nushagak and Kvichak River watersheds 3-37
Table 3-7 Average annual water surplus for historical and projected periods across the Bristol Bay
watershed and the Nushagak and Kvichak River watersheds 3-37
Table 4-1 Characteristics of past, existing, or potential large mines in Alaska 4-2
Table 4-2 Global grade and tonnage summary statistics for porphyry copper deposits 4-3
Table 5-1 Fish species reported in the Nushagak and Kvichak River watersheds 5-4
Table 5-2 Life history, habitat characteristics, and total documented stream length occupied for
Bristol Bay's five Pacific salmon species in the Nushagak and Kvichak River
watersheds 5-10
Table 5-3 Mean annual commercial harvest (number of fish) by Pacific salmon species and
Bristol Bay fishing district, 1990 to 2009 5-14
Table 5-4 Summary of regional economic expenditures based on salmon ecosystem services 5-26
Table 5-5 Life-history variation within Bristol Bay sockeye salmon populations 5-28
Table 6-1 Summary of scenarios considered in the assessment 6-2
Table 6-2 Mine scenario parameters 6-10
Table 6-3 Summary of annual water balance flows (million m3/year) during operations for the
three mine scenarios 6-16
Table 6-4 Characteristics of pipelines in the mine scenarios 6-19
Table 6-5 Estimated areas for individual mine components in the Pebble 0.25 scenario 6-21
Table 6-6 Estimated areas for individual mine components in the Pebble 2.0 scenario 6-22
Table 6-7 Estimated areas for individual mine components in the Pebble 6.5 scenario 6-23
Table 6-8 Summary of annual water balance flows (million m3/year) during the post-closure
period for the Pebble 6.5 scenario 6-34
Table 6-9 Stressors considered in the assessment and their relevance to the assessment's
primary endpoint (salmonids) and the U.S. Environmental Protection Agency's
regulatory authority 6-37
Table 6-10 Screening benchmarks for metals with no national ambient water quality criteria 6-39
Table 7-1 Highest reported index spawner counts in the mine scenario watersheds for each year,
2004 to 2008 7-14
Table 7-2 Average 2008 index spawner counts by stream reach 7-15
Table 7-3 Highest index counts of selected stream-rearing fish species from mainstem habitats of
the mine scenario watersheds 7-16
Bristol Bay Assessment
January 2014
-------
Table 7-4 Distribution of stream channel length classified by channel size (based on mean annual
streamflow in m3/s), channel gradient (%), and floodplain potential (based on %flatland
in lowland) for streams and rivers in the mine scenario watersheds 7-17
Table 7-5 Stream length (km) eliminated, blocked, or dewatered by the mine footprints in the
Pebble 0.25, 2.0, and 6.5 scenarios 7-26
Table 7-6 Distribution of stream channel length classified by channel size (based on mean annual
streamflow in m3/s), channel gradient (%), and floodplain potential (based on %flatland
in lowland) for streams lost to the Pebble 6.5 mine footprint 7-27
Table 7-7 Total documented anadromous fish stream length and stream length documented to
contain different salmonid species in the mine scenario watersheds 7-27
Table 7-8 Wetland, pond, and lake areas (km2) eliminated, blocked, or dewatered by the mine
footprints in the Pebble 0.25, 2.0, and 6.5 scenarios 7-29
Table 7-9 Stream gages and related characteristics for the South and North Fork Koktuli Rivers
and Upper Talarik Creek 7-36
Table 7-10 Measured mean monthly pre-miningstreamflow rates (m3/s) and estimated mean
monthly streamflow rates (m3/s) in the Pebble 0.25, 2.0, and 6.5 scenarios, for gages
along the South Fork Koktuli River 7-41
Table 7-11 Measured mean monthly pre-mining streamflow rates (m3/s) and estimated mean
monthly streamflow rates (m3/s) in the Pebble 0.25, 2.0, and 6.5 scenarios, for gages
alongthe North Fork Koktuli River 7-41
Table 7-12 Measured mean monthly pre-mining streamflow rates (m3/s) and estimated mean
monthly streamflow rates (m3/s) in the Pebble 0.25, 2.0, and 6.5 scenarios, for gages
along Upper Talarik Creek 7-42
Table 7-13 Measured minimum monthly pre-mining streamflow rates (m3/s) and estimated
minimum monthly streamflow rates (m3/s) in the Pebble 0.25, 2.0, and 6.5 scenarios,
for gages alongthe South Fork Koktuli River 7-42
Table 7-14 Measured minimum monthly pre-mining streamflow rates (m3/s) and estimated
minimum monthly streamflow rates (m3/s) in the Pebble 0.25, 2.0, and 6.5 scenarios,
for gages alongthe North Fork Koktuli River 7-43
Table 7-15 Measured minimum monthly pre-mining streamflow rates (m3/s) and estimated
minimum monthly streamflow rates (m3/s) in the Pebble 0.25, 2.0, and 6.5 scenarios,
for gages along Upper Talarik Creek 7-43
Table 7-16 Pre-mining watershed areas, mine footprint areas, and flows in the mine scenario
watersheds for the Pebble 0.25 scenario 7-44
Table 7-17 Pre-mining watershed areas, mine footprint areas, and flows in the mine scenario
watersheds for the Pebble 2.0 scenario 7-45
Table 7-18 Pre-mining watershed areas, mine footprint areas, and flows in the mine scenario
watersheds for the Pebble 6.5 scenario 7-46
Table 7-19 Estimated changes in streamflow (%) and subsequent stream lengths affected (km) in
the mine scenario watersheds in the Pebble 0.25, Pebble 2.0, and Pebble 6.5
scenarios 7-50
Table 8-1 Annual effluent and receiving water flows at each gage in the Pebble 0.25 scenario 8-5
Table 8-2 Effluent and receiving water flows at each gage in the Pebble 2.0 scenario 8-6
Table 8-3 Effluent and receiving water flows at each gage in the Pebble 6.5 scenario 8-7
Table 8-4 Aquatic toxicological screening of tailings supernatant against acute (criterion
maximum concentration) and chronic (criterion continuous concentration) water quality
criteria or benchmark values 8-8
Table 8-5 Aquatic toxicological screening of tailings humidity cell leachates against acute
(criterion maximum concentration) and chronic (criterion continuous concentration)
water quality criteria or benchmark values 8-9
Table 8-6 Aquatic toxicological screening of test leachate from Tertiary waste rock in the Pebble
deposit and quotients against acute (criterion maximum concentration) and chronic
(criterion continuous concentration) water quality criteria or benchmark values 8-10
Bristol Bay Assessment
January 2014
-------
Table 8-7 Aquatic toxicological screening of test leachate from Pebble East pre-Tertiary waste
rock and quotients against acute (criterion maximum concentration) and chronic
(criterion continuous concentration) water quality criteria or benchmark values 8-11
Table 8-8 Aquatic toxicological screening of test leachate from Pebble West pre-Tertiary waste
rock against acute (criterion maximum concentration) and chronic (criterion continuous
concentration) water quality criteria or benchmark values 8-12
Table 8-9 Estimated concentration of contaminants of concern in effluents from the wastewater
treatment plant, tailings, non-acid-generating waste rock, and potentially acid
generating waste rock 8-15
Table 8-10 Means and coefficients of variation for background surface water characteristics of the
mine scenario watersheds, 2004-2008 8-21
Table 8-11 Results of applyingthe biotic ligand model to mean water chemistries in the mine
scenario watersheds (Table 8-10) to derive acute (CMC) and chronic (CCC) copper
criteria specific to receiving waters 8-24
Table 8-12 Results of applyingthe biotic ligand model to mean water chemistries in waste rock
leachates (Appendix H) to derive effluent-specific acute (CMC) and chronic (CCC)
copper criteria 8-24
Table 8-13 Site-specific acute and chronic copper toxicity values for rainbow trout, derived by
applyingthe biotic ligand model to mean water chemistries in the mine scenario
watersheds (Table 8-10) 8-25
Table 8-14 Site-specific benchmarks for sensory effects in rainbow trout 8-26
Table 8-15 Hardness-dependent acute water quality criteria (CMC) and chronic water quality
criteria (CCC) for the three potential receiving streams in the mine scenarios 8-29
Table 8-16 Estimated concentrations of contaminants of concern and associated risk quotients for
the Pebble 6.5 scenario, assuming routine operations, at locations in the mine scenario
watersheds 8-37
Table 8-17 Estimated concentrations of contaminants of concern and associated risk quotients for
the Pebble 6.5 scenario, assuming wastewater treatment plant failure, at locations in
the mine scenario watersheds 8-38
Table 8-18 Estimated total toxicity of metals of concern for each mine scenario, under routine
operations and with wastewater treatment plant failure, at locations in the mine
scenario watersheds 8-39
Table 8-19 Background copper concentrations and, for each mine scenario, copper concentrations
in contributing loads and ambient waters (fully mixed reaches below each gage) and
associated risk quotients, assuming routine operations 8-40
Table 8-20 Background copper concentrations and, for each mine scenario, copper concentrations
in contributing loads and ambient waters (fully-mixed reaches below each gage) and
associated risk quotients, assuming wastewater treatment plant failure 8-41
Table 8-21 Description of stream reaches affected in the mine scenarios and sources of the
concentration estimates applied to the stream reaches 8-42
Table 8-22 Copper concentrations and benchmarks exceeded in ambient waters in each reach and
for each mine scenario, assuming routine operations 8-44
Table 8-23 Copper concentrations and benchmarks exceeded in ambient waters in each reach and
for each mine scenario, assuming a wastewater treatment plant failure 8-45
Table 8-24 Results of the spillway release scenario in terms of copper concentrations at North Fork
Koktuli stream gages downstream of TSF 1, estimated effects, and the length of the
associated reaches 8-49
Table 8-25 Length of stream in which copper concentrations would exceed levels sufficient to
cause toxic effects, assuming routine operations, wastewater treatment plant failure,
and spillway release, for each of the three mine scenarios 8-52
Table 9-1 Number and cause of tailings dam failures at active and inactive tailings dams 9-7
Table 9-2 Summary of Alaska's classification of potential dam failure hazards 9-10
Table 9-3 Summary of tailings dam failure probabilities in the three mine scenarios 9-12
Table 9-4 HEC-RAS model results for the Pebble 0.25 and Pebble 2.0 TSF dam failure analyses 9-17
Bristol Bay Assessment
XI
January 2014
-------
Table 9-5 Tailings mobilized and deposited in the Pebble 0.25 and Pebble 2.0 dam failures
analyses 9-18
Table 9-6 Sediment size distributions surveyed at the South and North Fork Koktuli Rivers, Upper
Talarik Creek, and 77 wadeable stream sites in the Nushagak and Kvichak River
watersheds 9-21
Table 9-7 Comparison of average metal concentrations of tailings (Appendix H) to threshold effect
concentration and probable effect concentration values for freshwater sediments and
sums of the quotients (XTU) 9-34
Table 9-8 Results of applying the biotic ligand model to mean water chemistries of tailings
leachates and supernatants to derive effluent-specific copper criteria 9-38
Table 9-9 Summary of evidence concerning risks to fish from the toxic effects of a tailings dam
failure 9-44
Table 10-1 Proportion of stream channel length in stream subwatersheds intersected by the
transportation corridor (Scale 5) classified according to stream size (based on mean
annual discharge in m3/s), channel gradient (%), and floodplain potential (based on %
flatland in lowland) 10-8
Table 10-2 Average number of spawning adult sockeye salmon at locations near the transportation
corridor 10-11
Table 10-3 Proximity of the transportation corridor to National Hydrography Dataset streams (USGS
2012) 10-17
Table 10-4 Proximity of the transportation corridor to National Wetlands Inventory wetlands, ponds,
and small lakes (USFWS 2012) 10-18
Table 10-5 Proximity of the transportation corridor to water, in terms of the length occurring within
200 m of National Hydrography Dataset streams (USGS 2012) or National Wetlands
Inventory wetlands, ponds, and small lakes (USFWS 2012) 10-19
Table 10-6 Road-stream crossings a long the transportation corridor, upstream lengths of streams
of different sizes likely to support salmonids (based on stream gradients of less than
12%), and downstream lengths to Iliamna Lake 10-22
Table 10-7 Stream lengths downstream of road-stream crossings, classified by stream size 10-25
Table 10-8 Lengths of different stream sizes that occur upstream of road-stream crossings and are
likely to support salmonids (based on stream gradients of less than 12%) 10-26
Table 11-1 Studies that examined pipeline failure rates 11-6
Table 11-2 Parameters for concentrate pipeline spills to Chinkelyes Creek and Knutson Creek 11-8
Table 11-3 Comparison of mean metal concentrations in product concentrate from the Aitik
(Sweden) porphyry copper mine (Appendix H) to threshold effect concentration and
probable effect concentration values for fresh water 11-12
Table 11-4 Aquatic toxicological screening of leachates from Aitik (Sweden) product concentrate
(Appendix H) based on acute (criterion maximum concentration) and chronic (criterion
continuous concentration) water quality criteria or equivalent benchmarks, and
quotients of concentrations divided by benchmark values 11-13
Table 11-5 Summary of evidence concerning risks to fish from a product concentrate spill 11-17
Table 11-6 Parameters for return water pipeline spills to Chinkelyes and Knutson Creeks 11-20
Table 11-7 Parameters for diesel pipeline spills to Chinkelyes and Knutson Creeks 11-21
Table 11-8 Toxicity of diesel fuel to freshwater organisms in laboratory tests 11-26
Table 11-9 Cases of diesel spills into streams 11-27
Table 11-10 Summary of evidence concerning risks to fish from a diesel spill 11-30
Table 13-1 Mining prospects (in addition to the Pebble deposit) with more than minimal recent
exploration in the Nushagak and Kvichak River watersheds 13-3
Table 13-2 Waters, fishes, and subsistence uses potentially affected by a mine at the Pebble
South/PEB prospect 13-11
Table 13-3 Waters, fishes, and subsistence uses potentially affected by a mine at the Big Chunk
South prospect 13-12
Bristol Bay Assessment
XII
January 2014
-------
Table 13-4 Waters, fishes, and subsistence uses potentially affected by a mine at the Big Chunk
North prospect 13-13
Table 13-5 Waters, fishes, and subsistence uses potentially affected by a mine at the Groundhog
prospect 13-14
Table 13-6 Waters, fishes, and subsistence uses potentially affected by a mine at the
AUDN/lliamna prospect 13-16
Table 13-7 Waters, fishes, and subsistence uses potentially affected by a mine at the Humble
prospect 13-18
Table 13-8 Streams, water bodies, and wetlands potentially eliminated by additional large-scale
mines in the Nushagak and Kvichak River watersheds 13-21
Table 14-1 Probabilities and consequences of potential failures in the mine scenarios 14-5
Table 14-2 Summary of estimated stream lengths potentially affected in the three mine size
scenarios, assuming routine operations 14-18
Bristol Bay Assessment
XIII
January 2014
-------
List of Figures
Figure 2-1 Conceptual model illustrating sources, stressors, and responses potentially associated
with large-scale mine development in the Bristol Bay watershed 2-6
Figure 2-2 The five geographic scales considered in this assessment 2-10
Figure 2-3 The Bristol Bay watershed (Scale 1), comprising the Togiak, Nushagak, Kvichak,
Naknek, Egegik, and Ugashik River watersheds and the North Alaska Peninsula 2-11
Figure 2-4 The Nushagak and Kvichak River watersheds (Scale 2) 2-12
Figure 2-5 The mine scenario watersheds—South Fork Koktuli River, North Fork Koktuli River, and
Upper Talarik Creek—within the Nushagak and Kvichak River watersheds (Scale 3) 2-13
Figure 2-6 Footprints of the major mine components for the three scenarios evaluated in the
assessment (Scale 4) 2-14
Figure 2-7 The transportation corridor area (Scale 5), comprising 32 subwatersheds in the Kvichak
River watershed thatdrain to Iliamna Lake 2-15
Figure 3-1 Hydrologic landscapes within the Nushagak and Kvichak River watersheds, as defined
by physiographic division and climate class 3-4
Figure 3-2 Distribution of mean annual precipitation (mm) across the Nushagak and Kvichak River
watersheds, 1971 to 2000 (SNAP 2012) 3-5
Figure 3-3 Generalized geology of the Bristol Bay watershed (adapted from Selkregg 1974) 3-6
Figure 3-4 Occurrence of permafrost in the Bristol Bay watershed (adapted from Selkregg 1974) 3-7
Figure 3-5 Dominant soils in the Bristol Bay watershed (adapted from Selkregg 1974) 3-8
Figure 3-6 Erosion potential in the Bristol Bay watershed (adapted from Selkregg 1974) 3-9
Figure 3-7 Dominant vegetation in the Bristol Bay watershed (adapted from Selkregg 1974) 3-10
Figure 3-8 Physiographic divisions of the Nushagak and Kvichak River watersheds of Bristol Bay 3-11
Figure 3-9 Groundwater resources in the Bristol Bay watershed (adapted from Selkregg 1974) 3-15
Figure 3-10 Mean monthly runoff for selected streams and rivers in the Nushagak and Kvichak
River watersheds 3-17
Figure 3-11 Examples of different stream size and gradient classes in the Nushagak and Kvichak
River watersheds 3-21
Figure 3-12 Channel gradient classes in the Nushagak and Kvichak River watersheds 3-22
Figure 3-13 Likelihood of floodplain potential, as measured by the percent flatland in lowland
areas, for the Nushagak and Kvichak River watersheds 3-28
Figure 3-14 Stream size classes in the Nushagak and Kvichak River watersheds as determined by
mean annual streamflow 3-29
Figure 3-15 Seismic activity in southwestern Alaska 3-32
Figure 3-16 Mean annual temperature across the Bristol Bay watershed under (A) historical
conditions (1971 to 2000) and (B) the A2 emissions scenario (2071 to 2099), and (C)
the temperature change between these two climate scenarios (SNAP 2012) 3-38
Figure 3-17 Mean annual precipitation across the Bristol Bay watershed under (A) historical
conditions (1971 to 2000) and (B) the A2 emissions scenario (2071 to 2099), and (C)
the precipitation change between these two climate scenarios (SNAP 2012) 3-39
Figure 3-18 Mean annual water surplus (precipitation minus evapotranspiration) across the Bristol
Bay watershed under (A) historical conditions (1971 to 2000) and (B) the A2 emissions
scenario (2071 to 2099), and (C) the water surplus change between these two climate
scenarios (SNAP 2012) 3-40
Figure 3-19 Relationship between time from fertilization to emergence and temperature for the five
Pacific salmon species 3-42
Figure 4-1 Porphyry copper deposits around the world 4-4
Figure 4-2 Neutralizing potential at the Bingham Canyon porphyry copper deposit in Utah 4-7
Figure 4-3 Simplified schematic of mined material processing 4-14
Bristol Bay Assessment • January 2014
-------
Figure 4-4 Cross-sections illustrating (A) upstream, (B) downstream, and (C) centerline tailings
dam construction 4-18
Figure 5-1 Approximate extents of popular Chinook and sockeye salmon recreational fisheries in
the vicinity of the Nushagakand Kvichak River watersheds 5-6
Figure 5-2 Subsistence harvest and harvest effort areas for salmon and other fishes within the
Nushagakand Kvichak River watersheds 5-7
Figure 5-3 Diversity of Pacific salmon species production in the Nushagak and Kvichak River
watersheds 5-16
Figure 5-4 Reported sockeye salmon stream distribution in the Nushagak and Kvichak River
watersheds 5-17
Figure 5-5 Reported Chinook salmon distribution in the Nushagak and Kvichak River watersheds 5-18
Figure 5-6 Reported coho salmon distribution in the Nushagak and Kvichak River watersheds 5-19
Figure 5-7 Reported chum salmon distribution in the Nushagakand Kvichak River watersheds 5-20
Figure 5-8 Reported pink salmon distribution in the Nushagak and Kvichak River watersheds 5-21
Figure 5-9 Proportion of total sockeye salmon run sizes by (A) region and (B) watershed within the
Bristol Bay region 5-22
Figure 5-10 Reported rainbow trout occurrence in the Nushagak and Kvichak River watersheds 5-24
Figure 5-11 Reported Dolly Varden occurrence in the Nushagak and Kvichak River watersheds 5-25
Figure 5-12 Subsistence use intensity for salmon, other fishes, wildlife, and waterfowl within the
Nushagakand Kvichak River watersheds 5-42
Figure 6-1 Footprint of the major mine components (mine pit, waste rock piles, and tailings
storage facility [TSF]) in the Pebble 0.25 scenario 6-5
Figure 6-2 Footprint of the major mine components (mine pit, waste rock piles, and tailings
storage facility [TSF]) in the Pebble 2.0 scenario 6-6
Figure 6-3 Footprint of the major mine components (mine pit, waste rock piles, and tailings
storage facilities [TSFs]) in the Pebble 6.5 scenario 6-7
Figure 6-4 Height of the dam at tailing storage facility (TSF) 1 in the Pebble 2.0 and Pebble 6.5
scenarios, relative to U.S. landmarks 6-12
Figure 6-5 Water management and water balance components for the three mine scenarios 6-14
Figure 6-6 Transportation corridor connecting the Pebble deposit area to Cook Inlet 6-18
Figure 6-7 Hydraulic conductivity in the Pebble depositarea 6-26
Figure 6-8 Water flow schematic for the Pebble 0.25 scenario 6-28
Figure 6-9 Water flow schematic for the Pebble 2.0 scenario 6-29
Figure 6-10 Water flow schematic for the Pebble 6.5 scenario 6-30
Figure 6-11 Approximate locations of stream gages and wastewater treatment plant (WWTP)
discharges represented in Figures 6-8 through 6-10 6-31
Figure 6-12 Conceptual model illustrating potential effects of routine mine construction and
operation on physical habitat 6-43
Figure 6-13 Conceptual model illustrating potential effects of routine mine construction and
operation on water chemistry 6-44
Figure 6-14 Conceptual model illustrating potential effects of unplanned events on physical habitat
and water chemistry 6-45
Figure 7-1 Conceptual model illustrating potential linkages between sources associated with the
mine scenario footprints, changes in physical habitat, and fish endpoints 7-3
Figure 7-2 Reported sockeye salmon distribution in the mine scenario watersheds 7-5
Figure 7-3 Reported coho salmon distribution in the mine scenario watersheds 7-6
Figure 7-4 Reported Chinook salmon distribution in the mine scenario watersheds 7-7
Figure 7-5 Reported chum salmon distribution in the mine scenario watersheds 7-8
Figure 7-6 Reported pink salmon distribution in the mine scenario watersheds 7-9
Figure 7-7 Reported Dolly Varden occurrence in the mine scenario watersheds 7-10
Figure 7-8 Reported rainbow trout occurrence in the mine scenario watersheds 7-11
Bristol Bay Assessment January 2014
-------
Figure 7-9 Cumulative frequency of stream channel length classified by (A) mean annual
streamflow (MAP) (m3/s), (B) channel gradient (%), and (C) floodplain potential (based
on % flatland in lowland) for the mine scenario watersheds (Scale 3) versus the
Nushagakand Kvichak River watersheds (Scale 2) 7-18
Figure 7-10 Streams and wetlands lost (eliminated, blocked, or dewatered) in the Pebble 0.25
scenario 7-20
Figure 7-11 Streams and wetlands lost (eliminated, blocked, or dewatered) in the Pebble 2.0
scenario 7-21
Figure 7-12 Streams and wetlands lost (eliminated, blocked, or dewatered) in the Pebble 6.5
scenario 7-22
Figure 7-13 Cumulative frequency of stream channel length classified by (A) mean annual
streamflow (m3/s), (B) channel gradient (%), and (C) floodplain potential (based on %
flatland in lowland) for the mine footprints (Scale 4) versus the Nushagak and Kvichak
River watersheds (Scale 2) 7-25
Figure 7-14 Stream segments in the mine scenario watersheds showing streamflow changes (%)
associated with the Pebble 0.25 footprint 7-37
Figure 7-15 Stream segments in the mine scenario watersheds showing streamflow changes (%)
associated with the Pebble 2.0 footprint 7-38
Figure 7-16 Stream segments in the mine scenario watersheds showing streamflow changes (%)
associated with the Pebble 6.5 footprint 7-39
Figure 7-17 Monthly mean streamflows for stream gages in the (A) South Fork Koktuli River,
(B) North Fork Koktuli River, and (C) Upper Talarik Creek watersheds, based on water
years 2 004 thro ugh 2010 7-47
Figure 7-18 Monthly mean pre-miningstreamflow for South Fork Koktuli River gage SK100F (bold
solid line), with 10 and 20%sustainability boundaries (gray lines) and projected
monthly mean streamflows, in the Pebble 0.25 scenario (dashed line) 7-54
Figure 8-1 Conceptual model illustrating the pathways linking water treatment, discharge, fate,
and effects 8-2
Figure 8-2 Processes involved in copper uptake as defined in the biotic ligand model (USEPA
2007) 8-23
Figure 8-3 Comparison of copper concentrations in leachates and background water to state
hardness-based acute (CMC) and chronic (CCC) water quality criteria for copper 8-34
Figure 9-1 Conceptual model illustrating potential pathways linking tailings dam failure and effects
on fish endpoints 9-3
Figure 9-2 Annual probability of dam failure due to slope failure vs. factor of safety (modified from
Si Iva eta I. 2008) 9-11
Figure 9-3 Representative particle size distributions for tailings solids (bulk and pyritic tailings)
and tailings dam rockfill 9-13
Figure 10-1 Streams, wetlands, ponds, and lakes a long the transportation corridor 10-3
Figure 10-2 High-impact areas along the transportation corridor 10-4
Figure 10-3 Conceptual model showing potential pathways linking the transportation corridor and
related sources tostressors and assessment endpoints 10-5
Figure 10-4 Cumulative frequency of stream channel length classified by (A) mean annual
streamflow (MAF) (m3/s), (B) channel gradient (%), and (C) floodplain potential (based
on % flatland in lowland) for stream subwatersheds intersected by the transportation
corridor (Scale 5) versus the Nushagak and Kvichak River watersheds (Scale 2) 10-9
Figure 10-5 Location of sockeye salmon surveys and number of spawners observed along the
transportation corridor 10-12
Figure 10-6 Reported salmon, Dolly Varden, and rainbow trout distributions along the transportation
corridor 10-14
Figure 11-1 Conceptual model illustrating potential stressors and effects resulting from a
concentrate pipeline failure 11-2
Bristol Bay Assessment • January 2014
-------
Figure 11-2 Conceptual model illustrating potential stressors and effects resulting from a return
water pipeline failure 11-3
Figure 11-3 Conceptual model illustrating potential stressors and effects resulting from a diesel
pipeline failure 11-4
Figure 12-1 Conceptual model illustrating potential effects on wildlife resulting from effects on
salmon 12-3
Figure 12-2 Conceptual model illustrating potential effects on Alaska Native cultures resulting from
effects on salmon and other fishes 12-4
Figure 13-1 Claim blocks with more than minimal recent exploration in the Nushagak and Kvichak
River watersheds 13-4
Figure 13-2 Conceptual model illustrating potential cumulative effects of multiple mines 13-5
Figure 13-3 Location of claim blocks in relation to subsistence use intensity for salmon, other
fishes, wildlife, and waterfowl in the Nushagak and Kvichak River watersheds 13-34
Bristol Bay Assessment ,, January 2014
-------
List of Boxes
Box 1-1 Stakeholder involvement in the assessment 1-5
Box 1-2 Overview of the assessment's peer review process 1-7
Box 2-1 Conceptual models 2-2
Box 2-2 Exploratory mining activities 2-7
Box 2-3 Key salmonids in the Bristol Bay watershed 2-8
Box 2-4 The National Hydrography Dataset 2-9
Box 3-1 Methods for characterizing channel gradient 3-20
Box 3-2 Methods for characterizing mean annual streamflow 3-24
Box 3-3 Methods for characterizing percent flatland in lowland 3-26
Box 3-4 Methods for climate change projections 3-35
Box 4-1 Reducing mining's impacts 4-8
Box 4-2 Permitting large mine projects in Alaska 4-9
Box 4-3 Financial assurance 4-10
Box 4-4 Block caving and subsidence 4-12
Box4-5 Chemicals used in ore processingand handling 4-15
Box 4-6 Use of cyanide in gold recovery 4-16
Box 4-7 Dry stack tailings management 4-19
Box 5-1 Cultural groups in the Bristol Bay watershed 5-2
Box 5-2 Subsistence use methodology 5-8
Box 5-3 Salmon in freshwater and terrestrial foodwebs 5-12
Box 5-4 Commercial fisheries management in the Bristol Bay watershed 5-15
Box 5-5 Testimony on the importance of subsistence use 5-38
Box 6-1 Cumulative impacts of a large-scale porphyry copper mine 6-3
Box 6-2 Mine pit drawdown calculations 6-25
Box 7-1 Calculation of streams and wetlands affected by mine scenario footprints 7-23
Box 7-2 Compensatory mitigation 7-34
Box 8-1 An accidental tailings water release: Nixon Fork Mine, Alaska, winter 2012 8-16
Box 8-2 Potential failures of reverse osmosis wastewater treatment plants 8-17
Box 8-3 Use of risk quotients to assess toxicological effects 8-33
Box 8-4 The Fraser River 8-51
Box 9-1 Examples of historical tailings dam failures 9-2
Box 9-2 Selecting earthquake characteristics for design criteria 9-6
Box 9-3 Interpretation of dam failure probabilities 9-8
Box 9-4 Methods for modelingtailings dam failures 9-15
Box 9-5 Background on relevant analogous tailings spill sites 9-31
Box 10-1 Calculation of stream lengths and wetland areas affected by transportation corridor
development 10-16
Box 10-2 Culvert mitigation 10-30
Box 10-3 Stormwater runoff and fine sediment mitigation 10-36
Box 10-4 Mitigation for invasive species 10-42
Box 10-5 Likely effectiveness of mitigation measures 10-46
Bristol Bay Assessment ,,, January 2014
-------
Box 12-1 Potential direct effects of mining 12-2
Box 12-2 Testimony on potential effects of mining on Alaska Native cultures 12-8
Box 13-1 Methods for estimating impacts of other mines 13-9
Box 13-2 Examples of mine characterization errors 13-31
Box 14-1 Failure probabilities 14-6
Box 14-2 Climate change and potential risks of large-scale mining 14-16
Bristol Bay Assessment • January 2014
-------
Acronyms and Abbreviations
AAC
ADEC
ADF&G
ADNR
ADOT
AFFI
ANCSA
AP
APDES
API
ASME
AUC
AVS
AW
AWC
BBAP
BLM
BMP
CCC
CFR
CH
CIBB
CMC
CPU
CWA
DBB
DEM
DOC
EC2o
ECso
ELso
E-R
ERA
FA
FERC
FK
FP
FR
FS
GCM
GIS
GMU
HEC-RAS
HUC
IA
1C
IC2o
ICso
IFIM
Alaska Administrative Code
Alaska Department of Environmental Conservation
Alaska Department of Fish and Game
Alaska Department of Natural Resources
Alaska Department of Transportation and Public Facilities
Alaska Freshwater Fish Inventory
Alaska Native Claims Settlement Act
acid-generation potential
Alaska Pollutant Discharge Elimination System
American Petroleum Institute
American Society of Mechanical Engineers
area under curve
acid volatile sulfides
ambient waters
Anadromous Waters Catalog
Bristol Bay Area Plan for State Lands
biotic ligand model
best management practice
criterion continuous concentration
Code of Federal Regulations
Channel
Cook Inlet-to-Bristol Bay
criterion maximum concentration
Climate Research Unit
Clean Water Act
Dillingham/Bristol Bay
digital elevation model
dissolved organic carbon
20% effective concentration
median effective concentration
median effective level
exposure-response relationship
ecological risk assessment
fish avoidance
Federal Energy Regulatory Commission
fish kill
high floodplain potential
fish reproduction
fish sensory
global climate model
geographic information system
Game Management Unit
Hydrologic Engineering Center's River Analysis System
hydrologic unit code
invertebrate acute
invertebrate chronic
20% inhibitory concentration
median inhibitory concentration
Instream Flow Incremental Methodology
Bristol Bay Assessment
xx
January 2014
-------
IGTT
LCso
LFP
MAP
MCE
MDE
MDN
Mi
MOA
NA
NAG
NANA
NDM
NED
NFP
NHD
NNP
NP
NPR
NWI
QBE
OHW
PAG
PEC
PEL
PET
PHABSIM
PLP
PRISM
Reclamation
RFP
SCADA
SEM
SNAP
SWATP
SWPPP
IDS
TEC
TEL
TLm
TSF
USAGE
USEPA
USFWS
USGS
WWTP
Intergovernmental Technical Team
median lethal concentration
left flood plain
mean annual streamflow
maximum credible earthquake
maximum design earthquake
marine-derived nutrients
Minerals
memorandum of agreement
not applicable
non-acid-generating
NANA Regional Corporation, Inc.
Northern Dynasty Minerals
National Elevation Dataset
no or low floodplain potential
National Hydrography Dataset
net neutralizing potential
neutralizing potential
neutralizing potential ratio
National Wetlands Inventory
operating basis earthquake
ordinary high water
potentially acid-generating
probable effect concentration
probable effect level
potential evapotranspiration
Physical Habitat Simulation
Pebble Limited Partnership
Parameter-elevation Regressions on Independent Slopes Model
Bureau of Reclamation
right floodplain
supervisory control and data acquisition
simultaneously extracted metals
Scenarios Network for Alaska and Arctic Planning
Southwest Alaska Transportation Plan
stormwater pollution prevention plan
total dissolved solids
threshold effect concentration
threshold effect level
equivalent to LCso
tailings storage facility
U.S. Army Corps of Engineers
U.S. Environmental Protection Agency
U.S. Fish and Wildlife Service
U.S. Geological Survey
wastewater treatment plant
Bristol Bay Assessment
XXI
January 2014
-------
Units of Measure
|jg microgram
|jS micro-Siemens
°C degrees Celsius
cm centimeter
g gram
ha hectare
kg kilogram
km kilometer
km2 square kilometers
km-yr kilometer-year
L liter
m meter
m2 square meter
m3 cubic meter
mg milligram
mm millimeter
s second
t ton
yr year
Bristol Bay Assessment ,, January 2014
-------
Unit of Measure Conversion Chart
Metric
1 [jg(microgram)
1 mg (milligram)
Ig(gram)
1 kg (kilogram)
1 metric ton
1 mm (millimeter)
1 cm (centimeter)
1 m (meter)
1 m2 (square meter)
1 m3 (cubic meter)
1 km (kilometer)
1 km2 (square kilometer) or 100 ha (hectares)
1 ha (hectare)
1L (liter)
1°C (degrees Celsius)
Standard
3.527396 x lQ-°8 ounces
3.527396 x 10-°5 ounces
0.035 ounce
2.202 pounds
1.103 tons
0.039 inch
0.39 inch
3.28 feet
10.764 square feet
35.314 cubic feet
0.621 mile
0.386 square mile
2.47 acres
0.264 gallon
1.8°C + 32° Fahrenheit
Bristol Bay Assessment
xxill
January 2014
-------
Elements and Chemical Symbols
Ag
Al
As
B
Ba
Be
Bi
Ca
CaCOs
Cd
Cl
CN
Co
Cr
Cu
F
Fe
Ga
Hg
In
K
Mg
Mn
Mo
Na
Ni
0
Pb
S
Sb
Se
Se+4
Se+6
Si
Si02
Sn
S04
Sr
Te
Th
Tl
U
V
Zn
silver
aluminum
arsenic
boron
barium
beryllium
bismuth
calcium
calcium carbonate
cadmium
chlorine
cyanide
cobalt
chromium
copper
fluorine
iron
gallium
mercury
indium
potassium
magnesium
manganese
molybdenum
sodium
nickel
oxygen
lead
sulfur
antimony
selenium
selenate
selenite
silicon
silicon dioxide
tin
sulfate
strontium
tellurium
thorium
thallium
uranium
vanadium
zinc
Bristol Bay Assessment
XXIV
January 2014
-------
This assessment represents a collaboration among the U.S. Environmental Protection Agency's (USEPA's)
Region 10, Office of Water, and Office of Research and Development. It was conducted as an ecological risk
assessment to evaluate the potential impacts of large-scale porphyry copper mine development on salmon and
other salmonid fishes and their habitats and consequent effects on wildlife and Alaska Native cultures in the
Nushagak and Kvichak River watersheds of Bristol Bay, Alaska. It is not an assessment of a specific mine
proposal for development, but the mine scenarios considered in the assessment are based on a published
plan to mine the Pebble deposit. The assessment does not outline or evaluate decisions made or to be made
by USEPA.
The first external review draft of this assessment (EPA 910-R-12-004) was released in May 2012 for a 60-day
public comment period and external peer review by 12 independent expert reviewers. The revised, second
external review draft was released in April 2013 (EPA 910-R-12-004B) for another 60-day public comment
period and follow-on review by the same 12 peer reviewers. All public and peer review comments on the two
drafts were considered in the development of this final assessment.
Bristol Bay Assessment
xxv
January 2014
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
Authors (listed alphabetically)
Rebecca Aicher, AAAS Fellow, USEPA-ORD, Washington, DC.
Greg Blair, ICF International, Seattle, WA
Barbara Butler, USEPA-ORD, Cincinnati, OH
Heather Dean, USEPA-Region 10, Anchorage, AK
Joseph Ebersole, USEPA-ORD, Corvallis, OR
Sheila Eckman, USEPA-Region 10, Seattle, WA
Tami Fordham, USEPA-Region 10, Seattle, WA
Jeffrey Frithsen, USEPA-ORD, Washington, DC
Ralph Grismala, ICF International, Lexington, MA
Michael Kravitz, USEPA-ORD, Cincinnati, OH
Phil North, USEPA-Region 10, Soldotna, AK (Retired)
Richard Parkin, USEPA-Region 10, Seattle, WA
Jim Rice, ICF International, Lexington, MA
Dan Rinella, University of Alaska, Anchorage, AK
Kate Schofield, USEPA-ORD, Washington, DC
Steve Seville, ICF International, Portland, OR
Glenn Suter, USEPA-ORD, Cincinnati, OH
Jason Todd, USEPA-ORD, Washington, DC
Michael Wiedmer, Malma Consulting, Anchorage, AK
Parker J. Wigington, Jr., USEPA-ORD, Corvallis, OR (Retired)
Contributors (listed alphabetically)
Dave Athens, Kenai River Center, Soldotna, AK
Alan Barnard, ICF International, Sacramento, CA
Deborah Bartley, ICF International, Seattle, WA
David Bauer, ICF International, Fairfax, VA
Alan Boraas, Kenai Peninsula College, Soldotna, AK
Philip Brna, USFWS, Anchorage, AK
Saadia Byram, ICF International, Irvine, CA
Laura Cooper, ICF International, Portland, OR
Eric Doyle, ICF International, Seattle, WA
John Duffield, Bioeconomics, Inc., Missoula, MT
Lorraine Edmond, USEPA-Region 10, Seattle, WA
Ginny Fay, University of Alaska, Anchorage, AK
Rachel Fertik, USEPA-OW, Washington, DC
Christopher Frissell, Pacific Rivers Council, Poison, MT
Cindi Godsey, USEPA-Region 10, Anchorage, AK
Oliver Scott Goldsmith, University of Alaska, Anchorage, AK
Michael Griffith, USEPA-ORD, Cincinnati, OH
Carol-Anne Hicks, ICF International, Sacramento, CA
Stephen Hoffman, USEPA-OSWER, Washington, DC
Palmer Hough, USEPA-OW, Washington, DC
Maureen Johnson, USEPA-ORD, Washington, DC
Gunnar Knapp, University of Alaska, Anchorage, AK
Catherine Knott, Kenai Peninsula College, Homer, AK
Bristol Bay Assessment
XXVI
January 2014
-------
Douglas Limpinsel, NOAA, Anchorage, AK
James Lopez-Baird, USEPA-Region 10, Seattle, WA
Michael McManus, USEPA-ORD, Cincinnati, OH
Chris Neher, Bioeconomics, Inc., Missoula, MT
Grant Novak, ICF International, Seattle, WA
Corrine Ortega, ICF International, Sacramento, CA
Aaron Park, Department of the Army, Anchorage, AK
David Patterson, Bioeconomics, Inc., Missoula, MT
Ryan Patterson, ICF International, Los Angeles, CA
Rori Perkins, ICF International, Portland, OR
Caroline Ridley, USEPA-ORD, Washington, DC
Ken Rock, ICF International, Fairfax, VA
Tobias Schworer, University of Alaska, Anchorage, AK
Robert Seal, USGS, Reston, VA
Sacha Selim, ICF International, San Francisco, CA
Rebecca Shaftel, University of Alaska, Anchorage, AK
Michael Slimak, USEPA-ORD, Washington, DC
Judy Smith, USEPA-Region 10, Portland, OR
Danny Stratten, ICF International, Bellingham, WA
Greg Summers, ICF International, Portland, OR
Jenny Thomas, USEPA-OW, Washington, DC
Lori Verbrugge, USFWS, Anchorage, AK
Barbara Wolf, ICF International, Sacramento, CA
Reviewers of Internal Review Drafts (listed alphabetically)
Dwight Atkinson, USEPA-OW, Washington, DC
Ned Black, USEPA-Region 9, San Francisco, CA (check)
Adrianne Fleek, USEPA-Region 10, Anchorage, AK
John Goodin, USEPA-OW, Washington, DC
Cami Grandinetti, USEPA-Region 10, Washington, DC
Scot Hagerthey, USEPA-ORD, Washington, DC
James Hanley, USEPA-Region 8, Denver, CO
Stephen Hoffman, USEPA-OSW, Washington, DC
Chris Hunter, USEPA-OW, Washington, DC
Thomas Johnson, USEPA-ORD, Washington, DC
Phil Kaufman, USEPA-ORD, Corvallis, OR
Stephen LeDuc, USEPA-ORD, Washington, DC
Julia McCarthy, USEPA-Region 8, Denver, CO
Caroline Ridley, USEPA-ORD, Washington, DC
Carol Russell, USEPA-Region 8, Denver, CO
Dave Tomten, USEPA-Region 10, Seattle, WA
Brian Topping, USEPA-OW, Washington, DC
Felicia Wright, USEPA-OW, Washington, DC
Reviewers of External Review Draft (listed alphabetically)
David Atkins, Watershed Environmental, LLC
Steve Buckley, WHPacific
Courtney Carothers, University of Alaska Fairbanks
Dennis Dauble, Washington State University
Gordon Reeves, USDA Pacific Northwest Research Station
Charles Slaughter, University of Idaho
Bristol Bay Assessment ,, January 2014
-------
John Stednick, Colorado State University
Roy Stein, The Ohio State University
William Stubblefield, Oregon State University
Dirk van Zyl, University of British Columbia
Phyllis Weber Scannell, Scannell Scientific Services
Paul Whitney, Independent Consultant
Bristol Bay Assessment xxviii January 2014
-------
PHOTO CREDITS
Front cover Main photo: Upper Talarik Creek (Joe Ebersole, USEPA)
Thumbnail 1: Brown bear (Steve Hillebrand, USFWS)
Thumbnail 2: Fishing boats at Naknek, Alaska (USEPA)
Thumbnail 3: Iliamna Lake (Lorraine Edmond, USEPA)
Thumbnail 4: Sockeye salmon in the Wood River (Thomas Quinn, University of Washington)
Title Pages
Executive Area of tailings storage facility 1 in the mine scenarios (Michael Wiedmer)
Summary Sockeye salmon near Gibraltar Lake (Thomas Quinn, University of Washington)
Tributary of Napotoli Creek, near the Humble claim (Michael Wiedmer)
Chapter 1 Kvichak River below Iliamna Lake and Igiugig (Joe Ebersole, USEPA)
Salmon art on a building in Dillingham (Alan Boraas, Kenai Peninsula College)
Sockeye salmon in Gibraltar Creek (Thomas Quinn, University of Washington)
Chapter 2 Pebble deposit area (Lorraine Edmond, USEPA)
Sockeye salmon in Wood River (Thomas Quinn, University of Washington)
Knutson Creek draining into the Knutson Bay area of Iliamna Lake (Keith Denton)
Chapter 3 Iliamna Lake (Lorraine Edmond, USEPA)
Homes in Nondalton (Alan Boraas, Kenai Peninsula College)
Groundwater upwellingnear Kaskanak Creek, Lower Talarik basin (Joe Ebersole, USEPA)
Chapter 4 Area of the Pebble deposit (Joe Ebersole, USEPA)
Brown bear feeding on salmon (Steve Hillebrand, USFWS)
Lodge on the Kvichak River (Joe Ebersole, USEPA)
Chapter 5 Nushagak River at Koliganek (Alan Boraas, Kenai Peninsula College)
Sockeye salmon in Wood River (Thomas Quinn, University of Washington)
Fishing boats at Naknek, Alaska (USEPA)
Chapter 6 Subsistence skiffs at New Stuyahok (Alan Boraas, Kenai Peninsula College)
Sockeye salmon near Pedro Bay, Iliamna Lake (Thomas Quinn, University of Washington)
Tributary near the Humble claim and Ekwok (Joe Ebersole, USEPA)
Chapter 7 Salmon drying at Koliganek (Alan Boraas, Kenai Peninsula College)
Beaver pond succession in Upper Talarik Creek (Joe Ebersole, USEPA)
Rainbow trout (USEPA)
Chapter 8 Area of tailings storage facility 1 in the mine scenarios (Michael Wiedmer)
Sockeye salmon near Gibraltar Lake (Thomas Quinn, University of Washington)
Tributary of Napotoli Creek, near the Humble claim (Michael Wiedmer)
Chapter 9 Kvichak River below Iliamna Lake and Igiugig (Joe Ebersole, USEPA)
Salmon art on a building in Dillingham (Alan Boraas, Kenai Peninsula College)
Sockeye salmon in Gibraltar Creek (Thomas Quinn, University of Washington)
Chapter 10 Pebble deposit area (Lorraine Edmond, USEPA)
Sockeye salmon in Wood River (Thomas Quinn, University of Washington)
Knutson Creek draining into the Knutson Bay area of Iliamna Lake (Keith Denton)
Chapter 11 Iliamna Lake (Lorraine Edmond, USEPA)
Homes in Nondalton (Alan Boraas, Kenai Peninsula College)
Groundwater upwellingnear Kaskanak Creek, Lower Talarik basin (Joe Ebersole, USEPA)
Chapter 12 Area of the Pebble deposit (Joe Ebersole, USEPA)
Brown bear feeding on salmon (Steve Hillebrand, USFWS)
Lodge on the Kvichak River (Joe Ebersole, USEPA)
Chapter 13 Nushagak River at Koliganek (Alan Boraas, Kenai Peninsula College)
Sockeye salmon in Wood River (Thomas Quinn, University of Washington)
Fishing boats at Naknek, Alaska (USEPA)
Bristol Bay Assessment
XXIX
January 2014
-------
Chapter 14 Subsistence skiffs at New Stuyahok (Alan Boraas, Kenai Peninsula College)
Sockeye salmon near Pedro Bay, Iliamna Lake (Thomas Quinn, University of Washington)
Tributary near the Humble claim and Ekwok (Joe Ebersole, USEPA)
Chapter 15 Salmon drying at Koliganek (Alan Boraas, Kenai Peninsula College)
Beaver pond succession in Upper Talarik Creek (Joe Ebersole, USEPA)
Rainbow trout (USEPA)
Bristol Bay Assessment January 2014
-------
ACKNOWLEDGEMENTS
Assistance for this assessment was provided by ICF International under USEPA contract numbers
EP-C-09-009 and EP-C-14-001, and by NatureServe under USEPA contract number EP-W-07-080.
Additional assistance was provided by Colleen Matt (USEPA Contract EP-12-H-000099), Northern
Ecologic LLC (USEPA Contract EPA-12-H-001001), Charles Schwartz (USEPA Contract EP-12-H-000105)
and Halcyon Research (USEPA Contract EP-12-H-000107). The initial external peer review of the
assessment was coordinated by Versar, Inc., under USEPA contract number EP-C-07-025. The external
peer review of specific supplemental materials provided during the public comment period was
coordinated by Versar, Inc., under USEPA contract number EP-C-12-045. Assistance with the
management of public comments was provided by Horsley-Witten under USEPA contract EP-C-08-018.
Contractors contributing to this report were required to certify that they had no organizational conflicts
of interest. As defined by Federal Acquisition Regulations subpart 2.101, an organizational conflict of
interest may exist when, "because of other activities or relationships with other persons, a person is
unable or potentially unable to render impartial assistance or advice to the Government, or the person's
objectivity in performing the contract work is or might otherwise be impaired or a person has an unfair
competitive advantage."
We would like to acknowledge the following people for their efforts in developing, completing, and
releasing this assessment: David Allnut, Liz Blackburn, Andrew Bostrom, David Bottimore, Betzy Colon,
Taukecha Cunningham, Kacee Deener, Brittany Ekstrom, Dave Evans, Ross Geredien, Rick Griffin, Rusty
Griffin, Marianne Holsman, Carol Hubbard, Cheryl Itkin, Mark Jen, Nick Jones, Hanady Kader, Denise
Keehner, Bill Kirchner, Matt Klasen, Terri Konoza, Rockey Louis, Don Maddox, Colleen Matt, Elizabeth
McKenna, Julie Michaelson, Chris Moller, Heidi Nalven, Tim O'Neil, Jim Pendergast, Melissa Revely-
Wilson, Tom Rothe, Stephanie Sarraino, Charles Schwartz, Vicki Soto, Gautam Srinivasan, Cara Steiner-
Riley, Lowell Suring, Michael Szerlog, Jerry Tande, Mike Weimer, Bill Wilen, and Amina Wilkins.
Bristol Bay Assessment
XXXI
January 2014
-------
The Bristol Bay watershed in southwestern Alaska supports the largest sockeye salmon fishery in the
world, is home to 25 federally recognized tribal governments, and contains significant mineral
resources. The potential for large-scale mining activities in the watershed has raised concerns about the
impact of mining on the sustainability of Bristol Bay's world-class commercial, recreational, and
subsistence fisheries and the future of Alaska Native tribes in the watershed, who have maintained a
salmon-based culture and subsistence-based way of life for at least 4,000 years.
The U.S. Environmental Protection Agency (USEPA) launched this assessment to determine the
significance of Bristol Bay's ecological resources and evaluate the potential impacts of large-scale mining
on these resources. It uses the well-established methodology of an ecological risk assessment, which is a
type of scientific investigation that provides technical information and analyses to foster public
understanding and inform future decision making. As a scientific assessment, it does not discuss or
recommend policy, legal, or regulatory decisions, nor does it outline or analyze options for future
decisions.
This assessment characterizes the biological and mineral resources of the Bristol Bay watershed. It is
intended to increase understanding of potential impacts of large-scale mining on the region's fish
resources and serve as a technical resource for the public and for federal, state, and tribal governments
as they consider how best to address the challenges posed by mining and ecological protection in the
Bristol Bay watershed. It will inform ongoing discussions of the risks of mine development to the
sustainability of the Bristol Bay salmon fisheries and thus will be of value to the many stakeholders in
this debate.
The assessment also will inform the consideration of options for future government action, including,
possibly, by USEPA, which has been petitioned by multiple groups to address mining activity in the
Bristol Bay watershed using its authority under the Clean Water Act (CWA). Should specific mine
projects reach the permitting stage, the assessment will enable state and federal permitting authorities
Bristol Bay Assessment
ES-1
January 2014
-------
to make informed decisions to grant, deny, or condition permits and/or conduct additional research or
assessment as a basis for such decisions. USEPA conducted this assessment consistent with its authority
under the CWA Section 104(a) and (b).
Scope of the Assessment
This assessment reviews, analyzes, and synthesizes information relevant to potential impacts of large-
scale mine development on Bristol Bay fisheries and consequent effects on wildlife and Alaska Native
cultures in the region. Given the economic, ecological, and cultural importance of the region's key
salmonids (sockeye, Chinook, coho, chum, and pink salmon, as well as rainbow trout and Dolly Varden)
and stakeholder and public concern that a mine could affect those species, the primary focus of the
assessment is the abundance, productivity, and diversity of these fishes. Because wildlife in Bristol Bay
are intimately connected to and dependent on these and other fishes, changes in these fisheries are
expected to affect the abundance and health of wildlife populations. Alaska Native cultures have strong
nutritional, cultural, social, and spiritual dependence on salmon, so changes in salmon fisheries are
expected to affect the health and welfare of Alaska Native populations. Therefore, wildlife and Alaska
Native cultures are also considered as assessment endpoints, but only as they are affected by changes in
salmonid fisheries.
The assessment considers multiple geographic scales. The largest scale is the Bristol Bay watershed,
which is a largely undisturbed region with outstanding natural, cultural, and mineral resources. Within
the larger Bristol Bay watershed, the assessment focuses on the Nushagak and Kvichak River
watersheds (Figure ES-1). These are the largest of the Bristol Bay watershed's six major river basins,
containing about 50% of the total watershed area, and are identified as mineral development areas by
the State of Alaska. Given its size and extent of characterization, the Pebble deposit is the most likely site
for near-term, large-scale mine development in the region. Because the Pebble deposit is located in the
headwaters of tributaries to both the Nushagak and Kvichak Rivers, both of these watersheds are
subject to potential risks from mining. The third geographic scale is the watersheds of the three
tributaries that originate within the potential footprint of a mine on the Pebble deposit: the South Fork
Koktuli River, which drains the Pebble deposit area and converges with the North Fork west of the
Pebble deposit; the North Fork Koktuli River, located to the northwest of the Pebble deposit, which
flows into the Nushagak River via the Koktuli and Mulchatna Rivers; and Upper Talarik Creek, which
drains the eastern portion of the Pebble deposit and flows into the Kvichak River via Iliamna Lake, the
largest undeveloped lake in the United States (Figure ES-1). The mine footprints in the three realistic
mine scenarios evaluated in the assessment make up the fourth geographic scale. These scenarios—
Pebble 0.25, Pebble 2.0, and Pebble 6.5—define three potential mine sizes, representing different stages
in the potential mining of the Pebble deposit. The final geographic scale is the combined area of the
subwatersheds between the mine footprints and the Kvichak River watershed's eastern boundary that
would be crossed by a transportation corridor linking the mine site to Cook Inlet.
Bristol Bay Assessment F52 January 2014
-------
Figure ES-1. The Nushagak and Kvichak River watersheds of Bristol Bay.
Togiakf
National*-
Wildlife
Refuge
CJ
Cook Inlet
Bristol Bay
IN
A
25
25
50
] Kilometers
50
] Miles
Approximate Pebble Deposit Location
Towns and Villages
Parks, Refuges, or Preserves
Watershed Boundary
Bristol Bay Assessment
ES-3
January 2014
-------
The assessment also addresses two periods for mine activities. The first is the development and
operation phase, during which mine infrastructure would be built and the mine would be operated. This
phase may last from 20 to 100 years or more. The second is the post-mining phase, during which the site
would be monitored and maintained. Water treatment and other waste management activities would
continue as necessary and any failures would be remediated. Because mine wastes would be persistent,
this period could continue for centuries and potentially in perpetuity.
We began the assessment with a thorough review of what is known about the Bristol Bay watershed, its
fisheries and wildlife populations, and its Alaska Native cultures. We also reviewed information about
copper mining and publicly available information outlining proposed mine operations for the Pebble
deposit. The Pebble deposit has been the focus of much exploratory study and has received significant
attention from groups in and outside of Alaska. With the help of regional stakeholders, we developed a
set of conceptual models to show potential associations between salmon populations and the
environmental stressors that might reasonably result from large-scale mining. Then, following the
USEPA's ecological risk assessment framework, we analyzed the sources and exposures that would
occur and potential responses to those exposures. Finally, we characterized the risks to fish habitats,
salmon, and other fish populations, as well as the implications of those risks for the wildlife and Alaska
Native cultures that use them.
This is not an in-depth assessment of a specific mine, but rather an examination of potential impacts of
reasonably foreseeable mining activities in the Bristol Bay region, given the nature of the watershed's
mineral deposits and the requirements for successful mine development. The assessment analyzes mine
scenarios that reflect the expected characteristics of mine operation at the Pebble deposit. It is intended
to provide a baseline for understanding potential impacts of mine development, not just at the Pebble
deposit but throughout the Nushagak and Kvichak River watersheds. The mining of other existing
porphyry copper deposits in the region would be expected to include the same types of activities and
facilities evaluated in this assessment for the Pebble deposit (open pit mining and the creation of waste
rock piles and tailings storage facilities [TSFs]), and therefore would present potential risks similar to
those outlined in this assessment. However, because the region's other ore bodies are believed to be
much smaller than the Pebble deposit, those mines would likely be most similar to the smallest mine
scenario analyzed in this assessment (Pebble 0.25).
This assessment considers many but not all potential impacts associated with future large-scale mining
in the Bristol Bay watershed. Although the mine scenarios assume development of a deep-water port on
Cook Inlet to ship product concentrate elsewhere for smelting and refining, impacts of port development
and operation are not assessed. The assessment does not evaluate impacts of the one or more large-
capacity electricity-generating power plants that would be required to power the mine and the port. We
recognize that large-scale mine development would induce the development of additional support
services for mine employees and their families, vacation homes and other recreational facilities, and
transportation infrastructure beyond the main corridor (i.e., airports, docks, and roads). The assessment
describes but does not evaluate the effects of induced development resulting from large-scale mining in
the region. Direct effects of mining on Alaska Natives and wildlife are not assessed. The assessment also
Bristol Bay Assessment ES 4 January 2014
-------
does not include a cost-benefit analysis and does not compare mining to other ongoing activities such as
commercial fishing.
Ecological Resources
The Bristol Bay watershed provides habitat for numerous animal species, including at least 29 fish
species, more than 40 terrestrial mammal species, and more than 190 bird species. Many of these
species are essential to the structure and function of the region's ecosystems and current economies.
The Bristol Bay watershed supports several wilderness compatible and sustainable economic sectors,
such as commercial, sport, and subsistence fishing; sport and subsistence hunting; and non-consumptive
recreation. Considering all these sectors, the Bristol Bay watershed's ecological resources generated
nearly $480 million in direct economic expenditures and sales in 2009 and provided employment for
over 14,000 full- and part-time workers.
Chief among these ecological resources are world-class commercial and sport fisheries for Pacific
salmon and other salmonids. The region's commercial salmon fishery generates the largest component
of economic activity. The watershed supports production of all five species of Pacific salmon found in
North America: sockeye [Oncorhynchus nerka], coho (0. kisutch], Chinook (0. tshawytscha), chum (0.
ketd), and pink (0. gorbuschd) (Figure ES-2). These fishes are anadromous, meaning that they hatch and
rear in freshwater systems, migrate to sea to grow to adult size, and return to freshwater systems to
spawn and die. Because no hatchery fish are raised or released in the watershed, Bristol Bay's salmon
populations are entirely wild.
The most abundant salmon species in the Bristol Bay watershed is sockeye salmon. The watershed
supports the largest sockeye salmon fishery in the world, with approximately 46% of the average global
abundance of wild sockeye salmon (Figure ES-3). Between 1990 and 2009, the annual average inshore
run of sockeye salmon in Bristol Bay was approximately 37.5 million fish. Annual commercial harvest of
sockeye over this same period averaged 25.7 million fish. Approximately half of Bristol Bay's sockeye
salmon production is from the Nushagak and Kvichak River watersheds, the main area of focus for this
assessment (Figure ES-3).
Chinook salmon are also abundant in the region. Chinook returns to the Nushagak River are consistently
greater than 100,000 fish per year and have exceeded 200,000 fish in 11 years between 1966 and 2010,
frequently placing Nushagak River Chinook runs at or near the world's largest. This is noteworthy given
the Nushagak River's small watershed area compared to other Chinook-producing rivers such as the
Yukon River, which spans Alaska and much of northwestern Canada, and the Kuskokwim River in
southwestern Alaska, just north of Bristol Bay.
Bristol Bay Assessment FS 5 January 2014
-------
Figure ES-2. Reported salmon (sockeye, Chinook, coho, pink, and chum combined) distribution in
the South and North Fork Koktuli River and Upper Talarik Creek watersheds. Designation of species
spawning, rearing, and presence is based on the Anadromous Waters Catalog (Johnson and Blanche
2012). Life-stage-specific reach designations are believed to be underestimates, given the challenges
inherent in surveying all streams that may support life-stage use throughout the year.
ORTH FORK KOKTULI
SOUTH FORK KOKTULI
Present
Spawning
Rearing
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Mine Scenario Watersheds
Watershed Boundary
Bristol Bay Assessment
ES-6
January 2014
-------
Figure ES-3. Proportion of total sockeye salmon run sizes by (A) region and (B) watershed in the
Bristol Bay region. Values are averages from (A) 1956 to 2005 from Ruggerone et al. 2010 and (B)
1956 to 2010 from Baker pers. comm.
• Bristol Bay
• Russia Mainland & Islands
• West Kamchatka
• East Kamchatka
• Western Alaska (excluding Bristol Bay)
• South Alaska Peninsula
D Kodiak
DCook Inlet
D Prince William Sound
D Southeast Alaska
D North British Columbia
D South British Columbia, Washington & Oregon
• Togiak
• Nushagak
DKvichak
DNaknek
DEgegik
DUshagik
Bristol Bay Assessment
ES-7
January 2014
-------
The Bristol Bay watershed also supports populations of non-salmon fishes that typically (but not
always) remain in the watershed's freshwater habitats throughout their life cycles. The region contains
highly productive waters for sport and subsistence fish species, including rainbow trout (0. mykiss),
Dolly Varden (Salvelinus malma), Arctic char (S. alpinus), lake trout (S. namaycush), Arctic grayling
[Thymallus arcticus), northern pike [Esox lucius), and humpback whitefish [Coregonus pidschiari). These
fishes occupy a variety of habitats in the watershed, from headwater streams to wetlands to large rivers
and lakes. The Bristol Bay region is especially renowned for the size and abundance of its rainbow trout:
between 2003 and 2007, an estimated 183,000 rainbow trout were caught in the Bristol Bay
Management Area.
The exceptional quality of the Bristol Bay watershed's fish populations can be attributed to several
factors, the most important of which is the watershed's high-quality, diverse aquatic habitats unaltered
by human-engineered structures and flow management controls. Surface and subsurface waters are
highly connected, enabling hydrologic and biochemical connectivity between wetlands, ponds, streams,
and rivers and thereby increasing the diversity and stability of habitats able to support fish. These
factors all contribute to making the Bristol Bay watershed a highly productive system. High aquatic
habitat diversity also supports the high genetic diversity offish populations. This diversity in genetics,
life history, and habitat acts to reduce year-to-year variability in total production and increase overall
stability of the fishery.
The return of spawning salmon from the Pacific Ocean brings marine-derived nutrients into the
watershed and fuels both aquatic and terrestrial foodwebs. Thus, the condition of Bristol Bay's
terrestrial ecosystems is intimately linked to the condition of salmon populations, as well as to almost
totally undisturbed terrestrial habitats. The watershed continues to support large carnivores such as
brown bears (Ursus arctos), bald eagles (Haliaeetus leucocephalus), and gray wolves (Cam's lupus);
ungulates such as moose [Alces alcesgigas) and caribou [Rangifer tarandus granti); and numerous
waterfowl and small mammal species. Brown bears are abundant in the Nushagak and Kvichak River
watersheds. Moose also are abundant, particularly in the Nushagak River watershed where felt-leaf
willow, a preferred forage species, is plentiful. The Nushagak and Kvichak River watersheds are used by
caribou, primarily the Mulchatna caribou herd. This herd ranges widely through these watersheds, but
also spends considerable time in other watersheds.
Alaska Native Cultures
The predominant Alaska Native cultures present in the Nushagak and Kvichak River watersheds—the
Yup'ik and Dena'ina—are two of the last intact, sustainable, salmon-based cultures in the world. In
contrast, other Pacific Northwest salmon-based cultures are severely threatened by development,
degraded natural resources, and declining salmon resources. Salmon are integral to these cultures'
entire way of life via the provision of subsistence food and subsistence-based livelihoods, and are an
important foundation for their language, spirituality, and social structure. The cultures have a strong
connection to the landscape and its resources. In the Bristol Bay watershed, this connection has been
Bristol Bay Assessment F58 January 2014
-------
maintained for at least 4,000 years and is in part both due to and responsible for the continued
undisturbed condition of the region's landscape and biological resources. The respect and importance
given salmon and other wildlife, along with traditional knowledge of the environment, have produced a
sustainable subsistence-based economy. This subsistence-based way of life is a key element of Alaska
Native identity and serves a wide range of economic, social, and cultural functions in Yup'ik and
Dena'ina societies.
There are 31 Alaska Native villages in the wider Bristol Bay region, 25 of which are located in the Bristol
Bay watershed. Fourteen of these communities are within the Nushagak and Kvichak River watersheds,
with a total population of 4,337 in 2010. Thirteen of these 14 communities have federally recognized
tribal governments and a majority Alaska Native population. Many of the non-Alaska Native residents in
the watersheds have developed cultural ties to the region and they also practice subsistence. Virtually
every household in the watersheds uses subsistence resources. In the Bristol Bay region, salmon
constitute approximately 52% of the subsistence harvest; for some communities this proportion is
substantially higher.
The subsistence-based way of life in many Alaska Native villages is augmented with activities that
support cash economy transactions, including commercial fishing. Alaska Native villages, in partnership
with Alaska Native corporations and other business interests, are considering a variety of economic
development opportunities. Some Alaska Native villages have decided that large-scale mining is not the
course they would like to pursue, whereas a few others are seriously considering this opportunity. All
are concerned with the long-term sustainability of their communities.
Geological Resources
In addition to significant and valuable ecological resources, the Nushagak and Kvichak River watersheds
contain considerable mineral resources. The potential for large-scale mine development in the region is
greatest for copper deposits and, to a lesser extent, for intrusion-related gold deposits. Because these
deposits are low-grade—meaning that they contain relatively small amounts of metals relative to the
amount of ore—mining will be economic only if conducted over large areas and will necessarily produce
large amounts of waste material.
The largest known and most explored deposit is the Pebble deposit. If fully mined, the claim holder
estimates that the Pebble deposit would produce more than 11 billion tons of ore, which would make it
the largest mine of its type in North America. A mine at the Pebble deposit could ultimately generate
revenues between $300 billion to $500 billion over the life of the mine, as well as provide more than
2,000 jobs during mine construction and more than 1,000 jobs during mine operation.
Although the Pebble deposit represents the most imminent site of mine development, other mineral
deposits with potentially significant resources exist in the Nushagak and Kvichak River watersheds. Ten
specific claims with more than minimal recent exploration (in addition to the Pebble deposit claim) have
Bristol Bay Assessment F59 January 2014
-------
been filed for copper deposits. Most of these claims are near the Pebble deposit. The potential impacts of
large-scale mining considered in this assessment are generally applicable to these other sites.
Mine Scenarios
Like all risk assessments, this assessment is based on scenarios that define a set of possible future
activities and outcomes. To assess mining-related stressors that would affect ecological resources in the
watershed, we developed realistic mine scenarios that include a range of mine sizes and operating
conditions. These mine scenarios are based on the Pebble deposit because it is the best-characterized
mineral resource and the most likely to be developed in the near term. The mine scenarios draw on
preliminary plans developed for Northern Dynasty Minerals, consultation with experts, and baseline
data collected by the Pebble Limited Partnership to characterize the mine site, mine activities, and the
surrounding environment. The exact details of any future mine plan for the Pebble deposit or for other
deposits in the watershed will differ from our mine scenarios. However, our scenarios reflect the general
characteristics of mineral deposits in the watershed, modern conventional mining technologies and
practices, the scale of mining activity required for economic development of the resource, and the
infrastructure needed to support large-scale mining. Therefore, the mine scenarios evaluated in this
assessment realistically represent the type of development plan that would be anticipated for a
porphyry copper deposit in the Bristol Bay watershed. Uncertainties associated with the mine scenarios
are discussed later in this executive summary.
The three mine scenarios evaluated in the assessment represent different stages of mining at the Pebble
deposit, based on the amount of ore processed: Pebble 0.25 (approximately 0.25 billion tons [0.23
billion metric tons] of ore over 20 years), Pebble 2.0 (approximately 2.0 billion tons [1.8 billion metric
tons] of ore over 25 years), and Pebble 6.5 (approximately 6.5 billion tons [5.9 billion metric tons] of ore
over 78 years). The major parameters of the three mine scenarios are presented in Table ES-1, and their
layouts are presented in Figure ES-4. The major components of each mine would be an open mine pit,
waste rock piles, and one or more TSFs. Other significant features include plant and ancillary facilities
(e.g., a water collection and treatment system, an ore-processing facility, and other facilities associated
with mine operations) and the groundwater drawdown zone (the area over which the water table is
lowered due to dewatering of the mine pit). An underground extension of the mine, which could
increase the size of the mine to 11 billion tons of ore, is not included in this assessment.
Each of these mine scenarios includes a 138-km (86-mile) transportation corridor; 113 km (70 miles) of
the corridor would fall within the Kvichak River watershed (Figure ES-5). This corridor would include a
gravel-surfaced road and four pipelines (one each for product concentrate, return water, diesel fuel, and
natural gas).
The assessment considers risks from routine operation of a mine designed using modern conventional
design, practices, and mitigation technologies, assuming no significant human or engineering failures.
The assessment also considers various types of failures that have occurred during the operation of other
Bristol Bay Assessment ES 10 January 2014
-------
mines and that could occur in this case, including failures of a wastewater treatment plant, a tailings
dam, pipelines, and culverts.
Table ES-1. Mine scenario parameters.
Parameter
Amount of ore mined (billion metric tons)
Approximate duration of mining (years)
Ore processing rate (metric tons/day)
Mine Scenario
Pebble 0.25
0.23
20
31,100
Pebble 2.0
1.8
25
198,000
Pebble 6.5
5.9
78
208,000
Mine Pit
Surface area (km2)
Depth (km)
1.5
0.30
5.5
0.76
17.8
1.24
Waste Rock Pile
Surface area (km2)
PAG waste rock (million metric tons)
NAG waste rock (million metric tons)
2.3
86
320
13.0
580
2,200
22.6
4,700
11,000
TSFl"
Capacity, dry weight (billion metric tons)
Surface area, exterior (km2)
Maximum dam height (m)
0.25
6.8
92
1.97
16.1
209
1.97
16.1
209
TSF 2"
Capacity, dry weight (billion metric tons)
Surface area, exterior (km2)
NA
NA
NA
NA
3.69
22.7
TSF3a
Capacity, dry weight (billion metric tons)
Surface area, exterior (km2)
Total TSF surface area, exterior (km2)
NA
NA
6.8
NA
NA
16.1
0.96
9.82
48.6
Notes:
a Final value, when TSF is full.
PAG = potentially acid-generating; NAG = non-acid-generating; TSF = tailings storage facility; NA = not applicable.
Bristol Bay Assessment
ES-11
January 2014
-------
Figure ES-4. Major mine components for the three scenarios evaluated in the assessment. Pebble
0.25 represents 0.25 billion tons of ore; Pebble 2.0 represents 2.0 billion tons of ore; Pebble 6.5
represents 6.5 billion tons of ore. Each mine footprint includes the mine components shown here, as
well as the drawdown zone and the area covered by plant and ancillary facilities. Light blue areas
indicate streams and rivers from the National Hydrography Dataset (USGS 2012) and lakes and
ponds from the National Wetlands Inventory (USFWS 2012); dark blue areas indicate wetlands from
the National Wetlands Inventory (USFWS 2012).
B*"> »T> *Sif". * < ^
Sgfcite^^F;' «*
oEr^'-V .?> -
..y
«(gfc/ ,, '+!^xWr*#3ijir
V~. -VX"' . ^-f'^'lJP' £r - ;'-~' "I**
C'tiP' •*,•• f^^v.ii^r^.-j- ---•
iVr^ • '^; V>,i£fV'- s- _J
[Auf v; •.
p'-'Mf dk i ^
wfr VS.- c -^ ^»,
*« ^;«i ^Sr«s
ssfi
».-v'^
• - A*» %>^
•^ V"\- ^
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Bristol Bay Assessment
ES-12
January 2014
-------
Figure ES-5. The transportation corridor area, comprising 32 subwatersheds in the Kvichak River watershed that drain to Iliamna Lake
Subwatersheds are defined by 12-digit hydrologic unit codes according to the National Hydrography Dataset (USGS 2012).
« Approximate Pebble Deposit Location
= = = = Transportation Corridor (Outside Assessment Area)
Transportation Corridor
Existing Roads
Transportation Corridor Area
Subwatersheds within Area
Bristol Bay Assessment
ES-13
January 2014
-------
Risks to Salmon and Other Fishes
Based on the mine scenarios, the assessment defines mining-related stressors that would affect the
Bristol Bay watershed's fish and consequently affect wildlife and human welfare. The scenarios include
both routine operations (Tables ES-2 and ES-3) and several potential failure scenarios (Table ES-4).
Mine Footprint
Effects on fish resulting from habitat loss and modification would occur directly in the area of mine
activity and indirectly downstream because of habitat destruction. These habitat loss estimates are
believed to be low due to incomplete delineation of streams, wetlands, and salmon distribution across
the region. However, it is possible that careful siting of mine facilities could reduce habitat losses to
some degree.
• Due to the mine footprint (the area covered by the mine pit, waste rock piles, TSFs, groundwater
drawdown zone, and plant and ancillary facilities), 38, 89, and 151 km (24, 55, and 94 miles) of
streams would be lost—that is, eliminated, blocked, or dewatered—in the Pebble 0.25, 2.0, and 6.5
scenarios, respectively (Table ES-2). This translates to losses of 8, 22, and 36 km (5,14, and 22
miles) of streams known to provide spawning or rearing habitats for coho salmon, sockeye salmon,
Chinook salmon, and Dolly Varden (Table ES-2, Figure ES-6).
• Altered streamflow due to retention and discharge of water used in mine operations, ore processing
and transport, and other mine activities would reduce the amount and quality offish habitat.
Streamflow alterations exceeding 20% would adversely affect habitat in an additional 15, 27, and
53 km (9.3,17, and 33 miles) of streams in the Pebble 0.25, 2.0, and 6.5 scenarios, respectively
(Table ES-2), reducing production of sockeye salmon, coho salmon, Chinook salmon, rainbow trout,
and Dolly Varden. Reduced streamflows would also result in the loss or alteration of an
unquantifiable area of riparian floodplain wetland habitat due to loss of hydrologic connectivity
with streams.
• Off-channel habitats for salmon and other fishes would be reduced due to losses of 4.5,12, and 18
km2 (1,200, 3,000 and 4,900 acres) of wetlands and 0.41, 0.93, and 1.8 km2 (100, 230, and 450
acres) of ponds and lakes to the mine footprints in the Pebble 0.25, 2.0, and 6.5 scenarios,
respectively (Figure ES-6). These losses would reduce availability of and access to hydraulically and
thermally diverse habitats that provide enhanced foraging opportunities and important rearing
habitats for juvenile salmon.
• Indirect effects of stream and wetland losses would include reductions in the quality of
downstream habitat for coho salmon, sockeye salmon, Chinook salmon, rainbow trout, and Dolly
Varden. Although these indirect effects cannot be quantified, such effects would be expected to
diminish fish production downstream of the mine site because fish depend on these habitats.
Indirect effects would be caused by the following alterations.
Bristol Bay Assessment ES 14 January 2014
-------
o Reduced food resources would result from the loss of organic material and drifting
invertebrates from streams and streamside wetlands lost to the mine footprint.
o The balance of surface water and groundwater inputs to downstream reaches would shift,
potentially reducing winter fish habitat and making streams less suitable for spawning and
rearing.
o Seasonal temperatures could be altered by water treatment and reduced groundwater
flowpaths, making streams less suitable for salmonids.
Water Quality
Leakage during Routine Operations
Water from the mine site would enter streams through wastewater treatment plant discharges and in
uncollected runoff and leakage of leachates from the waste rock piles and TSFs. Wastewater treatment is
assumed to meet all state standards and national criteria, or equivalent benchmarks for chemicals that
have no criteria. However, water quality would be diminished by uncollected leakage of tailings and
waste rock leachates from the containment system, which would occur during routine operations. Test
leachates from the tailings and non-acid-generating waste rocks are mildly toxic. They would require an
approximately two-fold dilution to achieve water quality criteria for copper, but are not estimated to be
toxic to salmonids. Waste rocks associated with the ore body are acid-forming with high copper
concentrations in test leachates, and would require 2,900- to 52,000-fold dilution to achieve water
quality criteria. Several metals could be sufficiently elevated to contribute to toxicity, but copper is the
dominant toxicant.
Uncollected leachates from waste rock piles and TSFs would elevate instream copper levels and cause
direct effects on salmonids ranging from aversion and avoidance of the contaminated habitat to rapidly
induced death of many or all fish (Table ES-2). Avoidance of streams by salmonids would occur in 24
and 34 to 57 km (15 and 21 to 35 miles) of streams in the Pebble 2.0 and Pebble 6.5 scenarios,
respectively. Rapidly induced death of many or all fish would occur in 12 km (7.4 miles) of streams in
the Pebble 6.5 scenario. Copper would cause death or reduced reproduction of aquatic invertebrates in
21, 40 to 62, and 60 to 82 km (13, 25 to 38, and 37 to 51 miles) of streams in the Pebble 0.25, 2.0, and
6.5 scenarios, respectively. These invertebrates are the primary food source for juvenile salmon and all
life stages of other salmonids, so reduced invertebrate productivity would be expected to reduce fish
productivity. These results are sensitive to the assumed efficiency of the leachate capture system, and a
more efficient system could be devised. However, greater than 99% capture efficiency would be
required to prevent exceedance of the copper criteria for the South Fork Koktuli River in the Pebble 6.5
scenario, which would require technologies beyond those specified in our scenarios or identified in the
most recent preliminary mine plan.
Bristol Bay Assessment ES 15 January 2014
-------
Wastewater Treatment Plant Failure
Based on a review of historical and currently operating mines, some failure of water collection and
treatment systems would be expected to occur during operation or post-closure periods. A variety of
water collection and treatment failures are possible, ranging from operational failures that result in
short-term releases of untreated or partially treated leachates to long-term failures to operate water
collection and treatment systems in perpetuity. A reasonable but severe failure scenario would involve a
complete loss of water treatment and release of average untreated wastewater flows into average
dilution flows. In that failure scenario, copper concentrations would be sufficient to cause direct effects
on salmonids in 27, 64 to 87, and 74 to 97 km (17, 40 to 54, and 46 to 60 miles) of streams in the Pebble
0.25, 2.0, and 6.5 scenarios, respectively. Aquatic invertebrates would be killed or their reproduction
reduced in 78 to 100 km (48 to 62 miles) of streams in all three scenarios. In the Pebble 2.0 and 6.5
scenarios, a fish kill would occur rapidly in 3.8 and 31 km (2.4 and 19 miles) of streams, respectively,
following treatment failure.
Spillway Release
In the event of TSF overfilling, supernatant water would be released via a spillway. If the water was
equivalent to the test tailings supernatant, 2.6 km (1.6 miles) of streams would be avoided by fish and
3.4 to 23 km (2.1 to 14 miles) of streams would be toxic to invertebrates, independent of other sources.
Transportation Corridor
Construction and Routine Operation
In the Kvichak River watershed, the transportation corridor would cross approximately 64 streams and
rivers. Of those, 55 are known or likely to support migrating and resident salmonids, including 20
streams designated as anadromous waters at the location of the crossing (Figure ES-7). The corridor
would run near Iliamna Lake and cross multiple tributary streams near their confluences with the lake.
These habitats are important spawning areas for sockeye salmon, putting sockeye particularly at risk
from the road. Diminished habitat quality in streams and wetlands below road crossings would result
primarily from altered streamflow, runoff of road salts, and siltation of habitat for salmon spawning and
rearing and invertebrate prey production (Tables ES-2 and ES-3).
Culvert Failure
Culverts commonly fail to allow free passage offish. They can become blocked by debris or ice that may
not stop water flow but that create a barrier to fish movement. Fish passage also may be blocked or
inhibited by erosion below a culvert that "perches" the culvert and creates a waterfall, by shallow water
caused by a wide culvert and periodic low streamflows, or by excessively high gradients. If blockages
occurred during adult salmon immigration or juvenile salmon emigration and were not cleared for
several days, production of a year-class (i.e., fish spawned in the same year) would be lost from or
diminished in the stream above the culvert.
Bristol Bay Assessment ES 16 January 2014
-------
Culverts can also fail to convey water due to landslides or, more commonly, floods that wash out
undersized or improperly installed culverts. In such failures, the stream would be temporarily
impassible to fish until the culvert is repaired or until erosion re-establishes the channel. If the failure
occurs during a critical period in salmon migration, effects would be the same as with a debris blockage
(i.e., a lost or diminished year-class).
Culvert failures also would result in the downstream transport and deposition of silt, which could cause
returning salmon to avoid a stream if they arrived during or immediately following the failure.
Deposition of silt would smother salmon eggs and alevins if they were present, and would degrade
downstream habitat for salmonids and the invertebrates that they eat.
Blockages of culverts could persist for as long as the intervals between culvert inspections. We assume
that the transportation corridor would be inspected daily and maintained during mine operation. The
level of surveillance along the corridor can be expected to affect the frequency of culvert failure
detection. Driving inspections would likely identify a single erosional failure of a culvert that damaged
the road, or a debris blockage sufficient to cause water to pool above the road. However, long-term fixes
may not be possible until conditions are suitable for culvert replacement, and these fixes may not fully
address fish passage, which may be reduced or blocked for longer periods. Extended blockage of
migration would be less likely if daily road inspections included stops to inspect each end of each
culvert.
After mine operations cease, the road would likely be maintained less carefully by the operator or may
be transferred to a government entity that would be expected to employ a more conventional inspection
and maintenance schedule. In either case, the proportion of impassable culverts at any one time would
be expected to revert to levels found in published surveys of public roads (mean of 48% [range of 30 to
61%] of culverts that had failed and not been repaired when surveyed). Of the approximately 45
culverts that would be required, 36 would be on streams that are believed to support salmonids. Hence,
11 to 22 streams would be expected to have impeded passage of salmon, rainbow trout, or Dolly Varden
for an indefinite period of time, and some proportion of those streams would have degraded
downstream habitat resulting from sedimentation following washout of the road.
Truck Accidents
Trucks would carry ore processing chemicals to the mine site and molybdenum product concentrate to
the port. Truck accident records indicate that truck accidents near streams are likely over the long
period of mine operation. These accidents could release sodium ethyl xanthate, cyanide, other process
chemicals, or molybdenum product concentrate to streams or wetlands, resulting in toxic effects on
invertebrates and fish. However, the risk of spills could be mitigated by using impact-resistant
containers.
Tailings Dam Failure
Tailings are the waste materials produced during ore processing. In our scenarios, these wastes would
be stored in TSFs consisting of tailings dams and impoundments. The probability of a tailings dam
Bristol Bay Assessment ES 17 January 2014
-------
failure increases with the number of dams. The Pebble 0.25 scenario would include one TSF with a
single dam, the Pebble 2.0 scenario would include one TSF with three dams, and the Pebble 6.5 scenario
would include three TSFs with a total of eight dams. Because their removal is not feasible, the TSFs and
their component dams would be in place for hundreds to thousands of years, long beyond the life of the
mine. Available reports from the Pebble Limited Partnership suggest a tailings dam as high as 209 m
(685 feet) at TSF 1 (Figure ES-8). We evaluated two potential dam failures at TSF 1 in this assessment:
one at a volume approximating the complete Pebble 0.25 scenario (92-m dam height) and one at a
volume approximating the complete Pebble 2.0 scenario (209-m dam height). In both cases we assumed
20% of the tailings would be released, a conservative estimate that is well within the range of historical
tailings dam failures. Failures of the TSF 2 and TSF 3 tailings dams were not analyzed but would be
expected to be similar in terms of types of effects.
Table ES-2. Summary of estimated stream lengths potentially affected in the three mine size
scenarios, assuming routine operations.
Effect
Eliminated, blocked, or dewatered
Eliminated, blocked, or dewatered— anadromous
>20% streamflow alteration3
Direct toxicity to fish3
Direct toxicity to invertebrates3
Downstream of transportation corridor
Stream Length Affected (km)
Pebble 0.25
38
8
15
0
21
Pebble 2.0
89
22
27
24
40-62
Pebble 6.5
151
36
53
34-57
60-82
272
Notes:
8 Stream reaches with streamflow alterations partially overlap those with toxicity.
Table ES-3. Summary of estimated wetland, pond, and lake area potentially affected in the three
mine size scenarios, assuming routine operations.
Effect
Lost to the mine footprint
Lost to reduced streamflow below mine footprint
Filled by roadbed
Influenced by the road (within 200 m)
Wetland, Pond, and Lake Area Affected (km2)
Pebble 0.25
4.9
Pebble 2.0
13
Unquantified
Pebble 6.5
20
0.11
4.7
Bristol Bay Assessment
ES-18
January 2014
-------
Table ES-4. Probabilities and consequences of potential failures in the mine scenarios.
Failure Type
Tailings dam
Product concentrate pipeline
Concentrate spill into a stream
Concentrate spill into a wetland
Return water pipeline spill
Diesel pipeline spill
Culvert, operation
Culvert post-operation
Truck accidents
Water collection and treatment,
operation
Tailings storage facility spillway
release
Water collection and treatment,
managed post-closure
Water collection and treatment,
after site abandonment
Probability3
4 x 1O4 to 4 x 1O6 per dam-year =
recurrence frequency of 2,500 to
250,000 yearsb
10'3 per km-year = 95% chance
per pipeline in 25 years
1.5 x 1Q-2 per year = 1 stream-
contaminating spill in 78 years
2.6 x lO-2 per year = 2 wetland-
contaminating spills in 78 years
Same as product concentrate
pipeline
Same as product concentrate
pipeline
Low
3 x 10-1 to ~6 x 10-1 per culvert;
instantaneous = 11 to 22 culverts
1.9 x 10-7 spills per mile of travel =
4 accidents in 25 years and
2 near-stream spills in 78 years
0.93 = proportion of recent U.S.
porphyry copper mines with
reportable water collection and
treatment failures
No data, but spills are known to
occur and are sufficiently frequent
to justify routine spillway
construction
Somewhat higher than operation
Certain, by definition
Consequences
More than 29 km of salmonid stream would be
destroyed or degraded for decades.
Most failures would occur between stream or
wetland crossings and might have little effect on fish.
Fish and invertebrates would experience acute
exposure to toxic water and chronic exposure to toxic
sediment in a stream and potentially extending to
Iliamna Lake.
Invertebrates and potentially fish would experience
acute exposure to toxic water and chronic exposure
to toxic sediment in a pond or other wetland.
Fish and invertebrates would experience acute
exposure to toxic water if return water spilled to a
stream or wetland.
Acute toxicity would reduce the abundance and
diversity of invertebrates and possibly cause a fish
kill if diesel spilled to a stream or wetland.
Frequent inspections and regular maintenance would
result in few impassable culverts, but for those few,
blockage of migration could persist for a migration
period, particularly for juvenile fish.
In surveys of road culverts, 30 to 61% are
impassable to fish at any one time. This would result
in 11 to 22 salmonid streams blocked at any one
time. In 10 to 19 of the 32 culverted streams with
restricted upstream habitat, salmon spawning may
fail or be reduced and the streams would likely not
be able to support long-term populations of resident
species.
Accidents that spill processing chemicals into a
stream or wetland could cause a fish kill. A spill of
molybdenum concentrate may also be toxic.
Water collection and treatment failures could result
in exceedance of standards potentially including
death offish and invertebrates. However, these
failures would not necessarily be as severe or
extensive as estimated in the failure scenario, which
would result in toxic effects from copper in more than
60 km of stream habitat.
Spilled supernatant from the tailings storage facility
could result in toxicity to invertebrates and fish
avoidance for the duration of the event.
Post-closure collection and treatment failures are
very likely to result in release of untreated or
incompletely treated leachatesfor days to months,
but the water would be less toxic due to elimination
of potentially acid-generating waste rock.
When water is no longer managed, untreated
leachates would flow to the streams. However, the
water may be less toxic.
a Because of differences in derivation, the probabilities are not directly comparable.
b Based on expected state safety requirements. Observed failure rates for earthen dams are higher (about 5 x 10 4 per year or a recurrence
frequency of 2,000 years).
Bristol Bay Assessment
ES-19
January 2014
-------
Figure ES-6. Streams and wetlands lost (eliminated, blocked, or dewatered) in the Pebble 6.5
scenario. Light blue areas indicate streams and rivers from the National Hydrography Dataset (USGS
2012) and lakes and ponds from the National Wetlands Inventory (USFWS 2012); dark blue areas
indicate wetlands from the National Wetlands Inventory (USFWS 2012).
<* <
r^V^?JB^ ?'
-ivirtMr :
«T
w f ^VMHF**^'
K v ^5%>:
wJp** ^; • ^ '*rP**ir . y^fcr^
-> *3W ^^ •* efnUr^^'*'*. -,f ^^i ^J*^
" '« i< mr '• * *S ^-V^
1 - • /<>^ - ',- •/ -.- A J_A "
2 .#*
%•-•
r^1
,b
-S~
rM
NUSHAGAK
l*^0 S^l ^
012 | ^_ 4J ^^,
] Kilometers
] Miles
T
rt
/"
KVICHAK^-
Pebble 6.5 Components
Drawdown Zone
Eliminated, Blocked, or Dewatered Streams
Eliminated, Blocked, or Dewatered
Lakes and Ponds
Eliminated, Blocked, or Dewatered Wetlands
Bristol Bay Assessment
ES-20
January 2014
-------
Figure ES-7. Reported salmon, Dolly Varden, and rainbow trout distribution along the transportation corridor. Salmon presence data are
from the Anadromous Waters Catalog (Johnson and Blanche 2012); Dolly Varden and rainbow trout presence data are from the Alaska
although these points are not indicated on this map.
10
Kilometers
5 10
Miles
Note: Sampling intensity is greatly reduced away from the Pebble deposit area
Streams without data points may not have been surveyed: thus, it is unknown
whether or not they contain these species..
Approximate Location of Pebble Deposit
Transportation Corridor (Outside Assessment Area)
Dolly Varden
Rainbow Trout
Transportation Corridor
Transportation Corridor Area
Subwatersheds within Area
Salmon
Bristol Bay Assessment
ES-21
January 2014
-------
Figure ES-8. Height of the dam at tailings storage facility (TSF) 1 in the Pebble 2.0 and Pebble 6.5
scenarios, relative to U.S. landmarks.
260-r
240-
220-
200
180
160
g 140
s 120
100
80
60
40
20
0
Transamerica Building - 260 Meters
Tailings Dam TSF 1-209 Meters
Gateway Arch -192 Meters
Washington Monument -169 Meters
Tailings Reservoir
The range of estimated dam failure probabilities is wide, reflecting the great uncertainty concerning
such failures. The most straightforward method of estimating the annual probability of a tailings dam
failure is to use the historical failure rate of similar dams. Three reviews of tailings dam failures
produced an average rate of approximately 1 failure per 2,000 dam-years, or 5 x 10~4 failures per dam-
year. Strictly speaking, these frequencies are properties that apply to a group of dams. However, by
extension, if there is one dam and it is typical of the population, it would be expected to fail, on average,
within a 2,000-year period. This does not mean it is expected to fail 2,000 years after it is built. Rather, it
indicates that, after 2,000 years have passed, it is more likely than not that the dam would have failed
and that expected failure could occur any year in that 2,000-year window with an average annual
probability of 0.0005.
The argument against this method is that the record of past failures does not fully reflect current
engineering practice. Some studies suggest that improved design, construction, and monitoring
practices can reduce the failure rate by an order of magnitude or more, resulting in an estimated failure
probability within the range assumed here (Table ES-4). The State of Alaska's guidelines suggest that an
applicant follow accepted industry design practices such as those provided by the U.S. Army Corps of
Engineers (USAGE), the Federal Energy Regulatory Commission (FERC), and other agencies. Based on
safety factors in USAGE and FERC guidance, we estimate that the probability of failure for all causes
requires a minimum factor of safety of 1.5 against slope instability for the loading condition
corresponding to steady seepage from the filled storage facility. An assessment of the correlation of dam
Bristol Bay Assessment
ES-22
January 2014
-------
failure probabilities with slope instability safety factors suggests an annual probability of failure of 1 in
250,000 per year for facilities designed, built, and operated with state-of-the-practice engineering
(Category I facilities) and 1 in 2,500 per year for facilities designed, built, and operated using standard
engineering practice (Category II facilities). The advantage of this approach is that it addresses current
regulatory guidelines and engineering practices. The disadvantage is that we do not know whether
standard practice or state-of-the practice dams will perform as expected, particularly given the potential
dam heights and subarctic conditions in these scenarios.
Failure of the dam at TSF 1 (the TSF included in all three mine scenarios) would result in the release of a
flood of tailings slurry into the North Fork Koktuli River. This flood would scour the valley and deposit
many meters of tailings fines in a sediment wedge across the entire valley near the TSF dam, with lesser
quantities of fines deposited as far as the North Fork's confluence with the South Fork Koktuli River. The
North Fork Koktuli River currently supports spawning and rearing populations of sockeye, coho, and
Chinook salmon; spawning populations of chum salmon; and rearing populations of Dolly Varden and
rainbow trout. The tailings slurry flood would continue down the mainstem Koktuli River with similar
effects, the extent of which cannot be estimated at this time due to model and data limitations.
The tailings dam failures evaluated in the assessment would be expected to have the following severe
direct and indirect effects on aquatic resources, particularly salmonids.
• It is expected that the North Fork Koktuli River below the TSF 1 dam and much of the mainstem
Koktuli River would not support salmonids in the short term (less than 10 years).
o In the tailings dam failure scenarios, spilled tailings would bury salmon habitat under meters of
fines along nearly the entire length of the North Fork Koktuli River valley downstream of the
dam (over 29 km or 18 miles in the Pebble 0.25 dam failure scenario), and beyond (in the Pebble
2.0 dam failure scenario).
o Deposited tailings would degrade habitat quality for both fish and the invertebrates they eat.
Based largely on their copper content, deposited tailings would be toxic to benthic
macroinvertebrates, but existing data concerning toxicity to fish are less clear.
o Deposited tailings would continue to erode from the North Fork Koktuli River and mainstem
Koktuli River valleys.
o Suspension and redeposition of tailings would be expected to cause serious habitat degradation
in the mainstem Koktuli River and downstream into the Mulchatna River; however, the extent of
these effects cannot be estimated at this time due to model and data limitations.
• The affected streams would provide low-quality spawning and rearing habitat for a period of
decades.
o Recovery of suitable substrates via mobilization and transport of tailings would take years to
decades, and would affect much of the watershed downstream of the failed dam.
Bristol Bay Assessment ES 23 January 2014
-------
o Ultimately, spring floods and stormflows would carry some of the tailings into the Nushagak
River.
o For some years, periods of high flow would be expected to suspend sufficient concentrations of
tailings to cause avoidance, reduced growth and fecundity, and even death of fish.
• Near-complete loss of North Fork Koktuli River fish populations downstream of the TSF and
additional fish population losses in the mainstem Koktuli, Nushagak, and Mulchatna Rivers would
be expected to result from these habitat losses.
o The Koktuli River watershed is an important producer of Chinook salmon. The Nushagak River
watershed, of which the Koktuli River watershed is a part, is the largest producer of Chinook
salmon in the Bristol Bay region, with annual runs averaging over 190,000 fish.
o A tailings spill could eliminate 29% or more of the Chinook salmon run in the Nushagak River
due to loss of the Koktuli River watershed population. An additional 10 to 20% could be lost due
to tailings deposited in the Mulchatna River and its tributaries.
o Sockeye are the most abundant salmon returning to the Nushagak River watershed, with annual
runs averaging more than 1.9 million fish. The proportion of sockeye and other salmon species
of Koktuli-Mulchatna origin is unknown.
o Similarly, the North Fork Koktuli River populations of rainbow trout and Dolly Varden would be
lost for years to decades if they could not successfully be maintained entirely in headwater
networks upstream of the affected zone. Quantitative estimates of these losses are not possible
given available information.
Effects would be qualitatively similar for both the Pebble 0.25 and Pebble 2.0 dam failures, although
effects from the Pebble 2.0 dam failure would extend farther and last longer. Failure of dams at the two
additional TSFs in the Pebble 6.5 scenario (TSF 2 and TSF 3) were not modeled, but would have similar
types of effects in the South Fork Koktuli River and downstream rivers.
Pipeline Failure
In the mine scenarios, the primary mine product would be a sand-like copper concentrate with traces of
other metals, which would be pumped via pipeline to a port on Cook Inlet. Water that carried the
concentrate would be returned to the mine site in a second pipeline. Based on the general record of
pipelines and further supported by the record of metal concentrate pipelines at existing mines, one
near-stream failure and two near-wetland failures of each of these pipelines would be expected to occur
over the life of the Pebble 6.5 scenario (approximately 78 years).
Failure of either the product or the return water pipeline would release water that is expected to be
highly toxic due to dissolved copper and possibly processing chemicals. Invertebrates and potentially
early fish life stages would be killed in the affected stream over a relatively brief period. If concentrate
spilled into a stream, it would settle and form highly toxic bed sediment based on its high copper content
and acid generation. The mean velocities of many streams crossed by the pipelines are sufficient to carry
Bristol Bay Assessment ES 24 January 2014
-------
the concentrate downstream to Iliamna Lake, but some would collect in low-velocity areas of the
receiving stream. If the spill occurred during low streamflows, dredging could recover some concentrate
but would cause physical damage to the stream. Concentrations in Iliamna Lake could not be predicted,
but near the pipeline route Iliamna Lake contains important sockeye salmon beach spawning areas that
would be exposed to a spill. Sockeye also spawn in the lower reaches of streams that could be directly
contaminated by a spill.
Based on petroleum pipeline failure rates, the diesel fuel pipeline also would be expected to spill near a
stream over the life of the Pebble 6.5 mine. Evidence from modeling the dissolved and dispersed oil
concentrations in streams, laboratory tests of diesel toxicity, and studies of actual spills in streams
indicates that a diesel spill at a stream crossing would be expected to immediately kill invertebrates and
likely fish as well. Remediation would be difficult but recovery would be expected to occur within 3
years. Failure of the natural gas pipeline would also be expected, but significant effects on fish would not
be expected.
Spills into wetlands that support fish would be expected to have greater toxic effects because
contaminants would be washed out slowly, if at all. However, retention of contaminants within the
wetland would make remediation by removal more practical.
Common Mode Failures
Multiple, simultaneous failures could occur due to a common event, such as a severe storm with heavy
precipitation (particularly precipitation that fell on spring snow cover) or a major earthquake. Over the
long period that tailings impoundments, a mine pit, and waste rock piles would be in place, the
likelihood of multiple extreme precipitation events, earthquakes, or combinations of these events
becomes much greater. Multiple events further increase the chances that facilities remaining in place
will weaken and eventually fail.
Such an event could cause multiple tailings dam failures that would spill tailings slurry into both the
South and North Fork Koktuli Rivers; road culvert washouts that would send sediments downstream
and potentially block fish passage; and pipeline failures that would release product slurry, return water,
or diesel fuel. The effects of each of these accidents individually would be the same as discussed
previously, but their co-occurrence would cause cumulative effects on salmonid populations and make
any remedial responses more difficult.
Fish-Mediated Risks to Wildlife
Although the effects of salmonid reductions on wildlife—that is, fish-mediated risks to wildlife—cannot
be quantified given available data, some reduction in wildlife would be expected in the mine scenarios.
Changes in the occurrence and abundance of salmon have the potential to change animal behavior and
reduce wildlife population abundances. The mine footprints would be expected to have local effects on
brown bears, wolves, bald eagles, and other wildlife that consume salmon, due to reduced salmon
Bristol Bay Assessment ES 25 January 2014
-------
abundance from habitat loss and degradation in or immediately downstream of the mine footprint. Any
of the accidents or failures evaluated would increase effects on salmon, which would further reduce the
abundance of their predators.
The abundance and production of wildlife also is enhanced by the marine-derived nutrients that salmon
carry upstream on their spawning migration. These nutrients are released into streams when the
salmon die, enhancing the production of other aquatic species that feed wildlife. Salmon predators
deposit these nutrients on the landscape, thereby fertilizing terrestrial vegetation that, in turn, provides
food for moose, caribou, and other wildlife. The loss of these nutrients due to a reduction in salmon
would be expected to reduce the production of riparian and upland species.
Fish-Mediated Risks to Alaska Native Cultures
Under routine operations with no major accidents or failures, the predicted loss and degradation of
salmonid habitat in the South and North Fork Koktuli Rivers and Upper Talarik Creek would be expected
to have some impact on Alaska Native cultures of the Nushagak and Kvichak River watersheds. Fishing
and hunting practices would be expected to change in direct response to the stream, wetland, and
terrestrial habitats lost to the mine footprints and the transportation corridor. It is also possible that
subsistence use of salmon resources would decline based on perceptions of reduced fish or water
quality resulting from mining.
The potential for significant effects on Alaska Native cultures is much greater from mine failures that
reduced or eliminated fish populations in affected areas, including areas significant distances
downstream from the mine. In the case of the tailings dam failures described in the assessment, the
significant loss of Chinook salmon populations would have severe consequences, especially for villages
in the Nushagak River watershed.
Any loss offish production from these failures would reduce the availability of these subsistence
resources to local Alaska Native villages, and the reduction of this highly nutritious food supply could
have negative consequences on human health. Because salmon-based subsistence is integral to Alaska
Native cultures, the effects of salmon losses go beyond the loss of food resources. If salmon quality or
quantity was (or was perceived to be) adversely affected, the nutritional, social, and spiritual health of
Alaska Natives would decline.
Cumulative Risks of Multiple Mines
This assessment has focused on the effects that a single large mine at the Pebble deposit would have on
salmon and other resources in the Nushagak and Kvichak River watersheds, including the cumulative
effects of multiple stressors associated with that mine. However, multiple mines and their associated
infrastructure may be developed in these watersheds. Each mine would pose risks similar to those
identified in the mine scenarios. Estimates of the stream and wetland habitats lost would differ across
different deposits, based on the size and location of mine operations within the watersheds. Individually,
Bristol Bay Assessment ES 26 January 2014
-------
each mine footprint would eliminate some amount offish-supporting habitat and, should operator or
engineering failures occur, affect fish habitats well beyond the mine footprint.
We considered development of mines at the Pebble South/PEB, Big Chunk South, Big Chunk North,
Groundhog, AUDN/Iliamna, and Humble claims in the Nushagak and Kvichak River watersheds. These
sites were chosen because all contain copper deposits that have generated exploratory interest. If all six
mine sites were developed, the cumulative area covered by these six mine footprints could be 37 to 57
km2 (9,100 to 14,000 acres). Stream habitats eliminated or blocked could be 43 to 70 km (27 to 43
miles). Cumulative wetland losses could be 7.9 to 27 km2 (2,000 to 6,700 acres).
These are conservative estimates of habitat loss, because we did not estimate the hydrologic drawdown
zones around each mine pit as was done for the Pebble scenarios. Inclusion of the drawdown area in the
Pebble 0.25 scenario increased the area of stream and wetlands losses by roughly 50%. A similar
increase might be expected at the other mine sites, depending on local geology. These mines also would
be expected to modify streamflows and diminish water quality to approximately the same extent as the
Pebble 0.25 scenario. Waters on these claim blocks include the Chulitna River and Rock, Jensen, Yellow,
Napotoli, Klutuk, and Kenakuchuk Creeks, as well as over 250 unnamed tributaries and over 50
unnamed lakes and ponds. Although not all support salmon, many do. Loss of substantial habitat across
the watersheds could contribute to diminishing the genetic diversity of salmon stocks and thereby
increasing annual variability in salmon returns.
Mitigation and Remediation
The mine scenarios assessed here include modern conventional mitigation practices as reflected in
Northern Dynasty Mineral's published plan for the Pebble deposit, plus practices suggested in the
mining literature and consultations with experts. These practices include, but are not limited to,
processing all potentially acid-generating waste rock before closure, managing effluent water
temperatures, inspecting and maintaining roads daily, and providing automatic monitoring and remote
shut-off for the pipelines. However, we recognize that risks could be further reduced by unconventional
or even novel mitigation measures, such as dry stack tailings disposal or the use of armored containers
on the trucks carrying process chemicals to the site. These practices may be unconventional because
they are expensive, unproven, or impractical. However, these obstacles to implementation might be
overcome and justified by the large mineral resource and the highly valued natural and cultural
resources of the Bristol Bay watershed.
Although remediation would be considered if spills contaminated streams, features of the Pebble
deposit area would make remediation difficult. Spilled tailings from a dam failure would flow into
streams, rivers, and floodplains that are in roadless areas and that are not large enough to float a barge-
mounted dredge. Recovery, transport, and disposal of hundreds of millions of metric tons of tailings
under those conditions would be extremely difficult and would result in additional environmental
damage. Compensatory mitigation measures could offset some of the stream and wetland losses,
although there are substantial challenges regarding the efficacy of these measures to offset adverse
Bristol Bay Assessment ES 27 January 2014
-------
impacts. Pipeline crossings of streams would be near Iliamna Lake, so the time available to block or
collect spilled material before it reached the lake would be short. Spilled return water and the aqueous
phase of the product concentrate slurry would be unrecoverable. The product concentrate itself would
resemble fine sand, and mean velocities in many receiving streams would be sufficient to suspend and
transport it. Hence, concentrate spilled or washed into streams could be recovered only where it
collected in low-velocity locations. Diesel spills would dissolve, vaporize, and flow as a slick to Iliamna
Lake. Booms and absorbents are not very effective in moderate- to high-velocity streams.
Summary of Uncertainties in Mine Design and Operation
This assessment considers realistic mine scenarios that are based on specific characteristics of the
Pebble deposit and preliminary plans proposed by Northern Dynasty Minerals. These scenarios are
generally applicable to copper deposits in the Bristol Bay watershed. If the Pebble deposit is mined,
actual events will undoubtedly deviate from these scenarios. This is not a source of uncertainty, but
rather an inherent aspect of a predictive assessment. Even an environmental assessment of a specific
plan proposed for permitting by a mining company would be an assessment of a scenario that
undoubtedly would differ from actual mine development.
Multiple uncertainties are inherent in planning, designing, constructing, operating, and closing a mine.
These uncertainties, summarized below, are inherent in any complex enterprise, particularly when that
enterprise involves an incompletely characterized natural system. However, the large spatial scales and
long durations required to mine the Pebble deposit make these inherent uncertainties more prominent.
• Mines are complex systems requiring skilled engineering, design, and operation. The uncertainties
facing mining and geotechnical engineers include unknown geological features, uncertain values of
geological properties, limited knowledge of mechanisms and processes, and human error in design
and construction. Models used to predict the behavior of engineered systems represent idealized
processes and by necessity contain simplifications and approximations that potentially introduce
errors.
• Accidents are unplanned and inherently unpredictable. Although systems can be put into place to
protect against system failures, seemingly logical decisions about how to respond to a given
situation can have unexpected consequences due to human error (e.g., the January 2012 overflow of
the tailings dam at the Nixon Fork Mine near McGrath, Alaska). Further, unforeseen events or events
that are more extreme than anticipated can negate apparently reasonable operation and mitigation
plans. Climate change will likely exacerbate this uncertainty. In the Bristol Bay region, climate
change is expected to lead to changes in snowpack and the timing of snowmelt, an increased chance
of rain-on-snow precipitation, and increased flooding. All of these changes are likely to affect
multiple aspects of any large-scale mining in the area, including mine infrastructure, the
transportation corridor, water treatment and discharge, and post-closure management, in unknown
and potentially unpredictable ways.
Bristol Bay Assessment ES 28 January 2014
-------
• The ore deposit would be mined for decades and wastes would require management for centuries or
even in perpetuity. Engineered mine waste storage systems have been in existence for only about 50
years, and their long-term behavior is not known. The response of current technology in tailings
dam construction is untested and unknown in the face of centuries of unpredictable events such as
extreme weather and earthquakes.
• Over the long time span (centuries) of mining and post-mining care, generations of mine operators
must exercise due diligence. Priorities could change in the face of financial circumstances, changing
markets for metals, new information about the resource, political priorities, or any number of
currently unforeseeable changes in circumstance.
Summary of Uncertainties and Limitations in the Assessment
The most important uncertainties concerning estimated effects of the mine scenarios, as judged by the
assessment authors, are identified below.
• Consequences of habitat loss and degradation for fish populations could not be quantified because of
the lack of quantitative information concerning salmonid populations in freshwater habitats. The
occurrence of salmonid species in rivers and major streams is known, but detailed and
comprehensive information on abundances, productivities, and limiting factors in each of the
watersheds is not available. Estimating fish population changes would require population modeling,
which requires knowledge of life-stage-specific survival and production and limiting factors and
processes. Further, it requires knowledge of how temperature, habitat structure, prey availability,
density dependence, and sublethal toxicity influence life-stage-specific survival and production.
Obtaining this information would require more detailed monitoring and experimentation. Salmon
populations naturally vary in size due to many factors that vary among locations and years. At
present, data are insufficient to establish reliable salmon population estimates, and obtaining such
data would take many years. Estimated effects of mining on fish habitat thus become the best
available surrogate for estimated effects on fish populations.
• Standard leaching test data are available for test tailings and waste rocks from the Pebble deposit,
but these results are uncertain predictors of the actual composition of leachates from waste rock
piles, tailings impoundments, or tailings deposited in streams and on their floodplains.
• Leachate capture efficiencies are uncertain. We assume 50% capture for waste rock leachates
outside of the mine pit drawdown zone. In the Pebble 2.0 scenario, for example, this would result in
capture of 84% of the leachate by the pit drawdown zone and the wells combined. To avoid
exceeding water quality criteria for copper, more than 99% capture would be required.
• The quantitative effects of tailings and product concentrate deposited in spawning and rearing
habitat are uncertain. It is clear that they would have harmful physical and toxicological effects on
salmonid larvae or sheltering juveniles, but the concentration in spawning gravels required to
reduce salmonid reproductive success is unknown.
Bristol Bay Assessment ES 29 January 2014
-------
• The estimated annual probability of tailings dam failure is uncertain because it is based on design
goals. Historical experience is presumed to provide an upper bound of failure probability. Features
that should reduce failure frequencies have not been tested for the thousands of years that they
must function properly. Hence, actual failure rates could be higher or lower than the estimated
probability.
• The proportion of tailings that would spill in the event of a dam failure could be larger than the
largest value modeled (20%).
• The long-term fate of spilled tailings in the event of a dam failure could not be quantified. It is
expected that tailings would erode from areas of initial deposition and move downstream over more
than a decade. However, the data needed to model that process and the resources needed to develop
that model are not available.
• The actual response of Alaska Native cultures to any impacts of the mine scenarios is uncertain.
Interviews with village Elders and culture bearers and other evidence suggest that responses would
involve more than the need to compensate for lost food, and would be expected to include some
degree of cultural disruption. It is not possible to predict specific changes in demographics, cultural
practices, or physical and mental health.
• Because we mention but do not evaluate potential direct effects of mining on wildlife or on Alaska
Natives, this assessment represents a conservative estimate of how these endpoints would be
affected by mine development and operation.
Uses of the Assessment
This assessment is a scientific investigation. It does not reflect any conclusions or judgments about the
need for or scope of potential government action, nor does it offer or analyze options for future
decisions. Rather, it is intended to provide a characterization of the biological and mineral resources of
the Bristol Bay watershed, increase understanding of the risks from large-scale mining to the region's
fish resources, and inform future government decisions. The assessment will also better inform
dialogues among interested stakeholders concerning the resources in the Bristol Bay watershed and the
potential impacts of large-scale mining on those resources.
Bristol Bay Assessment ES 30 January 2014
-------
The Bristol Bay watershed in southwestern Alaska supports the largest sockeye salmon fishery in the
world, is home to 25 federally recognized tribes, and contains abundant natural resources, including
significant mineral reserves. Worldwide attention to this watershed has increased because of
widespread mineral exploration activities and the discovery of a large ore deposit in the watershed's
northeast-central region. The potential for large-scale mining activities has raised concerns about the
quality and sustainability of Bristol Bay's world-class fisheries and the future of Alaska Natives who
have maintained a salmon-based culture and a subsistence-based way of life for at least 4,000 years.
Public interest in the Bristol Bay watershed has centered on the ecological goods and services provided
by the watershed and on potential mining activity. The watershed is most noted for its abundant fish
resources. The Bristol Bay watershed supports production of all five species of Pacific salmon found in
North America (sockeye, Chinook, chum, coho, and pink), including almost half of the world's
commercial sockeye salmon harvest. In 2009, Bristol Bay's ecosystems, which support the watershed's
commercial, recreational, and subsistence fisheries, generated $480 million in direct economic
expenditures in the region, and provided employment for over 14,000 full- and part-time workers
(Appendix E). This consistently large fish production results from the watershed's high hydrologic
diversity and pristine quality, both of which contribute to highly diverse fish populations.
In addition to these biological resources, 16 mine claim blocks have recently been explored in the Bristol
Bay's Nushagak and Kvichak River watersheds. Eleven of these claims are associated with porphyry
copper deposits, the largest belonging to the Pebble Limited Partnership (PLP). This partnership was
created in 2007 by co-owners Northern Dynasty Minerals, Ltd. and Anglo American to design, permit,
construct, and operate a long-life mine at the Pebble deposit (PLP 2013) (Anglo American withdrew
from the partnership in late 2013). Although PLP has not yet submitted a permit application for a mine,
preliminary mine plans have been developed and publicly available information strongly suggests that a
Bristol Bay Assessment
1-1
January 2014
-------
Chapter 1 Introduction
mine at the Pebble deposit has the potential to become one of the largest mining developments in the
world.
The Pebble deposit is a large, low-grade deposit containing copper, gold, and molybdenum-bearing
minerals. Extraction is expected to require the creation of a large open pit, the production of large
amounts of waste rock and mine tailings, the creation of a transportation corridor connecting the
deposit area to Cook Inlet, and the development of a deep-water port. Revenues from such a mine have
been estimated at between $300 billion and $500 billion over the mine's life (Chambers et al. 2012), and
more than 2,000 and 1,000 jobs could be created during mine construction and operation, respectively
(Ghaffari et al. 2011).
In light of these factors, nine Bristol Bay federally recognized tribes, the Bristol Bay Native Association,
the Bristol Bay Native Corporation, other tribal organizations, and many groups and individuals
petitioned the U.S. Environmental Protection Agency (USEPA) to use its authorities under Clean Water
Act (CWA) Section 404(c) to restrict or prohibit the disposal of dredged or fill material associated with
large-scale mining activities in the Bristol Bay watershed. These groups are concerned that large-scale
mining could adversely affect the region's valuable natural resources, particularly its fisheries. Four
Bristol Bay federally recognized tribes, other tribal organizations, the governor of Alaska, and other
groups and individuals, including PLP, have asked USEPA to wait until formal mine permit applications
have been submitted and an environmental impact statement has been developed.
USEPA initiated this assessment in response to these competing requests. The assessment's purpose is
to characterize the biological and mineral resources of the Bristol Bay watershed, increase
understanding of the potential impacts of large-scale mining, in terms of both day-to-day operations and
potential accidents and failures, on the region's fish resources, and inform future decisions, by
government agencies and others, related to protecting and maintaining the chemical, physical, and
biological integrity of the watershed. The assessment represents a review and synthesis of available
information to identify and evaluate potential risks of future large-scale mining development on the
Bristol Bay watershed's fish habitats and populations and consequent indirect effects on the region's
wildlife and Alaska Native cultures.
1.1 Assessment Approach
This assessment was conducted as an ecological risk assessment (ERA). ERA is a scientific process used
to determine whether exposure to one or more stressors may result in adverse ecological effects, the
findings of which are used to inform environmental decision making. USEPA routinely uses ERA
methods to evaluate the potential impacts of current and future actions when considering management
decisions (USEPA 2002a, 2002b, 2002c, 2008, 2011). USEPA is conducting this assessment consistent
with its authority under CWA Section 104. CWA Sections 104(a) and (b) provide USEPA with the
authority to study the resources of the Bristol Bay watershed, evaluate the effect of pollution from large-
scale mining development on those resources, and make such an assessment available to the public. This
Bristol Bay Assessment ,. 2 January 2014
-------
Chapter 1 Introduction
assessment is not an environmental impact assessment, an economic or social cost-benefit analysis, or
an assessment of any one specific mine proposal.
Risk assessors, decision makers, and community stakeholders determine the topical, spatial, and
temporal scope needed to effectively address the decisions the ERA is informing. Within this scope, risk
assessments consider the potential effects of an activity and use one or more scenarios, or sets of
assumptions, to identify how resources of interest (in this case, fish habitat and populations) could be
exposed to stressors generated by some activity (in this case, porphyry copper mining).
Assessment endpoints are explicit expressions of the environmental resources of interest to the risk
assessors, decision makers, and stakeholders. We selected fish, and specifically salmon and other
salmonids, as our primary assessment endpoint because of their critical importance to stakeholders and
future decision making in the watershed. The sustainability of the Bristol Bay fisheries is a concern
shared by all Bristol Bay stakeholders—including those who support mining—and the ecological,
economic, and cultural importance of the region's commercial, sport and subsistence fisheries has been
emphasized consistently by all stakeholders throughout the process. Our preliminary technical
consultations with federal, state, and tribal representatives indicated that evaluating the potential risks
of large-scale mining on the region's fishery resources was a top priority. During our public engagement
efforts, stakeholders consistently emphasized that fish are the crucial resource of concern.
We also considered two key secondary endpoints: wildlife and Alaska Native cultures. Fish-mediated
effects on wildlife were considered because fish, particularly salmon, are an important food resource for
wildlife, via both direct consumption and as a source of marine-derived nutrients that contribute to the
watershed's overall productivity. Fish-mediated effects on Alaska Natives were considered because
sustainability of the region's fish populations is critical to the future of Alaska Natives in the Bristol Bay
region, and because concern about the region's fishery resources prompted the original requests from
Alaska Natives that USEPA examine potential mine development in the Bristol Bay watershed.
Multiple geographic scales are considered in the assessment. Background and characterization
information is presented for the entire Bristol Bay watershed. The evaluation of potential large-scale
mining impacts focuses on the Nushagak and Kvichak River watersheds. These two watersheds are the
most likely to be affected by large-scale mining development, given the location of current mine claims
and current federal and state restrictions on development in other portions of the Bristol Bay
watershed. These two watersheds are responsible for approximately half of the Bristol Bay sockeye
salmon production, and are also home to approximately half of the region's Alaska Native communities
and federally recognized tribes. There are 31 federally recognized tribes in the larger Bristol Bay region,
and 25 of these tribal communities are within the Bristol Bay watershed boundary defined in this
assessment. Fourteen of these communities (13 of which have federally recognized tribes) are within
the Nushagak and Kvichak River watersheds.
The assessment also considers smaller geographic scales for risk analysis and characterization. Because
the Pebble deposit is the largest known and most explored deposit in the region, we use it as a case
study for potential risks. Because none of the parties holding mine claims in the Bristol Bay watershed
Bristol Bay Assessment ,. - January 2014
-------
Chapter 1 Introduction
has submitted a formal application and mine plan, we developed a set of realistic mine scenarios for the
assessment. The foundations for these scenarios are industry documents outlining approaches for
mining porphyry copper deposits, as well as specific documents from the PLP outlining a basic,
preliminary mine plan for the Pebble deposit. The mine scenarios were used to complete the risk
analyses and characterizations in the assessment. Although these mine scenarios were developed for the
Pebble deposit, the potential risks evaluated are expected to be qualitatively similar to potential risks
associated with any mine of the same resource type (porphyry copper) anywhere in the Nushagak and
Kvichak River watersheds.
Risk assessments are inherently uncertain, because they must predict the occurrence and consequences
of future actions. In this assessment, expressions of uncertainty are treated differently for accidents and
failures and for routine operations. Risks of accidents and failures are based on empirical frequencies
(summarized in Table 14-1), but we acknowledge the possibility of lower risks due to advances in
technology or practices. For example, data concerning risks of culvert failure provide frequencies of 0.3
to 0.6 per culvert. However, risks during operation are simply described as low, because our scenario
specifies daily inspections and there are no data quantifying failure rates under such intensive
maintenance programs. Risks that effects would occur due to routine operations are not described
probabilistically, because they are unintended results of planned actions. However, these risks are
uncertain due to lack of knowledge about the receiving environment and its response to mining
activities. Those uncertainties are described based on the professional judgment of the authors using
ordinary language such as "likely" and, when the evidence allows, in terms of possible deviations from
expectations (e.g., thresholds for effects could be at least a factor of 2 lower). The term "likely" is used
commonly as an abbreviation for "more likely than not" (>0.5 probability). The risk of a tailings storage
facility (TSF) spillway release is different in that it is a hybrid between a failure (TSFs should not
overflow during mine operation) and routine operations (spillways are installed to spill excess water,
because overflows are a reasonable expectation). No statistics are available on overflow frequencies, but
they are judged to be likely over the life of a mine and inevitable afterwards, if water treatment is not
continued in perpetuity.
Throughout the assessment, we have reached out to interested parties to ensure transparency of the
assessment process (Box 1-1). Through public comment opportunities and by engaging an
Intergovernmental Technical Team (IGTT) of federal, state, and tribal representatives, we were able to
identify additional information helpful for characterizing the biological and mineral resources of the
watershed. These interactions with community members were also helpful in narrowing the scope of
the assessment to issues that were most important to stakeholders.
Bristol Bay Assessment ,. , January 2014
-------
Chapter 1 Introduction
BOX 1-1. STAKEHOLDER INVOLVEMENT IN THE ASSESSMENT
Meaningful engagement with stakeholders was essential to ensure that the U.S. Environmental Protection Agency
(USEPA) heard and understood the full range of perspectives on the assessment and the potential effects of
mining in the region. USEPA involved and informed stakeholders throughout the assessment process. Community
involvement efforts included a project webpage and listserv to ensure that assessment-related information is
shared with the public. Additional ways in which stakeholders and tribal governments were involved in the
assessment process are summarized below.
• Public and stakeholder meetings. Throughout development of the assessment, USEPA visited many Bristol
Bay communities, including Ekwok, Dillingham, Kokhanok, NewStuyahok, Koliganek, Iliamna, Newhalen,
Nondalton, Naknek, King Salmon, Igiugig, and Levelock. USEPA also met with representatives from Bristol Bay
tribal governments and corporations, as well as organizations representing the mine industry, commercial
fishers, seafood processors, hunters and anglers, chefs and restaurant owners, jewelry companies,
conservation interests; members of the faith community; and elected officials from Alaska and other states.
USEPA heard from hundreds of people at these meetings and from thousands more via phone and email.
USEPA was also invited to numerous conferences and meetings to discuss the assessment.
• Intergovernmental Technical Team (IGTT). In August 2011, USEPA met with the IGTT, which was
established to provide USEPA with input on the structure of the assessment and to identify potential data
sources. IGTT participants included tribal representatives from Ekwok, Newhalen, Iliamna, South Naknek, New
Koliganek, Curyung, Nondalton, and Levelock and agency representatives from the Alaska Department of
Public Health, the Alaska Department of Fish and Game, the National Park Service, the U.S. Fish and Wildlife
Service, the National Oceanic and Atmospheric Administration, and the Bureau of Land Management.
Feedback from this workshop was used to inform the early stages of problem formulation. USEPA also updated
the IGTT on assessment progress in January 2012 via webinar.
• Tribal consultation. USEPA's policy is to consult on a government-to-government basis with federally
recognized tribal governments when USEPA actions and decisions may affect tribal interests. Consultation is a
process of meaningful communication and coordination between USEPA and tribal officials. In February 2011,
USEPA invited all 31 federally recognized tribal governments (tribes) of the Bristol Bay region to enter formal
consultation on the assessment, to ensure their involvement and to include their concerns and relevant
information in the assessment. Throughout development of the assessment there have been numerous
opportunities for tribes to participate in the tribal consultation process. Not all tribes elected to participate in
consultation. USEPA met with representatives from 20 of the 31 tribes (including all 13 tribes with federally
recognized tribal governments in the Nushagak and Kvichak River watersheds), either in person or on the
phone, during the consultation process.
• Alaska Native Claims Settlement Act (ANCSA) engagement. USEPA provided multiple engagement
opportunities for ANCSA Village and Regional Corporations throughout development of the assessment,
consistent with Public Law 108-199, Division H, Section 161, and Public Law 108-447, Division H, Title V,
Section 518. USEPA representatives traveled to King Salmon, Iliamna, and Anchorage for meetings at the
request of multiple ANCSA Corporations, to share information about and receive input on the assessment.
Additionally, ANCSA Corporation representatives were invited to participate in a webinar following the release
of April 2013 draft of the assessment. Throughout assessment development, ANCSA Corporations have
traveled numerous times to meet with USEPA officials in Anchorage, Seattle, and Washington, D.C. Seventeen
of the 26 ANCSA Corporations within the Bristol Bay region were engaged through these mechanisms.
• Public comments. USEPA released two drafts of the assessment for public comment. Approximately 233,000
and 890,000 comments were submitted to the USEPA docket during the 60-day public comment period for the
May 2012 and April 2013 drafts of the assessment, respectively. USEPA also held eight public comment
meetings in June 2012, in Dillingham, Naknek, New Stuyahok, Nondalton, Levelock, Igiugig, Anchorage, and
Seattle. Approximately 2,000 people attended these meetings. An overview of these meetings was shared via
two webinars in July 2012.
• Public involvement in peer review. USEPA provided multiple opportunities for stakeholder involvement in
the peer review process. In February 2012, the public was invited to nominate qualified scientists as potential
peer reviewers; these nominations were submitted to the peer review contractor for consideration. In March
2012, USEPA requested public comments on the questions to be given to peer reviewers, and these questions
were revised in response to comments received. In August 2012, the public was invited to participate in the
first 2 days of the peer review meeting in Anchorage, to provide oral comments to and observe discussions
among the peer reviewers.
Bristol Bay Assessment ,. c January 2014
-------
Chapter 1 Introduction
Detailed background characterizations of the watershed's resources are included in the assessment's
appendices. We used these characterization studies and input from the IGTT to develop a series of
conceptual models illustrating potential linkages between sources and stressors associated with large-
scale mining and the assessment endpoints. These models were then used to develop a plan for
analyzing and characterizing risks. In the risk analysis, available data were used to assess potential
exposures to stressors and potential effects on assessment endpoints stemming from those exposures.
In the final phase, results of these analyses were integrated to provide a comprehensive picture of the
risks to assessment endpoints within the defined scope of the assessment. The uncertainties and
limitations associated with these analyses were also identified.
This assessment has undergone extensive review throughout its development. Two earlier drafts of the
assessment, released in May 2012 and April 2013, were subjected to review by 12 independently
selected, expert peer reviewers (Box 1-2). Both of these drafts also had 60-day public comment periods,
during which interested parties could submit their comments on the assessment to USEPA (Box 1-1).
1.2 Uses of the Assessment
This assessment is a scientific investigation. It does not reflect any conclusions or judgments about the
need for or scope of possible government action, nor does it offer or analyze options for future decisions.
Rather, it is a scientific product intended to provide a characterization of the biological and mineral
resources of the Bristol Bay watershed, increase understanding of the potential risks to fish resources
from large-scale mining, and inform future government decisions.
USEPA and other stakeholders may use this assessment in several ways. The assessment will inform the
public and interested government entities about the resources of the Bristol Bay watershed. Much of the
information about these resources was previously found in a variety of sources. In this assessment, we
have synthesized and integrated available literature and provided a useful summary characterizing the
Bristol Bay watershed's resources.
The assessment also will inform the public and interested government entities about the potential
impacts of large-scale mining. USEPA recognizes the high level of interest concerning the impacts of
potential mine development on the watershed's ecological resources. That interest originates from
Alaska Native communities within the watershed, other Alaska residents, and interested parties
throughout the United States. It is expressed both by those interested in protecting the Bristol Bay
fishery and by those interested in developing the watershed's extensive mineral resources. This
assessment is a scientific and technical resource that is useful to members of the public as they weigh
the challenges of both mining and protecting the ecological resources in the Bristol Bay watershed in the
years ahead.
Bristol Bay Assessment ,. g January 2014
-------
Chapter 1 Introduction
BOX 1-2. OVERVIEW OF THE ASSESSMENT'S PEER REVIEW PROCESS
The peer review process is designed to provide a documented, independent, and critical review of a draft
assessment. Its purpose is to identify any problems, errors, or necessary improvements to a document prior
to it being published or otherwise released as a final document. To this end, the U.S. Environmental
Protection Agency (USEPA) tasked Versar, an independent contractor, with coordinating an external peer
review of the May 2012 draft assessment. Versar assembled 12 independent experts to serve as peer
reviewers. These reviewers were selected from a pool of candidates that included those suggested during a
public nomination process. In assembling the peer reviewers, Versar evaluated the qualifications of each
peer review candidate and conducted a thorough conflict of interest screening process.
The peer reviewers were asked to evaluate and provide a written review of the May 2012 draft of the
assessment (the main report and its appendices) by responding to 14 questions developed by USEPA with
input from public commenters. Peer reviewers were charged only with evaluating the quality of the science
included in the draft assessment and were not charged with making any regulatory recommendations,
commenting on any policy implications of USEPA's role or mine development in the region, or reaching
consensus in either their deliberations (during the peer review meeting, see below) or their written
comments. Peer reviewers were provided with a summary of public comments submitted during the 60-day
public comment period for the May 2012 draft and were given access to the public comments themselves.
A 3-day peer review meeting, coordinated by Versar, was held in Anchorage, Alaska, on August 7 through 9,
2012. On the first day of the meeting, peer reviewers heard testimony from approximately 100 members of
the public. Peer reviewers deliberated amongthemselves on the second and third days of the meeting;
these deliberations were open to the public on the second but not the third day.
Following the public peer review meeting, peer reviewers were given additional time to complete their
individual written reviews. Versar provided these final written comments to USEPA in their Final Peer Review
Meeting Summary Report for the May 2012 draft, which USEPA released to the public in November 2012.
USEPA considered these peer review comments, as well as comments received during the 60-day public
comment period, as they revised the May 2012 draft of the assessment.
In April 2013, USEPA released a revised draft of the assessment. The same 12 peer reviewers were asked
to conduct a follow-on peer review to evaluate whether the April 2013 draft of the assessment was
responsive to their original comments. USEPA provided reviewers with a draft response to comments
document, in which USEPA responses to peer review comments on the May 2012 draft assessment were
added to the Final Peer Review Meeting Summary Report submitted by Versar.
In the follow-on review, peer reviewers were asked to go through their comments on the May 2012 draft,
review USEPA's draft responses to their original comments, and evaluate whether their original review
comments had been addressed sufficiently and whether appropriate changes had been incorporated into
the April 2013 draft. USEPA received these follow-on peer review comments directly from the 12 peer
reviewers in August to September 2013. Again, USEPA considered these peer review comments, as well as
comments received during the 60-day public comment period, as they revised the April 2013 draft of the
assessment.
All drafts of the assessment (May 2012, April 2013, and final), as well as the peer review comments on the
May 2012 and April 2013 drafts and USEPA's responses to those comments, are available online.
Bristol Bay Assessment ,. -, January 2014
-------
Chapter 1 Introduction
Our findings concerning the potential impacts of large-scale mining will help to inform future
government decisions regarding mine development in the Bristol Bay watershed and potential actions
to protect and maintain the integrity of the watershed's aquatic resources. One of the initiators for the
assessment was the multiple petitions to USEPA to use its authority under CWA Section 404(c). It is
expected that the assessment will provide an important base of information for any agency decision
about whether or not to use Section 404(c), either now or in the future, and will facilitate a thoughtful
decision regarding whether application of this authority is or is not warranted.
The assessment may also assist federal and state scientists and resource managers involved in the
evaluation of future mine permit applications submitted for the deposits in the Bristol Bay watershed. It
is likely that future mines in the watershed would require the filling of streams and wetlands and thus
would require a Section 404 permit from the U.S. Army Corps of Engineers. USEPA reviews and
comments on proposed Section 404 permit applications, and this assessment will be a valuable resource
in the development and review of such permit applications.
If a Section 404 permit or other major federal action is required for a future mine in the watershed, it
would trigger review of the proposed mine under the National Environmental Policy Act. This
assessment, particularly in terms of its identification and analysis of potential direct, indirect, and
cumulative effects of large-scale mining, will be a valuable resource in the development and review of
any future environmental assessment related to mining in the Bristol Bay watershed.
Perhaps the most important use of this assessment is to better inform dialogues among interested
stakeholders concerning the resources in the Bristol Bay watershed and the potential impacts of large-
scale mining on those resources.
Bristol Bay Assessment ,. o January 2014
-------
2.1 Structure
We based this assessment on U.S. Environmental Protection Agency (USEPA) guidelines for ecological
risk assessment (ERA) (USEPA 1998). We began by reviewing existing literature to synthesize
background information on the Bristol Bay region, particularly the Nushagak and Kvichak River
watersheds. This information focused on several topics, including the ecology of Pacific salmon and
other fishes; the ecology of relevant wildlife species; mining and mitigation, particularly in terms of
porphyry copper mining; potential risks to aquatic systems due to road and pipeline crossings; fishery
economics; and Alaska Native culture. These detailed background characterizations are included as
appendices to this assessment.
In accordance with the different phases of an ERA, the assessment document itself is organized into two
main sections: Problem Formulation (Chapters 2 through 6) and Risk Analysis and Characterization
(Chapters 7 through 14). Problem formulation is the first phase of an ERA, during which the purpose
and scope of the assessment are defined (USEPA 1998). Risk assessors, decision makers, and
stakeholders determine the topical, spatial, and temporal scope needed to effectively address whatever
decision process the assessment is meant to inform. Assessment endpoints, or explicit expressions of the
environmental entities of interest (USEPA 1998), are identified. Conceptual models illustrating potential
linkages among sources, stressors, and endpoints considered in the assessment (Box 2-1), as well as a
plan for analyzing and characterizing risks, are developed.
The risk analysis and characterization phases follow problem formulation (USEPA 1998). During the
risk analysis phase, available data are used to assess potential exposures to stressors and exposure-
response relationships for those exposures and endpoint effects. In the risk characterization phase,
information on exposures and effects is integrated, and the uncertainties and limitations associated with
the assessment's analyses are identified.
Bristol Bay Assessment
2-1
January 2014
-------
Chapter 2
Overview of Assessment
step in causal
pathway
BOX 2-1. CONCEPTUAL MODELS
Throughout this assessment we use conceptual model diagrams to illustrate potential ways in which large-
scale mine development may adversely affect the Bristol Bay watershed's biota and Alaska Native cultures.
These conceptual model diagrams show hypothesized pathways linking common sources associated with
miningto potential stressors, and those stressors to responses of interest. Inclusion of a pathway indicates
that the pathway can occur, not that it w/7/ definitely occur. Thus, these diagrams are not meant to illustrate
worst-case scenarios in which all pathways occur simultaneously. Rather, they are meant to provide
overviews of potential linkages among sources, stressors, and responses, one or more of which may
plausibly result from mine development.
The conceptual model diagrams contain the following elements
(note that not all elements are found in each diagram).
Sources are entities associated with mining that may directly or
indirectly result in one or more stressors.
Steps in causal pathways are processes or states that may link
sources to stressors or stressors to responses.
Stressors are physical or chemical entities that may directly induce
a response of concern.
Modifying factors are processes, states, or other factors that may
influence the delivery, expression, or effect of stressors (e.g.,
temperature, time or duration of exposure, mitigation).
Biotic responses are potential effects on salmon, other fishes, and
wildlife.
Human responses are potential effects on Alaska Native people
and culture.
When viewing these diagrams, it helps to keep the following
principles in mind.
• Arrows leading from one shape to another indicate a
hypothesized cause-effect relationship, whereby the first (or
originating) shape could plausibly cause or result in the second
shape.
• Arrows leading from a shape to another arrow (or a general
section of the diagram) indicate that the originating shape
(always categorized as a modifying factor) could plausibly
influence the cause-effect relationships indicated (e.g., by
increasing or decreasing its probability or intensity of
occurrence).
• Shapes bracketed under another shape are specific components
of the more general shape under which they appear.
• Within a shape, 1" indicates an increase in the parameter, 4,
indicates a decrease in the parameter, and A indicates a change
in the parameter.
step in causal
pathway
V
(Abiotic response J)
V
(f human response^)
2.1.1 Data Used in the Assessment
An ERA requires data of sufficient quantity and quality, from a variety of sources. Throughout the
problem formulation, risk analysis, and risk characterization phases, relevant data are identified and
acquired. These data may result from different kinds of studies, including field studies at the site of
interest, field studies at other sites somehow relevant to the site or issue of interest, laboratory tests,
and modeling applications.
Bristol Bay Assessment
2-2
January 2014
-------
Chapter 2 Overview of Assessment
In this assessment, we prioritized peer-reviewed, publicly accessible sources of information to ensure
that the information and data we incorporated were of sufficient quality. In many cases, however, peer-
reviewed data—particularly those directly relevant to potential mining in the Bristol Bay region—were
not available. Thus, we incorporated credible, non-peer-reviewed data from multiple sources, including
state government agencies (e.g., the Alaska Department of Fish and Game [ADF&G], the Alaska
Department of Natural Resources [ADNR]), federal government agencies (e.g., the U.S. Geological Survey
[USGS], the U.S. Fish and Wildlife Service [USFWS]), and academic organizations (e.g., Scenarios Network
for Alaska and Arctic Planning [SNAP] data).
We also incorporated non-peer-reviewed data collected under the auspices of the Pebble Limited
Partnership (PLP) (e.g., as presented in Ghaffari et al. 2011, PLP 2011), as these sources contain data
directly relevant to the Pebble deposit and the surrounding region. Both Ghaffari et al. (2011) and the
PLP's environmental baseline document (PLP 2011) are cited numerous times throughout the
assessment. PLP is currently conducting its own peer review of the data presented in its baseline
document, but that review had not been completed when this assessment was released.
Other non-governmental organizations have collected data relevant to the assessment. USEPA subjected
some of these documents to external peer review and, where defensible, we have incorporated this
information into the assessment (e.g., Chambers and Higman 2011, Woody and Higman 2011,
Earthworks 2012).
In addition, some minor sources of information (e.g., permits and reports filed by mining companies)
were used without peer review. In all cases, sources of information and data included in the assessment
are appropriately cited (Chapter 15).
Throughout the assessment, we present numbers from the scientific literature or from PLP (2011) using
the number of significant figures in the original source. Numbers derived for this assessment are
presented with the appropriate number of significant figures given the precision of the input data and
uncertainties due to modeling and extrapolation.
2.1.2 Types of Evidence and Inference
As in other ERAs, the risk analysis and characterization phase of this assessment is based on weighing
multiple types of evidence. Available and relevant pieces of evidence from a variety of sources are used
to follow different lines of inference and reach the best-supported conclusions.
In this risk analysis, we use general scientific knowledge, mathematical and statistical models, and data
from the Bristol Bay region, other sites (e.g., mines in other regions), and laboratory studies to evaluate
potential consequences of three mine size scenarios—that is, realistic potential mines of different sizes,
the characteristics of which are based largely on a mining company report (Ghaffari et al. 2011)—in
terms of sources, exposure to different stressors, and exposure-response relationships. First, we
estimate the magnitude of exposures potentially resulting from both routine operation and accidents
and failures in the mine scenarios, such as elevated aqueous copper concentrations, kilometers of
streams eliminated, and kilometers of streams upstream of road crossings. Then, we consider the effects
Bristol Bay Assessment 23 January 2014
-------
Chapter 2 Overview of Assessment
of these exposures—that is, the exposure-response relationships—on our endpoints of interest (e.g., the
relationship between water withdrawal and loss of salmon habitat, concentration-response
relationships for copper and fish). We describe and quantify, where possible, exposure-response
relationships for the endpoints and estimated exposures. For some issues, multiple lines of evidence are
available (e.g., state standards, federal criteria, effects models, field studies, and toxicity tests as lines of
evidence for copper toxicity); for other issues, lines of evidence are more limited.
Evidence from existing mines and other analogous facilities is used where relevant. Prior mining
activities in comparable watersheds provide examples of what can happen to the environment when
metals are mined. Some components of our mine scenarios have analogues in other industries (e.g., oil
and gas pipelines). These inferences by analogy reduce the uncertainties that come with modeling and
prediction, but introduce other uncertainties related to industry-specific or site-specific differences in
environmental conditions and potential changes in practices. Because no analogue is similar in all
aspects to potential mines and their components in the Bristol Bay region, we choose analogues to fit the
specific issues being assessed and take care to use analogues that are defensible despite their
differences from our mine scenarios. For example, the Fraser River watershed could be considered an
analogous system to the Bristol Bay watershed because it has similar mines and a similar salmon
resource, but we recognize that there are important differences between these systems (e.g., extensive
urban development, forestry, and agriculture in the Fraser River watershed). Metal mines in the Rocky
Mountain metal belt (e.g., sites near the Coeur d'Alene River, Idaho, and the Clark Fork River, Montana)
were developed using mining practices that would not be allowed under current mining laws. However,
the fate and effects of tailings in streams and floodplains at these sites, which also supported trout and
salmon populations, offer some parallels to the fate and effects of tailings following potential tailings
dam failures in the Bristol Bay region, should they occur—even if the underlying causes of failure differ.
The use of data from the historical, operational records of mines, pipelines, and roads is necessary but
controversial. It is essential and conventional for risk assessments to use the history of a technology to
estimate failure rates. However, developers argue, with some justification, that the record of older
technology is not relevant because of technological advances. Despite advances, no technology is perfect,
and rates of past failures may be a better guide to future outcomes than the expectation that developers
can design a system that will not fail. A classic example is the National Aeronautics and Space
Administration (NASA) space shuttle program, which denied the relevance of the failure rate of solid
rocket boosters and declared that the shuttle's rate of failure on launch would be one in a million. The
Challenger failure showed that the prior failure rate was still relevant, despite updated technology.
For most potential failures, historical failure rates are the only available evidence. New technologies
typically have not been in use long enough or widely enough to provide failure rates, and measures to
correct past failure modes may unwittingly introduce new ones. Thus, in this assessment we choose
failure rates that are most relevant and interpret them cautiously, using them to provide an upper
bound estimate of future failure rates.
Bristol Bay Assessment 24 January 2014
-------
Chapter 2 Overview of Assessment
After these analyses and lines of evidence are presented, we characterize risk for each line of evidence
by combining exposures and exposure-response relationships to estimate effects and by considering
uncertainties. We weigh different lines and types of evidence based on evidence strength and quality.
The resulting qualitative or quantitative estimates of risk and uncertainty are based on either the best
line of evidence or a combined estimate from multiple lines of evidence and inferences. Bounding
analyses, which set upper and lower limits for key parameters, are used to express uncertainties
concerning future mine activities and their effects. In particular, multiple mine sizes and durations are
included in the mine scenarios (Chapter 6). Bounding is also used to express stochasticity. For example,
the occurrence and magnitude of tailings dam failures are random variables that cannot be reasonably
defined. Hence, a range of tailings dam failure probabilities and a range of tailings release magnitudes
are evaluated (Chapter 9).
2.2 Scope
2.2.1 Topical Scope
Construction and operation of a large-scale mining operation require the development of extensive
infrastructure and involve numerous processes and components, each of which may have repercussions
for receiving environments. In this assessment, we do not consider all potential sources of risk
associated with the development of large-scale mining in the Bristol Bay watershed, all the stressors
that may result from these sources, and all the endpoints that may be affected. Rather, we focus on a
more limited set of sources, stressors, and endpoints based on stakeholder concerns and potential
decision-maker needs (Chapter 1). These focal components are described in broad terms below. In
Chapters 3 through 6, we consider these components in greater detail, and more specifically define the
focus of the assessment—in terms of geographic region, type of mining development, and ecological
endpoints—for risk analysis and characterization purposes.
In terms of sources, we consider the mine infrastructure and transportation corridor components of a
large-scale surface mining operation (Figure 2-1). Exploratory mining activities are ongoing in the
region (Box 2-2), but these activities are considered outside the scope of the assessment. Certain sources
associated with mining but not directly related to mine operations are not evaluated here, including
power generation and transmission facilities and activities, ancillary facilities such as housing for mine
workers and wastewater treatment plants to serve an increased human population, and construction
and operation of a deep-water port at Cook Inlet (Figure 2-1). A thorough evaluation of induced
development—development that is not part of the mine project, but for which the mine project provides
the impetus or opportunity, such as residential and commercial growth resulting from increased
accessibility—is also outside the scope of this assessment, although its importance is considered
qualitatively in Chapter 13.
Bristol Bay Assessment 25 January 2014
-------
Chapter 2
Overview of Assessment
Figure 2-1. Conceptual model illustrating sources, stressors, and responses potentially associated with large-scale mine development in
the Bristol Bay watershed. Pathways explicitly evaluated in this assessment are in bold; dashed pathways may be considered qualitatively in
parts of the assessment, but are generally considered outside its scope. See Box 2-1 for a general discussion of how conceptual models are
used and structured in the assessment.
I power generation &
'- transmission facilities
[ other ancillary
<- facilities
•yl [transportation |
f) L corridor J
port
_„__¥_„-_
f ^
induced
development
LEGEND
Bold arrows and outlines indicate
topics within assessment's scope.
Dashed arrows and outlines indicate
topics outside scope of assessment.
environmental impacts
on freshwater habitats
V
v
environmental impacts
on terrestrial habitats
environmental impacts
on marine habitats
other
environmental impacts
economic
impacts
! cultural
\ impacts
effects on
Alaska Native culture
Y
effects on
other biota
"" other effects on
Alaska residents
V- --V.
effects on
recreational sectors
effects on
commercial fisheries
Bristol Bay Assessment
2-6
January 2014
-------
Chapter 2 Overview of Assessment
BOX 2-2. EXPLORATORY MINING ACTIVITIES
Exploratory activities associated with the Pebble deposit—including geophysical, geochemical, and
environmental surveys, geological mapping, and drilling—have been underway for several decades (Ghaffari
et al. 2011). For example, 1,158 holes were drilled on the Pebble property through 2010, totaling
948,638 feet (289,145 m) (Ghaffari et al. 2011). These holes are concentrated in the Pebble deposit area,
but occur throughout the Pebble claim block. According to the Pebble Limited Partnership's annual
reclamation reports (submitted to the State of Alaska by the Pebble Limited Partnership in accordance with
their land use permits), the total amount of land disturbed between 2009 and 2012 was approximately 3
acres.
Because these exploratory activities require water, power, personnel support, and the use of chemicals,
heavy machinery, helicopters, and other equipment in relatively undeveloped areas, they likely have had
some environmental impact on the region. Full evaluation of these effects is beyond the scope of this
assessment, and it is likely that any effects of exploratory activities would be small relative to the effects of
full mine development.
In terms of stressors, we focus on potential environmental effects on freshwater habitats (Figure 2-1).
We focus on freshwater habitats because the Bristol Bay watershed supports exceptional fish
populations, and these populations are intimately linked to the watershed's freshwater habitats.
Although we recognize that large-scale mining could also have significant direct impacts on terrestrial
and marine systems, as well as direct economic and cultural repercussions, we do not evaluate these
impacts here (Figure 2-1).
Given the ecological and cultural significance of fishery resources in the Bristol Bay watershed, and the
fact that the health and sustainability of the watershed's fish populations are primary concerns shared
by all stakeholders interested in the Bristol Bay area (including those who support mining), we focus on
effects on key salmonids (Box 2-3) and resulting effects on wildlife and Alaska Native cultures as
assessment endpoints (Chapter 5). Direct effects of mining on wildlife and Alaska Native cultures,
although potentially significant, are not evaluated in this assessment. For example, construction and
operation of a transportation corridor would likely directly affect wildlife populations (Forman and
Alexander 1998); however, because the assessment focuses on freshwater habitats, these direct wildlife
effects are not considered here. The only effects on wildlife and Alaska Native cultures evaluated in the
assessment are those resulting from impacts on fish populations (Chapter 12). We also recognize that
many other endpoints may be directly affected by large-scale mining operations, including other biota
(e.g., vegetation, small mammals), other recreational and commercial fisheries, and human health
(Figure 2-1), but these topics are also outside the scope of the assessment.
It is important to keep in mind that exclusion of a source, stressor, or endpoint from this assessment
does not imply that it would be insignificant or unaffected. We recognize that many of the pathways we
identify as outside of the assessment's scope could have significant repercussions for the region's biota
and people.
Bristol Bay Assessment 27 January 2014
-------
Chapter 2 Overview of Assessment
BOX 2-3. KEYSALMONIDS IN THE BRISTOL BAY WATERSHED
The Bristol Bay watershed's freshwater habitats support a diverse and robust assemblage of fishes,
dominated by the family Salmonidae. This family comprises three subfamilies—Salmoninae (salmon, trout,
and char), Thymallinae (grayling), and Coregoninae (whitefish)—all of which are represented in the region. In
this assessment, we focus on fishes in the subfamily Salmoninae, particularly the five North American
Pacific salmon species (sockeye, Chinook, coho, chum, and pink), rainbow trout, and Dolly Varden (a species
of char). Collectively, we refer to these seven species as salmonids throughout this report.
All Salmonidae spawn in freshwater, but they can differ in their life histories. Some populations (e.g., Bristol
Bay's Pacific salmon) are anadromous, meaning that individual fish migrate to marine waters to feed and
grow before returning to fresh waters to reproduce. Other Bristol Bay populations (e.g., lake trout, Arctic
grayling) are non-anadromous (resident), meaning that essentially all individuals remain in fresh waters to
feed. Other populations (e.g., rainbow trout, Dolly Varden) can exhibit either anadromous or non-
anadromous life histories.
2.2.2 Geographic Scales
Throughout this assessment, we consider data across five geographic scales (Table 2-1, Figure 2-2).
• The Bristol Bay watershed (Scale 1, Figure 2-3) includes all the basins and waterways that flow into
Bristol Bay.
• The Nushagak and Kvichak River watersheds (Scale 2, Figure 2-4) include those drainage areas that
contain stream segments flowing either directly or via downstream segments into the mainstem
Nushagak River or Kvichak River.
• The mine scenario watersheds (Scale 3, Figure 2-5) include the cumulative drainage areas of the
South and North Fork Koktuli Rivers to their junction and Upper Talarik Creek to its junction with
Iliamna Lake.
• The mine scenario footprints (Scale 4, Figure 2-6) include the footprints of the major mine
components (i.e., the mine pit, waste rock piles, and tailings storage facilities), the groundwater
drawdown zone, and plant and ancillary facilities for each mine size scenario (Chapter 6).
• The transportation corridor area (Scale 5, Figure 2-7) includes 32 subwatersheds in the Kvichak
River watershed that drain to Iliamna Lake and would be crossed by the transportation corridor
(Chapter 6); the transportation corridor does not cross into the Nushagak River watershed.
These geographic scales are defined using the USGS National Hydrography Dataset (USGS 2012) (Box
2-4, Table 2-1). In problem formulation, we use broader geographic scales to describe the physical,
chemical, and biological environment in the Bristol Bay region (Table 2-1); we also use broader scales to
consider the effects of multiple mines across the landscape. In risk analysis and characterization, we use
finer geographic scales to evaluate the potential effects of mining operations.
Bristol Bay Assessment 2g January 2014
-------
Chapter 2
Overview of Assessment
BOX 2-4. THE NATIONAL HYDROGRAPHY DATASET
The National Hydrography Dataset (NHD) is a publicly available database of surface water information for
the United States (USGS 2012). Within the NHD, the entire landscape of the United States is organized into
a six-tiered system of nested hydrologic units, each with their own identifiable codes (hydrologic unit codes,
or HUCs). These tiers are defined as regions (represented by 2-digit codes), subregions (4-digit codes),
basins (6-digit codes), subbasins (8-digit codes), watersheds (10-digit codes), and subwatersheds (12-digit
codes). In total, the entire United States is divided into roughly 160,000 subwatersheds (12-digit HUCs)
within roughly 21 regions (2-digit HUCs). Due to the hierarchical nature of the system, all subwatersheds
(12-digit HUCs) within the same watershed start with the same first 10 digits, all watersheds (10-digit HUCs)
within the same subbasin start with the same first 8 digits, and so on.
It is important to note that the NHD hydrologic units do not always delineate true hydrologic watersheds (i.e.,
their boundaries do not always accurately indicate where water drains to a particular point). Nevertheless,
these boundaries are useful in both water resource and land management and are used as a foundational
geographic layer in this assessment.
Table 2-1. Geographic scales considered in the assessment.
Scale
1
2
3
4
5
Description
Bristol Bay watershed
Nushagak and
Kvichak River
watersheds
Mine scenario
watersheds
Hydrologic Unit Codes (HUCs)a
19030202-19030206,
19030301-19030306,
1903010101-1903010113,
1903010201-1903010203,
1903020101-1903020110
19030301-19030304,
19030205, 19030206b
190303021103, 190303021104,
190303021101-190303021102
1903020607,
Area
(% of scale above)
116,000 km2 (NA)
59,900 km2 (5 2%)
925 km2 (2%)
Representative
Chapters
2, 3, 4, 5, 13
2, 3, 4, 5, 13
6, 7, 8, 9, 12
Mine scenario footprints
Pebble 6.5
Pebble 2.0
Pebble 0.25
Transportation
corridor area0
NA
NA
NA
190302051403-190302051406,
190302060101-190302060104,
190302060201-190302060206,
190302060301-190302060302,
190302060701-190302060702,
190302060704,
190302060901-190302060905,
190302060907, 190302060914d
103 km2 (11%)
45.3 km2 (5%)
18.9 km2 (2%)
2,340 km2(4%e)
6, 7, 8, 9, 12
6, 7, 8, 9, 12
6, 7, 8, 9, 12
6, 10, 11
Notes:
'" From the National Hydrography Dataset (NHD) (USGS 2012). Scale 1 is defined by 8-digit and 10-digit HUCs; Scale 2 by 8-digit and 12-digit
HUCs; Scale 3 by 10-digit and 12-digit HUCs; Scale 5 by 12-digit HUCs. See Box 2-4 for further discussion of the NHD.
b Except for 190302062301-190302062311.
c The transportation corridor would include a 113-km road in the Kvichak River watershed; the area presented here represents the area of the
12-digit HUCs incorporating this road.
d The 190302060914 area was clipped to remove the area of Iliamna Lake and any land area draining directly to Iliamna Lake.
e Represents % of Scale 2 encompassed by the transportation corridor area HUCs.
NA = not applicable
Bristol Bay Assessment
2-9
January 2014
-------
Chapter 2
Overview of Assessment
Figure 2-2. The five geographic scales considered in this assessment. Only selected towns and villages are shown on this map. See Figures
2-3 through 2-7 for detailed views of each scale.
170 'W
I
~if Approximate Pebble Deposit Location
• Towns and Villages
j | Scale 1: Bristol Bay Watershed
| | Scale 2: Nushagak & Kvichak River Watersheds
Scale 3: Mine Scenario Watersheds
I Scale 4: Mine Scenario Components
I Scale 5: Transportation Corridor Area
N
A
0 50 100
0 50 100
] Miles
i
Bristol Bay Assessment
2-10
January 2014
-------
Chapter 2
Overview of Assessment
Figure 2-3. The Bristol Bay watershed (Scale 1), comprising the Togiak, Nushagak, Kvichak, Naknek, Egegik, and Ugashik River
watersheds and the North Alaska Peninsula. Only selected towns and villages are shown on this map.
Cook Inlet
Approximate Pebble Deposit Location
Towns and Villages
Watershed Boundary
Parks, Refuges, or Preserves
Bristol Bay Assessment
2-11
January 2014
-------
Chapter 2
Overview of Assessment
Figure 2-4. The Nushagak and Kvichak River watersheds (Scale 2).
Cooh /n/et
Bristol Bay
N
A
25
25
50
] Kilometers
50
] Miles
if Approximate Pebble Deposit Location
• Towns and Villages
Parks, Refuges, or Preserves
Watershed Boundary
Bristol Bay Assessment
2-12
January 2014
-------
Chapter 2
Overview of Assessment
Figure 2-5. The mine scenario watersheds—South Fork Koktuli River, North Fork Koktuli River, and
Upper Talarik Creek—within the Nushagak and Kvichak River watersheds (Scale 3).
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Mine Scenario Watershed
Watershed Boundary
NUSHAGAK .L
Bristol Bay Assessment
2-13
January 2014
-------
Overview of Assessment
Figure 2-6. Footprints of the major mine components for the three scenarios evaluated in the
assessment (Scale 4). Pebble 0.25 represents 0.25 billion ton of ore; Pebble 2.0 represents 2.0
billion tons of ore; Pebble 6.5 represents 6.5 billion tons of ore. Each mine footprint includes the
footprints of the major mine components shown here, as well as the groundwater drawdown zone
and the area covered by plant and ancillary facilities. See Figures 6-1, 6-2, and 6-3 for more detailed
maps of the major mine components for each scenario. Light blue areas indicate streams and rivers
from the National Hydrography Dataset (USGS 2012) and lakes and ponds from the National
Wetlands Inventory (USFWS 2012); dark blue areas indicate wetlands from the National Wetlands
Inventory (USFWS 2012).
r,
*r
- .4?*
t^?Vz^^>
4We'^^:' 7
.-,
'^.£ -'i**^
***-2Sto***&>!*
$%"
*v^ ^x
? * *:'"
i i i ,/" - y
J' V
^^r ^' r^ vrT^f!
^^ \ ? ^*;- >vv- ( v, - .j4&
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Bristol Bay Assessment
2-14
January 2014
-------
Chapter 2
Overview of Assessment
A Approximate Pebble Deposit Location
Transportation Corridor (Outside Assessment Area)
Transportation Corridor
Existing Roads
Transportation Corridor Area
Subwatersheds within Area
5 10
Kilometers
0 5 10
Miles
Bristol Bay Assessment
2-15
January 2014
-------
Bristol Bay is a large gulf of the Bering Sea located in southwestern Alaska. The land area draining to
Bristol Bay consists of six major watersheds—from west to east, the Togiak, Nushagak, Kvichak, Naknek,
Egegik, and Ugashik River watersheds—and a series of smaller watersheds draining the North Alaska
Peninsula (Figure 2-3). The Bristol Bay region encompasses complex combinations of physiography,
climate, geology, and hydrology, which interact to control the amount, distribution, and movement of
water through a landscape shaped by processes such as tectonic uplift, glaciation, and fluvial erosion
and deposition. The region's freshwater habitats are varied and abundant, and support a diverse and
robust assemblage offish (Chapter 5).
The Nushagak and Kvichak River watersheds account for more than half the land area in the Bristol Bay
watershed (Table 2-1). The Pebble deposit, the largest known porphyry copper deposit in the region, is
located in the headwaters of both watersheds (Figure 2-4) and represents the most likely site for near-
term, large-scale mine development in the Bristol Bay watershed. In this chapter, we consider key
aspects of the Bristol Bay watershed's physical environment, with particular emphasis on the Nushagak
and Kvichak River watersheds (Figure 2-4).
3.1 Physiographic Divisions
The Nushagak and Kvichak River watersheds comprise five distinct physiographic divisions (Wahrhaftig
1965): the Ahklun Mountains, the Southern Alaska Range, the Aleutian Range, the Nushagak-Big River
Hills, and the Nushagak-Bristol Bay Lowland (Table 3-1, Figure 3-1). Precipitation is greatest in the
Southern Alaska Range, the Aleutian Range, and the Ahklun Mountains (Figures 3-1 and 3-2), and these
physiographic divisions serve as major water source areas for lower portions of the watersheds. Annual
water balance, especially in the mountains and hills, is dominated by snowpack accumulation and
subsequent melt, although late summer and fall rains are also important contributors to the hydrologic
Bristol Bay Assessment
3-1
January 2014
-------
Chapter 3 Region
cycle, particularly in the Nushagak-Bristol Bay Lowland division (Selkregg 1974). Additional key
attributes of each physiographic division are discussed below.
The Ahklun Mountain physiographic division, in the western portion of the Nushagak River watershed,
is dominated by rolling hills to sharp, steep, glaciated mountains that receive high snowfall (Table 3-1,
Figure 3-1) (Wahrhaftig 1965, Selkregg 1974, Gallant et al. 1995). Parent bedrock is deformed
sedimentary rocks, intruded in several locations by igneous batholiths and stocks (Figure 3-3). A few
small glaciers occur in high mountain cirques, and isolated masses of permafrost occur sporadically
(Figure 3-4). Glacially carved lowland valleys are now filled with large, deep lakes, and adjacent streams
are often incised in bedrock gorges. The surrounding area is mantled with colluvium, alluvium, and
glacial drift and moraines (Figure 3-3). Soils are generally well drained and have medium erosion
potential (Figures 3-5 and 3-6). Dwarf scrub is the dominant vegetation in the mountains and tall scrub
and herbaceous plants are common in the valleys and lower mountain slopes (Figure 3-7).
The Southern Alaska Range physiographic division comprises a series of high, steep, glaciated
mountains with land surfaces covered by rocky slopes, glacial drift and moraines, and glaciers
(Table 3-1, Figure 3-1) (Wahrhaftig 1965, Selkregg 1974). Bedrock is a complex of granitic batholiths
intruded into metamorphosed sedimentary and volcanic rock (Figure 3-3). Soils are shallow or not
present (Figure 3-5) and permafrost occurs as isolated masses (Figure 3-4). Alpine tundra is the
predominant vegetation (Figure 3-7). Streams are frequently swift and braided with several headwaters
originating in glaciers (Figure 3-8). Several large, deep lakes occur in the glaciated valleys within the
division (Figure 3-8). Braided, turbid streams flow into lakes, allowing sediment to settle, before flowing
into the Nushagak and Kvichak River systems.
Within the Bristol Bay watershed, the Aleutian Range physiographic division consists of rolling hills to
steep, glaciated mountains built of sedimentary, volcanic, and intrusive bedrock (Table 3-1, Figure 3-1)
(Wahrhaftig 1965, Selkregg 1974). Cirque glaciers remain atop mountains in the extreme southeast
corner of the Kvichak River watershed (Figure 3-3). This division is generally free of permafrost
(Figure 3-4). Soils have formed in volcanic ash over glacial deposits at lower elevations, whereas rocky
lands dominate at higher elevations (Figure 3-5). Erosion potential is high for some soils in the Aleutian
Range division (Figure 3-6). Large, deep, moraine- and sill-impounded lakes are found in the ice-carved
valleys. The Alagnak River, which drains most of the Aleutian Range physiographic division within the
Bristol Bay watershed, is highly braided as it flows across the Nushagak-Bristol Bay Lowland division to
the Kvichak River. Dwarf scrub vegetation is common (Figure 3-7) (Selkregg 1974, Gallant et al. 1995).
Bristol Bay Assessment 30 January 2014
-------
Chapter 3
Region
Table 3-1. Physiographic divisions (Wahrhaftig 1965) of the Nushagak and Kvichak River watersheds.
Physiographic Division
Ahklun Mountains
Southern Alaska Range
Aleutian Range
Nushagak-Big River Hills
Nushagak-Bristol Bay
Lowland
Description
Rolling hills to sharp, steep, glaciated mountains
separated by broad lowlands, with a few small
glaciers in high mountain cirques
Rolling hills to steep, glaciated mountains covered
by glacial drifts and moraines, rocky slopes, and
glaciers
Rolling hills to sharp, steep glaciated mountains,
separated by broad lowlands, with a few small
glaciers in high mountain cirques
Rounded ridges with broad, gentle slopes and
broad, flat or gently sloping valleys
Flat to rolling landscape with low local relief and
deep morainal, drift, and outwash deposits, but no
glaciers
Elevation
(meters)
10-1,600
14-2,800
14-1,600
14-1,300
0-800
Permafrost
Extent
Sporadic
Unknown
Unknown
Sporadic
Sporadic or
absent
Freshwater Habitats
Mix of unconstrained and constrained streams;
Wood and Tikchik Lakes in U-shaped valleys
Swift, braided streams and rivers, some with
glacial headwaters; Lake Clark and other large
lakes in glaciated valleys
Large lakes associated with ice-carved valleys and
terminal moraines; glacially fed lake tributaries
Glacial moraines and ponds in eastern part of
region; upper reaches of the Nushagak and
Mulchatna Rivers
Morainal and thaw lakes; western half of Iliamna
Lake; Kvichak, Alagnak, Nushagak, Nuyakuk, and
Mulchatna River mainstems
Bristol Bay Assessment
3-3
January 2014
-------
Chapter 3
Region
Figure 3-1. Hydrologic landscapes within the Nushagak and Kvichak River watersheds, as defined
by physiographic division and climate class. Physiographic divisions (Wahrhaftig 1965) are classified
as Ahklun Mountains, Nushagak-Bristol Bay Lowland, Aleutian Range, Nushagak-Big River Hills, and
Southern Alaska Range. Climate classes (Feddema 2005) were defined as very wet, wet, moist, dry,
and semiarid, and calculated using 30-year (1971-2000) mean annual precipitation averages from
the Scenarios Network for Alaska and Arctic Planning data (SNAP 2012). Points labeled A through H
indicate approximate locations where photos in Figure 3-8 were taken.
Cook Inlet
Bristol Bay
Approximate Pebble Deposit Location
Towns and Villages
Watershed Boundary
N
A
25
25
50
] Kilometers
50
U Miles
Ahklun Mountains
•
•
Moist
Wet
Very Wet
Nushagak-Big River Hills
•
Dry
Moist
Wet
Southern Alaska Range
•
•
•
•
Semiarid
Dry
Moist
Wet
Very Wet
Nushagak-Bristol Bay Lowlands
Dry
Moist
Wet
• Very Wet
Aleutian Range
•
•
•
Moist
Wet
Very Wet
Bristol Bay Assessment
3-4
January 2014
-------
Chapter 3
Region
Figure 3-2. Distribution of mean annual precipitation (mm) across the Nushagak and Kvichak River
watersheds, 1971 to 2000 (SNAP 2012).
KVICHAK J&
rf—'""^S.^l'W
ulltnaR've> • PortAlsworth
•Nondalton
Pedro B
Iliamna
Cook Inlet
Bristol Bay
11
A
0 25
25
50
] Kilometers
50
] Miles
Precipitation
3,725 mm/yr
2,025 mm/yr
325 mm/yr
~j[ Approximate Pebble Deposit Location
• Towns and Villages
I I Watershed Boundary
Bristol Bay Assessment
3-5
January 2014
-------
Chapter 3
Region
Figure 3-3. Generalized geology of the Bristol Bay watershed (adapted from Selkregg 1974).
"A" Approximate Pebble Deposit Location
• Towns and Villages
| | Watershed Boundary
Generalized Geology
Moraine and Drift
Glacio lacustrine
Glaciofluvial
Alluvial
Coastal
| Eolian
Undifferentiated
Quaternary Volcanics
Intrusives
Tertiary
| Jurassic to Cretaceous
Late Paleozoic to Middle Mesozoic
Triassic to Early Jurassic
| Paleozoic and Older
Glacier
Lake
f ?V^
giak ^ "J*-* KVICHApr Cooh/n/et
lanokolak'
NORTWALASK/UPENINSULA
N
A
0 50 100
0 50 100
] Miles
Bristol Bay Assessment
3-6
January 2014
-------
Chapter 3
Region
Figure 3-4. Occurrence of permafrost in the Bristol Bay watershed (adapted from Selkregg 1974).
Mountainous
Isolated masses of permafrost
Lowland
Thick to thin permafrost in fine-grained sediments
Isolated deep relicts or shallow lenses in fine-grained sediments
Permafrost free; coarse-grained sediments
Permafrost free except for few isolated masses
Glacier
Lake
if Approximate Pebble Deposit Location
• Towns and Villages
| | Watershed Boundary
KVICHAK ( Cook Inlet
NORTM ALASKAPENINSULA
N
A
50 100
] Kilometers
50 100
] Miles
Bristol Bay Assessment
3-7
January 2014
-------
Chapter 3
Region
Figure 3-5. Dominant soils in the Bristol Bay watershed (adapted from Selkregg 1974).
Entisols
Well-drained loamy or gravelly gray soils
| Well-drained soils in stratified materials on floodplains and low terraces
Inceptisols
Poorly drained soils with peaty surface layer; shallow permafrost table
Poorly drained soils; shallow to deep permafrost table
Well-drained dark soils formed in fine volcanic ash
Well-drained dark soils formed in fine volcanic ash; shallow bedrock
Well-drained soils formed in dominantly coarse volcanic ash or in shallow ash over other material
| Well-drained soils with dark, acid surface layer; shallow bedrock
^^| Well-drained thin soils with dark, acid surface layer; deep permafrost table
Spodosols
Well-drained thin strongly acid soils; deep permafrost table
Well-drained strongly acid soils; very dark subsoil
Histosols
Other
Poorly drained fibrous peat; freezes in winter
Poorly drained fibrous peat; shallow permafrost table
Poorly drained fibrous peat; lenses of volcanic ash or alluvial material; seldom freezes deeply
Very steep rocky or ice-covered la nd
Fresh volcanic ash or cinders; little or no vegetation
Cook Inlet
ENINSULA
"K" Approximate Pebble Deposit Location
• Towns and Villages
| Watershed Boundary
N
A
0 50 100
0 50 1OO
] Miles
Bristol Bay Assessment
3-8
January 2014
-------
Chapter 3
Region
Figure 3-6. Erosion potential in the Bristol Bay watershed (adapted from Selkregg 1974).
"A" Approximate Pebble Deposit Location
• Towns and Villages
I | Watershed Boundary
Erosion Potential
Low
Low-Medium
Medium
Medium-High
^B High
Undetermined
KVICHAK { Cook Inlet
N
A
0 50 100
0 50 100
] Miles
Bristol Bay Assessment
3-9
January 2014
-------
Chapter 3
Region
Figure 3-7. Dominant vegetation in the Bristol Bay watershed (adapted from Selkregg 1974)
« Approximate Pebble Deposit Location
' Towns and Villages
Watershed Boundary
Dominant Vegetation
| Coastal Western Hemlock-Sitka Spruce Forest
| Bottomland Spruce-Poplar Forest
Upland Spruce-Hardwood Forest
Lowland Spruce-Hardwood Forest
High Brush
Moist Tundra
Wet Tundra
Alpine Tundra and Barren Ground
KVICHAK ( c°°* '"let
NORTWALASK/UPENINSULA
N
A
0 50 100
0 50 100
] Miles
Bristol Bay Assessment
3-10
January 2014
-------
Chapter 3
Region
Figure 3-8. Physiographic divisions of the Nushagak and Kvichak River watersheds of Bristol Bay.
The Nushagak and Kvichak River watersheds contain a wide range of aquatic habitats within five
distinct physiographic divisions; see Figure 3-1 for a map of these divisions and the general location
where each photo was taken. All photos taken between August 2003 and August 2013, courtesy of
Michael Wiedmer.
Coastal plain south of the lower Mushagak River, Nushagak-Bristol Tributary to Nistilik Lake in the upper Nushagak River watershed
Bay Lowland division Ahklun Mountains division
Klutuk Creek in the lower Nushagak River watershed, western
Nushagak-Bristol Bay Lowland division
Confluence of the Upper Nushagak River and the Nuyakuk Rtver,
Nushagak-Bristol Bay Lowland division
Source of the Muichatna River, Southern Alaska Range division
Battle Lake outlet, Aleutian Range division
Lake Clark, Southern Alaska Rangedivision of the upper Kvichak River
watershed
Nushaga|c_Big River m divisjon
Bristol Bay Assessment
3-11
January 2014
-------
Chapter 3 Region
The Nushagak-Big River Hills physiographic division consists largely of rounded ridges that have
moderate elevations and broad, gentle slopes and broad, flat or gently sloping valleys (Table 3-1, Figure
3-1) (Wahrhaftig 1965, Selkregg 1974). Major geologic formations include graywacke, argillite,
conglomerate, and greenstone flows (Figure 3-3). No modern glaciers are present, but glacial drift and
moraines are common throughout lower elevations and colluvium and alluvium mantle higher
elevations. The Nushagak River headwaters are the only part of the Nushagak and Kvichak River
watersheds that have not been glaciated. In most of this division falling within the Nushagak and
Kvichak River watersheds, permafrost is found only in isolated masses or lenses (Figure 3-4). Soils
throughout the division are typically shallow, occur in well-drained to poorly drained conditions, and
have medium erosion potential (Figures 3-5 and 3-6). Rivers in the Mulchatna and Newhalen River
systems originate from glaciers in the Southern Alaska Range. Sediment from these glaciers is trapped in
large lakes, providing clearer water for downstream reaches.
The Pebble deposit is located in the eastern portion of the Nushagak-Big River Hills and is heavily
influenced by past glaciation (PLP 2011: Chapter 3). At various times, Pleistocene glaciers blocked the
South Fork Koktuli River, the North Fork Koktuli River, and Upper Talarik Creek, the three tributaries
draining the Pebble deposit area (Figure 2-5). Unconsolidated glacial deposits, ranging from a few to
several tens of meters in thickness, cover most of the area's lower elevations (Detterman and Reed
1973). All three of the stream valleys in the Pebble deposit area have extensive glacial sand and gravel
deposits (PLP 2011: Chapter 8). Based on studies in the Pebble area, the Pebble Limited Partnership
(PLP) (2011) concluded that the presence of permeable shallow aquifers, upward hydraulic gradients,
and strong local relief indicate that local and intermediate groundwater flow systems dominate regional
groundwater flow systems. Further, PLP (2011) noted the presence of many local, cross-cutting faults
with high hydraulic conductivities in the Pebble deposit area.
The Nushagak-Bristol Bay Lowland physiographic division (Table 3-1, Figure 3-1) is mantled with
glacial drift and moraine deposits up to hundreds of meters deep, forming a rolling landscape with low
local relief (15 to 75 m) and maximum elevations of 90 to 150 m near the transitions from the lowland
to adjacent mountains or hills (Wahrhaftig 1965, Detterman 1986, Lea et al. 1991, Stilwell and Kaufman
1996). Arc-shaped bands of morainal deposits ranging from 1.6 to 8 km wide enclose Iliamna Lake and
are frequent in the lowlands between the Nushagak River and the Ahklun Mountains division
(Figure 3-3). Steep outliers of the Wood River Mountains in the Ahklun Mountains physiographic
division arise from the western part of the lowland. A small area with sand dunes occurs east of the
Nushagak River (Lea and Waythomas 1990). Glacial drift is coarser near the mountains because of high
amounts of outwash and grades to fine sand along the coast (Wahrhaftig 1965). The remainder of the
lowland is dominated by low-relief (less than 20 m), rolling expanses of tundra underlain by Holocene
peat and wind-born deposits (Lea et al. 1991). Glaciers do not occur today in the Nushagak-Bristol Bay
Lowland division, and permafrost is sporadic or absent (Figure 3-4) (Wahrhaftig 1965). Morainal and
thaw lakes are common, and mainstem rivers draining this area exhibit high channel complexity
(Figure 3-8). Poorly drained soils dominate in the southern portions, whereas well-drained soils
dominate across the remainder of the physiographic division (Figure 3-5). Soil erosion potential is
Bristol Bay Assessment 312 January 2014
-------
Chapter 3 Region
moderate throughout the area (Figure 3-6). Extensive dwarf scrub communities occur on relatively well-
drained soils, and moist and wet tundra communities cover large areas as well (Figure 3-7) (Selkregg
1974, Gallant et al. 1995).
3.2 Hydrologic Landscapes
To better evaluate the influence of inherent river basin attributes on streamflows and thus fish
populations, we used the physiographic divisions discussed above to define different hydrologic
landscapes across the Nushagak and Kvichak River watersheds. These landscapes can be considered
hydrologic building blocks, in that they provide a broad-scale approach to spatially characterizing
climate and watershed factors controlling the amount, timing, and flowpaths of water within the
watersheds (Winter 2001).
We defined hydrologic landscapes by calculating water surplus (precipitation minus potential
evapotranspiration) across the basins in each of the five physiographic divisions, using Scenarios
Network for Alaska and Arctic Planning (SNAP) data (SNAP 2012) and procedures outlined by Feddema
(2005). Feddema (2005) defined six annual climate classes ranging from very wet to arid conditions.
The very wet, wet, and moist classes have an annual water surplus, whereas the dry, semi-arid, and arid
classes have an annual water deficit. Combining these climate classes with the physiographic divisions
(Section 3.1), we identified 18 different hydrologic landscapes across the Nushagak and Kvichak River
watersheds (Table 3-2, Figure 3-1), which represent the range of hydrologic characteristics across the
region.
3.3 Groundwater Exchange and Flow Stability
A key aspect of the Bristol Bay watershed's aquatic habitats is the importance of groundwater exchange.
Because salmon rely on clean, cold water flowing over and upwelling and downwelling through porous
gravels for spawning, egg incubation, and rearing (Bjornn and Reiser 1991), areas of groundwater
exchange create high-quality salmon habitat (Appendix A). For example, densities of beach spawning
sockeye salmon in the Wood River watershed were highest at sites with strong groundwater upwelling
and zero at sites with no upwelling (Burgner 1991). Portions of the Nushagak-Bristol Bay Lowland and
Nushagak-Big River Hills physiographic divisions, including the Pebble deposit area, contain coarse-
textured glacial drift with abundant, high-permeability gravels and extensive connectivity between
surface waters and groundwater (Figures 3-3, 3-4, and 3-9). Abundant wetlands and small ponds also
contribute disproportionately to groundwater recharge (Rains 2011). This strong connection between
groundwater and surface waters helps to moderate water temperatures and streamflows. For example,
groundwater contributions that maintain water temperatures above 0°C are critical for maintaining
winter refugia in streams that might otherwise freeze (Power et al. 1999).
Bristol Bay Assessment 313 January 2014
-------
Chapter 3
Region
Table 3-2. Distribution of hydrologic landscapes in the Nushagak and Kvichak River watersheds. Values represent percentage of total area
in the two watersheds.
Physiographic Division
Climate Class
Ahklun Mountains
V
W
M
Southern Alaska Range
V
W
M
D
Aleutian Range
V
W
M
Nushagak-Big River Hills
V
W
M
D
Nushagak-Bristol
Bay Lowland
V
W
M
Nushagak River Watershed
Nushagak River (whole watershed)
Nushagak River at Ekwoka
Nuyakuk River
Mulchatna River
Nushagak River at Mulchatna River
Koktuli River
South Fork Koktuli Riverb
North Fork Koktuli River0
7
4
19
8
-
16
9
43
18
-
1
2
1
-
1
2
4
-
-
2
3
7
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
25
40
3
53
30
99
100
100
9
14
22
9
-
-
-
-
-
1
-
-
24
27
32
14
35
1
-
15
1
-
-
-
Kvichak River Watershed
Kvichak River (whole watershed)
Kvichak River at lgiugigd
Kaskanak Creek near Igiugig8
Iliamna River near Pedro Bay'
Upper Talarik Creeks
-
-
-
-
-
-
-
-
-
-
-
-
16
25
94
-
13
20
6
-
8
12
-
-
1
2
-
-
2
-
-
-
11
-
-
-
2
6
-
-
-
-
-
-
7
10
21
-
100
7
11
-
-
-
1
-
-
-
-
-
-
3
-
28
-
-
28
11
50
-
-
Notes:
Dashes (-) indicate hydrologic landscapes that are not found in that portion of the Nushagak or Kvichak River watersheds. Climate classes are defined as very wet (V), wet (W), moist (M), and dry (D)
according to Feddema (2005); no semi-arid or arid climates are found in the region.
a USGS gage 15302500.
b USGS gage 15302200.
c USGS gage 15302250.
d USGS gage 15300500.
e USGS gage 15302520.
' USGS gage 15300300.
f- USGS gage 15300250.
Bristol Bay Assessment
3-14
January 2014
-------
Chapter 3
Region
Figure 3-9. Groundwater resources in the Bristol Bay watershed (adapted from Selkregg 1974). Yields are presented in gallons per minute.
Groundwater Resources
Unconsolidated Deposits; mostly sand and gravel (100-1,000 gal/min yield)
Unconsolidated Deposits; mostly sand and gravel, silt and clay (10-100 gal/min yield)
Unconsolidated Deposits; mostly sand and gravel, silt and clay (0-10 gal/min yield)
Bedrock (0-10 gal/min yield)
Approximate Pebble Deposit Location
Towns and Villages
Watershed Boundary
Cook Inlet
NINSULA
N
A
50
50
100
] Kilometers
100
] Miles
Bristol Bay Assessment
3-15
January 2014
-------
Chapter 3 Region
These groundwater contributions to streamflow, along with the influence of large and small lakes,
support flows in the region's streams and rivers that are more stable than those typically observed in
many other salmon streams (e.g., in the Pacific Northwest or southeastern Alaska). Greater groundwater
contributions to streams result in more moderated streamflow regimes with lower peak flows and
higher base flows, creating a less temporally variable hydraulic environment. The lower mainstem
Nushagak and Kvichak Rivers illustrate this tendency toward moderated, consistent streamflows
(Figure 3-10). Coarse-textured glacial drift in the Kaskanak and Upper Talarik Creek drainages promotes
high groundwater contributions to these streams, resulting in stable flows through much of the year
(Figure 3-10). High baseflow in the Nushagak River also is consistent with increased interactions
between surface water and groundwater, as water flows from the Southern Alaska Range, Ahklun
Mountains, and Nushagak-Big River Hills into the coarse-textured glacial drift of the Nushagak-Bristol
Bay Lowland (Figure 3-10).
Water storage in upstream lakes plays a role in flow stabilization, as well. For example, in the Kvichak
River watershed, Iliamna Lake dampens high flows from the Iliamna and Newhalen Rivers before they
reach the mainstem. The attenuating effect of upstream lakes on streamflow is also evident in the
Newhalen River, located downstream of Lake Clark (Figure 3-10).
3.4 Quantity and Diversity of Aquatic Habitats
Differences in hydrology, geology, and climate across the Bristol Bay watershed interact to create the
region's diverse hydrologic landscapes (Table 3-2, Figure 3-1) and ultimately shape the quantity, quality,
diversity, and distribution of aquatic habitats throughout the watershed. These diverse habitats, in
conjunction with the enhanced ecosystem productivity associated with anadromous salmon runs,
support a high level of biological complexity that contributes to the environmental integrity and
resilience of the watershed's ecosystems (Schindler et al. 2010, Ruff et al. 2011, Lisi et al. 2013).
In general, conditions in the Bristol Bay watershed are highly favorable for Pacific salmon. The
Nushagak and Kvichak River watersheds encompass an abundant and diverse array of aquatic habitats
and support a diverse salmonid assemblage (Section 5.2). Freshwater habitats range from headwater
streams to braided rivers, small ponds to large lakes, side channels to off-channel alcoves. These
watersheds contain over 54,000 km of streams, 14% of which have been documented as anadromous
fish streams (Johnson and Blanche 2012). This percentage is likely a significant underestimate of the
actual extent of anadromous waters across the watersheds (Box 7-1, Appendix A).
Bristol Bay Assessment 316 January 2014
-------
Chapter 3
Region
Figure 3-10. Mean monthly runoff for selected streams and rivers in the Nushagak and Kvichak
River watersheds. USGS gages and dates used to generate each line: A. Nushagak River watershed:
Nushagak River (15302500, Oct 1977-Sep 1993); Nuyakuk River (15302000, Jun 1953-Sep 2010);
North Fork (NF) Koktuli River (15302250, Sep 2004-Sep 2010); South Fork (SF) Koktuli River
(15302200, Sep 2004-Sep 2010). B. Kvichak River watershed: Kvichak River (15300500, Aug
1967-Sep 1987); Kaskanak Creek (15300520, Jun 2008-Sep 2011); Iliamna River (15300300, Jun
1996-Sep 2010); Upper Talarik Creek (15300250, Sep 2004-Sep 2010); Newhalen River
(15300000, Jul 1951-Sep 1986).
600
500
400
3= 300
o
200
100
-*-Nushagak River
Nuyakuk River
NF Koktuli River
SF Koktuli River
Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep
B
600
-•-Kvichak River
-•-Kaskanak Creek
-*-lliamna River
-•-Upper Talarik Creek
Newhalen River
Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep
Bristol Bay Assessment
3-17
January 2014
-------
Chapter 3 Region
Lakes and associated tributary and outlet streams are key spawning and rearing areas for sockeye
salmon. Lakes cover relatively high percentages of watershed area in the Bristol Bay region: 7.9% for the
entire Bristol Bay watershed area and 13.7% for the Kvichak River watershed (RAP 2011). In other
North Pacific river systems supporting sockeye salmon populations, from northern Russia to western
North America, these values tend to be much lower (e.g., 0.2 to 2.9%) (RAP 2011). Relatively low
watershed elevations (especially in the extensive Nushagak-Bristol Bay Lowland physiographic
division) and the absence of artificial barriers to migration (e.g., dams and roads) mean that not only are
streams, lakes, and other aquatic habitats abundant in the Bristol Bay region, but they also tend to be
accessible to anadromous salmonids. With very few exceptions, all major lakes in the watershed are
accessible to anadromous salmon (Appendix A). Lakes and ponds also play a key role in groundwater
dynamics and flow stability (Section 3.3).
Overall physical habitat complexity in the Bristol Bay watershed is higher than in many other systems
supporting sockeye salmon populations. Of 1,509 North Pacific Rim watersheds, the Kvichak, Wood, and
Nushagak (exclusive of Wood) Rivers ranked third, fourth, and forty-fourth, respectively, in physical
habitat complexity, based on an index that included variables such as lake coverage, stream junction
density, floodplain elevation and density, and human footprint (Luck et al. 2010, RAP 2011).
3.4.1 Stream Reach Characterization: Attributes
To characterize the stream and river habitats in the Nushagak and Kvichak River watersheds, we
described stream and river valley attributes for each of the 52,277 stream and river reaches (54,427
km) in the Nushagak and Kvichak River watersheds documented in the National Hydrography Dataset
(NHD) (USGS 2012). We excluded another 27,186 reaches (7,936 km) for which we could not identify
reach-specific drainage areas from the analysis. For each reach, we estimated the mean annual
streamflow (m3/s), mean channel gradient (%), and percent of flatland in the contributing watershed
lowland (% flat); each attribute is described in detail in the following sections. These attributes were
selected because they represent fundamental aspects of the physical and geomorphic settings in
streams, providing context for stream and river habitat development and subsequent fish habitat
suitability (Burnett et al. 2007). It also was feasible to obtain these attributes for the entire area given
available data. These attributes have been used to model habitat suitability for salmon at large scales, for
example via intrinsic potential modeling (Burnett et al. 2007, Shallin Busch et al. 2011). We did not
develop intrinsic potential models for salmon species in this assessment, as that effort would require
multiple years of field data collection for model validation and testing and those data are not currently
available. However, our characterization results do provide insights into the distribution of broad-scale
habitat conditions within the watersheds, and could provide the basis for future intrinsic potential
model development.
3.4.1.1 Channel Gradient
Channel gradient broadly characterizes channel steepness and geomorphic form. Channel gradient and
associated aspects of channel morphology influence channel capacity to transport sediment, affecting
channel response to disturbance (Montgomery and Buffington 1997). Channel morphology can strongly
Bristol Bay Assessment 318 January 2014
-------
Chapter 3 Region
influence suitability for salmon rearing and spawning. Specific substrate and hydraulic requirements
vary slightly by species (Appendix A), but stream-spawning salmon generally require relatively clean
gravel-sized substrates with interstitial flow, and sufficient bed stability to allow eggs to incubate in
place for months prior to fry emergence (Quinn 2005).
Montgomery and Buffington (1997) proposed a process-based classification of mountain streams. Field
data from their study indicated that gradients estimated by digital elevation models (OEMs) provide a
useful predictor of channel morphology. We estimated the channel gradient of each stream reach in the
Nushagak and Kvichak River watersheds by assessing the gradient of correlated flowpaths across a
30-m-cell National Elevation Dataset DEM (Gesch et al. 2002, Gesch 2007, USGS 2013) (Box 3-1). We
adapted the classification scheme put forth by Montgomery and Buffington (1997) to define four
gradient classes and predicted channel morphologies for stream reaches at different watershed scales.
• Less than 1%, dune-ripple or pool-riffle morphology.
• At least 1% and less than 3%, plane-bed morphology.
• At least 3% and less than 8%, step-pool morphology.
• At least 8%, cascade morphology.
The substrate and hydraulic conditions required by stream-spawning salmon are most frequently met in
stream channels with gradients less than 3% (Montgomery et al. 1999). At the lowest gradients, the
channel's capacity to transport fine sediments will be low and substrates may be dominated by sands
and other fines, providing suboptimal salmon spawning habitat. A notable exception to this generality
occurs in low-gradient, off-channel habitats and ponds that may be dominated by fine sediments but
that contain areas of upwelling. These areas are used by riverine-spawning (Eiler et al. 1992) and pond-
spawning (Quinn et al. 2012) sockeye salmon. At gradients above 3%, channels develop step-pool or
cascade morphologies and the size, stability, and frequency of pockets of suitable spawning substrates
decrease substantially (Montgomery and Buffington 1997). In the Bristol Bay region, gradients of
productive stream reaches for salmon are typically less than 3%, with gradients less than 1%
characterizing the most productive reaches; these habitats include lake outlets and lower tributary
reaches, and most of the major spawning reaches and tributaries of the Nushagak and Kvichak River
watersheds (Figures 3-11 and 3-12) (Demory et al. 1964). We note, however, that low-gradient
watersheds in the coastal plain region of the Nushagak-Bristol Bay Lowland that lack upland
headwaters are generally not productive salmon habitats. These streams tend to have lower dissolved
oxygen levels, be characterized by fine-textured substrates with high proportions of organic material,
and may lack substrates coarser than sand, presumably due to lack of higher-gradient source areas for
gravel recruitment (ADF&G 2012, Wiedmer pers. comm.).
Environmental conditions determining suitability for juvenile salmon and adult resident salmonids
(e.g., resident Dolly Varden; Box 2-3) are also influenced by gradient. Fish movement can be restricted
by the high water velocities and frequent drops found in streams with gradients exceeding 12%,
although Dolly Varden have been found at gradients exceeding 15% in southeast Alaska streams
Bristol Bay Assessment 319 January 2014
-------
Chapter 3 Region
(Wissmar et al. 2010). Gradient and channel roughness also influence the distribution of water velocities
and hydraulic conditions in streams, influencing food delivery rates and availability and subsequent
energetic demands of drift feeding fish (Hughes and Dill 1990).
BOX 3-1. METHODS FOR CHARACTERIZING CHANNEL GRADIENT
The valley gradient of each stream reach in the Nushagak and Kvichak River watersheds was estimated by
assessing the gradient of correlated flowpaths along across a 30-m cell National Elevation Dataset digital
elevation model (DEM) (Gesch et al. 2002, Gesch 2007, USGS 2013). We found the measured gradient of
the National Hydrography Dataset (NHD)flowlines (based on the elevation of the underlying DEM) was not an
accurate representation of channel gradient because of inconsistencies between the mapped streams and
rivers in the NHD and the topography described by the DEM. Channel traces in the NHD did not reliably follow
the valley floor, and upslope traces and misalignment with the DEM resulted in inaccurate measures of
stream gradients and sampled elevations.
We determined that the gradient of streams in a drainage network described by a flow analysis across the
DEM would more accurately represent channel morphology given the data available. The drainage network of
the DEM paralleled the network of NHD flowlines, but included or excluded some small tributaries and lacked
the sinuosity mapped in the NHD.
Gradients of flowlines across the DEM were determined using the hydrology tools of the Spatial Analyst
extension of ArcGIS. First, the hydraulic network was generated based on the topography of the NHD DEM.
Generation of the hydraulic network involved the following tools:
• Fill. Sinks in the DEM were filled so that continuous flowpaths could be described.
• Flow direction. The steepest path or flow direction was determined from each cell in the DEM.
• Flow accumulation. Based on the direction of flow, the total number of cells, or receiving area for each
cell in the DEM, was determined.
• Reclassify. A threshold value of 0.25 km2 was applied to the total receiving area output from the previous
step to distinguish streams from non-streams.
• Stream link. The resulting network was processed to assign unique identifiers to each link in the drainage
network.
To determine the gradient of each stream link in the drainage network, and to generate geometry that could
assign these values to the reaches of the NHD flowlines, the following tools were used:
• Extract by mask. Elevation values underlying the drainage network were isolated from the DEM so that
cross-valley slopes would not be measured when determining gradient.
• Slope. Gradient along the drainage network was measured between each cell of the isolated drainage
network DEM. The drainage DEM confined the slope measures to the flowpath of the drainage network,
providing an estimate of stream gradient at each 30-m cell.
• Watershed. The output of the Stream Link tool (see above) and the results of the flow direction analysis
were used to delineate the drainage basin for each stream link. This geometry was then used to transfer
gradient values to the NHD stream reaches.
• Zonal statistics. In the drainage basin for each stream segment, the average gradient was determined for
all cells with values (i.e., a mean gradient of the stream segment). Mean gradient values were then
assigned to the drainage basin geometry.
• Zonal statistics as table. The mean gradient for each drainage basin was used to calculate the channel
gradient for each NHD flowline. This tool measured the length-weighted mean of the gradients for each
reach (as defined by the NHD Reach Code attribute) from the means calculated for each drainage basin.
Typically, the NHD flowlines occupied no more than two drainage basins. The resulting gradient estimates
were appended to the table of NHD flowlines.
Bristol Bay Assessment 320 January 2014
-------
Chapter 3
Region
Figure 3-11. Examples of different stream size and gradient classes in the Nushagak and Kvichak
River watersheds.
<1% gradient headwater stream, Kaskanak Creek drainage, Kvichak
River watershed. Photo: Michael Wiedmer, ADF&G, 8/27/03
5% gradient headwater stream, North Fork Koktuli River drainage,
Nushagak River watershed. Photo: Michael Wiedmer, ADF&G, 8/19/06
1.2% gradient medium stream, Koktuli River drainage, Nushagak
River watershed. Photo: Michael Wiedmer, ADF&G, 8/24/03
3% gradient medium stream, Mulchatna River drainage, Nushagak
River watershed. Photo: Michael Wiedmer, USGS, 8/19/10
<1% gradient small river, South Fork Koktuli River, Nushagak River <1% gradient large river, Nushagak River between Portage Creek and
watershed. Photo: Joe Buckwalter, ADF&G, 8/17/06 Ekwok. Photo: Michael Wiedmer, ADF&G, 8/12/06
Bristol Bay Assessment
3-21
January 2014
-------
Chapter 3
Region
Figure 3-12. Channel gradient classes in the Nushagak and Kvichak River watersheds. Valley
gradient was assessed by measuring drainage channel slope across the watersheds' landscapes (Box
3-1).
Cook Inlet
Bristol Bay
"if Approximate Pebble Deposit Location
• Towns and Villages
I Watershed Boundary
< 1% Gradient
1-3% Gradient
3 - 8% Gradient
> 8% Gradient
N
A
25
25
50
] Kilometers
50
H Miles
Bristol Bay Assessment
3-22
January 2014
-------
Chapter 3 Region
3.4.1.2 Mean Annual Streamflow
Mean annual streamflow is a metric of stream size, an important determinant of available habitat space
(capacity) for stream fishes. The relationship between mean annual streamflow and habitat capacity for
rearing juvenile salmon can vary with streamflow regime and other limiting factors, but is generally
positive when other factors are not constraining.
Mean annual streamflow for each stream reach within the Nushagak and Kvichak River watersheds was
estimated using regression equations for the prediction of mean annual streamflow, based on drainage
area and historical mean annual precipitation in southwestern Alaska (Parks and Madison 1985) (Box
3-2). We defined four classes of stream size based on these mean annual streamflow calculations.
• Small headwater streams (less than 0.15 m3/s), including many of the tributaries of the South and
North Fork Koktuli Rivers and Upper Talarik Creek.
• Medium streams (0.15 to 2.8 m3/s), including the upper reaches and larger tributaries of the South
and North Fork Koktuli Rivers and Upper Talarik Creek.
• Small rivers (2.8 to 28 m3/s), including the middle to lower portions of South and North Fork
Koktuli Rivers, and Upper Talarik Creek, and the mainstem Koktuli River.
• Large rivers (greater than 28 m3/s), including the Mulchatna River below the confluence with the
Koktuli River, the Newhalen River, and other larger rivers.
All five species of Pacific salmon present in the Bristol Bay region use portions of large and small rivers
and medium streams for migration, spawning, and/or rearing habitat. Research in the Wood River
system suggests that larger stream sizes allow multiple salmon species to coexist, perhaps due to habitat
partitioning made possible by increased space and habitat diversity (Pess et al. 2013). Salmon also use
small streams in the Bristol Bay region for spawning and rearing, but use of these habitats may be
constrained by shallow depths, insufficient streamflow to allow passage, the unavailability of open
water in winter, or other limitations related to stream size.
Salmonid species differ in their propensities for small streams. Dolly Varden have been documented
using all stream sizes, including some of the smallest channels. Of the Pacific salmon species, coho
salmon are most likely to use small streams for spawning and rearing, and have been observed in many
of the smaller streams near the Pebble and other deposits. Larger-bodied Chinook salmon adults are less
likely to access smaller streams for spawning (Quinn 2005). However, juvenile Chinook salmon are
observed in small tributaries where spawning has not been documented.
Bristol Bay Assessment 323 January 2014
-------
Chapter 3 Region
BOX 3-2. METHODS FOR CHARACTERIZING MEAN ANNUAL STREAMFLOW
Mean annual streamflow for each stream reach in the Nushagak and Kvichak River watersheds was
estimated using regression equations, based on drainage area and historical mean annual precipitation
data in southwestern Alaska (Parks and Madison 1985). Total drainage area was determined for reaches
alongthe National Hydrography Dataset (NHD) flowlines by developinga drainage-corrected digital elevation
model (DEM) based on the National Elevation Dataset (NED). Although the underlying topography and
catchments described by the NED remained the same, the elevations underlying the NHD flowlines and in
their immediate vicinity were lowered and smoothed such that runoff conformed to the geometry of the NHD
flowlines.
Using the drainage-corrected DEM, we estimated total catchment area above any location in the drainage
network. The NED DEM was corrected to better conform to the NHD flowlines and drainage areas were
calculated using the following tools of the ArcHydro and Spatial Analyst tools of the ArcGIS suite:
• DEM reconditioning. The elevations of the DEM were altered alongthe NHD flowlines and in their
immediate vicinity. Parameters used for this tool were a 10-m reduction of elevations along the flowline, a
5-cell (150-m)-wide transition zone on either side of the flowline, and a post-process 1-km reduction in
elevations alongthe flowlines. The initial elevation reduction and transition width were found to
adequately capture flows and maintain those flows within the channel geometry. The post-processing
adjustment is a more arbitrary value intended to confine flows to the channels once captured.
• Fill. Sinks in the reconditioned DEM were filled so that continuous flowpaths could be described.
• Flow direction. The steepest path or flow direction was determined from each cell in the DEM.
• Flow accumulation (drainage area). Based on the direction of flow, the total number of cells, or receiving
area for each cell in the DEM, was determined. These values were multiplied by 0.0009 to convert the
area of each cell (900 m2) to square kilometers.
• Flow accumulation (accumulated precipitation). Due to variation in precipitation patterns across the
study area, the average accumulated precipitation was calculated by using the flow accumulation tool
with a weight assigned to each cell based on the average annual precipitation data for 1971 to 2001
(SNAP 2012). The result was divided by the total number of cells accumulated at each location on the
grid to determine the average accumulated annual precipitation.
The output drainage area raster and raster coverage of average annual precipitation were used as inputs for
the mean annual streamflow regression equation developed by Parks and Madison (1985) for southwestern
Alaska:
Q = (10-1.38)*(DA°-98)*(P1.13)
where Q is mean annual flow in cubic feet per second, DA is drainage basin area in square miles, and P is
mean annual precipitation in inches per year. We used the median mean annual streamflow value from the
cells within the drainage network that corresponded to each NHD flowline as the estimate of mean annual
streamflow for the stream segment.
3.4.1.3 Proportion of Flatland in Lowland
Stream channels in mountainous and foothill terrain are laterally constrained by their valley walls to
varying degrees. Degree of channel constraint influences channel form, including the development of off-
channel habitats, variability in local channel gradients, and hydraulic conditions during over-bank flows.
Unconstrained channels generally have higher complexity of channel habitat types and hydraulic
conditions and higher frequencies of off-channel habitats such as side channels, sloughs, and beaver
ponds. Such habitat complexity can be beneficial to salmon by providing a diversity of spawning and
rearing habitats throughout the year (Stanford et al. 2005).
Bristol Bay Assessment 324 January 2014
-------
Chapter 3 Region
To provide an index of the degree of channel constraint expected within each stream reach, we
estimated the percent of flatland (less than 1% slope) within lowland (area below median elevation) for
each stream reach's adjacent drainage basin (Box 3-3). Visual inspection of portions of the study area
where high-resolution aerial photographs were available showed that channels were typically
unconstrained when the proportion of flatland in lowland exceeded 5%. This threshold was used to
identify two classes:
• Less than 5% flatland in lowland, indicating reaches are constrained and have limited floodplain
area. These reaches are classified as having low or no floodplain potential.
• Greater than or equal to 5% flatland in lowland, indicating reaches are unconstrained and have high
likelihood for floodplain development. These reaches are classified as having floodplain potential.
In the Bristol Bay region, streams that are unconstrained and able to develop complex off-channel
habitats are more likely to provide a diversity of channel habitat types and hydraulic conditions,
creating favorable conditions, particularly for salmonid rearing. For Chinook and coho salmon, as well as
river-rearing sockeye salmon that may overwinter in streams, such habitats may be particularly
valuable. The percent flatland in lowland metric is not a perfect index of channel constraint, however.
Channels in flat lowlands such as the coastal Nushagak-Bristol Bay Lowlands physiographic division
(Figure 3-1) may actually be incised into fine-grained sediments with very little off-channel habitat
complexity. In the glacially worked landscapes of the Bristol Bay region, streams may be constrained by
relatively flat valley terraces and moraine deposits that are not distinguishable on the coarse-scale DEM
available for the region. Terraces are a common feature in portions of the region, but the degree to
which terrace constraint influences these results could not be determined from the existing DEM. In
steep, mountainous terrain, narrow valleys may occasionally allow for unconstrained stream channel
development across low-gradient floodplains, but these features are likely not always detected with the
DEM resolution currently employed for this effort.
3.4.2 Stream Reach Characterization: Results
We estimated the three stream-reach attributes discussed above in four geographically defined areas
that vary in scale and location (as described in Section 2.2.2).
• The Nushagak and Kvichak River watersheds (Scale 2).
• The mine scenario watersheds—that is, the South Fork Koktuli River, the North Fork Koktuli River,
and the Upper Talarik Creek watersheds (Scale 3).
• The streams lost to the Pebble 6.5 scenario footprint (Scale 4).
• The subwatersheds of the transportation corridor area (Scale 5).
In this section, we summarize results for the Nushagak and Kvichak River watersheds to broadly
characterize the region. Results for the other three geographic scales are reported later in the
assessment (Sections 7.2.1 and 10.2), where we evaluate potential impacts of large-scale mining.
Bristol Bay Assessment 325 January 2014
-------
Chapter 3 Region
BOX 3-3. METHODS FOR CHARACTERIZING PERCENT FLATLAND IN LOWLAND
The relative degree of channel constraint in the Nushagak and Kvichak River watersheds was estimated by
calculating the percent of flatland (<1% slope) within lowland (area below median elevation) in each stream
reach's adjacent drainage basin. These calculations included the delineation of drainage basins of the
drainage-corrected drainage network (developed for the mean annual streamflow analysis; see Box 3-2) as
well as elevation and slope analyses of the unaltered digital elevation model (DEM).
To establish the drainage basin geometry of the drainage-corrected flow analysis, the following Spatial
Analyst tools were applied within an ArcGIS workspace.
• Reclassify. A threshold value of 0.25 km2 was applied to the total receiving area output from the
drainage-corrected flow analysis to distinguish streams from non-streams.
• Stream link. The resulting network was processed to assign unique identifiers to each link in the
drainage network.
• Watershed. The output of the Stream Link tool (see above) and the results of the flow direction analysis
were used to delineate the drainage basin for each stream link. This geometry was used as the
geographic extent of analysis for each stream segment.
Areas of flatland and lowland were then identified for each drainage basin. The unaltered National Elevation
Dataset DEM was processed with the following Spatial Analyst tools from ArcGIS.
• Slope. The original (not drainage-corrected) DEM was analyzed to determine slope (%) across the extent
of the Nushagak and Kvichak River watersheds.
• Reclassify. A threshold value of 1% was applied to the slope analysis, and attributes were assigned
across the study area as meeting or not meetingthe flatland criteria.
• Zonal statistics. In the drainage basin for each stream segment, the minimum and maximum elevations
were determined usingthe Zonal Statistics tool. These values were used to identify the median elevation
for each watershed.
• Reclassify. The DEM was classified as meeting or not meetingthe lowland criteria based on results of the
previous step.
Finally, the percent flatland in lowland for each stream reach's drainage basin was calculated usingthe
following steps.
• Times. Areas of flatland outside of lowland areas were eliminated by multiplying the flatland and lowland
rasters. The flatland and lowland rasters used 1 and 0 values for true and false, respectively, so both
conditions were required to return a positive result for flatland in lowland.
• Zonal statistics. The total areas of lowland and flatland within lowland were calculated for each drainage
basin.
• Divide. The percent flatland in lowland was determined for each drainage basin by dividing the area of
flatland in lowland by the area of lowland in each drainage basin.
• Zonal statistics as table. The average value of percent flatland in lowland for each stream reach was
calculated and added to a table, which was then appended to the National Hydrography Dataset (NHD)
flowline data table. Although the mean statistic was used to ascertain these values for the NHD flowlines,
the flowlines typically had a one-to-one correlation with drainage basins, as the basins were based on the
drainage-corrected flow analysis.
We characterized 54,427 km of streams and 52,277 stream and river reaches in the Nushagak and
Kvichak River watersheds. Reach attributes reflected the hydrologic landscapes in which the reaches
occurred and upstream within each reach's drainage (Section 3.2). Relatively low-gradient stream
channels extend far up into the headwaters of the upper Mulchatna and Nushagak River watersheds
(Figure 3-12), allowing salmon to access headwater streams. High-gradient conditions are primarily
found in the headwaters of Lake Clark and Iliamna Lake tributaries and the headwaters of the Alagnak,
Bristol Bay Assessment 326 January 2014
-------
Chapter 3 Region
Wood, Kokwok, and Nuyakuk Rivers (Figure 3-12). Valley flatland is heavily concentrated in the
Nushagak-Bristol Bay Lowlands physiographic division and along the larger rivers, but includes
significant wider-valley reaches in the Nushagak-Big River Hills, Southern Alaska Range, and Aleutian
Range divisions (Figure 3-13).
The majority of stream channel length (75%) in the Nushagak and Kvichak River watersheds is
composed of low-gradient (less than 3%), medium and small (less than 2.8 m3/s mean annual
streamflow) streams (Table 3-3, Figures 3-12 and 3-14). The extent of flatland in valley lowlands is
strongly associated with gradient. For streams with less than 1% gradient, 55% have high floodplain
potential (i.e., greater than or equal to 5% flatland in lowland). In contrast, less than 5% of streams with
gradients greater than 1% have high floodplain potential. Stream reaches with greater than 3% gradient
were only found in landscapes where floodplain potential was low (i.e., less than or equal to 5% flatland
in lowland). Overall, these results reveal the high proportion of stream channels in these watersheds
that possess the broad geomorphic and hydrologic characteristics enabling the development of stream
and river habitats highly suitable for fishes such as Pacific salmon, Dolly Varden, and rainbow trout.
3.5 Water Quality
3.5.1 Water Chemistry
Water quality of streams near the Pebble deposit has been characterized extensively (PLP 2011,
Zamzow 2011). The streams draining the watersheds in the Pebble deposit area (Figure 2-5) are neutral
to slightly acidic, with low conductivity, hardness, dissolved solids, suspended solids, and dissolved
organic carbon (see Section 8.2.1.1 for more detailed discussion of water chemistry in streams draining
the mine scenario watersheds). In those respects, they are characteristic of undisturbed streams.
However, as would be expected for a metalliferous site, levels of sulfate and some metals (copper,
molybdenum, nickel, and zinc) are elevated, particularly in the South Fork Koktuli River. PLP (2011)
found that copper levels in some samples from the South Fork Koktuli River exceeded Alaska's chronic
water quality standard. However, most of the exceedances were in or close to the deposit and the
number and magnitude of exceedances decreased with distance downstream (PLP 2011: Figure 9.1-35,
60, 61, 65, and 66).
Bristol Bay Assessment 327 January 2014
-------
Chapter 3
Region
Figure 3-13. Likelihood of floodplain potential, as measured by the percent flat land in lowland
areas, for the Nushagak and Kvichak River watersheds. Flatland refers to land with less than 1%
slope; lowland areas are defined as areas below the midpoint elevation within the drainage basin of
each stream reach (Box 3-3).
Cook Inlet
Bristol Bay
« Approximate Pebble Deposit Location
* Towns and Villages
| Watershed Boundary
< 5% Flatland in Lowlands
5 -10% Flatland in Lowlands
10 - 25% Flatland in Lowlands
25-50% Flatland in Lowlands
> 50% Flatland in Lowlands
N
A
25
25
50
] Kilometers
50
D Miles
Bristol Bay Assessment
3-28
January 2014
-------
Chapter 3
Region
Figure 3-14. Stream size classes in the Nushagak and Kvichak River watersheds as determined by
mean annual streamflow. Mean annual streamflow for streams and rivers was estimated using
drainage area and mean annual precipitation (Box 3-2).
Cook Inlet
Bristol Bay
Approximate Pebble Deposit Location
Towns and Villages
Watershed Boundary
Small Headwater Stream (< 0.15 m3/s)
Medium Stream (0.15 - 2.8 m3/s)
Small River (2.8 - 28 m3/s)
Large River (> 28 m3/s)
I N
A
] Miles
Bristol Bay Assessment
3-29
January 2014
-------
Chapter 3
Region
Table 3-3. Proportion of stream channel length within the Nushagak and Kvichak River watersheds
classified according to stream size (based on mean annual streamflow in m3/s), channel gradient
(%), and floodplain potential (based on % flatland in lowland). Gray shading indicates proportions
greater than 5%; bold indicates proportions greater than 10%.
Stream Size
Small headwater streams3
Medium streams'5
Small rivers0
Large riversd
Gradient
<1%
FP
27%
20%
6%
2%
NFP
5%
3%
1%
0%
>1% and <3%
FP
3%
1%
0%
0%
NFP
13%
3%
0%
0%
>3% and <8%
FP
0%
0%
0%
0%
NFP
8%
2%
0%
0%
>8%
FP
0%
0%
0%
0%
NFP
3%
1%
0%
0%
Notes:
3 0-0.15 m3/s; most tributaries in the mine footprints.
b 0.15-2.8 m3/s; upper reaches and larger tributaries of the South Fork Koktuli, North Fork Koktuli, and Upper Talarik Creek.
c 2.8-28 m3/s; mid to lower portions of the South Fork Koktuli, North Fork Koktuli, and Upper Talarik Creek, including the mainstem Koktuli
River.
d >28 m3/s; the Mulchatna River below the Koktuli confluence, the Newhalen River, and other large rivers.
FP = high floodplain potential (>5% flatland in lowland); NFP = no or low floodplain potential (<5% flatland in lowland).
3.5.2 Water Temperature
Water temperature data (PLP 2011: Appendix 15.IE, Attachment 1) indicate significant spatial
variability in thermal regimes. Average monthly stream water temperatures in the Pebble deposit area
in July or August can range from 6°C to 16°C. Longitudinal profiles of temperature indicate that stream
temperatures in the Pebble deposit area do not uniformly increase with decreasing elevation (PLP
2011). This is often due to substantial inputs of cooler water from tributaries or groundwater (PLP
2011). Extensive glacially reworked deposits with high hydraulic conductivity allow for extensive
connectivity between groundwater and surface waters in the region (Power et al. 1999). This
groundwater-surface water connectivity has a strong influence on the hydrologic and thermal regimes
of streams in the Nushagak and Kvichak River watersheds, and provides a moderating influence against
both summer heat and winter cold extremes in stream reaches where this influence is sufficiently
strong. The range of spatial variability in temperatures in the Pebble deposit area (PLP 2011) is
consistent with streams influenced by a variety of thermal modifiers, including upstream lakes,
groundwater, or tributary contributions (Mellina et al. 2002, Armstrong et al. 2010).
3.6 Seismicity
The Alaska Earthquake Information Center and U.S. Geological Survey (USGS) collect data on
earthquakes occurring in Alaska at seismological monitoring stations throughout the state. Earthquakes
in Alaska range from minor events detected only by sensitive instruments, to the largest earthquake
ever recorded in North America (the 1964 Good Friday earthquake near Anchorage, magnitude 9.2)
(Table 3-4, Figure 3-15).
Bristol Bay Assessment
3-30
January 2014
-------
Chapter 3
Region
Table 3-4. Examples of earthquakes in Alaska.
Date
March 28, 1964
Novembers, 2002
September 25, 1985
July 13, 2007
March 25, 2012
Depth (km)
25
4.2
184
6.2
12
Magnitude3
9.2
7.2
4.9
4.3
3.0
Distance and Direction from
Deposit
the Pebble
469 km east-northeast
593 km northeast
61 km southeast
30 km west-southwest
122 km east
Notes:
'" Local magnitude as reported by the Alaska Earthquake Information Center. Note that earthquakes in the range of magnitudes
occur regularly in the Lake Clark area (data not shown). These earthquakes are centered at a depth of 100 km or greater.
1.5 to 3.6
Southwestern Alaska experiences a large number of earthquakes related to the presence of four active
moving blocks of crust associated with large fault systems. These faults are, from north to south, the
Tintina-Kaltag Fault, the Iditarod-Nixon Fork Fault, the Denali-Farewell Fault, the Lake Clark-Castle
Mountain Fault system, the Bruin Bay Fault, and the Border Ranges Fault (Figure 3-15). Some sections
along these faults are seismically active and have generated earthquakes in the past. The size of an
earthquake is directly related to the area of the fault that ruptures; thus, longer faults are capable of
producing larger earthquakes. The damage caused by an earthquake is related to the size of and distance
from the earthquake. The effects of an earthquake diminish with distance, so more damage occurs at the
epicenter than at a point several kilometers away.
The Lake Clark-Castle Mountain Fault system, with a mapped length of 225 km, is the fault located
nearest to the Pebble deposit. The northeast-southwest trending Lake Clark Fault is the western
extension of the Castle Mountain Fault (Koehler and Reger 2011). The western terminus of the Lake
Clark Fault has not been identified, but was originally interpreted to be near the western edge of Lake
Clark. Recent studies by USGS reinterpreted the position of the Lake Clark Fault further to the
northwest, potentially bringing it as close as 16 km to the Pebble deposit (Haeussler and Saltus 2004).
Haeussler and Saltus (2004) acknowledge that the fault could extend closer than 16 km, but data are not
available to support this interpretation.
There are few residents and no long-term seismic monitoring station records in the area of the Pebble
deposit, which make it difficult to assess accurately the recent seismic history of the area. As a result, the
paleoseismic history of the western part of the Lake Clark Fault is unknown (Koehler and Reger 2011).
USGS has concluded that there is no evidence for fault activity or seismic hazard associated with the
Lake Clark Fault in the past 1.8 million years, and no evidence of movement along the fault northeast of
the Pebble deposit since the last glaciations 11,000 to 12,000 years ago (Haeussler and Waythomas
2011).
Bristol Bay Assessment
3-31
January 2014
-------
Chapter 3
Region
Figure 3-15. Seismic activity in southwestern Alaska. Location and magnitude of significant, historic
earthquakes (USGS 2010) that caused deaths, property damage, and geological effects or were
otherwise experienced are shown. Fault lines are based on Haeussler and Saltus (2004), including
the preferred drawing of the Lake Clark Fault (dashed purple line).
X Approximate Pebble Deposit Location
Transportation Corridor
Fault
Watershed Boundary
Bristol Bay Assessment
3-32
January 2014
-------
Chapter 3 Region
The 1980 USGS map of the structural geology of the Iliamna Lake quadrangle shows several mapped
faults in the Tertiary-age volcanic rocks that host the area's mineral deposits. Geologic mapping
conducted by consulting firms for PLP identified numerous faults in the Pebble deposit area. The
mapped faults shown in both these sources are all considerably shorter than the Lake Clark Fault, and
therefore by themselves have a very limited capability to produce damaging earthquakes. The largest
mapped fault in the Pebble deposit area is an unnamed northwest-trending fault approximately 13 km
southwest of the deposit, approximately 16 km in length. There are several short (less than 4 km) faults
mapped within and near the mine scenario watersheds (the Z-series faults), about half of which have
northeast-southwest orientations. The faults show vertical displacement ranging from tens of meters to
over 900 m, and are interpreted to have formed coincident with mineralization (Ghaffari et al. 2011).
Although there is no evidence that the Lake Clark Fault extends closer than 16 km to the Pebble
depositor that there is a continuous link between the Lake Clark Fault and the northeast-trending faults
at the mine site, mapping the extent of subsurface faults over long, remote distances is difficult and has a
high level of uncertainty.
Not all earthquakes occur along the mapped sections of faults. In some instances, stresses build up and
cause earthquakes in rock outside of known pre-existing faults. Earthquakes can occur on previously
unidentified, minor, or otherwise inactive faults, or along deeper faults that are not exposed at the
surface. Although these floating earthquakes are generally smaller and less frequent than those
associated with faults, they may occur at locations closer to critical structures than the nearest mapped
capable fault. Small earthquakes can be induced when reservoirs or impoundments are constructed
(Kisslinger 1976), altering the soil and rock stresses and increasing pore pressure along pre-existing
zones of weakness. Induced earthquakes are generally small, but can occur frequently and cause
landslides and structural damage to earthen structures.
Interpreting seismicity in the Bristol Bay area is difficult because of the remoteness of the area, its
complex bedrock geology overlain by multiple episodes of glacial activity, and the lack of historical
records on seismicity. Thus, there is a high degree of uncertainty in determining the location and extent
of faults, their capability to produce earthquakes, whether these or other geologic features have been
the source of past earthquakes, and whether they have a realistic potential for producing future
earthquakes. Large earthquakes have return periods of hundreds to thousands of years, so there maybe
no recorded or anecdotal evidence of the largest earthquakes on which to base future predictions.
3.7 Existing Development
Unlike most other areas supporting Pacific salmon populations, the Bristol Bay watershed is
undisturbed by significant human development. It is located in one of the last remaining virtually
roadless areas in the United States (Section 6.1.3.1). Large-scale, human-caused modification of the
landscape—a factor contributing to extinction risk for many native salmonid populations (Nehlsen et al.
1991)—is absent, and development in the watershed consists of only a small number of towns, villages,
Bristol Bay Assessment 333 January 2014
-------
Chapter 3 Region
and roads. The Bristol Bay watershed also encompasses Iliamna Lake, the largest undeveloped lake in
the United States.
The primary human manipulation of the Bristol Bay ecosystem is the marine harvest of approximately
70% of salmon returning to spawn. However, commercial salmon harvests are the Alaska Department of
Fish and Game's (ADF&G's) second priority for fish management; its first priority is to ensure that
sufficient fish migrate into rivers to maintain a sustainable fishery, and thus sustainable salmon-based
ecosystems. No hatchery fish are reared or released in the Bristol Bay watershed, whereas
approximately 5 billion hatchery-reared juvenile salmon are released annually across the North Pacific
(Irvine et al. 2009). Given the potential for hatchery fish to have negative effects on wild fish (e.g., Araki
et al. 2009, Rand et al. 2012), this lack of hatchery fish is notable.
3.8 Climate Change
Thus far, this chapter has focused on the current physical environment in the Bristol Bay watershed. In
the future, over time scales at which large-scale mining will potentially affect these watersheds, this
physical environment is likely to change substantially—particularly in terms of climate and, by
extension, hydrology. Over the past 60 years, much of Alaska has been warming at twice the average
rate of the United States and many parts of the world (ACIA 2004). Throughout Alaska, changes such as
warmer temperatures, melting glaciers, declining sea ice, and declining permafrost have already
occurred (Serreze et al. 2000, Stafford et al. 2000, ACIA 2004, Hinzman et al. 2005, Liston and Hiemstra
2011, Markon et al. 2012). However, there is limited evidence over the last decade that suggests air
temperature in much of Alaska has cooled, due to changes in the Pacific Decadal Oscillation and
weakening of the Aleutian low (Wendler et al. 2012). Climate models suggest that warming throughout
Alaska is projected to continue, and it is likely to lead to changes in the type and timing of precipitation,
decreased snowpack and earlier spring snowmelt, and subsequent changes in hydrology similar to
projections in Arctic regions (Hinzman et al. 2005).
Using methods detailed in Box 3-4, we used the multi-model average A2 emissions scenario developed
by SNAP (2012) to generate 30-year means for future temperature and precipitation patterns in the
Bristol Bay region. We focused on characterizing possible climate change impacts using the A2
emissions scenario 30-year mean for the end of this century (2071-2100) as an upper bound estimate of
climate change effects expected for this region with current modeling. Similar trends in temperature and
precipitation, but with smaller magnitudes, are shown for effects earlier in the century or with more
benign emission scenarios.
Bristol Bay Assessment 334 January 2014
-------
Chapter 3 Region
BOX 3-4. METHODS FOR CLIMATE CHANGE PROJECTIONS
To project temperature and precipitation changes over the next century, we used data from the Scenarios
Network for Alaska and Arctic Planning (SNAP). A full description of the SNAP data and methodology used is
available on the SNAP website (SNAP 2012).
From the SNAP dataset, we used downscaled values of monthly mean temperature and precipitation. The
historical dataset is derived from the Climate Research Unit (CPU) at the University of East Anglia for 1901
to 2009 (CPU 2012). The CPU data are downscaled using the Parameter-elevation Regressions on
Independent Slopes Model (PRISM) 1971 to 2000 monthly climatologies for Alaska (PRISM Climate Group
2012), which take into account elevation, slope, and aspect. SNAP then developed downscaled monthly
projections of temperature and climate for Alaska under three emissions scenarios developed by the
Intergovernmental Panel on Climate Change for the Coupled Model Intercom pa rison Project. SNAP uses five
global climate models (GCMs) [cccma_cgcm31, mpi-echam5, gfdl_cm21, ukmo_hadcm3, and
miroc3_2_medres] that best characterize the Arctic region up to the year 2100 (Walsh et al. 2008). These
emissions scenarios are:
• the Bl scenario, which represents a best-case emissions scenario;
• the A1B scenario, which represents a middle-of-the-road emissions scenario; and
• the A2 scenario, which represents a worst-case emissions scenario.
For this assessment, we use the SNAP 5-model average for the A2 scenario of the best-performing GCMs to
consider a worst-case climate change scenario for the Bristol Bay region. Although uncertainty is inherent in
climate modeling due to many factors, the SNAP 5-model average tends to perform better than any single
model under the A2 scenario. Using the SNAP model, we calculated 30-year normal values, or average
values over a 30-year period, for temperature and precipitation over 1971 to 2000 (historical) and over
2011 to 2040, 2041 to 2070, and 2071 to 2100 under the three emissions scenarios. We focused on the
A2 scenario for the years 2071 to 2099 (the year 2100 is not included because one of the GCMs used in
the average did not include that year). Using the SNAP data, we calculated changes in temperature and
precipitation at three scales: the Bristol Bay watershed (Figure 2-3), the Nushagakand Kvichak River
watersheds (Figure 2-4), and the mine scenario watersheds (Figure 2-5). We also calculated annual
potential evapotranspiration (PET) (Hamon 1961) and annual water surplus (annual precipitation minus
PET) for the Bristol Bay watershed and the Nushagak and Kvichak River watersheds.
Data for the appropriate watersheds were extracted from the SNAP dataset, which covers the entire state of
Alaska. The resolution of the SNAP dataset is a 771-m grid. Any grid pixel intersecting a watershed boundary
was included, even if the intersection was minimal, to account for the full range of possible temperature and
precipitation values across the watersheds. In all cases, the values reported in the assessment represent
the geographic spatial average across the entire watershed over an average of 30 years. Precipitation and
temperature differences between the two periods were calculated as the geographic spatial average across
the entire watershed of the raster represent!ng the A2 scenario (2071 to 2099), minus the present period.
Precipitation percent differences were calculated as the geographic spatial average across the entire
watershed of the raster representing the difference between the A2 scenario (2071 to 2099) and the
present period, divided by the present period and multiplied by 100.
Water surpluses under historical and future periods were calculated for each calendar month and summed
to arrive at annual values. Differences between periods were calculated by subtracting the present value
from the A2 scenario (2071 to 2099) value. It is important to remember that surplus measurements were
calculated at the annual level and do not represent monthly or seasonal differences across a single scenario
or between multiple scenarios.
Uncertainty is an inherent issue when dealing with projected temperature, precipitation, and water surplus
values because of local variability and uncertainty in GCMs. Using average values for the five best-
performing GCMs for the Arctic and calculating mean values over 30-year periods helps to reduce
uncertainty; however, this averaging also decreases precision in predicting extreme events.
Bristol Bay Assessment 335 January 2014
-------
Chapter 3 Region
By the end of the century, based on SNAP (2012) data for the A2 emissions scenario, the multi-model
average annual air temperature in the Bristol Bay region is projected to increase by approximately 4°C,
with an approximately 6°C increase occurring in the winter months. Increases in air temperature are
likely to affect the accumulation and melt of snowpack, the extent of lake ice, and the timing of spring ice
break up, and result in increased water temperatures. Research from adjacent regions provides some
basis for estimating water temperature changes that may result from climate change. Kyle and Brabets
(2001) estimated that air temperature increases of 7.2°C to 8.5°C projected for Cook Inlet watersheds by
2100 would be associated with water temperature increases of 1.2°C to 7.1°C. It is important to note
that although air temperature can be a useful metric for modeling water temperature, other factors (e.g.,
quantity, type, and seasonally of precipitation, snow and glacier cover) can also be critical water
temperature drivers (Webb and Nobilis 1997, Mohseni and Stefan 1999).
Although we are unable to predict a change in extreme events, changes in precipitation patterns are
likely to occur (Salathe 2006, Christensen et al. 2007, Peacock 2012, Markon et al. 2012), with rain-on-
snow events becoming more common. The effect of increased rain-on-snow events on the frequency or
volume of floods is unclear. Storm patterns also may change, although the increased likelihood of
extreme events occurring and potential impacts on flooding are unknown. Changes in the seasonally of
precipitation, snowpack, and the timing of snowmelt will likely affect streamflow regimes and may
result in water availability changes, particularly in terms of decreased water availability in summer.
Based on temperature, precipitation, and evapotranspiration projections, the landscape will likely be
warmer and wetter annually; however, due to method limitations we are not able to determine how
evapotranspiration will affect water availability on the landscape seasonally (Box 3-4).
3.8.1 Climate Change Projections for the Bristol Bay Region
Across the entire Bristol Bay watershed, average temperature is projected to increase by approximately
4°C by the end of the century (Table 3-5, Figure 3-16), and winter temperature is projected to increase
the most (Table 3-5). Similar patterns are projected in the Nushagak and Kvichak River watersheds
(Table 3-5).
By the end of the century, precipitation is projected to increase roughly 30% across the Bristol Bay
watershed, for a total increase of approximately 250 mm annually (Table 3-6, Figure 3-17). In the
Nushagak and Kvichak River watersheds, precipitation is projected to increase roughly 30% as well, for
a total increase of approximately 270 mm of precipitation annually (Table 3-6). At both spatial scales,
increases in precipitation are expected to occur in all four seasons (Table 3-6). Based on
evapotranspiration calculations, annual water surpluses of 144 mm and 165 mm are projected for the
Bristol Bay watershed and the Nushagak and Kvichak River watersheds, respectively (Table 3-7,
Figure 3-18). Our simulated temperature and precipitation changes based on SNAP (2012) data for the
Bristol Bay region are within the range of changes projected by other studies concentrating on Alaska
and the Arctic (Christensen et al. 2007, Peacock 2012, Markon et al. 2012).
Bristol Bay Assessment 336 January 2014
-------
Chapter 3
Region
Table 3-5. Average annual and seasonal air temperature for historical and projected periods across
the Bristol Bay watershed and the Nushagak and Kvichak River watersheds. Values were calculated
using the SNAP (2012) dataset (Box 3-4). Temperature was calculated as average values over each
30-year period. Number in parentheses equals one standard deviation.
Scale
Bristol Bay Watershed
(Scale 1)
Nushagak and Kvichak
River Watersheds
(Scale 2)
Season
Annual
Winter
Spring
Summer
Fall
Annual
Winter
Spring
Summer
Fall
Historical Temperature
(1971-2000)
(°C)
1(1)
-8(2)
0(1)
11(2)
1(2)
1(1)
-9(1)
0(1)
11(2)
0(2)
Projected Temperature
(2017-2099)
(°C)
5(1)
-2(2)
4(1)
14(2)
5(2)
5(1)
-3(1)
4(1)
14(2)
5(2)
Difference
(°C)
4 (0.2)
6(1)
4 (0.2)
3 (0.07)
4 (0.3)
4 (0.2)
6 (0.4)
3 (0.2)
3 (0.05)
4 (0.07)
Table 3-6. Average annual and seasonal precipitation for historical and projected periods across
the Bristol Bay watershed and the Nushagak and Kvichak River watersheds. Values were calculated
using the SNAP (2012) dataset (Box 3-4). Precipitation was calculated as average values over each
30-year time period. Number in parentheses equals one standard deviation.
Scale
Bristol Bay Watershed
(Scale 1)
Nushagak and Kvichak
River Watersheds
(Scale 2)
Season
Annual
Winter
Spring
Summer
Fall
Annual
Winter
Spring
Summer
Fall
Historical Precipitation
(1971-2000)
(mm)
847 (421)
177 (121)
150 (91)
234 (97)
286 (141)
795 (336)
160 (79)
138 (67)
226 (84)
271 (123)
Projected Precipitation
(2017-2099)
(mm)
1,095 (512)
229 (143)
196 (112)
303 (117)
367 (170)
1,062 (430)
215 (97)
189 (90)
300 (107)
357 (152)
Difference
(mm)
248 (104)
52 (27)
45 (25)
69 (25)
81 (34)
267 (95)
55 (21)
51 (23)
75 (24)
86 (32)
Table 3-7. Average annual water surplus for historical and projected periods across the Bristol Bay
watershed and the Nushagak and Kvichak River watersheds. Values were calculated using the SNAP
(2012) dataset (Box 3-4). Number in parentheses equals one standard deviation.
Scale
Bristol Bay Watershed
(Scale 1)
Nushagak and Kvichak River Watershed
(Scale 2)
Historical Surplus
(1971-2000)
(mm)
400 (441)
341 (359)
Projected Surplus
(2017-2099)
(mm)
544 (534)
506 (456)
Difference
(mm)
144 (106)
165 (99)
Bristol Bay Assessment
3-37
January 2014
-------
Chapter 3
Region
Figure 3-16. Mean annual temperature across the Bristol Bay watershed under (A) historical conditions (1971 to 2000) and (B) the A2
emissions scenario (2071 to 2099), and (C) the temperature change between these two climate scenarios (SNAP 2012). See Box 3-4 for
additional details.
A. Historical Conditions
Temperature (°C)
C. Temperature Change
4.5 - 5.0
4.0-4.5
3.5-4.0
3.0-3.5
2.5-3.0
2.0-2.5
1.5-2.0
1.0 -1.5
0.5-1.0
0.0-0.5
Bristol Bay Assessment
3-38
January 2014
-------
Chapter 3
Region
Figure 3-17. Mean annual precipitation across the Bristol Bay watershed under (A) historical conditions (1971 to 2000) and (B) the A2
emissions scenario (2071 to 2099), and (C) the precipitation change between these two climate scenarios (SNAP 2012). See Box 3-4 for
additional details.
A. Historical Conditions
Precipitation (mm)
>2000
1,500 - 2,000
1,000 -1,500
750 -1,000
600 - 750
550 - 600
500 - 550
450 - 500
400 - 450
350 - 400
325 - 350
C. Precipitation Change
Precipitation Difference (mm)
>600
500-600
400 - 500
300-400
250 - 300
200 - 250
150-200
100 -150
Bristol Bay Assessment
3-39
January 2014
-------
Chapter 3
Region
Figure 3-18. Mean annual water surplus (precipitation minus evapotranspiration) across the Bristol Bay watershed under (A) historical
conditions (1971 to 2000) and (B) the A2 emissions scenario (2071 to 2099), and (C) the water surplus change between these two climate
scenarios (SNAP 2012). See Box 3-4 for description of surplus calculations.
A. Historical Conditions
Surplus (mm)
>2,000
1,500 - 2,000
1,000 -1,500
750 -1,000
500 - 750
400 - 500
300 - 400
200 - 300
100 - 200
0-100
-200 - 0
C. Water Surplus Change
Difference in Surplus (mm)
>500
250-500
200-250
150-200
100 -150
50 -100
0-50
-10-0
Bristol Bay Assessment
3-40
January 2014
-------
Chapter 3 Region
3.8.2 Potential Climate Change Effects
There are likely to be hydrological impacts associated with projected changes in temperature,
precipitation, and evapotranspiration in the Bristol Bay watershed, including changes in the magnitude
and timing of streamflow that are likely to affect salmon habitat and populations. When temperature
increases in freshwater environments, community structure, habitat, and salmon populations can be
affected (Eaton and Scheller 1996, Hauer et al. 1997). With warmer temperatures and changes in the
type, timing, and amount of precipitation, there likely will be changes in snowpack, a shift in the timing
of spring snowmelt, and changes in the type of precipitation falling (Barnett et al. 2005). With these
changes, there will be alterations to both the magnitude and timing of the natural streamflow regime
and a likely decline in seasonal water availability, mirroring already observed changes in other systems
such as the Pacific Northwest (Mote et al. 2003).
These hydrologic flow regime changes may affect salmon populations during spawning and smolt
migrations, and can scour streambeds leading to the loss of salmon eggs (Lisle 1989, Montgomery et al.
1996, Steen and Quinn 1999, Mote et al. 2003, Lawson et al. 2004, Stewart et al. 2004). Changes in
hydrology are likely to affect existing habitat via changes in water volume and velocity along with
channel forms, which may lead to declines in habitat availability for spawning and rearing salmon
populations. Changes to baseflow, depending on groundwater and surface water interactions, are likely
to affect the amount of wetlands in the Bristol Bay watershed, in that wetlands are likely to decrease
under drier baseflow conditions. Although we are unable to predict whether baseflow will increase or
decrease, any changes in baseflow will likely affect water temperature (in addition to the direct effects of
increased air temperature on water temperature).
Both the hydrology and water temperature of freshwater systems affect critical life stages of salmonid
species. Furthermore, these hydrological changes are likely to have different effects on salmon
populations depending on the amount of time they spend rearing in freshwater habitats, their life stage,
and their ability to adapt to changes in environmental conditions. Pink and chum salmon are likely to be
affected by temperature increases early in egg incubation, which can affect timing of emergence,
migration to the ocean, and potential mismatch in the timing of peak food abundance in the marine
environment (Bryant 2009). For example, the average migration time for one population of pink salmon
in southeast Alaska now occurs nearly 2 weeks earlier than it did 40 years ago (Kovach et al. 2012). For
sockeye salmon that typically rear in fresh water for 1 to 2 years, temperature increases may affect life-
stage timing, including spawning and fry emergence, as well as the growth and survival of lake-rearing
fry (Healey 2011, Martins et al. 2012). Across all five Pacific salmon species, time to fry emergence
decreases as water temperature increases (Figure 3-19); thus, warmer winters may result in earlier fry
emergence.
Bristol Bay Assessment 341 January 2014
-------
Chapter 3
Region
Figure 3-19. Relationship between time from fertilization to emergence and temperature for the
five Pacific salmon species. Data are from Quinn 2005.
350
•Chinook
Chum
-Pink
6 8 10
Temperature (°C)
12
14
16
Changes in precipitation and hydrology also may affect access to lakes and spawning locations, and high-
intensity rainfall may increase sedimentation in spawning streams and rearing lakes for sockeye salmon
(Bryant 2009). Rich et al. (2009) hypothesized that warmer temperature was a factor in poor sockeye
salmon recruitment in the Kvichak River watershed. For Chinook salmon, increases in temperature are
likely to affect incubation and fry emergence (Beer and Anderson 2001), which may affect growth,
survival, and timing of migration to the ocean (Heming et al. 1982, Taylor 1990, Berggren and Filardo
1993). Coho salmon incubation and timing of emergence are also affected by increases in temperature
(Tang etal. 1987).
Populations of Pacific salmon species are likely to respond and adapt to changes in temperature,
precipitation, and hydrology in different ways, and the geographic location of populations is likely to
affect their ability to adapt to these changes. Studies have predicted that the reproductive success of
salmon populations in Washington is likely to decline over the next century (Battin et al. 2007, Mantua
et al. 2010), and freshwater temperature increases in the Fraser River will negatively affect growth and
survival of sockeye salmon at all life stages (Healey 2011). The genetic and life history diversity within
and among the Bristol Bay Pacific salmon populations (Section 5.2.4) will likely be crucial for
maintaining the resiliency of the region's salmon stocks under a future environment characterized by
climate change and increased anthropogenic stressors (Hilborn et al. 2003, Schindler et al. 2010, Rogers
and Schindler 2011).
Bristol Bay Assessment
3-42
January 2014
-------
4.1 Mineral Deposits and Mining in the Bristol Bay Watershed
Significant mineral resources are located in Alaska, and the state has a long mining history. Russian
explorers began searching for placer gold in the early 1800s, and substantial placer deposits have been
found in many areas of the state. More recently, hard rock exploration has increased throughout the
region. Alaska mines range in size from small, recreational suction dredging operations to large-scale
commercial operations, for a variety of deposit types (Table 4-1).
Several known mineral deposits with potentially economically significant resources are located in the
Nushagak and Kvichak River watersheds, and active exploration of deposits is occurring in a number of
claim blocks (deposits other than Pebble are considered in greater detail in Chapter 13; see Table 13-1
and Figure 13-1 for the names and locations of these deposits). Of deposit types occurring or likely to
occur in the region, porphyry copper, intrusion-related gold, and copper and iron skarn may indicate
economically viable mining, thereby prompting large-scale development. Thus, the development of a
number of mines, of varying sizes, is plausible in this region—and once the infrastructure for one mine
is available, it would likely facilitate the development of additional mines (Chapter 13).
The potential for large-scale mining development within the Nushagak and Kvichak River watersheds is
greatest for porphyry copper deposits, most notably the Pebble deposit. Significant exploration activity
has been ongoing at this deposit for many years, and the information available provides the most
complete description of potential mining in the region. Because the Pebble deposit is the most likely
deposit to be developed in the near term, this assessment focuses exclusively on porphyry copper
deposits. However, much of the discussion of mining methods (Section 4.2.3) applies to all types of
disseminated ore deposits (i.e., ores with low concentrations of metal spread throughout the body of
rock).
Bristol Bay Assessment
4-1
January 2014
-------
Chapter 4
Type of Development
Table 4-1. Characteristics of past, existing, or potential large mines in Alaska.
Mine
Location
Target metals
Ore type
Ore grade
quality
Operational life
(years)
Extraction type
Total resource
(million metric
tons)
Ore processing
rate (metric
tons/day)
Total waste rock
(million metric
tons)
Tailings disposal
Tailings amount
(million metric
tons)
Tailings footprint
(km2)
Dam height (m)
Acid mine
drainage
potential
Kennecott
Copper River
basin, in
Wrangell-St.
Elias National
Park
Copper, silver
Massive sulfide
Very high
27 (1911-1938)
Underground
slope mining
-4.5
-91
<0.9
On Kennicott
Glacier
<0.9
NA
NA
No
Donlin
13 miles N of
village of
Crooked Creek
and Kuskokwim
River
Gold
Gold-bearing
quartz
Moderate
22
Open pits (2)
491b
48,524
1900
Dams/ponds (2)
426
5.4
143 (largest of
multiple dams)
Yes
Fort Knox
26 miles NEof
Fairbanks
Gold
Oxide ore body
Low
20
Open pit
401
33,000-45,000
338
Dam/pond
181
4.5
111
No
Greens Creek
18 miles SW of
Juneau, in
Admiralty Island
National
Monument
Zinc, lead, silver,
gold
Massive sulfide
High
35-50
Underground
slope mining
29
1,524
- 1.8
Drytailings
-13.6
0.25
NA
Yes
Kensington
45 miles NW of
Juneau, belween
Berners Bay and
Lynn Canal
Gold
Gold-bearing
quartz
Moderale
10
Underground slope
mining
24
1,134
1.5
Lake disposal
4.1
0.24
27b
No
Pogo
85 miles ESE of
Fairbanks
Gold
Gold-bearing
quartz
Moderale
11
Underground
slope mining
9.1
2,267
1.7
Drylailings
4.9
0.12
NA
No
Red Dog
Western Brooks
Range, 82 miles N
of Kolzebue and 46
miles from Ihe
Chukchi Sea
Zinc, lead
Massive sulfide
High
42(1989-2031)
Open pils (2)
171
7,500-8,300
142
Dam/pond
91
3
63
Yes
Pebble (78-yr)"
Headwalers of
Ihree slreams
running inlo Ihe
Nushagak and
Kvichak Rivers
Copper, gold,
molybdenum
Porphyry copper
Low
78
Open pil
5,920
208,000
14,600
Dams/ponds
(multiple)
5,860
46
209 (largeslof
multiple dams)
Yes
Notes:
a Ghaffarietal. 2011.
b Novagold 2012.
NA = not applicable.
Source: Levit and Chambers 2012, except as noted.
Bristol Bay Assessment
4-2
January 2014
-------
Chapter 4
Type of Development
4.2 Porphyry Copper Deposits and Mining Processes
4.2.1 Genesis of Porphyry Copper Deposits
Porphyry copper deposits are found around the world, often occurring in clusters (Lipman and Sawyer
1985, Singer et al. 2001, Anderson et al. 2009) in areas with active or ancient volcanism (Figure 4-1).
They are formed when hydrothermal systems are induced by the intrusion of magma into shallow rock
in the Earth's crust. Water carries dissolved sulfur-metallic minerals (sulfides) into crustal rock where
they precipitate (John et al. 2010). Minerals containing sulfur and metals are disseminated and
precipitate throughout the affected rock zone in concentrations typically less than 1% (Table 4-2)
(Singer et al. 2008). Porphyry copper deposits range in size from millions to billions of tons (Table 4-2).
The well-delineated Pebble deposit is at the upper end of the total size range; thus, any additional
deposits found in the Nushagak and Kvichak River watersheds are likely to be much smaller than the
Pebble deposit.
Table 4-2. Global grade and tonnage summary statistics for porphyry copper deposits.
Parameter
Tonnage (Mt)
Copper grade (%)
Molybdenum grade (%)
Silver grade (g/t)
Gold grade (g/t)
10th Percentile
30
0.26
0.0
0.0
0.0
50th Percentile
250
0.44
0.004
0.0
0.0
90th Percentile
1,400
0.73
0.023
3.0
0.20
Pebble Deposit3
10,777
0.34
0.023
unknown
0.31
Notes:
a Pebble deposit information is based on 0.3% copper cut-off grade, and includes measured, indicated, and inferred resources from Pebble
Limited Partnership.
Mt = million tons; g/t = grams per ton.
Sources: Singer et al. 2008; Appendix H.
4.2.2 Chemistry and Associated Risks of Porphyry Copper Deposits
Exposure to hazards associated with mining porphyry copper deposits can pose risks to aquatic and
terrestrial ecosystems and to human health. These risks can range from insignificant to extremely
harmful depending on a variety of factors that control the hazards, including site geology (both local and
regional), hydrologic setting, climate, and mining and ore processing methods. There are a variety of
geochemical models and approaches to understand and predict the water quality of releases to the
environment; however, our ability to make predictions is limited because of data insufficiency and the
inherent complexity of natural materials and their environment.
Bristol Bay Assessment
4-3
January 2014
-------
Chapter 4
Type of Development
mines and mining regions mentioned in the text also are shown on the map.
- Red Dog
- Pebble
Greens Creek
Frasier River
Coeur d'Alene River
Bingham Canyon
Safford
Clark Fork River
Soda Butte Creek
Kingston Fossil Plant
Los Frailes -
Chuquicamata
Bajo de la Alumbrera
• Other Mines/Mine Areas
Porphyry Copper Deposits
Ore (millions of metric tons)
0-500
- 501-2,000
I 2,001-8,000
• 8,001-16,000
C 16,001 - 24,000
N
A
0 1,500 3,000
0 1,500 3,000
I Miles
Bristol Bay Assessment
4-4
January 2014
-------
Chapter 4 Type of Development
Sources of hazards from porphyry copper mines can be grouped into four broad, interrelated categories:
acid-generating potential, trace elements and their mobilities, mining and ore processing methods, and
waste disposal practices. The relative importance of these categories will vary from deposit to deposit,
but some generalization can be made for porphyry copper deposits as a whole. In this section we
consider those categories related to environmental chemistry, acid-generating potential, and trace
elements (categories related to mining processes are described in Section 4.2.3).
Mining processes expose rocks and their associated minerals to atmospheric conditions that cause
weathering, which releases minerals (e.g., copper minerals) from the rock matrix. Grinding methods
used in these processes create materials that have high specific surface areas, which accelerates the rate
of weathering. Porphyry copper deposits are characterized by the presence of sulfide minerals, and
oxidation of sulfide minerals creates acidity, sulfate, and free metal ions (e.g., iron in the case of pyrite);
in addition, the acid produced can further accelerate weathering rates. Because most metals and other
elements become more soluble as pH decreases, the acid-generating or acid-neutralizing potentials of
waste rock, tailings, and mine walls are of prime importance in determining potential environmental
risks associated with exposure to metals and certain elements in the aquatic environment.
One way to predict if acid generation has the potential to occur is to perform acid-base accounting tests.
Acid-base accounting tests are rapid methods to determine the acid-generating potential (AP) and
neutralizing potential (NP) of a rock or mining waste material, independent of reaction rates. These
potentials are then compared to one another by either their differences or their ratios, with the net
neutralizing potential (NNP) being NP-AP and the neutralizing potential ratio (NPR) being NP/AP. AP,
NP, and NNP typically are expressed in units of kilograms of calcium carbonate per metric ton of waste
material (kg CaCOs/metric ton). Positive NNP values are net alkaline and negative values are net acidic.
Although methods used for acid-base accounting have known limitations, it is common industry practice
to consider materials that have an NPR of 1 or less as potentially acid-generating (PAG) and materials
that have an NPR greater than 4 as being non-acid-generating (NAG) (Brodie et al. 1991, Price and
Errington 1998). Materials that have a ratio between 1 and 4 require further testing via kinetic tests and
geochemical assessment for classification (Brodie et al. 1991, Price 2009, Price and Errington 1998).
This further testing and assessment are necessary because if neutralizing minerals react before acid-
generating minerals, the neutralizing effect may not be realized and acid might be generated at a later
time—that is, pH of the system may decrease over time as neutralizing materials are used up, resulting
in acid mine drainage. Additionally, some toxic elements (e.g., selenium and arsenic) may be released
from mining materials under neutral or higher pH conditions, which would be observed during kinetic
leaching tests conducted at variable pH values. Depending on the water chemistry of both a receiving
water body and any mine drainage, released elements may either be transported downstream as
dissolved ions or form precipitates that travel as suspended solids or settle to the streambed.
In general, the rocks associated with porphyry copper deposits tend to straddle the boundary between
being net acidic and net alkaline, as illustrated by Borden (2003) for the Bingham Canyon porphyry
copper deposit in Utah (Figure 4-2A). AP values for porphyry copper deposits typically correlate with
Bristol Bay Assessment January 2014
-------
Chapter 4 Type of Development
the distribution of pyrite. The pyrite-poor, low-grade core corresponds to the central part of the
Bingham Canyon deposit, where NNP values are greater than zero. Moving outward from the core to the
ore shell and pyrite shell, pyrite abundance increases and NNP values become progressively more
negative (Figure 4-2B).
4.2.3 Overview of the Mining Process
Developing a mine requires establishing surface or underground mine workings that allow access to the
ore body. The scope and complexity of development-related activities vary depending on the
characteristics of each project, but typically include the following components.
• Site preparation (clearing, stripping, and grading). Topsoil and overburden are removed and
typically stockpiled for later use in mine reclamation.
• Construction of mine site infrastructure. Specific requirements depend on the size and type of mine
operation, its location, and proposed mining, milling, and processing methods. Typical infrastructure
includes facilities for ore crushing, grinding, and other mineral separation processes; ore stockpiling
and waste rock disposal facilities; tailings storage facilities; water supply, treatment, and
distribution facilities; transportation infrastructure such as roads or railways; pipelines; conveyers;
and other infrastructure (e.g., offices, shops, housing).
• Establishment of mine workings. Once the site is prepared and infrastructure is constructed, mine
workings are established: ore is extracted and processed, water at the site is managed and treated,
and tailings and waste rock are stored and managed.
At each stage of mine development, potential impacts on the environment and human health can be
reduced by ensuring effective implementation of proper design, construction, operation, and
management techniques and protocols (Box 4-1).
Any mining company must comply with a number of federal, state, and local laws when developing and
operating a mine. Compliance is facilitated through the regulatory permitting process and involves
multiple state and federal agencies (see Box 4-2 for additional detail on these regulatory requirements).
Regulations also serve to hold an operator accountable for potential future impacts, through
establishment of financial assurance requirements and imposition of fines or compliance orders upon
non-compliance with permit requirements (Box 4-3).
Bristol Bay Assessment * January 2014
-------
Chapter 4
Type of Development
Figure 4-2. Neutralizing potential at the Bingham Canyon porphyry copper deposit in Utah. (A) Plot
of neutralizing potential (NP) vs. acid-generating potential (AP) for mineralized rock types. PAG
denotes potentially acid-generating. Note that the range of uncertainty is indicated as 1 to 2 in this
figure; in the assessment, we use the more conservative range of 1 to 4. (B) Plan view of the
distribution of net neutralizing potential (NNP) values. Plots modified from Borden (2003).
B
1,000
_ 100
§
o
o
o
&
1 £
Uncertain
non-PAG
X
/ •
,' •/'
•f -' ^'1
.' +
' r' 4
PAG
-H-
10 100
AP (kg CaCOyt)
1,000
Bristol Bay Assessment
4-7
January 2014
-------
Chapter 4 Type of Development
BOX 4-1. REDUCING MINING'S IMPACTS
Reducing mining's impacts on the environment and human health requires proper mine planning, design,
construction, and operation; appropriate management and closure of waste and water containment and
treatment facilities; and monitoring and maintenance over all mine-life phases, including post-closure. Some
general methods for reducing adverse impacts of mining are provided here, along with information about
how these concepts are incorporated into the assessment.
Best management practices refer to specific measures for managing non-point source runoff (40 CFR
130.2(m)). Measures for minimizing and controlling sources of pollution in other situations are often
referred to as best practices, state of the practice, or simply mitigation measures. These are not the best
possible or conceivable practices, but rather the current practices of the best operators. We assume that
these types of measures would be applied throughout a mine as it is constructed, operated, closed, and
post-closure. Although we describe some measures as they are relevant to a discussion, it is not necessary,
for the purpose of this assessment, to describe them all.
Mitigation refers to all steps taken to avoid, minimize, treat, or compensate for potential adverse impacts
on the environment from a given activity. One example of a mitigation measure for avoidance is to avoid
mining a particularly reactive type of rock that might make future leachate management too difficult.
Minimization of an impact is practiced when avoidance is not feasible, and includes measures taken to
lessen the amount of contaminant released. An example of a mitigation measure to minimize an impact is to
blend known acid-producing material with sufficient neutralizing material. Treatment is required when
contaminants are released. An example is the diversion and collection of seepage from a waste rock pile for
passage through a wastewater treatment plant to meet appropriate water quality criteria prior to release to
the environment. Many elements of our mine scenarios include mitigation measures and all are assumed to
meet minimum regulatory requirements. Appendix I contains further discussion of mitigation measures.
Compensatory mitigation refers to the restoration, establishment, enhancement, and/or preservation of
wetlands, streams, or other aquatic resources to offset environmental losses resulting from unavoidable
impacts on waters of the United States, as authorized by Clean Water Act Section 404 permits issued by the
U.S. Army Corps of Engineers (40 CFR 230.93(a)(l)). This becomes an option only after all opportunities for
aquatic resource impact avoidance and minimization have been exhausted. See Box 7-2 and Appendix J for
a more complete discussion of compensatory mitigation.
Reclamation refers to restoration of a disturbed area to an acceptable form and planned use following
closure of a mining operation. Our mine scenarios assume that the site would be reclaimed according to
statutory requirements and present some options that are feasible and common, but it is outside the scope
of this assessment to evaluate a specific post-closure plan.
Remediation refers to fixing a problem that has become evident, such as an accidental release or spill of
product or waste material. For example, a tailings slurry spill would require remediation. The dam may have
been designed and constructed to properly mitigate (i.e., avoid or minimize) the potential for a spill, but an
accident or failure could cause contaminant release, thereby creating the need for remediation.
Bristol Bay Assessment 40 January 2014
-------
Chapter 4
Type of Development
BOX 4-2. PERMITTING LARGE MINE PROJECTS IN ALASKA
Large mine projects in Alaska must comply with federal and state environmental laws, and many federal,
state, and local government permits and approvals are required before construction and operation of a
large hard rock mine can begin. The specific permits and approvals vary from project to project, depending
on the unique challenges posed by each mine.
Federal laws and agencies. The involvement of federal agencies varies for each mine, but most projects at
least require authorizations from the U.S. Army Corp of Engineers. Other agencies that may be involved
include (but are not limited to) the U.S. Environmental Protection Agency, the U.S. Fish and Wildlife Service,
the National Marine Fisheries Service, the U.S. Coast Guard, and the U.S. Department of Transportation.
Federal agency authorizations ensure that projects comply with the following applicable federal laws.
Clean Water Act
Clean Air Act
National Environmental Policy Act
National Historic Preservation Act
Resource Conservation and Recovery Act
Rivers and Harbors Act
Endangered Species Act
Bald Eagle Protection Act
Migratory Bird Act
Magnuson-Stevens Act
Mine Safety and Health Act
Alaska Department of Natural Resources permits and approvals. The Alaska Department of Natural
Resources (ADNR) Office of Project Management and Permitting coordinates the permitting of large mine
projects via the establishment of a large mine project team for each project. This project team is an
interagency group, coordinated by ADNR, that works cooperatively with large mine permit applicants and
operators, federal resource agencies, and the Alaskan public to ensure that projects are designed, operated,
and reclaimed in a manner consistent with the public interest.
ADNR may require the following permits and approvals.
Plan of operations approval
Reclamation plan and bond approval
Right-of-way for access and utilities (roads, power
lines, pipelines)
Millsite lease
Permit to appropriate water
Dam safety certification (certificates of approval
to construct and operate a dam)
Upland ortideland leases
Material sale
Winter travel permits
Cultural resource authorization
Mining license
Alaska Department of Environmental Conservation permits and approvals. The Alaska Department of
Environmental Conservation may require the following permits related to wastewater management and
water and air quality.
• Waste management permit
• Alaska pollutant discharge elimination permit
• Domestic and non-domestic wastewater disposal
permits
Air quality permits
Approval to construct and operate in a public
water supply system
Plan review for non-domestic wastewater
treatment system
Plan review and construction approval for
domestic sewage system
• Oil discharge prevention and contingency plan
Other state permits and approvals. The state may require the following permits and approvals.
Certificate of reasonable assurance for 404
permits
Stormwater discharge pollution prevention plan
Fish passage permit
Fish habitat permit
Utility permit on right of way
Driveway permit
Approval to transport hazardous materials
Life and fire safety plan check
State fire marshal plan review certificate
Certificate of inspection for fired and unfired
pressure vessel
Employer identification number
Bristol Bay Assessment
4-9
January 2014
-------
Chapter 4
Type of Development
BOX 4-3. FINANCIAL ASSURANCE
Many of the regulatory checks listed in Box 4-2 help to reduce potential impacts of mining on the
environment, but they do not ensure that a permitted mine will have negligible effects on the environment.
Even with the most stringent requirements, accidents and human error may cause mine systems to fail—and
the most unpredictable accidents and errors often result in the most economically and environmentally
costly failures. Thus, regulations also serve to hold an operator accountable during mine operations via both
the imposition of fines for non-compliance with permit regulations and the establishment of financial
assurance requirements for closure and reclamation of the mine. Financial assurance basically means that
operators must ensure that sufficient funds are available for future remediation, closure, and reclamation of
a mine.
Operators of Alaska's hard rock mining facilities, including copper and gold facilities, are required by the
state to demonstrate financial assurance for reclamation, waste management, and dam safety costs.
• Prior to the start of hard rock mining operations on state-owned, federal, municipal, or private land, the
Alaska Department of Natural Resources (ADNR) must approve a reclamation plan and financial
assurance must be demonstrated in an amount necessary to ensure performance of the plan (Alaska
Statute 27.19).
• The Alaska Department of Environmental Conservation may require hard rock mining operations that
dispose of solid or liquid waste material or heated process or cooling water under a waste management
and disposal permit to demonstrate financial assurance in an amount based on the estimated costs of
required closure activities and post-closure monitoring for the waste management area (Alaska Statute
46.03.100(f)).
• Operators of hard rock mines on state-owned or privately owned land seeking ADNR approval to construct
mine tailings dams must demonstrate financial assurance to cover the cost of reclamation and post-
closure monitoring and maintenance of the dam (Alaska Statute 46.17).
• Operators of hard rock miningfacilities on land managed by the Bureau of Land Management or U.S.
Forest Service can be required by these agencies to demonstrate additional financial assurance for
reclamation (43 CFR 3809 and 36 CFR 228 Subpart A, respectively).
• In addition to State of Alaska and Bureau of Land Management financial assurance requirements,
facilities operating under leases, permits, or other agreements for the development of hard rock minerals
on tribal lands can be required by the Bureau of Indian Affairs to demonstrate financial assurance to
ensure compliance with the terms and conditions of the mineral agreement and applicable statutes and
regulations (25 CFR 211.24 and 225.30).
Financial assurance calculations assume that a government entity would have to enter the site and
commence reclamation activities without the benefit of any equipment or labor that may be at the site. The
process determiningthe cost of every shovel, loader, gallon of fuel, and hour of labor is revisited and
adjusted as necessary every 5 years. The State of Alaska allows several types of assurance (e.g., cash, gold
bullion, surety bonds, reclamation trust funds, irrevocable letters of credit).
Example Financial Assurance Amounts for Alaska Mines
Mine
Fort Knox
Kensington
Pogo
Red Dog
Amount
$68,852,293
$28,727,011
$44,430,000
$305,150,000
It is important to note that effective financial assurance depends on accurate estimates of costs, which
poses challenges when dealing with the potentially long-term, unpredictable, and costly events that a hard
rock mining operation must consider. For example, current financial assurance requirements do not address
chemical or tailings spills because of the greater degree of uncertainty related to these accidents; whereas
the costs associated with reclamation and closure can be estimated, the cost of cleaning up a spill is
unpredictable. However, financial assurance calculations increasingly include long-term water treatment.
Bristol Bay Assessment
4-10
January 2014
-------
Chapter 4 Type of Development
4.2.3.1 Extraction Methods
The low concentrations of disseminated metals in porphyry copper deposits require large amounts of
ore to enable a return on investment. Bulk or large-scale mining methods have been developed for this
purpose, and specific mining methods depend on ore quality and depth. A long-range mining plan is
usually developed first to match the final mine design with the available ore reserves, weighing
economics against engineering restrictions. This plan is re-evaluated throughout the life of the mine to
reflect changes in the economy, increased knowledge of the ore body, and potential changes in mining
technology.
Porphyry copper deposits are most commonly mined using open pit and, less commonly, underground
mining methods (John et al. 2010). Open pit mining is typically used to extract ore where the top of a
deposit is within 100 m of the surface (Blight 2010). Excavation of a pit begins at the surface, with
drilling and blasting to strip overburden from the ore body surface. The equipment and materials used
will fit the economies of scale for the project (e.g., mine life, daily production). The ore is drilled and
blasted according to a blasting pattern. The size and spacing of the drill holes and the amount of
explosives used determine the size of the material that is loaded and hauled to the crushing plant. The
pit is successively enlarged until the pit limits are established by the extent of ore that can be profitably
mined.
Pit design depends on the material characteristics of the ore and waste rock. The moisture content,
strength, and load-bearing capacity of the ore and waste rock help determine the angle of the pit slopes,
which generally are designed to be as steep as possible while still maintaining stability. A properly
designed pit reduces the stripping ratio, or the volume of waste rock to ore, thereby increasing
efficiency, potentially decreasing costs, and optimizing the amount of ore that can be mined
economically.
Block caving is an underground mining method used for large deposits with rock mass properties
amenable to sustainable caving action (Singer et al. 2008, Lusty and Hannis 2009, Blight 2010). Such
deposits typically have mineralization throughout the rock (e.g., porphyry copper deposits) and are too
deep to be mined economically by open pit methods. Block caving uses gravity to reduce the amount of
drilling and blasting required to extract ore. It involves tunneling to the bottom of the ore and
undercutting it, so that the deposit caves under its own unsupported weight. As ore is removed from
below, fractures spread throughout the block, which breaks into fragments and is removed from the
bottom of the enlarging void (Box 4-4).
Underground mining via block caving has a different set of costs than open pit mining, because of the
extensive drilling of tunnels and shafts through non-ore-bearing rocks needed to gain access to the ore.
Once begun, block caving generally requires less drilling and blasting, allows for less ore selectivity in
the mining process, and may require less labor relative to open pit mining. As with other types of
mining, the economics of block caving are determined by the prices of the metals being extracted,
operational costs, and a number of other factors. If block caving allows the mining of additional ore that
could not be mined using open pit mining methods, it creates the need for additional tailings storage
Bristol Bay Assessment 411 January 2014
-------
Chapter 4 Type of Development
capacity, increased capacity at the mill, increased consumption of utilities such as water and power,
increased production of metal concentrates, and possible extension of the mine life.
BOX 4-4. BLOCK CAVING AND SUBSIDENCE
Subsidence at the ground surface is an inevitable result of the extraction of any underground resource (SME
2011). Block caving causes the surface above the worked-out mine to collapse into the void created by the
removed ore. The area of subsidence on the ground's surface generally is larger than the area actually
block-caved underground (Whittaker and Reddish 1989, USDA 1995). The extent and rate at which
subsidence occurs depend on a number of factors, including the strength and thickness of the overburden,
the extent of faulting and fracturing, and the depth of the mine workings (Whittaker and Reddish 1989).
In addition to altering surface topography, subsidence can affect both the quantity and quality of surface-
water and groundwater systems, either directly or indirectly. For example, Slaughter et al. (1995) observed
both increases and decreases in groundwater levels and changes in groundwater total dissolved solids
concentrations due to subsidence at a coal mine in Utah. The authors attributed the rise in the water table
to stream water seeping through fractures in thestreambed, the subsequent decrease in the water table to
connectivity between streambed fractures and the mine workings, and the total dissolved solids changes to
exposure of the water to mine workings (Slaughter etal. 1995).
Backfilling a mining void is known to reduce subsidence. However, this requires a sufficient amount of
suitable material, which may need to be imported in areas mined with methods that generate little waste
material (SME 2011). Void-filling grout also may be used to mitigate subsidence, as well as to minimize
oxidation of mined surfaces to reduce the potential for production of acid mine drainage.
4.2.3.2 Water Treatment and Management
Because mine workings must be kept dry for the duration of mining activities, dewatering is required for
both open pit mines and block caving operations. Dewatering is accomplished by pumping water either
directly from the pit or underground workings or from wells surrounding these areas. This pumping of
water may create a cone of depression, which is a cone-shaped reduction in water level extending
outward from the point of water withdrawal, where water levels are lowest. Water extracted during
dewatering typically is pumped to lined process water ponds for use in the milling process. Excess water
typically is tested and, if necessary, treated before discharge.
In hard rock metal mining, most water use occurs during milling and separation operations. This water
is obtained from the mine site area and then held in storage facilities until its use. However, much of the
water used in the mining process is recycled and reused. For example, the water used to pump tailings
slurry from the mill to the tailings storage facility (TSF) becomes available when the tailings solids settle
and excess overlying water is recycled back to the mill. Other water use needs include power plant
cooling and transport of metal concentrate slurry (where transport occurs via pipeline).
In general, stormwater runoff is diverted around mine components (e.g., the open pit or waste rock
piles) to keep it from becoming contaminated, and then collected in sedimentation ponds to settle out
suspended solids prior to use or discharge to a stream. Stormwater runoff that contacts mine
components may be contaminated with pollutants. Such water is directed to collection ponds and
treated before being used in mine processes or released. Seepage and leachate are directed to storage
ponds for containment, treated, and released to the environment. Tailings may be dewatered, and
reclaimed water directed to process water holding ponds for reuse. Surface water and groundwater are
Bristol Bay Assessment 412 January 2014
-------
Chapter 4 Type of Development
monitored for contamination throughout mine operations, and are routed to a treatment facility if
significant contamination is detected.
Water treatment options include physical or chemical methods—for example, reverse osmosis
(physical) and formation of precipitated solids (chemical)—used together or independently. The choice
of treatment methods and the chemicals used for treatment depends on the site's specific water
chemistry and the water's end use.
Once mining ceases, an open pit is typically allowed to fill with water. Acid-generating waste rock and
other potentially acid-generating (PAG) materials (e.g., pyrite-rich tailings) may be placed at the bottom
of the pit to submerge these materials and reduce the potential for acid mine drainage once the pit fills.
In block caving, ore is removed from the ground and the resulting void is filled by overlying materials
(Box 4-4). After mining operations cease, groundwater fills in the remaining pore spaces in the void.
4.2.3.3 Ore Processing
Generally, two streams of materials come from a mine: ore and waste rock (Figure 4-3). Ore is rock with
sufficient amounts of metals to be economically processed. Waste rock is material that has little or no
economic value at the time of disturbance, although it may have recoverable value at a future time
(i.e., under different technological or economic conditions).
Ore blasted from a porphyry copper mine typically is hauled to a crushing plant near or in the mine pit
(Figure 4-3). The crushing plant reduces ore to particle sizes manageable in the processing mill (e.g., less
than 15 cm; Ghaffari et al. 2011). Crushed ore is carried by truck or conveyer to a ball mill, where
particle size is further reduced (e.g., 80% to less than 200 um; Ghaffari et al. 2011) to maximize the
recovery of metals. The milled ore is subjected to a flotation process with an aqueous mixture of
chemical reagents (Box 4-5) to collect valuable copper, molybdenum, and gold minerals in a copper-
molybdenum concentrate, which also contains gold. Bulk tailings are the material remaining after the
first flotation circuit, which are directed to a TSF (Figure 4-3). Figure 4-3 assumes NAG bulk tailings;
however, if prior testing has indicated the potential for acid production, they can be treated further to
minimize this potential prior to their disposal. The copper-molybdenum (+gold) concentrate may be fed
through a second ball mill to regrind the particles (e.g., 80% to less than 25 um; Ghaffari et al. 2011).
Once sufficiently sized, the regrind concentrate is directed into a second flotation process and then to a
copper-molybdenum separation process. Final products are a copper concentrate that includes gold, a
molybdenum concentrate, and pyritic tailings (Figure 4-3).
The most profound influence that ore processing can have on long-term management of a mine site
centers on the fate of pyrite (Fuerstenau et al. 2007). Traditionally, PAG and NAG tailings were
discharged together, thereby contributing to the acid-generating potential of the TSF. It is possible to use
a technique called selective flotation to separate most of the pyrite into the cleaner circuit tailings (PAG)
with the rougher tailings (bulk tailings in Figure 4-3) comprising predominantly NAG minerals. The PAG
tailings would need to be stored separately and kept isolated from oxygen.
Bristol Bay Assessment 413 January 2014
-------
Chapter 4
Type of Development
Figure 4-3. Simplified schematic of mined material processing.
Open Pit Mining
Waste Rock
Waste Rock Pile
Processing Facility
Product
Destination
Crushing Plant
Ball Mill
Milled Ore
Flotation Mill
Rougher Concentrate
Regrind Ball Mill
Milled Concentrate
Flotation Mill
Bulk Tailings
(non-acid-generating)
Tailings Storage
Facility
f
Copper (+gold)
Concentrate
Port Site
(via Pipeline)
Molybdenum
Concentrate
Port Site
(via Truck)
Pyritic Tailings
(potentially acid-generating)
Tailings Storage
Facility
Bristol Bay Assessment
4-14
January 2014
-------
Chapter 4 Type of Development
BOX 4-5. CHEMICALS USED IN ORE PROCESSING AND HANDLING
After dry grinding and milling, water is added to the fine ore particles to create a slurry. This slurry undergoes
further beneficiation using chemical reagents to separate minerals from gangue (rock barren of target
minerals) and to separate one mineral from another. Reagents are added to the slurry at different points in
the process to chemically or physically modify the surface of particles and facilitate separation. The amounts
and types of reagents used are site-specific and depend on many factors such as particle size variation,
particle density, ore grade, and host rock character. The volume of reagents used per metric ton of ore is
closely monitored to optimize the mineral concentration process and minimize the unnecessary use of
reagents. Although highly site-specific, most reagents are used at a rate of 0.01 to 0.3 kg of reagent per
metric ton of ore (USEPA 1994a, Khoshdast and Sam 2011). To ensure the flotation system is optimized,
the incoming ore composition is monitored and the reagent mix is modified as changes occur due to
variations in the ore.
The reagents used in flotation generally fall into five categories.
• Collectors (e.g., xanthates, dithiophosphates) increase the ability of air bubbles to stick to a particle.
Toxicity of collectors varies widely within the group, but some commonly used collectors, such as sodium
ethyl xanthate, are toxic to freshwater organisms (Alto et al. 1977, Vigneault et al. 2009).
• pH regulators (e.g., lime, caustic soda, sulfuric acid) are added to maintain the proper pH level in the
slurry. If released, these reagents could affect pH in natural waters.
• Frothers (e.g., aliphatic alcohol, methylisobutyl carbinol, propylene glycol) increase the stability of air
bubbles so they do not burst before bringing a particle to the surface. These reagents are generally
considered to have lowtoxicity (Fuerstenau 2003).
• Flocculants and dispersants (e.g., polyacrylamides, aluminum salts, polyphosphate) promote settling of
fine materials and separation of fine gangue materials. They are generally considered to have low toxicity
(Vigneault etal. 2009).
• Modifiers (e.g., cyanide salts, carboxymethylcellulose) make collectors more effective by either activating
or depressing certain reactions. Toxicity of these reagents varies widely.
Although some of these reagents can be transported to a mine site as powder or pellets, most material
arrives in liquid form.
The gold in porphyry copper deposits is partitioned among the copper-sulfide minerals (chalcopyrite,
bornite, chalcocite, digenite, and covellite), pyrite, and free gold (Kesler et al. 2002). Gold associated
with the copper minerals would stay with the copper (+gold) concentrate and be recovered at an off-site
smelter. Gold associated with pyrite would end up in the TSF unless a separate pyrite concentrate were
produced, and gold could be recovered from this concentrate by a vat leaching cyanidation process
(Logsdon et al. 1999, Marsden and House 2006) (Box 4-6).
Porphyry copper deposits (and other metal deposits) often have marketable quantities of metals other
than the primary target metals. These metals are carried through the flotation process and might be
removed at some later point. As an example, the Pebble deposit is reported to have marketable
quantities of silver, tellurium, rhenium, and palladium (Ghaffari et al. 2011), which are not sufficiently
concentrated in the ore to warrant separation and production of an additional metal concentrate.
The process for removing metals from ore is not 100% efficient. At some point the cost of recovering
more metals exceeds their value, so the amount of metals left in the tailings represents a tradeoff
between revenues from more complete ore processing and extraction costs. The process proposed by
Ghaffari et al. (2011) would recover 86.1% of the copper, 83.6 % of the molybdenum and 71.2% of the
Bristol Bay Assessment 415 January 2014
-------
Chapter 4 Type of Development
gold from the Pebble deposit ore. The residual metals remaining with the tailings would be discharged
to a TSF along with the residue of blasting agents, flotation reagents, and inert portions of the ore.
BOX 4-6. USE OF CYANIDE IN GOLD RECOVERY
At mines producing both copper and gold, copper concentrate and gold dore (unrefined gold produced at the
mine site) are extracted usingstandard processes such as gravity separation and froth flotation. If enhanced
gold recovery is undertaken at the mine site, cyanide is universally used for such gold extraction (Marsden
and House 2006).
The gold recovery process involves a cyanide leach step. The solution that remains after the cyanidation
process is commonly passed through either a cyanide recovery unit or a cyanide destruction unit. Cyanide
recovery allows the recycling of cyanide for reuse in the cyanidation process. Cyanide destruction converts
the cyanide ion to less toxic cyanate, which is then treated in a wastewater treatment plant for discharge or
transferred to a tailings storage facility (TSF). Because the tailings from this process have high
concentrations of acid-generating sulfides, they are typically directed to the TSF, encapsulated in non-acid-
generating tailings, and kept saturated to minimize oxidation. If water is recycled from the TSF into the
copper process water system, cyanide can interfere with the flotation process; to prevent this interference,
some mines isolate cyanidation tailings in a separate TSF (Scott Wilson Mining 2005).
Once in the TSF, cyanide concentrations may decrease through natural attenuation (e.g., volatilization,
photodegradation, biological oxidation, precipitation) (Logsdon et al. 1999). Cyanide may escape the TSF
through seepage or as dust from tailings beaches. Because cyanide dissolves other metals such as copper,
fauna also may be exposed to high metal concentrations and toxic copper-cyanide complexes.
Reported rates of cyanide use at gold mines average about 0.15 to 0.50 kg of cyanide (as sodium cyanate)
per metric ton of concentrate after cyanide recovery (Stange 1999).
4.2.3.4 Tailings Storage
Tailings are a mixture of fine-grained particles, water, and residues of reagents remaining from the
milling process. The most common method of tailings storage is disposal in an impoundment (i.e., a TSF)
(Porter and Bleiwas 2003). Tailings are transported from the mill to a TSF as a slurry, of which solids—
silt to fine sand particles (0.001 to 0.6 mm) with concentrations of metals too low to interact with
flotation reagents—typically make up 30 to 50% by weight. Tailings may be thickened (dewatered)
prior to disposal. Thickening reduces evaporation and seepage losses and allows recycling of more
process water back to the processing plant, thereby reducing operational water demand. It also
minimizes the amount of water stored in the TSF.
Tailings impoundments are water-holding structures typically built by creating a dam in a valley.
Tailings dams are generally earthen or rockfill dams constructed from waste rock or the coarse fraction
of the tailings themselves. The majority of existing tailings dams are less than 30 m in height, but the
largest exceed 150m (McLeod and Murray 2003, National Inventory of Dams 2005, Rico et al. 2008).
The engineering principles governing the design and stability of tailings dams are similar to the
geotechnical principles for earthen and rockfill dams used for water retention. They are typically built in
sections, called lifts, over the lifetime of the mine, such that dam height increases ahead of reservoir
level, using upstream, downstream, or centerline methods (Figure 4-4). Tailings dams built by the
upstream method are less stable against seismic events than dams built by either the downstream or the
centerline method (ICOLD 2001). This is because part of the dam rests on the tailings, which have a
Bristol Bay Assessment 416 January 2014
-------
Chapter 4 Type of Development
lower density and a higher water saturation than the dam materials (USEPA 1994b). Although upstream
construction is considered unsuitable for impoundments intended to be very high or to contain large
volumes of water or solids (State of Idaho 1992), this method is still employed (Davies 2002). For
example, an upstream dam lift was recently designed and constructed on the Fort Knox Mine tailings
impoundment dam near Fairbanks, Alaska (USAGE 2011). The downstream method is considered more
stable from a seismic standpoint, but it is more expensive to implement than the upstream method.
Centerline construction has characteristics of both upstream and downstream types (USEPA 1994b,
Martin etal. 2002).
As they fill with tailings, TSFs must store immense quantities of water (Davies 2011). Water level is
controlled by removing excess water either for use in the mining process or for treatment and
subsequent discharge to local surface waters. Tailings are deposited against the embankment through
spigots or cyclones. Coarser-grained sands are directed at the embankment to create a beach, causing
water and fines to drain away from the dam to form a tailings pond. Care must be taken to prevent the
formation of low-permeability lenses or layers on tailings beaches, as these layers may perch water in
the TSF such that saturation of or flow through the dam may occur, leading to erosion or failure.
Although most of the tailings dam mass consists of fairly coarse and permeable material, the dams often
have a low permeability core to limit seepage, as well as internal drainage structures to collect seepage
water and to control pore pressures. Mitigation measures for seepage through or beneath a tailings dam
may include any combination of liners, seepage cutoff walls, under-drains, or decant systems.
Bristol Bay Assessment 417 January 2014
-------
Chapter 4
Type of Development
Figure 4-4. Cross-sections illustrating (A) upstream, (B) downstream, and (C) centerline tailings
dam construction. In each case, the initial dike is illustrated in light gray, with subsequent dike raises
shown in darker shades (modified from Vick 1983). Tailings dams in our mine scenarios are assumed
to use the downstream construction method initially and at some point change to centerline
construction.
Tailings
B
Tailings
Tailings
Liners may cover the entire impoundment area (e.g., as proposed for the Donlin Creek Mine TSF in
Alaska) or only the pervious bedrock or porous soils. Full liners beneath TSFs are not always used;
however, at least in Australia, mining companies are required to justify why a liner would not be
necessary (e.g., the foundation has a sufficiently low saturated hydraulic conductivity or the
groundwater has no beneficial use) (Commonwealth of Australia 2007). Full liners may not be
economically practicable, in which case partial liners may be used to cover areas of pervious bedrock or
porous soils.
Liners can include a high-density polyethylene, bituminous, or other type of geosynthetic material
(geomembrane) and/or a clay cover over an area of higher hydraulic conductivity. A clay liner may have
a saturated hydraulic conductivity of 10~8 m/s, whereas a geomembrane may have a hydraulic
conductivity of approximately W~w m/s (Commonwealth of Australia 2007). However, geomembrane
technology has not been available long enough to know the service life of these liners. Laboratory tests
and data from landfills suggest that high-density polyethylene liner lifespans range from 69 to 600
years, depending on whether it is the primary (upper) or secondary (lower or backup) liner (Rowe
Bristol Bay Assessment
4-18
January 2014
-------
Chapter 4 Type of Development
2005, Koerner et al. 2011). In general, longer lifespans are expected at lower temperatures and
exposures to light (Rowe 2005, Koerner et al. 2011). Breakdown of the liner material and punctures by
equipment or rocks may limit the effective life of liners (Rowe 2005). Overly steep slopes also may put
stresses on geomembranes and cause them to fail. Service life data for other types of geomembranes are
anecdotal and based on field performance, since no laboratory studies have been conducted (Koerner et
al. 2011).
If seepage is expected or observed, mitigation or remedial measures such as interception trenches or
seepage recovery wells can be installed around the perimeter and downstream of the TSF to capture
water and redirect it to a treatment facility. Precipitation runoff from catchment areas up-gradient of the
TSF is typically diverted away from the impoundment to reduce the volume of stored liquid.
Dry stack tailings management, in which tailings are filtered and "stacked" for long-term storage, is a
newer, less commonly used tailings disposal method. Dry stacked tailings require a smaller footprint,
are easier to reclaim, and have lower potential for structural failure and environmental impacts (Martin
et al. 2002) (Box 4-7). Dry stack technology has found greatest acceptance in arid regions where water is
scarce or expensive, although dry stacks are also used in wet climates or in cold regions where water
handling is difficult (Martin et al. 2002). Currently, the only mines in Alaska that use dry stack tailings
disposal are underground mines with high-grade ore and relatively low quantities of tailings (e.g.,
Greens Creek, a lead, silver, zinc mine in southeast Alaska; Pogo, a gold mine in eastern interior Alaska;
and Nixon Fork, a gold mine in west-central Alaska).
BOX 4-7. DRY STACK TAILINGS MANAGEMENT
In a dry stacking operation, tailings are dried usingfilter presses or vacuum technologies such that water
content typically falls below 20%. The dewatered tailings are either loaded into trucks or transported by
conveyer to the tailings storage facility (TSF), where they are spread in lifts and compacted, similar to a
traditional earth-moving operation.
The compacted tailings have a higher in-place bulk density than tailings placed using more conventional
slurry methods. We estimate that dry stacking would reduce the required volume for tailings storage by
approximately 15%. The lower water content of dry stack tailings means that less water is captured in the
void spaces between solid tailings particles, reducing the amount of water "lost" to the TSF by approximately
one-third. The additional water that is not captured in the TSF is available for treatment and release,
potentially reducing streamflow losses in local streams. The higher density and lower water content of the
tailings also increase their stability. In many cases, the need for a confining embankment and the risk of a
tailings dam failure and tailings liquefaction can be eliminated with dry stack management.
The additional capital costs for dewatering equipment and the high energy cost of dewatering have often
been barriers to adopting dry stack tailings management for low-grade ores such as porphyry copper.
However, higher production costs may be at least partially offset by cost savings in other areas. For example,
the increased stability of a dry stacked TSF may reduce closure costs, post-closure monitoring costs, and
post-closure financial assurance requirements.
Dry stacked tailings are typically placed in unsaturated conditions, which can increase the exposure of
tailings to oxygen. Thus, this type of storage may be less appropriate for potentially acid-generating tailings
or may require additional engineering controls to limit, collect, or treat acid drainage. Where TSFs are
typically used to store water as well as tailings, the use of dry stack tailings may not eliminate the need for
construction and operation of a separate water impoundment facility.
Bristol Bay Assessment 419 January 2014
-------
Chapter 4 Type of Development
4.2.3.5 Waste Rock
Waste rock is rock overlying or removed with the ore body that contains uneconomic quantities of
metals. A waste-to-ore ratio of 2:1—that is, the removal of 2 metric tons of waste rock for each metric
ton of ore—is not uncommon for porphyry copper deposits (Porter and Bleiwas 2003). Waste rock is
stored separately from tailings (Blight 2010), typically in large, terraced stockpiles. Some waste rock
that contains marketable minerals may be stored such that it can be milled if commodity prices increase
sufficiently or if higher than usual metal concentrations in ore require dilution to optimize mill
operation. However, the potential for environmental impacts must be managed if the waste rock is PAG,
via selective handling, drains, diversion systems, or other means. PAG waste rock also may be blended
with ore in the mill to maintain a steady and predictable composition of feed material for the flotation
process over time. NAG waste rock may be placed in piles near the open pit, with ditches to divert
stormwater around the piles and drains (or other systems) to capture leachate or direct it toward the
open pit. At closure, a dry cover (e.g., encapsulation) can be placed over the waste rock pile to isolate it
from water and oxygen, or the pile could be placed into the completed open pit and kept below the
water line if it contains PAG material, depending on site-specific characteristics (O'Kane and Wels 2003).
With small pits and in some settings, it is beneficial to fill the pit with waste rock and other waste
material and then construct a dry cover over the filled pit area.
4.2.4 Timeframes
The mining process described above can be thought of in terms of three distinct periods.
• Operation refers to the period during which the mine is active—that is, the period when mine
infrastructure is being built and ore is being extracted and processed.
• Closure refers to the period following completion of mining operations (either as planned or
prematurely) when mining has ceased and activities related to reclamation and preparation of the
site for future stability continue. During this period, waste areas are reclaimed and facilities needed
to support ongoing monitoring and maintenance activities—such as stormwater management
ditches, monitoring wells, engineered covers on waste materials (if required), wastewater treatment
plants, and roads—are created, retained from the operational period, or replaced or remediated if
they had become compromised.
• Post-closure refers to the extended period following closure activities when monitoring and
maintenance activities continue. During this time, water leaving the site is monitored and treated for
as long as contaminants are present at levels exceeding regulatory standards. The post-closure
phase may last decades, centuries, or longer, until only minimal oversight is required. Such minimal
oversight is necessary, perhaps in perpetuity, to ensure the remaining infrastructure's structural
integrity and to minimize environmental impacts. Given the limited lifetime of human institutions,
continued monitoring and maintenance of the site might become increasingly unlikely as the time
from mine closure increases.
Bristol Bay Assessment 420 January 2014
-------
5.1 Overview of Assessment Endpoints
Selection of assessment endpoints is a key component of the problem formulation stage of an ecological
risk assessment. Each endpoint is an explicit expression of the environmental values of concern in the
assessment, in terms of both the entity valued (e.g., a species, community, or ecological process) and a
potentially at-risk characteristic or attribute of that entity (USEPA 1998). Endpoints can be defined at
any level of ecological organization, from within an organism to across ecosystems, depending on the
needs of the assessment. In all cases, however, selected endpoints should be relevant to both ecology
and decision-maker needs, as well as susceptible to potential stressors (USEPA 1998).
We consider three endpoints in this assessment: (1) the abundance, productivity, or diversity of the
region's Pacific salmon and other fish populations; (2) the abundance, productivity, or diversity of the
region's wildlife populations; and (3) the health and welfare of Alaska Native cultures. Endpoint 1 is
evaluated in terms of direct effects of mining; endpoints 2 and 3 are evaluated indirectly, in terms of
effects resulting from fish-related impacts (i.e., via fish-mediated effects). Each of these endpoints meets
the criteria of ecological relevance, management relevance, and potential susceptibility to stressors
associated with large-scale mining.
The assessment focuses most heavily on Endpoint 1, which is the only endpoint for which direct effects
of mining are considered (Section 2.2.1). Most analyses center on Pacific salmon, rainbow trout, and
Dolly Varden. This focus reflects the ecological, economic, and cultural significance of these fish species,
as well as data availability. Other parts of the region's aquatic ecosystems, including algae, aquatic
invertebrates, and smaller resident fishes such as sculpins, also may be affected by large-scale mining.
However, these taxa are not as relevant to decision makers and data on their distributions, abundances,
and susceptibilities are more limited.
Bristol Bay Assessment
5-1
January 2014
-------
Chapter 5 Endpoints
We evaluate Endpoints 2 and 3 indirectly, in terms of the effects of large-scale mining on Endpoint 1 (i.e.,
via fish-mediated effects). This focus on indirect effects is not meant to suggest that mining would
directly affect only fish populations, or that direct effects of mining on wildlife and Alaska Native
populations would be inconsequential. Rather, it reflects the ecological and regulatory importance of the
region's fisheries and their susceptibility to potential impacts. Under Endpoint 2, we focus on wildlife
species that depend on salmon for food (e.g., brown bear, bald eagles, gray wolves, waterfowl) or that
are important subsistence foods for Alaska Natives (e.g., moose, caribou). Although Alaska Natives are
not the only people who would potentially be affected by mining in the region, Endpoint 3 focuses on
Alaska Native populations because of the centrality of salmon and other salmon-dependent resources to
their way of life and well-being, and because this assessment was initiated in response to requests from
federally recognized tribal governments to restrict large-scale mining in the watersheds. We focus on
the primary Alaska Native cultures of the Nushagak and Kvichak River watersheds, the Yup'ik and
Dena'ina. Sugpiaq people, who traditionally lived along the Alaska Peninsula within the greater Bristol
Bay watershed, still live in this region. However, because the Alaska Peninsula falls outside the
Nushagak and Kvichak River watersheds, these cultures were not included in the assessment (Box 5-1).
We also recognize that non-Native people have lived in the Bristol Bay region for hundreds of years, and
also consider salmon integral to their way of life. Further discussion of the scope of the assessment and
how this scope was defined can be found in Chapters 1 and 2.
In the following sections, we discuss each of the three assessment endpoints in greater detail. We
present information on the fish and wildlife species considered, including what is known about their life
histories, distributions, and abundances both across the Bristol Bay watershed (Scale 1) and within the
Nushagak and Kvichak River watersheds (Scale 2). We discuss the Alaska Native populations in the
region and examine why the region's salmon fisheries are an ecologically, economically, and culturally
important resource.
BOX 5-1. CULTURAL GROUPS IN THE BRISTOL BAY WATERSHED
Within the Bristol Bay watershed there are three main cultural groups: the Yup'ik, the Dena'ina, and the
Sugpiaq. Prior to western contact, these three groups tended to be seasonally dispersed, with large
populations periodically gathering in a central location. Westernization efforts by both Russia and the United
States promoted permanent communities with year-round occupation. Some communities grew around
traditional Alaska Native sites (e.g., Nondalton); other communities were built where resources were more
concentrated or accessible. Naknek is one of the older recorded communities in the Bristol Bay region, with
archaeological surveys indicating that Alaska Natives have occupied the Naknek area for at least 6,000
years.
Although there are descendants of the Sugpiaq that currently live both along the Alaska Peninsula and
within the Nushagak and Kvichak River watersheds, this assessment focuses on the primary cultural groups
found within the Nushagak and Kvichak River watersheds, the Yup'ik and the Dena'ina.
Bristol Bay Assessment c_2 January 2014
-------
Chapter 5 Endpoints
5.2 Endpoint 1: Salmon and Other Fishes
The Bristol Bay watershed is home to at least 29 fish species, representing at least nine different families
(Table 5-1). The region is renowned for its fish populations, and it supports world-class fisheries for
multiple species of Pacific salmon and other game fishes (Dye and Schwanke 2009). These resources
generate significant benefit for commercial fishers, support valued recreational fisheries (Figure 5-1),
and provide sustenance for Alaska Native populations and other rural residents (Figure 5-2, Box 5-2).
In this section we summarize key fish species found in the Bristol Bay watershed, their distributions and
abundances in the region, and some of the factors contributing to the significance of these resources.
This background information is provided to underscore the uniqueness of the region's fisheries and
support the assessment's focus on potential impacts of large-scale mining on these fishes. More detailed
discussion of the region's fishes can be found in Appendices A and B.
Bristol Bay Assessment c_2 January 2014
-------
Chapter 5
Endpoints
Table 5-1. Fish species reported in the Nushagak and Kvichak River watersheds. (H) indicates
species considered to be harvested—that is, they are well-distributed across these watersheds and are
or have been targeted by sport, subsistence, or commercial fisheries. This list does not include
primarily marine species that periodically venture into the lower reaches of coastal streams. See
Appendix B, Table 1, for references and additional information on the abundance and life history of
each species.
Family
Salmonids
(Salmonidae)
Lampreys
(Petromyzontidae)
Suckers
(Catostomidae)
Pikes
(Esocidae)
Species
Bering Cisco
(Coregonus laurettae)
Humpback whitefish (H)
(C. pidschian)
Least Cisco
(C. sardinella)
Pygmy whitefish
(Prosopium coulterii)
Round whitefish
(P. cylindraceum)
Coho salmon (H)
(Oncorhynchus kisutch)
Chinook salmon (H)
(0. tshawytscha)
Sockeye salmon (H)
(0. nerka)
Chum salmon (H)
(0. keta)
Pinksalmon (H)
(0. gorbuscha)
Rainbow trout (H)
(0. my kiss)
Arctic char (H)
(Salvelinus alpinus)
Dolly Varden(H)
(S. ma/ma)
Lake trout (H)
(S. namaycush)
Arctic grayling (H)
(Thymallus arcticus)
Arctic lamprey
(Lethenteron camtschaticum)
Alaskan brook lamprey
(L. a/askense)
Pacific lamprey
(Entosphenus tridentatus)
Longnose sucker
(Catostomus catostomus)
Northern pike (H)
(Esox /L/C/US)
Relative Abundance
Very few specific reports
Common in large upland lakes; locally and seasonally common in large
rivers
Locally common in some lakes (e.g., Lake Clark, morainal lakes near
Iliamna Lake); less common in Iliamna Lake and large slow-moving
rivers such as the Chulitna, Kvichak, and lower Alagnak
Locally common in a few upland lakes or adjacent streams
Abundant/widespread throughout larger streams in upland drainages;
not found in headwaters or coastal plain areas
Juveniles abundant/widespread in upland flowing waters of Nushagak
River watershed and in some Kvichak River tributaries downstream of
Iliamna Lake; present in some Iliamna Lake tributaries; not recorded in
the Lake Clark watershed
Juveniles abundant and widespread in upland flowing waters of
Nushagak River watershed and in Alagnak River; infrequent upstream of
Iliamna Lake
Abundant
Abundant in upland flowing waters of Nushagak River watershed and in
some Kvichak River tributaries downstream of Iliamna Lake; rare
upstream of Iliamna Lake
Abundant (in even years), with restricted distribution, in the Nushagak
River watershed and in some Kvichak River tributaries downstream of
Iliamna Lake; rare upstream of Iliamna Lake
Frequent/common; in summer, closely associated with spawning
salmon
Locally common in upland lakes
Abundant in upland headwaters and selected lakes
Common in larger upland lakes and seasonally present in lake outlets;
absent from the Wood River lakes
Abundant/widespread
Juveniles common/widespread in sluggish flows where fine sediments
accumulate3
Rare
Common in slower flows of larger streams
Common/widespread in still or sluggish waters
Bristol Bay Assessment
5-4
January 2014
-------
Chapter 5
Endpoints
Table 5-1. Fish species reported in the Nushagak and Kvichak River watersheds. (H) indicates
species considered to be harvested—that is, they are well-distributed across these watersheds and are
or have been targeted by sport, subsistence, or commercial fisheries. This list does not include
primarily marine species that periodically venture into the lower reaches of coastal streams. See
Appendix B, Table 1, for references and additional information on the abundance and life history of
each species.
Family
Mudminnows
(Umbridae)
Smelts
(Osmeridae)
Cods
(Gadidae)
Sticklebacks
(Gasterosteidae)
Sculpins
(Cottidae)
Species
Alaska blackfish
(Da ///a pectoralis)
Rainbow smelt
(Osmerus mordax)
Pond smelt
(Hypomesus olidus)
Eulachon
(Thaleichthys pacificus)
Burbot
(Lota lota)
Threespine stickleback
(Gasterosteus aculeatus)
Ninespine stickleback
(Pungitius pungitius)
Coastrange sculpin
(Cottus aleuticus)
Slimy sculpin
(C. cognatus)
Relative Abundance
Locally common/abundant in still or sluggish waters in flat terrain
Seasonally abundant in streams near the coast
Locally common in coastal lakes and rivers, Iliamna Lake, inlet spawning
streams, and the upper Kvichak River; abundance varies widely
interannually
No or few specific reports; if present, distribution appears limited
abundance low
and
Infrequent to common in deep, sluggish, or still waters
Locally abundant in still or sluggish waters; abundant in Iliamna Lake
Abundant/widespread in still or sluggish waters
Abundant/widespread15
Notes:
a These species are combined here, because juveniles, the most commonly encountered life stage for each, are indistinguishable.
b These species are combined here, because they are not reliably distinguished in field conditions, although slimy sculpin is thought to be more
abundant and widely distributed.
Bristol Bay Assessment
5-5
January 2014
-------
Chapter 5
Endpoints
Figure 5-1. Approximate extents of popular Chinook and sockeye salmon recreational fisheries in
the vicinity of the Nushagak and Kvichak River watersheds. Areas were digitized from previously
published maps (Dye et al. 2006). Recreational rainbow trout fisheries are also distributed
throughout the watersheds.
Cook Inlet
Bristol Bay
N
A
Approximate Pebble Deposit Location
Towns and Villages
Watershed Boundary
Chinook Salmon
Sockeye Salmon
Bristol Bay Assessment
5-6
January 2014
-------
Chapter 5
Endpoints
Figure 5-2. Subsistence harvest and harvest effort areas for salmon and other fishes within the
Nushagak and Kvichak River watersheds. Other fishes are defined as those non-salmon and
whitefish species discussed in the text. Each fish category is designated by a representative individual
color and includes all harvest points, lines, or polygons meeting that classification. See Box 5-2 for
more detailed discussion of methodology.
Manokotak
Cook Inlet
Clark's Point
Naknek
<
South Naknek
Bristol Bay
N
A
25
25
50
] Kilometers
50
] Miles
Approximate Pebble Deposit Location
Nonsurveyed Towns and Villages
Surveyed Towns and Villages
Other Fish Harvest Areas
Salmon Harvest Areas
Watershed Boundary
Existing Roads
\
Bristol Bay Assessment
5-7
January 2014
-------
Chapter 5 Endpoints
BOX 5-2. SUBSISTENCE USE METHODOLOGY
Subsistence use and harvest data were extracted from data collected by the Alaska Department of Fish and
Game in collaboration with Stephen R. Braund and Associates (Fall etal. 2006, Kriegetal. 2009, Holen and
Lemons 2010, Holen etal. 2011, Holen etal. 2012). These data are a compilation of a multi-year study to
document and examine baseline subsistence use and harvest (via both directed or targeted efforts and
incidental catches), along with demographic and economic data within the communities near the Pebble
deposit. Eighteen communities were interviewed: Aleknagik, Clark's Point, Dillingham, Igiugig, Iliamna, King
Salmon, Kokhanok, Koliganek, Levelock, Lime Village, Manokotak, Naknek, NewStuyahok, Newhalen,
Nondalton, Pedro Bay, Port Alsworth, and South Naknek.
Members of participating households within each community were asked to document where they hunted,
fished, and gathered subsistence resources during the previous year by adding points (used for harvest
locations), polygons (used for harvest effort areas), and lines (used to depict trap lines or courses travelled
during fish trolling) to various maps. Interviews were conducted from 2004 to 2011; not every community
was interviewed in the same year, so the reported years differed between communities. Following
completion of interviews, hand-drawn maps were digitized and data compiled for use within a geographic
information system. In this assessment, only towns and villages documenting subsistence use and harvest
within the Nushagak and Kvichak River watersheds were considered; data points or sections of polygons
and lines falling outside the boundary of these watersheds were omitted.
Subsistence use and harvest data were extracted for four representative use categories: salmon, other
fishes, wildlife, and waterfowl, based on tables found within each report (e.g., Holen et al. 2012: Table 1-16).
Species or other general classifications within each category include:
• Salmon: chum salmon, Chinook (king) salmon, pink salmon, salmon, coho (silver) salmon, sockeye
salmon, and spawning sockeye (red) salmon
• Other fishes (i.e., non-salmon fish species and whitefishes)'. Arctic char, Dolly Varden, humpback
whitefish, lake trout, least Cisco, rainbow trout, round whitefish, steelhead trout, trout, and whitefish
• Wildlife: black bear, brown bear, caribou, and moose
• Waterfowl: black scoter, brant, Canada goose, eggs, geese, gull eggs, lesser snow goose, mallard, pintail,
sandhill crane, teal, tern eggs, tundra swan, waterfowl, and white-fronted goose
Data were extracted for all points, lines, and polygons in each category, for each interviewed community.
Data were then summed across all communities to produce a cumulative layer for the entire Nushagak and
Kvichak River watersheds. Subsistence intensity across the landscape was derived by first generating a 1-
km square grid across the Nushagak and Kvichak River watersheds. Each documented point, line, and
polygon shapefile was spatially joined and summed across the 1-km grid to account for multiple or
overlapping points, lines, and polygons within the same 1-km pixel. Therefore, each pixel represents the total
number of points and sections of lines and polygons within its boundaries. Subsistence use was then
summed across the four representative use categories to derive total cumulative subsistence use across the
Nushagak and Kvichak River watersheds.
This subsistence use metric provides a coarse measure of areas that are used for subsistence uses more
than others within the watersheds. However, it is important to note some of the limitations of the
subsistence intensity metric. Points represent harvest locations, but the way these data are tabulated does
not confer abundance of species harvested within the pixel. Therefore, a point may represent either a single
capture or multiple captures of a given species. Although abundance information was collected by the
researchers, it was not consistently reported in the geospatial data. Further, the line and polygon files
represent general catch areas and not point of actual capture, allowing broad areas to have the same value
as an actual point of capture. Finally, since this assessment is focused on fish as the main assessment
endpoint, we focus on aquatic species and habitats. Many other plant and animal species included in the
subsistence use databases were not used to arrive at this subsistence intensity metric.
Bristol Bay Assessment CQ January 2014
-------
Chapter 5 Endpoints
5.2.1 Species and Life Histories
5.2.1.1 Salmon
Five species of Pacific salmon spawn and rear in the Bristol Bay watershed's freshwater habitats:
sockeye or red (Oncorhynchus nerka], coho or silver (0. kisutch), Chinook or king (0. tshawytschd), chum
or dog (0. ketd), and pink or humpback (0. gorbuschd). Because no hatchery fish are raised or released
in the watershed, Bristol Bay's salmon populations are entirely wild.
All five salmon species share a trio of life-history traits that contribute to the success and significance of
these species in the Bristol Bay region. First, they are anadromous: they hatch in freshwater habitats,
migrate to sea for a period of relatively rapid growth, and then return to freshwater habitats to spawn.
Second, the vast majority of adults return to their natal freshwater habitats to spawn. This homing
behavior fosters reproductive isolation, thereby enabling populations to adapt to the particular
environmental conditions of their natal habitats (Blair et al. 1993, Dittman and Quinn 1996, Eliason et al.
2011). Homing is not absolute, however, and this small amount of straying increases the probability that
suitable habitats will be colonized by salmon (e.g., Milner and Bailey 1989). Finally, each species is
semelparous: adults die after spawning a single time. After completing their upstream migration,
females excavate nests (redds) in the gravel and release eggs into them. These eggs are fertilized by one
or more competing males as they are released, and the females bury them in the nests. The females and
males then die, depositing the nutrients incorporated into their bodies in their spawning habitats
(Section 5.2.5).
The seasonally of spawning and incubation is roughly the same for all five species, although the timing
can vary somewhat by species, population, and region. In general, salmon spawn from summer through
fall, and fry emerge from spawning gravels the following spring to summer. Freshwater habitats used for
spawning and rearing vary across and within species, and include headwater streams, larger mainstem
rivers, side- and off-channel wetlands, ponds, and lakes (Table 5-2). With some exceptions, preferred
spawning habitat consists of gravel-bedded stream reaches of moderate water depth (30 to 60 cm) and
current (30 to 100 cm/s) (Quinn 2005). Sockeye are unique among the species, in that most populations
rely on lakes as the primary freshwater rearing habitat (Table 5-2).
Both chum and pink salmon migrate to the ocean soon after fry emergence (Heard 1991, Salo 1991).
Because sockeye, coho, and Chinook salmon spend a year or more rearing in the Bristol Bay watershed's
streams, rivers, and lakes before their ocean migration (Table 5-2), these species are more dependent on
upstream freshwater resources than chum and pink salmon. As a result, potential large-scale mining in
this region likely poses greater risks to sockeye, coho, and Chinook salmon.
Bristol Bay Assessment eg January 2014
-------
Chapter 5
Endpoints
Table 5-2. Life history, habitat characteristics, and total documented stream length occupied for
Bristol Bay's five Pacific salmon species in the Nushagak and Kvichak River watersheds.
Salmon
Species
Sockeye
Coho
Chinook
Chum
Pink
Freshwater
Rearing Period
(years)
0-3
1-3
1+
0
0
Freshwater Rearing
Habitat
Lakes, rivers
Headwater streams to
moderate-sized rivers,
headwater springs, beaver
ponds, side channels,
sloughs
Headwater streams to
large-sized mainstem
rivers
Limited
Limited
Ocean
Feeding
Period (years)
2-3
1+
2-4
2-4
1+
Spawning Habitat
Beaches of lakes, streams
connected to lakes, larger
braided rivers
Headwater streams to
moderate sized rivers
Moderate-sized streams
to large-sized mainstem
rivers
Moderate-sized streams
and rivers
Moderate-sized streams
and rivers
Documented
Stream Length
Occupied
(kilometers)
4,600
5,900
4,800
3,400
2,200
Notes:
Data compiled from Appendix A, pages 4-13.
5.2.1.2 Other Fishes
In addition to the five Pacific salmon species discussed above, the Bristol Bay region is home to at least
24 other fish species, most of which typically (but not always) remain within the watershed's freshwater
habitats throughout their life cycles. The region contains highly productive waters for such sport and
subsistence fish species as rainbow trout (0. mykiss}, Dolly Varden (Salvelinus malma), Arctic char (S.
alpinus), Arctic grayling (Thymallus arcticus), humpback whitefish (Coregonuspidschiari), northern pike
[Esox lucius), and lake trout [S. namaycush), as well as numerous other species that are not typically
harvested (Table 5-1). These fish species occupy a variety of habitats throughout the watershed, from
headwater streams to rivers and lakes.
In this assessment, we focus primarily on the five Pacific salmon species, rainbow trout, and Dolly
Varden (Box 2-3). This focus is not meant to imply that other fish species found in the Bristol Bay
watershed are not economically, culturally, or ecologically important, or that they are unlikely to be
affected by potential mining-related activities. Rather, it reflects the value of Pacific salmon, rainbow
trout, and Dolly Varden as both sport and subsistence fisheries throughout the region, the potential
sensitivity of these species to mine development and operation, and the relatively greater amount of
information available for these species, particularly in terms of their distributions and abundances.
The species 0. mykiss includes both a non-anadromous or resident form (commonly referred to as
rainbow trout) and an anadromous form (commonly referred to as steelhead). In the Bristol Bay
watershed, steelhead generally are restricted to a few spawning streams near Port Moller, on the Alaska
Peninsula; thus, most populations throughout the region of the assessment are the non-anadromous
form.
Bristol Bay Assessment
5-10
January 2014
-------
Chapter 5 Endpoints
The spawning habitat and behavior of rainbow trout are generally similar to that of the Pacific salmon
species, with a few key exceptions. First, rainbow trout are iteroparous, meaning that they can spawn
repeatedly. Second, spawning occurs in spring, versus summer and early fall for salmon. Juveniles
emerge from spawning gravels in summer (Johnson et al. 1994, ADF&G 2012), and immature fish may
remain in their natal streams for several years before migrating to other habitats (Russell 1977).
Rainbow trout in the Bristol Bay watershed exhibit complex migratory patterns, moving between
spawning, rearing, feeding, and overwintering habitats. For example, many adults in the region spawn in
inlet or outlet streams of large lakes, then migrate shortly after spawning to feeding areas within those
lakes. Some mature fish may seasonally move distances of 200 km or more (Russell 1977, Burger and
Gwartney 1986, Minard et al. 1992, Meka et al. 2003). Often, these migratory patterns ensure that
rainbow trout are in close proximity to the eggs and carcasses of spawning salmon, which provide an
abundant, high-quality food resource (Meka et al. 2003). The variety of habitat types utilized by rainbow
trout is reflected by different life-history types identified in the region, including lake, lake-river, and
river residents (Meka et al. 2003). See Appendix B (pages 11-16) for additional information on rainbow
trout life history.
Dolly Varden is a highly plastic fish species, with multiple genetically, morphologically, and ecologically
distinct forms that can co-exist in the same water bodies (Ostberg et al. 2009). Both anadromous and
non-anadromous Dolly Varden are found in the Bristol Bay watershed, and both life-history forms can
exhibit complex and extensive migratory behavior (Armstrong and Morrow 1980, Reynolds 2000,
Scanlon 2000, Denton et al. 2009). Anadromous individuals usually undertake three to five ocean
migrations before reaching sexual maturity (DeCicco 1992, Lisac and Nelle 2000, Crane et al. 2003).
During these migrations, Dolly Varden frequently leave one drainage, travel through marine waters, and
enter a different, distant drainage (DeCicco 1992, DeCicco 1997, Lisac 2009). Non-anadromous
individuals also may move extensively between different habitats (Scanlon 2000).
Dolly Varden spawning occurs in fall, upstream of overwintering habitats (DeCicco 1992). Northern-
form anadromous Dolly Varden (the geographic form of Dolly Varden found north of the Alaska
Peninsula) overwinter primarily in lakes and in lower mainstem rivers where sufficient groundwater
provides suitable volumes of free-flowing water (DeCicco 1997, Lisac 2009). Within the Nushagak and
Kvichak River watersheds, juveniles typically rear in low-order, high-gradient stream channels (ADF&G
2012). Because Dolly Varden occur in upland lakes and high-gradient headwater streams (ADF&G
2012)—farther upstream than many other fish species and above migratory barriers to anadromous
salmon populations—they may be especially vulnerable to mine development and operation in these
headwater areas. See Appendix B (pages 20-25) for additional information on Dolly Varden life history.
It is important to note that these endpoint species do not exist in isolation from other fish species. The
biomass carried into the Bristol Bay watershed's aquatic habitats by spawning salmon is a fundamental
driver of aquatic foodwebs (Box 5-3). Many of the species listed in Table 5-1 are prey for, predators of,
or competitors with the endpoint species. For example, sculpins, Dolly Varden, and rainbow trout are
well-known predators of salmon eggs and emergent fry, and northern pike can be effective predators of
Bristol Bay Assessment 5-11 January 2014
-------
Chapter 5
Endpoints
juvenile salmon and other fish species (Russell 1980, Sepulveda et al. 2013). Insectivorous and
planktivorous fishes (e.g., Arctic grayling and pond smelt, respectively) may prey on similar species as
juvenile salmonids (e.g., Hartman and Burgner 1972). Given these foodweb interactions, we recognize
that shifts in the relative abundance of species are likely to have repercussions throughout the aquatic
community; however, evaluation of the myriad foodweb interactions that could result from large-scale
mining is beyond the scope of this assessment.
BOX 5-3. SALMON IN FRESHWATER AND TERRESTRIAL FOODWEBS
Salmon are a cornerstone species in the Bristol Bay region, in that they comprise a significant portion of the
resource base upon which both aquatic and terrestrial ecosystems in the region depend (Willson etal.
1998). Adults returning to freshwater systems to spawn import marine-derived nutrients (MDN) back into
these freshwater habitats. These nutrients provide the foundation for aquatic and terrestrial foodwebs via
two main pathways: direct consumption of salmon in any of its forms (spawning adults, eggs, carcasses,
and/or juveniles) and nutrient recycling (Gende etal. 2002).
Because salmon are a seasonally abundant, high-quality food resource in the Bristol Bay watershed, many
aquatic and terrestrial species take advantage of this resource (e.g., see Sections 5.3 and 12.1). For
example, Willson and Halupka (1995) found that more than 40 species of mammals and birds feed on
salmon in southeastern Alaska. Salmon eggs and juveniles are eaten by many fishes, such as other salmon,
rainbow trout, northern pike, and Dolly Varden (Appendix B).
The nutrients incorporated into spawningsalmon biomass also can have a bottom-up effect on both
freshwater and terrestrial ecosystems via nutrient recycling (Gende et al. 2002). Given that these systems
tend to be nutrient-poor, MDN contributions play a significant role in the Bristol Bay region's productivity. In
lakes and streams, MDN help to fuel the production of algae, bacteria, fungi, and other microorganisms that
make up aquatic biofilms. These biofilms in turn provide food for aquatic invertebrates, which are preyed on
by juvenile salmon and other fishes. Terrestrial vegetation and invertebrates also receive a salmon-related
nutrient subsidy, in the form of carcasses and excreta deposited on land by mammal and bird consumers.
FRESHWATER ECOSYSTEM
TERRESTRIAL ECOSYSTEM
juvenile
salmon
mammals & birds
invertebrates
invertebrates
algae, bacteria, fungi &
other microorganisms
riparian & terrestrial
vegetation
spawningsalmon,
eggs, carcasses, &
associated MDN
i carcasses & [
1 associated MDN <
Note that the simplified foodweb above (modified from Willson et al. 1998) focuses on how salmon serve as
a resource base within and across freshwater and terrestrial ecosystems. Not all interactions, particularly
those mediated by other species (e.g., invertebrates) and those that cross between freshwater and
terrestrial ecosystems, are shown on this schematic. It also does not illustrate the role of salmon in
estuarine and marine foodwebs, as these habitats are outside the scope of this assessment.
Bristol Bay Assessment
5-12
January 2014
-------
Chapter 5 Endpoints
5.2.2 Distribution and Abundance
Fish populations throughout the Bristol Bay watershed have not been sampled comprehensively; thus,
estimates of total distribution and abundance across the region are not available. However, available
data (e.g., the Anadromous Waters Catalog, the Alaska Freshwater Fish Inventory, escapement and
harvest data) provide at least minimum estimates of where key species are found and how many
individuals of those species have been caught. More information on the distribution and abundance of
key fish species can be found in Appendices A and B. See Section 7.2.5 for additional information on the
interpretation of available fish distribution data.
5.2.2.1 Salmon
Most (63%) of the subwatersheds in the Nushagak and Kvichak River watersheds are documented to
contain at least one species of spawning or rearing salmon within their boundaries, and 12% are
documented to contain all five species (Figure 5-3). Reported distributions for each salmon species in
the Nushagak and Kvichak River watersheds are shown in Figures 5-4 through 5-8.
Sockeye is by far the most abundant salmon species in the Bristol Bay watershed (Table 5-3) (Salomone
et al. 2011). Bristol Bay is home to the largest sockeye salmon fishery in the world, with 46% of the
average global abundance of wild sockeye salmon between 1956 and 2005 (Figure 5-9A) (Ruggerone et
al. 2010). Between 1990 and 2009, the average annual inshore run of sockeye salmon in Bristol Bay was
approximately 37.5 million fish (ranging from a low of 16.8 million in 2002 to a high of 60.7 million in
1995) (Salomone et al. 2011). Annual commercial harvest of sockeye over this period averaged
25.7 million fish (Table 5-3), and 78% of the average annual subsistence salmon harvest
(140,767 salmon) over this period were sockeye (Dye and Schwanke 2009, Salomone et al. 2011).
Escapement goals—that is, the number of individuals allowed to escape the fishery and spawn, to
ensure long-term sustainability of the stock—vary by species and stock. The current sockeye
escapement goal for the Kvichak River ranged from 2 to 10 million fish (Box 5-4). Annual sport harvest
of sockeye in recent years has ranged from approximately 8,000 to 23,000 fish (Dye and Schwanke
2009).
More than half of the Bristol Bay watershed's sockeye salmon harvest comes from the Nushagak and
Kvichak River watersheds (Figure 5-9B). Sockeye returns to the Kvichak River averaged 10.5 million fish
between 1963 and 2011, and this number climbs to 12.1 million fish when returns to the Alagnak River
are included (Cunningham et al. 2012). Kvichak River sockeye runs have exceeded 30 million fish three
times since 1956, with 48.6, 34.9, and 37.9 million fish in 1965,1970, and 1980, respectively
(Cunningham et al. 2012).
Tributaries to Iliamna Lake, Lake Clark, and the Wood-Tikchik Lakes (Figure 2-4) are major sockeye
spawning areas, and juveniles rear in each of these lakes (Figure 5-4). Iliamna Lake provides the
majority of sockeye rearing habitat in the Kvichak River watershed, and historically has produced more
sockeye than any other lake in the Bristol Bay region (Fair et al. 2012). Riverine sockeye populations
spawn and rear throughout the Nushagak River watershed (Figure 5-4).
Bristol Bay Assessment 5-13 January 2014
-------
Chapter 5
Endpoints
Table 5-3. Mean annual commercial harvest (number of fish) by Pacific salmon species and Bristol
Bay fishing district, 1990 to 2009a. Number in parentheses indicates percentage of total found in
each district.
Salmon
Species
Sockeye
Chinook
Coho
Chum
Pinkb
Bristol Bay Fishing District
Naknek-
Kvichak3
8,238,895 (32)
2,816 (4)
4,436 (5)
184,399 (19)
73,661 (43)
Egegik
8,835,094 (34)
849 (1)
27,433 (33)
78,183(8)
1,489 (1)
Ugashik
2,664,738(11)
1,402 (2)
10,425 (12)
70,240 (7)
138 (<1)
Nushagak3
5,478,820 (21)
52,624 (80)
27,754 (33)
493,574 (50)
50,448 (30)
Togiak
514,970 (2)
8,803 (13)
14,234 (17)
158,879 (16)
43,446 (26)
Total
25,732,517
66,494
84,282
985,275
169,182
Notes:
a Naknek-Kvichak district includes the Alagnak River; Nushagak district includes the Wood and Igushik Rivers.
b Pink salmon data are from even-numbered years; harvest is negligible during odd-year runs.
Source: Appendix A, Table 1.
Chinook salmon spawn and rear throughout the Nushagak River watershed and in several tributaries of
the Kvichak River (Figure 5-5), and they are an important subsistence food for residents of both
watersheds. Although Chinook is the least common salmon species across the Bristol Bay region, the
Nushagak River watershed supports a large Chinook salmon fishery and its commercial and sport-
fishing harvests are greater than those of all other Bristol Bay river systems combined (Table 5-3).
Chinook returns to the Nushagak River are consistently greater than 100,000 fish per year, and have
exceeded 200,000 fish per year in 11 years between 1966 and 2010. This frequently places the
Nushagak at or near the size of the world's largest Chinook runs, which is notable given the Nushagak
River's small watershed area compared to other Chinook-producing rivers such as the Yukon River,
which spans Alaska and much of northwestern Canada, and the Kuskokwim River in southwestern
Alaska, just north of Bristol Bay.
Coho salmon spawn and rear in many stream reaches throughout the Nushagak and lower Kvichak River
watersheds (Figure 5-6). Juveniles distribute widely into headwater streams, where they are often the
only salmon species present (Woody and O'Neal 2010, King et al. 2012). Production of juvenile coho is
often limited by the extent and quality of available overwintering habitats (Nickelson et al. 1992, Solazzi
etal. 2000).
Chum salmon is the second most abundant salmon species in the Nushagak and Kvichak River
watersheds (Table 5-3). Both chum and pink salmon spawn throughout the Nushagak and lower
Kvichak River watersheds (Figures 5-7 and 5-8), but do not have an extended freshwater rearing stage.
Bristol Bay Assessment
5-14
January 2014
-------
Chapter 5
Endpoints
BOX 5-4. COMMERCIAL FISHERIES MANAGEMENT IN THE BRISTOL BAY WATERSHED
Commercial fisheries management in Alaska is largely focused on achieving escapement goals—management
goals based on the optimum range of fish numbers allowed to escape the fishery and spawn—rather than harvest
rates (Fair et al. 2012). Thus, management involves allowing an adequate number of spawners to reach each river
system while maximizing harvest in the commercial fishery (Salomone et al. 2011). Bristol Bay's commercial
salmon fisheries are considered a management success (Hilborn et al. 2003, Hilborn 2006). Several factors have
contributed to this success, including a clear management objective of maximum sustainable yield, the
escapement goal system, management responsibility fallingto a single agency, a permit system that limits the
number of fishers, and favorable freshwater habitats and ocean conditions (Hilborn et al. 2003, Hilborn 2006).
Escapement goals for sockeye salmon in the nine major rivers draining the Bristol Bay watershed are listed in the
table below. The Alaska Department of Fish and Game (ADF&G) regularly reviews escapement goals for the major
salmon stocks in Bristol Bay. These reviews include updates to escapement estimates, revisions to how catch is
partitioned to stocks, and revisions to stock-recruit models used to recommend escapement goals. For example,
data on sockeye genetic stock composition, age composition, and run timing were used to reconstruct brood
tables for the major stocks in 2012 (Cunningham et al. 2012, Fair et al. 2012).
The Kvichak River frequently did not meet its sockeye escapement goal from 1991 through 1999, and in 2001 it
was placed into special management status due to chronic low yields (Fair 2003). The cause of this low
productivity in Kvichak River sockeye is not entirely known, but marine conditions likely led to this decline (see
Appendix A, pages 31-33, for a more detailed discussion of this decline). However, the Kvichak River stock is
considered to be rebuilding: escapement goals have been met for the last 5 years, and in 2012 ADF&G
recommended that it be removed from special management status (Morstad and Brazil 2012).
Sockeye Salmon Escapement Goals in the Bristol Bay Watershed
River
Kvichak
Alagnak
Naknek
Egegik
Ugashik
Wood
Igushik
Nushagak-Mulchatna
Togiak
Escapement Range
(thousands of fish)
2,000-10,000
320 minimum
800-1,400
800-1,400
500-1,200
700-1,500
150-300
370-840
120-270
Once escapement goals are set, the timing and duration of commercial fishery openings are adjusted throughout
the fishing season to ensure that escapement goals are met and any additional fish are harvested. Fishery
openings are based on information from a number of sources, including pre-season forecasts (expected returns of
the dominant age classes in a given river system, based on the number of spawning adults that produced each
age class); the test fishery at Port Moller on the Alaska Peninsula; early performance of the commercial fishery;
and in-river escapement monitoring. At the beginning of the fishing season, the frequency and duration of
openings are primarily based on pre-season forecasts and are managed conservatively. As the season progresses
and additional information becomes available, fishing times and areas are continuously adjusted via emergency
orders. If the escapement goal is exceeded at a given monitoring station, the fishery is opened longer and more
frequently. If the escapement goal is not reached, the fishery is closed.
This type of in-season management is also used to meet a Chinook salmon escapement goal for the Nushagak
River (55,000-120,000 fish). There is a chum salmon escapement goal for the Nushagak River (200,000 fish
minimum) and there are Chinook salmon escapement goals for the Alagnak and Naknek Rivers; however, in-
season management is not used to help attain these goals (Baker et al. 2009).
See Appendix A for a more detailed discussion of historical and current fisheries management in the Bristol Bay
region.
Bristol Bay Assessment
5-15
January 2014
-------
Chapter 5
Endpoints
Figure 5-3. Diversity of Pacific salmon species production in the Nushagak and Kvichak River
watersheds. Counts of salmon species (sockeye, Chinook, coho, pink, and chum) spawning and
rearing, based on the Anadromous Waters Catalog (Johnson and Blanche 2012), are summed by 12-
digit hydrologic unit codes. See Section 7.2.5 for details on interpretation of distribution data.
Bristol Bay
Approximate Pebble Deposit Location Number of Species Documented
Towns and Villages
Mine Scenario Watersheds
Watershed Boundary
Cook Inlet
N
A
25
25
50
] Kilometers
50
] Miles
Bristol Bay Assessment
5-16
January 2014
-------
Chapter 5
Endpoints
Figure 5-4. Reported sockeye salmon stream distribution in the Nushagak and Kvichak River
watersheds. "Present" indicates species was present but life-stage use was not determined;
"spawning" indicates spawning adults were observed; "rearing" indicates juveniles were observed.
Present, spawning, and rearing designations are based on the Anadromous Waters Catalog (Johnson
and Blanche 2012). Life-stage-specific reach designations are likely underestimates, given the
challenges inherent in surveying all streams that may support life-stage use throughout the year. See
Section 7.2.5 for details on interpretation of fish distribution data.
Cook Inlet
Bristol Bay
"if Approximate Pebble Deposit Location
Present
Spawning
Rearing
Mine Scenario Watersheds
Watershed Boundary
IN
A
25
50
]Miles
Bristol Bay Assessment
5-17
January 2014
-------
Chapter 5
Endpoints
Figure 5-5. Reported Chinook salmon distribution in the Nushagak and Kvichak River watersheds.
"Present" indicates species was present but life-stage use was not determined; "spawning" indicates
spawning adults were observed; "rearing" indicates juveniles were observed. Present, spawning, and
rearing designations are based on the Anadromous Waters Catalog (Johnson and Blanche 2012).
Life-stage-specific reach designations are likely underestimates, given the challenges inherent in
surveying all streams that may support life-stage use throughout the year. See Section 7.2.5 for
details on interpretation offish distribution data.
Cook Inlet
Bristol Bay
Approximate Pebble Deposit Location
Present
Spawning
Rearing
Mine Scenario Watersheds
Watershed Boundary
IN
A
25
50
] Miles
Bristol Bay Assessment
5-18
January 2014
-------
Chapter 5
Endpoints
Figure 5-6. Reported coho salmon distribution in the Nushagak and Kvichak River watersheds.
"Present" indicates species was present but life-stage use was not determined; "spawning" indicates
spawning adults were observed; "rearing" indicates juveniles were observed. Present, spawning, and
rearing designations are based on the Anadromous Waters Catalog (Johnson and Blanche 2012).
Life-stage-specific reach designations are likely underestimates, given the challenges inherent in
surveying all streams that may support life-stage use throughout the year. See Section 7.2.5 for
details on interpretation offish distribution data.
Cook Inlet
Bristol Bay
"If Approximate Pebble Deposit Location
Present
Spawning
Rearing
Mine Scenario Watersheds
Watershed Boundary
A
25
50
] Miles
Bristol Bay Assessment
5-19
January 2014
-------
Chapter 5
Endpoints
Figure 5-7. Reported chum salmon distribution in the Nushagak and Kvichak River watersheds.
"Present" indicates species was present but life-stage use was not determined; "spawning" indicates
spawning adults were observed; "rearing" indicates juveniles were observed. Present, spawning, and
rearing designations are based on the Anadromous Waters Catalog (Johnson and Blanche 2012).
Life-stage-specific reach designations are likely underestimates, given the challenges inherent in
surveying all streams that may support life-stage use throughout the year. See Section 7.2.5 for
details on interpretation offish distribution data.
Cooh Inlet
Bristol Bay
Approximate Pebble Deposit Location
Present
Spawning
Rearing
Mine Scenario Watersheds
Watershed Boundary
IN
A
25
50
] Miles
Bristol Bay Assessment
5-20
January 2014
-------
Chapter 5
Endpoints
Figure 5-8. Reported pink salmon distribution in the Nushagak and Kvichak River watersheds.
"Present" indicates species was present but life-stage use was not determined; "spawning" indicates
spawning adults were observed. Present and spawning designations are based on the Anadromous
Waters Catalog (Johnson and Blanche 2012). Life-stage-specific reach designations are likely
underestimates, given the challenges inherent in surveying all streams that may support life-stage
use throughout the year. See Section 7.2.5 for details on interpretation of distribution data.
Cook Inlet
Bristol Bay
'if Approximate Pebble Deposit Location
Present
Spawning
Mine Scenario Watersheds
I I Watershed Boundary
A
25 50
25
] Kilometers
50
DMiles
Bristol Bay Assessment
5-21
January 2014
-------
Chapter 5
Endpoints
Figure 5-9. Proportion of total sockeye salmon run sizes by (A) region and (B) watershed within the
Bristol Bay region. Values are averages from (A) 1956-2005 from Ruggerone et al. 2010 and (B)
1956-2010 from Baker pers. comm. (Appendix A: Tables A2 and A3).
• Bristol Bay
• Russia Mainland & Islands
• West Kamchatka
• East Kamchatka
• Western Alaska (excluding Bristol Bay)
• South Alaska Peninsula
I .Kodiak
r Cook Inlet
D Prince William Sound
DSoutheast Alaska
D North British Columbia
D South British Columbia, Washington & Oregon
B /^""^^^
\ BTogiak
\ \ BNushagak
DKvichak
DNaknek
\ W DEgegik
DUshagik
Bristol Bay Assessment
5-22
January 2014
-------
Chapter 5 Endpoints
5.2.2.2 Other Fishes
Extensive sampling for rainbow trout and Dolly Varden has not been conducted throughout the Bristol
Bay region, so total distributions and abundances are unknown. Figures 5-10 and 5-11 show the
reported occurrence of rainbow trout and Dolly Varden throughout the Nushagak and Kvichak River
watersheds and provide minimum estimates of their extents.
Between 2003 and 2007, an estimated 183,000 rainbow trout were caught in the Bristol Bay
Management Area (Dye and Schwanke 2009). Radio telemetry, tagging, and genetic studies indicate that
multiple rainbow trout populations are found within Bristol Bay watersheds (Gwartney 1985, Burger
and Gwartney 1986, Minard et al. 1992, Krueger et al. 1999, Meka et al. 2003). The most popular
rainbow trout fisheries are found in the Kvichak River watershed, the Naknek River watershed, portions
of the Nushagak and Mulchatna River watersheds, and streams of the Wood River lakes system (Dye and
Schwanke 2009).
Dolly Varden populations are a significant subsistence resource. In the mid-2000s, subsistence harvests
of Dolly Varden and Arctic char combined (Alaska's fisheries statistics do not distinguish between the
two species) were estimated at 3,450 fish for 10 communities in the Nushagak and Kvichak River
watersheds (Fall et al. 2006, Krieg et al. 2009). From the mid-1970s to the mid-2000s, these two species
were estimated to represent between 16.2 and 26.9% of the total weight of the Kvichak River
watershed's non-salmon freshwater fish subsistence harvest (Krieg et al. 2005). Dolly Varden also
support a popular sport fishery.
Bristol Bay Assessment 5-23 January 2014
-------
Chapter 5
Endpoints
Figure 5-10. Reported rainbow trout occurrence in the Nushagak and Kvichak River watersheds.
Designation of species presence is based on the Alaska Freshwater Fish Inventory (AFFI point data,
ADF&G 2012). Note that points shown on land actually occur in smaller streams not shown on this
map. Absence cannot be inferred from this map. See Section 7.2.5 for details on interpretation of fish
distribution data.
Cook Inlet
Bristol Bay
Approximate Pebble Deposit Location
Present (AFFI)
Mine Scenario Watersheds
Watershed Boundary
N
A
0 25
25
50
] Kilometers
50
] Miles
Bristol Bay Assessment
5-24
January 2014
-------
Chapter 5
Endpoints
Figure 5-11. Reported Dolly Varden occurrence in the Nushagak and Kvichak River watersheds.
Designation of species presence is based on the Alaska Freshwater Fish Inventory (AFFI point data,
ADF&G 2012) and the Anadromous Waters Catalog (AWC line data, Johnson and Blanche 2012).
Note that points shown on land actually occur in smaller streams not shown on this map. Absence
cannot be inferred from this map. See Section 7.2.5 for details on interpretation of fish distribution
data.
Cook Inlet
Bristol Bay
Approximate Pebble Deposit Location
Present (AFFI)
Present (AWC)
Mine Scenario Watersheds
Watershed Boundary
N
A
25
50
] Miles
Bristol Bay Assessment
5-25
January 2014
-------
Chapter 5
Endpoints
5.2.3 Economic Implications
The Bristol Bay watershed supports several sustainable, wilderness-compatible economic sectors,
including commercial fishing, sport fishing, subsistence hunting and fishing, recreational hunting, and
wildlife viewing and other non-consumptive recreation. Each of these sectors generates expenditures or
sales that drive the region's economy, generating roughly $480 million (in 2009 dollars) in total direct
annual economic benefit (Table 5-4).
Table 5-4. Summary of regional economic expenditures based on salmon ecosystem services.
Values are regional expenditures in different economic sectors, expressed in 2009 dollars. Note that
estimates of certain year-specific total harvest and sales values vary slightly throughout this report,
due to differences in how data were aggregated and reported. See Appendix E for additional
information on these values.
Economic Sector
Commercial fisheries, wholesale value
Sport fisheries
Sport hunting
Wildlife viewing/ tourism
Subsistence harvest
TOTAL
Estimated Direct Expenditure
(sales per year, in $ millions)
300.2
60.5
8.2
104.4
6.3
479.6
Roughly 75% of this annual economic benefit results directly from the commercial, sport, and
subsistence fishing supported by the Bristol Bay watershed. The commercial salmon fishery currently
provides the region's greatest source of economic activity. From 2000 through 2010, the annual
commercial salmon catch averaged 23 million fish (170 million pounds). The average annual
commercial value of all Bristol Bay salmon fisheries from 1990 to 2010 totaled $116.7 million,
$114.7 million of which resulted from the sockeye harvest (Salomone et al. 2011). Thus, sockeye salmon
represent the principal species of economic value throughout the Bristol Bay region.
In 2009, fishers received $144 million for their catch, and fish processors received approximately
$300 million, which is referred to as the first wholesale value of the fish (Table 5-4, Appendix E). The
commercial salmon fishery, which is largely centered in the region's salt waters rather than its
freshwater streams and rivers, is closely managed for sustainability using a permit system (Box 5-4).
Approximately 26% of permit holders are Bristol Bay residents. The commercial fishery also provides
significant employment opportunities, directly employing over 11,000 full- and part-time workers at the
season's peak.
The uncrowded, pristine wilderness setting of the Bristol Bay watershed attracts recreational fishers,
and aesthetic qualities are rated as most important in selecting fishing locations by Bristol Bay anglers.
Sport fishing in Bristol Bay accounts for approximately $60.5 million in annual spending (Table 5-4),
$58 million of which is spent in the Bristol Bay region. In 2009, approximately 29,000 sport-fishing trips
were taken to the Bristol Bay region (12,000 trips by people living outside of Alaska, 4,000 trips by
Alaskans living outside the Bristol Bay area, and 13,000 trips by Bristol Bay residents). These sport
Bristol Bay Assessment
5-26
January 2014
-------
Chapter 5 Endpoints
fishing activities directly employ over 800 full- and part-time workers. In 2010, 72 businesses and
319 guides were operating in the Nushagak and Kvichak River watersheds alone, down from a peak of
92 businesses and 426 guides in 2008 (Appendix A, Table 4).
Many households participate in the subsistence harvest of fish, which generates regional economic
benefits when Alaskan households spend money on subsistence-related supplies. In total, individuals in
Bristol Bay communities harvest about 2.6 million pounds of subsistence foods per year. In 2010, the
U.S. Census Bureau reported an estimated 1,873 Alaska Native and 666 non-native households in the
Bristol Bay region. Goldsmith et al. (1998) estimated that Alaska Native households spend an average of
$3,054 on subsistence harvest supplies, whereas non-native households spend an estimated $796 on
supplies (values updated to 2009 price levels). Based on these estimates, subsistence harvest activities
resulted in expenditures of approximately $6.3 million (Table 5-4). It is important to note that these
estimates of expenditures reflect only the annual economic activity generated by these activities and not
the value of the subsistence resources harvested. It may be useful to consider calculations such as net
economic value, or the value of the resource or activity over and above regular expenditures associated
with it. These types of calculations, as well as the regional economic significance of Bristol Bay's salmon
fishery, are discussed in Appendix E.
5.2.4 Biological Complexity and the Portfolio Effect
As the previous sections illustrate, the Bristol Bay watershed supports world-class salmon fisheries.
These fisheries result from numerous, interrelated factors. Closely tied to the Bristol Bay region's
physical habitat complexity (Chapter 3) is its biological complexity, which greatly increases the region's
ecological productivity and stability. This biological complexity operates at multiple scales and across
multiple species, but it is especially evident in the watershed's Pacific salmon populations. As discussed
in Section 5.2.1.1, the five Pacific salmon species found in the Bristol Bay watershed vary in many life-
history characteristics (Table 5-5). This variability allows them to fully exploit the range of habitats
available throughout the watershed. Even within a single species, life histories can vary significantly. For
example, sockeye salmon may spend anywhere from 0 to 3 years rearing in freshwater habitats, then 1
to 4 years feeding at sea, before returning to the Bristol Bay watershed anytime within a 4-month
window (Table 5-5).
Bristol Bay Assessment 5-27 January 2014
-------
Chapter 5
Endpoints
Table 5-5. Life-history variation within Bristol Bay sockeye salmon populations.
Element of Biological Complexity
Location within the Bristol Bay watershed
Time of adult return to freshwater habitats
Time of spawning
Spawning habitat
Body size of adults
Body shape of adults
Egg size
Time between entry into spawning habitat and death
Time spent rearing in freshwater
Time spent at sea
Range of Traits or Options
7 major subwatersheds, ranging from maritime-influenced systems on
the Alaska Peninsula to more continental systems
June-September
July-November
Major rivers, small streams, spring-fed ponds, mainland beaches,
island beaches
130 to 190-mm body depth at450-mm male length
Sleek, fusiform to very deep-bodied, with exaggerated humps and jaws
88-116 mgat450-mm female length
Days-weeks
0-3 years
1-4 years
Notes:
Data from Hilborn et al. 2003.
This life-history variability, together with the Pacific salmon's homing behavior, results in distinct
populations adapted to their own specific spawning and rearing habitats (Hilborn et al. 2003). In the
Bristol Bay region, hydrologically diverse riverine and wetland landscapes provide a variety of large
river, small stream, floodplain, pond, and lake habitats for salmon spawning and rearing, and
environmental conditions can differ among habitats in close proximity. Variations in temperature and
streamflow associated with seasonally and groundwater-surface water interactions create a habitat
mosaic that supports a range of spawning times across the watersheds. Spawning adults return at
different times and to different locations, creating and maintaining a degree of reproductive isolation
and allowing development of genetically distinct stocks (Hilborn et al. 2003, McGlauflin et al. 2011).
These distinct stocks can occur at fine spatial scales, with sockeye salmon that use spring-fed ponds and
streams approximately 1 km apart exhibiting differences in spawn timing, spawn site fidelity,
productivity, and other traits that are consistent with discrete populations (Quinn et al. 2012).
Thus, the Bristol Bay watershed's sockeye salmon "population" is actually a sockeye salmon stock
complex—that is, a combination of hundreds of genetically distinct populations, each adapted to
specific, localized environmental conditions (Hilborn et al. 2003, Schindler et al. 2010). This stock
complex structure can be likened to a financial portfolio in which assets are divided among diverse
investments to increase financial stability. Essentially, it creates a biological portfolio effect (Schindler et
al. 2010), stabilizing salmon productivity across the watershed as a whole as the relative contribution of
sockeye with different life-history characteristics, from different regions of the Bristol Bay watershed,
changes over time in response to changes in environmental conditions (Hilborn et al. 2003). For
example, salmon stocks that spawn in small streams may be negatively affected by low-streamflow
conditions, whereas stocks that spawn in lakes may not be affected (Hilborn et al. 2003). Thus, any
population containing stocks that vary in spawning habitat is better able to persist as environmental
conditions change.
Bristol Bay Assessment
5-28
January 2014
-------
Chapter 5 Endpoints
Without this high level of system-wide biological complexity, annual variability in the size of Bristol
Bay's sockeye salmon runs would be expected to more than double and fishery closures would be
expected to become more frequent (Schindler et al. 2010). In other watersheds with previously robust
salmon fisheries, such as the Sacramento River's Chinook fishery, losses of biological complexity have
contributed to salmon population declines (Lindley et al. 2009). These findings suggest that even the
loss of a small stock within an entire watershed's salmon population may have more significant effects
than expected, due to associated decreases in biological complexity of the population's stock complex.
5.2.5 Salmon and Marine-Derived Nutrients
Adult salmon returning to their natal freshwater habitats import nutrients that they obtained during
their ocean feeding period—that is, marine-derived nutrients (MDN)—back into those habitats
(Cederholm et al. 1999, Gende et al. 2002). Because approximately 95 to 99% of the carbon, nitrogen,
and phosphorus in an adult salmon's body are derived from the marine environment (Larkin and Slaney
1997, Schindler et al. 2005), MDN from salmon account for a significant portion of nutrient budgets in
the Bristol Bay watershed (Kline et al. 1993). For example, sockeye salmon are estimated to import
approximately 12,700 kg of phosphorus and 101,000 kg of nitrogen into the Wood River system
annually, and 50,200 kg of phosphorus and 397,000 kg of nitrogen into the Kvichak River system
annually (Moore and Schindler 2004). The distribution and relative importance of the trophic subsidies
provided by MDN within salmon-bearing watersheds are not expected to be spatially or temporally
uniform (Janetski et al. 2009). The magnitude and density of spawning salmon and their by-products
(i.e., excreta and gametes) will be highest in areas of high spawning density and where carcasses
accumulate. In contrast, MDN influences on aquatic foodwebs may be negligible in headwater streams
above the upstream limit of anadromous fish distributions. In these systems, other sources of energy,
such as terrestrial inputs and benthic production, will be important (Wipfli and Baxter 2010).
Where salmon are abundant, productivity of the Bristol Bay region's fish and wildlife species is highly
dependent on this influx of MDN into the region's freshwater habitats (Box 5-3). When and where
available, salmon-derived resources—in the form of eggs, carcasses, and invertebrates that feed upon
carcasses—are important dietary components for many fishes (e.g., rainbow trout, Dolly Varden,
juvenile Pacific salmon, Arctic grayling). Eggs from spawning salmon are a major food source for Bristol
Bay rainbow trout and are likely responsible for much of the growth attained by these fish and the
abundance of trophy-sized rainbow trout in the Bristol Bay system. Upon arrival of spawning salmon in
the Wood River basin, rainbow trout shifted from consuming aquatic insects to primarily salmon eggs,
resulting in a five-fold increase in ration and energy intake (Scheuerell et al. 2007). With this rate of
intake, a bioenergetics model predicts a 100-g trout to gain 83 g in 76 days; without the salmon-derived
subsidy, the same fish was predicted to lose 5 g (Scheuerell et al. 2007). Rainbow trout in Lower Talarik
Creek were significantly fatter (i.e., had a higher condition factor) in years with high salmon spawner
abundance than in years with low abundance (Russell 1977). Research in Iliamna Lake suggests that
between 29 and 71% of the nitrogen in juvenile sockeye salmon, and even higher proportions in other
aquatic taxa, comes from marine-derived sources, and that the degree of MDN influence increases with
escapement (Kline et al. 1993).
Bristol Bay Assessment 5-29 January 2014
-------
Chapter 5 Endpoints
Terrestrial mammals (e.g., brown bears, wolves, foxes, minks), and birds (e.g., bald eagles, waterfowl)
also benefit from these subsidies (Box 5-3) (Brna and Verbrugge 2013; this document was originally
published as Appendix C of this assessment, but has since been released as a U.S. Fish and Wildlife
Service [USFWS] report). Availability and consumption of salmon-derived resources can have significant
benefits for these species, including increased growth rates, energy storage, litter size, nesting success,
and population density (Appendix A, Brna and Verbrugge 2013). Terrestrial systems of the Bristol Bay
watershed also benefit from these MDN (Cederholm et al. 1999, Gende et al. 2002) (Box 5-3). Bears,
wolves, and other wildlife transport carcasses and excrete wastes throughout their ranges (Darimont et
al. 2003, Helfield and Naiman 2006), which then provide food and nutrients for other terrestrial species.
Finally, by dying in the streams where they spawn, adult salmon subsidize the next generation by adding
their nutrients to the ecosystem that will feed their young. This positive feedback is missing from
freshwater systems with depleted salmon runs, which may inhibit attempts to renew those runs if
trophic resources are limiting those populations (Gresh et al. 2000). It is important to note that,
although there is ample evidence for the significant benefits provided by trophic subsidies associated
with spawning salmon in the Bristol Bay region, trophic limitations to fish population productivity
should not be assumed. For example, Schindler et al. (2005) showed that MDN are indeed important for
lake productivity in the Wood River system, but that interception of MDN inputs by the commercial
fishery did not appear to be a driver of sockeye salmon population dynamics—likely because spawning
habitat is a more limiting resource for this population.
5.2.6 Bristol Bay Fisheries in the Global Context
The Bristol Bay region is a unique environment supporting world-class fisheries, particularly in terms of
Pacific salmon populations. The region takes on even greater significance when one considers the status
and condition of Pacific salmon populations throughout their native geographic distributions. These
declines are discussed briefly below; for additional information on threatened and endangered salmon
stocks, see Appendix A (pages 37-41).
Although it is difficult to quantify the true number of extinct Pacific salmon populations around the
North Pacific, estimates for the western United States (California, Oregon, Washington, and Idaho) range
from 106 to 406 populations (Nehlsen et al. 1991, Augerot 2005, Gustafson et al. 2007). Pacific salmon
are no longer found in 40% of their historical breeding ranges in the western United States, and
populations tend to be significantly reduced or dominated by hatchery fish where they do remain (NRC
1996). For example, 214 salmon and steelhead stocks were identified as facing risk of extinction in the
western United States; 76 of those stocks were from the Columbia River basin alone (Nehlsen et al.
1991). In general, these losses have resulted from cumulative effects of habitat loss, water quality
degradation, climate change, overfishing, dams, and other factors (NRC 1996, Schindler et al. 2010).
Species with extended freshwater rearing periods—that is, species like sockeye, which dominates
salmon production in the Bristol Bay watershed—are more likely to be extinct, endangered, or
threatened than species which spend less time in freshwater habitats (NRC 1996). No Pacific salmon
Bristol Bay Assessment 5-30 January 2014
-------
Chapter 5 Endpoints
populations from Alaska are known to have gone extinct, although many show signs of population
declines.
The status of Pacific salmon throughout the United States highlights the value of the Bristol Bay
watershed as a salmon sanctuary or refuge (Rahr et al. 1998, Pinsky et al. 2009). The Bristol Bay
watershed contains intact, connected habitats that extend from headwaters to ocean with minimal
influence of human development. These characteristics, combined with the region's high Pacific salmon
abundance and life-history diversity, make the Bristol Bay watershed a significant resource of global
conservation value (Pinsky et al. 2009). Because the region's salmon resources have supported Alaska
Native cultures in the region for at least 4,000 years and continue to support one of the last intact wild
salmon-based cultures in the world (Appendix D), the watershed also has global cultural significance.
5.3 Endpoint 2: Wildlife
Unlike most terrestrial ecosystems, the Bristol Bay watershed has undergone little development and
remains largely intact. Thus, it still supports its historical complement of species, including large
carnivores such as brown bears (Ursus arctos), bald eagles [Haliaeetus leucocephalus), and gray wolves
(Cam's lupus); ungulates such as moose (Alces alces gigas) and caribou (Rangifer tarandusgranti); and
numerous waterfowl species. Wildlife populations tend to be relatively large in the region, due to the
increased productivity associated with Pacific salmon runs (Section 5.2.5). MDN provide a foundational
element for the foodwebs in these watersheds and are important for many species of wildlife. Wildlife,
in turn, distribute these nutrients from the aquatic to the terrestrial environment, cycling them through
the entire ecosystem (Box 5-3). Thus, interactions between salmon and wildlife species are complex and
reciprocal.
In this section we summarize key wildlife species in the Nushagak and Kvichak River watersheds, with
particular focus on how these species are related to salmon resources. The species selected for
characterization—brown bear, moose, barren-ground caribou, gray wolf, bald eagle, waterfowl (as a
guild), shorebirds (as a guild), and land birds (as a guild)—are important to ecosystem function, have a
direct link to salmon, and/or are important to Alaska Native and non-native residents. Within the
Nushagak and Kvichak River watersheds, there are no known breeding or otherwise significant
occurrences of any species listed as threatened or endangered under the Endangered Species Act, nor
any designated critical habitat. For additional information on wildlife species, readers should consult
Brna and Verbrugge (2013). In many cases, little abundance data specific to the Bristol Bay watershed
are available, but it is reasonable to assume that species distribution and abundance patterns in this
region mirror those observed in similar habitats across southwestern Alaska.
Although this assessment focuses on inland aquatic and nearshore habitats of the Bristol Bay watershed,
it should be noted that once the region's Pacific salmon populations migrate to the ocean, they also
provide food for marine predators (Appendix F). Marine mammals such as northern fur seals, harbor
seals, stellar sea lions, orcas and beluga whales are known to feed on Pacific salmon. These interactions
also can be important in freshwater habitats, as one of two freshwater harbor seal populations in North
Bristol Bay Assessment 5-31 January 2014
-------
Chapter 5 Endpoints
America is found in Iliamna Lake (Smith et al. 1996). Although this population is not evaluated in this
assessment, the National Oceanic and Atmospheric Administration is currently conducting a status
review on Iliamna Lake seals to determine if they represent a distinct population segment that may
warrant protection under the Endangered Species Act (Appendix F).
5.3.1 Life Histories, Distributions, and Abundances of Species
5.3.1.1 Brown Bears
Brown bears are wide-ranging and feed on many different plant and animal species. They typically
spend July through mid-September near streams supporting salmon runs, then move to higher
elevations in the fall to feed on berries and other food items before denning in October to November.
They emerge in spring and feed on vegetation and carrion, as well as moose and caribou calves. Because
of their wide-ranging behavior, they distribute MDN via both deposition of salmon carcasses and
excretion of wastes throughout their ranges.
Brown bear density estimates range from roughly 40 bears per 1,000 km2 in the northern Bristol Bay
region (Togiak National Wildlife Refuge and the Bureau of Land Management's Goodnews Block) (Walsh
et al. 2010) to 150 bears per 1,000 km2 along the shore of Lake Clark (Olson and Putera 2007). From
July 2006 to July 2007, 621 brown bears were reported harvested from the Alaska Department of Fish
and Game's (ADF&G's) Game Management Unit (GMU) 9, which includes the Kvichak River watershed
and the Alaska Peninsula. Brown bears are not as abundant in the Nushagak River watershed as the
Kvichak River watershed, and densities in both watersheds are lower than on the Alaska Peninsula's
Pacific coast, which is home to the highest documented brown bear density in North America (551 bears
per 1,000 km2) (Miller et al. 1997). Brown bears are reported as common in the area surrounding the
Pebble deposit, with a 2009 estimated density of 18.4 to 22.5 per 1,000 km2 (PLP 2011).
5.3.1.2 Moose
Moose habitat is determined by forage opportunities and includes both aquatic and upland areas.
Alluvial habitats along the Nushagak and Mulchatna Rivers, where willows and other plants regenerate
after scouring and subsequent deposit of river silt, support an abundant moose population. High-quality
summer forage, especially near wetlands, is important for nursing cows and calves. It is likely that MDN
contribute to increased plant productivity in these alluvial areas (Cederholm et al. 1999, Gende et al.
2002).
Moose abundance in the Nushagak and Kvichak River watersheds was estimated at 8,100 to 9,500 in
2004 (Butler 2004, Woolington 2004). Populations are especially high in the Nushagak River watershed
(ADF&G 2011), where felt-leaf willow, a preferred plant species, is abundant (Bartz and Naiman 2005).
Moose were considered "low density" (0.04 moose/km2) in the immediate area of the Pebble deposit
and the transportation corridor, but there is a large variance around this estimate (PLP 2011).
Bristol Bay Assessment 5-32 January 2014
-------
Chapter 5 Endpoints
5.3.1.3 Caribou
Caribou feed in open tundra, mountain, and sparsely forested areas and can travel for long distances.
The Nushagak and Kvichak River watersheds are primarily used by caribou from the Mulchatna herd,
one of 31 caribou herds found in Alaska. The Mulchatna herd ranges widely through the Nushagak and
Kvichak River watersheds, but also spends considerable time in other watersheds. It numbered roughly
200,000 in 1997 but had decreased to roughly 30,000 by 2008 (Valkenburg et al. 2003, Woolington
2009). Recent surveys reported only a few caribou near the Pebble deposit area and potential
transportation corridor (PLP 2011). However, caribou populations and ranges in the Bristol Bay region
fluctuate significantly over time, and in previous years the herd was much larger and there was higher-
density use of the Pebble deposit area (PLP 2011). Barren-ground caribou on the North Slope of Alaska
have demonstrated avoidance of exploration activities (Fancy 1983), and some tribal Elders in the
Nushagak and Kvichak River watersheds believe that mining exploration has contributed to avoidance
of the Pebble deposit area (Brna and Verbrugge 2013).
5.3.1.4 Gray Wolf
Gray wolf abundance is influenced by prey abundance and availability, but populations are primarily
limited by mortality caused by humans. Wolves have flexible diets and can shift to non-ungulate prey
species when ungulate prey are scarce, or take advantage of seasonally abundant species such as
salmon. Wolves often transport salmon away from streams for consumption or to feed pups through
regurgitation.
Gray wolf populations have not been well-studied in the Bristol Bay region, and it is difficult to assess
population numbers. Wolves are currently thought to be abundant in the Nushagak River watershed:
between 2003 and 2008, reported annual wolf harvest ranged from 60 to 141 in GMU 17, which includes
the Nushagak and Togiak River watersheds. In the Kvichak River watershed, numbers are believed to be
lower, although populations have increased since the 1990s (Butler 2009).
5.3.1.5 Bald Eagle
Bald eagles generally nest near riparian and beach areas and are primarily piscivorous, although they
have a variable diet. Nesting bald eagles rely on salmon resources (Hansen 1987), and inland bald eagles
nesting near spawning streams have higher nesting success than those with more distant nests (Gerrard
et al. 1975). Birds and non-salmon fishes are also important prey for bald eagles. Salmon abundance in
the Nushagak and Kvichak River watersheds affects bald eagle abundance, distribution, breeding, and
behavior. Bald eagles, in turn, distribute MDN in their excretions.
Although no comprehensive survey of bald eagles or bald eagle nests has been conducted in the Bristol
Bay watershed, limited count data are available for parts of the region. For example, 50 bald eagle nests
were recorded along portions of the Nushagak, Mulchatna, and Kvichak Rivers in 2006 (Brna and
Verbrugge 2013); approximately half of those nests were categorized as active. The USFWS Bald Eagle
Nest Database contains approximately 230 nest records for the Nushagak and Kvichak River
watersheds, with 169 of those records collected between 2003 and 2006. Raptor studies in the Pebble
Bristol Bay Assessment 5-33 January 2014
-------
Chapter 5 Endpoints
deposit area indicate that bald eagles were the most abundant nesting raptor (30% of all raptor nests in
2005) (PLP 2011).
5.3.1.6 Waterfowl
More than 30 species of waterfowl, including ducks (e.g., northern pintail, scaup, mallard, and green-
winged teal), geese (e.g., white-fronted, Canada), swans, and sandhill cranes, regularly use the Bristol
Bay region (PLP 2011). Diversity of habitat and extent of wetlands and waters provide habitat for
migrants and wintering waterfowl, and the region is an important staging area for many species,
including emperor geese, Pacific brant, and ducks, during spring and fall migrations.
The Alaska Yukon Waterfowl Breeding Population Survey found average late-May abundance indices of
497,000 ducks, 7,700 geese, 15,400 swans, and 5,300 sandhill cranes in the Bristol Bay Lowlands
between 2002 and 2011 (Brna and Verbrugge 2013). Salmon are used by some waterfowl as direct
sources of prey and carrion, and used indirectly through invertebrates and vegetation. Of the 24 duck
species in the Bristol Bay region, at least 11 prey on salmon eggs, parr, or smolts, or scavenge on salmon
carcasses (Brna and Verbrugge 2013).
5.3.1.7 Shorebirds
Thirty of 41 shorebird species or subspecies that regularly occur in Alaska can be found in the Bristol
Bay watershed (see Brna and Verbrugge [2013] for a summary of different shorebird surveys).
Shorebirds use the Bristol Bay watershed primarily during migration and breeding. Significant areas of
intertidal habitat exist at Kvichak Bay (530 km2) and Nushagak Bay (400 km2). Important foods include
abundant intertidal invertebrates and fruits and tubers in upland areas. Shorebirds likely play an
important role in the distribution of MDN to terrestrial ecosystems. Adults, young, and eggs also provide
a source of food for predatory birds and terrestrial mammals. Although there is not a strong direct link
between salmon and shorebirds, it is reasonable to assume that MDN contribute to the abundance of
invertebrates in the intertidal zone.
The Bristol Bay/Alaska Peninsula lagoon system, which includes the Nushagak and Kvichak River deltas,
is one of the most important migratory shorebird stop-over areas in Alaska. Surveys of the Pebble
deposit area in 2004 to 2005 identified 14 shorebird species in the Pebble deposit area (PLP 2011).
5.3.1.8 Land Birds
Approximately 80 species of land birds, both migratory and year-round residents, breed in and adjacent
to the Nushagak and Kvichak River watersheds. Land birds eat vegetation (e.g., seeds, berries),
invertebrates, and vertebrates. Studies indicate that the abundance of many songbird species is related
to the presence of salmon carcasses (Willson et al. 1998, Gende and Willson 2001, Christie and
Reimchen 2008). Salmon carcasses provide food for aquatic invertebrate larvae, and MDN contribute to
increased plant productivity (Cederholm et al. 1999, Gende et al. 2002), both important food sources for
land birds. Few abundance studies have focused on the Nushagak and Kvichak River watersheds, but
2004 to 2005 surveys identified 28 land bird species in the Pebble deposit area (PLP 2011).
Bristol Bay Assessment 5-34 January 2014
-------
Chapter 5 Endpoints
5.3.2 Recreational and Subsistence Activities
Many of the species discussed in the preceding sections are important subsistence resources. For
example, a 2002 survey of Bristol Bay residents found that 86% and 88% of respondents have
consumed moose and caribou meat, respectively (Ballew et al. 2004). Between 1983 and 2006, moose
harvest in GMU 17 increased from 127 to 380 moose per year; the upper Nushagak River watershed
alone (GMU 17B) had a mean annual harvest of 149 moose (Brna and Verbrugge 2013). Caribou harvest
ranged from 1,573 to 4,770 per year between 1991 and 1999, but this estimate is for the entire
Mulchatna herd, including those taken outside of the Nushagak and Kvichak River watersheds
(Valkenburg et al. 2003).
Waterfowl support recreational and subsistence harvests, as well as wildlife viewing opportunities.
There are no reliable estimates of recreational harvests specific to the Nushagak and Kvichak River
watersheds. Subsistence harvest of waterfowl is very important in the watershed. The spring harvest
provides fresh meat early in the season, after winter food supplies are depleted. Harvest data from 1995
through 2005 for the Dillingham, Nushagak River, and Iliamna subregions (Wentworth 2007, Wong and
Wentworth 1999) indicate annual harvests of roughly 10,000 ducks, 2,500 to 2,900 geese, and up to
300 tundra swans, as well as fewer than 500 waterfowl eggs (Brna and Verbrugge 2013).
Sport hunting for caribou, moose, brown bear, and other species also plays a role in the local economy of
the Bristol Bay region. In recent years, approximately 1,323 non-residents and 1,319 non-local residents
of Alaska traveled to the region to hunt. Miller and McCollum (1994) estimate that non-residents and
non-local residents spend approximately $5,170 and $1,319 per trip (values updated to 2009 dollars),
respectively. These hunting activities result in an estimated $8.2 million per year in direct hunting-
related expenditures (Table 5-4) and directly employ over 100 full- and part-time workers.
5.4 Endpoint 3: Alaska Natives
Alaska Natives are the majority population in the Bristol Bay region, and salmon has been central to
their health, welfare, and culture for thousands of years. In fact, Alaska Native cultures in the region
represent one of the last intact salmon-based cultures in the world (Appendix D). Much of the region's
population practices subsistence, with salmon making up a large proportion of subsistence diets—
making Alaska Natives particularly vulnerable to potential changes in salmon resources.
The effect on Alaska Natives resulting from potential mining-related changes in salmon and other fishes
was selected as an assessment endpoint because of the nutritional and cultural importance of salmon to
Alaska Natives, and because of the U.S. Environmental Protection Agency's (USEPA's) responsibilities to
work with federally recognized tribes on a government-to-government basis to protect, restore, and
preserve the environment. These responsibilities are set forth in Executive Order 13175, Executive
Order 12898, President Obama's 2009 Indian Policy, former USEPA Administrator Jackson's
Reaffirmation of USEPA's Indian Policy 2009, USEPA's Policy on Tribal Consultation and Coordination,
and USEPA's Region 10 Tribal Consultation and Coordination Procedures. Nine Bristol Bay federally
Bristol Bay Assessment 5-35 January 2014
-------
Chapter 5 Endpoints
recognized tribes and other tribal organizations petitioned the USEPA in 2010, requesting that the
agency use its authority under the Clean Water Act Section 404(c) to restrict or prohibit the disposal of
dredged or fill material associated with large-scale mining activities in the Bristol Bay watershed.
5.4.1 Alaska Native Populations
There are 31 Alaska Native villages in the wider Bristol Bay region, 25 of which are located in the Bristol
Bay watershed. Fourteen of these communities are within the Nushagak and Kvichak River watersheds,
with a total population of 4,337 in 2010 (U.S. Census Bureau 2010). Dillingham (population 2,329) is the
largest community; other communities range in size from two (year-round) residents (Portage Creek) to
510 residents (New Stuyahok). Because population in some communities is seasonal, these numbers
increase during the subsistence fishing season. Thirteen of these 14 villages—all but Port Alsworth—
have federally recognized tribal governments and had an Alaska Native population majority in 2010.
Overall population in the region grew 55% from 1980 to 2000, and remained relatively stable from
2000 to 2010 (U.S. Census Bureau 2010). Population has fluctuated in individual villages since 1980
(Appendix D, Table 2). From 2000 to 2010, nine villages decreased and five villages increased in
population. The extent to which these changes reflect natural population fluctuations or whether any
gains or losses indicate a long-term trend is unknown. Four of the villages that decreased in population
(Dillingham, Igiugig, Aleknagik, and Kokhanok) and one of villages that increased in population
(Iliamna) changed less than 10%. Port Alsworth has experienced steady population growth since 1980.
Its economy is more closely tied to Lake Clark National Park, and its population contains the smallest
proportion of Alaska Natives among the 14 villages. Portage Creek is the smallest village in the region,
and its year-round population has fluctuated significantly over the past 40 years (e.g., 48 in 1980, 5 in
1990, 36 in 2000, 2 in 2010), making it difficult to draw conclusions about trends.
5.4.2 Subsistence and Alaska Native Cultures
5.4.2.1 Importance of Salmon to Alaska Native Cultures
The primary Alaska Native cultures present in the Nushagak and Kvichak River watersheds—the Yup'ik
and Dena'ina (Box 5-1)—are part of the last intact, sustainable salmon-based cultures in the United
States (Appendix D). This is especially significant as other Pacific Northwest salmon-based cultures
struggle with degraded resources (Colombi and Brooks 2012). Cultures associated with salmon fishing
appeared in these watersheds as early as 4,000 before present (BP) and intensified around 1,000 BP
(Appendix D). Currently, the percentage of Alaska Native population in the region's villages ranges from
21.4% (Port Alsworth) to 95.7% (Koliganek) (Appendix D, Table 2). The Yup'ik and Dena'ina cultures
still provide the framework and values for everyday life in the region. Among the Yup'ik, over 40% of the
population continues to maintain their native language, one of the highest percentages among native
cultures in the United States (Appendix D).
In the Bristol Bay region, the subsistence way of life is irreplaceable. Subsistence resources provide high
quality foods, foster a healthy lifestyle, and form the basis for social relations for both Alaska Natives
Bristol Bay Assessment 5-36 January 2014
-------
Chapter 5 Endpoints
and non-Alaska Natives in the villages. These resources, particularly salmon, are integral to the entire
way of life in Yup'ik and Dena'ina cultures. The Alaska Federation of Natives (2010) describes
subsistence as follows.
The hunting, fishing, and gathering activities which traditionally constituted the economic base of
life for Alaska's Native peoples and which continue to flourish in many areas of the state
today...Subsistence is a way of life in rural Alaska that is vital to the preservation of communities,
tribal cultures, and economies. Subsistence resources have great nutritional, economical, cultural,
and spiritual importance in the lives of rural Alaskans...Subsistence, being integral to our
worldviewand among the strongest remaining ties to our ancient cultures, are as much spiritual
and cultural as it is physical.
For Alaska Natives today, subsistence is more than the harvesting, processing, sharing, and trading of
land and sea mammals, fish, and plants. Subsistence holistically subsumes the cultural, social, and
spiritual values that are the essence of Alaska Native cultures. There is a strong tradition and practice of
sharing and trading subsistence resources. Food is shared with tribal Elders, family living outside of the
watershed, and others who may not be able to fully participate in subsistence (Appendix D). This
practice was confirmed by tribal Elders interviewed for Appendix D and those who testified at public
meetings on the May 2012 draft of the assessment (Box 5-5).
Cultural and personal identity largely revolve around traditional cultural practices such as hunting,
fishing, and gathering of wild food resources—that is, subsistence. Tribal Elders and culture bearers
continue to instruct young people, particularly at fish camps where cultural values as well as fishing and
fish processing techniques are shared. The social system that forms the backbone of the culture, by
nurturing the young, supporting the producers, and caring for the tribal Elders, is based on the virtue of
sharing wild foods harvested from the land and waters. Sharing networks extend to family members
living far from home. The first salmon catch of the year is recognized with a prayer of thanks and shared
in a continuation of the ancient First Salmon Ceremony (Appendix D), when those who have caught the
first Chinook (king) salmon in the spring share them with tribal Elders and all those in need, as well as
friends and family.
Traditional and more modern spiritual practices place salmon in a position of respect and importance,
as exemplified by the First Salmon Ceremony and the Great Blessing of the Waters (Appendix D). The
salmon harvest provides a basis for many important cultural and social practices and values, including
the sharing of resources, fish camp, gender and age roles, and the perception of wealth. Although a small
minority of tribal Elders and culture bearers interviewed expressed a desire to increase market
economy opportunities (including large-scale mining), most equated wealth with stored and shared
subsistence foods (Appendix D). In interviews conducted for Appendix D, the Yup'ik and Dena'ina
communities of the Nushagak and Kvichak River watersheds consistently define a "wealthy person" as
one with food in the freezer, a large extended family, and the freedom to pursue a subsistence way of life
in the manner of their ancestors. Their ability to continue their reliance on subsistence and their concept
of wealth have contributed to the maintenance of vital and viable cultures for at least 4,000 years.
Bristol Bay Assessment 5-37 January 2014
-------
Chapter 5 Endpoints
BOX 5-5. TESTIMONY ON THE IMPORTANCE OF
SUBSISTENCE USE
The USEPA held a series of public meetings to collect input on the May 2012 draft of the assessment. Many
Alaska Natives, including tribal Elders and other tribal leaders, provided testimony on the importance of
salmon and the subsistence way of life to Alaska Native cultures in the region. The following are selected
quotes representative of this testimony; complete public meeting transcripts are available at
www.epa.gov/BristolBay.
• "Our subsistence way of life plays a substantial role in our health both spiritually and physically."
• "From traditional knowledge we keep our culture going. My subsistence life is with my family, which
consists of four boys and my wife. I also help my grandmother, grandfather, mother, father, and our other
family members. I hold a Bristol Bay drift permit, my family fishes with me both commercially and
subsistence. My family processes approximately 4,000 pounds of salmon, kings, reds, silvers, etc. We
start when the fish first come into the river, all the way to the very end. My family and I smoke, dry, and
freeze the salmon. I brought you some canned salmon to share that we keep year round."
• "The king salmon is a very important part of our fishery. If you cover that portion of the king [Chinook]
salmon spawning beds, it is going to make it very hard for us to maintain our culture of people who eat
king every year. Is the first fish of the year, it's a very important fish for us and we can't have that huge
loss."
• "Fishing is our life and our livelihood. It's what we do for healthy communities, healthy lifestyles. Going
out and catching the subsistence fish, smoking these. Passing the traditional knowledge on to younger
generations. You hear about how they will make you free, the fish. We have been doing this for 6,000
years and we will want to do it for 6,000 more."
• "The generations that are coming who can be fed from this resource and this land and it's a beautiful
interaction and it's one that we are losing around the world. When we realize that we have lost it we strive
to get it back, but it is taking a longtime for this beautiful balance between human, animal and
subsistence lifestyle to come about and evolve."
• "The survival of our culture directly depends on the health of our land, the fish and the wildlife. No
amount of money or jobs can replace our way of life and our culture."
• "I am a Dena'ina, and Athabascan Indian. This village is my home. We are very rich people in our culture,
our resources, plants, animals and salmon. They all need clean water. That includes us, the Dena'ina
people of the land. But only because we are so blessed to have clean water. Salmon have been a great
part of our diet for generations and will be in the future."
• "Right now we are getting excited for the kings to come up our river. For everyone works together cutting
fish. To dry, salt or vacuum pack for the winter. We do not waste anything, because we fish. Around here
it is gold, gold to us which we treasure. When we fill our dry rack, we go walking and help one another."
• "I've lived here for 30 years and I moved here by choice. My experience of living in this area is that people
choose to be here whether born or coming here. It's a choice. It is not a scientific fact, but three reasons
people choose to be in Bristol Bay is because clean water, the fishery and the lifestyle."
• "This environment has sustained our culture for thousands of years. It sustained jobs and commercial
fishing for hundreds of years, and recreation and sport fishing and everything."
The Alaska Native community is also dependent on the regional economy, which is primarily driven by
commercial salmon fishing and tourism. The commercial fishing and recreation market economies
provide seasonal employment for many residents, giving them both the income to purchase goods and
services needed for subsistence and the time to participate year-round in subsistence activities. The
fishing industry provides half of all jobs in the region, followed by government (32%), recreation (15%),
and mineral exploration (3%) (Appendix E). It is estimated that local Bristol Bay residents held one-
Bristol Bay Assessment 5-38 January 2014
-------
Chapter 5 Endpoints
third of all 2009 jobs and earned almost $78 million (28%) of the total income traceable to the Bristol
Bay salmon ecosystem (Appendix E).
5.4.2.2 Use of Subsistence Resources in the Bristol Bay Watershed
Alaska Native populations, as well as non-Alaska Native residents, have continual access to a range of
subsistence foods. As described by Fall et al. (2009), these subsistence resources are the most consistent
and reliable component of the local economies in the Bristol Bay watershed, even given the world-
renowned commercial fisheries and other recreational opportunities the region supports (Table 5-4).
Subsistence uses on state lands are given priority by state law and regulations (i.e., the 1978 State of
Alaska Subsistence Act). All citizens of Alaska benefit from a subsistence priority in areas specifically
designated as subsistence areas by the State of Alaska. State hunting and fishing regulations apply to
lands of the Alaska Native Corporations. These lands were often selected because of their significant
value for subsistence activities, and Alaska Native peoples have the exclusive right to occupy and use
these lands for subsistence. These rights are not recognized in the State of Alaska Constitution; however
the Alaska Federation of Natives has passed resolutions for several years asking for the constitution to
be revised. In addition, the Alaska Federation of Natives recommended improvements to management of
state and federal subsistence programs. Indigenous hunting and fishing rights are recognized by statute
only and therefore can be diminished over time. Their lack of special status makes these rights
vulnerable to constitutional challenges, especially challenges based on the right to equality (Duhaime
and Bernard 2008).
Virtually every household in the Nushagak and Kvichak River watersheds uses subsistence resources
(Appendix D: Table 12). No watershed data are available for the proportion of Bristol Bay watershed
residents' diets made up of subsistence foods, as most studies focus on harvest data and are not dietary
surveys. A study that included the nearby Yukon-Kuskokwim region found that 22.8 % of calories came
from Native (subsistence) foods (Johnson et al. 2009). In 2004 and 2005, annual subsistence
consumption rates in the Nushagak and Kvichak River watersheds were over 300 pounds per person in
many villages, and reached as high as 900 pounds per person (Appendix D, Table 12; for comparison, an
average American consumes 1,996 pounds of food per year). Villages with the highest per capita
subsistence usage were Koliganek, Ekwok, Newhalen, Kokhanok, Igiugig, and Levelock.
Subsistence use varies throughout the Bristol Bay watershed, as villages differ in the per capita amount
of subsistence harvest and the variety of subsistence resources used. Salmon and other fishes provide
the largest portion of subsistence harvests of Bristol Bay communities. On average, about 50% of the
subsistence harvest by local community residents (measured in pounds usable weight) is Pacific salmon,
and about 10% is other fishes (Fall et al. 2009). The percentage of salmon harvest in relation to all
subsistence resources ranges from 29 to 82% in the villages (Appendix D, Table 11). Salmon accounts
for an especially high percentage compared to all subsistence resources for Iliamna, Kokhanok, and
Pedro Bay. Igiugig, Levelock, and New Stuyahok show the lowest percentage of salmon usage relative to
other subsistence resources. Villages in the Nushagak River watershed, especially New Stuyahok,
Ekwok, and Dillingham, rely on Chinook salmon to a great extent, whereas villages in the Kvichak River
Bristol Bay Assessment 5-39 January 2014
-------
Chapter 5 Endpoints
watershed and Iliamna Lake area (e.g., Iliamna, Kokhanok, Iguigig, Newhalen, Nondalton, Pedro Bay, and
Port Alsworth) rely more on sockeye salmon. All communities also rely on non-salmon fishes (Table 5-
1), but to a lesser extent than salmon. These fishes are taken throughout the year by a variety of harvest
methods and fill an important seasonal component of subsistence cycles (Fall et al. 2009). For example,
whitefish and other freshwater species provide fresh fish during winter ice-fishing season (Appendix D).
The ADF&G overview of subsistence fisheries in the Bristol Bay watershed (Fall et al. 2009) provides the
following information.
• The number of Bristol Bay subsistence salmon permits issued has been stable since 1990, and the
recent 10-year average is 1,146 permits. Most permit holders (84%) are residents of Bristol Bay
communities, and most permits are issued for the Nushagak and Naknek/Kvichak districts. Sockeye
salmon make up the largest portion of the Bristol Bay subsistence salmon harvest (79% of the
1998-2007 average, based on subsistence salmon permits), followed by Chinook (19%), coho (5%),
chum (5%), and pink (2%).
• Annual subsistence harvests for the Bristol Bay management area vary from year to year. Salmon
harvest declined from the early 1990s to the early 2000s but has recovered slightly since 2002.
Since 1975, the average annual harvest was about 152,371 salmon; the recent 5-year average
(2003-2007) was 126,717 salmon.
• The largest decline over the last 15 years has occurred in the Kvichak River watershed subsistence
sockeye salmon fishery, historically the largest component of the Bristol Bay subsistence salmon
harvest. Declines are due to lower harvests per permit, rather than reduced fishing effort. Since
1996, harvest per day is down 26% in years of escapements under 2 million fish, compared to the
previous 13-year average. The long-term average (45 years, for which permit data are available) for
this fishery is 66,614 sockeye salmon.
• There has been an overall harvest decline in the Nushagak district from a high of 86,400 fish in 1986
to a low of 40,373 salmon in 2006. The 24-year average harvest (the time for which data are
available) is 50,740 fish. However, the number of subsistence salmon permits issued in the
Nushagak district has remained relatively stable since 1983.
• Subsistence salmon harvests in the Nushagak district are similar to those in the Kvichak district in
terms of harvest levels. For example, in 2007the communities in the Nushagak district harvested
44,944 salmon, compared to 47,538 salmon in the Kvichak River/Iliamna Lake subdistrict, based on
permit returns. However, there are differences in the two fisheries. Whereas salmon harvest in the
Kvichak River watershed is almost all sockeye salmon (47,473 out of 47,538 in 2007), salmon
harvest in the Nushagak district is more varied, with larger harvests of Chinook, coho, and chum
salmon. There are also larger communities in the Nushagak district, including Dillingham,
Manokotak, Aleknagik, New Stuyahok, and Koliganek.
• Chinook salmon returns are higher in the Nushagak River watershed than in the Kvichak River
watershed. In the upper portion of the Nushagak River, residents attempt to harvest large numbers
Bristol Bay Assessment 5-40 January 2014
-------
Chapter 5 Endpoints
of Chinook salmon, their traditionally preferred salmon resource. Chinook salmon spawn early in
the season, and it is important to put up these fish for subsistence before commercial fishing starts
in earnest (Holen et al. 2012). Substitution of Chinook for sockeye salmon accounts for some, but not
all, of the decline in the Nushagak district. Subsistence sockeye salmon harvests in the Kvichak River
watershed, including Iliamna Lake and Lake Clark (historically the largest component of the Bristol
Bay subsistence salmon fishery), declined by more than 50% during the 1990s and early 2000s.
Local subsistence fishers attributed these lowered harvests to poor returns and scarcities of salmon
in once reliable and abundant traditional harvest locations. Effort has increased in harvesting
salmon in these areas since the low harvest levels seen in early 2000.
Figures 5-2 and 5-12 show areas of subsistence use identified by ADF&G in the Nushagak and Kvichak
River watersheds. Clark's Point subsistence use areas overlap with Nushagak and Kvichak River
watersheds for caribou, coho salmon, and moose. Clark's Point high per capita harvest rate (1,210 Ibs
per capita) resulted from a high harvest rate of salmon in 2008. This was three times higher than the
harvest levels reported in 1973 and 1989 (Holen et al. 2012). Manokotak subsistence use areas overlap
with the Nushagak communities for caribou and moose. Aleknagik moose search areas include part of
Nushagak River area (Holen et al. 2012). South Naknek, Naknek, and King Salmon subsistence use areas
for waterfowl, rainbow trout, unspecified trout, moose, and berry picking, as well as caribou search
areas, overlap the Nushagak and particularly the Kvichak River watersheds (Holen et al. 2011). It should
be noted that available subsistence data are coarse and incomplete (Box 5-2), and it is likely that
subsistence activities occur outside of the areas identified on the figures. Data used to generate the
figures were collected in different years, and at least one village with high recorded subsistence harvests
(Ekwok) declined to be surveyed. Also note that these figures do not indicate abundance or harvest, only
use.
Although subsistence is a non-market economic activity that is not officially measured, the effort put
into subsistence activities is estimated to be the same or greater than full-time equivalent jobs in the
cash sector (Appendix E). There is a strong and complex relationship between subsistence and the
market economy (largely commercial fishing and recreation) in the area (Wolfe and Walker 1987, Krieg
et al. 2007). Market economy income funds goods and services purchased by households and used for
subsistence activities (e.g., boats, rifles, nets, snow mobiles, and fuel). In addition to the economic
activity generated by the purchase of subsistence goods, subsistence harvests are valued at
approximately $60 to $86 per pound, or 34 to 42% of the 2009 per capita income of regional residents
(Appendix E).
Bristol Bay Assessment 5-41 January 2014
-------
Chapter 5
Endpoints
Figure 5-12. Subsistence use intensity for salmon, other fishes, wildlife, and waterfowl within the
Nushagak and Kvichak River watersheds. See Box 5-2 for more detailed discussion of methodology.
Cook Inlet
South NakneK
Bristol Bay
IN
A
25
25
50
] Kilometers
50
] Miles
Approximate Pebble Deposit Location
Surveyed Towns and Villages
Nonsurveyed Towns and Villages
Existing Roads
Watershed Boundary
Subsistence Use Intensity
High
Low
Bristol Bay Assessment
5-42
January 2014
-------
Chapter 5 Endpoints
The salmon-dependent diet of the Yup'ik and Dena'ina benefits their physical and mental well-being in
multiple ways, in addition to encouraging high levels of fitness based on subsistence activities. The
interviews conducted for Appendix D confirm ADF&G harvest data that people of the Nushagak and
Kvichak River watersheds primarily eat two species of Pacific wild salmon, sockeye and Chinook. These
are consumed in different ways, including fresh, salted, pickled, canned, dried, and smoked. Salmon and
other traditional wild foods comprise a large part of the people's daily diet throughout their lives,
beginning as soon as they are old enough to eat solid food. (Appendix D). Subsistence foods consumed in
rural Alaska have demonstrated multiple nutritional benefits, including lower cumulative risk of
nutritionally mediated health problems such as diabetes, obesity, high blood pressure, and heart disease
(Murphy et al. 1995, Dewailly et al. 2001, Dewailly et al. 2002, Din et al. 2004, Alaska Department of
Health and Social Services 2005, Chan et al. 2006, Ebbesson and Tejero 2007) and provision of essential
micronutrients and omega-3 fatty acids (Murphy et al. 1995, Nobmann et al. 2005, Bersamin et al. 2007,
Ebbesson and Tejero 2007).
A disproportionately high amount of total diet protein and some nutrients comes from subsistence
foods. For example, a 2009 study of two rural regions found that 46% of protein, 83% of vitamin D, 37%
of iron, 35% of zinc, 34% of polyunsaturated fat, 90% of eicosapentaenoic acid, and 93% of
docosahexaenoic acid came from subsistence foods consumed by Alaska Natives (Johnson et al. 2009).
In summary, the roles of salmon as a subsistence food source and as the basis for Alaska Native cultures
are inseparable. The characteristics of these subsistence-based salmon cultures have been widely
documented (Appendix D). The cultures have a strong connection to the landscape and its resources,
and in the Bristol Bay watershed this connection has been maintained for centuries by the uniquely
pristine condition of the region's landscape and resources. In turn, the respect and importance given
salmon and other wildlife, along with Alaska Natives' traditional knowledge of the environment, has
produced a sustainable, subsistence-based economy (Appendix D). This subsistence-based way of life is
a key element of Alaska Native identity and serves a wide range of economic, social, and cultural
functions in Yup'ik and Dena'ina societies (Appendix D). Appendix D states the following:
... Salmon and clean water are foundational to the Yup'ik and Dena'ina cultures in the Nushagak
and Kvichak watersheds. The people in this region not only rely on salmon for a large proportion of
their highly nutritional food resources; salmon is also integral to the language, spirituality, and
social relationships of the culture. Because of this interconnection, the cultural viability, as well as
the health and welfare of the local population, are extremely vulnerable to a loss of either quality
or quantity of salmon resources.
It should be noted that, even though the scope of the assessment is focused on villages in the Nushagak
and Kvichak River watersheds, subsistence harvest areas do not necessarily correspond with watershed
boundaries. As noted previously, villages outside of these watersheds use areas within the watersheds
for subsistence activities.
Bristol Bay Assessment 5-43 January 2014
-------
6.1 Basic Elements of the Mine Scenarios
For this assessment, we used information on porphyry copper deposits and mining practices
summarized in Chapter 4 to develop three mine size scenarios: Pebble 0.25 with 0.25 billion ton
(0.23 billion metric tons) of ore, Pebble 2.0 with 2.0 billion tons (1.8 billion metric tons) of ore, and
Pebble 6.5 with 6.5 billion tons (5.9 billion metric tons) of ore. The word Pebble in the names of the
scenarios represents the fact that we place our scenarios at the Pebble deposit. These three mine size
scenarios, as well as other scenario types considered in later chapters of the assessment, are
summarized in Table 6-1.
The three mine size scenarios evaluated in the assessment represent realistic, plausible descriptions of
potential mine development phases, consistent with current engineering practice and precedent. The
scenarios are not mine plans: they are not based on a specific mine permit application and are not
intended to be the detailed plans by which the components of a mine would be designed. However, the
scenarios are based on preliminary mine details put forth in Northern Dynasty Minerals' Preliminary
Assessment of the Pebble Mine (Ghaffari et al. 2011), as well as information from scientific and industry
literature for mines around the world (see Chapter 4 and Appendix H for background information on
mining and the geology of porphyry copper deposits). Thus, the mine scenarios reflect the general
activities and processes typically associated with the kind of large-scale porphyry copper mine
development likely to be proposed once a specific mine application is developed. We use these scenarios
to benchmark potential risks resulting from this type of development, to provide decision makers with a
better understanding of potential risks associated with any specific action proposed in the future.
Bristol Bay Assessment
6-1
January 2014
-------
Chapter 6
Mine Scenarios
Table 6-1. Summary of scenarios considered in the assessment.
Scenario Type
Mine size
Water collection,
treatment, and
discharge
Tailings dam failure
Scenario
Pebble 0.25
Pebble 2.0
Pebble 6. 5
Routine operations3
Wastewater treatment
plant failure3
TSF spillway release
Pebble 0.25
Pebble 2.0
Transportation corridor
Pipeline failure
Product concentrate
pipeline failure15
Return water pipeline
failure15
Diesel pipeline failure15
Description
Mine size of 0.25 billion ton (0.23 billion metric ton) of
ore.
Mine size of 2.0 billion tons (1.8 billion metric tons) of
ore.
Mine size of 6.5 billion tons (5.9 billion metric tons) of
ore.
All water collection and treatment at site works
properly, and wastewater is treated to meet state and
national standards before release; however, some
leachate from waste rock and TSFs is not captured.
Wastewater treatment plant fails and releases
untreated wastewater through its two outfalls.
Excess water stored in TSF 1 is released over the
spillway.
Failure of 92-m dam at TSF 1.
Failure of 209-m dam at TSF 1.
113-km gravel road with four pipelines, within the
Kvichak River watershed.
Complete break or equivalent failure of the product
concentrate pipeline.
Complete break or equivalent failure of the return
water pipeline.
Complete break or equivalent failure of the diesel
pipeline.
Assessment
Chapters)
7, 8
8
9
10
11
Notes:
8 Scenario was considered for each mine size scenario.
b Each pipeline failure scenario was considered at two locations: Chinkelyes Creek and Knutson Creek.
TSF = tailings storage facility.
In the scenarios, we make decisions concerning mine placement; the size of the mine and the time over
which mining would occur; the size, placement, and chemistry of waste rock; the size, placement, and
chemistry of tailings storage facilities (TSFs); on-site processing of the ore; and the removal of processed
ore concentrate from the site. For comparison purposes, Table 4-1 provides similar information for
other past, existing, and potential large mines in Alaska. The mine components described in the
scenarios are placed on the landscape based on information either from Ghaffari et al. (2011) or where,
in our experience, modern mining practice suggests a component would be placed. For example, the pit
is located on the deposit; TSFs are placed in locations described by Ghaffari et al. (2011) and where
topography provides an efficient location to store a large volume of tailings; waste rock is placed around
the pit to minimize the cost of hauling millions to billions of metric tons of material (Table 6-2); and the
transportation system is located within the corridor described by Ghaffari et al. (2011).
We focus on the major mine components (mine pit, waste rock piles, and TSFs) that have the potential to
adversely affect aquatic resources regulated under the federal Clean Water Act (33 USC 1251-1387)
(Box 6-1). Smaller mine facilities such as crushing and screening areas, the mill, laydown areas,
Bristol Bay Assessment
6-2
January 2014
-------
Chapters Mine Scenarios
workshops, offices, and housing would be expected to be placed in uplands to avoid wetlands, ponds,
and streams; thus, they are only addressed as they relate to stormwater runoff.
BOX 6-1. CUMULATIVE IMPACTS OF A LARGE-SCALE PORPHYRY COPPER MINE
In this assessment, we focus on the areas of the major mine components (mine pit, tailings storage
facilities, and waste rock piles) and the transportation corridor. The actual infrastructure needed to operate
any large-scale mine would be significantly more extensive than these four components and would result in
larger cumulative impacts of a single mine. These additional infrastructure needs (based on Ghaffari et al.
2011) would include, but are not limited to, the following.
• Mining and processing facilities, including grinding mills, ore stockpiles, conveyers, a wastewater
treatment plant, and process water ponds and distribution lines.
• Drainage management structures, such as seepage cutoff walls, stream diversion channels, drainage
ditches, and sediment control ponds.
• Other storage and disposal facilities, such as overburden and topsoil stockpiles, explosives storage, a
non-hazardous waste landfill, process water storage tanks, waste incinerators, a fuel storage compound,
and hazardous waste storage.
• Other operational infrastructure, such as administrative buildings, dormitories, a sewage treatment
plant, a power generation plant, power distribution lines, potable water treatment plant and distribution
lines, and a truck shop.
These cumulative plant and ancillary areas are included in the total mine footprint for each scenario (Tables
6-5 through 6-7) but are not specifically placed on the landscape because of the greater uncertainty
regardingtheir placement.
The cumulative impacts of a large-scale mine at the Pebble deposit likely would be much larger than the
footprints evaluated in the mine scenarios.
• According to Ghaffari etal. (2011), the total area of direct impact for a 25-year mine at the Pebble
deposit would cover approximately 125 km2; in comparison, the mine footprint for the 25-year mine
scenario (Pebble 2.0) considered in this assessment covers approximately 45 km2 (Table 6-6).
• Net power generation for such a mine would be approximately 378 megawatts (Ghaffari et al. 2011). This
is more than 100 times the maximum electrical load of the largest population center in the Bristol Bay
watershed, the Dillingham/Aleknagik area (Marsik 2009), and slightly less than half of the combined
capacity of the two electric utilities that serve more than 40% of Alaska's total population (CEA 2011,
ML&P2012).
• Dormitories for such a mine would house more than 2,000 people during construction and more than
1,000 people during mine operation (Ghaffari et al. 2011). Thus, the mine site would rival Dillingham as
the largest population center in the Bristol Bay watershed during construction and would remain the
second largest population center during operation.
• The mine site could contain more than 19 km of main roads, as well as numerous pit and access roads,
and would depend on a fleet of 50 to 100 vehicles, in addition to 150 or more large ore-hauling trucks
(Ghaffari et al. 2011). Potential risks associated with these roads would be similar in type to those
described in Chapter 10.
We specify that all mine components would be developed using modern conventional design and
technologies and operated under standard industry practices. Our purpose in this assessment is to
evaluate the potential effects of mining porphyry copper deposits in the Nushagak and Kvichak River
watersheds given design and operation to these standards. We have included basic descriptions of
design features intended to mitigate potential adverse effects of mine operation.
Bristol Bay Assessment « January 2014
-------
Chapters Mine Scenarios
In spite of these design and operation standards, however, any large-scale mine in the Bristol Bay region
would have a footprint that would affect aquatic resources (Figures 6-1 through 6-3). These footprint-
related impacts are addressed in Chapter 7. Additional impacts that may result from human error,
mechanical failure, accidents, and other unplanned events are considered in Chapters 8 through 11.
Compensatory mitigation for effects on aquatic resources that cannot be avoided or minimized by mine
design and operation is discussed in Appendix J.
It is important to remember that this is an assessment of mine scenarios, and that like any predictive
assessment it is hypothetical. Although major features of the scenarios will undoubtedly be correct (e.g.,
a pit at the location of the ore body and the generation of a large volume of tailings), some specifics
would inevitably differ in an official mine plan submitted for permitting. All plans—even those
submitted to and approved by state and federal regulators—are scenarios, and unforeseen changes in
design and practice inevitably occur over the course of mine development and operation. The Fort Knox
Mine near Fairbanks, Alaska, provides an example. On October 1, 2012, an Alaska Pollution Discharge
Elimination System permit authorized the Fort Knox Mine to discharge wastewater to nearby Fish
Creek, although the mine was originally designed and permitted in 1994 as a no-discharge facility.
It is also important to note that the largest scenario considered in this assessment, based upon 6.5
billion tons (5.9 billion metric tons) of ore, does not represent complete extraction of the Pebble deposit.
Ghaffari et al. (2011) estimate the entire Pebble mineral resource at 11.9 billion tons (10.8 billion metric
tons); were a mine to be developed that fully extracted this amount of ore, potential effects could be
significantly greater than those estimated in the assessment.
This section describes the mine components common to the three mine size scenarios (and most other
mines of this type, as described in Chapter 4). Section 6.2 describes specific characteristics of each mine
size scenario relevant to our assessment, including water treatment and discharge. Section 6.3 describes
closure of the mines, and Section 6.4 provides conceptual models of the relationships between mine
components, potential stressors, and biotic responses.
Bristol Bay Assessment KA January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-1. Footprint of the major mine components (mine pit, waste rock piles, and tailings
storage facility [TSF]) in the Pebble 0.25 scenario. Light blue areas indicate streams and rivers from
the National Hydrography Dataset (USGS 2012a) and lakes and ponds from the National Wetlands
Inventory (USFWS 2012); dark blue areas indicate wetlands from the National Wetlands Inventory
(USFWS 2012).
* m*U -•+ ~f • ^V- • •«- , ' ~ •
•^^-V-T&l^''*
Bristol Bay Assessment
6-5
January 2014
-------
Mine Scenarios
Figure 6-2. Footprint of the major mine components (mine pit, waste rock piles, and tailings
storage facility [TSF]) in the Pebble 2.0 scenario. Light blue areas indicate streams and rivers from
the National Hydrography Dataset (USGS 2012a) and lakes and ponds from the National Wetlands
Inventory (USFWS 2012); dark blue areas indicate wetlands from the National Wetlands Inventory
(USFWS 2012).
,3>i *V- /' *• • >• • i
^%$er £/£ -v ' H ^^*^ ^ ^ "^V-^'
/^M^^JP1 ' _V-^-;^^
*^ - j--c ^Tv - -Xj ^
r^ ->i *^'k' - ' •
{£-& ^ ~ ^\ ^,
*V^C ]&-* --
«T N^ '1>^«^—
- ^y jiiki
Bristol Bay Assessment
6-6
January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-3. Footprint of the major mine components (mine pit, waste rock piles, and tailings
storage facilities [TSFs]) in the Pebble 6.5 scenario. Light blue areas indicate streams and rivers
from the National Hydrography Dataset (USGS 2012a) and lakes and ponds from the National
Wetlands Inventory (USFWS 2012); dark blue areas indicate wetlands from the National Wetlands
Inventory (USFWS 2012).
,NUSHAGAK
Bristol Bay Assessment
6-7
January 2014
-------
Chapters Mine Scenarios
6.1.1 Location
The mine scenarios considered in this assessment are sited at the Pebble deposit, in headwaters of the
Nushagak and Kvichak River watersheds where the South and North Fork Koktuli Rivers and Upper
Talarik Creek originate (Figure 2-5). The Pebble deposit represents the most likely site for near-term,
large-scale mine development in the Bristol Bay watershed. Many other mineral exploration sites in the
Nushagak and Kvichak River watersheds report findings consistent with a porphyry copper deposit
similar to the Pebble deposit (see Table 13-1 and Figure 13-1 for other mineral prospects in the area).
Non-porphyry copper deposits being explored in the area are likely to require similar mining facilities
such as an open pit, a tailings impoundment, and waste rock dumps, and may produce acid-generating
materials. Salmon and other fishes occur in streams throughout the Nushagak and Kvichak River
watersheds (Chapter 5; Appendices A and B). Thus, much of our analysis is transferable to other
portions of the two watersheds, in that a mining operation at any one of these sites could have
qualitatively similar impacts to a mine operation at the Pebble deposit. However, we recognize that
specific component placement could differ based on site-specific factors at each mine. Because our
scenarios are located at the Pebble deposit, we refer to them throughout the text as Pebble 0.25, Pebble
2.0, and Pebble 6.5. This distinguishes the site of the analysis from other potential mine sites in the
Nushagak and Kvichak River watersheds that are included in the evaluation of potential impacts of
multiple mines (Chapter 13).
6.1.2 Mining Processes
6.1.2.1 Extraction
Ore associated with the western portion of the Pebble deposit is near-surface and, in our scenarios,
would be mined via conventional open-pit mining methods of drilling and blasting. Pit depth and width
would be increased progressively to recover the ore. Pit walls and benches would be constructed to
stabilize slopes for safety and to direct runoff. Dusts would be controlled by wetting surfaces with site
water and covering truck beds during transport of excavated rock. Groundwater flow into the pit would
be managed by pumping to storage ponds or TSFs for later treatment or use in mine processes. Although
our scenarios describe open pit mining, underground methods could be used, particularly for the deeper
eastern portion of the ore body. Many of the impacts would be similar in type and magnitude to those of
surface mining (Section 4.2.3.1).
6.1.2.2 Ore Processing
In the mine scenarios, an in-pit crusher would reduce the ore to particles below a maximum size and a
conveyor would bring the crushed ore to processing facilities. Ore would be processed in a flotation
circuit similar to that described in Section 4.2.3.3. The milling process would generate two tailings
streams, one from the rougher flotation circuit (bulk tailings having undergone a single grind sequence)
and another from the secondary cleaner circuit (cleaner scavenger tailings) (Figure 4-3). Selective
flotation would be used to minimize the amount of potentially acid-generating (PAG) tailings. Copper
(+gold) and molybdenum concentrates would be produced as described in Section 4.2.3.3, with the
Bristol Bay Assessment go January 2014
-------
Chapters Mine Scenarios
copper (+ gold) slurry concentrate pumped via pipeline to Cook Inlet and the final molybdenum
concentrate dried, bagged, and trucked off site for processing. Gold associated with the copper minerals
in the slurry concentrate would be recovered at an off-site smelter. Pyrite tailings would be directed
either to the TSF for subaqueous disposal or to a vat leach cyanidation operation for removal of gold
(Box 4-6), after which sulfide-rich tailings would be directed to the TSF for subaqueous disposal. A
cyanide destruction unit would be used at the end of the leaching process (Box 4-6).
All chemical reagents used in ore processing (Box 4-5) would be transported to the mine site, then
prepared and stored in areas with secondary containment and instrumentation to detect any spills or
leaks. All pipelines would be designed to standards of the American Society of Mechanical Engineers
(ASME), which include the use of liners to minimize abrasion and corrosion, freeze protection,
secondary containment over water bodies, and leak monitoring and detection. Dusts would be
controlled in the processing area through use of cartridges, wet scrubbers, and/or enclosures.
6.1.2.3 Waste Rock
Waste rock consisting of both PAG and non-acid-generating (NAG) materials would be stored around the
mine pit, at least partially within the groundwater drawdown zone from mine pit dewatering. PAG waste
rock would be stored separately from NAG waste rock. Over the life of the mine, PAG waste rock would
be blended with processed ore to allow consistency in chemical usage and to remove material from
surface storage prior to its expected time of acid generation (e.g., within 20 years of its excavation). Any
PAG material remaining unprocessed at the end of mining would be processed separately prior to
closure.
During operation, waste rock piles would be constructed with a 2:1 slope for structural stability and
minimization of the amount of runoff requiring treatment. Waste rock piles would occupy
approximately 2.3,13.0, and 22.6 km2 in the Pebble 0.25, 2.0, and 6.5 scenarios, respectively (Table 6-2).
Water quality of the leachate from waste rock is described in Tables 8-6 and 8-7. Monitoring and
recovery wells and seepage cutoff walls would be placed downstream of waste rock piles to manage
seepage, with seepage and contaminated groundwater directed either into collection ponds for use in
mine processes or for treatment and release to the environment, or into the mine pit. Stormwater falling
upslope of waste rock piles would be diverted around the piles and directed toward sedimentation
ponds for settling of suspended solids prior to discharge to a nearby stream, or for treatment if
determined to be contaminated. Embankments would be constructed above the seepage cutoff walls to
contain any excess stormwater runoff that could not be contained in collection ponds. Water captured in
these embankments would be released or directed to treatment as appropriate. Because the Tertiary
volcanic rocks are classified as NAG (Ghaffari et al. 2011, PLP 2011), they may be useful for building
purposes such as TSF construction. However, because of the potential for metals leaching, use would be
appropriate only where leachate would be collected for treatment as necessary.
Bristol Bay Assessment q January 2014
-------
Chapter 6
Mine Scenarios
Table 6-2. Mine scenario parameters. These scenarios were developed by the U.S. Environmental
Protection Agency for the purposes of this assessment, but draw heavily on specifics put forth by
Ghaffari et al. (2011).
Parameter
Amount of ore mined (billion metric tons)
Ore volume (million m3)
Approximate duration of mining
Ore processing rate (metric tons/day)
Tailings produced, dry (billion metric tons)
Tailings produced, volume (million m3)
Mine Scenario
Pebble 0.25
0.23
86.9
20 years
31,100
0.225
158
Pebble 2.0
1.8
697
25 years
198,000
1.80
1,270
Pebble 6.5
5.9
2270
78 years
208,000
5.86
4,130
Mine Pit
Surface area (km2)
Depth (km)
1.5
0.30
5.5
0.76
17.8
1.24
Waste Rock Pile
Surface area (km2)
PAG waste rock (million metric tons)
PAG waste rock bulk density (metric tons/m3)
PAG waste rock area (km2)
NAG waste rock (million metric tons)
NAG waste rock bulk density (metric tons/m3)
NAG waste rock area (km2)
2.3
86
2.08
0.55
320
2.08
1.78
13.0
580
2.08
1.79
2,200
2.08
11.2
22.6
4,700
2.08
6.77
11,000
2.08
15.8
TSF1"
Capacity, dry weight (billion metric tons)
Surface area, interior (km2)b
Surface area, exterior (km2)
Maximum dam height (m)
Maximum number of dams
Capacity, volume (million m3)
Tailings dry density (metric tons/m3)0
NAG density, embankment (metric tons/m3)0
0.25
6.5
6.8
92
1
177
1.42
2.31
1.97
14.2
16.1
209
3
1,390
1.42
2.31
1.97
14.2
16.1
209
3
1,390
1.42
2.31
TSF 2"
Capacity, dry weight (billion metric tons)
Surface area, interior (km2)b
Surface area, exterior (km2)
Maximum dam height (m)
Maximum number of dams
Capacity, volume (million m3)
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
3.69
20.1
22.7
Not determined
3
2,600
TSF 3"
Capacity, dry weight (billion metric tons)
Surface area, interior (km2)b
Surface area, exterior (km2)
Maximum dam height (m)
Maximum number of dams
Capacity, volume (million m3)
Total TSF surface area, exterior (km2)
NA
NA
NA
NA
NA
NA
6.8
NA
NA
NA
NA
NA
NA
16.1
0.96
8.23
9.82
Not determined
2
680
48.6
Transportation Corridor
Total length (km)
Length in assessment watersheds (km)
138
113
138
113
138
113
Notes:
a Final value when TSF is full.
b Area does not include TSF dams.
c Values are the same for TSF 2 and TSF 3, so not repeated under those TSFs.
NA = not applicable; TSF = tailings storage facility; PAG = potentially acid-generating; NAG = non-acid-generating.
Bristol Bay Assessment
6-10
January 2014
-------
Chapters Mine Scenarios
6.1.2.4 Tailings Storage Facilities
In the mine scenarios, TSF dam design would proceed as described by Ghaffari et al. (2011). The number
and size of TSFs in each scenario (Figures 6-1 through 6-3) would be commensurate with tailings
storage requirements. The water rights application submitted by Northern Dynasty Minerals to the State
of Alaska in 2006 described several potential locations for TSFs (NDM 2006). Drawing on this
information, and given site-specific geotechnical, hydrological, and environmental considerations, we
assume that the higher mountain valleys similar to the site of TSF 1, on the flanks of Kaskanak Mountain,
are the most plausible TSF sites for a mine at the Pebble deposit. This placement does not imply that
these sites would not pose unacceptable environmental harm, or that they would be the least
environmentally damaging practicable alternatives for purposes of Clean Water Act permitting. Permit-
specific study, which is beyond the scope of this assessment, would determine if these or other sites met
these criteria.
At each TSF, a rockfill starter dam would be constructed, with a liner (high-density polyethylene
geomembrane on top of a geosynthetic clay liner) extending up the upstream dam face. Seepage capture
and toe drain systems would be installed at the upstream toe, with perpendicular drains installed to
direct seepage toward collection ponds. Each TSF would be unlined other than on the upstream dam
face, and there would be no impermeable barrier constructed between tailings and underlying
groundwater. As tailings accrued near the top of the starter dam, dam height would be raised using the
downstream construction method (Figure 4-4) (Ghaffari et al. 2011). At some point, dam construction
would shift to the centerline method (Figure 4-4), and a new stage would be constructed as the capacity
of each previous stage was approached. TSF 1 would require maximum dam heights of approximately
92 m for the Pebble 0.25 scenario and 209 m for both the Pebble 2.0 and Pebble 6.5 scenarios (Table 6-
2, Figure 6-4).
Given the low grade of ore expected in the region, our mine scenarios would produce large amounts of
tailings: approximately 99% of the mass of ore processed would be tailings, with 85% as NAG bulk
tailings and 14% as PAG (pyritic) tailings (Ghaffari et al. 2011). Both types of tailings would be directed
to TSFs (Figures 6-1 through 6-3). The discharge of bulk tailings would be managed such that the
coarsest materials (fine sand) would be discharged at intervals along the inside perimeter of the TSF to
form beaches, while finer materials (silt) would be carried with discharged water toward the center of
the impoundment. Pyritic tailings would be discharged below the water surface of the tailings pond and
encapsulated in NAG tailings to retard the rate of pyrite oxidation.
The capacity and dimensions of each TSF are listed in Table 6-2. Pebble 6.5, the largest size scenario
considered, would require the construction of TSFs 1, 2, and 3, with a combined tailings capacity
exceeding 6 billion metric tons. We estimate that these three TSFs would have a combined surface area
of more than 48 km2 (Table 6-2).
Bristol Bay Assessment 611 January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-4. Height of the dam at tailing storage facility (TSF) 1 in the Pebble 2.0 and Pebble 6.5
scenarios, relative to U.S. landmarks.
260-r
240-
220-
200
180
160
140
120
100
80
60
40
20
0
Transamerica Building - 260 Meters
Tailings Dam TSF 1-209 Meters
Gateway Arch -192 Meters
Washington Monument -169 Meters
Tailings Reservoir
During operation, water quality in TSF ponds would be similar to process water. At the end of mining,
process water would no longer enter the TSF, so it is expected that, over time, dilution from
precipitation would cause the composition of tailings pond water to approach that of local surface water.
Seepage from the base of the tailings impoundment, either during operation or after closure, would be
expected to be similar to water quality estimates based on pre-mining humidity-cell test results
(Appendix H). The low solubility of oxygen in water (less than 15 mg/L) limits the access of oxygen to
submerged unreacted sulfide minerals in the tailings, reducing dissolution reaction rates and thus the
concentration of solutes. In addition, trace amounts of carbonate or silicate minerals may partially
neutralize acid under anoxic conditions commonly encountered in sulfidic tailings, further limiting the
solubility of metals and other trace elements (Blowes et al. 2003). However, a good deal of uncertainty
exists because the humidity cell tests used to predict pore water chemistry represent a small sample of
the ore body. Thus, actual water quality in the tailings impoundment may differ significantly from what
is estimated (Appendix H). For example, lower concentrations of metals than those reported in humidity
cells tests would likely be seen in TSF water if pH was buffered by reactions with carbonate and silicate
minerals (see Section 8.1.1.1 for discussion of tailings leachate quality).
Well fields spanning the valley floor would be installed at the downstream base of all tailings dams to
monitor groundwater flow down the valley, including potential uncaptured seepage from the TSF. If
contaminated groundwater was detected, monitoring wells would be converted to collection wells or
Bristol Bay Assessment
6-12
January 2014
-------
Chapters Mine Scenarios
new recovery wells would be installed, and water from the well field would be pumped back into the
TSF or treated and released to stream channels.
6.1.2.5 Water Management and Treatment
Water uses in the mine scenarios would include ore processing, tailings slurry transport, and transport
of copper concentrate slurry in the product pipeline. In this section, we provide an overview of water
management and treatment in the three mine scenarios. Figure 6-5 presents a schematic illustration of
these components (note that, for clarity, diversions of stormwater around mine components are not
shown on the schematic).
• Stormwater runoff that did not contact potential contaminants would be diverted around mine
components (e.g., waste rock piles, processing facilities) in ditches directed toward sediment
settling ponds.
• Stormwater runoff from waste rock piles and water from pit dewatering would be pumped to lined
process water ponds; water reclaimed from tailings impoundments or tailings thickening also would
be stored in the process water storage ponds for reuse in ore processing.
• Stormwater falling onto TSFs would be stored in the tailings impoundments and used in the process
water cycle.
• Seepage collected from waste rock piles and TSFs would be directed to lined seepage collection
ponds or TSFs for later treatment.
• Seepage escaping the waste rock and TSF leachate collection systems would be monitored with
monitoring wells. If groundwater contamination was detected, wells would be converted to recovery
wells or new recovery wells would be installed, and the groundwater pumped to either a TSF or a
storage pond for later treatment.
• Water reclaimed from the copper concentrate after transport to the port would be returned to
process water storage ponds via pipeline from the port.
• Streams blocked by the mine pit or waste rock piles would be diverted, where practicable, around
and downstream of the mine. However, the zone of groundwater depression around the mine pit
and the slow filling of the post-operation pit would likely dewater these streams for as long as it
took the pit to fill, which could be hundreds of years.
• Prior to being discharged, water would be treated to meet effluent limits using chemical
precipitation methods and/or reverse osmosis. Water would be discharged to the South and North
Fork Koktuli Rivers according to permit conditions for composition, flow, and temperature. Sludge
and brine from the treatment process would be disposed in the TSF.
Bristol Bay Assessment 613 January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-5. Water management and water balance components for the three mine scenarios
Return Water
and Site Runoff
from Port
Bristol Bay Assessment
6-14
January 2014
-------
Chapters Mine Scenarios
Water balances for both the operation and post-closure phases of our mine scenarios are discussed in
detail in Section 6.2.2. Development of these water balances is important, because they estimate the
amount of water available to contribute to downstream flows. Calculating these water balance
components is challenging, however, and requires a number of assumptions (e.g., estimates of the
amount of water needed to support mining operations, the amount of water delivered to the site via
precipitation, the amount of water lost due to evaporation, and the net balance of water to and from
groundwater sources). Information exists to estimate precipitation and evaporation, and estimates of
water needed for mining operations are available based on typical mining practices (Ghaffari et al.
2011). More challenging—and potentially the largest source of uncertainty in these calculations—is the
net balance of water from groundwater sources.
Mining operations would affect the quantity, quality, timing, and distribution of surface flows. Mining
operations always consume some water, so there would be less water available in the landscape during
active mining than before the mine was present. Major stream flow reductions during mine operation
would result from the capture of precipitation falling on the mine pit, waste rock piles, and TSFs
(Table 6-3, Figure 6-5). The mine pit would capture precipitation directly, but pit dewatering would also
draw down the water table beyond the rim of the pit, creating a cone of depression that would extend
underneath the waste rock piles (Figure 6-5). Leachate recovery wells for any detected groundwater
contamination downstream of the waste rock piles would extend the cone of depression. Because the
mine pit would be located on a water divide, we estimate that there would be little net contribution from
groundwater flow into the area defined by the cone of depression, and that the cone of depression
would expand until water flow into the mine pit was balanced by recharge from precipitation over the
cone of depression. The cone of depression would lower the groundwater table, drying up streams,
ponds, and wetlands that depend on groundwater discharge and turning areas of groundwater
discharge into areas of groundwater recharge. Precipitation and other water collected in the mine pit or
from recovery wells would be pumped to a process water pond or to one of the TSFs. Water falling
within the perimeter of a TSF would be captured directly in the TSF, but runoff from catchment areas
up-gradient of the TSF would be diverted downstream. Runoff at the port site would be pumped to the
mine site in the return water pipeline, contributing to the mine's water supply and avoiding the need for
treatment at the port.
Prior to active mining, but after the starter dam was built for TSF 1, site water would be diverted to TSF
1 to allow sufficient water for process plant startup. During mine operation, groundwater and
precipitation would be pumped from the mine pit to prevent flooding of the mine workings (Figure 6-5).
Water would be needed for the flotation mill, to operate the TSF, and to maintain concentrated slurry in
the product pipeline.
Bristol Bay Assessment 615 January 2014
-------
Chapter 6
Mine Scenarios
Table 6-3. Summary of annual water balance flows (million m3/year) during operations for the three
mine scenarios.
Flow Component
Captured at mine pit area
Captured atTSF 1
Captured atTSF 2
Captured atTSF 3
Captured at mill & other facilities
Potable water supply well(s)
Water in ore (3%)
Total Captured
Cooling tower losses
In concentrate to port
In concentrate return
Runoff collected from port
Stored in TSF as pore water
Stored in mine pit
Crusher use
Total Consumptive Losses
Returned to streams via wastewater treatment plant
Returned as NAG waste rock leachate
Returned as PAG waste rock leachate
Returned as TSF leakage
Total Reintroduced
Percent of Captured Water Reintroduced
Pebble 0.25
9.77
5.86
NA
NA
0.629
0.031
0.340
16.6
0.211
0.166
-0.149
-0.125
3.72
0
0.113
3.93
10.9
0.676
0
1.11
12.7
76.3%
Pebble 2.0
22.4
13.8
NA
NA
2.69
0.124
2.17
41.2
1.32
1.04
-0.934
-0.251
23.8
0
0.722
25.7
10.3
2.58
0.216
2.35
15.4
37.5%
Pebble 6.5
44.1
13.8
19.5
8.43
2.69
0.124
2.27
91.0
1.32
1.04
-0.934
-0.251
24.9
0
0.758
26.8
51.0
4.97
1.03
7.20
64.2
70.5%
Notes:
TSF = tailings storage facility; NA = not applicable; NAG = non-acid-generating; PAG = potentially acid-generating.
In hard rock metal mining, most water use occurs during milling and separation operations; however,
much of this water is recycled and reused. For example, much of the water used to pump the tailings
slurry from the mill to a TSF becomes available when the tailings solids settle, and excess overlying
water is pumped back to the mill. Water losses occur when there is a consumptive use and that water is
no longer available for reuse (Table 6-3, Section 6.2.2). Consumptive losses would be made up by
withdrawing water stored in a TSF or by pumping directly from the mine pit. Some of this captured
water (approximately 38 to 76%, Table 6-3) would not be needed at the mine site. This excess captured
water would be treated to meet existing water quality standards and discharged to nearby streams
(Figure 6-5), partially mitigating streamflow lost due to eliminated or blocked upstream reaches
(Chapter 7).
6.1.3 Transportation Corridor
6.1.3.1 Roads
Development of any mine in the Bristol Bay watershed would require substantial expansion and
improvement of the region's transportation infrastructure. The Bristol Bay watershed is located in one
of the last remaining, virtually roadless regions in the United States. There are no improved federal or
Bristol Bay Assessment
6-16
January 2014
-------
Chapters Mine Scenarios
state highways, and no railroads, pipelines, or other major industrial transportation infrastructure.
Roadways presently link Iliamna Lake (Pile Bay) to Cook Inlet (tidewater at Williamsport) and the
Iliamna area (including the Iliamna airport) north to the site of a proposed bridge over the Newhalen
River near the village of Nondalton. Two other short road segments link Dillingham to Aleknagik and
Naknek to King Salmon (Figure 6-6). Local roads also exist in villages throughout the Nushagak and
Kvichak River watersheds. Most people travel by air or boat during the ice-free season, and by air or
snow machine in winter.
In our mine scenarios, a 138-km, two-lane (approximately 9-meter-wide), gravel surface, all-weather
permanent access road would connect the mine site to a new deep-water port on Cook Inlet (Figure 6-
6), from which concentrate would be shipped elsewhere for processing (Ghaffari et al. 2011). An
estimated 113 km of this corridor would fall within the Kvichak River watershed (this distance does not
include the portion of the road occurring within the potential mine site). This route would traverse
highly variable terrain and variable subsurface soil conditions, including extensive areas of rock
excavation in steep mountainous terrain.
The primary purpose of this road would be to transport freight by conventional highway tractor-trailers,
although critical design elements would be dictated by specific oversize and overweight loads associated
with project construction. Material sources for road embankment fill, road topping, and riprap (e.g.,
borrow and gravel pits and rock quarries) would be available at regular intervals along the road route.
We assume state-of-the-art practices for design, construction, and operation of the road infrastructure,
including design of bridges and culverts for fish passage. Permanent structures would be designed for a
service life of 50 years. Because the access road would be kept open for ongoing care, maintenance, and
environmental monitoring at the site post-closure, maintenance and resurfacing of the access road
would necessarily be required over the same time period, which may extend in perpetuity.
The transportation corridor would cross many streams (including unmapped tributaries), rivers,
wetlands, and extensive areas with shallow groundwater, all of which drain to Iliamna Lake (Figure 6-6,
Section 10.1). We used a mean annual streamflow threshold of >0.15 m3/s to designate stream crossings
that would be bridged (this threshold was also used to separate small headwater streams from medium
streams in broad-scale characterization of stream and river habitats; see Section 3.1.4.2). Bridges, with
spans ranging from approximately 12 to 183 m, would be constructed over 12 known anadromous
streams and seven additional streams likely to support salmonids. Culverts would be place at all
remaining stream crossings. In addition, there would be a 573-m (1,880-foot) causeway across the
upper end of Iliamna Bay, and approximately 8 km of embankment construction along coastal sections
in Iliamna Bay and Iniskin Bay (Ghaffari et al. 2011).
Bristol Bay Assessment 617 January 2014
-------
Chapter 6
Mine Scenarios
A Approximate Pebble Deposit Location
Transportation Corridor
= —: ~ Transportation Corridor (Outside Assessment Area)
Watershed Boundary
Existing Roads
Bristol Bay Assessment
6-18
January 2014
-------
Chapter 6
Mine Scenarios
Avalanche hazards exist in isolated locations along the alignment, but routing would attempt to avoid
any avalanche chutes and runout areas. Because of steep mountain slopes and the lack of significant
vegetation at high elevations, storm runoff can rapidly accumulate and result in intense local runoff
conditions. Road areas near the south slope of Knutson Mountain and the southeast slope of the
mountain above Lonesome Bay and Pile Bay (Figure 6-6) may be especially susceptible to these runoff
events, as demonstrated in late 2003 when storm runoff washed out several culverts on the state-
maintained Pile Bay Road.
6.1.3.2 Pipelines
The transportation corridor would include four pipelines, which would carry copper (+gold)
concentrate, return water, natural gas (to fuel a natural gas-fired power generating plant), and diesel
fuel between the mine site and the Cook Inlet port (Table 6-4). All pipelines would be designed following
ASME standards. Except at stream and river crossings, pipelines would be buried together in a trench
adjacent to the road alignment, in the right-of-way. At short stream and river crossings, pipelines would
be bored under channels to minimize waterway impacts. At longer crossings, pipelines would be
supported aboveground on road bridges. Any aboveground pipeline sections would be constructed of
double-walled pipe. Freeze protection would be provided by insulation (aboveground pipes) or burial
(1.5 meters below ground surface). External corrosion would be prevented by a cathodic protection
system. A leak detection system would be built into the pipelines, which would also assist in the
detection and prevention of slack flows. A supervisory control and data acquisition (SCADA) system
would monitor and control pumping facilities via a fiber optic line buried alongside the pipelines.
Instruments such as pressure and temperature transducers located along the pipeline route would be
tied into the fiber optic link.
Table 6-4. Characteristics of pipelines in the mine scenarios.
Pipeline
(number of pipes)
Route
Pipe Material
Nominal Diameter (cm)
Along Transportation Corridor
Copper (+gold) concentrate (1)
Reclaimed water (1)
Natural gas (1)
Diesel fuel (1)
Mine to port
Port to mine
Port to mine
Port to mine
HDPE-lined steel
HDPE-lined steel
Steel
Steel
20
18
5
13
At Mine Site
Bulk tailings (2)
Pyritic tailings (2)
Reclaimed water (1)
Reclaimed water (1)
Mine pit dewatering(l)
Process plant to TSF
Process plant to TSF
TSF barge to TSF head tank
TSF head tank to process
pond
Pit to process pond or TSF
Steel with liner
Steel with liner
HOPE
Steel
Steel
86
46
107
107
TBD
Notes:
HOPE = high-density polyethylene; TSF = tailings storage facility; TBD = to be determined.
Source: Ghaffari et al. 2011.
Bristol Bay Assessment
6-19
January 2014
-------
Chapters Mine Scenarios
On the mine site, pipelines would carry tailings slurry from the process plant to the TSFs and reclaimed
water from the TSFs to the process water storage ponds (Table 6-4). There also would be smaller
pipelines for water supply, firefighting, and process flows within the plant. In this assessment, we
assume that any leakage from pipelines in the process plant area would be captured and controlled by
the plant's drainage system and either be treated prior to discharge or pumped to the process water
storage pond or the TSFs. Failures of these on-site pipelines could result in uncontrolled releases in the
mine site, but these failures are not evaluated in this assessment. At mine closure, concentrate and
return water pipelines would be removed. Diesel and natural gas pipelines would be retained as long as
fuel was needed at the site for monitoring, treatment, and site maintenance. It is also possible that local
communities would select to retain the pipelines for continued use.
6.2 Specific Mine Scenarios
In this assessment we evaluate three specific mine scenarios representing mines of different sizes. The
smallest mine scenario, Pebble 0.25, represents a median-sized porphyry copper deposit of 250 million
tons (230 million metric tons) (Singer et al. 2008). The second mine scenario, Pebble 2.0, is based
largely on the 25-year, 2 billion tons (1.8 billion metric tons) case described by Ghaffari et al. (2011) for
initial development at the Pebble deposit. The third mine scenario, Pebble 6.5, is based largely on the
78-year, 6.5 billion tons (5.9 billion metric tons) case described by Ghaffari et al. (2011) for further
resource development at the Pebble deposit.
Pebble 2.0 and Pebble 6.5 reflect projects based on extensive exploration, assessment, and preliminary
engineering, which are described by Ghaffari et al. (2011) as "economically viable, technically feasible
and permittable." They are among the most likely to be developed in the Bristol Bay watershed and are
site-specific to the Pebble deposit. For the purposes of this assessment, we have also placed the Pebble
0.25 scenario at the Pebble deposit because of the availability of site-specific information. If mines are
developed at other exploration sites in the watershed (Figure 13-1), they are likely to have
characteristics and impacts much closer to those of the Pebble 0.25 scenario. Table 6-2 provides detailed
parameters for each of our three mine scenarios, and Figures 6-1 through 6-3 show the general layout of
each scenario's major mine components.
6.2.1 Mine Scenario Footprints
The major mine components contributing to each mine scenario footprint are the mine pit, waste rock
piles, and TSFs. Placement of these components for each of the scenarios is shown in Figures 6-1
through 6-3. In each case, these layouts represent one possible configuration for the mine. Other
configurations are possible, but would be expected to have impacts of similar types and magnitudes.
Each mine scenario footprint also includes two additional components: the groundwater drawdown
zone, or the area over which the water table is lowered due to pit dewatering (Figure 6-5), and the area
covered by plant and ancillary facilities (e.g., ore-crushing and screening areas, processing mill, storage
and stockpile areas, workshops, roads within the mine site, pipeline corridors, and other disturbed
Bristol Bay Assessment 620 January 2014
-------
Chapter 6
Mine Scenarios
areas). Summing these areas (mine pit, waste rock piles, TSFs, drawdown zone, and plant and ancillary
facilities) and correcting for any overlap among them yields an estimate for total mine footprint area in
each scenario (Tables 6-5 through 6-7).
6.2.1.1 Pebble 0.25 Footprint
Figure 6-1 shows the general layout of the mine pit, waste rock piles, and TSF for the Pebble 0.25
scenario. The TSF, identified as TSF 1, is located in a natural valley in a headwater tributary of the North
Fork Koktuli River located to the west of the Pebble deposit. The valley would be closed off by the
construction of a rockfill dam 92 m in height (Table 6-2). The waste rock pile area was determined by
calculating the area that would be covered by the expected volume of waste rock, assuming
approximately 100-m-high piles and taking advantage of natural landforms near the mine pit. In this
scenario, separate PAG and NAG waste rock would be created during mine operation. PAG waste rock
would be processed as mill conditions permit throughout the mine life, with the intent to process all of
the PAG waste rock before mine closure. The area of the plant and ancillary facilities is estimated to
account for approximately 4% of the total mine footprint area (Table 6-5). The drawdown zone (Table
6-5) includes the mine pit and the area beyond the mine pit perimeter, including some of the waste rock
piles, up to the limit of the cone of depression (see Box 6-2 for discussion of mine pit drawdown
calculations).
Table 6-5. Estimated areas for individual mine components in the Pebble 0.25 scenario.
Component
Drawdown zone
Mine pit
NAG waste rock in drawdown zone
PAG waste rock in drawdown zone
Other area in drawdown zone
NAG waste rock not in drawdown zone or TSFs
PAG waste rock not in drawdown zone
Cumulative plant and ancillary areas
TSFsa
TSF1
TOTAL MINE FOOTPRINT
Area (km2)
10.1
1.54
0.49
0.55
7.49
1.29
0.00
0.73
6.82
6.82
18.9
Notes:
a Exterior TSF area.
b NAG = non-acid-generating; PAG = potentially acid-generating; TSF = tailings storage facility.
6.2.1.2 Pebble 2.0 Footprint
Figure 6-2 depicts the general layout of the major mine components for the Pebble 2.0 scenario,
including the mine pit, the waste rock piles, and the TSF. The TSF is located in the same valley as TSF 1
in the Pebble 0.25 scenario (Figure 6-2), but it is increased in size to accommodate the additional
tailings expected with this larger mine size. Plant and ancillary facilities are estimated to account for
approximately 7% of the total disturbed area (Table 6-6).
Bristol Bay Assessment
6-21
January 2014
-------
Chapter 6
Mine Scenarios
Waste rock piles are located around the perimeter of the mine pit, with separate areas designated for
NAG and PAG waste rock. As in the Pebble 0.25 scenario, PAG and NAG waste rock would be stored in
separate waste rock piles during mine operation, and the PAG rock would be processed as mill
conditions permit throughout the mine life with the intent to process all of the PAG waste rock before
mine closure. Dewatering of the mine pit would generate a cone of depression around the pit, and more
than half of the area of the waste rock piles would fall within the resulting drawdown zone (Table 6-6).
Table 6-6. Estimated areas for individual mine components in the Pebble 2.0 scenario.
Component
Drawdown zone
Mine pit
NAG waste rock in drawdown zone
PAG waste rock in drawdown zone
Other area in drawdown zone
NAG waste rock not in drawdown zone or TSFs
PAG waste rock not in drawdown zone
Cumulative plant and ancillary areas
TSFs"
TSF1
TOTAL MINE FOOTPRINT
Area (km2)
21.4
5.50
7.08
1.29
7.52
4.14
0.50
3.13
16.1
16.1
45.3
Notes:
8 Exterior TSF area.
b NAG = non-acid-generating; PAG = potentially acid-generating; TSF = tailings storage facility.
6.2.1.3 Pebble 6.5 Footprint
The general layout of the Pebble 6.5 scenario is similar to that of the Pebble 2.0 scenario, with major
differences being a larger open pit, different and expanded areas for the waste rock piles, and the
inclusion of two additional TSFs (TSF 2 and TSF 3) to store the increased tailings volume (Figure 6-3,
Table 6-7). Placement of TSF 2 and TSF 3 in this scenario draws upon some of the TSF options presented
in Northern Dynasty Minerals' water rights application (NDM 2006) and takes advantage of natural
landforms in the Pebble deposit area.
The mine pit is located as shown by Ghaffari et al. (2011), based on evaluation of the Pebble deposit.
Waste rock piles are located around the perimeter of the expanded mine pit, with some portion of the
PAG waste rock stored in the mine pit to utilize storage within the drawdown zone prior to PAG waste
rock being taken to the surface for processing. This practice would reduce the amount of PAG waste rock
that must be stored outside the drawdown zone and, therefore, the amount of PAG leachate that could
seep into the South Fork Koktuli River.
Areas of the plant and ancillary facilities are the same as those described for the Pebble 2.0 scenario;
because production rates of the Pebble 2.0 and Pebble 6.5 scenarios are similar, no increase in these
areas is needed for the larger mine scenario.
Bristol Bay Assessment
6-22
January 2014
-------
Chapter 6
Mine Scenarios
Table 6-7. Estimated areas for individual mine components in the Pebble 6.5 scenario.
Component
Drawdown zone
Mine pit
NAG waste rock in drawdown zone
PAG waste rock in drawdown zone
Other area in drawdown zone
NAG waste rock not in drawdown zone or TSFs
PAG waste rock not in drawdown zone or mine pit
Cumulative plant and ancillary areas
TSFsa
TSF1
TSF2
TSF3
TOTAL MINE FOOTPRINT
Area (km2)
43.4
17.8
10.3
4.37
10.9
5.50
2.40
3.13
48.6
16.1
22.7
9.8
103
Notes:
"" Exterior TSF area.
b NAG = non-acid-generating; PAG = potentially acid-generating; TSF = tailings storage facility.
6.2.2 Water Balance
Many of the potentially significant impacts of large-scale mining relate to a mine's use of water and its
impact on water resources. To understand potential impacts of water use in our mine scenarios, we
developed an annual water balance for each scenario that accounts for major flows into and out of the
mine area. Three major categories of flows make up each water balance estimate: water inputs,
consumptive losses, and water outputs; these categories are discussed in detail in the following sections.
These water balances focus on changes in flows entering or leaving the mine site, relative to pre-mining
conditions. Changes are divided into flows that would be withdrawn or captured from the natural
system and flows that would be released to the natural system. Each water balance subtracts
consumptive water losses within mine operations from water inputs to determine the water available
for release. This water balance analysis does not attempt to describe or quantify internal flows among
mine components, although some are mentioned when necessary to explain the analysis. The water
balance analysis also does not attempt to quantify any flows that are recycled within the mine site,
because these do not capture water from the environment or release water to it.
6.2.2.1 Water Inputs
Water inputs for each of the three scenarios are summarized in Table 6-3. These inputs are derived
primarily from net precipitation (total precipitation minus any losses due to evapotranspiration) that
falls on the mine footprints and is captured by water collection and management systems within the
mine site. We assume that all captured flows would be available for use by the mine operator. Three
gages surrounding the mine site were used to calculate net precipitation at the mine site: gage SK100B
(USGS gage 15302200) on the South Fork Koktuli River, gage NK100A (USGS gage 15302250) on the
North Fork Koktuli River, and gage UT100B (USGS gage 15300250) on Upper Talarik Creek. Net
Bristol Bay Assessment
6-23
January 2014
-------
Chapters Mine Scenarios
precipitation (or measured runoff) at each gage represents precipitation minus evapotranspiration, plus
or minus interbasin storage, plus or minus internal groundwater storage. We assumed interbasin and
groundwater storage were zero since we were averaging across the three watersheds. Therefore, the
runoff measured at each gage represents net precipitation (precipitation minus evapotranspiration).
Monthly mean flows for each gage were summed across the year, producing an area-weighted average
of net runoff of 860 mm per year.
Water inputs resulting from the mine footprints are calculated as the product of footprint areas
multiplied by annual net precipitation. For the TSFs, the volume of water captured is based on the
interior area of the TSF, defined as the area within the dam crests and excluding the downstream faces
of the rockfill dams.
Dewatering the mine pit would create a cone of depression around the mine extending beyond the limits
of the mine pit. Because the mine pit would be located very close to the water divide between the South
Fork Koktuli River, North Fork Koktuli River, and Upper Talarik Creek watersheds, we assume that there
would be negligible net influx of groundwater from beyond the cone of depression. Most of the
groundwater outside the cone of depression would flow away from the site. Therefore, the area of the
cone of depression would be determined by matching net precipitation falling within the drawdown
zone with the calculated groundwater inflow into the mine pit (Box 6-2).
Precipitation falling on areas outside of these disturbed footprints would infiltrate as groundwater or
flow into streams without treatment. Flow in upstream tributaries blocked by the mine footprint would
be piped or otherwise diverted around the footprint and discharged back into streams without
treatment, where practicable. Because this diverted flow is not captured by the mine operations, it is not
explicitly included in the water balance tabulations and is assumed to remain part of the background
flow.
6.2.2.2 Consumptive Losses
Consumptive losses for each mine scenario are summarized in Table 6-3. To estimate the amount of
water available for release, we subtracted consumptive losses associated with mining activities from the
captured flows (Table 6-3). Consumptive losses would include water pumped to the port in the copper
(+gold) concentrate pipeline minus return water, cooling tower evaporation and drift losses, interstitial
water trapped in the pores of stored tailings, water used for dust suppression, and water stored in the
mine pit after closure. The tailings pore water accounts for over 90% of consumptive loss during mine
operations (Table 6-3). When the tailings settle, about 46% of the volume would consist of voids
between solid particles; the water trapped in these pore spaces would no longer be available for use at
the mine or release to streams.
Bristol Bay Assessment 624 January 2014
-------
Chapters Mine Scenarios
BOX 6-2. MINE PIT DRAWDOWN CALCULATIONS
Groundwater flow into the mine pit was calculated using a simplified model based on the Dupuit-Forcheimer
discharge formula for steady-state radial flow into a fully penetrating well in a phreatic aquifer with a
diameter equal to the average mine pit diameter. The hydraulic conductivity data gathered in the Pebble
deposit area during geologic investigations show significant scatter (Figure 6-7). We based our analysis on
the hydraulic conductivity (k) varying with depth, with logk varying linearly from the surface to a depth of
200 m (k = 1 xlO'4 m/s at the surface and k = 1 xlO'8 m/s at depths greater than or equal to 200 m). Given
these values, negligible flow occurs below a depth of 200 m, so our analytical model included a no-flow
boundary at that depth. To apply the Dupuit-Forcheimer formula, we needed to transform the cross-section
into an equivalent isotropic section by transforming the vertical dimension so that the thickness at any
depth was proportional to the hydraulic conductivity at that depth. The initial water table in our simplified
model was at the ground surface and assumed to be horizontal.
Our analysis assumed that the drawdown at the mine pit was 100 m, but we also verified that the results
were not very sensitive to this assumption. The radius of influence was determined by balancing the net
precipitation falling within the cone of depression with the calculated flow into the mine pit. Inflows were
calculated to be 0.274 m3/s (4,350 gpm), 0.584 m3/s (9,250 gpm) and 1.19 m3/s (18,800 gpm) for the
Pebble 0.25, 2.0, and 6.5 scenarios, respectively. The Pebble 2.0 mine inflow agrees closely with the
estimate provided by Ghaffari et al. (2011).
The cone of depression was determined to extend 1,148 m, 1,222 m, and 1,260 m from the edge of the
idealized circular mine pit in the Pebble 0.25, 2.0, and 6.5 scenarios, respectively. In a geographic
information system (GIS), we established the boundary of the cone of depression at those distances from
the actual perimeter of the mine pits to derive the drawdown zones presented in Tables 6-5 through 6-7.
The waste rock piles do not lie completely within the drawdown zones. This is important in assessing water
quality because precipitation falling on the waste rock piles within the drawdown zone is presumed to be
collected within the mine pit, whereas precipitation falling outside of the drawdown zone is presumed to
migrate away from the mine pit. To assess more accurately the waste rock pile positions relative to the
drawdown zones, we distorted the shape of the cone of depression by superimposingthe drawdown zone on
a uniform flow field with a southern gradient of 0.0354, approximately equal to the slope of the ground
surface across the mine pit from north to south. The effect of this distortion is a shift in the boundaries of
the cone of depression to the north, resulting in larger areas of waste rock outside of the drawdown zones.
Information on flows in the concentrate and return water pipelines and on cooling tower losses is
reported by Ghaffari et al. (2011). The return water pipeline reduces consumptive losses by returning
water from the port (e.g., from dewatering the copper [+gold] concentrate and from stormwater runoff
collected at the port site). We estimated the area of the port facilities over which runoff was likely to be
collected (137,160 m2) and multiplied that area by the precipitation rate at the port (1,830 mm/year) to
determine contributions from port site runoff (Table 6-3). We also included a consumptive loss at the
crusher and screening site for dust control equal to 1% of the mass of the material being crushed.
6.2.2.3 Water Outputs
When the amount of captured water exceeds consumptive losses, water would be available, after testing
and treatment, for release into area streams. This released water may differ from natural stream water
in chemistry and temperature, but would comply with permitted discharge requirements. Water may be
reintroduced at locations, flow rates, or times of year that differ from baseline conditions.
The water deficit for each scenario—that is, the amount of water extracted from the environment and
not returned to streams—is presented in Table 6-3. These water deficits equal the total consumptive
losses of approximately 3.9 million m3/year, 26 million m3/year, and 27 million m3/year for the Pebble
Bristol Bay Assessment 625 January 2014
-------
Chapter 6
Mine Scenarios
0.25, 2.0, and 6.5 scenarios, respectively. The percentage of water reintroduced to streams, including
uncaptured leachate, would equal approximately 76, 38, and 71% of the total water captured in the
Pebble 0.25, 2.0, and 6.5 scenarios, respectively.
Figure 6-7. Hydraulic conductivity in the Pebble deposit area. Data are from three test types:
bedrock packer (Lugeon) tests (blue diamonds, with error bars indicating upper and lower limits of
zone tested) (PLP 2011: Chapter 8 and Appendix 8.1N); overburden rising or falling head tests (red
squares) (PLP 2011: Chapter 8 and Appendix 8.1C); and bedrock rising or falling head tests (green
triangles) (PLP 2011: Chapter 8 and Appendix 8.1C). Red line indicates values used in the
assessment's mine pit drawdown and tailings storage facility leakage calculations.
HYDRAULIC CONDUCTIVITY (m/s)
l.E 09
0
50 -
100 -
J.
X i i-n
Q.
UJ
Q
200
250 -
1
<
l.E-08 l.E 07
*
1
I
i
4
1
1
i
*
3
w
.
t
* 1
f <
' 1
T
L|
-
'
•1
\
:[/
)
1 1
i
4
/
I
1
\
*
i
jrl
, I
4
/
\
i
-i
r<
r:
1
1
l.E-06
•{
A
r -
^
'
l.E-05 l.E 04 l.E 03
\\
i
/
/
/
/
i
X
"l
^
IA'
/
/
i
*
"^'
r
f • •
='
* Packer Tests (showing packer limits)
I
1 Overburden Rising/FallingHead Tests
Bedrock Rising/Falling Head Tests
Approximation used in Bristol Bay Assessment
m
6.2.2.4 Additional Water Balance Issues
During the early life of each mine, there is one other significant source of water that a mine operator
would need to manage that is not considered in Table 6-3: the water obtained from dewatering the
sandy and gravelly overburden overlying the waste rock and ore. Based on an average overburden
thickness of 30.5 m and a porosity of 0.40, dewatering the overburden would produce one-time
quantities of 19 million m3, 67 million m3, and 220 million m3 of water over the mine pit areas in the
Pebble 0.25, 2.0, and 6.5 scenarios, respectively. This water would be expected to be relatively clean and,
if properly managed to control turbidity, could most likely be released without chemical treatment to
maintain or augment stream flow.
Bristol Bay Assessment
6-26
January 2014
-------
Chapters Mine Scenarios
Water treated at the wastewater treatment plant (WWTP) might not be discharged to the same streams
that were dewatered. In accordance with the WWTP discharge points shown by Ghaffari et al. (2011),
the WWTP is assumed to discharge to the South and North Fork Koktuli Rivers, but not to Upper Talarik
Creek (Figures 6-8 through 6-11).
6.3 Closure and Post-Closure Site Management
As discussed in Section 4.2.4, the assessment examines potential impacts both during mine operations
and after mining activities have ceased, either as planned or prematurely. In this section, we consider
how the mine scenarios would be handled during and after closure of the mine.
We assume that the mine would be closed after all economically profitable ore was removed from the
site, leaving behind the mine pit, NAG waste rock piles, and TSFs. Water at the site would require
capture and treatment for as long as it did not meet water quality standards. Weathering of exposed
waste rock and pit walls would release ions of potential concern, such as sulfates and metals.
Weathering to the point where these contaminants decreased toward their pre-mining background
concentrations would likely take hundreds to thousands of years, resulting in the need for monitoring
and management of exposed materials and leachate over that time (Blight 2010). To minimize exposure
of waste rock and pit walls to weathering, we assume that they would be reclaimed. We also assume that
existing water management structures and the WWTP would be monitored and maintained as part of
post-closure operations.
Seepage and leachate monitoring and collection systems, as well as the WWTP, might need to be
maintained for hundreds to thousands of years. It is impossible to evaluate the success of such long-term
collection and treatment systems for mines. No examples exist, because these timeframes exceed both
existing systems and most human institutions. Throughout this section, we refer to the potential need
for treatment over extended periods. The uncertainty that human institutions have the stability to apply
treatment for these timeframes applies to all treatment options.
Bristol Bay Assessment 627 January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-8. Water flow schematic for the Pebble 0.25 scenario. Flows include water from the non-acid-generating waste rock pile and
tailings storage facility (TSF) 1 (dashed black arrows), discharge from the wastewater treatment plant (solid black arrows), flow along the
stream channels (solid blue arrows), and known groundwater transfers (dashed blue arrow). For clarity, only flows greater than 5% of total
outflows from the TSF and waste rock pile are shown. Gage locations are based on U.S. Geological Survey (2012b) and Pebble Limited
Note that the spatial orientation of streams and mine components is for schematic purposes only and is not to scale (see Figure 6-11 for a
spatially accurate map).
NK100A1
NK100A
LEGEND
> Surface Flow (Streams)
> Treated Water Discharge
> Contaminated Groundwater Transfer
— ->• Known GroundwaterTransfer
• Stream Gages
• Confluence Points
NAG Non-Acid-Generating Waste Rock Pile
PAG Potentially-Acid-Generating Waste Rock Pile
TSF Tailings Storage Facility
WWTP Wastewater Treatment Plant
SK South Fork Koktuli
NK North Fork Koktuli
UT Upper Talarik Creek
NK100B
UT100E
UT100D
1 UT100C2
•SK1°°G
SK100F
UT100C1
•
SK100B
SK100B1
•UT119A
UT100C
UT100B
Bristol Bay Assessment
6-28
January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-9. Water flow schematic for the Pebble 2.0 scenario. Flows include water from the potentially acid-generating and non-acid-
generating waste rock piles and tailings storage facility (TSF) 1 (dashed black arrows), discharge from the wastewater treatment plant (solid
flows greater than 5% of total outflows from the TSF and waste rock pile are shown. Gage locations are based on U.S. Geological Survey
(2012b) and Pebble Limited Partnership (2011). Confluence points represent virtual gages that were created for analysis purposes (see
Section 7.3 for additional details). Note that the spatial orientation of streams and mine components is for schematic purposes only and is
not to scale (see Figure 6-11 for a spatially accurate map).
NK100B
NK100A1
UT100E
NK100A
LEGEND
> Surface Flow (Streams)
> Treated Water Discharge
> Contaminated GroundwaterTransfer
— ->• Known Groundwater Transfer
0 Stream Gages
• Confluence Points
NAG Non-Acid-Generating Waste Rock Pile
PAG Potentially-Acid-Generating Waste Rock Pile
TSF Tailings Storage Facility
WWTP Wastewater Treatment Plant
SK South Fork Koktuli
NK North Fork Koktuli
UT Upper Talarik Creek
UT100C2
UT100C1
SK119CP* SK124CP*
•L. +-
SK100CP1
SK100B1
SK100B
UT100C
UT100B
Bristol Bay Assessment
6-29
January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-10. Water flow schematic for the Pebble 6.5 scenario. Flows include water from the potentially acid-generating and non-acid-
generating waste rock piles and tailings storage facilities (TSFs) 1, 2, and 3 (dashed black arrows), discharge from the wastewater treatment
plant (solid black arrows), flow along the stream channels (solid blue arrows), and known groundwater transfers (dashed blue arrow). For
clarity, only flows greater than 5% of total outflows from the TSFs and waste rock piles are shown. Gage locations are based on U.S.
Geological Survey (2012b) and Pebble Limited Partnership (2011). Confluence points represent virtual gages that were created for analysis
only and is not to scale (see Figure 6-11 for a spatially accurate map).
NK100B
UT100E
NK100A
LEGEND
)» Surface Flow (Streams)
> Treated Water Discharge
>• Contaminated Groundwater Transfer
— ^- Known Groundwater Transfer
• Stream Gages
• Confluence Points
NAG Non-Acid-Generating Waste Rock Pile
PAG Potentially-Acid-Generating Waste Rock Pile
TSF Tailings Storage facility
WWTP Wastewater Treatment Plant
SK South Fork Koktuli
NK North Fork Koktuli
UT Upper Talarik Creek
• UT100C2
UT100C1
UT100C
SK100B1
SK100B
UT100B
Bristol Bay Assessment
6-30
January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-11. Approximate locations of stream gages and wastewater treatment plant (WWTP) discharges represented in Figures 6-8
through 6-10. Gages denoted with CP indicate confluence points, where virtual gages were created for analysis purposes. Footprint of the
major mine components of the Pebble 6.5 scenario are shown for reference. Gage locations are based on U.S. Geological Survey (2012b) and
Pebble Limited Partnership (2011).
NK119CP1 V
' NK100C
. NK119B
NK119CP2S
v/
SK100G UT10
Waste Rock
3&
2k
Rivers & Streams (National Hydrography Dataset)
Bristol Bay Assessment
6-31
January 2014
-------
Chapters Mine Scenarios
6.3.1 Mine Pit
Upon mine closure, pit dewatering pumps would be turned off. The cone of depression would persist
around the pit for a time, and groundwater would flow toward the pit in response to the local gradient.
Eventually, the water level in the pit would recover toward equilibrium with the surrounding water
table. Any water exiting the pit through surface channels or pumped from the pit would be tested, and
treated if necessary, prior to discharge to surrounding surface waters. Based on our calculations for
groundwater and precipitation inflows to the pit after operations have ceased, we estimate that the time
required for the pit to fill ranges from approximately 20 years for the Pebble 0.25 scenario to more than
200 years for the Pebble 6.5 scenario. If additional runoff or TSF discharges were directed to the pit
instead of allowed to flow into streams, these time frames would be considerably shorter (e.g.,
approximately 100 years for the Pebble 6.5 scenario).
Upper benches of the pit would be partially backfilled, regraded, covered with plant-growth medium,
and vegetated. Some areas may be flattened to enable construction of wetlands for passive water
treatment. At least portions of the pit walls, as well as rocks on the pit bottom or on side benches, would
consist of mineralized rock that was not economical to mine. Any exposed rock containing sulfide
minerals would likely be acid-generating for as long as it remained above the water surface in the pit,
resulting in water with low pH and dissolved metals running down the sides of the pit into the water
body at the bottom. As water level in the pit rose, pit walls would become submerged and exposure to
oxygen would be reduced. Eventually, acid generation would be expected to cease from rocks below the
water's oxic zone. Exposed rock above the water surface or within the oxic zone would continue to
produce acidic metal-sulfate salts that would run into the pit lake with precipitation and snowmelt.
Surfaces anticipated to produce acidic drainage could be sealed against exposure to oxygen. However,
this might not be effective for a pit of this size, since it might be difficult to seal all cracks and fissures in
the pit walls. There could be degradation of sealants from exposure to sun and air, and freeze-thaw
fracturing of rock could reduce acid-preventing efficacy over time. Predicting pit water quality has a
high degree of uncertainty (Section 8.1.4; Appendix I) (Gammons et al. 2009), but water would need to
be monitored and treated to meet effluent requirements prior to being discharged to streams, for as long
as the water remained contaminated.
6.3.2 Tailings Storage Facilities
At closure, tailings beaches in the TSFs would be covered with NAG waste rock and a plant-growth
medium, then vegetated with native species (Ghaffari et al. 2011). Embankments and crests also would
be covered with a growth medium and vegetated. The tailings pond would be drawn down to prevent
flooding and to maintain stability, but a pond of sufficient depth would be retained to keep the core of
PAG tailings hydrated and minimize oxidation. Retaining water in the tailings maintains a higher
potential for tailings dam failure than if the tailings were drained; however, draining the tailings to
stabilize them could allow oxygen-rich water to percolate through the tailings and oxidize the sulfides.
As long as a cover of water is maintained, oxygen movement into the tailings would be retarded,
Bristol Bay Assessment 632 January 2014
-------
Chapters Mine Scenarios
minimizing acid generation. Drawing down the water level in the TSF would also provide capacity for
unusual precipitation events, reducing the likelihood that a storm would provide enough precipitation
to overwhelm capacity and cause tailings dam failure or overtopping. Additionally, wetlands might be
included in reclamation to provide additional stormwater retention, passive water treatment, and
significantly increased evapotranspiration (Reeve and Gracz 2008).
TSFs would require active management for hundreds to thousands of years (Blight 2010). A tailings dam
is an engineered structure that requires monitoring to ensure structural and operational integrity. An
assumption in the mining industry is that tailings continue to compact, expelling interstitial water and
becoming more stable over time. However, there appears to be little data available that document the
magnitude of this stability gain. A recent analysis suggests that densification of oil sands tailings may
stop after a period of time (Wells 2011). Although oil sands tailings are different from porphyry copper
tailings, the principle is the same. Lack of data specific to porphyry copper tailings suggests a cautious
approach, so we do not assume that tailings consolidate to a fully stable land form. Even if the tailings
did consolidate over time, they would remain susceptible to erosion if the tailings dam were
compromised. Thus, the system may require continued monitoring to ensure hydraulic and physical
integrity in perpetuity.
6.3.3 Waste Rock
Some NAG waste rock would be used to cover tailings beaches, and some would be used to backfill
upper portions of the mine pit. The remaining NAG waste rock would be sloped to a stable angle (e.g.,
less than 15 degrees [Blight and Fourie 2003]), covered with soil and plant-growth medium, and
vegetated with native species. No PAG waste rock would remain on the surface, as it would have been
processed either as blending material during operations or at the end of operations.
6.3.4 Water Management
Table 6-8 summarizes the flow components of the water balance after closure, both during the period in
which the mine pit is filling and the steady state condition after the mine pit reaches its maximum water
level. During the post-closure period, the mine would still capture water from precipitation over the
mine pit, waste rock piles, and the TSFs. Groundwater would continue to flow into the mine pit, so
precipitation over the cone of depression would continue to contribute to the captured water.
Consumptive losses from operation would cease, but water stored in the mine pit would constitute a
new consumptive loss until the mine pit water level reaches equilibrium with the surrounding
groundwater level.
The footprint of the mine would be reduced as land occupied by production facilities is reclaimed. For
purposes of estimating water inputs, we assume that 80% of the areas disturbed by the plant and
ancillary facilities would be reclaimed, but that some facilities (e.g., the fuel depot, the WWTP, some
pipelines, and part of the camp) would remain.
Bristol Bay Assessment 633 January 2014
-------
Chapter 6
Mine Scenarios
Table 6-8. Summary of annual water balance flows (million m3/year) during the post-closure period
for the Pebble 6.5 scenario.
Flow Component
Captured at mine pit area
Captured atTSFl
Captured atTSF2
Captured atTSF3
Captured at mill & other facilities
Potable water supply well(s)
Water in ore (3%)
Total Captured
Cooling tower losses
Water in concentrate to port
Water in concentrate return
Runoff collected from port
Stored in TSF as pore water
Stored in mine pit
Crusher use
Total Consumptive Losses
Returned to streams via wastewater treatment plant
Returned as NAG waste rock leachate
Returned as PAG waste rock leachate
Returned as TSF leakage
Total Reintroduced
Percent of Captured Water Reintroduced
During Mine Pit Filling
39.7
13.0
18.4
7.76
0.538
0
0
79.4
0
0
0
0
0
37.3
0
37.3
33.9
0.947
0
7.20
42.1
53.0%
Post-Closure
26.1
13.0
18.4
7.76
0.538
0
0
65.8
0
0
0
0
0
0
0
0
57.6
0.947
0
7.20
65.8
100%
Notes:
TSF = tailings storage facility; NAG = non-acid-generating; PAG = potentially acid-generating.
As the mine pit fills, the cone of depression would shrink to the point that most or all of the waste rock
would be outside of the drawdown zone. Runoff from the reclaimed NAG waste rock piles would either
seep into the ground, travel as overland flow, or be diverted to streams. Some precipitation would be
expected to infiltrate through the NAG waste rock cover, drain through the waste rock piles, and become
groundwater. Runoff from the reclaimed NAG waste rock piles is not anticipated to require treatment,
but would be monitored periodically to confirm this assumption.
The elevation of the north rim of the Pebble 6.5 mine pit would be over 100 m higher than the elevation
of the south rim, so that even when the mine pit reaches its maximum water level there would still be
seepage into the pit from the higher ground. For water balance purposes, we estimate that the post-
closure cone of depression would extend an average of 100 m beyond the pit rim as a result of surface
outflow or pumping.
Precipitation falling on the post-closure tailings would be monitored and discharged downstream or
diverted for treatment in the WWTP, as necessary, to meet water quality standards. Stormwater
diversions and collection systems from the operations phase would be maintained and water directed
away from the TSF, or, if risk of contamination existed, toward the WWTP for treatment prior to
discharge to streams. Interstitial water within the tailings would continue to seep into naturally
Bristol Bay Assessment
6-34
January 2014
-------
Chapters Mine Scenarios
fractured bedrock below the TSF. The well field placed downstream from the TSF during operations
would be retained and monitored post-closure, with water pumped and treated if determined to be
contaminated by leachate from the TSF. The pit water would be monitored and treated prior to being
released to streams, for as long as concentrations of contaminants exceeded effluent limits.
6.3.5 Premature Closure
Many mines close before their ore reserves are exhausted. In one study of international mine closures
between 1981 and 2009, 75% of the mines considered were closed before the mine plan was fully
implemented (Laurence 2011). The Illinois Creek and Nixon Fork mines are examples of mines that have
closed prematurely in Alaska.
Closure before originally planned—that is, premature closure—may occur for many reasons, including
technical issues, project funding, deteriorating markets, operational issues, or strategic financial issues
of the owner. Premature closures can range from cessation of mining with continued monitoring of the
site to complete abandonment of the site. As a result, environmental conditions at a prematurely closed
mine may be fully reclaimed or equivalent to those under a planned closure, may be severely
contaminated and require extensive remediation, or may fall anywhere between these extremes.
Environmental impacts associated with premature closure may be more significant than impacts
associated with planned closure, as mine facilities may not be at the end condition anticipated in the
closure plan and there may be uncertainty about future re-opening of the mine. For example, PAG waste
rock in our mine scenarios would likely still be on the surface in the event of a premature closure. If the
mine closed because of a drop in commodity price, there would be little economic incentive to incur the
cost of moving or processing millions of metric tons of PAG waste rock, and water treatment systems
might be insufficient to treat the volume of low pH water containing high metal concentrations from this
previously unplanned source. Some method of financial assurance generally is required by state and
federal agencies to ensure closure if a mine company defaults on its responsibility (Box 4-3). To be
effective, financial assurance must be based on accurate estimates of reclamation costs. In the past,
financial assurance often has not been adequate, and taxpayers have been left with substantial cleanup
costs (USEPA 1997). This may be changing, as agencies update bonding requirements to reflect cleanup
costs more accurately, but projecting these costs far into the future is a difficult task.
When a mine re-opens after premature closure, the owners might change the mining plan, implement
different mitigation practices, or negotiate new effluent permits. An example is the Gibraltar copper
mine in British Columbia. The Gibraltar mine began operations permitted as a zero-discharge operation.
However, when it was re-opened under new ownership after having closed prematurely, the new permit
allowed treated water to be discharged to the Fraser River with a 92-m dilution zone for copper and
other metals.
Bristol Bay Assessment 635 January 2014
-------
Chapters Mine Scenarios
6.4 Conceptual Models
The development of conceptual models is a key component of the problem formulation stage of an
ecological risk assessment (USEPA 1998), and in Chapter 2 we introduced the use of conceptual models
as tools to help structure ecological risk assessments. At the outset, we broadly define the scope of this
assessment to be potential effects of a large-scale mine and a transportation corridor on freshwater
habitats, resulting effects on fish, and consequent fish-mediated effects on wildlife and Alaska Native
populations (Section 2.2.1, Figure 2-1). To conduct a risk analysis, this scope needs to be refined and the
specific sources, stressors, and endpoints to be evaluated must be explicitly identified.
In this section, we summarize the specific sources, stressors, and endpoints considered in the
assessment, as informed by the background information on the region, type of development, and
endpoints of interest presented in the preceding chapters, and based on the mine scenarios described in
this chapter. We then integrate these components into conceptual model diagrams that illustrate
hypothesized cause-effect linkages among these sources, stressors, and endpoints.
6.4.1 Sources Evaluated
The two main sources considered in this assessment are the mine and the transportation corridor, each
of which can be subdivided into several components. These components are summarized below, and
discussed in greater detail in Section 6.1.
• The mine infrastructure includes the major mine components (open mine pit, waste rock piles,
TSFs), the groundwater drawdown zone associated with the mine pit, and plant and ancillary
facilities (e.g., water collection and storage facilities, a WWTP, ore-processing facilities, and chemical
storage facilities).
• The transportation corridor comprises a road and four pipelines (one each for product slurry, diesel
fuel, natural gas, and return water) connecting the mine site area to Cook Inlet.
6.4.2 Stressors Evaluated
As discussed above and in Chapter 4, large-scale mining is a complex process that typically involves both
physical alteration of the environment and the release of pollutants. The specific stressors considered
for inclusion in the assessment were identified based on their potential to significantly affect our
primary endpoint of interest—the region's salmon resources—and their relevance to the U.S.
Environmental Protection Agency's (USEPA's) regulatory authority and decision-making context.
Stakeholders also identified potential stressors of concern, which were considered by the assessment
team. These stressors are summarized in Table 6-9 and discussed in detail below. Those stressors that
are analyzed in the assessment or are of particular concern to stakeholders are discussed in the
following subsections.
Bristol Bay Assessment 636 January 2014
-------
Chapter 6
Mine Scenarios
Table 6-9. Stressors considered in the assessment and their relevance to the assessment's primary
endpoint (salmonids) and the U.S. Environmental Protection Agency's regulatory authority.
Stressor
Excavation
Filling
Water diversion
and withdrawal
Water
temperature
Product metal
(copper)
Other metals
PH
Process
chemicals
Nitrogen
Tailings and
other fine
sediment
Diesel fuel
Natural gas
Dust
Noise
Rock slide
Blocked or
perched culvert
Washed out
culvert
Invasive plants
Climate change
Description
Removal of streams and wetlands due to creation
of the mine pit and other excavations.
Filling in of streams and wetlands due to waste
rock piles, tailings impoundments, and roads.
Reduced flow in streams and wetlands due to
removal of water.
Changes in water temperature associated with
discharges of treated water or reduced
groundwater flows.
Copper occurring in the product concentrate,
waste rock, or tailings could enter streams and
wetlands.
Metals other than copper occurring in the product
concentrate, waste rock, or tailings could enter
streams and wetlands.
Oxidation of sulfides could result in acidification
of waste and receiving waters.
Chemicals used in ore processing would occur in
tailings and product concentrate and could spill.
Nitrogen compounds released during blasting
would deposit on the landscape. Nitrates could
also reach groundwater via leachate from waste
rock piles.
Tailings, product concentrate, and other fine
particles could fill streams or wetlands or, at
lower concentrations, could change substrate
texture and abrade fish gills.
Spilled diesel fuel could enter streams and
wetlands.
Leaking natural gas could combust.
Dust from blasting, tailings beaches, and vehicle
traffic could deposit on the landscape and wash
into streams.
Noise from blasting or other activities.
Slides from waste rock piles or roads.
Inhibition offish passage due to malfunctioning
culverts.
Downstream siltation or inhibition offish passage
due to washed out culverts.
Changes in habitat quality due to invasion by
plants carried by road traffic.
Altered risk of mine failures, and changes in
marine and freshwater habitat quality and life
history timing, associated with increased
precipitation and temperature.
Relevance to
Salmonids
Relevant
Relevant
Relevant
Relevant
Relevant
Relevant
Relevant
Relevant
Weakly relevant
Relevant
Relevant
Not relevant
Weakly relevant
Not relevant
Relevant
Relevant
Relevant
Weakly relevant
Indirectly relevant
Relevance to Decision-Making
Directly relevant to Section 404
of the Clean Water Act
Directly relevant to Section 404
of the Clean Water Act
Consequence of excavation and
filling
Consequence of excavation and
filling
Consequence of excavation and
filling
Consequence of excavation and
filling
Consequence of excavation and
filling
Consequence of excavation and
filling
Consequence of excavation and
filling
Directly relevant to Section 404
of the Clean Water Act (if
particles act as fill) and
consequence of excavation and
filling
Necessary for excavation and
filling
Peripheral to excavation and
filling
Consequence of excavation and
filling
Consequence of excavation and
filling
Consequence of excavation and
filling
Consequence of excavation and
filling for a road
Consequence of excavation and
filling for a road
Peripheral to excavation and
filling
Not related to excavation and
filling, but modifies other
consequences of excavation
and filling
Bristol Bay Assessment
6-37
January 2014
-------
Chapters Mine Scenarios
6.4.2.1 Physical Habitat Alteration
Large-scale mining in the Bristol Bay region would necessarily involve the destruction of streams and
wetlands through excavation and filling associated with the mine pit, waste rock piles, TSFs, borrow pits,
and the transportation corridor. This excavation and filling would directly affect anadromous and
resident salmonid habitats and directly involve USEPA under Section 404 of the Clean Water Act.
Mining-related excavation and filling would also result in water diversion and withdrawal. Stream and
overland flow must be diverted around the mine site to keep it dry and minimize erosion; the mine pit
must be dewatered to continue excavation; and water must be obtained for use in ore processing,
tailings and product transport, and other purposes. These diversions and withdrawals would redirect
and reduce flow and plausibly affect fish via reduced habitat quality or quantity.
6.4.2.2 Water Temperature
Stream and wetland water temperatures could be affected by the capture, storage, use, treatment, and
discharge of water throughout the mining process. Elevated temperatures could result from warm water
discharges or, in summer, from reduced groundwater inputs. In winter, reduced groundwater inputs
could result in reduced temperatures. Because water temperature affects fish development and habitat,
any temperature changes could plausibly influence fish populations.
6.4.2.3 Chemical Contaminants
A range of chemical contaminants associated with mining may enter surface waters and pose risks to
fish. These contaminants include rock-derived inorganic contaminants (metals and acidity), ore-
processing chemicals, fuels, and nitrogen compounds.
Rock-Derived Inorganic Contaminants
Mines are developed because rocks at the site have high metal concentrations, which are further
concentrated as ore is isolated from waste rock and as product concentrate is created from the ore.
These metals may enter surface waters from uncollected leachate and runoff, from WWTP discharges, or
from spills of product concentrate and its associated water. Metals are known to cause toxic effects on
aquatic biota, including fish; however, when combined with low pH (acidity), metals become especially
problematic. Acid rock drainage occurs when PAG rocks are present at the mine site. Acidity can be
directly deleterious to aquatic biota, but it also increases the solubility of minerals, which results in
increased concentrations of metals in solution.
Because copper is the major resource metal in the Pebble deposit and is particularly toxic to aquatic
organisms, it is the metal most likely to cause toxic effects at this site. Copper toxicity also has been a
primary concern of stakeholders, including the National Oceanic and Atmospheric Administration, the
federal agency responsible for salmon management. Thus, copper criteria, standards, and toxicity are
considered in detail in the assessment.
Bristol Bay Assessment 638 January 2014
-------
Chapter 6
Mine Scenarios
Other metals are considered if their concentrations in test leachates from the Pebble deposit indicate
that they are potentially toxic, based on benchmark values. When possible, national ambient water
quality criteria are used as screening benchmarks. Both criterion maximum concentrations (CMCs) and
criterion continuous concentrations (CCCs) are used to account for acute and chronic exposures,
respectively. When U.S. criteria are not available, the most similar available value is used (e.g., Canadian
benchmarks, the lowest acute and chronic values from the USEPA's ECOTOX database, or the European
Chemical Agency and Organization for Economic Cooperation and Development's eChemPortal)
(Table 6-10).
Table 6-10. Screening benchmarks for metals with no national ambient water quality criteria.
Metal
B
Ba
Co
Fea
Mn
Mo
Sb
Acute/Chronic Benchmarks
(Mg/L)
29,000/1,500
46,000/8,900
89/2.5
350/-
760/693
32,000/73
14,400/1,600
Source and Notes
Canadian acute and chronic guidelines based on SSDs (CCME 2009)
Austroptamobius pallipes 96-hour LCso (Boutet and Chaisemartin 1973) and
Daphnia magna 21-day reproductive ECso (Biesinger and Christensen 1972)
Acute value is the lowest acute test datum and the chronic value is the 5th centile
of a chronic species sensitivity distribution (Environment Canada and Health
Canada 2011)
Chronic data were inadequate to set a value, but the Canadian authors believed
that it would not be much lower than this acute value (BC 2008)
Hardness adjusted (for 20 mg/L) acute and chronic guidelines (BC 2001)
Daphnia magna 48-hour LCso (Kimball 1978) and Canadian chronic guideline
(CCME 1999)
Lowest acute and chronic values from a fathead minnow early life-stage test
(USEPA 1980, Swedish Chemicals Inspectorate 2008, Environment Canada and
Health Canada 2010)
Notes:
a The listed U.S. iron criterion, from the 1976 Red Book (USEPA 1976), is less reliable than this more recent benchmark.
SSD = species sensitivity distribution; LC5o = median lethal concentration; EC&O = median effective concentration.
Some metals, such as calcium, magnesium, and sodium, are not screened because of their low toxicity.
Molybdenum is treated as a contaminant of concern because it is a specific product of the mine, even
though it would not be retained based on the comparison of test leachates with benchmark values.
Molybdenum concentrate would be trucked to the port, and spills of the sand-like material could occur.
Gold is also a product, but is not evaluated because it has very low solubility and toxicity and would not
be transported in a form likely to result in aqueous exposures.
Screening against tailings and waste rock leachates are presented in Tables 8-4 through 8-8. The metals
of concern are aluminum, cadmium, cobalt, copper, manganese, nickel, lead, selenium, and zinc based on
average concentrations exceeding either acute or chronic benchmarks for at least one leachate.
However, most of the estimated total toxicity is due to copper.
Major Ions (Total Dissolved Solids)
Total dissolved solids (TDS) comprise all organic and inorganic materials dissolved in a water sample,
which can be measured directly or estimated from conductivity measurements (specific conductance is
the term for conductivity values that have been temperature-compensated to 25°C). Mining inevitably
Bristol Bay Assessment
6-39
January 2014
-------
Chapters Mine Scenarios
involves crushing rocks, and the leaching of crushed rock results in enhanced dissolution and elevated
concentrations of dissolved major ions (calcium, magnesium, sodium, potassium, chlorine, sulfate, and
bicarbonate). These major ions generally contribute the most mass to TDS measurements, especially
sulfate in waters influenced by metal mining. Some metals, such as calcium, magnesium, potassium, and
sodium are not screened because of their low toxicity, but they contribute to ionic stress. Thus, even if
this mixture of TDS is not acidic, it can be toxic to aquatic biota, particularly in this region's waters,
which have low ambient concentrations of these ions. Examples of toxicity due to leaching of major ions
from mine-derived waste rock are discussed in USEPA (2011) and Chapman et al. (2000). Also, the
history of TDS compliance problems at the Red Dog Mine near Kotzebue, Alaska, suggests that dissolved
major ions should be a stressor of concern.
Ore-Processing Chemicals
Chemicals used to process the ore and separate product from tailings have the potential to enter the
environment as a result of truck wrecks, on-site spills, tailings slurry spills, product concentrate slurry
spills, or water collection and treatment failures. Tests of the Pebble deposit ore used alkaline flotation
to separate product concentrate from tailings (Ghaffari et al. 2011). The collector was sodium ethyl
xanthate, the frother was methyl isobutyl carbinol, and lime was used to adjust pH. Molybdenum
separation also requires fuel oil as a collector (Box 4-5). Of these, xanthate is clearly a contaminant of
concern because it is highly toxic to aquatic life (Hidalgo and Gutz 2001). Methyl isobutyl carbinol has
been poorly tested but appears to have relatively low toxicity (acute lethality to African clawed frogs
and goldfish requires a relatively high concentration, 360 to 656 mg/L [USEPA 2013]). Lime would
contribute to the risk from major ions (TDS). Fuel oil use for this purpose would be small relative to its
use as fuel.
In addition, cyanide might be used to recover gold from pyritic tailings (Box 4-6). It is expected that a
cyanide destruction unit would be used at the end of the leaching process to achieve the acute and
chronic water quality criteria for free cyanide of 22 and 5.2 ug/L, respectively. Cyanide in the TSF is
likely to be rapidly diluted and degraded. Accidental releases and on-site spills, as recently occurred at
the Fort Knox mine (ADEC 2012), are possible but are not judged to be as directly significant to our
endpoints as other accidents considered. However, because cyanide is assumed to be transported as a
solid, as is common at other mines, truck accidents could result in cyanide spills to streams.
Fuels
Both diesel oil and natural gas would be piped to the mine site and could enter the environment via
pipeline leaks or failures. Diesel spills could enter surface waters and have been known to adversely
affect aquatic biota, so diesel is considered in the assessment. Natural gas could combust, but a natural
gas fire is unlikely to significantly affect salmon populations.
Nitrogen Compounds
Nitrogen compounds, expected to be predominantly nitrate due to combustion, would be released
during the blasting associated with excavation. Some of these compounds would deposit on waste rock
Bristol Bay Assessment 640 January 2014
-------
Chapters Mine Scenarios
piles and the landscape and could enter surface water and groundwater. However, it is likely that these
streams are phosphorus-limited, not nitrogen-limited (Goldman 1960, Moore and Schindler 2004), and
the consequences of an increase in nitrogen/phosphorus ratio for salmonids are unknown but judged to
be minimal. Thus, nitrogen residues are not considered in the assessment.
6.4.2.4 Fine Sediment
If tailings, product concentrate, unpaved road materials, or other fine particles are spilled or eroded,
they could fill streams and wetlands, alter streambed substrates, or abrade the gills offish.
6.4.2.5 Dust
Blasting and vehicle traffic, both at the mine site and along the transportation corridor, would generate
dust. Exposed tailings beaches within the TSFs also could result in dust generation. This dust could
contribute to the sedimentation of streams and, depending on the composition of the rock, could
contribute toxic metals to surface waters. Dust from unpaved roads is known to affect streams, so it is
included in this assessment. In contrast, the occurrence of dust from blasting and tailings beaches is
poorly documented, highly site-specific, and its effects are unknown. We anticipate that much of the dust
generated from blasting and tailings beaches would settle on the site and be collected with runoff water.
Wind may carry dust off site, but would also disperse it across the landscape. We do not judge dust from
blasting or tailings beaches to be an important contributor to risks to salmonids (although this judgment
is uncertain), and do not consider it in the assessment.
6.4.2.6 Noise
Noise would be generated by blasting at the mine site and vehicle traffic along the transportation
corridor. Although noise may directly affect wildlife, it is unlikely to affect salmonids and is not
considered in the assessment.
6.4.2.7 Culverts
Blocked or perched culverts could significantly reduce fish passage, thereby reducing salmon migrations
or movement among habitats by resident salmonids. Culverts also may wash out during floods,
temporarily inhibiting fish movement and reducing habitat due to siltation by the deposited roadbed
material. Culverts are a component of roads that fill wetlands and the floodplains of streams. They may
significantly affect salmon in the surface waters they intersect and thus are considered in the
assessment.
6.4.2.8 Invasive Species
Several dozen species of plants, animals, and micro-organisms are considered to be or have the potential
to be invasive in Alaska (ADF&G 2013, Eddmaps 2013). Of those currently present, reed canarygrass
(Phalaris arundinaced) is widespread on the Kenai Peninsula (HSWCD 2007) and elodea (Elodea
canadensis) exists in Stormy Lake on the northern Kenai Peninsula (Etcheverry 2012). These plants have
the potential to degrade salmon habitat (Merz et al. 2008). The improved and expanded road from Cook
Bristol Bay Assessment 641 January 2014
-------
Chapters Mine Scenarios
Inlet may facilitate the spread of reed canary grass and elodea from the Kenai Peninsula to the Bristol
Bay watershed, where they may adversely affect salmon habitat.
6.4.3 Endpoints Evaluated
In this assessment, the primary endpoint of interest is the region's key salmonid populations (Pacific
salmon, rainbow trout, and Dolly Varden) in terms of abundance, productivity, or diversity. Given the
importance of salmonids to the region's ecosystems and culture, we also consider the effects of potential
changes in fish populations on wildlife abundance, productivity, or diversity and on Alaska Native
culture. These endpoints are discussed in detail in Chapter 5.
6.4.4 Conceptual Model Diagrams
To frame the assessment, we developed conceptual model diagrams illustrating potential pathways
linking the sources, stressors, and endpoints detailed above (see Box 2-1 for an overview of how the
assessment's conceptual models are structured). These diagrams went through several iterations, from
initial brainstorming of all potential pathways associated with large-scale mine development in the
Bristol Bay region (both with the assessment team and other stakeholders) to focusing on those
pathways considered both within the assessment's scope (Chapter 2) and likely to affect endpoints of
interest.
Through this iterative process, we developed a series of three conceptual model diagrams illustrating
hypothesized cause-effect relationships leading from mine-related sources to endpoints of interest.
These diagrams illustrate potential effects of routine mine construction and operation on physical
habitat (Figure 6-12), potential effects of routine mine construction and operation on water chemistry
(Figure 6-13), and potential effects of unplanned events on physical habitat and water chemistry (Figure
6-14). Note that the distinction between physical habitat and water chemistry was made for
presentation purposes, though we recognize that water chemistry can be an important component of
the physical habitat. These diagrams provide a framework for the analysis sections of the assessment,
and the relevant portions of these diagrams evaluated in each analysis section are highlighted
throughout the remaining chapters of the assessment. Note that not all pathways included in each
conceptual diagram are necessarily evaluated in the assessment. For example, in some cases, we
hypothesized pathways that may be significant, but data were not sufficient for quantitative analysis.
We also developed three more general conceptual model diagrams for specific topics (wildlife, Alaska
Native cultures, and cumulative effects of multiple mines) that were defined as outside of the
assessment's scope but that are of key importance to stakeholders (Chapters 12 and 13).
Bristol Bay Assessment 642 January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-12. Conceptual model illustrating potential effects of routine mine construction and operation on physical habitat.
transportation corrid
wasterock If tamngsstorage 1 ' oreprocessmg
piles ) k facilities J I facilities
water collection & I -^»f water treatment
Agroundwater -
surface water interactions
Tsuspended
sediment
A stream
eeomorphology
4- floodplain
connectivity
A downstream water flows
a downstream walci temperatures
Amagnitude &
frequency of high flows
4> production and export of
food & other resources
4- macroiiivertebrale
"I" alteration of channel
morphology & floodpiain
connectivity
4> production &
export of food &
other resources
T physiological
stress
T competition
& predation
T inhibition of
fish passage
A riparian
vegetation
rearing habitat
(quality or quantity)
spawning habitat
[quality or quantity]
overwintering habitat
(quality or quantity}
incubation habitat
(quality or quantity)
salmon
(abundance, productivity or diversity)
4< other fishes
(abundance, productivity or diversity)
•i- marine-derived
nutrients
4, ecosystem
productivity
additional step in
causal pathway
additional step in
can sal pathway
Within a shape, '[" indicates an increase in the parameter,
>[,indicates a decrease in a parameter, and A indicates a
change in the parameter.
Arrows leading from one shape to another indicate a
hypothesized cause-effect relationship.
Shapes bracketed under another shape ate specific
components of the more general shape under which they
appear.
Bristol Bay Assessment
6-43
January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-13. Conceptual model illustrating potential effects of routine mine construction and operation on water chemistry.
ore processing
facilities
/
\
\
/
: water treatment j
facilities J
\
t \
\
/
/^
f chemical storage
I facilities
/
^ water
withdrawals
I transportation corridor
water collection &
water
discharges
adsorbed or precipitated metals
A metal speciation
& bioavailability
Aother water
chemistry parameters
^tissue
concentration
chronic
toxicity
bioaccumulation&
biomagnification
additional step in
can sal pathway
additional step in
can sal path way
Within a shape, ^ indicates an increase in the parameter,
1- indicates a decrease in a parameter, and A indicates a
change in the parameter.
Arrows leading from one shape to another indicate a
hypothesized cause-effect relationship.
Shapes bracketed under another shape are specific
components of the more general shape under which they
appeal.
>!• salmon
(abundance, productivity or diversity)
other fishes
(abundance, productivity or diversity)
A
marine-derived
nutrients
•^ ecosystem
productivity
mitigation
i
Bristol Bay Assessment
6-44
January 2014
-------
Chapter 6
Mine Scenarios
Figure 6-14. Conceptual model illustrating potential effects of unplanned events on physical habitat and water chemistry.
transportation corridor
| return water | f slurry
I pipeline J I pipeline
\ / \ /
tailings storage |
facilities J
/
am |
singsJ
\
/ \
/
V
V
ore processing
facilities
| water collection & |-
l storage systems J
/\
V
Vvater treatment I
storage facilities
\/
\/
failure of water collection
& storage systems
chemical storage
facilities
V
V
failure of water
treatment facilities
chemical
spill
\/
chemical or fuel spill
during transport
pipeline breaks
or leaks
tailings dam
fa i I LI re
contaminated water
discharges
culvert blockage
or perching
channel erosion
& entrenchment
1s dissolved metal;
1" adsorbed or precipitated metal:
A metal speciatior
& bioavailability
quality parameter!
physiologica
stress
T inhibition of
fish passage
macroinvertebratc
prey
Tmacroinvertebrate
toxicity
J/ rearing habitat
•I, spawning habitat
(quality orquantity)
•i- overwintering habitat
incubation habitat
T bioaccumulation&
biomagnification
T tissue
concentration
(quality or quantity)
[quality or quantity)
(quality or quantity
additional step in
causal pathway
additional step in
causal pathway
Within a shape, "I* indicates an increase in the parameter,
•^indicates a decrease in a parameter, and A indicates a
change in the parameter.
Arrows leading from one shape to another indicate a
hypothesized cause-effect r elationship.
Shapes bracketed under another shape are specific
componentsof the more general shape under which they
appear.
•i- salmon
(abundance, productivity or diversity)
V
•^ other fishes
(abundance, productivity or diversity)
•i- marine-derived
nutrients
s|/ ecosystem
productivity
stage of mining
operations
seasonal timing of
unplanned event
Bristol Bay Assessment
6-45
January 2014
-------
This chapter addresses the stream habitat and streamflow risks associated with routine operations of
the mine scenarios described in Chapter 6. It considers the unavoidable environmental effects
associated with the footprint of each mine scenario, in the absence of failures of water collection or
treatment facilities, tailings storage facilities (TSFs), the transportation corridor, or pipelines. This is not
meant to suggest that the absence of failures is a realistic possibility, because accidents and failures do
happen in complex and long-lasting operations. The risks and potential impacts of these failures are
described in Chapters 8, 9,10, and 11. In this chapter we evaluate the inevitable effects of the mine
scenarios, rather than those that are the result of accidents and failures.
Potential pathways linking mine components, stream habitat and streamflow alterations, and biotic
responses are highlighted in Figure 7-1. Key stressors associated with routine mine development and
operation include elimination and modification of habitat (Section 7.2) and changes in downstream
streamflow (Section 7.3), both of which can affect fish populations. The pathways associated with
stream and wetland elimination highlighted in Figure 7-1 primarily reflect linkages occurring within the
spatial extent of the mine footprint (Scale 4). Linkages and effects associated with streamflow
alterations primarily operate from the edge of the footprint downstream to the extent of detectable
streamflow changes (Scale 3). Effects on fish populations due to these modifications could extend
beyond these geographic scales and into the larger Nushagak and Kvichak River watersheds (Scale 2),
depending on the types and severity of impacts; these effects could not be quantified and are discussed
qualitatively (see also Chapter 14). Routine effects of water collection, treatment, and discharge and the
transportation corridor are discussed in Chapters 8 and 10, respectively.
Bristol Bay Assessment
7-1
January 2014
-------
Chapter 7 Mine Footprint
7.1 Abundance and Distribution of Fishes in the Mine
Scenario Watersheds
Potential effects of the mine footprint (addressed in this chapter) and of routine mine operations and
failures (addressed in Chapters 8 through 11) on the assessment endpoints depend on the abundance
and distribution of salmonids in the streams and rivers of the three watersheds draining the Pebble
deposit area: the South Fork Koktuli River, North Fork Koktuli River, and Upper Talarik Creek
watersheds (hereafter referred to as the mine scenario watersheds).
7.1.1 Fish Distribution
The mine scenario watersheds have been sampled extensively for summer fish distributions over
several years. These data, collected by the Alaska Department of Fish and Game (ADF&G) and various
consultants and non-profits, are captured in the Catalog of Waters Important for Spawning, Rearing, or
Migration ofAnadromous Fishes—Southwestern Region (also known as the Anadromous Waters Catalog
[AWC]) (Johnson and Blanche 2012) and the Alaska Freshwater Fish Inventory (AFFI) (ADF&G 2012).
The AWC is the State of Alaska's official record of anadromous fish distributions and, if available, the life
stages present (categorized as spawning, rearing, or present but life stage unspecified) in individual
stream reaches. The AFFI includes all fish species, including resident fishes, found at specific sampling
points. The catalogued distributions of the five Pacific salmon species (sockeye, coho, Chinook, chum,
and pink), Dolly Varden (both anadromous and non-anadromous forms are present), and resident
rainbow trout in the mine scenario watersheds are shown in Figures 7-2 through 7-8. In addition,
Alaskan or Arctic brook lamprey, longnose sucker, northern pike, humpback whitefish, least cisco, round
whitefish, Arctic char, Arctic grayling, burbot, threespine stickleback, ninespine stickleback, and slimy
sculpin occur in these watersheds (ADF&G 2012). Details of these species, including information on
distributions, abundances, habitats, life cycles, predator-prey relationships, and harvests, are provided
in Appendix B. AWC stream reach designations and AFFI observation points should be interpreted with
care, because not all streams could be sampled and there are potential errors associated with fish
identification and mapping. Additional caveats and uncertainties concerning interpretation of AWC and
AFFI data are discussed in Section 7.2.5.
Bristol Bay Assessment -, ~ January 2014
-------
Chapter 7
Mine Footprint
Figure 7-1. Conceptual model illustrating potential linkages between sources associated with the mine scenario footprints, changes in physical habitat, and fish endpoints.
Agroundwater-
surface water interactions
\/
\/
A downstream water flows
A downstream water temperatures
A magnitude &
frequency of high flows
\/
T" alteration of channel
morphology & floodplain
connectivity
t flow
intermittency
V
1s aquatic habitat
fragmentation
A flow timing
& seasonally
A other flow
parameters
\/
A migration
patterns
V
v
•!• rearing habitat
(quality or quantity)
V
sb spawning habitat
(quality orquantity)
xb overwintering habitat
(quality orquantity)
production and export of
food& other resources
climate
change
xb incubation habitat
(quality orquantity)
V
^ salmon
(abundance, productivity or diversity)
4- other fishes
(abundance, productivity or diversity)
LEGEND
additional step in
causal pathway
additional step in
can sal path way
Within a shape, 'f indicates an increase in the parameter,
•I'indicates a decrease in a parameter, and A indicates a
change in the parameter.
Arrows leading from one shape to another indicate a
hypothesized cause-effect relationship.
Shapes bracketed under another shape are specific
components of the more general shape under which they
appear.
Bristol Bay Assessment
7-3
January 2014
-------
This page intentionally left blank
-------
Chapter 7
Mine Footprint
Figure 7-2. Reported sockeye salmon distribution in the mine scenario watersheds. "Present"
indicates species was present but life-stage use was not determined; "spawning" indicates spawning
adults were observed; "rearing" indicates juveniles were observed. Present, spawning, and rearing
designations are based on the Anadromous Waters Catalog (Johnson and Blanche 2012). Life-stage-
specific reach designations are likely underestimates, given the challenges inherent in surveying all
streams that may support life-stage use throughout the year (see Section7.2.5 for additional notes on
interpretation of fish distribution data). Footprints of the major mine components for the three mine
scenarios and the drawdown zone for the Pebble 6.5 scenario are shown for reference.
•^v,
V
v^r
Ji
NORTH FORK KOKTULT
UPPER TALARIK
NUSHAGAK
i j m
.,
<
FORK KOKTULI
• '
"it
1 1
Present
Spawning
Rearing
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Pebble 6.5 Drawdown Zone
Mine Scenario Watersheds
Watershed Boundary
Iliamna Lake
NUSHAGAK
i KVICHAKO-
0 2.5 5
0 2.5
] Kilometers
]Miles
Bristol Bay Assessment
7-5
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-3. Reported coho salmon distribution in the mine scenario watersheds. "Present" indicates
species was present but life-stage use was not determined; "spawning" indicates spawning adults
were observed; "rearing" indicates juveniles were observed. Present, spawning, and rearing
designations are based on the Anadromous Waters Catalog (Johnson and Blanche 2012). Life-stage-
specific reach designations are likely underestimates, given the challenges inherent in surveying all
streams that may support life-stage use throughout the year (see Section 7.2.5 for additional notes
on interpretation of fish distribution data). Footprints of the major mine components for the three
mine scenarios and the drawdown zone for the Pebble 6.5 scenario are shown for reference.
Present
Spawning
Rearing
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Pebble 6.5 Drawdown Zone
Mine Scenario Watersheds
Watershed Boundary
Bristol Bay Assessment
7-6
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-4. Reported Chinook salmon distribution in the mine scenario watersheds. "Present"
indicates species was present but life-stage use was not determined; "spawning" indicates spawning
adults were observed; "rearing" indicates juveniles were observed. Present, spawning, and rearing
designations are based on the Anadromous Waters Catalog (Johnson and Blanche 2012). Life-stage-
specific reach designations are likely underestimates, given the challenges inherent in surveying all
streams that may support life-stage use throughout the year (see Section 7.2.5 for additional notes
on interpretation of fish distribution data). Footprints of the major mine components for the three
mine scenarios and the drawdown zone for the Pebble 6.5 scenario are shown for reference.
Present
Spawning
Rearing
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Pebble 6.5 Drawdown Zone
Mine Scenario Watersheds
Watershed Boundary
Bristol Bay Assessment
7-7
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-5. Reported chum salmon distribution in the mine scenario watersheds. "Present"
indicates species was present but life-stage use was not determined; "spawning" indicates spawning
adults were observed; "rearing" indicates juveniles were observed. Present, spawning, and rearing
designations are based on the Anadromous Waters Catalog (Johnson and Blanche 2012). Life-stage-
specific reach designations are likely underestimates, given the challenges inherent in surveying all
streams that may support life-stage use throughout the year (see Section 7.2.5 for additional notes
on interpretation of fish distribution data). Footprints of the major mine components for the three
mine scenarios and the drawdown zone for the Pebble 6.5 scenario are shown for reference.
Present
Spawning
Rearing
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Pebble 6.5 Drawdown Zone
Mine Scenario Watersheds
Watershed Boundary
Bristol Bay Assessment
7-8
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-6. Reported pink salmon distribution in the mine scenario watersheds. "Present" indicates
species was present but life-stage use was not determined; "spawning" indicates spawning adults
were observed. Present and spawning designations are based on the Anadromous Waters Catalog
(Johnson and Blanche 2012). Life-stage-specific reach designations are likely underestimates, given
the challenges inherent in surveying all streams that may support life-stage use throughout the year
(see Section 7.2.5 for additional notes on interpretation of fish distribution data). Footprints of the
major mine components for the three mine scenarios and the drawdown zone for the Pebble 6.5
scenario are shown for reference.
Present
Spawning
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Pebble 6.5 Drawdown Zone
Mine Scenario Watersheds
I I Watershed Boundary
Bristol Bay Assessment
7-9
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-7. Reported Dolly Varden occurrence in the mine scenario watersheds. Designation of
species presence is based on the Alaska Freshwater Fish Inventory (ADF&G 2012). Absence cannot
be inferred from this map (see Section 7.2.5 for additional notes on interpretation of fish distribution
data). Footprints of the major mine components for the three mine scenarios and the drawdown zone
for the Pebble 6.5 scenario are shown for reference.
NORTH FORK KOKTULI
Present
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Pebble 6.5 Drawdown Zone
Mine Scenario Watersheds
Watershed Boundary
Bristol Bay Assessment
7-10
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-8. Reported rainbow trout occurrence in the mine scenario watersheds. Designation of
species presence is based on the Alaska Freshwater Fish Inventory (ADF&G 2012). Absence cannot
be inferred from this map (see Section 7.2.5 for additional notes on interpretation of fish distribution
data). Footprints of the major mine components for the three mine scenarios and the drawdown zone
for the Pebble 6.5 scenario are shown for reference.
Present
Pebble 0.25 Components
Pebble 2.0 Components
Pebble 6.5 Components
Pebble 6.5 Drawdown Zone
Mine Scenario Watersheds
Watershed Boundary
Bristol Bay Assessment
7-11
January 2014
-------
Chapter 7 Mine Footprint
Sockeye salmon use mainstem reaches of the mine scenario watersheds for spawning and rearing,
including a portion of Upper Talarik Creek that is within the waste rock piles of the Pebble 2.0 and
Pebble 6.5 scenarios (Figure 7-2). Coho salmon have the most widespread distribution of the five
salmon species in the mine scenario watersheds, and make extensive use of mainstem and tributary
habitats (Figure 7-3). Coho spawn and rear in headwater streams that would be eliminated, blocked, or
dewatered by the mine pits, waste rock piles, and TSFs of the Pebble 0.25, 2.0, and 6.5 scenarios (Figure
7-3). Chinook salmon have been documented throughout mainstem reaches of the mine scenario
watersheds (Figure 7-4). Chinook are known to use small streams for rearing habitat, and juveniles have
been observed in streams that would be within the footprints of TSF 1 (North Fork Koktuli River), TSF 2
(South Fork Koktuli River), and the waste rock piles and mine pits (Upper Talarik Creek) (Figure 7-4).
The distributions of chum and pink salmon are generally restricted to mainstem reaches where
spawning and migration occur. Chum salmon have been found in all three mine scenario watersheds and
in a stream within the footprint of TSF 2 (Figure 7-5). Pink salmon have only been reported at very low
numbers in the lowest section of Upper Talarik Creek and in the Koktuli River below the confluence of
the north and south forks (Figure 7-6, Figure 5-8). Dolly Varden are found throughout the mine scenario
watersheds, and fish surveys indicate that they are commonly found in the smallest streams (i.e., first-
order tributaries), including streams within the footprints of each of the TSFs (Figure 7-7). Rainbow
trout have been collected at many mainstem locations, especially in Upper Talarik Creek, and their
reported distribution extends upstream throughout the TSF 1 area and in the portions of Upper Talarik
Creek within the waste rock pile footprints (Figure 7-8).
7.1.2 Spawning Salmon Abundance
Index estimates of relative spawning salmon abundance are available for sockeye, coho, Chinook, and
chum salmon in the mine scenario watersheds. Aerial index counts of spawning salmon are available
from ADF&G and the Pebble Limited Partnership (PLP). This type of survey is used primarily as an index
to track variation in run size over time. We recognize that survey values tend to underestimate true
abundance for several reasons. An observer in an aircraft is not able to count all fish in dense
aggregations or those concealed under overhanging vegetation or undercut banks, and only a fraction of
the fish that spawn at a given site are present at any one time (Bue et al. 1988, Jones et al. 2007).
Weather, water clarity, and other factors that influence fish visibility can also contribute to
underestimates. In addition, surveys intended to capture peak abundance may not always do so. For
example, aerial surveys counted, on average, only 44% of the pink salmon counted by surveyors walking
the same Prince William Sound spawning streams (Bue et al. 1988). Peak aerial counts of pink salmon in
southeastern Alaska are routinely multiplied by 2.5 to represent more accurately the number offish
present at the survey time (Jones et al. 2007). Helicopter surveys of Chinook salmon on the Kenai
Peninsula's Anchor River over 5 years counted only 5 to 10% of the fish counted by a concurrent
sonar/weir counting station (Szarzi et al. 2007).
ADF&G conducts aerial index counts that target peak sockeye salmon spawning periods on Upper
Talarik Creek and peak Chinook salmon spawning periods on the Koktuli River system. Sockeye salmon
counts have been conducted in most years since 1955 (Morstad 2003), and Chinook salmon counts in
Bristol Bay Assessment 712 January 2014
-------
Chapter 7 Mine Footprint
most years since 1967 (Dye and Schwanke 2009). Between 1955 and 2011, sockeye salmon counts in
Upper Talarik Creek ranged from 0 to 70,600, with an average of 7,021 over 49 count periods (Morstad
pers. comm.) Between 1967 and 2009, Chinook salmon counts in the Koktuli River system ranged from
240 to 10,620, with an average of 3,828 over 29 count periods (Dye and Schwanke 2009).
PLP (2011) provides aerial index counts for Chinook, chum, coho, and sockeye salmon adults in the mine
scenario watersheds from 2004 to 2008. Surveys on the South and North Fork Koktuli Rivers began at
the confluence and extended upward to the intermittent reach or Frying Pan Lake on the South Fork
Koktuli River and upward to Big Wiggly Lake or river kilometer 56 on the North Fork Koktuli River.
Surveys on Upper Talarik Creek ran from the mouth and extended upstream to Tributary 1.350 (just
east of Koktuli Mountain) or to the headwaters. Multiple counts were usually made for each stream and
species in a given year (Table 7-1). Repeat surveys of this type can be used to estimate the size of
spawning populations using an area under the curve (AUC) approach if estimates of stream life (i.e., the
number of days that salmon are present on the spawning grounds) and observer efficiency are available
(Hilborn et al. 1999). However, PLP was unable to make reliable estimates of stream life and observer
efficiency (PLP 2011: 15.1-14), a common shortcoming given the data-intensive demands of AUC
estimates (Holt and Cox 2008). Mean index counts can be reliable indicators of spawning coho salmon
abundance trends in simulation studies (e.g., Holt and Cox 2008), but optimum reliability is contingent
on sampling date and frequency. Peak index counts have been used to monitor trends in spawner
abundance, but these counts also have shortcomings associated with survey design and execution and
require area- and species-specific expansion factors to allow escapement estimates (e.g., Parken et al.
2003). In addition, trend analysis needs to account for the high interannual variability in escapement
estimates noted above, and likely requires many years of data. Streams or river segments lacking long-
term survey data require a larger watershed and population context to approximate baseline conditions
for those locations and populations.
Table 7-1 reports the highest of each year's index counts for each population, approximated from figures
in PLP (2011: Chapter 15). We report peak index counts because only a portion of the spawning
population is present on the spawning grounds on any given day. Thus, the highest index count is
mathematically closer to the true abundance than is the average of multiple surveys, and it more closely
matches ADF&G's index methods based on a single count that targets peak spawning. The highest peak
index counts for coho and sockeye salmon were in Upper Talarik Creek, whereas the highest counts for
Chinook and chum salmon were in the South and North Fork Koktuli Rivers (Table 7-1). The overall
highest count was for sockeye salmon in Upper Talarik Creek and Tributary 1.60 in 2008, when
approximately 82,000 fish were estimated (Table 7-1).
Bristol Bay Assessment 713 January 2014
-------
Chapter 7
Mine Footprint
Table 7-1. Highest reported index spawner counts in the mine scenario watersheds for each year,
2004 to 2008.
Mine Scenario Watershed
South Fork Koktuli River
North Fork Koktuli River
Upper Talarik Creek
Salmon
Species
Chinook
Chum
Coho
Sockeye
Chinook
Chum
Coho
Sockeye
Chinook
Chum
Coho
Sockeye
Highest Index Spawner Count Per Year
(Number Of Counts)3
2004
2,750 (3)
(0)
250 (2)
1,400 (2)
2,800 (3)
400 (1)
300 (3)
550 (2)
275 (2)
(0)
3,000 (4)
33,000 (2)
2005
1,500 (4)
350 (4)
550 (4)
2,000 (5)
2,900 (4)
350 (4)
350 (1)
1,100(5)
100 (3)
3(1)
(0)
15,000 (4)
2006
250 (5)
850 (7)
1,375 (3)
2,700 (8)
750 (4)
750 (4)
1,050 (4)
1,400 (7)
80(3)
13(2)
6,300 (3)
10,000 (6)
2007
300 (8)
200 (11)
250 (10)
4,000 (11)
600 (8)
800 (9)
125 (8)
2,200 (10)
150 (9)
8(8)
4,400 (9)
10,000 (14)
2008
500 (9)
950 (7)
1,875 (20)
6,000 (13)
500 (8)
1,400 (7)
1,700 (15)
2,000 (12)
100 (8)
18(5)
6,300 (14)b
82,000 (14)b
Notes:
a Values likely underestimate true spawner abundance.
b Tributary 1.60, a major tributary to Upper Talarik Creek, was included in this count.
Source: PLP 2011.
The spatial distribution of spawner counts in the study streams during the 2008 return year was
provided by PLP (2011). Spawner counts were summarized by individual stream reaches throughout
the mainstem of each of the mine scenario watersheds. Data were reported for three reaches in the
South Fork Koktuli River (A through C, extending from the confluence upstream to the intermittent
reach), five reaches in the North Fork Koktuli River (A through E, extending from the confluence
upstream to beyond Big Wiggly Lake), and seven reaches in Upper Talarik Creek (A through G, extending
from the mouth to the headwaters) (Figure 15.1-2 in PLP [2011] illustrates the stream reaches; Table 7-
2 provides river kilometer boundaries for each reach). Count data (approximated from figures in PLP
[2011]) and location (in river kilometers) for each of these reaches are shown in Table 7-2 to
demonstrate the relative spatial distribution of salmon during the 2008 spawning period.
Bristol Bay Assessment
7-14
January 2014
-------
Chapter 7
Mine Footprint
Table 7-2. Average 2008 index spawner counts by stream reach3.
Stream
South Fork
Koktuli
River
North Fork
Koktuli
River
Upper
Talarik
Creek
Salmon
Species
Reach
Boundaries
(river km)
Chinook
Chum
Coho
Sockeye
Reach
Boundaries
(river km)
Chinook
Chum
Coho
Sockeye
Reach
Boundaries
(river km)
Chinook
Chum
Coho
Sockeye
Stream Reach, Downstream to Upstream
A
0-24.9
200
90
200
800
0-13.7
110
50
100
530
0-5.9
<10
<10
100
10,000
B
24.9-34.3
70
190
250
1,510
13.7-21.1
40
50
70
<10
5.9-16.8
<10
<10
50
4,500
C
34.3-51.7
0
0
8
1
21.1-36.6
50
320
210
220
16.8-24.8
20
<10
40
3,000
D
-
-
36.6-48.4
0
0
30
60
24.8-36.3
<10
<10
180
3,000
E
-
-
48.4-52.5
0
0
60
0
36.3-45.1
20
<10
280
500
F
-
-
-
-
45.1-59.1
<10
10
180
47
G
-
-
-
-
59.1-62.4
0
0
<10
0
Notes:
Dashes (-) indicate no applicable stream reach.
8 Values likely underestimate true spawner abundance.
Source: PLP 2011.
7.1.3 Juvenile Salmon and Other Salmonid Abundance
PLP (2011) reports counts of juvenile salmon and other salmonids in the South and North Fork Koktuli
Rivers and Upper Talarik Creek based on extensive sampling efforts from 2004 through 2008. Snorkel
surveys were the primary data collection method, but electrofishing, minnow traps, beach seines, gill
nets, angling, and dip netting were used in certain situations. It is not always possible to determine
which survey methods generated which counts in PLP (2011). Raw field counts were frequently
expressed as densities (count per 100-m reach was the only unit reported for all three streams). These
counts should not be viewed as quantitative abundance estimates. They are very likely underestimates
because of the extreme difficulty of observing or capturing all fish in complex habitats (Hillman et al.
1992). Density estimates with confidence bounds (e.g., mark-recapture or depletion estimates) were
generated for some parts of the PLP (2011) studies (e.g., PLP 2011: Appendix 15.ID), but such efforts
were uncommon as they are much more time-consuming and labor-intensive.
Reported fish densities summarized over the 5-year period vary widely by stream, sample reach, and
habitat type (PLP 2011: Figures 15.1-23,15.1-52, and 15.1-82). Species that attain densities of several
hundred per 100-m reach in one setting were often absent or sparse in other habitat types or reaches in
the same stream, which is typical for fish in heterogeneous stream environments. Table 7-3 presents
Bristol Bay Assessment
7-15
January 2014
-------
Chapter 7
Mine Footprint
maximum fish densities in the mainstem of each mine scenario watershed, approximated from figures in
PLP (2011), for species that rear for extended periods in the surveyed streams and for which data are
available: Chinook and coho salmon, Arctic grayling, and Dolly Varden. We report maximum density to
give a sense of the magnitude attained in the surveyed streams, but it should be stressed that abundance
varied widely by stream reach and habitat type within a given stream (PLP 2011: Figures 15.1-23,15.1-
52, and 15.1-82). Highest reported densities were approximately 2,500 Arctic grayling and 1,600 coho
salmon per 100 m from adjacent reaches on Upper Talarik Creek, and 1,400 coho salmon per 100 m
from a reach on the North Fork Koktuli River.
Table 7-3. Highest index counts of selected stream-rearing fish species from mainstem habitats of
the mine scenario watersheds.
Highest Reported Density (count per 100 m)a
Stream
South Fork Koktuli River
North Fork Koktuli River
Upper Talarik Creek
Chinook
Salmon
450
500
400
Coho
Salmon
600
1,400
1,600
Arctic
Grayling
275
40
2,500
Dolly
Varden
55
40
10
Source
Figure 15. 1-52 (PLP 2011)
Figure 15. 1-23 (PLP 2011)
Figure 15. 1-82 (PLP 2011)
Notes:
a Values were approximated from figures listed in the source column.
7.2 Habitat Modification
The footprints of the major mine components (mine pit, waste rock piles, and TSFs) would directly
modify the amount of habitat available to salmon, rainbow trout, and Dolly Varden by eliminating
headwater streams and wetlands within and up-gradient of their footprints. Potential effects of this
habitat modification are described for the three mine scenarios in Sections 7.2.2, 7.2.3, and 7.2.4, and
uncertainties and assumptions are described in Section 7.2.5.
7.2.1 Stream Segment Characteristics in the Mine Scenario Watersheds
The mine scenario watersheds encompass an area of 925 km2 and contain 930 km of stream channels
mapped for this analysis (methods described in Section 3.4). In this section, we summarize stream
segment characteristics in the mine scenario watersheds to better characterize stream environments in
and downstream of the mine footprints. In Section 7.2.2, we summarize the characteristics of stream
segments that would be lost to the footprints of the major mine components themselves. Stream
segments for the entire Nushagak and Kvichak River watersheds (Scale 2) are characterized in Chapter
3. This characterization is provided to help readers understand variation in the relative size (mean
annual streamflow), channel gradient, and floodplain potential (proportion of flatland in lowland)
among stream segments in the mine scenario watersheds. Because these characteristics can strongly
influence the quality and suitability of stream habitats as fish habitat, they provide a way to evaluate the
coarse-scale characteristics of streams at risk of impacts at various scales. This characterization helps
highlight the fact that not all stream kilometers in these watersheds are equal in their potential to
support salmon carrying capacity or productivity.
Bristol Bay Assessment
7-16
January 2014
-------
Chapter 7
Mine Footprint
Results from this analysis are presented in tables that summarize the proportion of stream channel
length within each stream size, gradient, and floodplain potential category. To allow direct visual
comparison of the distribution of stream characteristics across scales, we present cumulative frequency
plots (e.g., Figure 7-9). These plots show a frequency curve for each attribute at different geographic
scales. Attributes are grouped into meaningful classes (Chapter 3), denoted by the vertical red
classification bars. For example, the lowest gradient streams are classified as having gradients of less
than 1% (Table 7-4), as shown by the vertical classification bar at 1% in Figure 7-9B. Cumulative
frequency plots can be interpreted by evaluating the height at which the frequency curve is intersected
by the red vertical classification bar. In Figure 7-9B, the 1% gradient classification bar intersects the
Scale 3 frequency curve (solid black line) at a cumulative frequency value of approximately 50%. Thus,
approximately 50% of the stream kilometers in the mine scenario watersheds (Scale 3) have less than
1% gradient. In comparison, approximately 64% of the stream kilometers in the Nushagak and Kvichak
River watersheds (Scale 2) have less than 1% gradient.
Table 7-4. Distribution of stream channel length classified by channel size (based on mean annual
streamflow in m3/s), channel gradient (%), and floodplain potential (based on % flatland in lowland)
for streams and rivers in the mine scenario watersheds. Gray shading indicates values greater than
5%; bold indicates values greater than 10%.
Channel Size
Small headwater streams3
Medium streams'5
Small rivers0
Large riversd
Gradient
<1%
FP
15%
14%
8%
0%
Notes:
a 0-0.15 m3/s; most tributaries in the mine fc
b 0.15-2.8 m3/s; upper reaches and larger tr
c 2.8-28 m3/s; middle to lower portions of thf
d >28 m3/s; the Mulchatna River below the Kc
FP = high floodplain potential (>5% flatland in 1
additional explanation).
NFP
5%
6%
2%
0%
>1% and <3%
FP
5%
0%
0%
0%
NFP
28%
3%
1%
0%
>3% and <8%
FP
0%
0%
0%
0%
NFP
12%
1%
0%
0%
>8%
FP
0%
0%
0%
0%
NFP
0%
0%
0%
0%
otprints.
butaries of the South and North Fork Koktuli Rivers and Upper Talarik Creek.
5 South and North Fork Koktuli Rivers and Upper Talarik Creek, including mainstem Koktuli River.
jktuli confluence, the Newhalen River, and other large rivers.
Dwland); NFP = no or low floodplain potential (<5% flatland in lowland) (see Chapter 3 for
Similar to the larger Nushagak and Kvichak River watersheds (summarized in Table 3-3), streams in the
mine scenario watersheds are generally low-gradient, with extensive flat floodplains or terraces in the
larger valleys (Figure 7-9; also see PLP 2011: Chapter 15 and Appendix 15.IB). There are no large rivers
(greater than 28 m3/s mean annual streamflow) in the mine scenario watersheds (Table 7-4). Compared
to the larger Nushagak and Kvichak River watersheds, streams in the mine scenario watersheds have
fewer very low gradient streams (mean gradient 0.7% versus 0.4%) and a higher proportion (58%
versus 39%) of stream length flowing through valleys with low floodplain potential (i.e., less than 5% of
flatland in lowland) (Table 7-4, Figure 7-9).
Bristol Bay Assessment
7-17
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-9. Cumulative frequency of stream channel length classified by (A) mean annual streamflow
(MAP) (m3/s), (B) channel gradient (%), and (C) floodplain potential (based on % flatland in lowland)
for the mine scenario watersheds (Scale 3) versus the Nushagak and Kvichak River watersheds
(Scale 2). See Section 3.4 for further explanation of MAP, gradient, and floodplain potential
classifications.
100%
s oi
JS —i
3 C
£ TO 40%
u Jj
to
0%
•ScaleB
•MAP Classification
Scale 2
10 20 30 40
Mean Annual Streamflow (m3/s)
50
B
M
100%
80%
60%
40%
<-> i:
to
£U/0
n%
'
0% 4%
•ScaleS
•Gradient Classification
Scale 2
8% 12%
Channel Gradient
16%
20%
100%
0%
20%
40% 60%
Flatland in Lowland
80%
100%
Bristol Bay Assessment
7-18
January 2014
-------
Chapter 7 Mine Footprint
Broadly classified, streams and rivers in the Nushagak and Kvichak River watersheds that are likely to
provide high capacity and quality habitats for salmonids include streams with gradients less than 3%
and of medium stream size (0.15 to 2.8 m3/s mean annual streamflow) or greater. Such streams and
rivers account for 36% of the stream network in the larger Nushagak and Kvichak River watersheds
(Table 3-3), and account for 34% of the stream network in the mine scenario watersheds (Table 7-4).
Smaller, steeper streams provide seasonal (and some year-round) habitat, and provide important
provisioning services to downstream waters (Section 7.2.3). Although streams in the mine scenario
watersheds are smaller and slightly steeper than streams and rivers throughout the entire Nushagak
and Kvichak River watersheds, these results show the high proportion of stream channels in these
basins with the broad geomorphic and hydrologic characteristics that support stream and river habitats
highly suitable for fish species such as Pacific salmon, Dolly Varden, and rainbow trout.
7.2.2 Exposure: Habitat Lost to the Mine Scenario Footprints
For each mine scenario, the total mine footprint consists of the area devoted to the major mine
components (mine pit, waste rock piles, and TSFs), the groundwater drawdown zone associated with
the mine pit, and plant and ancillary facilities (e.g., ore-processing facilities and water collection and
treatment facilities) (see Chapter 6 for additional details on each mine scenario). Portions of the mine
scenario watersheds would be affected by mining activity in this footprint. Stream and wetland habitats
would be lost within and upstream of the footprint (Figures 7-10 through 7-12), and downstream
habitat would be degraded by the loss of the headwater streams and wetlands. Streams under or
upstream of each mine footprint would be inaccessible by fish from downstream reaches because of the
following factors.
• Elimination of streams and wetlands within the mine footprints, either due to removal (e.g.,
excavation of streams or wetlands in the mine pit area) or burial under a TSF or waste rock pile.
• Dewatering by capture into a groundwater drawdown zone associated with the pit. This effect is
distinct from the effect of water removal and capture on streamflows downstream of the mine
footprint, which is covered in Section 7.3.
• Blockage due to either of the above or channel diversion in a manner that prevents fish passage
(e.g., via pipes or conveyances too steep for fish passage).
Streams and wetlands removed or altered via these various mechanisms are collectively referred to as
"lost" in this assessment. Methods used to estimate these losses are described in Box 7-1.
Bristol Bay Assessment 719 January 2014
-------
Mine Footprint
Figure 7-10. Streams and wetlands lost (eliminated, blocked, or dewatered) in the Pebble 0.25
scenario. Light blue areas indicate streams and rivers from the National Hydrography Dataset (USGS
2012a) and lakes and ponds from the National Wetlands Inventory (USFWS 2012); dark blue areas
indicate wetlands from the National Wetlands Inventory (USFWS 2012). See Box 7-1 for definitions
and methods used for delineation.
.r
•
' " '
Pebble 0.25 Components
Drawdown Zone
Eliminated, Blocked, or Dewatered Streams
Eliminated, Blocked, or Dewatered
Lakes and Ponds
Eliminated, Blocked, or Dewatered Wetlands
Bristol Bay Assessment
7-20
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-11. Streams and wetlands lost (eliminated, blocked, or dewatered) in the Pebble 2.0
scenario. Light blue areas indicate streams and rivers from the National Hydrography Dataset (USGS
2012a) and lakes and ponds from the National Wetlands Inventory (USFWS 2012); dark blue areas
indicate wetlands from the National Wetlands Inventory (USFWS 2012). See Box 7-1 for definitions
and methods used for delineation.
•^E-jfcU
Pebble 2.0 Components
Drawdown Zone
Eliminated, Blocked, or Dewatered Streams
Eliminated, Blocked, or Dewatered
Lakes and Ponds
Eliminated, Blocked, or Dewatered Wetlands
Bristol Bay Assessment
7-21
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-12. Streams and wetlands lost (eliminated, blocked, or dewatered) in the Pebble 6.5
scenario. Light blue areas indicate streams and rivers from the National Hydrography Dataset (USGS
2012a) and lakes and ponds from the National Wetlands Inventory (USFWS 2012); dark blue areas
indicate wetlands from the National Wetlands Inventory (USFWS 2012). See Box 7-1 for definitions
and methods used for delineation.
-" J
^/
Pebble 6.5 Components
Drawdown Zone
Eliminated, Blocked, or Dewatered Streams
Eliminated, Blocked, or Dewatered
Lakes and Ponds
Eliminated, Blocked, or Dewatered Wetlands
Bristol Bay Assessment
7-22
January 2014
-------
Chapter 7 Mine Footprint
To calculate kilometers of streams eliminated, blocked, or experiencingstreamflow alteration due to the
footprints of the major mine components and the groundwater drawdown zone associated with the mine pit,
we used the Alaska National Hydrography Dataset (NHD) (USGS 2012a). The scale of this dataset is
1:63,360. In this assessment, a stream segment is classified as eliminated if it falls within the boundaries
of the mine pit, waste rock pile, or tailings storage facility (TSF). A segment is classified as blocked if it or a
downstream segment it connects to directly intersects the mine pit, waste rock pile, or TSF. A stream
segment not otherwise eliminated is classified as dewatered if it falls within the groundwater drawdown
zone associated with the mine pit, or is classified as blocked and dewatered if it falls within the groundwater
drawdown zone and a downstream segment it connects to directly intersects the mine pit. For calculation of
stream kilometers either eliminated or blocked that are inhabited by anadromous and resident fish species,
we used the Anadromous Waters Catalog (AWC) (Johnson and Blanche 2012) and the Alaska Freshwater
Fish Inventory (AFFI) (ADF&G 2012). Stream lengths blocked, eliminated, or dewatered were summed across
each classification for both NHD and AWC fish-in habited stream segments (Table 7-5).
Estimates of wetland, pond, and lake areas eliminated, blocked, or dewatered by the mine scenario
footprints were derived from the National Wetlands Inventory (NWI) (USFWS 2012). For the State of Alaska,
the scale of this dataset is 1:63,360. In this assessment, wetland, pond, or lake area is classified as
eliminated if it falls within the boundaries of the mine pit, waste rock pile, or TSF; dewatered if it falls within
the groundwater drawdown zone associated with the mine pit; and blocked if it directly intersects a
previously categorized blocked NHD stream. Wetland, pond, and lake areas blocked, eliminated, or
dewatered were summed across each classification (Table 7-8).
The area covered by plant and ancillary facilities associated with mine site development (e.g., housing,
crushing plant, wastewater treatment plant) is not considered in the calculation of eliminated and blocked
streams and wetlands due to lack of knowledge about the orientation and placement of these structures on
the landscape. Thus, the values reported in Tables 7-5 and 7-8 are conservative estimates, as additional
development on the landscape would likely impact additional stream length and wetland area due to the
abundance of aquatic habitats in this region.
It is important to note that estimates of stream length and wetland, pond, and lake areas affected represent
a lower bound on the estimate. The NHD does not capture all stream courses and may underestimate
channel sinuosity, resulting in underestimates of affected stream length. In addition, the AWC and the AFFI
do not necessarily characterize all potential fish-bearing streams, because it is not possible to sample all
streams and there may be errors in identification and mapping. The Alaska Department of Fish and Game, in
its on-line AWC database, states: "Based upon thorough surveys of a few drainages it is believed that this
number represents less than 50% of the streams, rivers and lakes actually used by anadromous species"
(ADF&G 2013). The characterization of wetland area is limited by resolution of the available NWI data
product, which is not available for the full assessment area. Other investigations have revealed high spatial
variability in wetland density across the region (e.g., Hall etal. 1994). Others have conducted enhanced
wetland inventories. For example, the Pebble Limited Partnership (2011) used multiple sources of high
resolution remote imagingand ground-truthingto map wetlands in their mine mappingarea, which focused
on the proposed mine working area and major stream valleys. They reported wetland densities of
approximately 29% for the mapping area (PLP 2011: Table 14.1-3), whereas preliminary NWI mapping
identified approximately 20% of this same area as wetland (PLP 2011: Table 14.1-1). Furthermore, the
major mine components of the mine scenarios often bisected wetland, pond, or lake features, and areas
falling outside the boundary were assumed to maintain their functionality. We were also unable to determine
the effect that mine site development may have on wetlands with no direct surface connection to a blocked
NHD stream segment, but with a potential connection via groundwater pathways. Given these limitations,
these estimates could be enhanced with improved, higher-resolution mapping, increased sampling of
possible fish-bearing waters, and ground-truthing.
Bristol Bay Assessment 723 January 2014
-------
Chapter 7 Mine Footprint
7.2.2.1 Stream Losses
In the Pebble 0.25 scenario, 38 km of streams would be eliminated, blocked, or dewatered by the mine
footprint (Table 7-5, Figure 7-10). In the Pebble 2.0 scenario, over 89 km of streams would be
eliminated, blocked, or dewatered by the mine footprint (Table 7-5, Figure 7-11). In the Pebble 6.5
scenario, an additional 20 km of streams in the pit and waste rock pile areas and 41 km of streams under
TSF 2 and TSF 3 would be eliminated or blocked, for a total of 151 km of streams lost to the mine
footprint (Table 7-5, Figure 7-12). These scenarios represent 4, 8, and 14% of the total stream length
within the mine scenario watersheds. Of the streams lost to the mine footprint in the Pebble 6.5
scenario, 82% are headwater streams (less than 0.15 m3/s mean annual streamflow); 76% have less
than 3% gradient, and 26% have less than 1% gradient (Table 7-6, Figure 7-13). The majority (74%) of
smaller streams lost to the mine footprint in the Pebble 6.5 scenario flow through valleys with limited
flatland (Table 7-6).
Compared to the larger Nushagak and Kvichak River watersheds, streams lost to the mine footprints are
smaller: 9% of stream length in the Nushagak and Kvichak River watersheds exceeds 2.8 m3/s mean
annual streamflow (Table 3-3), whereas no streams lost to the mine footprints exceed this size (Figure
7-13A). Streams within the mine footprints also have a lower proportion of stream length with less than
1% gradient (26% versus 64% of stream length in the Nushagak and Kvichak River watersheds) (Figure
7-13B), and more stream length with low floodplain potential (74% versus 39%) (Figure 7-13C).
These results provide some indication of the relative size, steepness, and geomorphic setting of streams
that would be lost to the mine footprints. The streams that would be lost include a range of stream types
that provide a variety of habitat functions for salmon, including as year-round or seasonal habitat for
salmonids or other fishes or as important sources of water, macroinvertebrates, and other materials to
downstream waters (Section 7.2.3). Of the 151 km of streams lost to the Pebble 6.5 footprint, 36 km are
currently cataloged as anadromous fish streams in the AWC (Johnson and Blanche 2012). Most of these
cataloged anadromous streams are in the medium stream size class (0.15 to 2.8 m3/s), with gradients
less than 3%. These include the upper reaches and larger tributaries of the South and North Fork Koktuli
Rivers and Upper Talarik Creek, including smaller streams with documented occurrence of coho salmon
(Figure 7-3). Many of the smaller, steeper tributaries have been documented to contain Dolly Varden
(Figure 7-7).
Bristol Bay Assessment 724 January 2014
-------
Chapter 7
Mine Footprint
Figure 7-13. Cumulative frequency of stream channel length classified by (A) mean annual
streamflow (m3/s), (B) channel gradient (%), and (C) floodplain potential (based on % flatland
in lowland) for the mine footprints (Scale 4) versus the Nushagak and Kvichak River
watersheds (Scale 2).
100% -i
A
ono/
ni +•»
5 u>
— c fin%
™ 2
3 £
£ m API0/
§ 3 4U/o
u £
1/1
9n°/,
no/
0
__ mnn/
D
«(•)%
QJ fn
> M
'^ . sn%
_ro Jj
c ro /in°/
3 ™ 4U/o
<-» £
to
7n%
n%
0
C1 nn%
sno/,
5 "a
•— C cno/
+j ... DU /o
J2 -i
£ m An°/i
2 OJ
i:
1/1 ?n°/
no/,
0
X1
i
1
1
I
1
, 1
'J
>/0
/
/
/
/
/
%
»•• "*
Scale4
MAP Classification
Scale2
10 20 30 40 50
Mean Annual Streamflow (m3/s)
*-s**^
7 /
/ Scale4
/ Gradient Classification
/ C/--^lci "5
f
4% 8% 12% 16% 20%
Channel Gradient
^^-^ ~ ~-~ ~--~~"~
r ~ - "*
~~ ~
_• ** r r -i \ o ,1
*
s ' Floodplain Potential Classification
r
Scale 2
20% 40% 60% 80% 100%
Flatland in Lowland
Bristol Bay Assessment
7-25
January 2014
-------
Chapter 7
Mine Footprint
Table 7-5. Stream length (km) eliminated, blocked, or dewatered by the mine footprints in the Pebble 0.25, 2.0, and 6.5 scenarios. See Box
7-1 for methods.
Component
Stream Length3
Eliminated
by Footprint
Blocked by
Footprint
Dewatered
by Footprint
Blocked and
Dewatered
by Footprint
TOTAL
Stream
Length Lostc
to Footprint
AWC Stream Length"
Eliminated
by Footprint
Blocked by
Footprint
Dewatered
by Footprint
Blocked and
Dewatered
by Footprint
Salmon
Species
Present in
Lostc Streams
TOTAL AWC
Stream
Length
Lostc to
Footprint
Pebble 0.25
Pit
Waste rock
TSFld
TOTAL
3.0
5.1
12.3
20.4
0.0
<0.1
2.7
2.8
13.4
NA
NA
13.4
1.4
NA
NA
1.4
17.8
5.1
15.0
38.0
0.0
0.0
6.1
6.1
0.0
0.0
0.0
0.0
1.5
NA
NA
1.5
0.0
NA
NA
0.0
Chinook, coho
Chinook, coho
1.5
0.0
6.1
7.7
Pebble 2.0
Pit + waste
rock
TSFld
TOTAL
56.9
15.4
72.3
10.2
0.2
10.4
1.9
NA
1.9
4.8
NA
4.8
73.8
15.5
89.4
11.3
6.3
17.6
1.7
0.0
1.7
0.0
NA
0.0
2.4
NA
2.4
Chinook,
coho, sockeye
Chinook, coho
15.4
6.3
21.7
Pebble 6.5
Pit + waste
rock
TSF1
TSF2
TSF3
TOTAL
76.9
15.4
28.3
10.2
130.8
5.9
0.2
2.2
0.7
9.0
3.4
NA
NA
NA
3.4
7.7
NA
NA
NA
7.7
93.9
15.5
30.5
10.9
150.9
18.7
6.3
6.1
3.3
34.4
0.0
0.0
0.0
0.0
0.0
0.0
NA
NA
NA
0.0
1.6
NA
NA
NA
1.6
Chinook,
coho, sockeye
Chinook,
coho,
Chinook,
chum, coho
Coho
20.3
6.3
6.1
3.3
36.0
Notes:
'" From the National Hydrography Dataset (USGS 2012a).
b From the Anadromous Waters Catalog (Johnson and Blanche 2012)
c Lost = eliminated + blocked + dewatered.
d TSF 1 expands in size in the Pebble 2.0 scenario.
TSF = tailings storage facility; AWC = Anadromous Waters Catalog; NA = not applicable as the mine pit dewatering zone does not overlap these individual components.
Bristol Bay Assessment
7-26
January 2014
-------
Chapter 7
Mine Footprint
Table 7-6. Distribution of stream channel length classified by channel size (based on mean annual
streamflow in m3/s), channel gradient (%), and floodplain potential (based on % flatland in lowland)
for streams lost to the Pebble 6.5 mine footprint. Gray shading indicates proportions greater than
5%; bold indicates proportions greater than 10%.
Channel Size
Small headwater streams3
Medium streams'5
Small rivers0
Large riversd
Gradient
<1%
FP
10%
6%
0%
0%
NFP
4%
6%
0%
0%
>1% and <3%
FP
9%
1%
0%
0%
NFP
35%
5%
0%
0%
>3% and <8%
FP
0%
0%
0%
0%
NFP
23%
0%
0%
0%
>8%
FP
0%
0%
0%
0%
NFP
1%
0%
0%
0%
Notes:
3 0-0.15 m3/s; most tributaries in the mine footprints.
b 0.15-2.8 m3/s; upper reaches and larger tributaries of the North and South Fork Koktuli Rivers and Upper Talarik Creek.
c 2.8-28 m3/s; middle to lower portions of the North and South Fork Koktuli Rivers and Upper Talarik Creek, including the mainstem Koktuli
River.
d >28 m3/s; the Mulchatna River below the Koktuli confluence, the Newhalen River, and other large rivers.
FP = high floodplain potential (>5% flatland in lowland); NFP = no or low floodplain potential (<5% flatland in lowland) (see Chapter 3 for
additional details).
Table 7-7 provides a summary of the total documented anadromous fish stream length in the mine
scenario watersheds (Johnson and Blanche 2012). Approximately 2, 7, and 11% of the total anadromous
fish stream length in the mine scenario watersheds would be eliminated, blocked, or dewatered in the
Pebble 0.25, 2.0, and 6.5 scenarios, respectively (Table 7-5). In addition to these direct losses, loss of
these headwater habitats would have indirect impacts on fishes and their habitats in downstream
mainstem reaches of each watershed (Section 7.2.3).
Table 7-7. Total documented anadromous fish stream length and stream length documented to
contain different salmonid species in the mine scenario watersheds.
Total mapped streams3
Total anadromous fish streams'1
South Fork Koktuli
River (km)
315
95
North Fork Koktuli
River (km)
343
104
Upper Talarik
Creek (km)
427
123
Total (km)
1,085
322
By species
Chinook salmon
Chum salmon
Coho salmon
Pink salmon
Sockeye salmon
Dolly Varden0
59
37
93
0
64
48
61
31
103
0
47
0
63
45
122
7
80
26
183
113
318
7
191
75
Notes:
a From the National Hydrography Dataset (USGS 2012a).
b From the Anadromous Waters Catalog (Johnson and Blanche 2012).
c Listed as Arctic char in some cases, but assumed to be Dolly Varden (Appendix B).
Bristol Bay Assessment
7-27
January 2014
-------
Chapter 7 Mine Footprint
7.2.2.2 Wetland, Pond, and Lake Losses
In addition to the stream losses detailed above, 4.5,12, and 18 km2 of wetlands and 0.41, 0.93, and 1.8
km2 of ponds and lakes would be lost in the Pebble 0.25, 2.0, and 6.5 scenarios, respectively. (Table 7-8,
Figures 7-10 through 7-12). Methods used to estimate these losses are described in Box 7-1.
7.2.3 Exposure-Response: Implications of Stream and Wetland Loss for
Fish
7.2.3.1 Fish Occurrence in Streams and Wetlands Lost to the Mine Scenario Footprints
Tables 7-5 and 7-8 provide an estimate of salmon habitat directly affected by the mine footprint in the
three mine scenarios. A total of 8, 22, and 36 km of documented anadromous fish streams would be
eliminated, blocked, or dewatered by the mine footprints in the Pebble 0.25, 2.0, and 6.5 scenarios,
respectively. The distribution of anadromous Dolly Varden in the Nushagak and Kvichak River
watersheds is not fully known, making our estimate of the total anadromous fish habitat affected by the
mine scenarios incomplete. Of the total wetland area eliminated, blocked, or dewatered by each
footprint, the proportion used by anadromous salmonids or resident fish species is unknown. Fish
access to and use of wetlands are likely to be extremely variable in the deposit area, due to differences in
the duration and timing of surface water connectivity with stream habitats, distance from the main
channel, and physical and chemical conditions (e.g., dissolved oxygen concentrations) (King et al. 2012).
Wetlands can provide refuge habitats (Brown and Hartman 1988) and important rearing habitats for
juvenile salmonids by providing hydraulically and thermally diverse conditions. Wetlands can also
provide enhanced foraging opportunities (Sommer et al. 2001). Given our insufficient knowledge of how
fish use wetlands in the deposit area, it is not possible to calculate the effects of lost wetland
connectivity and abundance on stream fish populations.
Among the Pacific salmon species, coho salmon occupy the highest proportion of designated AWC
stream segments in the mine scenario watersheds (Table 7-7). Spawning habitat for coho salmon would
be lost in the South and North Fork Koktuli River watersheds under TSF 1 and TSF 2, respectively
(Figure 7-3); sockeye and coho salmon spawning habitat would be lost in the Upper Talarik Creek
watershed under the waste rock piles (Figures 7-2 and 7-3) (Johnson and Blanche 2012). In other
regions, anadromous and resident forms of Dolly Varden have been observed in the most upstream and
high-gradient habitats available for spawning, indicating that headwaters may be important source
areas for downstream populations (Bryant et al. 2004). Under the Pebble 6.5 footprint, 99% of stream
kilometers are estimated to have gradients less than 8% and 76% are estimated to have gradients less
than 3%, well within the range of gradients used by these species.
Bristol Bay Assessment 728 January 2014
-------
Chapter 7
Mine Footprint
Table 7-8. Wetland, pond, and lake areas3 (km2) eliminated, blocked, or dewatered by the mine footprints in the Pebble 0.25, 2.0, and 6.5 scenarios.
Component
Eliminated by Footprint
Wetland
Pond
Lake
Sum
Blocked by Footprint
Wetland
Pond
Lake
Sum
Dewatered by Footprint
Wetland
Pond
Lake
Sum
Blocked and Dewatered by
Footprint
Wetland
Pond
Lake
Sum
Total Area Lost to Footprint
Wetland
Pond
Lake
Sum
Pebble 0.25
Pit
Waste rock
TSF1
Total
0.26
0.29
2.33
2.88
0.02
0.07
<0.01
0.09
0.00
0.00
0.00
0.00
0.27
0.36
2.34
2.97
0.00
0.00
0.31
0.31
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.31
0.31
1.30
NA
NA
1.30
0.28
NA
NA
0.28
0.02
NA
NA
0.02
1.60
NA
NA
1.60
0.03
NA
NA
0.03
0.03
NA
NA
0.03
0.00
NA
NA
0.00
0.06
NA
NA
0.06
1.59
0.29
2.64
4.52
0.32
0.07
<0.01
0.39
0.02
0.00
0.00
0.02
1.93
0.36
2.64
4.93
Pebble 2.0
Pit + waste
rock
TSF1»
Total
5.86
3.56
9.42
0.63
<0.01
0.64
0.12
0.00
0.12
6.61
3.57
10.18
1.67
0.00
1.67
0.06
0.00
0.06
0.00
0.00
0.00
1.73
0.00
1.73
0.33
NA
0.33
0.08
NA
0.08
0.00
NA
0.00
0.41
NA
0.41
0.15
NA
0.15
0.04
NA
0.04
0.00
NA
0.00
0.19
NA
0.19
8.01
3.56
11.57
0.81
<0.01
0.82
0.12
0.00
0.12
8.94
3.57
12.51
Pebble 6.5
Pit + waste
rock
TSF1
TSF2
TSF3
Total
10.16
3.56
1.94
0.54
16.20
0.88
<0.01
<0.01
0.01
0.90
0.70
0.00
0.00
0.00
0.70
11.74
3.57
1.94
0.55
17.80
0.24
0.00
0.02
0.02
0.28
0.01
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.26
0.00
0.02
0.02
0.30
0.73
NA
NA
NA
0.73
0.18
NA
NA
NA
0.18
0.00
NA
NA
NA
0.00
0.91
NA
NA
NA
0.91
0.72
NA
NA
NA
0.72
0.03
NA
NA
NA
0.03
0.00
NA
NA
NA
0.00
0.75
NA
NA
NA
0.75
11.86
3.56
1.96
0.56
17.94
1.10
<0.01
<0.01
0.01
1.11
0.70
0.00
0.00
0.00
0.70
13.66
3.57
1.96
0.57
19.77
Notes:
a Based on the National Wetlands Inventory (USFWS 2012).
b TSF 1 expands in size in the Pebble 2.0 scenario.
TSF = tailings storage facility; NA = not applicable as the mine pit dewatering zone does not overlap with the footprints of these individual components.
Bristol Bay Assessment
7-29
January 2014
-------
Chapter 7 Mine Footprint
In addition to spawning, streams in each mine footprint provide rearing habitat for fish of the mine
scenario watersheds. Species known to rear in habitats in and upstream of the mine footprints are
sockeye salmon (Figure 7-2), coho salmon (Figure 7-3), Chinook salmon (Figure 7-4), chum salmon
(Figure 7-5), Dolly Varden (Figure 7-7), rainbow trout (Figure 7-8), Arctic grayling, slimy sculpin,
northern pike, and ninespine stickleback (ADF&G 2012, Johnson and Blanche 2012).
7.2.3.2 Importance of Headwater Stream and Wetland Habitats
The majority of streams lost to the footprint of the Pebble 6.5 scenario are classified as small headwater
streams (less than 0.15 m3/s mean annual streamflow) (Table 7-6). Because of their narrow width,
headwater streams receive proportionally greater inputs of organic material from the surrounding
terrestrial vegetation than larger stream channels (Vannote et al. 1980). This material is either used
locally (Tank et al. 2010) or transported downstream as a subsidy to larger streams in the network
(Wipfli et al. 2007). Consumers in headwater stream foodwebs, such as invertebrates and juvenile
salmon, can rely heavily on terrestrial inputs that enter the stream (Doucett et al. 1996, Dekar et al.
2012). Headwater streams also encompass the upper limits of anadromous fish distribution. These
streams may receive fewer or no marine-derived nutrients (MDN) from spawning salmon relative to
downstream portions of the river network, making terrestrial nutrient sources relatively more
important (Wipfli and Baxter 2010). Because of their shallow depths and propensity to freeze,
headwater streams may be largely uninhabitable in the winter (but see discussion of overwintering
below), and fish distribution in headwater systems in southwestern Alaska is likely most extensive in
late summer and early fall (Elliott and Finn 1983). This coincides with maximum growth periods for
rearing juvenile salmon, as both stream temperatures and food availability increase (Quinn 2005).
Data on riparian vegetation communities specific to the mine footprints were not available, but
vegetation in the deposit area is described generally by PLP (2011). Shrub vegetation communities
account for 81% of the total area, with four dominant vegetation types: dwarf ericaceous shrub tundra,
dwarf ericaceous shrub lichen tundra, open willow low shrub, and closed alder tall shrub (PLP 2011:
Chapter 13:10). Riparian areas are dominated by willow and alder shrub communities (PLP 2011:
Chapter 13:11). Deciduous shrub species such as alder and willow provide abundant and nutrient-rich
leaf litter inputs, which are used more rapidly in stream foodwebs than coniferous plants or grasses
(Webster and Benfield 1986). In addition, alder is a nitrogen-fixing shrub known to increase headwater
stream nitrogen concentrations (Compton et al. 2003, Shaftel et al. 2012), which can result in more rapid
litter processing rates (Ferreira et al. 2006, Shaftel et al. 2011). The presence of both willow and alder in
headwater stream riparian zones implies high-quality basal food resources for stream fishes in the
deposit area.
In addition to providing summer rearing habitat, lower-gradient headwater streams and associated
wetlands may also provide important habitat for stream fishes during other seasons. Loss of wetlands is
a common result of land development (Pess et al. 2005), and in more developed regions has been
associated with reductions in habitat quality and salmon abundance, particularly for coho salmon
(Beechie et al. 1994, Pess et al. 2002). Thermally diverse habitats in off-channel wetlands can provide
Bristol Bay Assessment 730 January 2014
-------
Chapter 7 Mine Footprint
rearing and foraging conditions that may be unavailable in the main stream channel (Sommer et al.
2001, Henning et al. 2006), increasing capacity for juvenile salmon rearing (Brown and Hartman 1988).
Winter habitat availability for juvenile rearing has been shown to limit salmonid productivity in streams
of the Pacific Northwest (Nickelson et al. 1992, Solazzi et al. 2000, Pollock et al. 2004) and may be
limiting for fish in the mine scenario watersheds given the relatively cold temperatures and long winters
in the region.
Overwintering habitats for stream fishes must provide suitable instream cover, dissolved oxygen, and
protection from freezing (Cunjak 1996). Beaver ponds and groundwater sources in headwater streams
and wetlands in the mine footprints likely meet these requirements. In winter, beaver ponds typically
retain liquid water below the frozen surface, which makes them important winter refugia for stream
fishes (Nickelson et al. 1992, Cunjak 1996). Beavers preferentially colonize headwater streams because
of their shallow depths and narrow widths, and several studies have indicated that dam densities are
reduced significantly at stream gradients above 6 to 9% (Collen and Gibson 2001, Pollock et al. 2003).
Beaver ponds provide excellent habitat for rearing salmon by trapping organic materials and nutrients
and creating structurally complex, large capacity pool habitats with potentially high macrophyte cover,
low streamflow velocity, and/or moderate temperatures (Nickelson et al. 1992, Collen and Gibson 2001,
Lang et al. 2006). Additionally, beaver dams, including ponds at a variety of successional stages, provide
a mosaic of habitats for not just salmon but other fish and wildlife species.
An aerial survey of active beaver dams in the deposit area, conducted in October 2005 (PLP 2011:
Chapter 16:16.2-8), mapped 113 active beaver colonies. The area surveyed did not include streams
draining the TSF 1 area (PLP 2011: Figure 16.2-20). Several active beaver colonies were mapped in
streams that would be eliminated or blocked by the mine pit and waste rock piles. These are lower-
gradient habitats than the headwater streams draining the TSF areas. Beaver ponds provide important
and relatively abundant habitat within the deposit area and may be particularly important for
overwinter rearing of species such as coho salmon and for providing deeper pool habitats for additional
species during low streamflow conditions (PLP 2011: Appendix 15.ID). Loss of beaver pond habitats in
the headwaters of the South Fork Koktuli River and Upper Talarik Creek watersheds would reduce both
summer and winter rearing opportunities for anadromous and resident fish species.
Inputs of groundwater-influenced streamflow from headwater tributaries likely benefit fish by
moderating mainstem temperatures and contributing to thermal diversity in downstream waters
(Cunjak 1996, Power et al. 1999, Huusko et al. 2007, Armstrong et al. 2010, Brown et al. 2011). PLP
(2011) collected temperature data from stream sampling sites using in-situ field meters (PLP 2011:
Appendix 15.1-E). Maximum summer (June through August) water temperatures recorded at gage
NK119A, which drains the TSF 1 area, were approximately 5°C colder than the mainstem reach that it
flows into (PLP 2011: Tables 15.1 through 15.4). This difference was not as pronounced at gage SK119A,
which drains the TSF 2 area and where recorded maximum summer water temperatures were
approximately 2°C colder than the mainstem reach that it flows into (PLP 2011: Tables 15.1 through
15.21).
Bristol Bay Assessment 731 January 2014
-------
Chapter 7 Mine Footprint
Longitudinal temperature profiles for the South and North Fork Koktuli River watersheds from August
and October indicate that the mainstem reaches just downstream of the tributaries draining TSF 1 and
TSF 2 experience significant summer cooling and winter warming compared to adjacent upstream
reaches (PLP 2011: Figures 15.1-11 and 15.1-41). Such thermal diversity can be an important attribute
of stream systems in the region, providing localized water temperature patches that may offer differing
trade-offs for species bioenergetics. For example, salmon may select relatively cold-temperature sites—
often associated with groundwater upwelling—for spawning, whereas juvenile salmon rearing in those
same streams may take advantage of warm-temperature patches for optimal food assimilation (e.g.,
Armstrong and Schindler 2013). Headwater streams in the South and North Fork Koktuli River
watersheds may provide a temperature-moderating effect and serve as sources of thermal
heterogeneity, providing cooler temperatures in summer and warmer temperatures in winter.
It has long been recognized that, in addition to providing habitat for stream fishes, headwater streams
and wetlands serve an important role in the stream network by contributing water, nutrients, organic
material, macroinvertebrates, algae, and bacteria downstream to higher-order streams in the watershed
(Vannote et al. 1980, Meyer et al. 2007). However, only recently have specific subsidies from headwater
streams been extensively quantified (Wipfli and Baxter 2010). Headwater contributions to downstream
systems result from the high density of headwaters in the dendritic stream network. Headwater streams
can also have high instream rates of nutrient processing and storage, thereby influencing downstream
water chemistry due to relatively large organic matter inputs, high retention capacity, high primary
productivity, bacteria-induced decomposition, and/or extensive hyporheic zone interactions
(Richardson et al. 2005, Alexander et al. 2007, Meyer et al. 2007).
Wipfli and Gregovich (2002) demonstrated that invertebrates and detritus are exported from
headwaters to downstream reaches and provide an important energy subsidy for juvenile salmonids.
Wipfli and Baxter (2010) describe how the relative importance of energy subsidies from headwaters,
terrestrial inputs, benthic production, and marine sources varies within salmon watersheds based on
spatial and temporal context. For example, foodwebs in small headwater streams of the mine scenario
watersheds may be proportionally more dependent on local terrestrial energy subsidies, whereas
stream communities in downstream waters may be more dependent on large seasonal fluxes of MDN.
Small headwater streams can be important exporters of subsidies to downstream waters, but the
relative value of this contribution will depend on the quantity and energy content of headwater-derived
subsidies relative to other energy sources (e.g., MDN, benthic production) that vary in time and space
(Wipfli and Baxter 2010).
The export value of headwater streams can be mediated by the surrounding vegetation. In southeastern
Alaskan streams, riparian alder (a nitrogen-fixing shrub) was positively related to aquatic invertebrate
densities and the export rates of invertebrates and detritus (Piccolo and Wipfli 2002, Wipfli and
Musslewhite 2004). In south-central Alaskan streams on the Kenai Peninsula, grass-dominated
headwater wetlands and associated vegetation can also be important sources of dissolved organic
matter, particulate organic matter, and macroinvertebrate diversity, contributing to the chemical,
physical, and biological condition of streams draining these landscapes (Shaftel et al. 2011, Dekar et al.
Bristol Bay Assessment 732 January 2014
-------
Chapter 7 Mine Footprint
2012, King et al. 2012, Walker et al. 2012). Because of their crucial influence on downstream water flow,
chemistry, and biota, impacts on headwaters reverberate throughout entire watersheds downstream
(Freeman et al. 2007, Meyer et al. 2007).
7.2.4 Risk Characterization
Direct loss of streams and wetlands to the mine footprints would make these habitats unavailable to
fishes. Such losses would be unavoidable for projects of the sizes described in our mine scenarios, due to
the density of streams and wetlands in the deposit area (combined 33% of the mine mapping area [PLP
2011: Table 14.1-5 and Figure 14.1-3]). Stream blockage is not necessarily unavoidable, but would
require appropriate engineering and maintenance. Indirect effects of headwater stream and wetland
losses due to the mine footprints would include reduced inputs of organic material, nutrients, water, and
macroinvertebrates to downstream reaches, but the relative effects of losses of upstream subsidies
would be highly context-dependent (Section 7.2.3).
The net effects of headwater stream and wetland losses would reduce the capacity and productivity of
stream habitats. Together, these reductions would result in adverse impacts on fish populations (Figure
7-1). These streams provide known spawning and rearing habitats for anadromous and resident fish
species, and their watersheds support some of region's highest diversity of salmonid species (Figure 5-
3). The lengths of streams lost directly to the Pebble 0.25, 2.0, and 6.5 mine footprints represent losses
of approximately 2, 7, and 11%, respectively, of the total AWC length in the mine scenario watersheds
(Table 7-7). Stream habitat losses leading to losses of local, unique populations would erode the
population diversity that is crucial to the stability of the overall Bristol Bay salmon fishery (Hilborn et al.
2003, Schindler et al. 2010).
Impact avoidance and minimization measures would not eliminate all the footprint impacts associated
with the mine scenarios, given the large extent and wide distribution of wetlands and streams in the
watersheds, the substantial infrastructure needed to support porphyry copper mining in this vast
undeveloped area, and the constraints that the ore body location puts on infrastructure siting options.
Compensatory mitigation measures could offset some of the stream and wetland losses described here
(Box 7-2), although the potential efficacy, applicability, and sustainability of these measures to
successfully offset adverse impacts face substantial challenges. Appendix J presents a more detailed
discussion of these compensatory mitigation issues.
Bristol Bay Assessment 733 January 2014
-------
Chapter 7 Mine Footprint
BOX 7-2. COMPENSATORY MITIGATION
Compensatory mitigation refers to the restoration, establishment, enhancement, and/or preservation of
wetlands, streams, or other aquatic resources. Compensatory mitigation regulations jointly promulgated by
the U.S. Environmental Protection Agency (USEPA) and the U.S. Army Corps of Engineers (USAGE) state that
"the fundamental objective of compensatory mitigation is to offset environmental losses resulting from
unavoidable impacts to waters of the United States authorized by [Clean Water Act Section 404 permits
issued by the USAGE]" (40 Code of Federal Regulations [CFR] 230.93(a)(l)). Compensatory mitigation
enters the analysis only after a proposed project design has incorporated all appropriate and practicable
means to avoid and minimize adverse impacts on aquatic resources (40 CFR 230.91(c)). Compensatory
mitigation measures are usually not part of project design, but are considered necessary to maintain the
integrity of the nation's waters. In addition, guidance issued by the USAGE Alaska District in 2009 clarifies
that fill placed in streams or in wetlands adjacent to anadromous fish streams in Alaska will require
compensatory mitigation (USAGE 2009). A 2011 supplement to the Alaska District's 2009 guidance further
recommends that projects in "difficult to replace" wetlands, fish-bearing waters, or wetlands within 500 feet
of such waters will also likely require compensatory mitigation, as will "large scale projects with significant
aquatic resource impacts," such as "miningdevelopment" (USAGE 2011).
The mine scenarios evaluated in this assessment identify that the mine footprints alone will result in the
loss (i.e., filling, blocking or otherwise eliminating) of high-functioning wetlands and tens of kilometers of
salmon-supporting streams. Appendix J provides an overview of Clean Water Act (CWA) Section 404
compensatory mitigation requirements for unavoidable impacts on aquatic resources and discusses the
likely efficacy of these potential compensation measures at offsetting potential adverse impacts. Note that
any formal determinations regarding compensatory mitigation can only take place in the context of a
regulatory action. This assessment is not a regulatory action, and thus a complete evaluation of
compensatory mitigation is outside the scope of this assessment.
Potential compensatory mitigation measures discussed in Appendix J include mitigation bank credits, in-lieu
fee program credits, and permittee-responsible compensatory mitigation projects, such as aquatic resource
restoration and enhancement within the South and North Fork Koktuli River and Upper Talarik Creek
watersheds as well as more distant portions of the Nushagak and Kvichak River watersheds. As discussed in
Appendix J, there are significant challenges regardingthe potential efficacy, applicability, and sustainability
of compensation measures for use in the Bristol Bay region, raising questions as to whether compensation
measures could realistically address impacts of this type and magnitude.
7.2.5 Uncertainties
Losses of anadromous fish-bearing streams in the mine scenario watersheds (Table 7-5) are likely
underestimated because of the difficulty of accurately capturing data on all streams that may support
fish use throughout the year. We rely on the AWC (Johnson and Blanche 2012) and the AFFI (ADF&G
2012) for documentation of species distributions, but these records are incomplete—not all stream
reaches have been surveyed—and may be subject to errors in fish identification. Additionally, depictions
of species and life history distributions in the AWC reflect a wide range of mapping policies, and it is
difficult to interpret under which policies a particular water body was mapped. That said, the fish
sampling documented by PLP (2011) is one of the highest-density efforts conducted to date in this
portion of Alaska, such that estimates of anadromous fish distributions are likely better represented
here than elsewhere in Alaska.
Losses of headwater streams and anadromous fish-bearing streams in the mine scenario watersheds
may also be underestimated because of challenges associated with stream network mapping. Estimates
of headwater stream extent were derived from the National Hydrography Dataset (NHD) for Alaska
Bristol Bay Assessment 734 January 2014
-------
Chapter 7 Mine Footprint
(USGS 2012a), which does not capture all stream courses and may underestimate channel sinuosity,
resulting in underestimates of stream length. A stream network map derived from a light detection and
ranging (LiDAR) mapping system would likely yield substantially different results than those presented
here. Similarly, actual wetland loss or blockage due to the mine footprints (Table 7-8) would likely be
higher than estimated here, as the National Wetlands Inventory (USFWS 2012) is based on remotely-
sensed imagery and generally underestimates wetland area. See Box 7-1 for additional discussion of
uncertainties associated with stream and wetland mapping.
In the Bristol Bay region, hydrologically diverse riverine and wetland landscapes provide a variety of
large river, floodplain, pond, and lake habitats for salmon spawning and rearing. Environmental
conditions can be very different among habitats in close proximity. The spatial separation and unique
spawning habitat features within the Bristol Bay watershed are associated with variation in life-history
characteristics and body morphology (Blair et al. 1993), and have influenced genetic divergence among
spawning populations of sockeye salmon at multiple spatial scales (Gomez-Uchida et al. 2011). These
distinct populations can occur at very fine spatial scales, with sockeye salmon that use spring-fed ponds
and streams approximately 1 km apart exhibiting differences in traits, such as spawn timing, spawn site
fidelity, and productivity, that are consistent with discrete populations (Quinn et al. 2012). In the Bristol
Bay region, phenotypic variation with apparent adaptive significance has been illustrated for sockeye
salmon egg size and spawning gravel size (Quinn et al. 1995), and for sockeye salmon body shape and
predation risk from brown bears (Quinn et al. 2001). Olsen et al. (2003) proposed that the fine-scale
genetic differentiation they observed in Alaskan coho salmon may be associated with local adaptation to
locally diverse freshwater selective pressures, but they did not examine phenotypic variation. These
results highlight the potential for fine-scale salmon population structure in the Bristol Bay watershed.
Current monitoring approaches are inadequate to fully assess population-level trends across the Bristol
Bay watershed (Rand et al. 2007). Additional genetic and ecological research is needed to clarify the
spatial scale of this population structure and the varying vulnerabilities of populations across the
landscape.
7.3 Streamflow Modification
7.3.1 Exposure: Streamflow
In this section, we describe projected changes in the hydrology of the mine scenario watersheds and
associated effects on downstream flows that would result from mine development and operation. We
assume that streams in and downstream of the mine footprints would experience Streamflow alterations
due to water collection, treatment, and discharge to streams via wastewater treatment plant (WWTP)
outfalls; leakage from TSFs; and leachate from waste rock piles. See Chapter 6 for a full description of
water flows through the mine facilities.
Streamflow alterations resulting from mine operations were estimated by reducing the streamflows
recorded at existing stream gages in the mine scenario watersheds (Table 7-9, Figures 7-14 through 7-
Bristol Bay Assessment 735 January 2014
-------
Chapter 7
Mine Footprint
16) by the percentage of expected surface area lost to each mine footprint and water yield efficiencies
for each watershed. Reductions also included losses to the drawdown zone, caused by the cone of
depression at the mine pit, or other locations of dewatering (Table 7-9, Section 6.2.2). Discharges
through the WWTP resulted in streamflow additions. Net effects on resulting streamflows were mapped
and summarized for individual stream and river segments (Figures 7-14 through 7-16).
Table 7-9. Stream gages and related characteristics for the South and North Fork Koktuli Rivers
and Upper Talarik Creek.
Stream and Gage
Drainage Area
(km2)
Mean Annual Streamflow3
(mVs)
Mean Annual Unit Runoff
(m3/s*km2)
South Fork Koktuli River
SK100G
SK100F
SK124A
SK100C
SK119A
SK100B1
SK100B"
14
31
22
99
28
141
179
0.4
0.8
0.5
1.3
1.0
3.7
5.1
0.026
0.026
0.024
0.013
0.036
0.026
0.029
North Fork Koktuli River
NK119A
NK119B
NK100C
NK100B
NK100A1
NK100AC
20
11
65
99
222
279
0.7
0.1
1.3
2.4
5.8
7.0
0.034
0.012
0.020
0.024
0.026
0.025
Upper Talarik Creek
UT100E
UT100D
UT100C2
UT100C1
UT100C
UT119A
UTIOOB"
10
31
133
159
185
10
222
0.3
0.8
2.9
3.4
4.5
0.8
6.2
0.027
0.025
0.022
0.022
0.024
0.079
0.028
Notes:
"" Calculated from stream gage data from PLP 2011.
b USGS 15302200.
c USGS 15302250.
d USGS 15300250.
Bristol Bay Assessment
7-36
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-14. Stream segments in the mine scenario watersheds showing streamflow changes (%) associated with the Pebble 0.25
footprint. Streamflow modification class is shown for each stream segment to indicate degree and direction of change. These classes are
assigned at a gage and extend upstream to the next gage, confluence point, or mine footprint. Channels and tributaries not classified are
119CP
SK100C
\ \
2k r-~
UTHSA^r/Bs,/^
UT100B •
PLP/USGS Gages
© Confluence Points
, .
J Pebble 0.25 Components
Y//X Drawdown Zone
5-10% Decrease
^^— 10-20% Decrease
>20% Decrease
0-5% Decrease or Increase
5-10% Increase
•^^— 10-20% Increase
o
>20% Increase _
0
M
N
A
-j
•=
i
•=
2
H Kilometers
2
1 Miles
Bristol Bay Assessment
7-37
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-15. Stream segments in the mine scenario watersheds showing streamflow changes (%) associated with the Pebble 2.0 footprint.
gage and extend upstream to the next gage, confluence point, or mine footprint. Channels and tributaries not classified are shown for
informational purposes. Gage locations based on U.S. Geological Survey (2012b) and Pebble Limited Partnership (2011).
WWTP
/ Discharge
K124A
T
SK124CP *SK100CP2
SK100C
• PLP/USGS Gages
0 Confluence Points
J Drawdown Zone
Pebble 2.0 Components
5-10% Decrease
10-20% Decrease
>20% Decrease
0-5% Decrease or Increase
5-10% Increase
10-20% Increase
>20% Increase
Bristol Bay Assessment
7-38
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-16. Stream segments in the mine scenario watersheds showing streamflow changes (%) associated with the Pebble 6.5 footprint.
Streamflow modification class is shown for each stream segment to indicate degree and direction of change. These classes are assigned at a
informational purposes. Gage locations based on U.S. Geological Survey (2012b) and Pebble Limited Partnership (2011).
SK119CP
SK100CP1 •
B
PLP/USGS Gages
Confluence Points
J Drawdown Zone
Pebble 6.5 Components
5-10% Decrease
10-20% Decrease
>20% Decrease
0-5% Decrease or Increase
5-10% Increase
10-20% Increase
>20% Increase
N
A
Bristol Bay Assessment
7-39
January 2014
-------
Chapter 7 Mine Footprint
Daily streamflow data were obtained using data from seven gages in the South Fork Koktuli River, six
gages in the North Fork Koktuli River, and seven gages in Upper Talarik Creek (Table 7-9) (PLP 2011).
We calculated mean and minimum monthly streamflows for each gage under pre-mining baseline
conditions (Tables 7-10 through 7-15, Figure 7-17). The periods of record varied for gages in the three
mine scenario watersheds, but generally covered the period from 2004 to 2010.
In addition, we estimated streamflow at six confluence points where mining-related streamflow impacts
were expected but where established stream gage records were lacking. This allowed for more discrete
estimation of baseline streamflow, as well as expected streamflow modification in each mine scenario
due to withdrawal, addition, or footprint loss. The tributary area to each stream gage or confluence
point was calculated based on the National Elevation Dataset digital elevation model (Gesch et al. 2002,
Gesch 2007, USGS 2013) in a geographic information system. We determined the area of each mine
component (i.e., the mine pit, waste rock piles, plant and ancillary areas, and TSFs) (Tables 6-5 through
6-7) in each drainage basin (Tables 7-16 through 7-18), and calculated the percentage of watershed area
covered by the mine components for each gage and confluence point subwatershed. Using the calculated
percentage of watershed area covered by the mine components, mean annual streamflow records for
each of the gages and confluence point subwatersheds were adjusted downward. Next, the annual
volume of return streamflow expected to reach each gage was added back to the adjusted streamflow
calculations based on the mine scenarios.
We assessed expected changes to surface water flows for the three mine scenarios (Tables 7-10 through
7-15). We also considered water balance issues for the post-closure period, but streamflow estimates
were not assessed for this period. The Pebble 0.25 mine footprint consists of the mine pit, its drawdown
zone (Section 6.2.2), one waste rock pile, plant and ancillary facilities, and TSF 1 (Table 6-5). The Pebble
2.0 footprint would add a second or expanded waste rock pile, larger areas for plant and ancillary
facilities, an expanded TSF 1, and a larger drawdown zone from groundwater flow to the pit (Table 6-6).
The Pebble 6.5 footprint would add effects associated with the fully expanded mine footprint (including
TSF 2 and TSF 3) to accommodate expanded mine operations (Table 6-7). We assume that during the
post-closure period, active dewatering of the pit would cease as the pit fills. Once the pit is filled, the
water level would be maintained below equilibrium level by pumping or gravity drainage to maintain a
gradient toward the pit. The pumped water would be treated for as long as it did not meet water quality
standards. When treatment was no longer necessary, the pit would be allowed to have a natural outlet if
the water level required one.
Bristol Bay Assessment 740 January 2014
-------
Chapter 7
Mine Footprint
Table 7-10. Measured mean monthly pre-mining streamflow rates (m3/s) and estimated mean monthly streamflow rates (m3/s) in the Pebble 0.25, 2.0, and 6.5 scenarios, for gages along the South Fork Koktuli River.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
SK100G
Pre
0.23
0.14
0.11
0.18
0.72
0.50
0.29
0.42
0.55
0.64
0.35
0.28
0.25
0.11
0.06
0.05
0.08
0.33
0.23
0.13
0.19
0.25
0.29
0.16
0.13
2.0
0.07
0.04
0.03
0.05
0.20
0.14
0.08
0.12
0.15
0.18
0.10
0.08
6.5
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
SK100F
Pre
0.44
0.25
0.19
0.24
1.95
1.38
0.59
0.83
1.20
1.47
0.75
0.53
0.25
0.32
0.18
0.14
0.18
1.44
1.02
0.44
0.62
0.89
1.08
0.55
0.39
2.0
0.28
0.16
0.12
0.16
1.26
0.90
0.38
0.54
0.78
0.95
0.48
0.35
6.5
0.17
0.09
0.07
0.09
0.74
0.53
0.23
0.32
0.46
0.56
0.28
0.20
SK124A
Pre
0.12
0.02
0.01
0.05
1.91
1.08
0.29
0.59
0.83
0.98
0.33
0.18
0.25
0.16
0.03
0.01
0.06
2.54
1.43
0.39
0.78
1.10
1.29
0.43
0.24
2.0
0.16
0.03
0.01
0.06
2.49
1.41
0.38
0.77
1.08
1.27
0.43
0.23
6.5
0.26
0.05
0.01
0.10
4.10
2.32
0.63
1.27
1.79
2.10
0.70
0.38
SK100C
Pre
0.37
0.03
<0.01
0.13
4.30
2.77
0.73
1.17
2.05
2.80
1.04
0.54
0.25
0.39
0.03
<0.01
0.13
4.52
2.90
0.77
1.23
2.15
2.93
1.09
0.57
2.0
0.38
0.03
<0.01
0.13
4.37
2.81
0.75
1.19
2.08
2.84
1.06
0.55
6.5
0.51
0.04
<0.01
0.18
5.87
3.77
1.00
1.60
2.79
3.81
1.42
0.74
SK119A
Pre
0.30
0.16
0.11
0.22
3.02
1.71
0.74
1.15
1.75
1.61
0.72
0.40
0.25
0.30
0.16
0.11
0.22
3.02
1.71
0.74
1.15
1.75
1.61
0.72
0.40
2.0
0.30
0.15
0.11
0.22
2.97
1.69
0.73
1.13
1.73
1.59
0.70
0.39
6.5
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
SK100B1
Pre
1.54
0.79
0.57
0.80
10.75
6.67
2.56
4.05
5.18
6.12
2.84
1.92
0.25
1.52
0.78
0.57
0.79
10.62
6.59
2.53
4.00
5.11
6.05
2.81
1.89
2.0
1.48
0.76
0.55
0.77
10.32
6.40
2.46
3.89
4.97
5.88
2.73
1.84
6.5
1.40
0.72
0.52
0.73
9.74
6.04
2.32
3.67
4.69
5.55
2.58
1.74
SKIOOB"
Pre
2.47
1.40
1.09
1.41
12.70
8.56
3.85
5.92
7.75
9.08
4.44
3.02
0.25
2.44
1.39
1.07
1.39
12.53
8.44
3.80
5.84
7.64
8.96
4.38
2.98
2.0
2.39
1.35
1.05
1.36
12.24
8.25
3.71
5.70
7.47
8.76
4.28
2.91
6.5
2.23
1.27
0.98
1.27
11.46
7.73
3.48
5.34
6.99
8.20
4.01
2.73
Notes:
a USGS 15302200.
NA = not applicable: SK100G would be eliminated by tailings storage facility (TSF) 2, and SK119A would be eliminated byTSFS in the Pebble 6.5 scenario.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
NK119A
Pre
0.15
0.10
0.08
0.21
2.28
1.15
0.55
0.71
1.10
1.10
0.52
0.24
0.25
0.10
0.07
0.06
0.15
1.63
0.82
0.39
0.51
0.79
0.78
0.37
0.17
2.0
0.05
0.04
0.03
0.08
0.86
0.43
0.21
0.27
0.42
0.41
0.20
0.09
6.5
0.06
0.04
0.03
0.08
0.87
0.44
0.21
0.27
0.42
0.42
0.20
0.09
NK119B
Pre
0.03
0.01
<0.01
0.03
0.54
0.20
0.05
0.10
0.20
0.26
0.09
0.04
0.25
0.03
0.01
<0.01
0.03
0.52
0.20
0.05
0.10
0.19
0.25
0.09
0.04
2.0
0.03
0.01
<0.01
0.03
0.48
0.18
0.04
0.09
0.18
0.23
0.08
0.04
6.5
0.02
0.01
<0.01
0.02
0.39
0.15
0.04
0.08
0.15
0.19
0.07
0.03
NK100C
Pre
0.71
0.48
0.39
0.54
3.48
1.91
1.11
1.24
1.75
2.20
1.24
0.88
0.25
0.80
0.54
0.44
0.61
3.93
2.16
1.25
1.40
1.98
2.49
1.40
0.99
2.0
0.79
0.53
0.43
0.61
3.90
2.15
1.24
1.38
1.96
2.47
1.39
0.98
6.5
1.13
0.76
0.62
0.87
5.58
3.07
1.78
1.98
2.81
3.53
1.99
1.40
NK100B
Pre
1.04
0.67
0.54
0.88
7.03
3.64
2.04
2.44
3.31
4.01
2.12
1.35
0.25
1.06
0.68
0.55
0.89
7.13
3.69
2.07
2.48
3.35
4.07
2.16
1.37
2.0
0.97
0.63
0.50
0.82
6.57
3.40
1.91
2.29
3.09
3.75
1.99
1.27
6.5
1.23
0.79
0.64
1.04
8.29
4.29
2.41
2.88
3.90
4.73
2.51
1.60
NK100A1
Pre
2.08
1.44
1.23
2.17
16.57
9.48
5.13
6.21
7.98
9.40
4.79
2.89
0.25
2.09
1.45
1.23
2.18
16.64
9.51
5.15
6.23
8.02
9.44
4.81
2.90
2.0
2.01
1.39
1.19
2.10
16.01
9.16
4.96
6.00
7.72
9.09
4.63
2.79
6.5
2.22
1.54
1.31
2.32
17.70
10.12
5.48
6.63
8.53
10.04
5.11
3.09
NK100A"
Pre
2.85
1.88
1.55
2.66
20.10
11.39
5.88
7.40
9.35
11.14
5.95
3.84
0.25
2.86
1.89
1.56
2.68
20.19
11.44
5.91
7.43
9.39
11.19
5.97
3.85
2.0
2.78
1.83
1.52
2.60
19.59
11.10
5.74
7.21
9.11
10.86
5.80
3.74
6.5
3.02
1.99
1.65
2.82
21.29
12.06
6.23
7.83
9.90
11.80
6.30
4.06
Notes:
a USGS 15302250.
Bristol Bay Assessment
7-41
January 2014
-------
Chapter 7
Mine Footprint
Table 7-12. Measured mean monthly pre-mining streamflow rates (m3/s) and estimated mean monthly streamflow rates (m3/s) in the Pebble 0.25, 2.0, and 6.5 scenarios, for gages along Upper Talarik Creek.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
UT100E
Pre
0.15
0.13
0.12
0.18
0.61
0.30
0.21
0.23
0.31
0.36
0.25
0.20
0.25
0.14
0.13
0.11
0.17
0.57
0.28
0.19
0.22
0.29
0.34
0.23
0.19
2.0
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
6.5
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
UT100D
Pre
0.32
0.28
0.22
0.55
1.95
1.02
0.62
0.78
1.03
1.18
0.74
0.52
0.25
0.32
0.28
0.22
0.55
1.95
1.02
0.62
0.78
1.03
1.18
0.74
0.52
2.0
0.29
0.25
0.20
0.50
1.77
0.93
0.56
0.71
0.94
1.07
0.68
0.47
6.5
0.18
0.16
0.12
0.31
1.09
0.57
0.34
0.43
0.58
0.66
0.41
0.29
UT100C2
Pre
0.05
0.04
0.03
0.08
0.28
0.15
0.09
0.11
0.15
0.17
0.11
0.07
0.25
1.32
1.15
0.93
2.06
6.64
4.04
2.40
2.81
4.21
4.69
2.98
2.05
2.0
1.29
1.13
0.91
2.02
6.50
3.96
2.35
2.75
4.12
4.59
2.91
2.01
6.5
1.18
1.03
0.84
1.85
5.96
3.63
2.16
2.52
3.78
4.21
2.67
1.84
UT100C1
Pre
1.05
0.92
0.74
1.64
5.28
3.22
1.91
2.24
3.35
3.73
2.37
1.64
0.25
1.74
1.55
1.28
2.51
7.43
4.29
2.76
3.30
4.67
5.26
3.67
2.61
2.0
1.71
1.52
1.26
2.47
7.30
4.22
2.72
3.25
4.59
5.17
3.60
2.57
6.5
1.59
1.41
1.17
2.30
6.79
3.93
2.53
3.02
4.27
4.81
3.35
2.39
UT100C
Pre
1.44
1.28
1.06
2.08
6.16
3.56
2.29
2.74
3.87
4.36
3.04
2.17
0.25
2.45
2.25
1.98
3.44
9.11
5.63
3.77
4.38
6.09
6.67
4.59
3.37
2.0
2.41
2.22
1.95
3.38
8.97
5.55
3.72
4.32
6.00
6.57
4.52
3.31
6.5
2.27
2.08
1.83
3.18
8.43
5.21
3.49
4.06
5.63
6.18
4.25
3.12
UT119A
Pre
2.09
1.92
1.69
2.93
7.76
4.80
3.21
3.73
5.19
5.68
3.91
2.87
0.25
0.76
0.75
0.74
0.78
0.88
0.82
0.80
0.81
0.86
0.88
0.84
0.80
2.0
0.69
0.69
0.67
0.71
0.80
0.75
0.72
0.74
0.78
0.81
0.76
0.73
6.5
0.67
0.66
0.65
0.69
0.77
0.72
0.70
0.72
0.76
0.78
0.74
0.71
UTIOOB"
Pre
0.60
0.60
0.59
0.62
0.69
0.65
0.63
0.64
0.68
0.70
0.66
0.63
0.25
3.62
3.31
2.88
4.79
12.80
7.40
5.13
6.48
7.82
9.08
6.33
5.00
2.0
3.54
3.23
2.81
4.68
12.49
7.22
5.00
6.32
7.63
8.86
6.18
4.88
6.5
3.34
3.05
2.66
4.42
11.81
6.83
4.73
5.97
7.21
8.37
5.84
4.61
Notes:
a USGS 15300250.
NA = not applicable: UT100E would be blocked by the waste rock pile in the Pebble 2.0 scenario (Figure 7-15), and by the mine pit in the Pebble 6.5 scenario (Figure 7-16).
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
SK100G
Pre
0.11
0.08
0.07
0.04
0.08
0.20
0.08
0.08
0.06
0.22
0.18
0.12
0.25
0.05
0.04
0.03
0.02
0.04
0.09
0.04
0.04
0.03
0.10
0.08
0.05
2.0
0.03
0.02
0.02
0.01
0.02
0.05
0.02
0.02
0.02
0.06
0.05
0.03
6.5
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
SK100F
Pre
0.20
0.15
0.11
0.11
0.14
0.46
0.22
0.16
0.08
0.63
0.34
0.21
0.25
0.15
0.11
0.08
0.08
0.10
0.34
0.16
0.12
0.06
0.47
0.25
0.16
2.0
0.13
0.10
0.07
0.07
0.09
0.30
0.14
0.10
0.05
0.41
0.22
0.14
6.5
0.08
0.06
0.04
0.04
0.05
0.18
0.08
0.06
0.03
0.24
0.13
0.08
SK124A
Pre
0.00
0.00
0.00
0.00
0.00
0.09
0.00
0.00
0.00
0.23
0.04
0.00
0.25
0.00
0.00
0.00
0.00
0.00
0.12
0.00
0.00
0.00
0.30
0.05
0.00
2.0
0.00
0.00
0.00
0.00
0.00
0.11
0.00
0.00
0.00
0.29
0.05
0.00
6.5
0.00
0.00
0.00
0.00
0.00
0.19
0.00
0.00
0.00
0.49
0.08
0.00
SK100C
Pre
0.00
0.00
0.00
0.00
0.00
0.12
0.00
0.00
0.00
0.71
0.12
0.00
0.25
0.00
0.00
0.00
0.00
0.00
0.12
0.00
0.00
0.00
0.74
0.12
0.00
2.0
0.00
0.00
0.00
0.00
0.00
0.12
0.00
0.00
0.00
0.72
0.12
0.00
6.5
0.00
0.00
0.00
0.00
0.00
0.16
0.00
0.00
0.00
0.96
0.16
0.00
SK119A
Pre
0.12
0.08
0.05
0.05
0.07
0.45
0.23
0.13
0.09
0.45
0.23
0.13
0.25
0.12
0.08
0.05
0.05
0.07
0.45
0.23
0.13
0.09
0.45
0.23
0.13
2.0
0.11
0.08
0.05
0.05
0.07
0.45
0.23
0.13
0.09
0.45
0.23
0.13
6.5
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
SK100B1
Pre
0.60
0.40
0.27
0.27
0.38
1.51
1.12
0.67
0.51
2.10
1.16
0.66
0.25
0.60
0.40
0.26
0.26
0.37
1.49
1.10
0.66
0.50
2.07
1.14
0.65
2.0
0.58
0.39
0.26
0.26
0.36
1.45
1.07
0.64
0.49
2.01
1.11
0.64
6.5
0.55
0.37
0.24
0.24
0.34
1.37
1.01
0.61
0.46
1.90
1.05
0.60
SK100B"
Pre
1.13
0.85
0.65
0.65
0.79
2.49
1.64
1.25
1.02
3.54
1.93
1.22
0.25
1.12
0.84
0.64
0.64
0.78
2.46
1.62
1.23
1.01
3.49
1.90
1.20
2.0
1.09
0.82
0.63
0.63
0.76
2.40
1.58
1.20
0.98
3.41
1.86
1.17
6.5
1.02
0.77
0.59
0.59
0.72
2.25
1.48
1.12
0.92
3.19
1.74
1.10
Notes:
a USGS 15302200.
NA = not applicable: SK100G would be eliminated by tailings storage facility (TSF) 2 and SK119A would be eliminated byTSFS in the Pebble 6.5 scenario.
Bristol Bay Assessment
7-42
January 2014
-------
Chapter 7
Mine Footprint
Table 7-14. Measured minimum monthly pre-mining streamflow rates (m3/s) and estimated minimum monthly streamflow rates (m3/s) in the Pebble 0.25, 2.0, and 6.5 scenarios, for gages along the North Fork Koktuli River.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
NK119A
Pre
0.08
0.07
0.06
0.04
0.01
0.30
0.21
0.14
0.12
0.20
0.12
0.10
0.25
0.05
0.05
0.04
0.03
0.01
0.22
0.15
0.10
0.09
0.14
0.09
0.07
2.0
0.03
0.03
0.02
0.02
0.00
0.11
0.08
0.05
0.05
0.08
0.05
0.04
6.5
0.03
0.03
0.02
0.02
0.00
0.11
0.08
0.05
0.05
0.08
0.05
0.04
NK119B
Pre
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.25
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
2.0
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
6.5
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NK100C
Pre
0.33
0.34
0.23
0.13
0.38
0.71
0.54
0.46
0.37
0.99
0.50
0.41
0.25
0.37
0.38
0.26
0.15
0.43
0.80
0.61
0.52
0.42
1.11
0.56
0.46
2.0
0.37
0.38
0.26
0.15
0.43
0.80
0.60
0.52
0.41
1.10
0.56
0.46
6.5
0.53
0.54
0.37
0.21
0.61
1.14
0.86
0.74
0.59
1.58
0.80
0.66
NK100B
Pre
0.43
0.44
0.33
0.18
0.54
1.31
1.04
0.96
0.91
1.53
0.71
0.57
0.25
0.44
0.45
0.34
0.19
0.55
1.33
1.05
0.97
0.92
1.55
0.72
0.58
2.0
0.41
0.41
0.31
0.17
0.51
1.23
0.97
0.89
0.85
1.43
0.67
0.53
6.5
0.51
0.52
0.39
0.22
0.64
1.55
1.23
1.13
1.07
1.80
0.84
0.67
NK100A1
Pre
0.93
0.95
0.80
0.84
1.16
3.75
2.57
2.02
1.89
3.19
1.51
1.21
0.25
0.93
0.95
0.80
0.84
1.16
3.76
2.58
2.03
1.90
3.20
1.51
1.21
2.0
0.90
0.92
0.77
0.81
1.12
3.62
2.48
1.96
1.83
3.08
1.46
1.17
6.5
0.99
1.01
0.85
0.89
1.23
4.00
2.75
2.16
2.02
3.41
1.61
1.29
NKIOOA"
Pre
1.10
1.13
0.91
0.96
1.44
4.27
2.35
1.93
1.76
4.39
1.98
1.53
0.25
1.11
1.14
0.91
0.97
1.45
4.29
2.36
1.93
1.76
4.41
1.99
1.54
2.0
1.08
1.10
0.88
0.94
1.41
4.17
2.29
1.88
1.71
4.28
1.93
1.49
6.5
1.17
1.20
0.96
1.02
1.53
4.53
2.49
2.04
1.86
4.65
2.10
1.62
Notes:
a USGS 15302250.
Table 7-15. Measured minimum monthly pre-mining streamflow rates (m3/s) and estimated minimum monthly streamflow rates (m3/s) in the Pebble 0.25, 2.0, and 6.5 scenarios, for gages along Upper Talarik Creek.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
UT100E
Pre
0.09
0.09
0.08
0.07
0.10
0.15
0.14
0.12
0.11
0.17
0.16
0.12
0.25
0.09
0.08
0.07
0.06
0.10
0.14
0.13
0.11
0.10
0.16
0.15
0.12
2.0
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
6.5
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
UT100D
Pre
0.12
0.10
0.12
0.11
0.22
0.23
0.21
0.22
0.20
0.33
0.31
0.22
0.25
0.11
0.10
0.11
0.10
0.20
0.21
0.19
0.20
0.18
0.30
0.28
0.20
2.0
0.07
0.06
0.07
0.06
0.12
0.13
0.12
0.12
0.11
0.18
0.17
0.12
6.5
0.02
0.02
0.02
0.02
0.03
0.03
0.03
0.03
0.03
0.05
0.04
0.03
UT100C2
Pre
0.53
0.47
0.53
0.50
0.91
1.46
1.46
1.35
1.29
1.71
1.34
0.91
0.25
0.52
0.46
0.52
0.48
0.89
1.43
1.43
1.32
1.26
1.67
1.32
0.89
2.0
0.47
0.42
0.47
0.44
0.81
1.31
1.31
1.21
1.15
1.53
1.21
0.81
6.5
0.42
0.37
0.42
0.39
0.72
1.16
1.17
1.07
1.02
1.36
1.07
0.72
UT100C1
Pre
0.80
0.73
0.80
0.76
1.25
1.57
1.37
1.58
1.52
2.24
2.04
1.25
0.25
0.78
0.71
0.78
0.75
1.23
1.54
1.35
1.55
1.49
2.20
2.00
1.23
2.0
0.73
0.66
0.73
0.70
1.14
1.44
1.25
1.44
1.39
2.05
1.86
1.14
6.5
0.66
0.60
0.66
0.63
1.04
1.30
1.14
1.31
1.26
1.86
1.69
1.04
UT100C
Pre
1.55
1.48
1.37
1.42
2.02
2.85
2.50
2.40
2.37
3.03
2.36
1.83
0.25
1.53
1.46
1.35
1.40
1.99
2.81
2.46
2.36
2.33
2.98
2.32
1.80
2.0
1.44
1.37
1.27
1.32
1.87
2.64
2.31
2.22
2.19
2.80
2.18
1.69
6.5
1.32
1.26
1.17
1.21
1.72
2.43
2.13
2.04
2.02
2.58
2.01
1.56
UT119A
Pre
0.72
0.71
0.72
0.71
0.63
0.62
0.65
0.62
0.67
0.69
0.78
0.74
0.25
0.65
0.65
0.65
0.65
0.58
0.56
0.59
0.57
0.61
0.62
0.71
0.67
2.0
0.63
0.63
0.63
0.63
0.56
0.54
0.57
0.55
0.59
0.60
0.68
0.65
6.5
0.57
0.56
0.57
0.56
0.50
0.49
0.51
0.49
0.53
0.54
0.61
0.58
UTIOOB"
Pre
2.09
1.98
2.09
2.04
2.83
2.58
2.55
2.97
2.83
3.82
3.68
2.83
0.25
2.04
1.93
2.04
1.99
2.76
2.51
2.49
2.90
2.76
3.73
3.59
2.76
2.0
1.93
1.83
1.93
1.88
2.61
2.38
2.35
2.74
2.61
3.52
3.39
2.61
6.5
1.78
1.68
1.78
1.73
2.40
2.19
2.16
2.52
2.40
3.24
3.12
2.40
Notes:
a USGS 15300250.
NA = not applicable: UT100E would be blocked by the waste rock pile in the Pebble 2.0 scenario (Figure 7-15) and by the mine pit in the Pebble 6.5 scenario (Figure 7-16).
Bristol Bay Assessment
7-43
January 2014
-------
Chapter 7
Mine Footprint
Table 7-16. Pre-mining watershed areas, mine footprint areas, and flows in the mine scenario watersheds for the Pebble 0.25 scenario.
Stream and Gage
Pre-Mining
Watershed Area (km2)
Mean Annual Unit Runoff
(m3/s*km2)
Runoff per Unit Area
(m3/yr*m2)
Mean Annual Runoff
(mVyr)
Returned Flow in Each Pathway (%)
Volume from Water Balance (m3/yr)
Total Mine Footprint
Drainage Area (km2)
Mine Footprint other
than TSF, NAG, or PAG
(km2)
TSF 1 Footprint (km2)
NAG Waste Rock
Footprint (km2)
PAG Waste Rock
Footprint (km2)
85.9
10,909,000
Flow Volume Returned
through WWTP (m3/yr)a
8.8
1,113,000
Flow Volume Returned
as TSF Leakage (m3/yr)
75.3
676,000
Flow Volume Returned
as NAG Waste Rock
Leachate (m3/yr)
0.0
0
Flow Volume Returned
as PAG Waste Rock
Leachate (m3/yr)
Operational Flows
Flow Volume Remaining
Due to Mine Footprint
with No Returns (m3/yr)
Captured Flow Volume
Returned from Footprint
(m3/yr)
Total Flow Volume in
Stream During
Operations (m3/yr)
Change in Average
Annual Runoff (%)
South Fork Koktuli River
SK100G
SK100F
SK100CP2b (total runoff)
SK100CP2b (losses to UTC)C
SK100CP2b (net streamflow at gage)
SK124A
SK124CPb
SK100C
SK100CPlb
SK119A
SK119CPb
SK100B1
SK100B"
14
31
54
54
54
22
24
99
99
28
30
141
179
0.026
0.026
0.026
0.009
0.018
0.024
0.024
0.013
0.013
0.036
0.036
0.026
0.029
0.82
0.83
0.83
0.28
0.55
0.76
0.76
0.42
0.42
1.12
1.12
0.82
0.91
11,618,000
25,842,000
44,681,000
-14,894,000
29,788,000
16,811,000
17,937,000
41,858,000
42,029,000
31,268,000
33,124,000
115,110,000
162,122,000
8.0
8.8
8.8
8.8
8.8
0.0
0.0
8.8
8.8
0.0
0.0
8.8
8.8
7.5
<0.1
-
-
-
-
-
-
-
-
-
-
-
-
-
0.5
0.8
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
5,454,000
-
-
-
-
-
-
-
-
-
-
-
207,000
350,000
-
-
-
-
-
-
-
-
-
-
-
-
-
5,080,000
18,499,000
-12,446,000
24,892,000
16,811,000
17,937,000
38,117,000
38,288,000
31,268,000
33,124,000
107,911,000
154,112,000
207,000
556,000
-185,000
371,000
5,454,000
5,454,000
5,825,000
5,825,000
-
5,825,000
5,825,000
5,287,000
19,055,000
-
25,263,000
22,265,000
23,391,000
43,942,000
44,113,000
31,268,000
33,124,000
113,737,000
159,937,000
-54
-26
-
-15
32
30
5
5
0
0
-1
-1
North Fork Koktuli River
NK119A
NK119CP2b
NK119B
NK119CPlb
NK100C
NK100B
NK100A1
NK100A6
20
22
11
33
65
99
222
279
0.034
0.034
0.012
0.027
0.020
0.024
0.026
0.025
1.08
1.08
0.38
0.85
0.64
0.77
0.82
0.79
21,515,000
24,155,000
4,081,000
28,431,000
41,853,000
76,408,000
182,297,000
220,715,000
6.8
6.9
0.3
7.2
0.0
7.2
7.3
7.3
<0.1
0.1
0.3
<0.1
<0.1
<0.1
<0.1
-
6.5
-
-
-
0.3
-
-
-
-
-
-
-
-
-
5,454,000
-
-
1,113,000
-
-
-
120,000
-
-
-
-
-
-
-
14,146,000
16,691,000
3,975,000
22,279,000
41,828,000
70,826,000
176,335,000
214,981,000
1,233,000
1,233,000
-
1,233,000
5,454,000
6,687,000
6,687,000
6,687,000
15,378,000
17,923,000
3,975,000
23,512,000
47,282,000
77,513,000
183,022,000
221,668,000
-29
-26
-3
-17
13
1
<1
<1
Upper Talarik Creek
UT100E
UT100D
UT100C2
UT100C1
UT100C
UT119A (local runoff)
UT119A (gains from SFK)C
UT119A (net streamflow at gage)
UT100B'
10
31
133
159
185
10
10
10
222
0.027
0.025
0.022
0.022
0.024
0.033
0.046
0.079
0.028
0.84
0.78
0.70
0.68
0.76
1.04
1.45
2.48
0.88
7,996,000
24,201,000
92,734,000
107,971,000
141,213,000
10,655,000
14,894,000
25,549,000
196,182,000
0.6
2.8
2.8
2.8
2.8
0.0
0.0
0.0
2.8
0.6
2.2
0.0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
7,474,000
22,008,000
90,768,000
106,050,000
139,053,000
10,655,000
12,446,000
23,101,000
191,238,000
-
-
-
185,000
185,000
185,000
7,474,000
22,008,000
90,768,000
106,050,000
139,053,000
-
23,286,000
191,423,000
-7
-9
-2
-2
-2
-
-9
-2
Notes:
Dashes (-) indicate that values are either not applicable or are equal to zero.
a WWTP discharges 50% of flow to South Fork Koktuli River, 50% of streamflow to North Fork Koktuli River (no WWTP flows are directed to Upper Talarik Creek).
b Confluence point where virtual gage was created because physical gage does not exist.
c 1/3 of total return flow is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location. Interbasin transfer flows are represented by negative flow values from SK100CP2 (losses to UTC) and equivalent positive flow values for UT119A (gains from SFK).
d USGS 15302200.
e USGS 15302250.
' USGS 15300250.
TSF = tailings storage facility; PAG = potentially acid-generating; NAG = non-acid-generating; WWTP = wastewater treatment plant; UTC = Upper Talarik Creek; SFK = South Fork Koktuli.
Bristol Bay Assessment
7-44
January 2014
-------
Chapter 7
Mine Footprint
Table 7-17. Pre-mining watershed areas, mine footprint areas, and flows in the mine scenario watersheds for the Pebble 2.0 scenario.
Stream and Gage
Pre-Mining
Watershed Area
(km2)
Mean Annual Unit
Runoff (m3/s*km2)
Runoff per Unit Area
(m3/yr*m2)
Mean Annual Runoff
(m3/yr)
Returned Flow in Each Pathway (%)
Volume from Water Balance (m3/yr)
Total Mine Footprint
Drainage Area
(km2)
Mine Footprint other
than TSF, NAG, or PAG
(km2)
E
Q.
5
£
•H sr
u. E
£ !
NAG Waste Rock
Footprint (km2)
PAG Waste Rock
Footprint (km2)
66.7
10,304,000
Flow Volume Returned
through WWTP (m3/yr)a
15.2
2,351,000
Flow Volume Returned
as TSF Leakage (m3/yr)
16.7
2,576,000
Flow Volume Returned
as NAG Waste Rock
Leachate (m3/yr)
1.4
216,000
Flow Volume Returned
as PAG Waste Rock
Leachate (m3/yr)
Operational Flows
Flow Volume Remaining
Due to Mine Footprint
with No Returns (m3/yr)
Captured Flow Volume
Returned from Footprint
(m3/yr)
Total Flow Volume in
Stream During
Operations (m3/yr)
Change in Average
Annual Runoff (%)
South Fork Koktuli River
SK100G
SK100F
SK100CP2b (total runoff)
SK100CP2" (losses to UTC)=
SK100CP2b (net streamflow at gage)
SK124A
SK124CPb
SK100C
SKIOOCPI"
SK119A
SK119CPb
SK100B1
SKIDDED
14
31
54
54
54
22
24
99
99
28
30
141
179
0.026
0.026
0.026
0.009
0.018
0.024
0.024
0.013
0.013
0.036
0.036
0.026
0.029
0.82
0.83
0.83
0.28
0.55
0.76
0.76
0.42
0.42
1.12
1.12
0.82
0.91
11,618,000
25,842,000
44,681,000
-14,894,000
29,788,000
16,811,000
17,937,000
41,858,000
42,029,000
31,268,000
33,124,000
115,110,000
162,122,000
11.2
12.6
12.6
12.6
12.6
0.1
0.1
12.7
12.7
0.6
0.6
13.3
13.3
9.2
0.2
0.1
-
<0.1
-
0.1
-
-
<0.1
-
0.1
1.5
1.2
0.1
-
0.4
0.5
<0.01
-
-
-
-
5,152,000
-
-
-
-
-
-
2,000
-
-
-
21,000
633,000
507,000
22,000
-
-
-
151,000
213,000
3,000
-
-
-
-
2,420,000
15,389,000
-11,409,000
22,819,000
16,702,000
17,829,000
36,472,000
36,643,000
30,602,000
32,458,000
104,262,000
150,051,000
846,000
1,356,000
-452,000
904,000
5,175,000
5,175,000
6,079,000
6,079,000
172,000
172,000
6,251,000
6,251,000
3,266,000
16,745,000
23,723,000
21,878,000
23,004,000
42,552,000
42,722,000
30,774,000
32,630,000
110,513,000
156,302,000
-72
-35
-20
30
28
2
2
-2
-1
-4
-4
North Fork Koktuli River
NK119A
NK119CP2b
NK119B
NK119CPlb
NK100C
NK100B
NK100A1
NK100A6
20
22
11
33
65
99
222
279
0.034
0.034
0.012
0.027
0.020
0.024
0.026
0.025
1.08
1.08
0.38
0.85
0.64
0.77
0.82
0.79
21,515,000
24,155,000
4,081,000
28,431,000
41,853,000
76,408,000
182,297,000
220,715,000
14.9
15.3
1.2
16.5
0.2
16.6
17.3
17.3
0.1
0.4
1.1
-
0.2
<0.1
0.1
13.9
<0.1
-
-
-
0.1
0.9
<0.1
<0.1
-
-
0.5
-
-
-
-
-
-
-
-
-
5,152,000
2,305,000
1,000
-
-
-
23,000
402,000
13,000
3,000
-
-
204,000
-
-
-
-
-
5,405,000
7,627,000
3,638,000
14,346,000
41,753,000
63,577,000
168,068,000
207,031,000
2,707,000
2,720,000
3,000
2,723,000
5,152,000
7,875,000
8,102,000
8,102,000
8,111,000
10,347,000
3,641,000
17,069,000
46,905,000
71,452,000
176,169,000
215,132,000
-62
-57
-11
-40
12
-6
-3
-3
Upper Talarik Creek
UT100E
UT100D
UT100C2
UT100C1
UT100C
UT119A (local runoff)
UT119A (gains from SFK)C
UT119A (net streamflow at gage)
UT100B'
10
31
133
159
185
10
10
10
222
0.027
0.025
0.022
0.022
0.024
0.033
0.046
0.079
0.028
0.84
0.78
0.70
0.68
0.76
1.04
1.45
2.48
0.88
7,996,000
24,201,000
92,734,000
107,971,000
141,213,000
10,655,000
14,894,000
25,549,000
196,182,000
3.2
14.5
14.6
14.6
14.6
14.6
3.2
9.8
0.1
-
-
-
-
-
-
-
-
1.5
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
642,000
-
-
-
-
-
-
-
-
5,290,000
12,839,000
82,573,000
98,042,000
130,049,000
10,655,000
11,409,000
22,064,000
179,795,000
-
642,000
642,000
642,000
642,000
452,000
452,000
1,094,000
5,290,000
13,481,000
83,215,000
98,684,000
130,691,000
22,516,000
180,889,000
-34
-44
-10
-9
-7
-12
-8
Notes:
Dashes (-) indicate that values are either not applicable or are equal to zero. UT100E is blocked by the mine footprint in this scenario.
a WWTP discharges 50% of flow to South Fork Koktuli River, 50% of flow to North Fork Koktuli River (no WWTP flows are directed to Upper Talarik Creek).
b Confluence point where virtual gage was created because physical gage does not exist.
c 1/3 of total return flow from is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location. Interbasin transfer flows are represented by negative flow values from SK100CP2 (losses to UTC) and equivalent positive flow values for UT119A (gains from SFK).
d USGS 15302200.
e USGS 15302250.
' USGS 15300250.
TSF = tailings storage facility; PAG = potentially acid-generating; NAG = non-acid-generating; WWTP = wastewater treatment plant; UTC = Upper Talarik Creek; SFK = South Fork Koktuli.
Bristol Bay Assessment
7-45
January 2014
-------
Chapter 7
Mine Footprint
Stream and Gage
Pre-Mining
Watershed Area
(km2)
Mean Annual Unit
Runoff (m3/s*km2)
Runoff per Unit
Area (m3/yr*m2)
Mean Annual
Runoff
(mVyr)
Returned Flow in Each Pathway (%)
Volume from Water Balance (m3/yr)
(D
UO
CB
c
E
o Q
II
— ^ ^r*
(n +^ (C 7-
•S O (D P
o o % 5
1— U- <. Zt-
Mine Footprint
other than TSF,
NAG, or PAG (km2)
TSF 1 Footprint
(km2)
TSF 2 Footprint
(km2)
+->
,c
Q.
*
O
u_
CO £P
u. C
£ I.
NAG Waste Rock
Footprint (km2)
PAG Waste Rock
Footprint (km2)
79.4
50,988,000
Flow Volume
Returned through
WWTP (m3/yr)a
11.2
7,203,000
Flow Volume
Returned as TSF
Leakage(m3/yr)
7.7
4,971,000
Flow Volume
Returned as NAG
Waste Rock
Leachate (m3/yr)
1.6
1,032,000
Flow Volume
Returned as PAG
Waste Rock
Leachate (m3/yr)
Operational Flows
Flow Volume
Remaining Due to
Mine Footprint with
No Returns (m3/yr)
Captured Flow
Volume Returned
from Footprint
(m3/yr)
Total Flow Volume
in Stream During
Operations (m3/yr)
Change in Average
Annual Runoff (%)
South Fork Koktuli River
SK100G
SK100F
SK100CP2b (total runoff)
SK100CP2b (losses to UTC)C
SK100CP2b (net flow at gage)
SK124A
SK124CPb
SK100C
SKIOOCPI"
SK119A
SK119CPb
SK100B1
SK100Bd
14
31
54
54
54
22
24
99
99
28
30
141
179
0.026
0.026
0.026
0.009
0.018
0.024
0.024
0.013
0.013
0.036
0.036
0.026
0.029
0.82
0.83
0.83
0.28
0.55
0.76
0.76
0.42
0.42
1.12
1.12
0.82
0.91
11,618,000
25,842,000
44,681,000
-14,894,000
29,788,000
16,811,000
17,937,000
41,858,000
42,029,000
31,268,000
33,124,000
115,110,000
162,122,000
14.0
22.1
22.1
22.1
22.1
11.4
11.4
33.6
33.6
18.0
19.2
54.3
54.3
14.0
2.5
-
-
-
0.1
0.1
-
<0.1
-
-
-
-
-
<0.1
0.1
-
-
-
-
-
-
-
1.8
<0.1
17.2
0.3
0.9
-
0.1
-
-
7.8
-
-
-
3.0
<0.1
-
-
1.7
0.1
0.6
1.0
0.6
-
2.4
-
-
-
-
-
-
-
-
-
25,494,000
-
-
-
20,000
-
-
-
1,626,000
2,930,000
50,000
145,000
-
1,278,000
5,000
-
-
713,000
54,000
242,000
413,000
260,000
-
1,032,000
-
-
-
-
-
-
95,000
7,480,000
26,309,000
-8,770,000
17,540,000
8,216,000
9,342,000
27,627,000
27,798,000
11,091,000
11,537,000
70,839,000
112,863,000
2,330,000
2,335,000
-778,000
1,557,000
27,833,000
27,833,000
29,443,000
29,443,000
3,171,000
3,635,000
33,482,000
33,482,000
9,810,000
-
-
19,096,000
36,049,000
37,175,000
57,070,000
57,241,000
15,172,000
104,322,000
146,346,000
-62
-
-
-36
114
107
36
36
-54
-9
-10
North Fork Koktuli River
NK119A
NK119CP2b
NK119B
NK119CPlb
NK100C
NK100B
NK100A1
NK100A6
20
22
11
33
65
99
222
279
0.034
0.034
0.012
0.027
0.020
0.024
0.026
0.025
1.08
1.08
0.38
0.85
0.64
0.77
0.82
0.79
21,515,000
24,155,000
4,081,000
28,431,000
41,853,000
76,408,000
182,297,000
220,715,000
14.9
15.3
3.3
18.6
0.5
19.1
19.8
19.8
0.1
0.4
2.7
0.5
-
0.1
-
13.9
<0.1
-
-
0.1
-
-
-
-
-
0.3
-
-
-
-
0.9
<0.1
0.3
-
-
0.5
-
-
-
-
-
25,494,000
-
-
-
2,360,000
1,000
48,000
-
-
23,000
-
402,000
13,000
144,000
-
-
204,000
-
-
-
-
-
5,405,000
7,627,000
2,812,000
12,506,000
41,559,000
61,683,000
166,049,000
205,090,000
2,762,000
2,775,000
192,000
2,967,000
25,494,000
28,461,000
28,688,000
28,688,000
8,167,000
10,402,000
3,004,000
15,473,000
67,053,000
90,144,000
194,738,000
233,778,000
-62
-57
-26
-46
60
18
7
6
Upper Talarik Creek
UT100E
UT100D
UT100C2
UT100C1
UT100C
UT119A (local runoff)c
UT119A (gains from SFK)
UT119A (net flow at gage)
UT100B'
10
31
133
159
185
10
10
10
222
0.027
0.025
0.022
0.022
0.024
0.033
0.046
0.079
0.028
0.84
0.78
0.70
0.68
0.76
1.04
1.45
2.48
0.88
7,996,000
24,201,000
92,734,000
107,971,000
141,213,000
10,655,000
14,894,000
25,549,000
196,182,000
7.4
27.8
29.0
29.0
29.0
-
-
-
29.1
6.6
18.7
0.8
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.8
1.7
0.4
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
346,000
739,000
160,000
-
-
-
-
-
-
-
-
-
-
1,779,000
2,398,000
72,570,000
88,266,000
119,058,000
10,655,000
8,770,000
19,425,000
164,453,000
346,000
1,085,000
1,245,000
1,245,000
1,245,000
-
778,000
778,000
2,023,000
2,125,000
3,482,000
73,815,000
89,511,000
120,303,000
-
-
20,203,000
166,476,000
-73
-86
-20
-17
-15
-
-
-21
-15
Notes:
Dashes (-) indicate that values are either not applicable or are equal to zero. UT100E is blocked and SK100G and SK119A are eliminated by the mine footprint in this scenario.
a WWTP discharges 50% of flow to South Fork Koktuli River, 50% of flow to North Fork Koktuli River (no WWTP flows are directed to Upper Talarik Creek).
b Confluence point where virtual gage was created because physical gage does not exist.
c 1/3 of total return flow from is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location. Interbasin transfer flows are represented by negative flow values from SK100CP2 (losses to UTC) and equivalent positive flow values for UT119A (gains from SFK).
d USGS 15302200.
e USGS 15302250.
' USGS 15300250.
TSF = tailings storage facility; PAG = potentially acid-generating; NAG = non-acid-generating; WWTP = wastewater treatment plant; UTC = Upper Talarik Creek; SFK = South Fork Koktuli.
Bristol Bay Assessment
7-46
January 2014
-------
Chapter 7
Mine Footprint
Figure 7-17. Monthly mean streamflows for stream gages in the (A) South Fork Koktuli River,
(B) North Fork Koktuli River, and (C) Upper Talarik Creek watersheds, based on water years 2004
through 2010.
A
B
•SK100G
•SK100F
•SK100C
•SK100B1
•SK100B
•SK119A
SK124A
•NK119A
•NK100B
•NK100A1
•NK100A
•NK119B
•NK100C
•UT100D
•UT100C1
•UT100C
•UTIOOB
-UT100E
-UT100C2
LIT119A
Bristol Bay Assessment
7-47
January 2014
-------
Chapter 7 Mine Footprint
For the three mine scenarios, it was assumed that some water captured from each mine footprint would
be treated and reintroduced to downstream areas. For the Pebble 0.25, 2.0, and 6.5 scenarios, we
estimated that 76.3, 37.5, and 70.5% of the total water captured, respectively, would be reintroduced
(Table 6-3). Figures 6-8 through 6-10 illustrate the various flowpaths expected in the three mine
scenarios. For each of the watersheds, reintroduced flow was returned to the appropriate gage based on
the expected flowpath as defined by the mine scenarios. Some upper tributaries would experience
reduced streamflows from watershed area losses, whereas others would experience increased annual
runoff from mining operation discharges.
Although some surface runoff might be collected, most of the precipitation in the drawdown zone would
flow as groundwater into the mine pit and be removed by pumping to the WWTP. Much of the flow from
components outside the drawdown zone, such as leachate from TSFs and waste rock piles, would be
captured and directed to the WWTP, but some would escape the collection systems and flow back to the
downstream receiving waters (Tables 7-16 through 7-18, Figures 6-8 through 6-10). It is important to
note that the WWTP is assumed to discharge to the South and North Fork Koktuli River watersheds via
the WWTP outfalls (after Ghaffari et al. 2011), so no treated flow would be reintroduced to streams in
the Upper Talarik Creek watershed. An area of interbasin groundwater transfer has been observed
between the South Fork Koktuli River and Upper Talarik Creek (PLP 2011: Chapter 7). This transfer was
accounted for by allowing one-third of the flow at gage SK100F to transfer to gage UT119A (Tables 7-16
through 7-18, Figures 6-8 through 6-10). The spatial extent of these projected changes in streamflow
and implications for fish and aquatic habitat are discussed in Section 7.3.2.
7.3.1.1 Pebble 0.25 Scenario
Water balance estimates for the Pebble 0.25 scenario considered an operational facility that intercepts
precipitation from a footprint encompassing portions of the mine scenario watersheds (Table 7-16,
Figure 7-14). Based on these conditions, we estimate that in each watershed the uppermost gages
closest to the mine footprint would experience the most significant streamflow reductions. Overall, it is
projected that 76.3% of captured watershed flows would be returned (Table 6-3), but the location of
return would vary depending on mine needs for process water and the location of mine facilities and
water treatment (Table 7-16). In the Upper Talarik Creek watershed in the Pebble 0.25 scenario,
streamflow would be reduced by 7% at gage UT100E and 9% at gage UT100D due to capture in the mine
footprint. The most significant streamflow reductions in the South Fork Koktuli River would be expected
at gages SK100G (54%) and SK100F (26%) (Table 7-16). In the North Fork Koktuli River, the greatest
changes would be expected at gage NK119A (29% reduction) (Table 7-16), as much of the watershed
would be occupied by TSF 1 (Figure 7-14).
Streamflow reductions due to water capture in the mine footprint would be partially offset by water
return via the WWTP, leakage through the TSF, and leaching through the waste rock piles. Water
balance calculations for these water budget components are described in Chapter 6. Excess captured
water would be treated at the WWTP and discharged upstream of gage SK124A in the South Fork
Koktuli River and gage NK100C in the North Fork Koktuli River (Figure 7-14). It is assumed that the
Bristol Bay Assessment 748 January 2014
-------
Chapter 7 Mine Footprint
WWTP would discharge equally to both outfalls, creating a 50/50 volume split for treated flows on an
annual basis, but that on-site storage would allow management of environmental streamflows to match
seasonal hydrographs to the degree possible. Flows from the WWTP outfalls would be projected to
increase streamflows by 32% at gage SK124A, in a tributary to the South Fork Koktuli River. In the
North Fork Koktuli River watershed, streamflows would be projected to increase by 13% at gage
NK100C, downstream of the WWTP outfall. In the mainstem South and North Fork Koktuli Rivers
downstream of these points, WWTP outfall flows (approximately 5.4 million m3/year from each outfall)
(Table 7-16), leakage from the TSF, and waste rock leaching would partially offset streamflow
reductions expected from water capture within the mine footprint. Projected streamflow changes for
gages farther downstream of the WWTP outfalls are within 5% of pre-project streamflows (Tables 7-16
and 7-19, Figure 7-14).
Because of the natural interbasin streamflow transfer from the South Fork Koktuli River watershed to
the Upper Talarik Creek watershed (described above), decreased streamflows in the South Fork Koktuli
River resulting from capture by the mine footprint would translate to decreased rates of interbasin
transfer. As a result, there would be a projected 9% decrease in streamflow to the tributary of Upper
Talarik Creek where the interbasin transfer flows emerge (gage UT119A) (Tables 7-16 and 7-19, Figure
7-14).
7.3.1.2 Pebble 2.0 Scenario
In the Pebble 2.0 scenario, area lost to the mine footprint would increase from the addition of a second
or expanded waste rock pile that would occupy much of the Upper Talarik Creek valley between gages
UT100E and UT100D (Figure 7-15). An expanded groundwater drawdown zone would develop around
the larger mine pit and further reduce water flowing to surrounding streams, and TSF 1 would expand
in size (Figure 7-15). Approximately 37.5% of the total water captured would be returned to the three
watersheds (Table 6-3). However, as in the Pebble 0.25 scenario described above, flow returns in the
upper watersheds via the WWTP outfalls would not necessarily be returned to their source stream
reaches.
After accounting for water captured in the footprint, leakage, leachate, and reintroduced water,
streamflow reductions in Upper Talarik Creek would be most severe for gage UT100D (44% reduction)
(Tables 7-17 and 7-19). In the South Fork Koktuli River, gages SK100G, SK100F, and confluence point
SK100CP2 would experience reductions of 72, 35, and 20%, respectively. In the North Fork Koktuli
River, the most severe effects would be seen in the watershed occupied by TSF 1, with gages on this
tributary predicted to experience streamflow reductions ranging from 40 to 62% (Tables 7-17 and 7-19,
Figure 7-15). Contributions of the WWTP flow to the South Fork Koktuli River watershed would cause
an increase in streamflow at gage SK124A (30%) and the associated confluence point SK124CP (28%).
WWTP contributions to the North Fork Koktuli River watershed would cause a 12% streamflow
increase at gage NK100C. At the lowermost gages in each watershed, projected reductions in streamflow
would be 8% (Upper Talarik Creek), 4% (South Fork Koktuli River), and 3% (North Fork Koktuli River)
(Tables 7-17 and 7-19).
Bristol Bay Assessment 749 January 2014
-------
Chapter 7
Mine Footprint
Table 7-19. Estimated changes in streamflow (%) and subsequent stream lengths affected (km) in
the mine scenario watersheds in the Pebble 0.25, Pebble 2.0, and Pebble 6.5 scenarios. Italics
indicates changes greater than 10% (minor effects on salmon populations expected); bold indicates
changes greater than 20% (moderate to major effects on salmon populations expected).
Stream and
Gage
Pebble 0.25
Estimated
Change in
Streamflow
Stream Length
Affected
Pebble 2.0
Estimated
Change in
Streamflow
Stream Length
Affected
Pebble 6.5
Estimated
Change in
Streamflow
Stream
Length
Affected
South Fork Koktuli River— Ma instem
SK100G
SK100F
SK100CP2
SK100C
SK100CP1
SK100B1
SK100B3
-54
-26
-15
5
5
-1
-1
1.9
3.3
10.7
6.3
1.2
4.3
4.5
-72
-35
-20
2
2
-4
-4
0.5
3.3
10.7
6.3
1.2
4.3
4.5
NA
-62
-36
36
36
-9
-10
NA
0.8
10.7
6.3
1.2
4.3
4.5
South Fork Koktuli River— Tributaries
SK119A
SK119CP
SK124A
SK124CP
0
0
32
30
7.0
1.6
5.0
2.6
-2
-1
30
28
6.7
1.6
5.0
2.6
NA
-54
114
107
NA
0.7
4.2
2.6
North Fork Koktuli River— Ma instem
NK100O
NK100B
NK100A1
NK100AC
13
1
0
0
4.5
0.8
20.4
8.4
12
-6
-3
-3
4.5
0.8
20.4
8.4
60
18
7
6
4.5
0.8
20.4
8.4
North Fork Koktuli River-Tributaries
NK119A
NK119CP2
NK119B
NK119CP1
-29
-26
-3
-17
0.8
1.3
6.8
0.4
-62
-57
-11
-40
0.7
1.3
6.8
0.4
-62
-57
-26
-46
0.7
1.3
6.5
0.4
Upper Talarik Creek— Mainstem
UT100E
UT100D
UT100C2
UT100C1
UT100C
UTlOOBd
-7
-9
-2
-2
-2
-2
2.3
7.1
6.1
6.9
7.5
4.3
NA
-44
-10
-9
-7
-8
NA
2.1
6.1
6.9
7.5
4.3
NA
-86
-20
-17
-15
-15
NA
0.3
6.1
6.9
7.5
4.3
Upper Talarik Creek Tributary— Tributaries
UT119A
-9
6.5
-12
6.5
-21
6.5
Notes:
Stream lengths are typically calculated from the gage upstream to the next gage or the mine footprint (but see below); stream lengths affected
do not include portions of stream lost in the pit drawdown zone.
For gages UT100D, SK100G, SK100F, SK119A, SK124A, and NK119A, stream lengths include mainstem length upstream to edge of mine
footprint only, and do not include upstream lengths, including tributaries, that would be blocked or eliminated by the mine footprint.
a USGS 15302200.
b Upstream to wastewater treatment plant outfall point.
c USGS 15302250.
d USGS 15300250.
NA = not applicable; the stream at the gage would be eliminated or blocked by the mine footprint
Bristol Bay Assessment
7-50
January 2014
-------
Chapter 7 Mine Footprint
7.3.1.3 Pebble 6.5 Scenario
In the Pebble 6.5 scenario, area lost to the mine footprint would increase with inclusion of a larger pit
and its associated drawdown zone, a substantially larger waste rock pile, and the development of TSF 2
on a tributary of South Fork Koktuli River upstream of gage SKI GOBI and TSF 3 on a tributary upstream
of gage SK124A (Table 7-18, Figure 7-16). Gage SK100G would be eliminated under the Pebble 6.5 waste
rock piles, gage UT100E would be isolated upstream of the mine footprint, and gage SK119A would be
buried under the TSF 2 dam (Figure 7-16). Although the larger mine footprint would result in the
capture of much greater quantities of water in the Pebble 6.5 scenario, annual water consumption would
not be appreciably higher than in the Pebble 2.0 scenario. Thus, an estimated 70.5% of the captured
water would be available for reintroduction to streams (Table 6-3). The net effects of lost effective
watershed area and the reintroduction of treated water would result in streamflow reductions that
would be most severe for gages UT100D (86% reduction), SK100F (62% reduction), and NFK119A
(62% reduction) (Tables 7-18 and 7-19).
WWTP flows would be increased greatly over the Pebble 2.0 scenario and would create increased
streamflow at SK124A (114%) and SK124CP (107%). This increase would continue to influence
streamflows downstream to gage SK100C (36% increase), but the large reduction attributed to the TSF
on the tributary measured by gage SK119A again creates streamflow deficits downstream at gages
SK100B1 and SK100B relative to pre-mining conditions (9 and 10% reductions, respectively) (Tables 7-
18 and 7-19, Figure 7-16). In the North Fork Koktuli River watershed, WWTP contributions would lead
to streamflow increases of 60% at gage NK100C and increased streamflows at all downstream gages
(Table 7-18). Upper Talarik Creek would experience streamflow reductions of 15% or more at all
mainstem gages. Upper Talarik Creek tributary gage UT119A would experience a 21% decrease in
streamflow due to reduced interbasin transfer resulting from streamflow losses in the South Fork
Koktuli River watershed. At the lowermost gages in each watershed, projected streamflow changes
would be a 15% reduction for Upper Talarik Creek, a 10% reduction for the South Fork Koktuli River,
and a 6% increase for the North Fork Koktuli River (Tables 7-18 and 7-19).
7.3.1.4 Post-Closure
After the mine closes, pit dewatering would cease, leading to pit filling. As the pit fills, water from the pit
that had been returned to streams via pumping to the WWTP would no longer be available for
streamflow. This period is projected to last from about 20 years for the Pebble 0.25 scenario to over 200
years for the Pebble 6.5 scenario, after which the pit would approach equilibrium with surrounding
groundwater. The pit water level could be controlled by pumping or gravity drainage to maintain a
hydraulic gradient toward the pit for as long as water needed treatment. When treatment was no longer
necessary and active control was abandoned, water from the filled mine pit would eventually discharge
to down-gradient streams, ponds, and wetlands (Section 6.3) under steady-state flow conditions. Given
uncertainties in the post-closure water balance, we have not attempted to estimate streamflows during
that period.
Bristol Bay Assessment 751 January 2014
-------
Chapter 7 Mine Footprint
7.3.1.5 Uncertainties and Assumptions
Our assessment of streamflow changes distributes losses according to the percentage of the area lost to
the mine footprint in a given watershed. The analysis uses flow per unit area derived from stream gage
data, and allocates water routing through the three mine scenarios based on decisions about mine
processes that will consume and reintroduce water to the watersheds. We assume that water captured
within the footprint and requiring treatment would be routed through the WWTP and discharged to the
two locations specified by Ghaffari et al. (2011). We assume that reduced streamflows would follow the
same spatial patterns of gaining or losing groundwater reaches as would initial (pre-mine) conditions.
We acknowledge, however, that mine operations could alter the relative importance of groundwater
flowpaths, and thus result in a different spatial distribution of streamflow changes than we have
reported.
7.3.2 Exposure-Response: Streamflow
Water from streams originating upstream of the mine footprints (i.e., blocked streams) could be
captured at the footprint for use or stored on site for eventual treatment and return to the stream
downstream of the footprint, either directly or via the WWTP. Water from blocked streams could be
returned to downstream stream segments via diversion channels or pipes. Habitat upstream of the
footprint would no longer be accessible to fish downstream because of the inability offish to move
upstream through diversion channels or pipes.
7.3.2.1 Altered Streamflow Regimes
Altered streamflows can have various effects on aquatic life. Short-term effects include reduced habitat
availability resulting from water withdrawal (effects on winter habitat reviewed by West et al. 1992,
Cunjak 1996) and reduced habitat quality resulting from extreme and rapid fluctuations in streamflow if
withdrawals are intermittent (Curry et al. 1994, Cunjak 1996). Temporal variability in streamflows is a
natural feature of stream ecosystems (Poff et al. 1997), although the degree of variability differs
depending on hydrologic controls such as climate, geology, landform, human land use, and relative
groundwater contributions (Poff et al. 2006). Fish populations may be adapted to periodic disturbances
such as droughts and may quickly recover under improved hydrologic conditions, but this is contingent
on many factors (Matthews and Marsh-Matthews 2003). Longer-term effects of prolonged changes in
streamflow regime can have lasting impacts on fish populations (Lytle and Poff 2004).
The natural flow paradigm is widely supported and based on the premise that natural streamflow
variability, including the magnitude, frequency, timing, duration, rate of change, and predictability of
streamflow events and the sequence of streamflow conditions, is crucial to maintaining healthy aquatic
ecosystems (Postel and Richter 2003, Arthington et al. 2006, Poff et al. 2009). However, numerous
human demands can directly alter natural streamflows, potentially affecting ecosystem function and
structure. Guidelines for minimizing impacts of altered hydrologic regimes have been offered by several
researchers (Poff et al. 1997 and 2009, Richter 2010). Determining the natural streamflow regime is a
Bristol Bay Assessment 752 January 2014
-------
Chapter 7 Mine Footprint
data-intensive process, but it is crucial to understanding how to manage streamflows within a system
(Arthington et al. 2006).
Given the high likelihood of complex groundwater-surface water connectivity in the deposit area,
predicting and regulating flows to maintain key ecosystem functions associated with groundwater-
surface water exchange would be particularly challenging. PLP has invested in a relatively intensive
network of stream gages, water temperature monitoring sites, fish assemblage sampling sites,
groundwater monitoring wells, and geomorphic cross-section locations. The integration of information
gathered by these efforts will help identify relationships among surface-water flow, groundwater and
surface-water temperatures, and instream fish habitat (Bartholow 2010). However, until linkages
between biology, groundwater, surface water, and potential mining activities can be better evaluated,
predicted, and understood, a protective approach is warranted to maintain surface-water and
groundwater flows and natural streamflow regimes across the mine scenario watersheds.
The sustainability boundary approach offers such a protective approach for balancing the maintenance
of aquatic ecosystems with human demands (Richter et al. 2012). Under this approach, percentage-
based deviations from natural conditions are used to set streamflow alteration limits. These percentages
are based on the natural flow regime and do not focus on the more simplistic approach of setting a
percentage based on a high-streamflow or low-streamflow event. Rather than a salmon-specific
instream flow habitat model, this is a system-based approach targeting the entire aquatic ecosystem.
Numerous case studies have tested this type of approach, and the percentage bounds of streamflow
alteration around natural daily streamflow that caused measurable ecological harm were determined to
be similar regardless of geographic location (Richter et al. 2012). Based on these studies, Richter et al.
(2012) proposed that streamflow alteration be managed based on the following thresholds of daily
percentage alteration.
• Streamflow alteration below 10% would cause minor impacts on the ecosystem with a relatively
high level of ecosystem protection.
• Streamflow alteration of 11 to 20% would cause measurable changes in ecosystem structure and
minor impacts on ecosystem function.
• Streamflow alteration greater than 20% would cause moderate to major changes in ecosystem
structure and function. Increasing alteration beyond 20% would cause significant losses of
ecosystem structure and function.
Losses of ecosystem structure and function could include reduced habitat availability for salmon and
other stream fishes, particularly during low-streamflow periods (West et al. 1992, Cunjak 1996);
reductions in macroinvertebrate production (Chadwick and Huryn 2007); and increased stream habitat
fragmentation due to increased frequency and duration of stream drying. Increases in streamflow above
background levels could result in altered sediment transport dynamics with increased scour and
transport of gravels. Increased streamflows could also be associated with altered distributions of water
velocities favorable for various fish life stages. These alterations, depending on magnitude, could
significantly decrease salmon habitat quantity and quality in these watersheds (Figure 7-1).
Bristol Bay Assessment 753 January 2014
-------
Chapter 7
Mine Footprint
We compared predicted streamflows for the Pebble 0.25, 2.0, and 6.5 scenarios (Tables 7-10 through 7-
15) with the sustainability boundary limits of 10 and 20% streamflow alteration around mean monthly
flow. As an example, mean monthly streamflows for the South Fork Koktuli River at gage SK100F during
the pre-mining period, projected streamflows in the Pebble 0.25 scenario and the 10 and 20%
sustainability boundaries for the baseline streamflow are shown in Figure 7-18.
Figure 7-18. Monthly mean pre-mining streamflow for South Fork Koktuli River gage SK100F (bold
solid line), with 10 and 20% sustainability boundaries (gray lines) and projected monthly mean
streamflows, in the Pebble 0.25 scenario (dashed line).
2.5
2.0
1.0
0.0
SK100F
10%
-10%
20%
20%
Projected
We used this sustainability boundary approach to evaluate risks associated with potential streamflow
alterations throughout the mine scenario watersheds. To estimate the spatial extent of potential
deleterious streamflow alterations, we calculated the length of stream network upstream of the
uppermost stream gage to the edge of each mine footprint, and the length of each segment between
stream gages in each mine scenario watershed. This stream length is in addition to the length of stream
that would be eliminated or blocked by the mine footprint—that is, this and all subsequent references to
stream lengths affected by flow modification reflect stream lengths downstream of the mine footprint
for each scenario, and thus do not include stream lengths eliminated, blocked, or dewatered by each
footprint (Section 7.2). Table 7-19 summarizes estimated percent changes in streamflow at each gage
location, and the length of stream affected by each streamflow alteration in each mine scenario. Figures
7-14 through 7-16 illustrate the spatial extent and location of streamflow alterations in relation to gage
sites. These estimates are for direct effects only. Stream sections throughout the stream network could
be affected indirectly, via streamflow reductions downstream that could preclude use of downstream
habitats by fish that move seasonally between headwater and mainstem habitats. Similarly, these stream
Bristol Bay Assessment
7-54
January 2014
-------
Chapter 7 Mine Footprint
sections could be isolated by downstream flow reductions that reduce or eliminate the potential for fish
movement into those areas from downstream.
Pebble 0.25 Scenario
During operation of the Pebble 0.25 scenario, streamflow reductions exceeding 20% sustainability
boundaries would occur in 7 km of streams beyond the mine footprint. Substantial reductions in fish
habitat capacity and productivity could be expected for these streams. Streamflow increases greater
than 20% are expected for 8 km of streams downstream of the WWTP outfall, and would likely lead to
substantial changes in sediment dynamics and habitat suitability for fish. An additional 16 km of streams
would experience streamflow alterations of 13 to 17%, with anticipated minor effects on ecosystem
structure and function.
In the upper South Fork Koktuli River, gages SK100G and SK100F would experience 54 and 26%
reductions in streamflow, respectively, affecting 5 km of streams (Table 7-19). The tributary to the
South Fork Koktuli River receiving outfall from the WWTP would experience increased streamflows (28
to 30%), affecting 8 km of streams. In the North Fork Koktuli River, the tributary downstream of TSF 1
would experience 17 to 26% reductions in streamflow, affecting 2 km of streams (Table 7-19).
Several sections of the South Fork Koktuli River and tributaries below Frying Pan Lake are losing
reaches (i.e., discharge decreases in a downstream direction), which under pre-mine conditions
experience periods of zero minimum monthly discharge (e.g., gage SK100C and WWTP-receiving stream
gage SK124A) (Table 7-13). We assumed that streamflow increases due to the WWTP would follow the
natural hydrograph, reflecting the amount of precipitation and runoff that must be captured and treated.
As a result, WWTP outfall flows would be lowest during periods when these streams typically go dry
based on pre-mine baseline data, and would be highest during period of snowmelt runoff and fall
storms.
Pebble 2.0 Scenario
In the Pebble 2.0 scenario, streamflow reductions exceeding 20% sustainability boundaries would occur
in 19 km of streams downstream of the mine footprint. For these streams, substantial reductions in fish
habitat capacity and productivity would be expected. Increases in streamflow of 28 to 30% would be
expected for 8 km of streams downstream of the WWTP in the South Fork Koktuli River, and increases
of 12% would be expected for 4 km of the WWTP-receiving tributary to the North Fork Koktuli River,
leading to changes in sediment dynamics and habitat suitability for fish. An additional 6 km of streams in
Upper Talarik Creek and 7 km of streams in the North Fork Koktuli River would experience flow
reductions of 10 to 11%, with anticipated minor effects on ecosystem structure and function.
In the Pebble 2.0 scenario, the mine footprint captures 47% of the Upper Talarik Creek watershed above
gage UT100D (Table 7-17). As a result, most of the total stream length in its upstream reaches, including
the mainstem and all tributaries above gage UT100D, would experience either total loss of habitat from
the mine footprint or indirect effects of fragmentation (Section 7.2, Figure 7-15). Of this stream length,
2 km of mainstem downstream of the footprint would experience a significant loss of habitat and decline
Bristol Bay Assessment 755 January 2014
-------
Chapter 7 Mine Footprint
in habitat quality from the predicted 44% streamflow reduction at gage UT100D (Figure 7-15).
Downstream of gage UT100D in Upper Talarik Creek, streamflow reductions would range from 8 to 10%
(Table 7-19). Impacts on salmon habitat from streamflow reductions would be moderated by tributary
and groundwater inputs that may help ameliorate flow losses originating upstream, assuming that
groundwater sources and flowpaths are not also altered by the mine footprint. This assumption is
questionable (Section 7.3.2.3). For instance, the groundwater-dominated Upper Talarik Creek tributary
monitored at gage UT119A would experience a 12% streamflow reduction due to reduced flow in
portions of the South Fork Koktuli River resulting from losses to the mine footprint. This was the only
case of interbasin hydrologic connectivity explicitly modeled, but other undocumented connections are
likely to occur.
In the South Fork Koktuli River, streamflow reductions would exceed the 20% sustainability threshold
at gages SK100G, SK100F, and SK100CP2 (Table 7-19, Figure 7-15). In the South Fork Koktuli River
mainstem and tributaries upstream of gage SK100G, the majority of stream length would be eliminated
by the mine footprint (Figure 7-15), resulting in severe streamflow reductions at gages SK100G (72%)
and SK100F (35%) (Table 7-19). Streamflows in the South Fork Koktuli River at gage SK100C would
increase by 2% because of WWTP releases discharged at tributary gage SK124A, which would
experience a 28% increase in streamflow at the confluence with the South Fork Koktuli River (Table 7-
19, Figure 7-16).
In the North Fork Koktuli River, the majority of stream length above gage NK119A would be eliminated
by construction of TSF 1 (Figure 7-15), resulting in substantial streamflow losses (62% reduction at
gage NK119A) for approximately 2 km of streams between TSF 1 and the North Fork Koktuli River
(Table 7-19, Figure 7-15). Approximately 7 km of streams in the tributary measured by gage NK119B
would experience 11% reductions in streamflow. Increases in streamflow downstream of the WWTP
discharge point would increase Streamflows by 12% in 4 km of the North Fork Koktuli River upstream
of gage NK100C (Table 7-19, Figure 7-15).
Pebble 6.5 Scenario
The Pebble 6.5 scenario would capture an even larger portion of the South and North Fork Koktuli
Rivers and Upper Talarik Creek watersheds in its footprint. During operation of the Pebble 6.5 scenario,
streamflow reductions exceeding 20% sustainability boundaries would occur in 34 km of streams. For
these streams, reductions in fish habitat capacity and productivity could be expected. An additional 19
km of streams in Upper Talarik Creek would experience streamflow reductions exceeding 10%, with
anticipated minor effects on ecosystem structure and function. Increases in streamflow exceeding 20%
are expected for 14 km of streams downstream of the WWTP in the South Fork Koktuli River and for 4
km of the WWTP-receiving tributary to the North Fork Koktuli River, and would likely lead to
substantial changes in sediment dynamics and habitat suitability for fish.
In the Upper Talarik Creek watershed, substantial streamflow reductions are projected at gages UT100D
(86%) and UT100C2 (20%), affecting 6 km of streams. Streamflow alterations exceeding 10% would
occur in an additional 19 km of streams at gages UT100C1, UT100C, and UT100B (Table 7-19). In the
Bristol Bay Assessment 756 January 2014
-------
Chapter 7 Mine Footprint
South Fork Koktuli River, gages SK100G and SK119A would be buried under the expanded mine
footprint. A 62% reduction in streamflow would be expected for 1 km of the upper South Fork Koktuli
River downstream from the edge of the waste rock to gage SK100F (Table 7-19, Figure 7-16).
In the Pebble 6.5 scenario, the WWTP is estimated to discharge over 50 million m3 of water per year
(Table 7-18). This discharge would result in a 36% increase in streamflow for 8 km in the South Fork
Koktuli River above gage SK100CP1, and a 107% increase in streamflow for 7 km of streams above gage
SK124CP (Table 7-19, Figure 7-16). In the North Fork Koktuli River, WWTP outfalls would result in a
60% increase in streamflows for 4 km of streams above gage NK100C, and an 18% increase in
streamflows for 1 km of streams upstream of gage NK100B.
Streamflow reductions and stream habitat losses of the magnitudes estimated in the Pebble 0.25, 2.0,
and 6.5 scenarios represent substantial risks to spawning and rearing habitat for populations of coho,
sockeye, and Chinook salmon, Dolly Varden, and rainbow trout in the upper portions of the mine
scenario watersheds. Habitat quantity and quality would be significantly diminished by the loss of
streamflow from the mine footprint, via multiple mechanisms such as direct reduction in habitat area
and volume, the loss of channel to off-channel habitat connectivity, increased periods of zero
streamflow, and reduced food production. Streamflow increases could alter channel morphologies,
induce higher rates of sediment transport and erosion, and change the distribution of water velocities
within habitats used by spawning and rearing salmon and other fishes. Although the loss of salmonid
production has not been estimated, streamflow alterations greater than 20% would be expected to have
substantial effects (Richter et al. 2012).
7.3.2.2 Connectivity, Timing, and Duration of Off-Channel Habitats
Losses of streamflow resulting from the mine footprints and potential water withdrawals described
above would affect connectivity between the main channel and off-channel habitats important to
juvenile salmonids. Losses of flood peaks could alter groundwater recharge rates and influence
characteristics of floodplain percolation channels, seeps, or other expressions of the hyporheic zone
(Hancock 2002). Rapid streamflow reductions that exceed recession rates typically experienced by fish
in these systems could result in stranding or isolation offish in off-channel habitats (Bradford et al.
1995). Off-channel habitats, particularly those with groundwater connectivity, are critical rearing
habitats for several species of juvenile salmonids and can be important sockeye salmon spawning
habitats (Quinn 2005). Maintaining connectivity and the physical and chemical attributes of these
habitats in conditions similar to baseline conditions would be important for minimizing risks to salmon
and other native fishes.
Wetlands that are hydrologically connected to affected streams would also respond to alterations in
streamflow and groundwater. Fish access to and use of wetlands are likely to be extremely variable in
the mine footprint areas because of differences in the duration and timing of surface water connectivity
with stream habitats, distance from the main channel, or physical and chemical conditions (e.g.,
dissolved oxygen concentrations) (King et al. 2012). Projecting the effects of lost wetland connectivity
Bristol Bay Assessment 757 January 2014
-------
Chapter 7 Mine Footprint
and abundance on stream fish populations is beyond the scope of this assessment, but could be a
significant unknown.
Flow regulation through the WWTP could be designed to somewhat approximate natural hydrologic
regimes during periods when sufficient water and water storage capacity were available, which could
provide appropriate timing and duration of connectivity with off-channel habitats. Channel cross-
section data and gage data (PLP 2011) would provide useful insights into streamflow connectivity
relationships and could help guide a streamflow management plan.
7.3.2.3 Changes in Groundwater Inputs and Importance to Fish
There is limited information describing potential surface water-groundwater interactions in the mine
scenario watersheds, but groundwater is likely the dominant source of streamflow in these streams
(Rains 2011) and can be very important locally. High baseflow levels in the monthly hydrographs of the
mine scenario watersheds illustrate groundwater's important influence on these streams (Figure 3-10).
Aerial winter open-water surveys consistently suggest the presence of upwelling groundwater, which
maintains ice-free conditions in portions of area streams and rivers. Highly permeable glacial outwash
deposits create a complex mosaic within less permeable, clay to silt-dominated Pleistocene lake deposits
and bedrock outcrops, which can control surface water-groundwater interactions in landscapes like this
one (Power et al. 1999). Mine operations that reduce surface water contributions in the natural drainage
course or that lower groundwater tables may influence groundwater paths and connections within and
among streams in the mine area in ways that are not predicted in this assessment, but that could have
significant impacts on fish. In our analyses of the water management regimes for the mine scenarios, we
project increasing proportions of streamflow derived from water released from the WWTP as the mine
develops. These increased releases would result from increased interception of groundwater associated
with the mine pit cone of depression, rainwater, and surface runoff collection. Water treated and
discharged would replace a portion of the groundwater that would otherwise be feeding stream
systems, and could have substantially different chemical characteristics (Chapter 8).
Fish in the region are highly attuned to groundwater signals in the hydrologic and thermal regimes
(Power et al. 1999). Spatial heterogeneity in streamflow and temperature, largely mediated by
groundwater-surface water exchange, provides a template for diverse sockeye salmon life histories and
migration timing (Hodgson and Quinn 2002, Rogers and Schindler 2008, Ruff et al. 2011). For example,
groundwater moderates winter temperatures, which strongly control egg development and hatch and
emergence timing (Brannon 1987, Hendry et al. 1998). Spatial thermal heterogeneity allows diverse
foraging strategies for consumers of sockeye salmon and their eggs, such as brown bear and rainbow
trout, thereby benefitting not only sockeye salmon populations but also the larger foodweb (Armstrong
etal. 2010, Ruff etal. 2011).
Altered groundwater contributions to surface waters in the mine area could have profound effects on
the thermal regimes and thermally cued life histories of aquatic biota. Curry et al. (1994) examined the
influence of altered hydrologic regimes on groundwater-surface water interchange at brook trout
Bristol Bay Assessment 758 January 2014
-------
Chapter 7 Mine Footprint
spawning locations in an Ontario stream. Responses of groundwater-surface water exchange to changes
in river discharge varied among sites, precluding predictable responses. The complexity that can be
inherent in groundwater-surface water interactions can make regulating or controlling such
interactions during large-scale landscape development very difficult (Hancock 2002). Adequately
protecting the critical services that groundwater provides to fish is complicated by the fact that
flowpaths vary at multiple scales, and connections between distant recharge areas and local
groundwater discharge areas are difficult to predict (Power et al. 1999).
7.3.2.4 Stream Temperature
Projecting specific mine-associated changes to groundwater and surface water interactions and
corresponding effects on surface water temperature in the mine area is not feasible at this time.
Disruptions or changes to groundwater flowpaths could have significant adverse effects on winter
habitat suitability for fish, particularly if groundwater-dominated stream reaches are converted to
surface water-dominated systems. Irons et al. (1989 in Reynolds 1997) reported that groundwater-
mediated unfrozen refugia were dependent on fall rains maintaining groundwater, but that during a dry
year, groundwater levels declined and allowed full freezing of stream surface waters and the streambed.
This suggests that the threshold between completely frozen and partially frozen streams can be a
narrow one, particularly for small streams with low winter discharge. The duration of freezing and the
extent and type of ice formation, including anchor ice, frazil, or surface ice (Slaughter 1990), can
severely limit habitat availability during the winter and spring months.
Two aerial surveys of the mine scenario watersheds provide additional information on groundwater
inputs to headwater streams and ice cover conditions in streams draining the mine footprints (PLP
2011, Woody and Higman 2011). PLP conducted aerial and foot surveys during late-winter low-flow
conditions in 2006, 2007, and 2008 to determine the extent of open water and ice cover (PLP 2011:
Appendix 7.2B). Open-water reaches were consistently observed in strongly gaining reaches in the
South and North Fork Koktuli Rivers and Upper Talarik Creek. Open-water reaches corresponded to
areas of relatively warm groundwater that helped keep portions of the river network relatively ice-free
(PLP 2011: Appendix 15.IE). Aerial surveys documented by Woody and Higman (2011) in March 2011
showed broadly similar patterns of open water, suggesting that the general patterns reflect consistent
areas of strong groundwater-surface water interaction. Maintaining winter groundwater connectivity
may be critical for fish in such streams (Cunjak 1996, Huusko et al. 2007, Brown et al. 2011).
7.3.3 Risk Characterization
The water consumption predicted for our mine scenarios would require large volumes of water from
surface streams or groundwater, inevitably resulting in alterations to streamflows. Streamflow
alterations exceeding 20% would occur in 15, 27, and 53 km of streams in the Pebble 0.25, 2.0, and 6.5
scenarios, respectively, leading to significant adverse effects on fish and other aquatic life. The seasonal
timing and magnitude of streamflow alterations would be contingent on water storage and management
systems and strategies, but would be constrained by the fundamental needs for water at specific times
and locations in the mining process (Chapter 6). Impacts on fish habitat and fish populations would
Bristol Bay Assessment 759 January 2014
-------
Chapter 7 Mine Footprint
likewise depend on the magnitude and timing of streamflow changes, but would be most severe for
streams close to the mine footprint.
The volume of water that would require treatment by the mine's WWTP would range from
11 million m3/yr for Pebble 0.25 (Table 7-16) to over 50 million m3/yr for Pebble 6.5 (Table 7-18). To
avoid or minimize risks associated with altered streamflows in downstream effluent-receiving areas,
water storage and release capacities would be required to maintain natural streamflow regimes or to
maintain any minimum streamflows required by regulatory agencies. Application of the Instream Flow
Incremental Methodology (IFIM) Physical Habitat Simulation (PHABSIM) system modeling approach
(Bovee 1982, Bovee et al. 1998) is being used by PLP to assess streamflow-habitat relationships (PLP
2011: Chapter 15), and could provide additional guidance for establishing streamflow requirements
(Estes 1998) beyond those identified in this document.
Maintenance of mine discharges, in terms of water quality, quantity, and timing, to avoid adverse
impacts would require long-term monitoring and facility maintenance commitments. As with other long-
term maintenance and monitoring programs, the financial and technological requirements could be very
large, and the cumulative risks (and likely instantaneous consequences) of potential accidents, failures,
and human error would increase with time. In addition, climate change and projected changes in
temperature and precipitation in the region (Section 3.8) would result in potential changes in
streamflow magnitude and seasonally. These climate-related changes would interact with mining-
related flow impacts (Box 14-2), requiring adaptation to potentially new streamflow regimes. We know
of no precedent for the long-term management of water quality and quantity on this scale at an inactive
mine.
7.3.4 Uncertainties and Assumptions
Projecting changes to groundwater-surface water interactions in the mine footprint area with any
specificity is not feasible at this time. Local geology and stream hydrographs are indicative of systems
that are largely driven by groundwater. Disruptions or changes to groundwater flowpaths in the
footprint area could have significant adverse effects on winter habitat suitability for fish, particularly if
groundwater-dominated stream reaches are converted to stream reaches dominated by WWTP effluent.
Given the high likelihood of complex groundwater-surface water connectivity in the mine area,
predicting and regulating streamflows to maintain key ecosystem functions associated with
groundwater-surface water exchange would be particularly challenging.
Our approach for assessing potential risks of streamflow alteration rests on simplifying assumptions
regarding changes to the natural streamflow regime in the three mine scenarios (Section 7.3.2). The
natural streamflow regime consists of multiple components, including flow magnitude, frequency,
duration, timing, and rate of change, all of which can have important implications for fish and other
aquatic life (Poff et al. 1997). We were unable to anticipate changes to the streamflow regime beyond
simplistic alterations in flow magnitude. However, it is very likely that other aspects of the streamflow
regime would be modified as well, depending on how flows respond to water management at the mine
site. In addition, any changes in the duration of open-water freezing conditions associated with mining
Bristol Bay Assessment 760 January 2014
-------
Chapter 7 Mine Footprint
activities could alter seasonal streamflow regimes differently than we assume here. Our analysis does
not account for these possibilities.
We assumed that streamflow modifications would follow the natural hydrograph, reflecting the amount
of precipitation and runoff that was intercepted and thus must be captured and treated. As a result,
WWTP outfall flows are lowest during periods when these streams typically go dry based upon pre-
mine baseline data, and are highest during snowmelt runoff and fall storms. Alternative flow
management strategies may be feasible, depending on the capacity to store and release flows to meet
environmental streamflow objectives (see Appendix J for additional discussion).
Additionally, we assume that larger deviations from the natural streamflow regime pose greater risks of
ecological change. The scientific literature supports this assumption as a general trend (Poff et al. 2009,
Poff and Zimmerman 2010, Richter etal. 2012). However, as pointed out by Poff and Zimmerman
(2010), specific responses to changes in streamflow vary. Although all stream studies reviewed by Poff
and Zimmerman (2010) showed declines in fish abundance, diversity, and demographic rates with any
level of streamflow modification, other ecological responses (e.g., macroinvertebrate abundance,
riparian vegetation metrics) sometimes increased. Responses offish populations and other ecological
metrics to streamflow modification would depend on a suite of interacting factors, including but not
limited to stream structural complexity, trophic interactions, and the ability offish to move seasonally
(Anderson et al. 2006).
Potential impacts of the mine footprints discussed in this chapter do not explicitly take into account the
effects of climate change. Over the time scale at which large-scale mining would potentially affect the
assessment area, projected increases in temperature and precipitation may substantially change the
physical environment (Section 3.8 and Box 14-2). Such changes could significantly alter the variability
and magnitude of streamflows. Seasonal transitions between frozen and unfrozen conditions can
strongly influence groundwater-surface water interactions and streamflow dynamics (Callegary et al.
2013). Duration of freezing conditions and timing of snowmelt may be highly sensitive to climate
change, with significant implications for flow regimes. Increases in rain-on-snow events are likely, but
the potential implications for flooding are unclear. Nevertheless, these changes in streamflow regime
would likely lead to changes in sediment transport, bed stability, and channel morphology with potential
adverse impacts to fish habitat and population genetic diversity and resiliency.
7.4 Summary of Footprint Effects
Streams eliminated, blocked, or dewatered by the mine footprints in the Pebble 0.25, 2.0, and 6.5
scenarios would result in the loss of 8, 22, or 36 km, respectively, of documented anadromous waters as
defined in the AWC (Johnson and Blanche 2012). These lengths represent a loss of 2 to 11% of the total
AWC length in the mine scenario watersheds (total AWC length = 322 km) (Johnson and Blanche 2012).
An additional 30 to 115 km of headwater streams supporting habitat for non-anadromous fish species
would be lost to the mine footprint in these scenarios. Loss of headwater streams to the footprints
would alter groundwater-surface water hydrology, nutrient processing, and export rates of resources
Bristol Bay Assessment 761 January 2014
-------
Chapter 7 Mine Footprint
and materials to downstream aquatic ecosystems. Losses of wetlands would be 4.5,12, and 18 km2 in
the Pebble 0.25, 2.0, and 6.5 scenarios, respectively. In addition, the Pebble 0.25, 2.0, and 6.5 scenarios
would result in losses of 0.41, 0.93, and 1.8 km2 of ponds and lakes, respectively. An unquantified area of
riparian floodplain wetland habitat would either be lost or suffer substantial changes in hydrologic
connectivity with streams because of reduced streamflow from the mine footprint.
Reduced streamflow resulting from water consumption in mine operations, ore processing, transport,
and other processes, would further reduce the amount and quality offish habitat downstream of the
mine footprints. Changes in streamflow exceeding 20% would adversely affect habitat in an additional
15, 27, and 53 km of streams in the Pebble 0.25, 2.0, and 6.5 scenarios, respectively, reducing production
of coho salmon, sockeye salmon, Chinook salmon, rainbow trout, and Dolly Varden. Losses of stream
habitat leading to losses of local, unique populations would erode the population diversity that is
essential to the stability of the overall Bristol Bay salmon fishery (Schindler et al. 2010).
Bristol Bay Assessment 762 January 2014
-------
The water collection, treatment, and discharge scenarios presume that, under routine operations, runoff
water, leachate, and wastewater would be collected and properly treated before release to meet state
standards, federal criteria, and permit requirements. However, some leachate would escape collection,
supernatant water may be spilled from tailings storage facilities (TSFs), and some treatment failures
would be expected to occur. This chapter begins with a description of potential sources of contaminants
(Section 8.1). It then describes potential routes and magnitudes of exposure to contaminated water and
the exposure-response relationships used to screen leachate constituents (Section 8.2), with particular
focus on the major contaminant of concern, copper. This section ends with a characterization of the
potential risks from aqueous effluents and a discussion of potential additional remediation and
uncertainties. Potential effects of water temperature changes associated with water collection,
treatment, and discharge are discussed in Section 8.3. Figure 8-1 illustrates potential linkages between
sources, stressors, and responses associated with water treatment, discharge, fate, and effects that are
considered in this chapter.
8.1 Water Discharge Sources
Discharges were calculated for routine operations and wastewater treatment plant (WWTP) failure in
the Pebble 0.25, 2.0, and 6.5 scenarios; post-closure discharges are discussed qualitatively. Sources of
water discharge under routine operations associated with each mine scenario include effluents
discharged from the WWTP, uncollected leachates from the TSFs and waste rock piles, and spillway
releases from the TSFs. Other routine sources, including domestic wastewater, are outside the scope of
this assessment and thus not analyzed here. In addition, we evaluate a WWTP failure scenario in which
the system releases untreated wastewater. This failure represents one potential failure among many
accidents and failures that could occur. We specify that under routine operations, the WWTP would
meet permit limits. In the event of a complete treatment failure, flows would pass through the WWTP at
Bristol Bay Assessment
8-1
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
the estimated influent concentrations. These two water collection, treatment, and discharge scenarios
bound the likely range of water treatment operation, but do not encompass the worst case. For example,
treatment might fail when wastewater composition is worse than average, or an extreme accident like
dumping reverse-osmosis brine could occur.
Figure 8-1. Conceptual model illustrating the pathways linking water treatment, discharge, fate,
and effects.
f water treatment | ( waste rock | | tailings storage |
I facilities J I piles J I facilities J
mitigation
(e.g., line waste rock piles) \
[effluent dilution \
& transport
LEGEND
( source ) X""biotic"~'1\
Vresponss"/
additional step in
can sal pathway
l 1 modifying
proximate factor
Within a shape, 'I* indicates an
increase in the parameter, 4-
indicates a decrease in a
parameter, and A indicates a
change in the parameter.
Arrows leading from one shape to
another indicate a hypothesized
cause-effect relationship.
Shapes bracketed under another
shape are specific components of
the more general shape under
which they appear.
i V \
' NX
| water failure of water l<
1 treatment treatment facilities I
i
! NX \
i
! treated water untreated water
i discharges discharges
1
-H
\/ NX
' t metals t total
i dissolved sol
i
i
i
! V N
, 'f chronic toxicity 1" acute tc
i
i
i
i \
J
' ( ~\
i -J/ invertebrate
i abundance
i
! \
,/ \x
NX NX
jakageof 1 spillway |
eachate 1 release 1
NX NX
contaminated water
discharges
ds
/
ixicity
j s^~ 4, salmonid fishes •,
• vjabundance, productivity or diversity)^/
In addition to the discharge of treated water, water treatment generates wastes that are likely to be
hazardous due primarily to the copper and other metals removed from the wastewaters. The treatment
process is unspecified and it is unclear whether treatment wastes would be transported off site or
deposited in the TSFs or another on-site facility. Therefore, this assessment does not include risks of
water pollution resulting from spills of those waste materials and does not include them when
estimating chemical concentrations in the TSF leakage or spillage.
Bristol Bay Assessment
8-2
January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
Following the termination of mine operations, it is expected that water collection and treatment would
continue for waste rock and tailings leachates. If the water is nontoxic, in compliance with all criteria
and standards, and its composition is stable or improving, the collection and treatment system may be
shut down under permit. Otherwise, treatment would continue in perpetuity—that is, until untreated
water quality was acceptable or institutional failures ultimately resulted in abandonment of the system.
If the mine operator abandons the site, the State of Alaska should assume operation of the treatment
system; if both the mine operator and the State of Alaska abandon the site, untreated leachates would
flow to streams draining the site.
The promulgated state water quality standards are enforceable numeric limits on the concentrations
and durations of exposure for ambient waters, biotic communities, and associated designated uses. They
would be applied to permits for the discharges discussed here. National ambient water quality criteria
are contaminant limits that are recommended to the states. However, states such as Alaska may lag in
adopting the latest criteria. In particular, the U.S. Environmental Protection Agency (USEPA) (2007) has
published copper criteria based on the biotic ligand model (BLM), but Alaska still uses the hardness-
based criteria for copper. We use the current USEPA copper criteria in this assessment based on the
assumption that, before permitting a copper mine in the Bristol Bay watershed, Alaska would adopt
those criteria at the state level or would apply them on a site-specific basis to any discharge permits.
8.1.1 Routine Operations
Under the mine scenarios, water in contact with tailings, waste rock, ore, product concentrate, or mine
walls would leach minerals from those materials (Section 6.1.2.5). In addition, chemicals would be
added to the water used in ore processing (Box 4-5). Most of the water used to transport tailings or
concentrate or used in ore processing would be reused. Leachates collected from TSFs or waste rock
piles would be stored in the TSF or treated for use or discharge, but leachate that escaped collection
would flow to streams (Figure 6-5). Waste rock used in the construction of berms, roads and other mine
structures would be leached by rain and snowmelt, but that source is assumed to be small relative to the
waste rock piles and dams. Waste rock leachates are assumed to have the mean concentrations of
reported humidity cell tests (Appendix H and PLP 2011). Mine pit water would also be used or treated
for disposal. Surplus water on the site would be treated to meet applicable standards and other permit
limits and discharged. Based on Alaskan Water Quality Standards defined in the Alaska Administrative
Code, Title 18, Section 70, no mixing zones would be authorized for anadromous streams or spawning
habitat for most game or subsistence fish species. Thus, it is expected that effluents would be required to
meet state standards that are equivalent to national criteria and other permit limits (i.e., no exemptions
would be granted).
During mine operations, water available on the site would exceed operational needs, and approximately
11 to 51 million m3 of treated water would be discharged per year (Table 6-3). The mine scenarios
specify that effluent would be discharged to the South and North Fork Koktuli Rivers as proposed by
Ghaffari et al. (2011) (Tables 8-1 through 8-3). The effluent could contain treated tailings leachate,
Bristol Bay Assessment go January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
waste rock leachate, mine pit water, runoff, and excess transport or process waters. Tailings leachate
would come from the TSFs as either excess water in the impoundment or leakage captured below the
dams. The primary concern during routine operations would be waste rock leachate. Captured waste
rock leachate would become more voluminous as the waste rock piles increased during operation. After
mine closure, that leachate would be a major component of routinely generated wastewater, along with
water pumped from the TSFs and the pit (after it has filled). In addition, because the waste rock piles
and TSFs would not be lined, some leachates from both would not be captured and would flow to the
three receiving streams.
Risk quotients are used to determine whether the leachates are potentially toxic and, if so, which
constituents are most responsible (Tables 8-4 through 8-8). A risk quotient equals the exposure level
divided by an ecotoxicological benchmark. For screening, the undiluted leachate concentration is treated
as an exposure level. The benchmarks are national ambient water quality criteria or equivalent values
(Section 6.4.2.3). These benchmarks are for either acute (the criterion maximum concentration, or CMC)
or chronic (the criterion continuous concentration, or CCC) exposures—that is, CMCs are intended to be
thresholds for significant lethality in short-term exposures, whereas CCCs are intended to be thresholds
for significant lethal or nonlethal effects in long-term exposures. If the quotient is less than 1, the
leachate or constituent can be eliminated as a chemical of potential concern because instream
concentrations would not exceed the undiluted concentrations.
8.1.1.1 Tailings Leachate
Estimation of potential flow through the substrate located under and around proposed TSFs requires
estimation of hydraulic conductivity. The hydraulic conductivity of the substrate material located near
possible dam sites varies greatly with depth and location. Ghaffari et al. (2011) report a range from 10~6
to 10-5 m/s in the upper bedrock, with a general decrease with depth and a range on the order of W~7 to
10~9 m/s in the lower portions of bedrock with some zones of higher hydraulic conductivity (Figure 6-7).
In addition, the presence of fractured bedrock allows for localized discontinuities in the rate of
groundwater movement that can greatly influence overall groundwater conveyance (Ghaffari et al.
2011).
We estimated leachate flow from the TSFs using a hydraulic conductivity of 10~6 m/s in the upper 100 m
of overburden and bedrock, with no flow below that depth. We allowed vertical downward flow in the
tailings and radial flow outward in all directions from the TSF, with the excess head dissipating over a
horizontal distance of 1,200 m, comparable to the distance of the mine pit drawdown beyond the pit rim.
The interior surface area of TSF 1 would be 6.5 km2 for the Pebble 0.25 scenario and 14.2 km2 for the
Pebble 2.0 and 6.5 scenarios (Table 6-2). The Pebble 6.5 scenario would include two additional
impoundments, with interior surface areas of 20.1 km2 (TSF 2) and 8.2 km2 (TSF 3) (Table 6-2).
Bristol Bay Assessment 04 January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-1. Annual effluent and receiving water flows at each gage in the Pebble 0.25 scenario. All values are presented in m3/yr.
Stream and Gage
Flow Returned
Through WWTP"
Flow Returned as TSF
Leakage
Flow Returned as NAG
Waste Rock Leachate
Flow Returned as PAG
Waste Rock Leachate
Flow of Interbasin
Transfer
TOTAL FLOW
South Fork Koktuli River
SK100G
SK100F
SK100CP2b'c
SK124A
SK124CP"
SK100C
SKIOOCPI"
SK119A
SK119CP"
SK100B1
SK100B"
-
-
5,454,000
-
-
-
-
-
-
-
-
-
-
207,000
350,000
-185,000
-
-
-
-
-
-
-
-
-
-
-
-12,446,000
-
-
-
-
5,287,000
19,055,000
25,263,000
22,265,000
23,391,000
43,942,000
44,113,000
31,268,000
33,124,000
113,737,000
159,937,000
North Fork Koktuli River
NK119A
NK119CP2"
NK119B
NK119CPlb
NK100C
NK100B
NK100A1
NK100A6
-
5,454,000
-
1,113,000
-
-
-
120,000
-
-
-
-
-
-
-
-
-
15,378,000
17,923,000
3,975,000
23,512,000
47,282,000
77,513,000
183,022,000
221,668,000
Upper Talarik Creek
UT100E
UT100D
UT100C2
UT100C1
UT100C
UT119AC
UT100B'
-
-
-
-
-
-
-
-
-
185,000
-
-
-
-
-
-
12,446,000
7,474,000
22,008,000
90,768,000
106,050,000
139,053,000
23,286,000
191,423,000
Notes:
Dashes (-) indicate that values are either not applicable or are equal to zero.
a WWTP discharges 50% of flow to South Fork Koktuli River, 50% of flow to North Fork Koktuli River (no WWTP flows are directed to Upper Talarik Creek).
b Confluence point where virtual gage was created because physical gage does not exist.
c 1/3 of total return flow is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location. Interbasin transfer flows are represented by negative flow values for SK100CP2
(losses to Upper Talarik Creek) and equivalent positive flow values for UT119A (ga ns from South Fork Koktuli).
d USGS 15302200.
e USGS 15302250.
' USGS 15300250.
WWTP = wastewater treatment plant; TSF = tailings storage facility; PAG = potentially acid-generating; NAG = non-ac d-generating.
Bristol Bay Assessment
8-5
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-2. Effluent and receiving water flows at each gage in the Pebble 2.0 scenario. All values are presented in m3/yr.
Stream and Gage
Flow Returned
Through WWTP"
Flow Returned as
TSF Leakage
Flow Returned as NAG
Waste Rock Leachate
Flow Returned as PAG
Waste Rock Leachate
Flow of Interbasin
Transfer
TOTAL FLOW
South Fork Koktuli River
SK100G
SK100F
SK100CP2».c
SK124A
SK124CPb
SK100C
SKIOOCPI"
SK119A
SK119CP"
SK100B1
SKlOOBd
-
-
5,152,000
-
-
-
-
-
2,000
-
-
21,000
-
633,000
507,000
-380,000
22,000
-
-
151,000
-
213,000
3,000
-72,000
-
-
-
-
-
-
-11,409,000
-
-
-
-
3,266,000
16,745,000
23,723,000
21,878,000
23,004,000
42,552,000
42,722,000
30,774,000
32,630,000
110,513,000
156,302,000
North Fork Koktuli River
NK119A
NK119CP2"
NK119B
NK119CPlb
NK100C
NK100B
NK100A1
NK100A6
-
-
5,152,000
-
-
2,305,000
1,000
-
-
23,000
-
402,000
13,000
3,000
-
204,000
-
-
-
-
-
-
-
-
-
8,111,000
10,347,000
3,641,000
17,069,000
46,905,000
71,452,000
176,169,000
215,132,000
Upper Talarik Creek
UT100E
UT100D
UT100C2
UT100C1
UT100C
UT119AC
UT100B'
-
-
-
-
642,000
-
380,000
-
-
72,000
-
-
11,409,000
5,290,000
13,481,000
83,215,000
98,684,000
130,691,000
22,516,000
180,889,000
Notes:
Dashes (-) indicate that values are either not applicable or are equal to zero.
a WWTP discharges 50% of flow to South Fork Koktuli River, 50% of flow to North Fork Koktuli River (no WWTP flows are directed to Upper Talarik Creek).
b Confluence point where virtual gage was created because physical gage does not exist.
c 1/3 of total return flow is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location. Interbasin transfer flows are represented by negative flow values for SK100CP2
(losses to Upper Talarik Creek) and equivalent positive flow values for UT119A (gains from South Fork Koktuli).
d USGS 15302200.
e USGS 15302250.
' USGS 15300250.
WWTP = wastewater treatment plant; TSF = tailings storage facility; PAG = potentially acid-generating; NAG = non-acid-generating.
Bristol Bay Assessment
8-6
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-3. Effluent and receiving water flows at each gage in the Pebble 6.5 scenario. All values are presented in m3/yr.
Stream and Gage
Flow Returned
Through WWTP"
Flow Returned as TSF
Leakage
Flow Returned as NAG
Waste Rock Leachate
Flow Returned as PAG
Waste Rock Leachate
Flow of Interbasin
Transfer
TOTAL
South Fork Koktuli River
SK100G
SK100F
SK100CP2b'c
SK124A
SK124CPb
SK100C
SKIOOCPI"
SK119A
SK119CPb
SK100B1
SK100BC
-
-
25,494,000
-
-
-
20,000
-7,000
1,626,000
-
-
2,930,000
50,000
145,000
-
1,278,000
-423,000
713,000
-
54,000
242,000
413,000
260,000
-
1,032,000
-344,000
-
-
-
-
-8,770,000
-
-
-
95,000
9,810,000
19,096,000
36,049,000
37,175,000
57,070,000
57,241,000
14,262,000
15,172,000
104,322,000
146,346,000
North Fork Koktuli River
NK119A
NK119CP2b
NK119B
NK119CPlb
NK100C
NK100B
NK100A1
NKlOOAe
-
-
-
25,494,000
-
2,360,000
1,000
48,000
-
23,000
402,000
13,000
144,000
-
204,000
-
-
-
-
-
-
-
-
8,167,000
10,402,000
3,004,000
15,473,000
67,053,000
90,144,000
194,738,000
233,778,000
Upper Talarik Creek
UT100E
UT100D
UT100C2
UT100C1
UT100C
UT119AC
UT100B'
-
-
-
-
-
-
-
7,000
-
346,000
739,000
160,000
428,000
-
-
-
-
344,000
-
-
-
-
8,770,000
-
2,125,000
3,482,000
73,815,000
89,511,000
120,303,000
20,203,000
166,476,000
Notes:
Dashes (-) indicate that values are either not applicable or are equal to zero. SK100G and SK119A are eliminated by the mine footprint in this scenario.
a WWTP discharges 50% of flow to South Fork Koktuli River, 50% of flow to North Fork Koktuli River (no WWTP flows are directed to Upper Talarik Creek).
b Confluence point where virtual gage was created because physical gage does not exist.
c 1/3 of total return flow is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location. Interbasin transfer flows are represented by negative flow values for SK100CP2
(losses to Upper Talarik Creek) and equivalent positive flow values for UT119A (gains from South Fork Koktuli).
d USGS 15302200.
e USGS 15302250.
' USGS 15300250.
WWTP = wastewater treatment plant; TSF = tailings storage facility; PAG = potentially acid-generating; NAG = non-acid-generating.
Bristol Bay Assessment
8-7
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-4. Aquatic toxicological screening of tailings supernatant against acute (criterion
maximum concentration) and chronic (criterion continuous concentration) water quality criteria or
benchmark values. Values are in ug/L unless otherwise indicated. Average leachate values are from
Appendix H.
Analyte
pH (standard units)
Alkalinity (mg/LCaCOa)
Hardness (mg/L CaCOa)
S04
Ag
Al
As
Ca
Cd
Co
Cr
Cua
Cub
Fe
Hg
K
Mg
Mn
Mo
Na
Ni
Pb
Sb
Se
Tl
Zn
Sum of metals
Average Value
7.9
75
320
320,000
0.018
72
17
116,000
<0.1
<0.1
<1.0
7.8
7.8
17
<0.037
26,000
8,000
72
70
44,000
<0.8
0.2
6.0
7.6
0.0
4.3
-
CMC or
equivalent
-
24
750
340
-
6.3
89
1,500
40
7.2
350
1.4
-
-
760
32,000
1,300
220
14,000
-
-
316
-
CCC or
equivalent
-
-
87
150
-
0.55
2.5
190
24
4.4
-
0.77
-
-
690
73
140
8.8
1,600
5
-
316
-
Acute Quotient
-
0.0007
0.096
0.051
-
<0.012
<0.0011
<0.0007
0.19
1.1
0.048
<0.026
-
-
0.095
0.0022
<0.0006
0.0010
0.0004
-
-
0.014
0.50a : 1.4b
Chronic
Quotient
-
-
0.82
0.11
-
<0.14
<0.040
<0.0051
0.32
1.8
-
<0.048
-
-
0.10
0.96
<0.0056
0.026
0.0038
1.5
-
0.014
4.3a : 5.8b
Dashes (-) indicate that criteria are not available.
a Acute and chronic criteria from Alaska's hardness-based standard.
b Acute and chronic criteria from the national ambient water quality criteria based on the biotic ligand model.
CMC = criterion maximum concentration; CCC = criterion continuous concentration.
Bristol Bay Assessment
8-8
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-5. Aquatic toxicological screening of tailings humidity cell leachates against acute
(criterion maximum concentration) and chronic (criterion continuous concentration) water quality
criteria or benchmark values. Values are in ug/L unless otherwise indicated. Average concentrations
are from Appendix H.
Analyte
pH (standard units)
Alkalinity (mg/LCaCOa)
Hardness (mg/L CaCOa)
Cl
F
S04
Ag
Al
As
B
Ba
Be
Bi
Ca
Cd
Co
Cr
Cua
Cub
Fe
Hg
K
Mg
Mn
Mo
Na
Ni
Pb
Sb
Se
Sn
Tl
V
Zn
Sum of metals
Average Value
7.8
60
67
520
450
17,000
0.01
24
5.5
11
9.2
0.20
0.49
23,000
0.05
0.19
0.50
5.3
5.3
30
0.01
4,000
2,500
44
33
2,100
0.54
0.06
1.8
1.5
2.9
0.05
0.78
3.2
-
CMC or
equivalent
-
-
-
1.6
750
340
29,000
46,000
-
1.5
89
445
10
2.5
350
1.4
-
760
32,000
360
46
14,000
-
3,600
1,370
91
-
CCC or
equivalent
6.5-9
-
-
-
87
150
1,500
8,900
-
0.20
2.5
58
6.9
1.6
0.77
-
693
73
40
1.8
1,600
5.0
75
0.8
120
91
-
Acute Quotient
-
-
-
0.0062
0.031
0.016
0.0004
0.0002
-
0.038
0.0021
0.0012
0.58
1.1
0.0071
-
0.058
0.0010
0.0016
0.0015
-
0.0008
0.0006
0.038
0.78a: 1.3b
Chronic
Quotient
-
-
-
0.27
0.036
0.0071
0.0010
-
0.28
0.076
0.0094
0.84
1.8
0.013
-
0.064
0.45
0.014
0.039
0.30
0.039
0.0065
0.038
2.5a: 3.4b
Notes:
Dashes (-) indicate that criteria are not available.
a Acute and chronic criteria from Alaska's hardness-based standard.
b Acute and chronic criteria from the national ambient water quality criteria based on the biotic ligand model.
CMC = criterion maximum concentration; CCC = criterion continuous concentration.
Bristol Bay Assessment
8-9
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-6. Aquatic toxicological screening of test leachate from Tertiary waste rock in the Pebble
deposit and quotients against acute (criterion maximum concentration) and chronic (criterion
continuous concentration) water quality criteria or benchmark values. Values are in ug/L unless
otherwise indicated. Average leachate concentrations are from Appendix H.
Parameter
PH
Alkalinity (mg/LCaCOa)
Hardness (mg/L CaCOa)
Cl
F
S04
Ag
Al
As
B
Ba
Be
Bi
Ca
Cd
Co
Cr
Cua
Cub
Fe
Hg
K
Mg
Mn
Mo
Na
Ni
Pb
Sb
Se
Sn
Tl
V
Zn
Sum of metals
Average Value
7.2
66
74
530
62
28,000
0.011
80
2.7
18
57
0.31
0.54
21,000
0.22
3.9
0.55
3.2
3.2
140
0.010
1,900
5,100
100
6.3
7,200
4.4
0.12
2.1
1.9
1.3
0.068
1.8
16
-
CMC or
equivalent
-
-
-
1.9
750
340
29,000
46,000
-
-
1.5
89
445
10
2.5
350
1.4
-
760
32,000
360
46
14,000
-
3,600
-
1,370
91
-
CCC or
equivalent
6.5-9
-
-
-
87
150
1,500
8,900
-
-
0.20
2.5
58
6.9
1.6
0.77
-
693
73
40
1.8
1,600
5.0
75
0.8
120
91
-
Acute
Quotients
-
-
-
0.0059
0.11
0.0081
0.0006
0.0012
-
-
0.15
0.044
0.0012
0.32
1.3
0.40
0.0073
-
0.13
0.0002
0.012
0.0025
0.0002
-
0.0003
-
0.0013
0.17
1.4a : 2.4b
Chronic
Quotients
-
-
-
0.92
0.018
0.012
0.0064
-
-
1.1
1.6
0.0094
0.46
2.0
0.013
-
0.15
0.086
0.11
0.06
0.0013
0.38
0.017
0.085
0.15
0.17
5.3a : 6.8b
Notes:
Dashes (-) indicate that criteria are not available.
a Acute and chronic criteria from Alaska's hardness-based standard.
b Acute and chronic criteria from the national ambient water quality criteria based on the biotic ligand model.
CMC = criterion maximum concentration; CCC = criterion continuous concentration.
Bristol Bay Assessment
8-10
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-7. Aquatic toxicological screening of test leachate from Pebble East pre-Tertiary waste
rock and quotients against acute (criterion maximum concentration) and chronic (criterion
continuous concentration) water quality criteria or benchmark values. Values are in ug/L unless
otherwise indicated. Average leachate values are from Appendix H.
Parameter
pH (standard units)
Alkalinity (mg/LCaCOa)
Hardness (mg/L CaCOa)
Cl
F
S04
Ag
Al
As
B
Ba
Be
Bi
Ca
Cd
Co
Cr
Cua
Cub
Fe
Hg
K
Mg
Mn
Mo
Na
Ni
Pb
Sb
Se
Sn
Tl
V
Zn
Sum of metals
Average Value
4.8
9.9
22
910
110
52,000
0.019
380
8.0
13
4.5
0.55
0.63
6,300
3.2
9.7
1.6
1,400
1,400
10,000
0.010
960
1,500
340
4.3
2,100
10
0.35
0.78
3.2
1.9
0.088
2.4
480
-
CMC or
equivalent
-
-
-
0.24
750
340
29,000
46,000
-
-
0.46
89
160
3.20
0.043
350
1.4
-
760
32,000
130
12
14,000
-
3,600
-
1,370
32
-
CCC or
equivalent
6.5-9
-
-
-
87
150
1,500
8,900
-
-
0.085
2.5
21
2.4
0.027
0.77
-
693
73
14
0.47
1,600
5.0
75
0.8
120
32
-
Acute Quotients
-
-
-
0.082
0.51
0.023
0.0004
0.0001
-
-
7.0
0.11
0.0096
440
33,000
0.0072
-
0.44
0.0001
0.081
0.029
0.0001
-
0.0005
-
0.0018
15
460a : 33,000b
Chronic
Quotients
-
-
-
4.4
0.053
0.0084
0.0005
-
-
38
3.9
0.073
580
52,000
0.013
-
0.49
0.059
0.73
0.75
0.0005
0.65
0.024
0.110
0.020
15
640a : 52,000b
Notes:
Dashes (-) indicate that criteria are not available.
a Acute and chronic criteria from Alaska's hardness-based standard.
b Acute and chronic criteria from the national ambient water quality criteria based on the biotic ligand model.
CMC = criterion maximum concentration; CCC = criterion continuous concentration.
Bristol Bay Assessment
8-11
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-8. Aquatic toxicological screening of test leachate from Pebble West pre-Tertiary waste
rock against acute (criterion maximum concentration) and chronic (criterion continuous
concentration) water quality criteria or benchmark values. Values are in ug/L unless otherwise
indicated. Average leachate values are from Appendix H.
Parameter
pH (standard units)
Alkalinity (mg/LCaCOa)
Hardness (mg/L CaCOa)
Cl
F
S04
Ag
Al
As
B
Ba
Be
Bi
Ca
Cd
Co
Cr
Cua
Cub
Fe
Hg
K
Mg
Mn
Mo
Na
Ni
Pb
Sb
Se
Sn
Tl
V
Zn
Sum of metals
Average Value
6.6
18
59
520
120
61,000
0.027
320
1.5
16
14
0.33
0.69
13,000
0.40
7.0
0.69
1,600
1,600
1,700
0.011
1,400
6,700
730
1.8
2,100
6.8
0.17
3.1
3.8
0.14
0.41
0.68
56
-
CMC or
equivalent
-
-
-
1.3
750
340
29,000
46,000
-
1.2
89
370
8.2
0.88
350
1.4
-
760
32,000
300
36
14,000
-
3,600
1,370
75
-
CCC or
equivalent
6.5-9
-
-
-
87
150
1,500
8,900
-
0.17
2.5
48
5.7
0.55
0.77
-
690
73
33
1.4
1,600
5.0
75
0.8
120
75
-
Acute Quotients
-
-
-
0.021
0.42
0.0044
0.0005
0.0003
-
0.33
0.079
0.0019
190
1,800
4.8
0.0076
-
0.96
0.0001
0.023
0.0047
-
0.00004
0.0005
0.74
200a : l,800b
Chronic
Quotients
-
-
-
3.7
0.0100
0.011
0.0015
-
2.3
2.8
0.014
280
2,900
0.014
-
1.1
0.025
0.20
0.12
0.76
0.0019
0.52
0.0057
0.74
290a : 2,900b
Notes:
Dashes (-) indicate that criteria are not available.
a Acute and chronic criteria from Alaska's hardness-based standard.
b Acute and chronic criteria from the national ambient water quality criteria based on the biotic ligand model.
CMC = criterion maximum concentration; CCC = criterion continuous concentration.
Bristol Bay Assessment
8-12
January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
Total leakage amounts for the three mine scenarios are 1.1 x 106 m3/yr (Pebble 0.25), 2.4 x 106 m3/yr
(Pebble 2.0), and 7.2 x 106 m3/yr (Pebble 6.5) (Tables 8-1 through 8-3). These estimates are based on a
simple assessment of seepage from the TSFs. Actual hydraulic conductivity would likely span several
orders of magnitude, from rapid flow in large fractures to essentially no flow in tight formations. Even a
small number of flowpaths with higher than expected hydraulic conductivity could significantly affect
the direction and quantity of flow.
Two potential estimates of tailings leachate composition are presented in Tables 8-4 and 8-5. Tailings
leachate from the humidity cell tests (Table 8-5) is judged to better represent effluent from a tailings
impoundment than the supernatant (Table 8-4); thus, these values are used to represent leachate from
the bottom of the TSFs and excess water from the TSFs routed to the WWTP.
The tailings slurry would also contain ore-processing chemicals. We use an estimated concentration of
sodium ethyl xanthate, the primary ore-processing contaminant of concern, of 1.5 mg/L in the tailings
slurry (NICNAS 1995). Process chemicals could enter the environment in TSF leachate or WWTP
effluent. The potential for process chemicals in product concentrate slurry is considered in Chapter 11.
8.1.1.2 Waste Rock Leachate
Tertiary rock would be used for construction of tailings dams and berms and potentially other
structures requiring fill, but most would be piled near the mine pit. It is classified as non-acid-generating
(NAG) and its leachate is neutral (Table 8-6). Pre-Tertiary rock is classified as potentially acid-
generating (PAG) and its leachate is acidic (Tables 8-7 and 8-8). PAG waste rock would be piled
separately and blended with ore, as needed, to maintain consistent composition in the processing plant
feed. Incomplete collection of pre-Tertiary waste rock leachate would result in acid mine drainage.
The mine scenarios (and the plan put forth for Northern Dynasty Minerals by Ghaffari et al. [2011]) do
not include liners for the waste rock piles. Instead, leachate within the mine pit's drawdown zone would
be captured in the pit and pumped to the WWTP. Outside the drawdown zone, we estimate that half the
leachate would be captured by extraction wells or other means and the rest would flow to surface
waters. This is considered reasonable given the likelihood that water would flow between wells and
below their zones of interception in the relatively permeable overburden materials and upper bedrock.
Wells would not catch all flows from the mine site given its geological complexity and the permeability
of surficial layers. As a result, we estimate that 84% of PAG leachate and 82% of total waste rock
leachate would be captured by the pit and the wells for the Pebble 2.0 scenario.
8.1.1.3 Mine Pit and Runoff Water
Water pumped from the mine pit would consist of captured waste rock leachate and leachate from the
pit walls as precipitation passes over them and groundwater flows through them. The pit wall leachate
is estimated from the maximum groundwater concentration at the mine site, because rainwater flowing
through the ore body and rocks in its vicinity is assumed to be similar to rainwater flowing over the pit
walls. The estimated concentration of the critical contaminant, 3.2 ug/L copper, is almost identical to the
mean Tertiary (NAG) waste rock test leachate. Other constituent concentrations are 37 ug/L aluminum,
Bristol Bay Assessment 813 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
0.05 ug/L cadmium, 0.63 ug/L cobalt, 45 ug/L manganese, 3.2 ug/L nickel, 0.86 ug/L lead, 0.30 ug/L
selenium, 7.9 ug/L zinc, 1600 mg/L total dissolved solids, and 5.6 pH. This means that the mine pit
water is much cleaner than PAG pre-Tertiary leachate (e.g., copper in estimated pit wall leachate is only
0.2% of PAG waste rock leachate).
Runoff from the ore-crushing and screening area is assumed to have the composition of pre-Tertiary
(PAG) waste rock test leachate (Table 8-7). All other plant and ancillary area runoff is assumed to have
the composition of the maximum background stream water. All of these waters would be captured and
routed to a TSF or the WWTP.
8.1.1.4 Wastewater Discharge
Under the three mine scenarios, the WWTP would be designed and sized to treat the expected volume
and composition of inflow water based on estimated groundwater flow from the mine pit and runoff
from other site areas (waste rock piles, TSFs, and plant and ancillary facilities). The WWTP would be fed
by pipelines that pump water to the plant from the mine pit, crusher area, waste rock and TSF leachate
collection systems, and other operating areas of the site. However, mine pit water represents the largest
component of flow into the WWTP in our scenarios. The flow volume contributed by each mine
component has been estimated for each scenario (Table 6-3). If the volume or composition of untreated
water exceeded plant specifications, it could be stored temporarily in a TSF process pond or even the
mine pit, and fed into the plant as needed to balance flows and meet permit effluent quality
requirements.
We specify that the WWTP would operate under a permit that would require meeting all national
criteria and Alaskan standards. We also assume that the Alaskan Pollutant Discharge Elimination System
wastewater discharge permit for a mine would include requirements that all other potentially toxic
contaminants be kept below concentrations equivalent to national chronic criteria. This use of non-
standard benchmarks in permitting is not normal practice, but the importance of the aquatic resources
and the degree of public concern would justify that action. The equivalent benchmark values used in this
assessment for metals with no criteria or standards appear in Table 6-10. Assumed discharge
concentrations are the minimum of the input water concentration and the chronic criterion, standard or
benchmark value. Influent and effluent concentrations of contaminants of concern are presented in
Table 8-9. WWTP discharge rates for the Pebble 0.25, 2.0, and 6.5 scenarios are estimated to be
approximately 11,10, and 51 million m3/year, respectively, equally distributed to the South and North
Fork Koktuli Rivers.
8.1.1.5 Sources of Total Dissolved Solids
Neither total dissolved solids (TDS) nor specific conductance data are available for waste rock or tailings
leachates from the Pebble deposit. However, TDS can be estimated by summing the concentrations of
leachate analytes after converting alkalinity to bicarbonate. Estimated TDS concentrations for the
tailings leachates, waste rock leachates, and WWTP effluents are summarized in Table 8-9.
Bristol Bay Assessment 814 January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-9. Estimated concentration of contaminants of concern in effluents from the wastewater
treatment plant, tailings, non-acid-generating waste rock, and potentially acid generating waste
rock. Values are in ug/L unless otherwise indicated.
Contaminant
IDS (mg/L)
Zn
Se
Pb
Ni
Mn
Co
Cd
Al
Cu
WWTP Influent and Failure
Effluent3
0.25"
312
17
1.6
0.22
2.3
67
0.99
0.14
44
75
2.0"
297
26
1.4
0.28
3.1
92
2.2
0.22
66
101
6.5'
529
33
1.5
0.33
3.3
105
2.1
0.26
73
150
WWTP Effluent3
0.25"
2.0"
6.5'
280d
17
1.6
0.22
2.3
67
0.99
0.064
44
23
1.4
0.28
3.2
92
2.2
0.064
66
23
1.4
0.29
3.3
100
2.0
0.064
73
Hd.e
Tailings
Leachate
123
3.2
1.5
0.064
0.54
44
0.19
0.052
24
5.3
NAG Waste
Rock
Leachate
145
16
1.9
0.12
4.4
101
3.9
0.22
80
3.2
PAG Waste
Rock
Leachate
100
270
3.5
0.26
8.6
530
8.4
1.8
350
1,500
Notes:
3 Concentrations for the Pebble 0.25 scenario.
b Concentrations for the Pebble 2.0 scenario.
c Concentrations for the Pebble 6.5 scenario.
d When only one value is shown across all three scenarios, it means that the contaminant is above the chronic criterion and must be lowered to
the criterion under all three scenarios.
e Chronic water quality criterion based on the biotic ligand model using mean North Fork Koktuli River water.
WWTP = wastewater treatment plant; IDS = total dissolved solids; NAG = non-acid-generating; PAG = potentially acid-generating.
8.1.2 Wastewater Treatment Plant Failure
There are innumerable ways in which wastewater treatment could fail in the mine scenarios, in terms of
failure type (e.g., breakdown of treatment equipment, ineffective leachate collection, wastewater
pipeline failure), location, duration, and magnitude (e.g., partial vs. no treatment). Box 8-1 presents an
example wastewater collection failure, and mechanisms of treatment failure are discussed in Box 8-2. To
bound the range of reasonable possibilities, we assess a serious failure in which the WWTP allows
untreated water to discharge directly to streams. This type of failure could result from a lack of storage
or treatment capacity or treatment efficacy problems. Chronic releases would occur during operation if
a lengthy process were required to repair a failure. We evaluate potential effects of this type of failure
using the following assumptions.
• The effluent is untreated water that is released to discharge points on tributaries to the South and
North Fork Koktuli Rivers.
• Untreated water composition is a flow-weighted average of concentrations from multiple
wastewater sources, including mine pit dewatering, waste rock leachates, runoff from crusher and
ancillary areas, and TSF leachates.
• Discharge rates are based on the sum of component flow volumes from the wastewater sources,
developed as part of the scenario water balances (Section 6.2.2).
• Discharge rates and concentrations were calculated for each of the three mine scenarios
(Pebble 0.25, 2.0 and 6.5) and account for shifts in the relative contribution and concentration of
different wastewater sources for different mine sizes.
Bristol Bay Assessment
8-15
January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
• Duration of a release could range from a few hours to several months, depending on the nature of
the failure and the difficulty of repair and replacement.
BOX 8-1. AN ACCIDENTAL TAILINGS WATER RELEASE: NIXON FORK MINE, ALASKA, WINTER 2012
The Nixon Fork Mine is an underground gold mine that was intermittently mined between 1917 and 1950.
The modern mine opened in 1995, then closed in 1999 (ADNR 2012) and reopened under new ownership
again in 2007. The mine is located on federal lands managed by the Bureau of Land Management. The
mine operates under authorizations from the Bureau of Land Management, the Alaska Department of
Natural Resources (ADNR), and the Alaska Department of Environmental Conservation (ADEC).
In January and February 2012, the tailings impoundment at the Nixon Fork Mine overtopped. Below is the
chronology of events described by the mine operator that led to this event, based on a March 15, 2012
memo to the Alaska State Mine Safety Engineer from Mystery Creek Resources, Inc.
• Prior to October 25, 2011, mine staff monitored the freeboard in the tailings impoundment per
requirements of agency authorizations.
• After October 25, 2011, staff decided to waive gage observation until spring melt because the gage was
frozen in ice.
• During a mid-January trip to the site, the president of Mystery Creek Resources, Inc. noticed insufficient
freeboard in the tailings pond. He notified the Bureau of Land Management, ADNR, and ADEC.
• Corrective action was taken and the pond level began to drop.
• In late February 2012, mill operations that had been completed in batches were switched to continuous
operation without recognizing the implications for water balance (i.e., more water would be flowing to
the tailings impoundment).
• On March 9, 2012, mine personnel noticed evidence of dam overtopping. The Bureau of Land
Management, ADNR, and ADEC were notified and action was taken to draw down the pond and stop the
overtopping.
• On March 10, 2012, agency inspections began. It was found that water from the tailings impoundment
was not likely to have reached nearby streams. An estimated 32,400 gallons of tailings water were
discharged from the impoundment.
On dam inspection it was found that the engineered spillway for the dam had been frozen over by a
previously undiscovered tailings water release. The ice prevented the spillway from operating as designed,
such that the later spill overtopped the dam at another location not designed for overflow. This case
illustrates the diversity of potential failures that can happen and suggests the practical impossibility of
predicting all possible failure modes.
Water treatment also would generate sludges or brines containing material removed from the
wastewaters plus materials added to the water, such as precipitating agents. These materials are
expected to be deposited in the TSFs. Because the mine scenarios do not include a specific water
treatment technology, no spill scenario for these wastes was developed. However, copper and other
metal concentrations in these wastes would be high, so they likely would be significantly toxic if spilled
into surface waters.
If a gold-processing facility was added at the site, a separate water treatment system would decompose
or recycle the cyanide used in the separation of gold (Box 4-6). That system would have the potential to
fail, releasing the cyanide solution to a stream or groundwater. Cyanide in the tailings would flow to a
TSF, where it could degrade or combine with copper or other metals. However, a cyanide-processing
system has not been described, and we do not consider a water treatment failure scenario for this
potential source.
Bristol Bay Assessment 816 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
BOX 8-2. POTENTIAL FAILURES OF REVERSE OSMOSIS WASTEWATER TREATMENT PLANTS
Because the high-quality receiving waters in the mine scenario watersheds would require extremely low
copper criteria and standards, reverse osmosis has been discussed as a potential treatment technology for
wastewater at the Pebble site. Studies of wastewater treatment plant (WWTP) efficiency and design
considerations show that reverse osmosis water treatment systems can be compromised by fouling and
scalingfrom calcium, iron, barium, strontium, silica, microbial growth, and silt (Mortazavi 2008). The
Bingham Canyon WWTP in Utah treats groundwater contaminated with sulfate and total dissolved solids
from copper mining by reverse osmosis. Pilot tests and optimization studies have shown that the structural
integrity of its reverse osmosis membranes can be damaged by abrasive materials (e.g., silt) or chlorine
(ITRC 2010). Changes in water composition could increase the concentration of chlorine if the mine pit
encounters a large flow of brine transmitted to the pit through deep fracture systems, or from localized
areas of mineralized rock with anomalous water quality. An example of WWTP failure due to highly variable
chemical composition of inflow wastewater has been documented at a copper mine in Chile: when silica
concentrations exceeded the design range, the whole reverse osmosis system could not be operated and
was therefore shut down until feed water quality improved (Shao et al. 2009).
8.1.3 Spillway Release
The spillway release scenario considered here involves the controlled release of water from TSF 1 to the
North Fork Koktuli River. Spillway releases are not part of routine operations; however, because
overflow is a sufficiently likely event, spillways are considered a routine feature of operating TSFs. This
spillway release is not a worst-case spill, in that it does not involve overfilling of the TSF with
wastewaters that would be diverted to the TSF during a WWTP shutdown or failure. It is, however, a
severe case.
For this spillway release analysis, we assume that TSF 1 has reached its maximum interior area of 14.2
km2. A spillway constructed in or near the dam on the north side of the TSF would discharge towards the
North Fork Koktuli River. This spillway may be either a temporary construction spillway for emergency
releases or the permanent spillway. We assume that the pond within the TSF has reached its maximum
safe operating level for the current dam height and that any additional precipitation requires the release
of a volume of water equal to the precipitation volume. We further assume that the volume of water
released exceeds the capacity of the WWTP and the conveyance mechanisms to transfer water from the
TSF to the mine pit or other on-site locations, resulting in all released water discharging directly into the
stream with no treatment.
8.1.4 Post-Closure Wastewater Sources
The post-closure period includes two distinct phases with respect to water management (Section 6.3.4).
The first phase would be from the time the mine ceased operations until the mine pit was effectively full
of water. The second phase would be after the mine pit filled until treatment was no longer necessary.
During both phases, the quality of the water captured at the mine site would be substantially better than
the water captured during mine operations. During operations, leachate from the PAG waste piles would
account for between 80% and 94% of the total copper load in the captured water, depending on the
scenario. Since the mine scenarios specify that all of the PAG waste rock would be processed by the close
of operations and the PAG areas rehabilitated during site closure, the remaining flows would carry a
Bristol Bay Assessment 817 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
much lower concentration of copper. The expected reduction in copper concentration in the loading to
the WWTP would be greater than 90%, with substantial reductions also expected for other metals.
During pit filling, the mine operator would potentially need to treat water captured from the surface of
the TSFs, captured leachate from the TSFs, captured leachate from the NAG waste rock piles outside the
drawdown zone, and runoff from remaining facility areas that support the ongoing water treatment.
Based on our drawdown model, the drawdown zone would not begin to shrink until pit water level was
within about 100 m of its final level. As the remainder of the pit filled, the drawdown zone would shrink
until the pit reached its final level. If water in the pit required treatment, the final pit level would be
maintained below the level that allowed natural outflow by pumping water to the WWTP. We assume
that this drawdown would result in drainage toward the pit for about 100 m beyond the pit perimeter. If
or when the pit water and other sources met the discharge criteria, all flows could be discharged
without treatment and the pit water level would be allowed to rise until natural discharge was
established at the low point of its perimeter.
Because post-closure water quality is expected to be better than water quality during operation, the
assessment does not model or evaluate water quality during this period. In addition, post-closure
conditions are much more uncertain than conditions during mine operations, so it is more difficult to
defend a particular set of conditions and assumptions. It is important to note, however, that although
post-closure water treatment failures would be less consequential, they also would be less likely to be
promptly detected and corrected. In addition, because site hydrology and chemistry would change over
time, particularly as the pit filled, treatment requirements would change and responses might be slow.
The pit lake is a novel feature of the post-closure period, and, because it has been a subject of
stakeholder and reviewer concern, it requires more specific consideration. After closure, the time
required for the mine pit to fill with water would range from approximately 20 years (in the Pebble 0.25
scenario) to more than 200 years (in the Pebble 6.5 scenario). Eventually, the pit water would be a
source of leached minerals to streams, if it were not collected and treated. Precipitation on the pit walls,
groundwater entering the pit, and water collected in the pit would dissolve metals and anions from the
rock walls and any waste rock returned to the pit, resulting in leachate.
Leachate composition would be approximated by some mixture of the waste rock test leachates
(Section 8.1.1.2), with some dilution by ambient water. These tests were run in oxidizing conditions, so
they maximize leaching rates. Although oxygen would be provided in the pit by atmospheric diffusion
from the surface, precipitation, shallow groundwater, and vertical mixing of water in the pit during
turnover, oxygen levels are expected to be lower in the pit than in the leachate tests. Flow of waste rock
leachate to the pit after closure would contribute to the mixture in the pit. However, as the drawdown
zone shrinks, most waste rock would be outside of the drawdown zone and much would be
downgradient of the pit, so its leachate would flow away from the pit. Pit water composition cannot be
predicted with any confidence, but some degree of leaching is inevitable. The experience with closed pit
mines is quite variable, but some mine pit lakes (e.g., the Berkeley Pit in Montana) are acidic and have
Bristol Bay Assessment 818 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
high metal concentrations. Water flowing out of the full pit would be expected to flow to Upper Talarik
Creek, where it would mix with waste rock leachate and water diverted from upstream.
In sum, failure to collect and treat waters from the waste rock piles, TSFs, or mine pit could expose biota
in the streams draining the post-closure mine site to contaminated water. There is little information on
failure rates for post-closure wastewater management at mines. If the closure occurs as described in this
assessment, toxic effects could occur but they are unlikely to be severe. However, premature closures of
mines do occur and such closures are likely to leave acid-generating materials on the surface. Further, it
is much too soon to know whether mines that are permitted for perpetual water collection and
treatment (e.g., the Red Dog Mine in Alaska) can actually carry out those functions in perpetuity.
8.1.5 Probability of Contaminant Releases
Water collection and treatment failures are a common feature of mines. A review of the 14 porphyry
copper mines that have operated for at least 5 years in the United States found that all but one (93%)
had experienced reportable aqueous releases (the definition of a reportable release is determined by
local regulations and differs among mines), with the number of events ranging from three to 54
(Earthworks 2012). Mine water releases range from chronic releases of uncaptured leachate to acute
events caused by equipment malfunctions, heavy rains, or power failures. The USEPA has observed that
some operators continue to discharge when they know that treatment is ineffective and not meeting
standards. Hence, the record of analogous mines indicates that releases of water contaminated beyond
permit limits would be likely over the life of any mine at the Pebble deposit.
The probability of the specific WWTP failure analyzed here cannot be estimated. It is improbable in that
it requires that wastewater not be treated and not be diverted to storage. However, it is plausible that
such an event would result from equipment failures, inadequate storage or human errors. It is more
likely that a partial failure (e.g., incomplete treatment) would occur, but any one of the innumerable
incomplete treatment scenarios is also unlikely. Hence, the WWTP failure scenario analyzed here
provides a reasonable upper bound.
8.2 Chemical Contaminants
8.2.1 Exposure
8.2.1.1 Effluent Dilution and Transport
Under the mine scenarios, treated wastewater discharges would be divided between the South and
North Fork Koktuli Rivers. South Fork Koktuli River flows include interbasin transfer to Upper Talarik
Creek. Tailings water leakage and any uncontrolled leachate from the NAG rockfill dams would
discharge to the South Fork Koktuli River (except in the Pebble 0.25 scenario, in which no tailings would
be placed in that watershed) and the North Fork Koktuli River from TSF 1 and TSF 3 and to the South
Fork Koktuli River from TSF 2. Leachate from the waste rock piles that is not captured and treated
would flow to Upper Talarik Creek (except in the Pebble 0.25 scenario, in which no waste rocks would
Bristol Bay Assessment 819 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
be placed in that watershed) and the South Fork Koktuli River. NAG waste rock leachate would be the
only direct source of wastewater to Upper Talarik Creek during routine operations in the Pebble 2.0 and
Pebble 6.5 scenarios, and no wastewater would directly enter Upper Talarik Creek in the Pebble 0.25
scenario.
WWTP effluents would be released at the surface, entering receiving waters as a plume and gradually
being diluted. Input of contaminated groundwater from waste rock or tailings leachates would be
introduced via upwelling through cobble and gravel substrates (i.e., via hyporheic input). In either case,
a gradient would occur between full-strength effluents and fully-mixed ambient waters.
Fully-mixed ambient concentrations for each scenario are calculated by diluting the estimated discharge
(i.e., contributing loads) in the background receiving waters using ambient flows and concentrations
from Pebble Limited Partnership (PLP) (2011), after adjusting baseflows for the reductions in
watershed areas due to the mine footprints (Tables 8-1 through 8-3; note that constituent flows at a
gage are less than total flows because mine-related flows from upstream are carried forward in the
model). Concentrations of contaminants of concern in wastewater discharges, waste rock leachates, and
tailings leachates are presented in Table 8-9. Discharge flow rates are based on the water balances
described in Section 6.2.2, and include reduced streamflows due to water use in the mine scenarios and
interbasin transfers. Contaminant flows were blended with adjusted ambient water flows and tracked
downstream from one stream gage to the next.
Working from the upstream-most point in each mine scenario watershed, ambient contaminant mass
flows were added to discharge contaminant mass flows and divided by total flow at each stream gage to
determine the diluted concentration. Moving to the next downstream stream gage the process was
repeated, each time adding the mining process flows at their expected concentrations, assuming that
background concentrations at each stream gage would be capturing all concentration inputs other than
mining inputs. This implies that mining processes cause no other degradation or metal contributions
through other mechanisms, such as surface erosion or mobilization of metals from in situ minerals by
acidic leachates.
Because the streams draining a mine site are the receptors for wastewaters, their water quality
constitutes the dilution water quality. The water quality of streams in the Pebble deposit area has been
extensively characterized (PLP 2011, Zamzow 2011). Streams in the mine scenario watersheds are
neutral to slightly acidic with low conductivity, hardness, dissolved solids, suspended solids, and
dissolved organic carbon (DOC) (Table 8-10). In this respect, they are characteristic of undisturbed
streams. However, as would be expected for a metalliferous site, levels of sulfate and some metals
(copper, molybdenum, nickel, and zinc) are elevated, particularly in the South Fork Koktuli River. PLP
(2011) found that copper levels in some samples from the South Fork Koktuli River exceeded Alaska's
chronic water quality standard. However, most of the exceedances were "in sampling locations within or
in proximity to the general deposit location" and the number and magnitude of exceedances decreased
with distance downstream (PLP 2011: Figure 9.1-35). Therefore, the stream reaches with significantly
Bristol Bay Assessment 820 January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
elevated copper concentrations would largely be destroyed by the mine footprints and by water
diversions.
The chemical fate of metals in receiving streams may be complex. Aluminum, iron, and manganese are
commonly precipitated in streams receiving acid mine drainage, diminishing or destroying stream
habitats with deposited floes but also reducing the aqueous toxicity of those metals. Acidic leachates
would form from the PAG waste rocks, but concentrations of precipitating metals in PAG waste rock
leachates are not particularly high (Tables 8-7 and 8-8). Other metals do not precipitate to a significant
degree but may have reduced bioavailability due to receiving water chemistry. That issue is largely dealt
with by use of the BLM for copper, which includes a metal speciation submodel.
Table 8-10. Means and coefficients of variation for background surface water characteristics of the
mine scenario watersheds, 2004-2008.
Analyte
IDS (mg/L)
pH (field)
DO (mg/L)
Temperature (°C)
Specific Conductivity (pS/cm)
TSS (mg/L)
Ca (mg/L)
Mg(mg/L)
Na (mg/L)
K (mg/L)
Alkalinity (mg/L)
S04 (mg/L)
Cl (mg/L)
F (mg/L)
Hardness (mg/L)
Al (Mg/L)
As (Mg/L)
Ba (Mg/L)
Cd (Mg/L)
Cu (Mg/L)
Fe (Mg/L)
Mn(Mg/L)
Mo (Mg/L)
Ni (Mg/L)
Pb (Mg/L)
Zn (Mg/L)
CN (Mg/L)
DOC (mg/L)
South Fork
Koktuli River
44 (0.41)
7.0 (0.045)
10.2 (0.21)
4.77 (1.02)
ND
2.21(2.13)
6.34 (0.42)
1.41 (0.56)
2.35 (0.35)
0.38 (0.47)
17.4 (0.43)
8.78 (0.87)
0.69 (0.26)
ND
19.6 (0.47)
11 (0.68)
ND
4.1(0.48)
ND
1.3 (0.88)
120 (1.01)
20(1.14)
0.66 (0.98)
0.41(0.61)
ND
2.7 (1.02)
ND
1.36 (0.62)
North Fork
Koktuli River
37 (0.035)
6.74(0.10)
10.2 (0.2)
4.39(1.12)
ND
1.39 (1.9)
5.09 (0.35)
1.32 (0.44)
2.38 (0.23)
0.41 (0.39)
20.5 (0.38)
2.26 (0.56)
0.66 (0.25)
ND
14.4 (0.36)
13 (0.82)
ND
3.1(0.35)
ND
0.39 (0.84)
110 (0.68)
10 (1.67)
0.19(1.2)
0.30(1.14)
ND
1.8 (0.64)
ND
1.5 (0.51)
Upper
Talarik Creek
51.2(0.37)
6.99 (0.091)
10.5(0.19)
4.04 (0.98)
73.4 (0.34)
2.52(1.58)
8.77 (0.3)
2.12(0.46)
2.82(0.32)
0.44 (0.43)
31.8(0.34)
5.48(1.43)
0.7 (0.29)
ND
26.5(0.42)
13(1.1)
ND
5.5 (0.41)
ND
0.42 (0.89)
110 (0.83)
21(1.09)
0.2 (0.51)
0.63(1.04)
ND
2.0 (1.09)
ND
1.57 (0.82)
Notes:
Filtered concentrations are used for hardness and trace elements.
ND = ana lytes detected in less than half of samples.
Source: PLP 2011.
Bristol Bay Assessment
8-21
January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
8.2.1.2 Biological Exposures
Aquatic biota would be directly exposed to contaminants in discharged waters. Fish embryos and larvae
(e.g., salmon eggs and alevins) would be exposed to benthic pore water, which would be provided by
groundwater in areas of upwelling and otherwise by surface water. In this chapter we assume that
sediments would not be contaminated by tailings, waste rock, or other mine-derived particles. Juvenile
fish (e.g., salmon fry and parr) would be exposed to surface water. Adult resident salmonids would also
be exposed to surface water, but unlike the early life stages, they would occur in the smallest streams
only during spawning. Adult anadromous salmonids would have brief exposures to waters near the site.
Aquatic insects would be exposed in all juvenile stages, which constitute most of their life cycles. They
would be exposed to benthic pore water or surface water depending on their habits.
8.2.2 Exposure-Response
We screened potential contaminants against ecotoxicological benchmarks to identify the most
potentially toxic constituents and indicated the degree of treatment that would be required and the
types of effects that might occur due to mining emissions. Criteria and equivalent screening benchmarks
are presented in Tables 8-4 through 8-8, and the sources of non-criteria screening benchmarks are
presented in Table 6-10. Benchmarks were derived from the literature to be as similar to criteria as
possible, given the available data (Section 6.4.2.3). Criteria for many of the metals are functions of
hardness, and copper criteria are a function of multiple water properties. For those metals, criteria are
calculated for each leachate based on its chemistry and for each receiving stream based on its
background chemistry.
8.2.2.1 Copper
Although the ore and waste rock from porphyry copper mines contain a mixture of metals, copper is the
major resource metal and is particularly toxic to aquatic organisms. Hence, it is the most likely to cause
toxic effects, and actions taken to prevent copper's effects are likely to mitigate, to some extent, effects
from co-occurring metals. For these reasons, we focus on copper criteria, standards, and toxicity in this
assessment.
Copper Standards and Criteria
The State of Alaska's copper standard is a function of hardness and is based on a prior national criterion
(USEPA 1985a). The formulas for Alaska's acute value (CMC) and chronic value (CCC), in micrograms
per liter and based on hardness in milligrams per liter, are:
Copper acute criterion = e°-9422(ln hardness) -1.700 x Q.96
Copper chronic criterion = e°-8545(ln hardness) -1.702 x Q.96
Note that the formulae are similar and yield similar values—that is, when copper causes toxic effects,
the effects occur relatively quickly. At 20 mg/L hardness (i.e., in soft water typical of the Bristol Bay
region), Alaska's acute and chronic values for copper are 2.95 and 2.26 ug/L, respectively.
Bristol Bay Assessment 822 January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
The federal government has developed new National Ambient Water Quality Criteria for Protection of
Aquatic Life (hereafter, criteria) for copper (USEPA 2007). These criteria are calculated using the BLM,
which derives the effects of copper as a function of the amount of metal bound to biotic ligands on gills
or other receptor sites on an aquatic organism. The ligands bind free copper ions and, to a lesser degree,
copper hydroxide ions (Figure 8-2). Copper competes for ligands with calcium and other cations. The
competitive binding model for the biotic ligand requires input from a metal speciation submodel and
user-input values for basic water chemistry parameters (i.e., pH, temperature, DOC, humic acid, calcium,
magnesium, sodium, potassium, sulfate, sulfide, chloride, and alkalinity). The BLM is an advance over
hardness normalization, because it more fully accounts for the mechanisms controlling variance in
toxicity. In practice, its most important consequence is to estimate the often large reduction in toxicity
resulting from the binding of copper by dissolved organic matter. The BLM is freely available from
USEPA (http://water.epa.gov) and from the model's developer Hydroqual Inc.
(http://www.hydroqual.com/blm).
Figure 8-2. Processes involved in copper uptake as defined in the biotic ligand model (USEPA
2007).
inorganic
Complexes
Gill Surface
(biotic ligand)
Active Metal
Sites
e.g. : Cu - Hydroxides
Cu - Carbonates
The results of applying the BLM to mean water chemistries of the South and North Fork Koktuli Rivers
and Upper Talarik Creek are presented in Table 8-11. These values are lower than Alaska's hardness-
based values and the variance among streams is potentially significant.
Bristol Bay Assessment
8-23
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-11. Results of applying the biotic ligand model to mean water chemistries in the mine
scenario watersheds (Table 8-10) to derive acute (CMC) and chronic (CCC) copper criteria specific
to receiving waters. Values are in ug/L.
Stream
South Fork Koktuli River
North Fork Koktuli River
Upper Talarik Creek
CMC
2.4
1.7
2.7
CCC
1.5
1.1
1.7
Notes:
CMC = criterion maximum concentration; CCC = criterion continuous concentration.
Biotic ligand model source: USEPA 2007.
The results of applying the BLM to mean chemistries of the waste rock leachates are presented in
Table 8-12. Model runs used mean water chemistries from the PLP tests (Appendix H). These effluent-
specific values differ from each other and from the values for ambient waters due to differences in water
chemistries.
Table 8-12. Results of applying the biotic ligand model to mean water chemistries in waste rock
leachates (Appendix H) to derive effluent-specific acute (CMC) and chronic (CCC) copper criteria.
Values are in ug/L
Leachates
Pebble Tertiary
Pebble West pre-Tertiary
Pebble East pre-Tertiary
CMC
2.5
0.88
0.043
CCC
1.6
0.55
0.027
Notes:
CMC = criterion maximum concentration; CCC = criterion continuous concentration.
Biotic ligand model source: USEPA 2007.
For both the background waters and the leachates, temperature was set to the mean from three streams
on the site (4.5°C) (PLP 2011). For the leachates, DOC was set to 1 mg/L (a reasonable value given the
absence of DOC in the leachate, which would mix with ambient water containing approximately 1.5
mg/L of DOC) and humic acid was set to the default value (10% of DOC).
Both the state standards and national criteria for copper are derived from the 5th centile of the aquatic
genera sensitivity distribution. The most sensitive 33% of genera in acute tests and 42% of genera in
chronic tests are all invertebrates (USEPA 2007). Hence, the regulatory benchmarks are determined by
invertebrate sensitivities. However, the most sensitive vertebrates in both types of tests are fishes in the
genus Oncorhynchus, which includes rainbow trout and the five Pacific salmon species. Rainbow trout is
a standard test species that is at least as sensitive to copper as Chinook and coho salmon in acute tests
(Chapman 1975,1978). Acute and chronic values for rainbow trout can be derived for background
water quality using the BLM (Table 8-13). BLM-estimated acute values could also be calculated for three
cladoceran species in the three streams draining the mine scenario footprints: Daphnia magna (8.68 to
13.02 ug/L), Daphnia pulex (4.28 to 6.63 ug/L), and Ceriodaphnia dubia (5.99 to 9.13 ug/L). These
zooplankters are less directly relevant to the receiving streams, but they are relevant to ponds and
Iliamna Lake and they illustrate the sensitivity of aquatic arthropods to copper.
Bristol Bay Assessment
8-24
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-13. Site-specific acute and chronic copper toxicity values for rainbow trout, derived by
applying the biotic ligand model to mean water chemistries in the mine scenario watersheds (Table
8-10).
Stream
South Fork Koktuli River
North Fork Koktuli River
Upper Talarik Creek
Acute Toxicity
(LCso in Mg/L)
63
59
75
Chronic Toxicity
(CV in Mg/L)
22
21
26
Notes:
LCso = median lethal concentration; CV = chronic value, calculated using the species-specific acute to chronic ratio of 2.88.
Biotic ligand model source: USEPA 2007.
A test conducted with juvenile Chinook salmon showed greater sensitivity to subchronic copper
exposures than is suggested by Table 8-13 (Mebane and Arthaud 2010). After 120 days, reductions in
both length (5.6%) and weight (21%) were observed in salmon exposed to 7.4 ug/L copper. A BLM
could not be developed for the salmon, but the test water chemistry was relevant to the Pebble site
(hardness = 25.4 mg/L, pH = 7.32, DOC = 1.2 mg/L). Mebane and Arthaud (2010) applied these growth
effects to a population demographic model for a threatened Chinook salmon population spawning in
Idaho. They found that the observed reductions in individual growth would reduce population growth
due to increased mortality of smaller out-migrating fish (Mebane and Arthaud 2010). However, it should
be noted that the sensitivity of juvenile Chinook salmon was still less than that of sensitive invertebrates
(USEPA 2007).
Alternative Copper Endpoints for Fish
Copper standards and criteria are based on conventional test endpoints of survival, growth, and
reproduction. However, research has shown that salmonid olfactory systems are affected at low copper
concentrations (Hecht et al. 2007). Salmon use olfaction to find their spawning streams, detect and avoid
predators, find food, detect reproductive and alarm pheromones, and perform other life processes.
Although effects on fish olfaction have not been shown to affect the viability of field populations, it is
reasonable to expect that interference with these essential processes would have population-level
consequences (DeForest et al. 2011a).
Meyer and Adams (2010) applied the hardness-corrected criteria and the BLM to data from multiple
laboratory tests for olfactory effects. They found that the BLM accounted well for variance among tests,
and that BLM-based criteria were consistently protective of olfactory effects in the test systems. In
contrast, hardness-corrected criteria were not consistently protective (Meyer and Adams 2010).
DeForest et al. (2011b) extended those results by applying the same models to 133 ambient waters in
the western United States (including Alaska) that exhibited a wide range of water chemistries. Using the
20% inhibitory concentration (IC2o) for coho salmon olfaction from Mclntyre et al. (2008a, 2008b) as the
endpoint, they found that the hardness-corrected criteria were not consistently protective, but that the
BLM-based chronic criteria were protective of this chronic effect in 100% of waters. Even the acute
BLM-based criteria were protective of this chronic effect in 98% of waters, since the criteria are
Bristol Bay Assessment
8-25
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
determined by sensitive invertebrates that experience diminished survival, growth, or reproduction at
even lower levels than those that inhibit fish olfactory receptors.
Although the criteria are protective screening values for sensory effects of copper on salmonids, it is
necessary to consider potential effects when criteria are exceeded. Meyer and Adams (2010) adapted
the BLM to sensory data and derived IC2o:BLM factors that can be used to convert site-specific criteria
into estimates of the copper concentration at which 20% of rainbow trout avoid the contaminated water
or at which they experience 20% inhibition of their olfactory senses (Table 8-14). These values bracket
the threshold for growth effects in juvenile Chinook salmon (7.4 ug/L copper) described above.
Table 8-14. Site-specific benchmarks for sensory effects in rainbow trout. Values are derived by
applying IC2o:BLM ratios from Meyer and Adams (2010) to the acute values in Table 8-8.
Stream
Avoidance
(IC2o in
Sensory Inhibition
(IC2o in
South Fork Koktuli River
5.2
26
North Fork Koktuli River
3.8
19
Upper Talarik Creek
5.9
30
IC2o = 20% inhibitory concentration; BLM = biotic ligand model.
Avoidance cannot prevent severe toxic effects of copper on salmonid fish, unless they encounter low
concentrations before high concentrations (e.g., if they are swimming up a concentration gradient). At
concentrations sufficient to cause mortality or reproductive failure, copper damages the sensory organs
and avoidance does not occur (Hansen et al. 1999).
Neurobehavioral effects may be responsible for findings that low-level exposures to copper reduce out-
migration success. Lorz and McPherson (1977) pre-exposed coho salmon to 0, 5,10, 20, or 30 ug/L of
copper for between 6 and 165 days and released them into a coastal Oregon stream on four dates.
Percent successful out-migration was reduced relative to controls by copper exposure at all
concentrations and durations, with greater effects observed at higher exposures.
Dietary Copper Exposure-Response for Fish
Dietary exposure to metals, particularly at mine sites, has become a topic of investigation in recent years
(Meyer et al. 2005). Studies of the tailings-contaminated Clark Fork River in Montana and the Coeur
d'Alene River in Idaho have indicated that macroinvertebrates can accumulate metals at levels that
result in toxicity and reduced growth in fish that consume them (Farag et al. 1994, Woodward et al.
1994, Woodward et al. 1995, Farag et al. 1999). Although those effects were shown to be most
correlated with exposure to copper, subsequent studies suggest that the effects were primarily caused
by co-occurring arsenic (Hansen et al. 2004, Erickson et al. 2010).
Participants in a recent Pellston Workshop (a workshop series convened by the Society for
Environmental Toxicology and Chemistry to examine environmental toxicological issues) reviewed the
literature and developed an estimate of the degree to which aqueous toxicity thresholds should be
adjusted to account for dietary exposures in rainbow trout (Borgmann et al. 2005). The estimate is
Bristol Bay Assessment
8-26
January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
based on an average bioconcentration factor of 2,000 L/kg and an average dietary chronic value of 646
ug/g for rainbow trout. The resulting factor is 0.95, so the adjustment is small. If the factor is applied to
the lowest chronic value for rainbow trout (11.3 ug/L) (USEPA 2007), the result (10.7 ug/L) is still much
higher than the national ambient water quality criteria and state standards, due to the relative
insensitivity offish. This result applies to aqueous-only exposures (i.e., it does not include contaminated
sediment or allochthonous material). Because this dietary exposure factor has little influence on risks to
fish from direct aqueous exposure and adds another source of uncertainty, it is not applied to the risk
estimates in this chapter. However, dietary exposure offish to copper in sediments, where direct
aqueous exposures of post-larval fish may be minor, is considered in Chapter 9.
Copper and Algal Production
Although copper sulfate is used as an algicide, a relatively small amount of high-quality toxicity data is
available for algae or other aquatic plants (USEPA 2007, European Copper Institute 2008). Freshwater
algae and aquatic vascular plants are generally less sensitive than invertebrates or fish, with No
Observed Effects Concentrations for growth ranging from 15.7 to 510.2 ug/L in high-quality data
sources (European Copper Institute 2008); these values are for dissolved copper but are not corrected
for water chemistry. However, a few whole ecosystem studies suggest that algal production may be
reduced at lower copper concentrations and this may contribute to the sensitivity of insects to copper in
the field (Hedtke 1984, Leland and Carter 1984,1985, Leland et al. 1989, Brix et al. 2011). The effects of
copper are complex and involve competition among algal taxa that vary in their sensitivities and
changes in grazing intensity, so in some systems algal production is relatively resistant (Le Jeune et al.
2006, Roussel et al. 2007). It appears that criteria based on toxicity to invertebrates would also be
protective of algal production, but the data are unclear. Risks to algal production from copper are not
considered further, because the uncertainties are so large relative to risks to fish and invertebrates.
Copper Exposure-Response Data from Analogous Sites
Evidence concerning exposure-response relationships for copper and other metals in streams at metal
mines also comes from field studies. Because the mine scenarios presume that water quality criteria
would be met during routine operations, the critical question is whether effects are observed at those
levels. The most relevant high-quality studies are those performed in the Colorado metal belt,
particularly near the Animas and Arkansas Rivers. These sites are contaminated predominantly by mine
drainage and mine waste leachates, and field and laboratory experiments have confirmed that aqueous
metals, not tailings or other particles, cause the observed effects (Courtney and Clements 2002). These
studies have identified effects on aquatic insect populations and invertebrate communities at
concentrations below water quality criteria for the dominant metals (cadmium, copper, and zinc)
(Buchwalter et al. 2008, Schmidt et al. 2010). Application of the BLM and an additive combined effects
model reduced the discrepancy but did not eliminate it, suggesting that chronic criteria for metals are
not protective against effects on invertebrates (Schmidt et al. 2010). In particular, although the
combined criteria approximated thresholds for taxa richness, abundances of sensitive taxa were
reduced at exposures below the combined criteria (Griffith et al. 2004, Schmidt et al. 2010). This result
Bristol Bay Assessment 827 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
was supported by a study funded by Rio Tinto, which concluded that "aquatic insects are indeed very
sensitive to some metals and in some cases may not be protected by existing WQC [water quality
criteria]" (Brix et al. 2011). Potential reasons for the discrepancy are the absence of sensitive species or
life stages from the criteria, less-than-life-cycle exposures, and the absence of dietary exposures.
Unexpected field effects might be caused by an unknown factor that is correlated with both the
concentration of metals and the biological effects (i.e., a confounding variable). However, no such factor
is known for this assessment, and the hypothesized mechanisms for the greater sensitivity of field
communities are supported by evidence from laboratory and field experiments (Brix et al. 2011). It also
must be noted that the occurrence of biological effects below criteria concentrations does not
necessarily indicate that criteria are not adequately protective. By design, the criteria allow acute or
chronic effects on as much as 5% of species (USEPA 1985b).
Copper Exposure-Response Uncertainties
The copper criteria are based on a large body of data and a mechanistic model of exposure and effects.
Hence, at least with respect to acute toxicity, it is one of the best-supported criteria. However, it is
always possible that it would not be protective in particular cases due to unstudied conditions or
responses. Because the most sensitive taxa are aquatic invertebrates, unknown aspects of invertebrates
are most likely to be influential. In particular, field studies, including studies of streams draining metal
mine sites, show that Ephemeroptera (mayflies) are often the most sensitive species and that smaller
instars are particularly sensitive (Kiffney and Clements 1996, Clements et al. 2000). However, the
copper criteria do not include any Ephemeroptera in the sensitivity distribution (USEPA 2007). If the
mayfly, stonefly, caddisfly, or other invertebrate species in the streams draining the mine footprints are
more sensitive than cladocerans (the most sensitive tested species), then they may not be protected by
the criteria.
In addition, the chronic copper criterion is derived by applying an acute-to-chronic ratio to the BLM-
derived final acute value (USEPA 2007). Because of the complex dynamics of chronic uptake,
distribution, and sequestration of metals in aquatic insects, the BLM, which focuses on binding to a
surface ligand, may not adequately adjust chronic toxicity (Luoma and Rainbow 2005, Buchwalter et al.
2008). Brix et al. (2011) reviewed the toxicity testing literature and found that aquatic insects are highly
sensitive to copper and some other metals in chronic relative to acute exposures and may not be
protected by current criteria. Thus, the protectiveness of the chronic criterion is more uncertain than
that of the acute criterion.
Based on the literature cited above, resolution of this uncertainty through additional research and
testing is likely to lower the chronic criterion. Therefore, this uncertainty biases downward the
estimated length of streams experiencing toxic effects and could change our conclusions with respect to
relatively low toxicity materials such as tailings and NAG waste rock. The naturally elevated copper
concentrations in the highest reaches of some of the South Fork Koktuli River tributaries further
complicate the assessment of copper toxicity. Sensitive taxa may not occur in those reaches.
Alternatively, the biota in those reaches may be somewhat resistant to copper additions, although
Bristol Bay Assessment 828 January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
studies in the Colorado metal belt (see previous subsection) suggest that significant adaptation does not
occur. However, in the mine scenarios, the reaches with the highest natural copper levels would be
destroyed and effluents and leachates would enter downstream or in other tributaries or watersheds, so
this source of uncertainty is largely moot.
Another source of uncertainty is the assumption that the State of Alaska would adopt the national
copper criterion as a state standard or apply it on a site-specific basis to any mine in the Bristol Bay
watershed. If the state retains and applies the current standard, the effects of copper on salmon and
other aquatic organisms in permitted effluents would be greater by a factor of approximately 1.3 to 2.0,
based on differences among receiving streams (Table 8-11).
8.2.2.2 Other Metals
Chronic national ambient water quality criteria, state standards, or equivalent benchmarks were used to
screen the constituents of tailings, waste rock, and product concentrate leachates (Section 8.1.1). Those
that were retained in the screening were carried forward to release, transport, and dilution modeling.
For hardness-dependent criteria, screening values were calculated for each receiving stream using mean
hardness (Table 8-15). Finally, those potential contaminants that exceeded screening values in streams
after dilution, or that were otherwise of concern, are discussed in more detail below.
Table 8-15. Hardness-dependent acute water quality criteria (CMC) and chronic water quality
criteria (CCC) for the three potential receiving streams in the mine scenarios. All values are in ug/L.
Criteria
CdCMC
CdCCC
PbCMC
PbCCC
NiCMC
NiCCC
ZnCMC
ZnCCC
South Fork Koktuli River
0.45
0.085
12
0.46
130
14
32
32
North Fork Koktuli River
0.30
0.064
7.4
0.29
91
10
23
23
Upper Talarik Creek
0.50
0.089
15
0.58
150
17
38
38
Notes:
CMC = criterion maximum concentration; CCC = criterion continuous concentration.
Aluminum
The environmental chemistry and resulting toxicity of aluminum are complex (Gensemer and Playle
1999). Aluminum is more soluble at acidic and alkaline pHs and less soluble at circumneutral pH. It
occurs in a variety of forms, including soluble complexes with common anions and humic and fulvic
acids, but in most streams soluble and insoluble hydroxide compounds dominate. Free ionic aluminum
is expected to be a small component of dissolved aluminum at the circumneutral pHs found in the
streams draining the mine footprints, unless mine drainage acidifies them. Aluminum is most toxic in
mixing zones where acidic waters mix with neutral or basic ambient waters, apparently due to
precipitation at the gill surface. In general, fish are more sensitive to aluminum than invertebrates.
Bristol Bay Assessment
8-29
January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
Cadmium
Cadmium is an uncommon but highly toxic divalent metal (Mebane 2010). A series of rainbow trout
acute median lethal concentration (LCso) values for cadmium, at hardness values from 7 to 32 mg/L,
ranged from 0.34 to 1.3 ug/L (Mebane 2010, Mebane et al. 2012). A 53-day early-life-stage test of
rainbow trout (at 21 mg/L hardness) gave a chronic value for survival and growth of 0.88 ug/L, but the
test was interrupted prior to completion due to quality control issues (Mebane et al. 2008). A later test
in the same series (but without those quality control issues and at 29 mg/L hardness) gave a higher
rainbow trout chronic value of 1.6 ug/L. Acute tests with mayflies, stoneflies, and caddisflies all resulted
in values that were much higher than the trout values (Mebane et al. 2012). The tests by Mebane etal.
(2008, 2012) were conducted for the State of Idaho to support the derivation of site-specific criteria for
the Coeur d'Alene River. BLM-derived acute values for Ceriodaphnia dubia were 37 to 51 ug/L for the
three streams draining the mine footprints. This is consistent with the relative insensitivity of
invertebrates to acute lethality. Although these and other tests in the literature show fish to be more
sensitive to cadmium than invertebrates in acute exposures, invertebrates were more sensitive in
chronic exposures (Mebane 2010). In particular, mortality of the amphipod Hyalella azteca increased at
0.16 ug/L cadmium at relevant hardness (17 mg/L) (Mebane 2010).
Cobalt
Current studies of the aquatic toxicity of cobalt can be found in a recent review (Environment Canada
and Health Canada 2011). Acutely lethal concentrations range from 89 to 585,800 ug/L. Chronically
toxic concentrations for invertebrates range from 2.9 to 155 ug/L (with the exception of a 1972 4-day
test for rotifers, which resulted in 59,000 ug/L). Only three fish species have been tested for chronic
effects, yielding values of 340 to 2,171 ug/L; the least sensitive of these three species was rainbow trout.
In experimental studies, Chinook salmon avoided waters with cobalt concentrations of 24 ug/L but
rainbow trout were less sensitive, avoiding concentrations of 188 ug/L (Hansen et al. 1999). It is
expected that the same water quality parameters that modify copper toxicity also affect cobalt toxicity,
but existing data are insufficient to perform adjustments.
Lead
Lead is a divalent metal with national criteria and state standards based on water hardness (USEPA
1986, Eisler 2000). A BLM is available that estimates acute LCso values for fathead minnows in the South
and North Fork Koktuli Rivers as 382 and 383 ug/L, respectively. In comparison, a rainbow trout test at
hardness similar to that in the South and North Fork Koktuli Rivers (20 mg/L) resulted in an LCso of 120
ug/L; the closely related cutthroat trout produced an LCso as low as 47 ug/L at a hardness of 11 mg/L
(Mebane et al. 2012). Tests at similar hardness levels for mayflies, stoneflies, caddisflies, and chironomid
midges gave higher LCsos (253 to more than 1,255 ug/L) (Mebane et al. 2012). This indicates that, for
acute lethality, trout species are more sensitive to lead than aquatic insect larvae, which is consistent
with BLM-derived acute values of 523 to 748 ug/L for Daphnia magna. Chronic tests gave values for
reduced rainbow trout weight and length of 36.0 and 12.1 ug/L at 21 and 29 mg/L hardness,
respectively, and for the midge Chironomus tentans of 65.4 ug/L at 32 mg/L hardness (Mebane et al.
Bristol Bay Assessment 830 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
2008). Note that we use tests performed for the State of Idaho (Mebane et al. 2008, 2012) for cadmium,
lead, and zinc, because they are high-quality tests that use species and water chemistries relevant to the
Bristol Bay watershed.
Manganese
The toxicity of manganese is strongly related to water hardness. Acutely lethal concentrations for
manganese in soft water range from 0.8 to 4.83 mg/L for invertebrates and 2.4 to 3,350 mg/L for fish.
The most sensitive acutely tested fish was coho salmon. In four soft-water tests of rainbow and brown
trout, chronic values ranged from 0.79 to 14.6 mg/L. More details can be found in recent reviews
(Reimer 1999, IPCS 2004).
Selenium
Selenium is a bioaccumulative and moderately biomagnifying element. Dissolved oxyanions of selenate
(Se+4) and selenite (Se+6) are taken up by microbes, algae, and plants and converted to organic forms. In
streams, periphyton growing on rocks and woody debris are the primary community that performs this
conversion and conversion rates are relatively low. Selenium causes deformities and death in fish
embryos and larvae, which are exposed to selenium accumulated by their mothers. Therefore, potential
selenium toxicity is of concern for resident but not anadromous fish. Effects of selenium on salmonids
have been studied below mines in British Columbia. For example, cutthroat trout embryos from a pond
with selenium concentrations of 93 ug/L at a coal mine in British Columbia showed effects ranging from
larval deformities to mortality (Rudolph et al. 2008). The probability of mortality was correlated with
selenium concentrations in the embryos. Invertebrates are less sensitive to selenium than fish.
The complex dynamics of selenium and its various forms have led to complex water quality criteria
(USEPA 2004). The acute national criteria, based on the proportions of selenite and selenate,
are!85.9 ug/L selenite and 12.82 ug/L selenate. The chronic criterion is 5.0 ug/L total selenium.
However, because the transformations and bioaccumulative processes are so complex, a chronic
criterion for fish tissue concentrations (7.91 ug/g whole body dry weight) has been proposed based on
juvenile bluegill sunfish mortality. The genus mean chronic value for rainbow trout is a little higher
(9.32 ug/g dry weight). These tissue-based values are believed to be more accurate than benchmarks
based on water concentrations. However, implementing the criterion or using the dietary toxicity test
and field data that were used in its derivation would require a model of selenium bioaccumulation that
is applicable to streams and lakes in the Bristol Bay watershed. No such model is currently available.
Zinc
Zinc, like copper, is a divalent metal and trace nutrient that is a common aquatic toxicant. The national
criteria and state standard are based on water hardness (USEPA 1987), but a BLM is available that
provides more accurate predictions of acute toxicity, at least for some test species (DeForest and Van
Genderen 2012). The BLM-based LCso estimates for rainbow trout in the South and North Fork Koktuli
Rivers are 64 and 63 ug/L, respectively. In comparison, results of a series of 17 rainbow trout LCso tests
(at hardnesses of 7 to 71 mg/L) ranged from 20 to 289 ug/L (Mebane et al. 2012). Acute tests at
Bristol Bay Assessment 831 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
14 mg/L hardness for two mayfly species and a caddisfly species resulted in values greater than
2,926 ug/L, and a stonefly species test resulted in values greater than 1,526 ug/L (Mebane et al. 2012).
These results suggest that an endpoint fish species is considerably more sensitive to zinc than relevant
stream invertebrates in acute exposures. BLM-derived acute values for Daphnia magna were 407 to
502 ug/L for the three receiving streams, which is consistent with the relative insensitivity of
invertebrates. The chronic value (20% effective concentration [EC2o] for survival) from a 69-day, early-
life-stage test of rainbow trout in 21 mg/L hardness water was 147 ug/L (Mebane et al. 2012).
8.2.2.3 Total Dissolved Solids
The Alaskan Water Quality Standard for Growth and Propagation of Fish, Shellfish, Other Aquatic Life
and Wildlife states: "TDS may not exceed 1,000 mg/L. A concentration of TDS may not be present in
water if that concentration causes or reasonably could be expected to cause an adverse effect on aquatic
life" (ADEC 2011). Meeting the state standard for TDS proved difficult at the Red Dog zinc and lead mine
(USEPA 1998, 2008). Laboratory tests of synthetic TDS for effluents from Red Dog and Kensington
Mines caused no statistically significant effects on rainbow trout embryo viability or fry survival or
weight, but did show statistically significant effects on chironomid larvae at 2,089 mg/L (Red Dog) and
1,750 mg/L (Kensington) (Chapman et al. 2000). However, the toxicity of TDS depends on the specific
mixture composition, and chironomids are relatively tolerant of major ion mixtures (USEPA 2011). Also,
the rainbow trout tests did not include the State of Alaska's endpoint of concern, egg fertilization.
8.2.2.4 Whole Leachates and Effluents
Metals and other aqueous contaminants have combined toxic effects that may be concentration additive,
effects additive, or more or less than additive. The assumption of concentration additivity is considered
to provide the best general approximation of combined metals effects. If, as in this assessment, the
number of metals potentially discharged is large, less than and more than additive interactions may
roughly average out. However, pairwise laboratory tests of defined metal mixtures indicate the
complexity of potential interactions. For example, Chinook salmon avoided a mixture of 0.9 ug/L cobalt
and 1.0 ug/L copper, which suggests that cobalt has no effect on copper avoidance at low levels (Hansen
et al. 1999). However, at overtly toxic copper levels (43 ug/L) cobalt did increase avoidance (Hansen et
al. 1999).
As discussed above with respect to copper (Section 8.2.2.1), field studies of streams contaminated by
copper and other metals indicate that laboratory-based criteria are not fully protective of aquatic
communities. Hence, the screening of metal mixtures by applying an additive model to criteria and
equivalent benchmarks probably does not overestimate effects on aquatic communities.
8.2.2.5 Ore-Processing Chemicals
Of the proposed ore-processing chemicals, sodium ethyl xanthate is the primary contaminant of concern
(Section 6.4.2.3). An assessment by Environment Australia generated a predicted no effect concentration
of 1 ug/L (NICNAS 2000). Australia and New Zealand have established a trigger value of 0.05 ug/L to
protect aquatic life (ANZECC 2000). However, because relatively little testing has been done, this is a
Bristol Bay Assessment 832 January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
"low reliability" value that "may not protect the most sensitive species." Rainbow trout appear to be
relatively tolerant of sodium ethyl xanthate, with lethal concentrations ranging from 1 to 50 mg/L
depending on test conditions (Fuerstenau et al. 1974, Webb et al. 1976). Other fishes had median lethal
concentrations of 0.01 to 10 mg/L (emerald shiner) and 0.32 to 3.2 mg/L (fathead minnow) (NICNAS
1995). Aquatic invertebrates are represented by only Daphnia magna, which has a median effective
concentration (EC50) of 0.35 mg/L (Xuetal. 1988).
8.2.3 Risk Characterization
Risk characterization was performed in stages. First, screening was performed against mean
concentrations in tailings and waste rock leachates to determine whether leachates pose a potential risk
and which constituents contribute to risks. Second, contaminants of concern from the initial screening
were screened against estimated ambient concentrations for routine operations and for WWTP failure.
The implications of potential toxic effects are discussed in terms of their spatial distributions.
Contaminants were screened for risks to aquatic biota by comparing exposure levels to criteria or other
ecotoxicological benchmarks using a risk quotient (Box 8-3). This conventional approach was used to
determine which contaminants and materials are likely to be toxic (Tables 8-4 through 8-8).
BOX 8-3. USE OF RISK QUOTIENTS TO ASSESS TOXICOLOGICAL EFFECTS
A risk quotient (Q) equals the exposure level divided by an ecotoxicological benchmark. If the quotient
exceeds 1, the effect implied by the benchmark is expected to occur, but with some uncertainty (see below).
Quotients much larger than 1 suggest larger effects than those that define the benchmark, with greater
confidence that an adverse effect would occur. Quotients much smaller than 1 suggest that even small
effects are unlikely. The acute criterion, or criterion maximum concentration (CMC), estimates a
concentration at which 5% of aquatic species experience some mortality among later life-stages in short-
term exposures. The chronic criterion, or criterion continuous concentration (CCC), estimates a
concentration at which 5% of aquatic species experience decreased survival, growth, or reproduction in
longer-term exposures. The criteria, or equivalent numbers when criteria are not available, are relatively well-
accepted as approximate thresholds for significant effects. Thus, these values are the ecotoxicological
benchmarks used as the divisor for calculating quotients in the screening portion of this assessment.
To describe the results of screening using chronic criteria or equivalent benchmarks in a consistent manner
(acute criteria and less protective benchmarks would be interpreted differently), the following scale was
developed:
• Q < 1 = not overtly toxic
• 2 > Q > 1 = marginally toxic
• 10 > Q > 2 = moderately toxic
• 100 > Q > 10 = highly toxic
• Q > 100 = extremely toxic
8.2.3.1 Screening Leachate Constituents
The results of screening inorganic leachate constituents from tailings and waste rock tests are presented
in Tables 8-4 through 8-8. All have at least moderate chronic toxicity based on their estimated total
toxicity (sum of chronic toxic units), and all are predicted to be acutely toxic if the BLM-based copper
criterion is used instead of the state standard. In all cases, copper is the dominant source of toxicity. The
Bristol Bay Assessment
8-33
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
acidic pre-Tertiary leachate is estimated to be extremely toxic, with copper concentrations thousands of
times higher than the chronic criterion. Figure 8-3 shows the copper concentrations of leachates and
ambient water in relation to state standards.
Tailings slurry would also contain processing chemicals, particularly sodium ethyl xanthate. The
predicted concentration in the slurry (1.5 mg/L) is above or within the range of acute lethality for fish
and well above the level for Daphnia magna. Therefore, the aqueous phase of the slurry delivered to the
TSF would be moderately toxic due to xanthate alone.
Figure 8-3. Comparison of copper concentrations in leachates and background water to state
hardness-based acute (CMC) and chronic (CCC) water quality criteria for copper. North Fork Koktuli
River = background water; Tails HCT = leachate from humidity tests of tailings; Supernatant =
leachate from column tests of tailings; PWZ = Pebble West pre-Tertiary; FEZ = Pebble East pre-
Tertiary; CMC = criterion maximum concentration; CCC = criterion continuous concentration. Copper
concentrations in tailings leachate in the field would be expected to lie in the lower blue triangle.
Copper concentrations in waste rock leachate would be expected to lie in the upper blue triangle.
Data sources: Appendix H and PLP 2011.
10,000
1,000
^ 100
Q.
a.
8
-a
01
"o
10 =
1
Supernatant
Mean
North Fork Kokluli
50
100 150 200 250
Hardness (mg/L CaCO3)
300
350
8.2.3.2 Screening Contaminants in Receiving Waters
Concentrations of contaminants of concern were calculated at the gages on all three receiving streams
for each mine scenario (Pebble 0.25, 2.0, and 6.5), and those concentrations were screened against
chronic criteria and benchmarks. Because the Pebble 6.5 scenario resulted in the highest concentrations,
screening results for this mine size are presented in Tables 8-16 and 8-17 to show which contaminants
remain of concern after dilution. In the Pebble 6.5 scenario under routine operations, copper is
estimated to exceed chronic water quality criteria at all stations on the South Fork Koktuli River, two of
six stations on the North Fork Koktuli River, and three of seven stations on Upper Talarik Creek (Table
Bristol Bay Assessment
8-34
January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
8-16). The pattern of exceedance is the same for the Pebble 6.5 scenario with WWTP failure, except that
all stations on the North Fork Koktuli River exceed the copper criterion (Table 8-17). Cadmium and zinc
also exceed chronic criteria, but at fewer stations and by much smaller magnitudes. No other metal
exceeded a criterion or benchmark.
Concentrations of major ions are a particular concern at mine sites because of the leaching of large
volumes of crushed rock. However, estimates of TDS, both with and without WWTP failure, are within
state standards (Tables 8-16 and 8-17). Without toxicity information on the dissolved solids mixtures
that would occur at the site, we must assume that the standard is protective.
The concentration of sodium ethyl xanthate was not estimated in the receiving streams. Although the
aqueous phase of the tailings slurry would be toxic due to xanthate, we expect that xanthate would occur
at non-toxic levels in ambient waters below TSFs due to degradation and dilution (Xu et al. 1988).
8.2.3.3 Screening Total Metal Toxicity in Receiving Waters
Table 8-18 presents the sums of quotients across the nine metals of concern (excluding selenium, which
has a different mode of exposure and toxicity), for all three mine scenarios and gage locations. In
addition, the sums of quotients for background water are presented. As is expected for streams draining
a surficial ore body, background metal concentrations are elevated. Although total metal toxicities are
estimated to be significantly higher than any individual metal, copper is responsible for most of the
estimated toxicity. Therefore, copper concentrations in contributing loads and ambient waters, as well
as quotients with respect to chronic criteria for the receiving waters, are presented for all three mine
scenarios and gage locations in Table 8-19. The same information but with WWTP failure is presented in
Table 8-20.
8.2.3.4 Screening for Severity of Effects
Tables 8-19 and 8-20 indicate that water management, assuming both routine operations and WWTP
failure, results in exceedance of the BLM-based chronic copper criteria at several locations in the mine
scenario watersheds. However, they do not provide an indication of the severity of effects. For that
purpose, we estimated copper concentrations in individual stream reaches. The reaches are defined by
flow gages and major confluences for each of the receiving streams (Table 8-21). Those fully-mixed
copper concentrations were screened against a series of benchmarks of increasing severity, beginning
with the national criteria, as follows.
• Invertebrate chronic (1C). The BLM-derived chronic ambient water quality criterion (Table 8-11),
based on toxicity to sensitive aquatic invertebrates in extended exposures. It implies reduced
survival, growth, or reproduction of copper-sensitive invertebrates.
• Invertebrate acute (IA). The BLM-derived acute ambient water quality criterion (Table 8-11), based
on lethality to sensitive aquatic invertebrates in short-term exposures. It implies greatly reduced
survival of copper-sensitive invertebrates.
Bristol Bay Assessment 835 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
• Fish avoidance (FA). The BLM-derived concentration at which 20% of rainbow trout avoid the
contaminated water (Table 8-14). It implies loss of habitat due to aversion.
• Fish sensory (FS). The BLM-derived concentration at which the olfactory sensitivity of rainbow
trout is reduced by 20% (Table 8-14). It implies an inability to identify natal streams, reduced
predator avoidance, and other behavioral effects.
• Fish reproduction (FR). The BLM-derived chronic value for rainbow trout, the concentration at
which their fecundity or the survival and growth of their larvae are reduced (Table 8-13). It implies
partial or complete reproductive failure of salmonids.
• Fish kill (FK). The BLM-derived rainbow trout LCso, the concentration at which half of adults and
juveniles are killed in short-term exposures (Table 8-13). It implies a fish kill and, in the long term,
local extirpation offish populations.
These copper benchmarks were applied to streams reaches (Table 8-21) rather than the point values in
the prior screening assessment. The combinations of stream reaches and mine scenarios at which these
copper benchmarks are exceeded are shown in Tables 8-22 and 8-23. The range of effect severity
extends from no overt effects expected (-) through the full range of effects up to numerous dead post-
larval salmonids (IC/IA/FA/FS/FR/FK).
Bristol Bay Assessment 836 January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-16. Estimated concentrations of contaminants of concern and associated risk quotients for the Pebble 6.5 scenario, assuming routine operations, at
how risk quotients were calculated.
Stream and
Gage
Copper
Mg/L
Quotient
Aluminum
Mg/L
Quotient
Cadmium
Mg/L
Quotient
Cobalt
Mg/L
Quotient
Manganese
Mg/L
Quotient
Nickel
Mg/L
Quotient
locations in the mine scenario watersheds. See Box 8-3 for a description of
Lead
Mg/L
Quotient
Selenium
Mg/L
Quotient
Zinc
Mg/L
Quotient
Total Dissolved Solids
mg/L
Quotient
South Fork Koktuli River
SK100G
SK100F
SK100CP2a'b
SK124A
SK124CPa'c
SK100C
SKlOOCPla
SK119A
SK119CPa
SK100B1
SK100B"
NA
160
56
1.4
1.4
20
20
NA
1.5
11
7.9
NA
150
52
1.3
1.3
18
18
NA
1.4
10
7.3
NA
57
28
56
55
46
45
NA
18
28
22
NA
0.65
0.32
0.65
0.63
0.52
0.52
NA
0.21
0.32
0.25
NA
0.23
0.09
0.06
0.23
0.07
0.07
NA
0.03
0.04
0.03
NA
3.6
1.4
0.91
3.6
1.1
1.1
NA
0.44
0.69
0.54
NA
1.5
0.55
1.6
1.5
1.2
1.2
NA
0.29
0.69
0.50
NA
0.58
0.22
0.63
0.62
0.47
0.47
NA
0.12
0.28
0.20
NA
98
58
82
80
62
61
NA
15
35
26
NA
0.14
0.08
0.12
0.12
0.09
0.09
NA
0.02
0.05
0.04
NA
1.8
0.89
2.5
2.5
1.9
1.9
NA
0.49
1.2
0.89
NA
0.18
0.090
0.25
0.25
0.19
0.19
NA
0.05
0.12
0.09
NA
0.11
0.10
0.23
0.23
0.19
0.19
NA
0.08
0.12
0.10
NA
0.38
0.33
0.80
0.78
0.65
0.64
NA
0.29
0.39
0.35
NA
0.73
0.35
1.2
1.2
0.88
0.88
NA
0.49
0.61
0.47
NA
0.15
0.07
0.24
0.24
0.18
0.17
NA
0.10
0.12
0.09
NA
33
13
18
18
15
15
NA
3.3
9.5
7.6
NA
1.4
0.58
0.79
0.78
0.67
0.67
NA
0.14
0.41
0.33
NA
52
62
49
390
260
260
NA
57
170
120
NA
0.05
0.06
0.05
0.39
0.26
0.26
NA
0.06
0.17
0.12
North Fork Koktuli River
NK119A
NK119CP23
NK119B
NK119CP13
NK1000
NK100B
NK100A1
NK100A6
1.9
1.6
0.63
1.2
0.62
0.74
0.74
0.54
1.8
1.5
0.59
1.1
0.58
0.69
0.69
0.51
23
22
20
21
34
30
18
20
0.26
0.25
0.23
0.24
0.39
0.34
0.20
0.23
0.03
0.03
0.02
0.02
0.03
0.03
0.02
0.02
0.52
0.45
0.31
0.38
0.53
0.45
0.29
0.29
0.30
0.26
0.22
0.22
0.83
0.65
0.32
0.28
0.12
0.10
0.09
0.09
0.33
0.26
0.13
0.11
23
19
8.0
15
49
37
18
21
0.03
0.03
0.01
0.02
0.07
0.05
0.03
0.03
0.59
0.54
0.46
0.49
1.5
1.2
0.66
0.63
0.06
0.05
0.05
0.05
0.15
0.12
0.07
0.06
0.07
0.07
0.06
0.06
0.16
0.16
0.07
0.09
0.23
0.23
0.20
0.21
0.54
0.55
0.24
0.29
0.62
0.52
0.24
0.41
0.66
0.57
0.35
0.31
0.12
0.10
0.05
0.08
0.13
0.11
0.071
0.06
2.7
2.5
2.2
2.3
9.9
8.0
5.0
4.2
0.12
0.11
0.09
0.10
0.43
0.35
0.22
0.18
63
57
38
49
230
180
100
92
0.06
0.06
0.04
0.05
0.23
0.18
0.10
0.09
Upper Talarik Creek
UT100E
UT100D
UT100C2
UT100C1
UT100C
UT119Ab
UT100B'
0.81
1.3
0.38
0.34
0.41
27
3.6
0.75
1.3
0.36
0.32
0.38
25
3.3
19
35
13
9.7
11
17
10
0.22
0.40
0.15
0.11
0.13
0.20
0.12
0.05
0.08
0.01
0.01
0.01
0.05
0.02
0.72
1.2
0.19
0.18
0.18
0.72
0.29
0.67
1.3
0.14
0.11
0.09
0.27
0.11
0.27
0.52
0.06
0.04
0.03
0.11
0.05
20
60
24
14
10
28
17
0.03
0.09
0.03
0.02
0.02
0.04
0.03
1.1
1.8
0.50
0.45
0.44
0.61
0.46
0.11
0.18
0.05
0.05
0.04
0.06
0.05
0.06
0.07
0.04
0.04
0.04
0.08
0.07
0.22
0.24
0.14
0.13
0.14
0.26
0.24
0.43
0.69
0.19
0.18
0.18
0.24
0.17
0.09
0.14
0.04
0.04
0.04
0.05
0.03
4.3
6.3
1.9
1.3
1.5
7.0
2.4
0.19
0.27
0.09
0.06
0.06
0.30
0.10
75
89
48
49
48
49
47
0.08
0.09
0.05
0.05
0.05
0.05
0.05
Notes:
a Confluence point where virtual gage was created because physical gage does not exist.
b 1/3 of total return flow is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location.
c Wastewater treatment plant discharges 50% of its flow at this site.
d USGS 15302200.
e USGS 15302250.
' USGS 15300250.
NA = not applicable; the stream at the gage would be destroyed.
Bristol Bay Assessment
8-37
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-17. Estimated concentrations of contaminants of concern and associated risk quotients for the Pebble 6.5 scenario, assuming wastewater treatment plant failure, at locations in the mine scenario watersheds. Upper Talarik
Creek would be unchanged from Table 8-16. See Box 8-3 for a description of how risk quotients were calculated.
Stream and
Gage
Copper
Mg/L
Quotient
Aluminum
Mg/L
Quotient
Cadmium
Mg/L
Quotient
Cobalt
Mg/L
Quotient
Manganese
Mg/L
Quotient
Nickel
Mg/L
Quotient
Lead
Mg/L
Quotient
Selenium
Mg/L
Quotient
Zinc
Mg/L
Quotient
Total Dissolved Solids
mg/L
Quotient
South Fork Koktuli River
SK100G
SK100F
SK100CP2a'b
SK124A
SK124CPa'c
SK100C
SKlOOCPla
SK119A
SK119CPa
SK100B1
SK100B"
NA
160
56
110
100
86
86
NA
1.5
47
34
NA
150
52
100
97
80
80
NA
1.4
44
32
NA
57
28
56
56
46
45
NA
18
28
22
NA
0.65
0.32
0.65
0.63
0.52
0.52
NA
0.21
0.32
0.25
NA
0.23
0.09
0.20
0.19
0.16
0.15
NA
0.03
0.09
0.07
NA
3.6
1.3
3.1
3.0
2.4
2.4
NA
0.44
1.4
1.1
NA
1.5
0.55
1.6
1.5
1.2
1.2
NA
0.29
0.70
0.50
NA
0.58
0.22
0.64
0.62
0.47
0.47
NA
0.12
0.28
0.20
NA
98
58
82
80
62
61
NA
15
35
26
NA
0.14
0.08
0.12
0.12
0.09
0.09
NA
0.02
0.05
0.04
NA
1.8
0.90
2.5
2.5
1.9
1.9
NA
0.49
1.2
0.89
NA
0.18
0.09
0.25
0.25
0.19
0.19
NA
0.05
0.12
0.09
NA
0.11
0.10
0.26
0.26
0.21
0.21
NA
0.08
0.13
0.11
NA
0.38
0.33
0.90
0.89
0.71
0.71
NA
0.29
0.43
0.37
NA
0.73
0.35
1.2
1.2
0.88
0.88
NA
0.49
0.61
0.47
NA
0.15
0.07
0.24
0.24
0.18
0.18
NA
0.10
0.12
0.09
NA
33
13
26
25
20
20
NA
3.3
12
9.5
NA
1.4
0.58
1.1
1.1
0.88
0.88
NA
0.14
0.52
0.41
NA
62
49
390
380
260
260
NA
57
170
120
NA
0.06
0.05
0.39
0.38
0.26
0.26
NA
0.06
0.17
0.12
North Fork Koktuli River
NK119A
NK119CP23
NK119B
NK119CP13
NK1000
NK100B
NK100A1
NK100A6
0.07
0.07
0.06
1.2
57
43
20
17
0.23
0.23
0.20
1.1
54
40
19
16
0.63
0.52
0.24
21
34
30
18
20
0.12
0.10
0.05
0.24
0.39
0.34
0.20
0.23
0.03
0.03
0.02
0.02
0.11
0.08
0.04
0.04
0.52
0.45
0.31
0.38
1.7
1.3
0.69
0.63
0.30
0.26
0.22
0.22
0.83
0.65
0.32
0.28
0.12
0.10
0.09
0.09
0.33
0.26
0.13
0.11
23
19
8.0
15
49
37
18
21
0.03
0.03
0.01
0.02
0.07
0.05
0.03
0.03
0.59
0.54
0.46
0.49
1.5
1.2
0.66
0.63
0.06
0.05
0.05
0.05
0.15
0.12
0.07
0.06
0.07
0.07
0.06
0.06
0.17
0.17
0.08
0.09
0.23
0.23
0.20
0.21
0.60
0.60
0.26
0.31
0.62
0.52
0.24
0.41
0.66
0.57
0.35
0.31
0.12
0.10
0.05
0.08
0.13
0.11
0.07
0.06
2.7
2.5
2.2
2.3
14
11
6.3
5.3
0.12
0.11
0.09
0.10
0.60
0.48
0.27
0.23
63
57
38
49
230
180
100
92
0.06
0.06
0.04
0.05
0.23
0.18
0.10
0.09
Notes:
a Confluence point where virtual gage was created because physical gage does not exist.
b 1/3 of total return flow is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location.
c Wastewater treatment plant discharges 50% of its flow at this site.
d USGS 15302200.
e USGS 15302250.
NA = not applicable; the stream at the gage would be destroyed.
Bristol Bay Assessment
8-38
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-18. Estimated total toxicity of metals of concern for each mine scenario, under routine
operations and with wastewater treatment plant failure, at locations in the mine scenario
watersheds. Values are the sums of the toxic quotients for the metals of concern.
Stream and
Gage
Background
Pebble 0.25
Routine
Operations
WWTP
Failure
Pebble 2.0
Routine
Operations
WWTP
Failure
Pebble 6.5
Routine
Operations
WWTP
Failure
South Fork Koktuli River
SK100G
SK100F
SK100CP2a'b
SK124A
SK124CP"
SK100C
SK100CP13
SK119A
SK119CPa
SK100B1
SK100B"
3.1
2.4
-
2.5
2.3
1.2
-
1.2
1.1
3.4
2.7
2.6
3.2
3.1
2.7
2.7
1.2
1.2
1.3
1.2
3.4
2.7
2.6
20
20
11
11
1.2
1.2
4.7
3.6
100
22
12
3.4
3.4
7.9
7.9
1.3
1.3
3.3
2.6
100
22
12
26
25
20
19
1.3
1.3
7.8
5.8
NA
160
55
5.7
5.6
22
22
2.6
2.8
12
9.2
NA
157
55
107
104
86
86
2.6
2.8
47
34
North Fork Koktuli River
NK119A
NK119CP23
NK119B
NK119CP13
NK100O
NK100B
NK100A1
NK100A6
1.0
-
1.0
-
1.1
1.2
1.1
1.1
1.6
1.5
1.0
1.4
1.6
1.6
1.3
1.2
1.6
1.5
1.0
1.4
9.7
6.6
3.4
3.0
3.2
2.8
1.1
2.1
1.7
1.8
1.4
1.3
3.2
2.8
1.1
2.1
12
8.8
4.2
3.6
3.2
2.8
1.6
2.3
3.2
2.9
1.9
1.8
3.2
2.8
1.6
2.3
57
43
21
17
Upper Talarik Creek
UT100E
UT100D
UT100C2
UT100C1
UT100C
UT119Ab
UT100B'
0.93
1.3
0.90
0.76
0.89
0.75
0.99
1.0
1.3
0.93
0.80
0.92
1.8
1.1
0.96
1.3
0.93
0.80
0.92
1.8
1.1
1.0
1.8
1.0
0.86
1.0
6.5
1.8
1.0
1.8
1.0
0.9
1.0
6.5
1.8
2.6
4.3
1.1
0.9
1.0
27
4.2
2.6
4.3
1.1
0.94
1.0
27
4.2
Notes:
8 Confluence point where virtual gage was created because physical gage does not exist; dash (-) indicates that no background value is
available.
b 1/3 of total return flow is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location.
c Wastewater treatment plant discharges 50% of its flow at this site.
d USGS 15302200.
e USGS 15302250.
' USGS 15300250.
WWTP = wastewater treatment plant; NA = not applicable, because the stream at the gage would be destroyed.
Bristol Bay Assessment
8-39
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-19. Background copper concentrations and, for each mine scenario, copper concentrations in contributing loads and ambient
waters (fully mixed reaches below each gage) and associated risk quotients, assuming routine operations. See Box 8-3 for a description of
how risk quotients were calculated. All concentrations are in ug/L
Stream and
Gage
Background
Pebble 0.25
CL
AW
Quotient
Pebble 2.0
CL
AW
Quotient
Pebble 6.5
CL
AW
Quotient
South Fork Koktuli River
SK100G
SK100F
SK100CP2a."
SK124A
SK124CP"
SK100C
SK100CP13
SK119A
SKllQCPa
SK100B1
SK100B"
2.4
1.6
1.6
1.4
1.4
1.4
1.4
0.42
0.42
0.62
0.54
3.2
3.2
1.7
1.1
-
-
-
-
2.4
1.7
1.7
1.3
1.3
1.4
1.4
0.42
0.42
0.54
0.47
2.2
1.6
1.5
1.2
1.2
1.3
1.3
0.39
0.39
0.51
0.44
380
12
11
1.1
-
-
3.5
-
100
21
11
1.3
1.3
6.5
6.5
0.43
0.43
2.5
1.9
94
20
10
1.2
1.2
6.1
6.0
0.41
0.41
2.4
1.8
670
56
1.4
-
3.2
-
5.2
3.4
4.0
NA
160
56
1.4
1.4
20
20
1.5
1.5
11
7.9
NA
150
52
1.3
1.3
18
18
1.4
1.4
10
7.3
North Fork Koktuli River
NK119A
NK119CP2a
NK119B
NK119CP13
NK1000
NK100B
NK100A1
NKlOOAe
0.31
0.31
0.41
0.33
0.35
0.40
0.61
0.41
5.1
-
-
1.1
-
0.70
0.64
0.41
0.58
0.43
0.52
0.66
0.45
0.65
0.60
0.39
0.55
0.41
0.49
0.61
0.42
5.0
3.3
3.2
-
1.1
-
3.4
1.9
1.5
0.42
1.1
0.43
0.62
0.70
0.49
1.8
1.4
0.39
1.0
0.40
0.58
0.65
0.46
5.0
3.3
3.7
-
1.1
-
3.4
1.9
1.6
0.63
1.2
0.62
0.74
0.74
0.54
1.8
1.5
0.58
1.1
0.58
0.69
0.69
0.51
Upper Talarik Creek
UT100E
UT100D
UT100C2
UT100C1
UT100C
UT119Ab
UT100B'
0.34
0.50
0.34
0.30
0.38
0.21
0.34
-
-
-
1.7
0.34
0.50
0.34
0.30
0.38
0.99
0.43
0.32
0.47
0.31
0.28
0.35
0.93
0.40
3.2
-
-
11
0.34
0.63
0.36
0.32
0.39
5.8
1.04
0.32
0.59
0.33
0.30
0.37
5.4
0.97
3.2
3.2
3.2
-
56
0.81
1.3
0.38
0.34
0.41
27
3.6
0.75
1.3
0.36
0.32
0.38
25
3.3
Notes:
NA = not applicable, because stream at gage location would be destroyed. Dashes (-) indicate there are no contributing loads at that gage under that scenario.
a Confluence point where virtual gage was created because physical gage does not exist.
b 1/3 of total return flow is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location.
c Wastewater treatment plant discharges 50% of its flow at this site.
d USGS 15302200.
e USGS 15302250.
' USGS 15300250.
CL = contributing loads; AW = ambient waters; quotient = predicted/criterion.
Bristol Bay Assessment
8-40
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-20. Background copper concentrations and, for each mine scenario, copper concentrations in contributing loads and ambient
waters (fully-mixed reaches below each gage) and associated risk quotients, assuming wastewater treatment plant failure. Upper Talarik
Creek would be unchanged from Table 8-19. See Box 8-3 for a description of how risk quotients were calculated. All concentrations are in
ug/L.
Stream and
Gage
Background
Pebble 0.25
CL
AW
Quotient
Pebble 2.0
CL
AW
Quotient
Pebble 6.5
CL
AW
Quotient
South Fork Koktuli River
SK100G
SK100F
SK100CP2a."
SK124A
SK124CPa'c
SK100C
SK100CP13
SK119A
SK119CP3
SK100B1
SKIDDED
2.4
1.6
1.6
1.4
1.4
1.4
1.4
0.42
0.42
0.62
0.54
3.2
3.2
1.7
75
-
-
-
2.4
1.7
1.7
20
19
11
11
0.42
0.42
4.1
3.0
2.2
1.6
1.5
18
17
9.9
9.9
0.39
0.39
3.8
2.8
380
12
11
100
-
-
3.5
-
100
21
11
25
24
19
19
0.43
0.43
7.2
5.2
94
20
10
23
22
17
17
0.41
0.41
6.7
4.8
-
670
56
140
3.2
-
5.2
3.4
4.0
NA
160
56
110
100
86
86
1.5
1.5
47
34
NA
150
52
100
97
80
80
1.4
1.4
44
32
North Fork Koktuli River
NK119A
NK119CP23
NK119B
NK119CP13
NK100O
NK100B
NK100A1
NK100A6
0.31
0.31
0.41
0.33
0.35
0.40
0.61
0.41
5.1
-
-
75
-
-
0.70
0.64
0.41
0.58
9.0
5.8
2.9
2.3
0.65
0.60
0.39
0.55
8.4
5.4
2.7
2.1
5.0
3.3
3.2
100
-
3.4
-
1.9
1.5
0.42
1.1
11
7.8
3.6
2.9
1.8
1.4
0.39
1.0
11
7.3
3.4
2.7
5.0
3.3
3.7
150
-
3.4
-
1.9
1.6
0.63
1.2
57
43
20
17
1.8
1.5
0.58
1.1
54
40
19
16
Notes:
NA = not applicable, because stream at gage location would be destroyed. Dashes (-) indicate there are no contributing loads at that gage under that scenario.
"" Confluence point where virtual gage was created because physical gage does not exist.
b 1/3 of total return flow is transferred from SK100CP2 to UT119Ato represent interbasin transfer at this location.
c Wastewater treatment plant discharges 50% of its flow at this site.
d USGS 15302200.
B USGS 15302250.
CL = contributing loads; AW = ambient waters; quotient = predicted/criterion.
Bristol Bay Assessment
8-41
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Reach Designation3
Reach Description11
Length (km)c
Concentration Assigned and Qualifiers'1
South Fork Koktuli River—Ma instem
SK100B
SK100B to confluence of the South and
North Fork Koktuli Rivers
23
SK100B, overestimates lower end due to dilution
SK100B1
SKlOOBltoSKlOOB
4.5
SK100B1, small overestimate of lower end due to dilution
SK100CP1/
SK119CP
SK100CP1/ SK119CP confluence to
SK100B1
4.3
Mixed SK100CP1 and SK119CP, little dilution downstream
SK100C
SK100C to SK100CP1
1.2
SK100C, negligible further dilution in short reach
SK100CP2/
SK124CP
SK100CP2/SK124CP confluence to
SK100C
6.4
Mixed SK100CP2 and SK124CP, little dilution downstream
SK100F
SK100F to SK100CP2
11
Mean SK100F and SK100CP2 due to significant dilution
SK100G
SK100G to SK100F (not Pebble 6.5)
3.3/3.3/NA
Mean SK100G and SK100F due to significant dilution
SK Rock
Waste rock to SK100F (Pebble 6.5 only)
NA/NA/0.83
SK 100F, assuming input near base of rock pile and short reach
SK Halo/Rock
Dewatering halo and rock pile to SK100G
(Pebble 0.25 and 2.0)
1.87/0.54/NA
SK 100G, assuming input near base of rock pile and short reach
South Fork Koktuli River—Tributaries
SK Headwaters
Headwaters to SK119A (Pebble 0.25)
7.0
Background for Pebble 0.25 scenario
SKTSF1
TSF1 to SK119A (Pebble 2.0)
6.8
SK119A, significant dilution so underestimate
SK119A
SK119A to SK119CP
1.6/1.6/1.5
SK119A, for Pebble 0.25 and Pebble 2.0 scenarios, dilution is minimal;
SK119CP for remnant reach in Pebble 6.5 scenario, when SK119A destroyed
SK124A
SK124A to SK124CP
2.6
SK124A, no dilution in this reach within precision
SKWWTP
WWTP to SK124A
5.0
SK124A, underestimate of upper end from dilution of WWTP and, in Pebble 6.5 scenario,
TSF 3 leachate
North Fork Koktuli River—Ma instem
NK100A
NK100A to confluence of the South and
North Fork Koktuli Rivers
4.7
NK100A, little dilution
NK100A1
NKlOOAltoNKlOOA
3.4
N100A1, which has a small contributing load in the Pebble 2.0 and Pebble 6.5
scenarios, so small overestimate
NK100B
NKlOOBtoNKlOOAl
20
NK100B, approximately two times dilution over long reach in Pebble 0.25 scenario but
no change in Pebble 2.0 and Pebble 6.5 scenarios due to balance of dilution by tailings
leachate
NK100CP1/NK100C
NK100CP1/NK100C confluence to
NK100B
0.79
Mixed NK100CP1 and NK100C, little dilution downstream
NK100C
NK100C to confluence NK119A stream
0.19
NK100C, negligible further dilution in tiny reach
NKWWTP
WWTP discharge to NK100C
4.3
NK100C, underestimate of upper end, but assuming negligible dilution
Bristol Bay Assessment
8-42
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Reach Designation3
Reach Description11
Length (km)c
Concentration Assigned and Qualifiers'1
North Fork Koktuli River-Tributaries
NK119B/NK100CP2
NK119A
NKTSF1
NK Headwaters
NK119B/NK100CP2 confluence to
NK119CP1
NK119AtoNK119CP2
TSFlto NK119A
Headwaters or dewatering halo to
NK119B
0.43
1.3
0.6
6.8/6.8/6.6
Mixed NK119B and NK100CP2, little dilution downstream
NK119A, little dilution downstream
NK119A, assuming input near toe of dam and short reach
NK119B, which has a small contributing load in the Pebble 6.5 scenario from tailings
leachate at its upper end, so small underestimate
Upper Talarik Creek—Mainstem
UT100B
UT100C
UT100C1
UT100C2
UT100D
UT100E
UT Rock
UTIOOBto Iliamna Lake
UT100C to UT119 confluence
UT100C1 to UT100C
UT100C1 to UT100C2
UT100D to UT100C2
UT100E to UT100D (Pebble 0.25 only)
Waste rock to UT100D (not Pebble 0.25)
23
4.3
7.6
6.9
6.1
7.1/NA/NA
NA/2.1/0.15
UT100B, considerable dilution would occur in this long reach, so only the upper end
would not be overestimate
UT100C, some unquantified dilution at lower end so overestimate there
UT100C1, minimal change in concentration
UT100C2, dilution and loading balance in reach
Mean of UT100D and UT100C2 because of significant dilution in the reach
UT100E flows at background concentrations for the Pebble 0.25 scenario
UT100D, assuming input near base of rock pile and short reach
Upper Talarik Creek—Tributaries
UT Headwaters
Headwaters to UT119A
6.5
UT119A receives interbasin transfer; assumed along nearly all of length but
overestimates at upper end
Notes:
"" Reaches are designated by the gage or other feature at their heads. Designations in the form G1/G2 indicate the confluence of a stream and tributary with gages Gland G2 above the confluence.
b Upper and lower bounds of the reach.
c Lengths that differ among mine sizes are presented as Pebble 0.25/Pebble 2.0/Pebble 6.5.
d Concentrations are point estimates at upstream gages from Table 8-20, flow-weighted mixtures of concentrations at upstream gages, or means of upstream and downstream gages. Qualifiers
explain the possibility of over or underestimation.
WWTP = wastewater treatment plant; NA = not applicable
Bristol Bay Assessment
8-43
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-22. Copper concentrations and benchmarks exceeded in ambient waters in each reach and
for each mine scenario, assuming routine operations. Reaches are described in Table 8-21.
Reach Designation3
Pebble 0.25
Copper
(Mg/L)
Effects
Pebble 2.0
Copper
(Mg/L)
Effects
Pebble 6.5
Copper
(Mg/L)
Effects
South Fork Koktuli River— Ma instem
SK100B
SK100B1
SK100CP1/ SK119CP
SK100C
SK100CP2/SK124CP
SK100F
SK100G
SK Rock
SK Halo/Rock
<0.47
0.54
0.95
1.4
1.5
1.7
2.0
NA
>2.4
-
-
1C
1C
1C
NA
IC/IA
<1.9
2.5
3.9
6.5
6.1
16
61
NA
>100
1C
IC/IA
IC/IA
IC/IA/FA
IC/IA/FA
IC/IA/FA
IC/IA/FA/FS/FR
NA
IC/IA/FA/FS/FR/FK
<7.9
11
16
20
20
110
NA
>160
NA
IC/IA/FA
IC/IA/FA
IC/IA/FA
IC/IA/FA
IC/IA/FA
IC/IA/FA/FS/FR/FK
NA
IC/IA/FA/FS/FR/FK
NA
South Fork Koktuli River— Tributaries
SK Headwaters
SKTSF1
SK119A
SK124A
SKWWTP
0.42
NA
0.42
1.3
1.3
-
NA
-
-
NA
>0.44
0.44
1.3
1.3
NA
-
-
NA
NA
1.5
1.4
1.3
NA
NA
-
-
North Fork Koktuli River— Ma instem
NK100A
NK100A1
NK100B
NK119CP1/NK100C
NK100C
NKWWTP
0.45
0.66
0.52
0.48
0.44
>0.44
-
-
-
0.44
0.70
0.62
0.61
0.43
>0.43
-
-
-
0.54
0.74
0.74
0.74
0.62
>0.62
-
-
-
North Fork Koktuli River-Tributaries
NK119B/NK119CP2
NK119A
NKTSF1
NK Headwaters
(NK119B)
0.60
0.70
>0.70
>0.41
-
-
1.1
1.8
>1.8
>0.42
1C
IC/IA
IC/IA
-
1.4
1.9
>1.9
>0.63
1C
IC/IA
IC/IA
-
Upper Talarik Creek— Mainstem
UT100B
UT100C
UT100C1
UT100C2
UT100D
UT100E
UT Rock
<0.42
0.38
0.30
0.34
0.42
<0.34
NA
-
-
-
NA
<1.0
0.39
0.32
0.36
0.49
NA
0.63
-
-
-
NA
-
<3.6
0.41
0.34
0.38
0.86
NA
1.5
IC/IA
-
-
NA
-
Upper Talarik Creek— Tributaries
UT Headwaters (119A)
>0.98
-
>5.8
IC/IA
>27
IC/IA/FA
Notes:
Dashes (-) indicate that no effects are expected.
a Reaches are designated by the gage or other feature at their heads. Designations in the form G1/G2 indicate the confluence of a stream and
tributary with gages Gl and G2 above the confluence.
1C = invertebrate chronic; IA = invertebrate acute; FA = fish avoidance; FS = fish sensory; FR = fish reproduction; FK = fish kill;
NA = not applicable.
Bristol Bay Assessment
8-44
January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-23. Copper concentrations and benchmarks exceeded in ambient waters in each reach and
for each mine scenario, assuming a wastewater treatment plant failure. Reaches are described in
Table 8-21.
Reach Designation3
Pebble 0.25
Copper
(Mg/L)
Effects
Pebble 2.0
Copper
(Mg/L)
Effects
Pebble 6.5
Copper
(Mg/L)
Effects
South Fork Koktuli River— Ma instem
SK100B
SK100B1
SK100CP1/SK119CP
SK100C
SK100CP2/SK124CP
SK100F
SK100G
SK Rock
SK Halo/Rock
<3.0
4.1
6.2
11
9.8
1.7
1.9
NA
>2.4
IC/IA
IC/IA
IC/IA/FA
IC/IA/FA
IC/IA/FA
1C
1C
NA
IC/IA
<5.1
7.2
11
19
17
16
60
NA
>100
IC/IA/FA
IC/IA/FA
IC/IA/FA
IC/IA/FA
IC/IA/FA
IC/IA/FA
IC/IA/FA/FS/FR/FK
NA
IC/IA/FA/FS/FR/FK
<34
47
68
86
87
110
NA
>160
NA
IC/IA/FA/FS/FR
IC/IA/FA/FS/FR
IC/IA/FA/FS/FR/FK
IC/IA/FA/FS/FR/FK
IC/IA/FA/FS/FR/FK
IC/IA/FA/FS/FR/FK
NA
IC/IA/FA/FS/FR/FK
NA
South Fork Koktuli River— Tributaries
SK Headwaters
SKTSF1
SK119A
SK124A
SKWWTP
>0.42
NA
0.42
19
>19
-
NA
IC/IA/FA
IC/IA/FA
NA
>0.43
0.43
25
>25
NA
IC/IA/FA/ FR
IC/IA/FA/ FR
NA
NA
1.5
110
>110
NA
NA
1C
IC/IA/FA/FS/FR/FK
IC/IA/FA/FS/FR/FK
North Fork Koktuli River— Ma instem
NK100A
NK100A1
NK100B
NK119CP1/NK100C
NK100C
NKWWTP
2.3
2.9
5.8
6.2
9.0
>9.0
IC/IA
IC/IA
IC/IA/FA
IC/IA/FA
IC/IA/FA
IC/IA/FA
2.9
3.6
7.8
8.7
11
>11
IC/IA
IC/IA
IC/IA/FA
IC/IA/FA
IC/IA/FA
IC/IA/FA
17
20
43
47
57
>57
IC/IA/FA
IC/IA/FA/FS
IC/IA/FA/FS/FR
IC/IA/FA/FS/FR
IC/IA/FA/FS/FR
IC/IA/FA/FS/FR
North Fork Koktuli River-Tributaries
NK119B/NK119CP2
NK119A
NKTSF1
NK Headwaters
0.60
0.70
>0.70
>0.41
-
-
1.3
1.9
>1.9
>0.41
1C
IC/IA
IC/IA
-
1.4
1.9
>1.9
>0.63
1C
IC/IA
IC/IA
-
Notes:
Dashes (-) indicate that no effects are expected.
a Reaches are designated by the gage or other feature at their heads. Designations in the form G1/G2 indicate the confluence of a stream and
tributary with gages Gl and G2 above the confluence.
1C = invertebrate chronic; IA = invertebrate acute; FA = fish avoidance; FS = fish sensory; FR = fish reproduction; FK = fish kill;
NA = not applicable.
8.2.3.5 Dilution Zones
Analyses in Sections 8.2.3.3 and 8.2.3.4 dealt with risks from concentrations in fully-mixed locations or
reaches. Prior to achieving full mixing, the effluent plume would create a gradient from undiluted to fully
diluted, within which exposures would be higher than the fully-mixed concentrations. This should not be
an issue for a plume of properly treated wastewater, but could result in locally high exposures under
WWTP failure. The untreated wastewater concentrations of copper alone (Table 8-9) would be sufficient
Bristol Bay Assessment
8-45
January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
to cause lethality in trout and other toxic endpoint effects among survivors, in both receiving streams
under all three mine scenarios.
For waste rock and tailings leachates, effluent can enter a stream below the waste rock pile or TSF dam.
Leachate that drains to shallow aquifers would enter a stream through its cobble and gravel substrate.
Where leachates enter, benthic invertebrates and fish eggs and larvae could be exposed to a range of
concentrations, from undiluted to highly diluted leachate (the scenarios include significant dilution by
groundwater). Undiluted concentrations of metals of concern are listed in Table 8-9. NAG leachate,
which would enter Upper Talarik Creek in the Pebble 2.0 and Pebble 6.5 scenarios at 3.2 ug/L copper,
would be sufficient to cause invertebrate mortality unless it was significantly diluted by groundwater
first. The NAG and PAG leachate, which would enter the South Fork Koktuli River in the Pebble 2.0
scenario at 395 ug/L copper, would be more than six times the acute lethal concentration for trout. The
Pebble 6.5 scenario, at 735 ug/L copper, would require dilution by more than a factor of 10 to avoid
acute lethality of trout. At the more sensitive end of the spectrum, invertebrates would require dilution
of NAG and PAG leachate in the Pebble 6.5 scenario by a factor of 490 to avoid chronic toxicity.
Tailings leachates would enter tributaries of the South and North Fork Koktuli Rivers at an undiluted
copper concentration of 5.3 ug/L unless they were significantly diluted by groundwater first. This is
sufficient to kill invertebrates and to cause avoidance by trout. It would require dilution by factors of
3.5 to 5 to avoid chronic toxicity to invertebrates.
8.2.3.6 Spatial Distribution of Estimated Effects
The results of screening for total metals and copper and the analysis of severity are presented in
Tables 8-18 through 8-23 and summarized below. They are best understood by consulting the maps of
the three mine scenario footprints showing streams and gages in Figures 7-14 through 7-16.
Pebble 0.25 Scenario—Routine Operations
• South Fork Koktuli River. Copper loading from NAG waste rock in reaches SK Halo/Rock, SK100G,
and SK100 F would slightly increase the naturally high levels of copper and other metals and would
increase estimated concentrations to chronically toxic levels for invertebrates in the first 22 km.
WWTP effluent would enter the SK124 tributary. That effluent would slightly decrease copper
concentration due to treatment to achieve criteria. Concentrations would decline downstream to
SK100B due to dilution.
• North Fork Koktuli River. Input of TSF 1 leachate to the tributary above NK119A would increase
copper levels from background, but no copper criteria or benchmarks would be exceeded. Input of
water treatment effluent at NK100C would increase metal concentrations over background such
that, although copper would meet the criterion, the total metal risk quotient would rise to 1.7. At the
confluence of the TSF- and WWTP-influenced streams (NK100B), copper concentrations would be
below criteria and total metal toxicity would be marginal and decline downstream.
• Upper Talarik Creek. Copper loading would come entirely from interbasin transfer to the UT119
tributary and copper would not reach toxic levels.
Bristol Bay Assessment 846 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
Pebble 2.0 Scenario—Routine Operations
• South Fork Koktuli River. Input of NAG and PAG waste rock leachate entering below the waste rock
pile (reach SK Halo/Rock) would raise copper concentrations to levels sufficient to kill trout and
other salmonids and would be sufficient to inhibit reproduction for another 3.3 km (reach SK100G).
Levels at SKI OOF and for 18 km downstream would be sufficient to cause avoidance by trout and
severely deplete invertebrates. Levels sufficient to cause acute lethality to invertebrates would
extend another 8.9 km and chronic toxicity would extend for some distance beyond that. However,
levels would be relatively low in the SK124 tributary due to dilution by the WWTP effluent.
• North Fork Koktuli River. The pattern of input would be the same as for the Pebble 0.25 scenario,
but copper and total metals would be highly toxic to invertebrates in the NK119A tributary because
the larger TSF would release more leachate. Concentrations would decrease below the confluence of
the tributary and the mainstem below NK100C due to WWTP effluent and background water, so that
by NK100A concentrations would be close to background.
• Upper Talarik Creek. Metals from NAG waste rock leachate would enter at UT100D and raise the
naturally marginally toxic total metal levels but not copper. Concentrations would decline
downstream to non-toxic levels in the mainstem. Interbasin transfers would raise copper and total
metal concentrations to levels that would be highly toxic to invertebrates in the UT119 tributary
(reach UT Headwaters).
Pebble 6.5 Scenario—Routine Operations
• South Fork Koktuli River. SK100G would be buried by waste rock and SK119A would be buried by
tailings. SK100F would exceed the copper criterion by more than 100-fold due to NAG and PAG
waste rock leachate, achieving levels sufficient to kill juvenile and adult trout and other salmonids
for 12 km. For another 16 to 39 km, aversion and acute toxicity to invertebrates would occur. On the
SK119 and SK124 tributaries toxicity would be low despite TSF leakage and WWTP effluent.
• North Fork Koktuli River. TSF leakage would enter both the NK119A and NK119B tributaries,
resulting in copper and total metal toxicity to invertebrates for 2.4 km. Due to the WWTP, no copper
toxicity would occur at or below NK100C but total metal toxicity to invertebrates would occur.
• Upper Talarik Creek. Due to interbasin transfer from the South Fork Koktuli River, copper in the
UT119 tributary would be highly toxic to invertebrates and aversive to trout. Below the confluence
of that tributary, the mainstem would be toxic to invertebrates. NAG waste rock leachate entering
the stream from the base of the expanded waste rock pile would increase copper concentrations but
would not be expected to cause toxicity.
Wastewater Treatment Plant Failure
The WWTP failure scenario would turn the WWTP effluent from a diluent for tailings leachates to a toxic
input that would be diluted by tailings leachate. The effects of releasing untreated wastewater would, of
course, be greatest at the points of release (on the SKI 24 tributary of the South Fork Koktuli River below
the tailings dam location and at the head of the North Fork Koktuli River above gage NK100C)
Bristol Bay Assessment 847 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
(Table 8-17). Under the Pebble 6.5 scenario, the copper quotient at SK124A would increase from 1.3
(marginal toxicity) with routine operation to 100 (high toxicity) with the WWTP failure (Table 8-20),
resulting in levels sufficient to cause a fish kill extending down the South Fork Koktuli mainstem (Table
8-23). Untreated wastewater input above gage NK100C would increase the copper risk quotient from
0.58 to 54 (Table 8-20), resulting in early-life-stage toxicity to trout and other salmonids. The effects
would increase as mine size increases. The most severe effects on trout in the SK124 tributary are
estimated to be aversion (FA), early-life-stage toxicity (FR), and lethality to all life stages (FK) for the
Pebble 0.25, 2.0, and 6.5 scenarios, respectively. The most severe effects on trout below the North Fork
Koktuli outfall are estimated to be aversion for the Pebble 0.25 and Pebble 2.0 scenarios and early-life-
stage toxicity for the Pebble 6.5 scenario. That implies a shift in severity from a depleted invertebrate
community (which would reduce fish production) and fish aversion to loss offish reproduction and
death of all fish (Table 8-23). Toxicities for total metals are slightly higher (Table 8-18).
Toxic effects are functions of both duration of exposure and concentration. Because concentrations
would be so high, toxic effects on salmonids for the Pebble 2.0 and Pebble 6.5 scenarios with WWTP
failure would be severe in the South Fork Koktuli River, even if the failure was of short duration.
However, in the North Fork Koktuli or downstream of the area analyzed, the effects of WWTP failure
would depend on the duration of exposure for the Pebble 0.25 scenario. The WWTP failure described in
this chapter could last from hours to months depending on the mechanics of the failure and whether
replacement of components would be required. Alternatively, WWTP failure could be a result of an
inadequately designed water treatment system, which could result in the release of inadequately treated
water as at the Red Dog Mine, Alaska (Ott and Weber Scannell 1994, USEPA 1998, 2008). In that case,
the failure could continue for years, until a new or upgraded treatment system could be designed,
approved, and constructed. However, such failures would be much less severe than the upper bound
failure scenario evaluated here.
Spillway Release
For the spillway release scenario, we assume that the TSF pond is deep relative to the amount of
precipitation so that no appreciable dilution occurs within the TSF, and that the released water has the
same chemical characteristics as the TSF supernatant (Table 8-4). Dilution would occur downstream
due to runoff from the watersheds along the North Fork Koktuli River (Table 8-24). We assume that
precipitation is uniform over the area and that all precipitation results in runoff to the streams. We also
assume that runoff would not contribute any additional metal concentrations. The amount of dilution
would be proportional to the areas of the contributing watersheds compared to the interior area of TSF
1 and would be independent of the amount or intensity of precipitation.
Bristol Bay Assessment 848 January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
Table 8-24. Results of the spillway release scenario in terms of copper concentrations at North Fork
Koktuli stream gages downstream of TSF1, estimated effects, and the length of the associated
reaches.
Stream Gage3
NK100A
NK100A1
NK100B
NK119CP1
NK119CP2
NK119A
NKTSF1
Copper Concentration (Mg/L)
0.4
0.5
1.1
3.3
4.9
5.5
7.8
Effects
-
1C
IC/IA
IC/IA/FA
IC/IA/FA
IC/IA/FA
Reach Length (km)
4.7
8.4
20
0.79
0.43
1.5
0.64
Notes:
Dashes (-) indicate that no effects are expected.
a Stream reaches and associated gages are described in Table 8-21.
TSF = tailings storage facility; 1C = invertebrate chronic; IA = invertebrate acute; FA = fish avoidance.
Of the measured tailings supernatant constituents, only copper concentrations are estimated to exceed
water quality criteria or equivalent benchmarks (Table 8-4). The spilled supernatant immediately below
the dam (NK TSF 1) would be lethal to invertebrates and would cause avoidance by salmonids. Those
effects would continue downstream for approximately 2.6 km through the reach below NK119CP1.
Below that, invertebrate lethality would continue for another 0.79 km. The chronic criterion for the
North Fork (1.1 ug/L) would be equaled at gage NK100B, so effects would not be expected to extend far
down the 20 km reach due to dilution by tributaries. Note, however, that these estimates are based on
the assumption that the spillway release would be the only source. If a spillway release was added to
routine releases (Table 8-22), exceedance of chronic water quality criteria and chronic toxicity to
invertebrates would be likely in all reaches.
8.2.3.7 Analogous Mines
Water quality degradation has been commonly associated with mining in the United States and
elsewhere. In particular, the phenomenon known as acid mine or acid rock drainage has severely
damaged many streams due to high acidity and dissolved metals and, as the effluent is neutralized, the
formation of aluminum, iron, and manganese oxide precipitates. Pre-Tertiary waste rock at the Pebble
deposit could produce such effluents (Table 8-8). Although published studies have emphasized the
severe effects of acidic waters, it should not be assumed that neutral or alkaline leachates, such as would
be expected from the Pebble deposit Tertiary rock (Table 8-6), would have no effects.
Water quality degradation at metal mines in the United States has been reviewed and summarized in a
recent report (Earthworks 2012). Earthworks (2012) reviewed the 14 porphyry copper mines operating
in the United States and found that all but one had reported failures to collect and treat seepage that
resulted in water quality degradation. Such degradation has not been uncommon at mines due to
various factors, including inadequate pre-mining data, poor prediction of mitigation needs, inadequate
design, improper operation, and equipment failure (Earthworks 2012). Although past frequencies of
Bristol Bay Assessment
8-49
January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
water quality degradation are not predictive of future frequencies due to changes in engineering
practices, they do provide a reasonable upper bound.
Unfortunately, biological or ecological monitoring has not been routinely conducted at operating mines,
so ecological consequences are not reported by Earthworks (2012). Where biological monitoring has
occurred, acid drainage has been shown to eliminate fish and invertebrates from streams and, after
dilution, to reduce abundance, production, and diversity in stream and river ecosystems (Marchand
2002, Jennings et al. 2008). For example, acid drainage from an abandoned copper mine in Britannia
Creek, British Columbia, resulted in pH levels below 6 and spring copper concentrations greater than
1,000 ug/L (Barry et al. 2000). The abundance of chum salmon fry was lower in the creek than in
reference areas, and 100% of Chinook salmon smolts died when placed in cages in the creek. In addition,
sustained discharges have resulted in the loss of habitat through precipitation of metal hydroxides. This
case illustrates the sensitivity of salmon to acid drainage from a copper mine.
The Fraser River watershed in British Columbia has been recommended as an example of how salmon
can coexist with metal mining, and therefore suggested as a model for potential mining in the Bristol Bay
watershed. However, a long and dramatic decline in Fraser River sockeye salmon led to an official
investigation of causes, with inconclusive results (Box 8-4). In any case, it is clear that the Fraser River is
not a good analogue because, unlike potential Bristol Bay development, Fraser River mines are located
away from salmon spawning and rearing habitats. The many other activities occurring in the Fraser
River watershed confound efforts to pinpoint specific causes of salmon population decline, and the
dramatic variability in Fraser River sockeye abundance is not an example that would reassure Alaskans
accustomed to the more productive and stable Bristol Bay sockeye salmon fishery.
Bristol Bay Assessment 850 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
BOX 8-4. THE FRASER RIVER
The Fraser River watershed, which supports sockeye and other salmon and contains multiple copper mines,
could serve as an analogue for proposed mine development in the Bristol Bay watershed. Mining proponents
have argued that the Fraser River fishery demonstrates that mining and fishing can co-exist (Joling 2011).
However, the Fraser River is much less productive per unit of habitat than the Bristol Bay watershed's rivers.
In addition, the fishery has been closed in some recent years and most of its salmon runs are listed as
threatened or endangered (Cohen 2010, O'Neal and Woody 2011).
The Cohen Commission for Inquiry into the Decline of Sockeye Salmon in the Fraser River commissioned
scientific projects to investigate potential causes of decline. The report on freshwater ecological factors
considered mining as one issue (Nelitz etal. 2011). The authors concluded that metal miningwas a minor
issue for sockeye habitat relative to other development in the watershed, because there are only five active
metal mines and only one (Endako) was near sockeye rearing habitat. Other developments in the Fraser
River watershed that potentially affect habitat include logging; pulp, paper, and other wood product
manufacturing; coal, placer, and gravel mining; urbanization; hydroelectricity generation; oil and gas drilling;
agriculture; and water withdrawal. Although the authors argued that acid and metal drainage from closed
mines pose a risk to salmon, they did not analyze those exposures. They concluded that, based on
sedimentation of stream habitats, mining was a plausible contributor but not the major contributor to
declines in sockeye salmon.
Another Cohen Commission report that addressed contaminants listed mine-related contaminants, but
could not specifically quantify the effects of mines (MacDonald etal. 2011). The authors concluded that
concentrations of six metals (including copper) and phenols were sufficient to reduce survival, growth, or
reproduction of sockeye salmon in the Fraser River. The final report concluded that contaminants could be a
secondary contributor, but data were insufficient (Cohen 2012).
In light of this information, Cohen Commission reports on the Fraser River do not provide evidence that
mining and salmon co-exist. The fishery declined from 1990 to 2007 and has fluctuated widely since.
Recent fluctuations have been associated with marine conditions, but available evidence is insufficient to
conclude whether harvesting, habitat degradation, or contaminants have been significant contributors.
Neither the Cohen Commission nor the U.S. Environmental Protection Agency's contractor, ICF International,
was able to assess the effects of metal mines in the Fraser River watershed, because compliance
documents are not readily available and monitoring data are insufficient. Some raw monitoring data show
episodes of low pH and frequently elevated dissolved copper in waters at the Gibraltar and Mount Polly
Mines. Other effects have been associated with closed mines. In particular, a tailings impoundment failed
during reclamation activities at the Pinchi Lake Mine in 2004, releasing tailings and leachateto Pinchi Lake.
This accident, along with prior releases, resulted in a fish consumption advisory related to mercury
bioaccumulation.
In sum, mines in the Fraser River watershed are not located in salmon habitat (Cohen 2012, Gustafson
2012) and other development activities in the watershed obscure any effects of mines at the watershed
scale. This diverse and relatively intensive development and the spatial discontinuity between mining and
salmon habitat make the Fraser River watershed a poor analogue for potential mine development in the
Bristol Bay watershed.
8.2.3.8 Summary
The risks to salmon, rainbow trout, Arctic grayling, and Dolly Varden can be summarized in terms of the
total stream kilometers likely to experience different types of effects (Table 8-25). Based on toxicity to
rainbow trout, the endpoint salmonids are estimated to be at risk of mortality at all life stages in 0.54 km
in the Pebble 2.0 scenario and 12 km in the Pebble 6.5 scenario, assuming routine operations. The
waters would be aversive for a much greater length. It is not clear how much resident fish might
acclimate to the copper, but newly arriving salmon would not be acclimated and would lose spawning
habitat. Hence, salmon could lose 24 km (Pebble 2.0) and 34 to 57 km (Pebble 6.5) of spawning habitat
Bristol Bay Assessment 851 January 2014
-------
Chapter 8
Water Collection, Treatment, and Discharge
due to copper contamination, assuming that they are as sensitive as rainbow trout. Additional habitat
would be lost in tributaries that would not be accessed due to aversion.
Table 8-25. Length of stream in which copper concentrations would exceed levels sufficient to cause
toxic effects, assuming routine operations, wastewater treatment plant failure, and spillway release,
for each of the three mine scenarios. Intervals account for the unknown but apparently significant
dilution in reach SK100B.
Toxic Effect3
Invertebrate chronic
Invertebrate acute
Fish avoidance
Fish sensory
Fish reproduction
Fish kill
Length of Stream Potentially Affected (km)
Pebble 0.25
Routine
Operations
21
1.9
-
-
-
WWTP
Failure
78-100
65-87
27
-
-
Pebble 2.0
Routine
Operations
40-62
39
24
3.8
3.8
0.54
WWTP
Failure
80-100
79-100
64-87
27
11
3.8
Pebble 6.5
Routine
Operations
60-82
59-82
34-57
12
12
12
WWTP
Failure
78-100
76-99
74-97
70-92
61-84
31
Pebble 2.0 and 6.5
Spillway Release11
3.4-23
3.4
2.6
-
-
Notes:
8 Effects are defined in Section 8.2.3.4.
b Spillway releases are independent routine releases.
Dashes (-) indicate that no stream lengths would likely be affected.
Intervals account for the unknown but apparently significant dilution in reach SK100B.
WWTP = wastewater treatment plant.
The effects of a WWTP failure would depend on its timing and duration. If it occurred during the period
of salmon return, more than 64 km (Pebble 2.0) and 74 km (Pebble 6.5) of habitat could be lost due to
aversion alone. Mortality of all fish life stages would occur in 3.8 km (Pebble 2.0) and 31 km (Pebble
6.5). Mortality or inhibited development of early fish life stages would occur in 11 km (Pebble 2.0) and
61 to 84 km (Pebble 6.5), where the interval distances account for dilution in the SK 100B reach by
excluding and including its 23 km length.
Under routine operations, toxic effects from copper on aquatic invertebrates would occur in 21 km
(Pebble 0.25), 40 to 62 km (Pebble 2.0), and 60 to 82 km (Pebble 6.5) of streams (Table 8-25). These
effects are highly relevant to protecting salmon and other valued fishes. Immature salmon rely on
invertebrates as food, and all post-larval life stages of resident rainbow trout and Dolly Varden feed on
invertebrates. In streams, these invertebrates are primarily aquatic insects, but immature sockeye
salmon in lakes are dependent on zooplankton. Hence, protection offish requires protection of sensitive
invertebrates. These estimated effects are based on metal concentrations in fully mixed reaches. Locally,
in mixing zones below outfalls or in areas of upwelling of contaminated water, effects would be more
severe.
Because available data do not quantify fish production in the potentially affected reaches, it is not
possible to estimate the lost production of salmon, trout, Arctic grayling, or Dolly Varden. However, the
semi-quantitative surveys performed by PLP (2011) and summarized in Section 7.1 provide some
indication of the relative amounts offish potentially affected. The focal species are those that rear for
Bristol Bay Assessment
8-52
January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
extended periods in the receiving streams: Chinook salmon, coho salmon, Arctic grayling, and Dolly
Varden.
The South Fork Koktuli River, which would be the most severely affected stream, has the lowest
reported density of focal species that rear for extended periods in the receiving streams and for which
data are available (roughly 14,000 fish/km for Chinook and coho salmon, Arctic grayling and Dolly
Varden) (Table 7-3), as well as chum and sockeye salmon. Because 28 to 50 km of the South Fork Koktuli
River would have copper levels sufficient to directly affect fish in the Pebble 6.5 scenario, more than a
half million individuals of the focal species would be exposed to copper levels sufficient to cause
aversion, sensory inhibition, inhibited development, or death. In the Pebble 2.0 scenario, copper levels
in 22 km of the South Fork Koktuli River would have direct effects on more than 300,000 individuals of
the focal species. Direct effects on fish would not be expected in the Pebble 0.25 scenario.
The North Fork Koktuli River has a focal species density of roughly 20,000 fish/km (Table 7-3), plus
unenumerated rainbow trout and chum and sockeye salmon. Since 2.4 km of the North Fork Koktuli
River would have copper levels sufficient to be toxic to invertebrates in the Pebble 2.0 and Pebble 6.5
scenarios, more than 47,000 individuals of the focal species would experience reduced food resources.
Upper Talarik Creek has the highest density of the focal species at 45,000 fish/km (Table 7-3) plus
unenumerated rainbow trout and sockeye and chum salmon. The 6.5 km of the tributary receiving South
Fork Koktuli River interbasin transfers would be expected to have avoidance effects on fish in the Pebble
6.5 scenario and reduced invertebrates in the Pebble 2.0 scenario. In the mainstem below the confluence
of that tributary, less than 23 km of stream would experience effects on invertebrates.
For the WWTP failure in the Pebble 2.0 and Pebble 6.5 scenarios, as under routine operations, 40 to
50 km of the South Fork Koktuli River would have copper levels sufficient to directly affect more than a
half million of the focal fish. However, effects would be more severe than under routine operations and
include acute lethality to all life stages in most reaches. For the Pebble 0.25 scenario, 20 km would
experience aversive effects on fish and, in 40 to 62 km, toxicity to invertebrates would result in reduced
food resources for more than a half million of the focal fishes.
Due to the uncertainties in the fish density data and the compounding uncertainties in exposure and
toxicity, these effects estimates are rough. However, it appears that the number of fish experiencing
death or an equivalent effect, such as loss of habitat, would be between 10,000 and 1 million for the
Pebble 2.0 and Pebble 6.5 scenarios.
For the WWTP failure in the Pebble 0.25, 2.0, and 6.5 scenarios, 27, 64 to 87, and 74 to 97 km of streams,
respectively, would have copper concentrations sufficient to directly affect fish (Table 8-25). Toxicity
would result in reduced survival or inhibited development for early salmonid life stages in 61 to 84 km
in the Pebble 6.5 scenario, potentially affecting more than a half million fish, depending on the season.
Sensory inhibition or aversion would affect 600,000 to 1.4 million individuals of the focal fish species in
the three mine scenarios until the failure was corrected.
Bristol Bay Assessment 853 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
For the spillway release in the Pebble 2.0 and 6.5 scenarios, copper concentrations would be sufficient to
cause avoidance by fish in 2.6 km in the North Fork Koktuli River (Table 8-25). Effects on invertebrate
survival would be expected in more than 3.4 km, depending on dilution by tributaries in the lowest
reach (Table 8-25).
8.2.4 Additional Mitigation of Leachates
The high metal concentrations in the South Fork Koktuli River due to PAG waste rock leachate suggest
that mitigation measures beyond those described in the scenarios or the preliminary Northern Dynasty
mining case (Ghaffari et al. 2011) should be considered. Although that design may be sufficient for a
typical porphyry copper mine (e.g., equivalent to the Pebble 0.25 scenario), it likely is sufficient for not
the massive Pebble 2.0 and 6.5 mine sizes. To avoid exceeding copper criteria, a leachate barrier or
collection system for the Pebble 6.5 scenario would require more than 99% effectiveness. Wells,
trenches, or walls are not likely to achieve that. Lining the PAG waste rock piles might be effective, but
liners have some leakage due to imperfect installation, punctures, and deterioration. An alternative
mitigation measure would be to ensure that all PAG waste rock is stored within the drawdown zone for
the mine pit. In that way, most acidic and high-metal leachate would be collected and treated before
discharge. If PAG waste rock was processed before or at closure, the risk of an acidic pit lake would be
minimized (Section 8.1.4). Moving all PAG waste rock near the pit would mean an increase in NAG waste
rock leachate leakage to streams as NAG waste rock is moved out of the drawdown zone. If all of the
leakage of waste rock leachate for the Pebble 6.5 mine were NAG and if mining did not affect the
background copper concentration, the copper concentration would be approximately 1.5 ug/L. That
would be a great improvement, but would still equal the chronic criterion for the stream and affect
sensitive invertebrates. Hence, it would also be necessary to improve the 50% efficiency of leachate
capture assumed here. The magnitude and extent of these predicted effects suggest the need for
additional mitigation measures to reduce the input of copper and other metals, beyond the conventional
practices assumed in the scenarios. Simply improving capture well efficiency, making the cutoff walls
more extensive, or adding a trench is unlikely to achieve water quality criteria under those scenarios.
Additional measures might include lining the waste rock piles, reconfiguring the piles, or processing
more of the waste rock as it is produced.
8.2.5 Uncertainties
Although it is highly likely that mine operations would adversely affect water quality at the mine site,
several factors make it difficult to predict the level of effects and consequent risks to fish.
One component of this uncertainty is associated with the likelihood of water collection and treatment
failure. Water collection and treatment failures have been documented at 13 of 14 porphyry copper
mines in the United States (Earthworks 2012). These 13 cases represent instances in which engineering
uncertainties led to prediction failures, despite the fact that mine permits included mitigation measures
intended to prevent such occurrences. These results indicate that failures are not uncommon at modern
Bristol Bay Assessment 854 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
U.S. copper mines; however, they cannot be used to quantitatively predict the likelihood of water
collection and treatment failures in this or future assessments.
Even in the absence of failures, predicting the effects of mining on water quality is difficult and results
are uncertain. Further, the effects of water quality changes on aquatic communities are uncertain. The
following factors contribute to these uncertainties.
• The range of potential failures is wide and the probability of occurrence for any of them cannot be
estimated from available data. Therefore, we can only state that, based on the record of the mining
industry, treatment failures of some sort are likely to occur.
• The waste rock leachate concentrations used in the assessment are from humidity cell tests. Because
these tests involve repeated flushing of rock under oxic conditions, they may reasonably represent
waste rock piles or pit walls leached intermittently by precipitation and snowmelt. However,
laboratory tests of relatively small samples are imperfect models of large rock piles in the field. This
uncertainty may be minimally estimated by comparing the humidity cell tests with barrel tests
conducted in the field with more realistic rock sizes and test conditions (PLP 2011: Section 11.7.1).
These tests give qualitatively similar results, but the initial flush of high leachate concentrations in
the barrel tests seems to be persisting past the date of data compilation (PLP 2011). Therefore, the
concentrations reported for humidity cells are used because they are likely to be closer to long-term
values, and the magnitude of uncertainty cannot be estimated.
• The tailings leachate concentrations are also from laboratory tests. Such tests of relatively small
samples are imperfect models of the processes in tailings slurries, TSF surface water, near surface-
deposited tailings, deeply buried tailings, and leakage into groundwater in the field. It is not clear
whether these tests tend to over or underestimate leachates in the field or how large the
discrepancy might be.
• The tailings test data do not include pyritic tailings, which are strongly acid-generating. This would
tend to underestimate the metal content of tailings leachate, but the effects on leachates from a TSF
are likely to be small due to the relatively small proportion of pyritic tailings.
• The available leach testing appears to be preliminary and should be augmented with additional and
more realistic testing if mine planning proceeds.
• The surface-water and groundwater hydrology of the potential mine site is complex and the
hydrological models used to estimate exposures are inevitably simplifications. This is one of the
greatest sources of uncertainty for the water quality risks. More information is needed concerning
the movement of water from precipitation to groundwater and surface water, including seasonally
and storm and melt events.
• The water quality models assume that mining would not affect background water quality. That is
unlikely, but any changes could not be estimated. This assumption is expected to result in
overestimation of copper levels, particularly in the South Fork Koktuli River. However, as mining
Bristol Bay Assessment 855 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
reduces background levels, it would increase levels from leachate input even more. It could change
the expected effects at the margins of toxicity but would not significantly affect the conclusions.
• The use of average receiving water flows neglects the potential consequences of low dilution during
low-flow periods. This would be difficult to model, because low streamflows would be associated
with low leachate formation and low groundwater levels. This consideration suggests that the low
dilution and low leaching rates might balance to some extent, but the degree of balance is unknown.
• Chemical criteria and other single chemical benchmarks do not address interactions or combined
effects of individual constituents. The additivity model used here is a reasonable default, but the lack
of test data for the actual mixture adds uncertainty. This is a concern of some reviewers but is
judged to be a relatively small contributor to uncertainty. Strong interactions tend to occur when all
constituents are at or near toxic levels. Only cadmium and zinc reach toxic levels and only in the
WWTP failure scenario. Given the overwhelming dominance of copper toxicity, this uncertainty
appears to be relatively minor.
• Studies of streams receiving mine effluents and laboratory studies suggest that the abundance of
important insect taxa could be reduced even if criteria are met. The implications of this uncertainty
are discussed at the end of Section 8.2.2.1.
• Criteria for chemicals other than copper either do not address site water chemistry or address it in a
simple way (e.g., via hardness normalization). Hence, they may be inaccurate estimates of threshold
concentrations for toxic effects in these highly pure waters. The example of copper suggests that the
criteria and screening benchmarks could be too high by a factor of 2 (see discussion below).
• Some leachate and process water constituents have water quality criteria, standards or benchmarks
for aquatic life that are based on old or sparse literature. Additional data are likely to reveal more
sensitive species or responses. This would result in lower benchmarks and criteria and higher risks
for the poorly studied chemicals. However, this is unlikely to affect conclusions, because the
relatively well-studied metal copper dominates the toxicity.
• If the State of Alaska uses its standards in effluent permitting rather than the national criteria, and if
the toxicity of chemicals with no state standards is not considered in the permit, toxicity of the
effluents would be significantly higher than estimated in this assessment. This could result in an
underestimate of effects by more than a factor of 2, primarily due to using the hardness-adjusted
copper standard.
• The concentrations of xanthate and other ore-processing chemicals in ambient waters are roughly
estimated to be below toxic levels, but studies in the laboratory or at mine sites are insufficient to
determine whether that would actually be the case. If xanthate does not degrade rapidly in the
tailings, the estimate that it would not leach into streams at toxic concentrations could be incorrect.
• If the tested rock and tailings samples are not representative, other wastewater constituents may be
of concern. Some waste rocks or tailings may have high levels of elements other than those
identified in the screening analysis for mean concentrations. For example, selenium concentrations
Bristol Bay Assessment 856 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
are not high on average but are well above criteria in some individual leachate samples. This
uncertainty might be estimated from statistical analyses of sampling results and modeling of waste
rock piles with variance in concentrations among locations, but that is beyond the scope of this
assessment.
• The separation of PAG and NAG into separate waste rock piles will inevitably be imperfect. Ghaffari
et al. (2011) estimate that 5% of rock in the NAG piles would be PAG. The humidity cell tests were
not reported to include any PAG, so humidity cell tests on NAG mixed with 5% PAG would be
expected to have higher leachate concentrations. This causes underestimation of risks.
• Although Alaska has a standard for TDS from any source, the toxicity of mixtures of major ions
depends on the constituents. Most studies of TDS are based on sodium chloride, which is less toxic
than mining leachates that have been studied. Hence, the degree of protection provided by the state
standard is uncertain. The toxicity of different salt mixtures may vary by at least a factor of 3.
However, estimated TDS levels are not high enough for this to be a major uncertainty.
One method for quantifying uncertainty is provided by comparing the benchmarks for copper toxicity
that might be used as thresholds for minimum risk (Section 8.2.2). The national ambient water quality
criteria for copper are based on the BLM, which better accounts for the influence of water quality on
bioavailability than the hardness-derived state standard. These two benchmarks differ by a factor of 1.7
for acute values and 2.1 for chronic values. Four other metals of concern (cadmium, lead, nickel, and
zinc) have hardness-dependent standards but no BLM-based criteria. If they also are too high by a factor
of approximately 2, then those metals and total metal toxicity are more of a concern than suggested by
the screening assessment. The same applies to relevant conventional thresholds for acute and chronic
toxicity in salmonids (the LCso and chronic value for rainbow trout), which are BLM-corrected for
copper (Table 8-13) but not for other metals. However, the threshold for avoidance of copper exposures
(IC2o) is 12 to 28 times lower than the LCso and 4.2 to 5.4 times lower than the chronic value (Table 8-
14). Hence, the concentration at which a stream would no longer be suitable for trout is considerably
underestimated by conventional endpoints. The effects thresholds for less well-studied metals are likely
to be equivalently underestimated. Even copper toxicity is likely to be underestimated, due to the
absence of tests for sensitive insect species—much less tests of the most sensitive responses of those
species (Section 8.2.2.1).
8.3 Temperature
Changes in water temperature associated with mine development activities are a concern given the
importance of suitable water temperatures for Pacific salmon. This section begins with a description of
current thermal regimes in the mine scenario watersheds and potential alterations due to WWTP
discharges under routine operations (Section 8.3.1). It then describes exposure-response relationships
for temperature (Section 8.3.2). It ends with a characterization of potential risks associated with the
thermal regime of water treatment effluents (Section 8.3.3) and a discussion of uncertainties
(Section 8.3.4).
Bristol Bay Assessment 857 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
8.3.1 Exposure
8.3.1.1 Thermal Regimes in the Mine Scenario Watersheds
Water temperature data collected by PLP (2011: Appendix 15.IE, Attachment 1) indicate significant
spatial variability in thermal regimes. Average monthly stream temperatures in the Pebble deposit area
in July or August can range from 6°C to 16°C. Extensive glacially reworked deposits with high hydraulic
conductivity allow for extensive connectivity between groundwater and surface waters in the region
(Power et al. 1999). This groundwater-surface water connectivity has a strong influence on the
hydrologic and thermal regimes of streams in the Nushagak and Kvichak River watersheds, providing a
moderating influence against both summer heat and winter cold extremes in stream reaches where this
influence is sufficiently strong. The range of spatial and temporal variability in temperatures provided
by PLP (2011) is consistent with streams influenced by a variety of thermal modifiers, including
upstream lakes, groundwater, or tributary contributions (Mellina et al. 2002, Armstrong et al. 2010).
Longitudinal profiles of temperature indicate that summertime stream temperatures in the Pebble
deposit area do not uniformly increase with decreasing elevation, often due to substantial inputs of
cooler water from tributaries or groundwater inputs (PLP 2011). An example of combined tributary and
groundwater inflow contributing to significant cooling in summer mainstem temperatures is the South
Fork Koktuli River downstream of gage SK100C (Figure 7-14). This is the section of the South Fork
Koktuli River fed by the tributary gaged by SK119A, on which a WWTP outfall would be located and to
which a portion of the WWTP flows would be directed. As reported by PLP (2011: Appendix 15.IE),
combined groundwater and tributary contributions between gages SK100C and SK100B1, including
contributions from the tributary gaged by SK119A, contributed to a cooling of 5.4°C, with a gain in flow
of 1.36 m3/s on August 24, 2007. Other examples of spatial variability in summer temperatures are
detailed by PLP (2011: Appendix 15.IE, Attachment 1).
Winter water temperatures are also spatially variable, as indicated by instream temperature monitoring
data (PLP 2011). The same reach of the South Fork Koktuli River that was cooled in August by
groundwater and tributary inflows experienced warming in October. Contributions of relatively warmer
groundwater were observed to maintain ice-free conditions in some areas, as revealed by patchiness in
ice cover seen in aerial surveys (PLP 2011, Woody and Higman 2011).
8.3.1.2 Thermal Regime Alterations
Mine development and operation would result in alteration of surface-water and groundwater flows and
water collection, treatment, and discharge, all of which would affect water temperatures. Some streams
would experience increased streamflows, whereas others would experience significant reductions
(Table 7-19). Increased streamflows due to additions of effluent from the WWTP would alter
temperatures significantly, depending on effluent temperature and quantity. Changes in the source of
water supplying streams would also influence thermal responses. For example, reductions in the
proportion of thermally-moderated groundwater inputs would result in surface-water temperatures
that would be warmer in summer and colder in winter. Conversely, active thermal management (i.e.,
heating or cooling of effluent) and timed releases from the WWTP could be used to attempt to
Bristol Bay Assessment 858 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
compensate for mine-related thermal modifications. However, the plan for a Pebble mine outlined by
Ghaffari et al. (2011) does not include temperature control by the planned WWTP. The mine scenarios
include temperature control to meet state standards, but not to match natural water temperature
regimes.
Treated water would be released to tributaries of the South and North Fork Koktuli Rivers and would
influence streamflows and water temperatures in downstream reaches. Thermal effects of WWTP
effluent would be greatest in the receiving tributaries. Effects would moderate with distance from the
WWTP outfall, due to mixing with surface-water and groundwater inputs and heat exchange. Due to the
substantial increases in discharge over baseline levels associated with the WWTP (up to 114% increases
in monthly mean flow depending on mine scenario and location; Table 7-19), the thermal loads
attributable to WWTP discharges would potentially influence temperatures downstream in the South
and North Fork Koktuli Rivers. For example, WWTP discharge is expected to comprise 11 to 38% of
mean annual flow in the North Fork Koktuli River at gage NK100C (calculated from Tables 8-1 through
8-3). Sensitivities of downstream reaches to WWTP outfall temperatures were not evaluated in this
assessment due to uncertainties in the timing and temperature of WWTP discharges and heat exchange
processes in downstream reaches. Managing treated water temperatures to maintain baseline thermal
regimes would be most protective offish populations adapted to local thermal regimes, but would
require temperature and hydrologic modeling informed by baseline monitoring and the ability to
control temperatures and quantities of discharged flows to meet temperature targets. Baseline data
collected by PLP contractors (PLP 2011) for the purposes of developing and applying surface-water
temperature models would be useful for managing flows and temperatures to minimize impacts on
aquatic life.
8.3.2 Exposure-Response
Water temperature controls the metabolism and behavior of salmon, and, if temperatures are stressful,
fish can be more vulnerable to disease, competition, predation, or death (McCullough et al. 2009).
Recognizing the importance of water temperature to healthy salmon populations, the State of Alaska
requires that maximum water temperatures not exceed 20°C at any time, with specific maximum
temperatures for migration routes and rearing areas (15°C) and spawning areas and egg and fry
incubation (13°C). For all other waters, the weekly average temperature may not exceed site-specific
requirements needed to preserve normal species diversity or to prevent the appearance of nuisance
organisms (ADEC 2012).
This standard is designed to protect against increases in summer temperature, a serious concern for
salmon populations, particularly in light of projected climate change effects on streamflow and
temperatures (Section 3.8) (Bryant 2009). Elevated summer temperatures are a management concern
due to potential adverse effects including increased risk of direct mortality, disease, elevated metabolic
costs, and altered community interactions. Sockeye salmon are particularly sensitive to high
temperatures during spawning, being limited to temperatures between 2°C and 7°C (Weber Scannell
1991). Summer, however, is not the only period during which salmon are sensitive to temperatures
Bristol Bay Assessment 859 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
(Poole et al. 2004). Salmon and other native fishes in the mine scenario watersheds rely on suitable
temperature regimes to successfully complete their life cycles (Quinn 2005). The period of salmon egg
incubation in gravels can be particularly sensitive to temperature changes, and changes of just a few
degrees Celsius in winter mean temperature can change emergence timing of young salmon by months
(Figure 3-19) (Brannon 1987, Beacham and Murray 1990, Quinn 2005). For locally adapted populations,
timing of key life-history events (i.e., spawning, incubation, and out-migration) can be closely tied to the
timing of other ecosystem functions that provide critical resources for salmon (Brannon 1987, Quinn
and Adams 1996). Thus, changes to thermal and hydrologic regimes that disrupt life-history timing cues
can result in mismatches between fish and their environments or food resources, adversely affecting
survival (Jensen and Johnsen 1999, Angilletta et al. 2008).
8.3.3 Risk Characterization
Stream temperatures in the mine scenario watersheds could be substantially altered due to changes in
streamflow, sources of streamflow (e.g., relative importance of groundwater versus WWTP
contributions), or other changes to the heat balance of WWTP discharges. We expect treated water
returned to streams would have different thermal characteristics than water derived from groundwater
sources (the dominant water source prior to mining). The extent and duration of temperature effects
would depend not only on source water temperatures, but also on the quantity and timing of water
contributed from various additional sources, such as tributaries and groundwater inputs. Simple mixing
models can be used to estimate stream temperatures below the confluence of multiple sources with
known temperatures and discharges. However, we do not use such models here, because we cannot
account for all sources of heat transfer. In the absence of models, we have relied on available literature
to identify the most likely risks to fish associated with deviations from current thermal regimes in the
Pebble deposit area.
Interception of groundwater that is collected then released as a point-source through a WWTP would
alter the ways in which groundwater feeds stream channels through dispersed and complex pathways.
Groundwater-surface water interactions in streams can create thermal heterogeneity, enhancing the
diversity of habitats available to fish (Power et al. 1999). Migration, spawning, and incubation timing are
closely tied to seasonal water temperatures. Diversity of thermal habitats can allow a diversity of
spawning migration timing to persist (Hodgson and Quinn 2002). For the Bristol Bay region, this
asynchrony in spawning timing helps buffer Bristol Bay salmon populations from climatic events or
other environmental changes that may adversely affect a particular run (Schindler et al. 2010). An
additional benefit of staggered spawner return timing is the extended availability of spawning sockeye
salmon to mobile consumers like brown bear (Schindler et al. 2010). Depending on the degree to which
adaptation and compensatory strategies may mitigate thermal effects on life-history development and
spawning timing, deviations from the thermal regime to which local populations of salmon may be
adapted could have serious population-level consequences (Angilletta et al. 2008).
The volume of water that would require treatment ranges from roughly 10 to 51 million m3/yr across
the three mine scenarios (Tables 8-1 through 8-3). To avoid or minimize risks associated with altered
Bristol Bay Assessment 860 January 2014
-------
Chapter 8 Water Collection, Treatment, and Discharge
thermal regimes in downstream effluent-receiving areas, capacity for thermal control of effluent would
be required to maintain natural thermal regimes or temperatures required by regulatory agencies.
Water temperature modeling is being used by PLP to assess thermal characteristics of streams in the
Pebble deposit area (PLP 2011: Chapter 15, Appendix 15.IE) and could provide additional guidance for
establishing a temperature management plan for the WWTP.
8.3.4 Uncertainties
The temperature of waters discharged from the mine, whether directly from the WWTP or indirectly
through changes in groundwater or surface-water runoff, would be influenced by a number of factors
controlling heat exchange that cannot be known with confidence at this point. Likewise, the influence of
these discharges on stream temperatures downstream of the mine site is unknown. Because exchange
with groundwater is so important to surface-water properties in the mine area, simple models that
assume primarily surface-water heat exchange would be incomplete and inaccurate.
Projecting changes to temperature due to changes in groundwater-surface water interactions in the
mine area was not attempted for this assessment. Local geology and stream hydrographs are indicative
of systems that are largely driven by groundwater. Disruptions or changes to groundwater flowpaths
and mechanisms of thermal exchange in the mine area could have significant adverse effects on winter
habitat suitability for fish, particularly if groundwater-dominated stream reaches are converted to
stream reaches dominated by WWTP effluent with a novel thermal regime. Given the high likelihood of
complex groundwater-surface water connectivity in the mine area, predicting and regulating
temperatures to maintain key ecosystem functions associated with groundwater-surface water
exchange would be particularly challenging.
Maintenance of mine discharges in terms of water quality, quantity, and timing to avoid adverse impacts
would require long-term commitments for monitoring and facility maintenance. As with other long-term
maintenance and monitoring programs, the financial and technological requirements could be large, and
the cumulative risks (and likely instantaneous consequences) of facility accidents, failures, and human
error would increase with time. Additionally, climate change and the predicted increases in water
surplus for the region (Section 3.8) will result in potential changes in streamflow magnitude and
seasonally, requiring adaptation to potentially new water management regimes for the water
processing facilities. We know of no precedent for the long-term management of water temperature on
this scale at a mine.
Finally, whereas the bioenergetics of the endpoint fish species are relatively well known, how these
species would respond to changes to thermal regimes is poorly understood—particularly with regard to
sublethal effects, behavior, adaptation, effects of fitness on the population, and other effects ranging
from the molecular to the ecosystem level (McCullough et al. 2009). The existing information consists
largely of field studies of salmonid distributions with respect to temperatures, supplemented by
laboratory studies of development, growth, and survival at controlled temperatures. Monitoring studies
to help confirm relationships between temperature alterations of various magnitudes and durations and
population consequences are desirable.
Bristol Bay Assessment 861 January 2014
-------
In this chapter, we describe risks to stream habitats and salmonid populations from potential failures of
tailings storage facility (TSF) dams. Specifically, we consider tailings dam failures at TSF 1 and potential
physical and toxicological effects on fish and fish habitat (Figure 9-1). Similar types of effects would
occur following tailings dam failures at TSF 2 or TSF 3, or at TSF locations in other parts of the Bristol
Bay watersheds, although the specifics of a failure at these locations would differ.
A breach of a TSF 1 dam would result in a flood wave and subsequent tailings deposition that would
greatly alter the downstream channel and floodplain (Figure 9-1). The initial flood wave for the tailings
dam failure scenarios modeled here would far exceed the typical flood event currently experienced in
these watersheds. The flood itself would have the capacity to scour the channel and floodplain and alter
the landscape, and the amount of tailings that could discharge from the TSF could bury the existing
channel and floodplain system with meters of fine-grained tailings material. The existing channel and
floodplain would be eliminated and a new channel form would develop in the resulting topography.
Given the size of these new fine-grain deposits, sediment would be highly mobile under typical
streamflow events and channel form would remain unstable. Sediment deposited on floodplains and
remaining behind the breached dam would create a concentrated source of highly mobile material that
does not currently exist in the mine scenario watersheds. Although a sediment transport study would be
required to quantify the temporal and spatial extent of effects, it is likely that the sediment regime of the
affected stream and downstream waters would be greatly altered, and that the existing and well-defined
gravel-bed stream would be transformed to an unstable, silt- and sand-dominated channel.
Remediation is possible following a tailings spill, but the occurrence and effectiveness of these measures
would be uncertain. A tailings spill would be flowing into a roadless area with streams and rivers that
are too small to float a dredge, so the proper course of remediation is not obvious. The remediation
process could be delayed by planning, litigation, and negotiation, particularly concerning the proper
removal and disposal of excavated tailings. If the operator was no longer present at the site or was no
Bristol Bay Assessment
9-1
January 2014
-------
Chapter 9 Tailings Dam Failure
longer in existence, the response would, at best, be delayed further. Once started, the building of a road
and support facilities and the excavation, hauling, and disposal of tailings could take years, particularly
given the long winter season. Therefore, the extent to which tailings exposure would be diminished by
remediation cannot be estimated. Given this uncertainty, the assessment assumes that significant
amounts of tailings would remain in the receiving watersheds for some time, and remediation would
never be complete.
9.1 Tailings Dam Failures
9.1.1 Causes
A tailings dam failure occurs when a tailings dam loses its structural integrity and releases tailings
material from the impoundment. Released tailings flow under the force of gravity as a fast-moving flood
that contains a dense mixture of solids and liquids, often with catastrophic results. This flood can
contain several million cubic meters of material traveling at speeds in excess of 60 km/hour. At dam
heights ranging from 5 to 50 m—substantially less than the 92-m and 209-m tailings dam failures
considered here—the flood wave can travel many kilometers over land and more than 100 km along
waterways (Rico et al. 2008). There are many international examples of such failures (Box 9-1),
involving dams that were significantly smaller than those considered in our mine scenarios.
BOX 9-1. EXAMPLES OF HISTORICAL TAILINGS DAM FAILURES
The examples below illustrate the characteristics and potential consequences of a tailings dam failure.
Details of the design, construction, or operation of any tailings dams constructed for mines in the Bristol Bay
watershed would not be the same as these mine tailings dams. However, these examples demonstrate that
tailings dam failures can occur and illustrate how these failures may affect downstream areas. In addition,
the dams in these failure examples were significantly smaller than the dams in our mine scenarios.
Stava, Italy, 1985. Two tailings impoundments were built, one upslope from the other, in the mountains of
northern Italy. The upslope dam had a height of 29 m; the downslope dam had a height of 26 m. A stability
failure of the upper dam released tailings, which then caused the lower dam to fail. The 190,000 m3 of
tailings, traveling at up to 60 km/hour, reached the village of Tesero (4 km downslope from the point of
release) in approximately 5 minutes. The failure killed 269 people (ICOLD 2001).
Aznalcollar Tailings Dam, Los Frailes Mine, Seville, Spain, 1998. A foundation failure resulted in a 45-m-
long breach in the 27-m-high, 600-m-long tailings dam, releasing up to 6.8 million m3 of acidic tailings that
traveled 40 km and covered 2.6 million ha of farmland (ICOLD 2001).
Aurul S.A. Mine, Baia Mare, Romania, 2000. A 5-km-long, 7-m-high embankment on flat land enclosed a
tailings impoundment containing a slurry with high cyanide and heavy metal concentrations. Heavy rains and
a sudden thaw caused overtopping of the embankment, cutting a 20-to 25-m breach and releasing
100,000 m3 of contaminated water into the Somes and Tisza Rivers. Flow continued into the Danube River
and eventually reached the Black Sea. The contamination caused an extensive fish kill and the destruction
of aquatic species over 1,900 km of the river system (ICOLD 2001).
Tennessee Valley Authority Kingston Fossil Plant, Roane County, Tennessee, USA, 2008. After receiving
nearly 20 cm of rain in less than 4 weeks, an engineered 18-m-high earthen embankment of a 34-ha
storage impoundment failed, producing a 14-m-high surge wave and releasing 4.1 million m3 of coal fly ash
slurry. The release covered over 121 ha with slurry containing arsenic, cobalt, iron, and thallium. Over 2.7
million m3 of coal ash and sediment were dredged from the Emory River to prevent further downstream
contamination (AECOM 2009).
Bristol Bay Assessment a 2 January 2014
-------
Chapter 9
Tailings Dam Failure
Figure 9-1. Conceptual model illustrating potential pathways linking tailings dam failure and effects on fish endpoints. Not all potential pathways are analyzed in this assessment.
tailingsstorage |
facilities J
T inhibition of
fish passage
/jv physiological
stress
V
macroinvertebrate
prey
V
A other water
quality parameters
V
"f dissolved metals
A
^adsorbed or precipitated metals
V
•^ macroinvertebrate
toxicity
V
xi- rearing habitat
(quality or quantity)
-spawning habitat
(quality or quantity)
xb overwintering habitat
(quality or quantity)
A metal spedation
& bioavailability
v
v
s- bioaccumulation&
biornagnification
1s tissue
concentration
•t chronic
toxicity
f- acute
toxicity
LEGEND
Within a shape, ^ indicates an increase in the parameter,
4-indicates a decrease in a parameter, and A indicates a
change in the parameter.
Arrows leading from one shape to another indicate a
hypothesized cause-effect relationship.
Shapes bracketed under another shape are specific
components of the more general shape under which they
appear.
V
xL- salmon
(abundance, productivity or diversity)
V
xb other fishes
(abundance, productivity or diversity)
>j, marine-derived
nutrients
xl/ ecosystem
productivity
climate
change
V
A
volcanic activity
seismic activity
hydrologic event
landslide/avalanche
engineering failure
human error
fire
stage of mining
operations
seasonal timing of
unplanned event
Bristol Bay Assessment
9-3
January 2014
-------
This page intentionally left blank
-------
Chapter 9 Tailings Dam Failure
Causes of tailings dam failure are similar to those for earthfill and rockfill water retention dams, and
include the following.
• Overtopping. Overtopping occurs when sufficient freeboard (the distance between the top of a dam
and the impounded water level) is not maintained and the water level behind a dam rises due to
heavy rainfall, rapid snowmelt, flooding, or operator error.
• Slope instability. A slope instability failure occurs when shear stresses in a dam exceed the shear
resistance of the dam material, most frequently resulting in a rotational or sliding failure of a
portion of the downstream slope, leading to overtopping or breaching of the dam.
• Earthquake. Earthquake-induced shaking (Box 9-2; Section 3.6) causes additional shear forces on a
dam that can lead to a slope instability failure.
• Foundation failure. Weak soil or rock layers and high pore pressures below the base of a dam can
lead to shear failures in the foundation, causing the entire dam to slide forward or rotate out of
position.
• Seepage. Seepage through an earthfill embankment increases interstitial pore pressures and
reduces the intergranular effective stresses and shear resistance, potentially leading to a slope
instability failure. Seepage can also cause internal erosion and piping within a dam leading to a
hydraulic failure.
• Structural failure. Tailings dams often contain structural components such as drainage systems or
spillways that, if they fail to function properly, can cause overtopping or slope instability failure.
• Erosion. Erosion, especially along the toe of a dam, can reduce slope stability to the point of failure.
Erosion near the crest can reduce freeboard and increase the chance of overtopping.
• Subsidence. If a tailings dam is built on compressible soils or overlies cavities such as underground
mining works (Box 4-4), subsidence can cause displacement or cracking of the dam. Cracking can
lead to a direct hydraulic breach or to slope instability. Settlement can reduce freeboard and
increase the chance of overtopping.
Bristol Bay Assessment Q ,- January 2014
-------
Chapter 9
Tailings Dam Failure
BOX 9-2. SELECTING EARTHQUAKE CHARACTERISTICS FOR DESIGN CRITERIA
Design criteria for dams specify that an evaluation be conducted to determine the effect of seismicity on
stability and performance of the dam. This seismic evaluation must establish the operating basis
earthquake (QBE) and maximum design earthquake (MDE). One important characteristic of determining
earthquake sizes is the return period (recurrence period) over which the event is likely to occur. If long return
periods are used in the analysis of earthquake size, the likelihood of a larger earthquake increases and the
resulting design basis earthquake will have a greater margin of safety.
The QBE represents the characteristic earthquake with a reasonable probability of occurring during the
functional lifetime of a project. Critical structures should be designed to withstand the effects of the QBE
and remain functional, with little, easily repairable damage. The QBE can be defined using a probabilistic
approach based on the likelihood that an earthquake of a certain magnitude and ground motion will be
exceeded during a particular period. Fora dam in Alaska with a Class II hazard potential, the return period
that must be considered for the QBE is 70 to 200 years—that is, the OBE represents the largest earthquake
likely to occur in 70 to 200 years.
The MDE represents the most severe earthquake considered at the site for which acceptable consequences
of damage would result. All critical structures such as tailings dams must be designed to resist the effects of
the MDE, so underestimating the MDE could increase the risk of a catastrophic tailings dam failure. The
MDE can be determined based on historical earthquake patterns or through a rigorous probabilistic analysis.
For a Class II dam, the return period considered appropriate for the MDE is 1,000 to 2,500 years.
A third category of earthquake design level is the maximum credible earthquake (MCE). The term is not
defined in the Alaska dam safety regulations, but supporting guidance defines it as the greatest earthquake
that reasonably could be generated by a specific seismic source, based on seismologic and geologic
evidence and interpretations. Design engineers sometimes use the MCE to represent a floating earthquake
(i.e., an earthquake not associated with a known fault) located directly under a critical structure.
The return periods stated in Alaska dam safety guidance are inconsistent with the expected lifetime of a
tailings dam for a porphyry copper mine developed in the Bristol Bay watershed, and represent a minimal
margin of safety. The mine scenarios evaluated in this assessment represent approximately 25 to 78 years
of mineral extraction, with additional long-term operations likely required for closeout and maintenance of
the mine. This period is barely within the minimum OBE return period for Class II dams. The MDE analysis
presents a potentially greater chance of underestimating the size of a characteristic earthquake. Tailings
storage facilities would operate during the active mining period and could have a life expectancy of
10,000 years after operations cease. Because the return period for the MDE is 1,000 to 2,500 years, this
could lead to significantly underestimating the largest earthquake that is likely to occur during the period
over which the tailings dams would be in place.
The Initial Application Package for Approval to Construct a Dam submitted by Northern Dynasty Minerals
(NDM) to the Alaska Department of Natural Resources (NDM 2006) included a seismic safety and design
analysis prepared by Knight Piesold Consulting that identified the following design criteria for the tailings
dams at the storage facility.
• OBE return period of 200 years, magnitude 7.5.
• MDE return period of 2,500 years, magnitude 7.8, with maximum ground acceleration of 0.3g (based on
Castle Mountain Fault data).
NDM used a deterministic evaluation to select the MDE and MCE, which were deemed equivalent for the
preliminary safety design. In the application, NDM reports that the preliminary design incorporates
additional safety factors, including design of storage facility embankments to withstand effects of the MDE
and a distant magnitude 9.2 event (NDM 2006). Ghaffari et al. (2011) state that an MCE of magnitude 7.5
with 0.44gto 0.47g maximum ground acceleration was used in the stability calculations for the tailings dam
design. Although the design specifications proposed by Ghaffari etal. (2011) exceed the minimum
requirements for dams in Alaska, the deterministic dataset used is small and contains considerable
uncertainties, which could lead to an underestimate of potential seismic risks.
Bristol Bay Assessment
9-6
January 2014
-------
Chapter 9
Tailings Dam Failure
A number of studies have attempted to analyze the historical record to determine proximate causes and
probabilities of tailings dam failures (Davies et al. 2000, ICOLD 2001, Davies 2002, Rico et al. 2008,
Chambers and Higman 2011). These efforts have been hindered by the lack of a worldwide inventory of
tailings dams, incomplete reporting of tailings dam failures, and incomplete data for known failures. The
National Inventory of Dams (2005) lists 1,448 tailings dams in the United States, and the worldwide
total is estimated at over 3,500 (Davies et al. 2000). The International Commission on Large Dams
compiled a database of 221 tailings dam incidents (events potentially leading to failure) and failures
(events in which dams stop retaining tailings as designed) that occurred from 1917 through 2000
(ICOLD 2001). Causes of incidents and failures were reported for 220 of these, comprising 85 incidents
and 135 failures. Causes of the 135 reported failures are summarized in Table 9-1.
Perhaps most noteworthy is the relatively high number of failures at active versus inactive tailings
dams, primarily resulting from slope instability and failure (Table 9-1). This suggests that the stability of
tailings dams and impoundments may increase with time, as dewatering and consolidation of tailings
occurs and additional loads are no longer applied. However, failures do occur after operation. For
example, rehabilitation of the Gull Bridge Mine in Newfoundland, Canada, occurred in 1999. In 2010, an
inspection found that the tailings dam at the closed mine was deteriorating (Stantec Consulting 2011),
and in 2012 the dam failed, leaving a 50-m gap the height of the dam (Fitzpatrick 2012). The primary
cause of failure for inactive tailings dams is overtopping, which accounts for 80% of recorded failures
with known causes (Table 9-1).
Table 9-1. Number and cause of tailings dam failures at active and inactive tailings dams.
Failure
Failure cause
Overtopping
Slope instability
Earthquake
Foundation
Seepage
Structural
Erosion
Mine subsidence
Unknown
TOTALS
Number of Tailings Dam Failures3
Active Dams
20
30
18
11
10
12
3
3
15
122
Inactive Dams
8
1
0
1
0
0
0
0
3
13
Total
28
31
18
12
10
12
3
3
18
135
Notes:
a Data are presented for 135 tailings dam failures for which causes were reported, from 1917 to 2000.
Source: ICOLD 2001.
9.1.2 Probabilities
It is difficult to estimate the probability of low-frequency events such as tailings dam failures, especially
when each tailings dam is a unique structure subject to unique loading conditions. In addition, failure
probabilities may be estimated and interpreted in different ways (Box 9-3).
Bristol Bay Assessment
9-7
January 2014
-------
Chapter 9 Tailings Dam Failure
BOX 9-3. INTERPRETATION OF DAM FAILURE PROBABILITIES
There are two fundamental types of probability interpretations: frequentist and subjectivist.
Frequentist probabilities are based on observed frequencies of past events. For example, based on the
observed frequency of tailings dam failures (88 in 176,000 dam-years, where dam-year is the existence of
one dam for one year), we estimate a frequency of 1 failure in 2,000 dam-years (or 0.00050 failures per
dam-year). In conventional risk probabilities, this means the following.
• Each year, there is a 5x 10'4 probability of failure per dam.
• Out of 200 dams, one fails each decade on average; out of 2,000 dams, one fails each year on average.
Strictly speaking, frequentist probabilities are properties of populations. However, by extension, if there is
one dam and it is typical of the population, it would be expected to fail, on average, within a 2,000-year
period. This does not mean it is expected to fail 2,000 years after it is built; a failure could occur during any
year. Rather, it indicates that, after 2,000 years have passed, it is more likely than not that the dam would
have failed (i.e., half of a population of such dams would have failed 2,000 years after they were built),
although the actual failure could occur any year in that 2,000-year window.
Subjectivist probabilities are based on degree of belief. For example, if engineers design a dam using novel
methods, they cannot make use of frequencies when estimating failure risks. They may, however, use a
model or best professional judgment to support a statement that the annual probability of failure is some
value (e.g., 1 x 10'6, or 1 failure in a million dam-years). As with frequentist probabilities, this does not mean
that the dam is expected to fail only after a million years have passed. Because subjective probabilities are
not based on frequencies, they are typically described as equivalent to betting odds—that is, the engineers
would be willing to accept a bet in which, if the dam stands for a year they win $1, but if it fails they pay $1
million. Rather than present subjective probabilities of failure, designs are more commonly said to conform
to standard or best engineering practices.
Despite these difficulties, several studies have calculated the frequency of past tailings dam failures,
resulting in the following failure frequencies.
• An estimated 0.00050 failures per dam-year (where dam-year is the existence of one dam for one
year), or 1 tailings dam failure every 2,000 dam-years, based on 88 failures from 1960 to 2010
(Chambers and Higman 2011).
• An estimated 0.00049 failures per dam-year, or 1 tailings dam failure every 2,041 dam-years, based
on 3,500 appreciable tailings dams that experienced an average of 1.7 failures per year from 1987 to
2007 (Peck 2007).
• An estimated 0.00057 to 0.0014 failures per dam-year, or 1 tailings dam failure every 714 to
1,754 dam-years, based on a database (including many unpublished failures) that showed 2 to
5 major tailings dam failures per year from 1970 to 2001 (Davies et al. 2000, Davies 2002).
Available data do not permit reliable estimation of failure rates for different causes of failure or stages of
activity. Although most failures have occurred while the tailings dams were actively receiving tailings
(Table 9-1), the dam inventories do not indicate whether the thousands of dams in the inventory are
active or inactive and do not include years of operation. This prevents estimation of the proportion in
each category and makes it impossible to calculate the number of active dam-years. Low failure
frequencies and incomplete datasets also make any meaningful correlations between failure
probabilities and dam height or other characteristics questionable. For example, although the 1,448
tailings dams listed in the National Inventory of Dams create a statistically large and fairly complete
Bristol Bay Assessment qo January 2014
-------
Chapter 9 Tailings Dam Failure
database that includes dam heights, the International Commission on Large Dams failure database
includes only 49 U.S. tailings dam failures—too small a dataset to develop a meaningful correlation
between dam height and failure probability. Very few existing rockfill dams approach the size of the
structures in our mine scenarios, and none of these large dams have failed.
The historical frequencies of tailings dam failures presented above may be interpreted as an upper
bound on the failure probability of a modern tailings dam. Morgenstern (2011), in reviewing data from
Davies and Martin (2009), did not observe a substantial downward trend in failure rates over time.
However, improvements in the understanding of dam behavior, dam design, construction techniques,
construction quality control, dam monitoring, and dam safety assessment would be expected to reduce
the probability of failure for dams designed, constructed, and operating using more modern or advanced
engineering techniques.
Silva et al. (2008) reported on over 75 earthen dams, tailings dams, natural and cut slopes, and some
earth-retaining structures to illustrate the relationship between the level of engineering, the annual
probability of slope failure in earthen structures, and factors of safety. They grouped projects into the
following four categories based on the level of engineering applied to design, site investigation,
materials testing, analysis, construction control, operation, and monitoring of each project.
• Category I: Facilities designed, built, and operated with state-of-the-practice engineering. Generally,
these facilities are constructed to higher standards because they have high failure consequences.
• Category II: Facilities designed, built, and operated using standard engineering practice. Many
ordinary facilities fall into this category.
• Category III: Facilities without site-specific design and substandard construction or operation.
Temporary facilities and those with low failure consequences often fall into this category.
• Category IV: Facilities with little or no engineering.
The State of Alaska regulates its tailings dams under Alaska Administrative Code (AAC) Title 11, Chapter
93, Article 3, Dam Safety (11 AAC 93). Each dam is assigned to a class based on the potential hazards of a
tailings dam failure (Table 9-2). Given that anadromous fish would be affected but no loss of human life
is expected under the tailings dam failure scenarios, Class II would be applicable, although a mine
operator might choose to exceed state requirements and meet Class I. Therefore, the tailings dams in the
mine scenarios would be classified as either Hazard Class I or II, both of which require a detailed
computer stability analysis with verification by other methods, and may require more sophisticated
finite element analyses in special circumstances. This analysis considers the effects of earthquakes
based on a site-specific evaluation of seismicity in the area (Section 3.6). Box 9-2 describes the selection
of earthquake characteristics for design criteria.
Bristol Bay Assessment qq January 2014
-------
Chapter 9
Tailings Dam Failure
Table 9-2. Summary of Alaska's classification of potential dam failure hazards.
Hazard Class
1 (High)
II (Significant)
III (Low)
Effect on Human Life
Probable loss of one or
more lives
No loss of life expected,
although a significant
danger to public health
may exist
Insignificant danger to
public health
Effect on Property
Irrelevant for classification, but may include the same losses indicated in Class
II or III
Probable loss of or significant damage to homes, occupied structures,
commercial or high-value property, major highways, primary roads, railroads, or
public utilities, or other significant property losses or damage not limited to the
owner of the barrier
Probable loss of or significant damage to waters identified under 11 AAC
195.010(a) as important for spawning, rearing, or migration of anadromous fish
Limited impact on rural or undeveloped land, rural or secondary roads, and
structures
Loss or damage of property limited to the owner of the barrier
Notes:
Tailings dams in the mine scenarios would be classified as Hazard Class 1 or II.
AAC = Alaska Administrative Code.
Source: ADNR 2005.
The Guidelines for Cooperation with the Alaska Dam Safety Program (ADNR 2005) do not specify a
minimum safety factor for dams, but rather allow the applicant to propose one. Guidelines suggest that
the applicant follow accepted industry design practices such as those provided by the U.S. Army Corps of
Engineers (USAGE), the Bureau of Reclamation, the Federal Energy Regulatory Commission (FERC), and
other agencies. Both USAGE and FERC require a minimum safety factor of 1.5 for the loading condition
corresponding to steady seepage from the filled storage facility (FERC 1991, USAGE 2003). Based on the
correlations among level of engineering, factor of safety, and slope failure probability derived from Silva
et al. (2008), application of a 1.5 safety factor yields an expected annual probability of slope failure
between 0.0001 (Category II) and 0.000001 (Category I) (Figure 9-2). This translates to one tailings dam
failure due to slope failure every 10,000 to 1 million dam-years.
The upper bound of this range (0.0001) is lower than the historical average of 0.00050 (1 failure every
2,000 dam-years) for tailings dams. This is partly because slope failure is only one of several possible
failure mechanisms, but it also suggests that some past tailings dams may have been designed for lower
safety factors or designed, constructed, operated, or monitored to lower than Category II engineering
standards. As shown in Table 9-1, slope failures only account for about 25% of all tailings dam failures
with known causes. Thus, the probability of failure from all causes may be about four times higher than
dam failures from slope instability alone (yielding an expected annual probability of failure between
0.0004 and 0.000004, or one tailings dam failure every 2,500 to 250,000 dam-years), although it is
important to recognize that this small dataset may not be representative. Because 90% of tailings dam
failures have occurred in active dams (Table 9-1), the probability of a tailings dam failure after TSF
closure would be expected to be lower than the historical average for all tailings dams.
Bristol Bay Assessment
9-10
January 2014
-------
Chapter 9
Tailings Dam Failure
Figure 9-2. Annual probability of dam failure due to slope failure vs. factor of safety (modified from
Silva et al. 2008).
CL'
1E+0
Category I Projects
Category II Projects
1.2 1.3
Factor of Safety
These low probabilities are based on failure frequencies within categories of engineering practice and
safety factors, but the authors describe results as "semiempirical" due to the judgment involved in
categorizing the dams and creating the curves to describe the relationships (Silva et al. 2008). Modern,
high earthen dams do not exist in large numbers and have not existed for long periods of time, and the
frequencies and time courses of failures may differ from both the historical record and design goals. In
particular, the failure rates of large earthen dams that are hundreds of years old are not known.
Given an annual probability of failure per dam-year, we can calculate the probability of the failure of any
project dam over any number of years. The three mine scenarios have different numbers of dams and
different operating lives: the Pebble 0.25 scenario has a single tailings dam and an operating life of
20 years; the Pebble 2.0 scenario has three tailings dams and an operating life of 25 years; and the
Pebble 6.5 scenario has eight tailings dams and an operating life of 78 years. Using an upper bound
annual probability of failure of 0.0004, the probability of dam failure would range from 0.8% to 22%
over the operating life of each scenario (Table 9-3). This range decreases to 0.008% to 0.25% when a
lower bound annual failure probability of 0.000004 is used (Table 9-3). If the tailings in theTSFs remain
saturated (e.g., to keep the pyritic tailings covered with water), the potential for dam failure over a
longer period needs to be considered. The probability that any of the dams would fail during a post-
closure period of 1,000 years ranges from upper bounds of 33 to 96% to lower bounds of 0.4 to 3%
across the three scenarios (Table 9-3).
Bristol Bay Assessment
9-11
January 2014
-------
Chapter 9
Tailings Dam Failure
Table 9-3. Summary of tailings dam failure probabilities in the three mine scenarios.
Time Period
Operational life
1,000 year post-closure period
Annual Failure Probability
0.0004
0.000004
0.0004
0.000004
Probability of Failure
Pebble 0.25'
0.8
0.008
33
0.4
Pebble 2.0"
3
0.03
70
1.2
Pebble 6.5C
22
0.25
96
3
Notes:
"" Operational life of 20 years; 1 tailings dam.
b Operational life of 25 years; 3 tailings dams.
c Operational life of 78 years, 8 tailings dams.
9.1.3 Uncertainties
The variability in published probabilities of tailings dam failure reflects the uncertainty inherent in these
estimates. Much of this uncertainty is due to incomplete data. TSFs may remain in place for long periods.
Most dams are created as water-holding dams with limited expected lifespans (generally about
50 years). TSFs can be drained after mine closure to reduce the probability and consequences of tailings
dam failures, but draining a thick layer of fine-grained material can be difficult. In the mine scenarios,
only 17 to 28% of net precipitation (depending on the TSF) would need to infiltrate into the tailings to
maintain full saturation with steady-state downward flow, so draining the TSFs would require
maintaining a high runoff percentage. Furthermore, if tailings ponds need to be maintained to keep
pyritic tailings hydrated and isolated from oxidation, tailings dams must retain solid and liquid materials
behind them in perpetuity—meaning that the dams must be maintained in perpetuity, in the face of
uncertain seismic and weather events that may occur over thousands of years and have cumulative
effects.
9.2 Material Properties
9.2.1 Tailings Dam Rockfill
In the mine scenarios, TSFs would be enclosed by rockfill dams constructed primarily of well-graded,
non-acid-generating (NAG) waste rock obtained from the mine pit during operations. The starter dike
would contain material excavated from the upstream toe trench and local quarry. Waste rock from the
mine pit would be used as it became available. The size of the rock used to construct the dam would
depend on the rock's fracture characteristics, the methods used to blast and remove it from the mine pit,
and the lift thickness specified for adequate compaction. Particle sizes used to construct tailings dams
typically range from sand to large boulders (Blight 2010). For a large rockfill dam with a high or
significant hazard potential, lift thickness would be expected to be limited to 1.5 m to guarantee
adequate compaction, which limits the maximum particle size to about 1 m (Breitenbach 2007).
Well-graded rock would have a coefficient of uniformity (Deo/Dio) greater than 4 and would have a
coefficient of curvature (Dso/ (D6o*Dio)) between 1 and 3. Combining these coefficients with Dawson
Bristol Bay Assessment
9-12
January 2014
-------
Chapter 9
Tailings Dam Failure
and Morin's (1996) report of a Dso particle size greater than 200 mm for waste rock, one can generate a
representative particle size distribution curve for the bulk of the tailings dam material (Figure 9-3).
Figure 9-3. Representative particle size distributions for tailings solids (bulk and pyritic tailings) and
tailings dam rockfill. Tailings distributions are based on particle sizes specified by Ghaffari et al.
(2011) and on typical tailings particle size distributions.
100
-bulktailings
•pyritictailings
•combined tailings
-rockfill
0
0.001
0.01 0.1 1 10
Particle Size (mm)
100
1000
9.2.2 Tailings Solids and Liquids
The tailings solids would include both bulk and pyritic tailings. The bulk tailings would consist largely of
sand and silt-sized particles (Dso = 200 um) and have an average dry density of 1.36 metric tons/m3. The
pyritic tailings would consist of predominantly silt-sized particles (Dso = 30 um) and have an average
dry density of 1.76 metric tons/m3. The mass of the bulk tailings and the pyritic tailings would equal 85
and 14% of the ore mass, respectively (Ghaffari et al. 2011). Representative particle size distribution
curves for bulk, pyritic, and combined tailings are shown in Figure 9-3.
Given the dry density of the bulk tailings reported above and the specific gravity reported for the ore
(2.63 for the solids) (Ghaffari et al. 2011), the bulk tailings would be 52% solids and 48% liquid by
volume. The pyritic tailings, given the dry density reported above and a solids-specific gravity of 3.00
(Ghaffari et al. 2011), would be 59% solids and 41% water. Based on the proportions of bulk and pyritic
tailings, the combined material in the TSF would be 53% solids and 47% water by volume, exclusive of
any ponded water above the settled tailings. As the tailings consolidate, the bulk density of the deeper
tailings would be expected to increase above the average density, although this consolidation may be
limited (Section 6.3.2).
Bristol Bay Assessment
9-13
January 2014
-------
Chapter 9 Tailings Dam Failure
9.3 Modeling a Tailings Dam Failure
Although a tailings dam failure is a low-probability event, the probability is not zero. Should such an
unlikely event occur, it is important to understand its potential impacts on the Bristol Bay watershed. In
this assessment, we consider the effects of two potential dam failures at TSF 1: one at a volume
approximating the complete Pebble 0.25 scenario (92-m dam height, with 158 million m3 of tailings
produced) and one at a volume approximating the complete Pebble 2.0 scenario (209-m dam height,
with, 1,270 million m3 of tailings produced). In both cases, we assumed 20% of the impounded tailings
(solids and pore water) would be mobilized (Azam and Li 2010, Dalpatram 2011). Although it is
reasonable to expect that 30 to 66% of the impounded tailings material could contribute to debris flow
following a tailings dam failure, given the particle size distribution of the tailings (Browne 2011), we
used a conservative estimate of 20% to account for the fact that the volume of material mobilized, the
distance it travels downstream, and the amount of deposition can vary greatly based on numerous
factors (e.g., dam height, material size distribution, material water content at time of failure) (Rico et al.
2008).
As detailed in Box 9-4, we used the USAGE Hydrologic Engineering Center's River Analysis System (HEC-
RAS) to model hydrologic characteristics of the dam failures. This tool requires the selection of one of
two failure initiation mechanisms: overtopping or piping failure. We selected overtopping as the
initiating event for final model runs for several reasons.
• The assessment TSF dam includes a liner (Section 6.1.2.4) that would reduce the risk of
embankment failure due to seepage and piping (Section 9.1.1).
• Many of the failure mechanisms listed in Section 9.1.1 involve failure via breaching or overtopping,
and thus are better approximated by the overtopping modeling approach in HEC-RAS.
• Overtopping could plausibly occur, for example, if storage freeboard was exceeded due to excessive
precipitation, settlement over time, or a landslide or seismic event, or if any designed overflow
spillway became blocked by ice or debris.
Although we modeled an overtopping failure, sensitivity analysis showed that model results were
insensitive to initiation type relative to failure duration (Box 9-4)—that is, the mechanism of failure
initiation did not significantly influence potential effects. The overtopping failure outputs were
compared to similar piping outputs generated by subsequent HEC-RAS model runs. Comparison of peak
discharges at the dam indicated that failure by overtopping generated the smallest expected flood wave
peaks, and did not create a situation in which the selection of model assumptions overestimated the
potential for flooding. Comparison of peak discharges was also reviewed by varying the time to full dam
breach from 30 minutes to 4 hours (Gee 2008). Results indicated that magnitude of the peak flood wave
was sensitive to breach development time, so we selected 2 hours as a reasonable time to full dam
breach (Box 9-4).
Bristol Bay Assessment 914 January 2014
-------
Chapter 9 Tailings Dam Failure
BOX 9-4. METHODS FOR MODELING TAILINGS DAM FAILURES
We modeled hydrologic characteristics of tailings dam failures at tailings storage facility (TSF) 1 in the
Pebble 0.25 and Pebble 2.0 scenarios using the U.S. Army Corps of Engineers Hydrologic Engineering
Center's River Analysis System (HEC-RAS). Under both dam failure scenarios we modeled hydrologic
conditions (e.g., water discharges, depths, and velocities) in the stream channel and floodplain during and
immediately following dam failure, and then used these outputs to estimate tailings transport and
deposition along the stream network. We limited the extent of the model to a 30-km reach downstream of
the TSF (i.e., from the face of the TSF 1 tailings dam down the North Fork Koktuli River valley to the
confluence of the South and North Fork Koktuli Rivers); extension of the simulation beyond this point would
have introduced significant error and uncertainty associated with the contribution of the South Fork Koktuli
River flows.
HEC-RAS inputs included geometry of an inline structure to simulate the dam cross-section and stream
channel geometry data, both of which we derived from a 30-m digital elevation model. Flow calculations are
completed between successive cross-sections in the model, balancing the hydraulic energy to determine the
water surface elevations and flow velocity, and then moving to the next cross-section in the sequence and
repeatingthe process. Because HEC-RAS is most often used to simulate clear water flows, it is appropriate
to increase the channel roughness coefficient (i.e., Manning's n coefficient) to better emulate flow
characteristics of the sediment-rich water released during a tailings dam failure; thus, we used a Manning's
n = 0.09 for analyses.
We present model outputs for overtopping failures at both the Pebble 0.25 scenario dam (92-m dam height)
and the Pebble 2.0 scenario dam (209-m dam height), assuming a 2-hour failure duration and the release of
20% of available tailings storage capacity in each failure. In HEC-RAS, options for initiating a dam failure are
limited to overtopping or piping failure. Both initiation types were modeled to examine sensitivity to initiation
conditions. In addition, a range of dam failure durations (30 minutes to 4 hours) was examined (Gee 2008).
Results showed that peak flows during a failure were much more sensitive to failure duration than to
initiation type. The 2-hour duration to full failure generated peak flows that fell within the middle range of
potential peak flows to consider. Overtopping generated the smallest peaks in the 2-hour simulation group
(Qmax = 39,100 m3/s). Piping failure in the 2-hour group was tested for failures initiating near the base of the
dam, at mid-elevation, and near the top of the dam face, generating Qmax values of 92,263 m3/s, 85,747
m3/s, and 48,868 m3/s, respectively. The 30-minute simulation group average Qmax values were 222%
greater, and the 4-hour simulation group average Qmax values were 38% lower.
We assumed a particle size distribution of 0.001- to 1.0-m diameter for the dam construction material, and
less than 0.01-to just over 1.0-mm diameter for the impounded tailings material (Figure 9-3). We focused
on transport and deposition of fine-grained (less than 1.0-mm diameter) tailings material, since larger dam
construction material would likely deposit within the first few kilometers downstream of the failure. Based on
the Hjulstrom curve—which estimates when a stream or river will erode, transport, or deposit sediment
based on flow speed and sediment grain size—all mobilized tailings would remain in suspension at water
velocities greater than 0.05 m/s. Thus, the channel would transport tailings under typical stormflow
conditions, and deposited tailings from floodplain terraces could be suspended and transported during
typical storm events following the failure.
Based on historical failures, we assumed that sediment deposition would be greatest near the dam, forming
an initial "wedge" that would be deposited rapidly and extend from the lowest elevation of the breach. Given
the potential mobility of the fine-grained tailings, we held the initial modeled slope to 1.6%, the valley slope
near the dam. We determined this slope was a reasonable estimate based on comparison with a publicly
available, simple tailings flow calculator that predicts flow depths for tailings with a variety of viscosities
(WISE 2012). Extending this slope from the dam breach, calculated sediment depths ranged from 1 m to 20
m 1.4 km downstream of the dam for both failure scenarios. We modeled that, on average, approximately 1
m of deposited tailings would remain on valley surfaces (i.e., in the channel and on the floodplain)
downstream of the dam following each failure; this created a conservative, uniform estimate of sediment
deposition. Deposition at each cross-section at this 1 m meter depth was used to calculate the volume
between modeled river sections, and this volume was subtracted from the volume released from the tailings
dam failure. We assumed that the remaining sediment in the tailings dam failure flow was available to
deposit at the next downstream section, and this logic was carried downstream until the end of the modeled
river length was reached.
Bristol Bay Assessment 915 January 2014
-------
Chapter 9 Tailings Dam Failure
9.3.1 Hydrologic Characteristics
Model results for hydrologic characteristics of the Pebble 0.25 and Pebble 2.0 tailings dam failures are
shown in Table 9-4. In both cases, estimated peak flows during a TSF dam failure would be much larger
than streamflows typically experienced in this watershed. This is because the impounded tailings would
create a flood wave far larger than any that could result from a precipitation event alone. The tailings
dam failure and subsequent release of massive quantities of impounded tailings and associated pore
water would produce a peak flood immediately downstream of the dam. Maximum depths of the flood
wave would exceed 10 m and 25 m, with peak velocities of approximately 4 and 10 m/s (14 and 36
km/hr), for the Pebble 0.25 and Pebble 2.0 dam failures, respectively (Table 9-4).
Peak discharges would exceed 5,000 m3/s for the Pebble 0.25 dam failure and 39,000 m3/s for the
Pebble 2.0 dam failure. If the failure occurred during an intense rainfall or rapid snowmelt event
discharges would be negligibly higher, due to the small watershed area of the TSF 1 dam. A dam-break
flood of this magnitude would dwarf the peak flows of even the largest rivers in the region. For example,
a local flood event measured by a U.S. Geological Survey (USGS) gage on the Nushagak River located near
the village of Ekwok, Alaska (Figure 2-4), experienced a record peak flood of 3,313 m3/s. Peak flows
predicted in the North Fork Koktuli River valley from the TSF dam failures would be more than 1.6 times
(Pebble 0.25) and 11.8 times (Pebble 2.0) greater than the flood of record on the Nushagak River at
Ekwok. Although we recognize that these are not analogous watersheds, this observed flood does
provide a point of reference for the flood magnitudes that would result from tailings dam failures.
9.3.2 Sediment Transport and Deposition
Dam failure flood waves (Table 9-4) and post-failure recessional flows in the Pebble 2.0 failure scenario
suggest that transport and deposition of tailings material could occur throughout (and beyond) the 30-
km modeled reach (Table 9-5). Deposition in the Pebble 0.25 failure scenario could extend for over 29
km, to within 1 km of the confluence with the South Fork Koktuli (Table 9-5). After the initial deposition
event, concentrated channel flows and floodplain conveyance areas would continue to transport
sediment further downstream, as channel and valley morphology would re-establish in the newly
deposited substrate.
Even with only 20% of impounded tailings mobilized, the flood wave and tailings deposition that would
result from a tailings dam failure under either dam failure scenario could significantly alter the
downstream channel and floodplain. The initial flood itself would have the capacity to scour the channel
and floodplain, as the wave of tailings slurry would travel down the valley at velocities of up to
approximately 10 m/s (Table 9-4). The quantity of mobilized sediments that could be released from the
TSF would bury the existing channel and floodplain under meters of fine-grained sediment in an initial
wedge near the dam; this material would move across the downstream valley as the flood wave receded
and water velocities slowed (Box 9-4, Tables 9-4 and 9-5). The sediment regime of the affected stream
and downstream waters would be greatly altered, with calculated sediment depths of up to 20 m
(Pebble 2.0) and 1 m (Pebble 0.25) extending 1.4 km downstream of the dam.
Bristol Bay Assessment 916 January 2014
-------
Chapter 9
Tailings Dam Failure
river stations along a 30-km length of the North Fork Koktuli River; representative river stations along that length are shown here, listed by
distance upstream from the confluence of the South and North Fork Koktuli Rivers (River Station 30 km = foot of the TSF 1 dam).
River
Station
(km)
30.0
29.0
25.0
20.4
15.5
10.3
4.8
0.0
Pebble 0.25 Dam Failure3
Discharge
(m3/s)
5,270
5,270
4,990
4,190
3,610
2,940
2,650
1,710
Depth
(m)
11.2
10.7
9.8
7.5
10.0
12.3
4.1
5.8
Velocity (m/s)
LFP
1.6
1.2
0.4
0.5
0.8
0.8
0.3
0.1
CH
3.1
4.2
0.8
0.9
1.3
1.6
0.5
0.4
RFP
1.4
1.4
0.6
0.5
0.7
0.7
<0.1
<0.1
Pebble 2.0 Dam Failure11
Discharge
(mVs)
39,100
39,100
39,000
38,000
34,900
29,800
25,800
18,600
Depth
(m)
23.6
19.8
20.2
19.0
25.8
27.2
10.1
10.6
Velocity (m/s)
LFP
4.1
5.2
1.3
1.5
1.5
2.5
0.9
0.5
CH
7.2
10.5
2.1
2.5
2.6
4.2
1.4
0.9
RFP
4.2
5.8
1.4
1.5
1.8
2.3
0.5
<0.1
Notes:
3 Dam height = 92 m, maximum volume of tailings and water expected to be stored = 158 million m3.
b Dam height = 209 m, maximum volume of tailings and water expected to be stored = 1,270 million m3.
HEC-RAS = Hydrologic Engineering Center's River Analysis System; LFP = leftfloodplain; CH = channel; RFP = right floodplain; TSF = tailings storage facility.
Bristol Bay Assessment
9-17
January 2014
-------
Chapter 9
Tailings Dam Failure
Table 9-5. Tailings mobilized and deposited in the Pebble 0.25 and Pebble 2.0 dam failures analyses. The mobilized tailings include
material within the dam cross-section that has failed, plus a percentage (5 to 20%) of the stored tailings material. See Box 9-4 for additional
information on how the dam failures were modeled.
Failure Scenario
Pebble 0.25
Pebble 2.0
Volume of Tailings3
(million m3)
158
1,270
% Mobilized"
20
15
10
5
20
15
10
5
Mobilized Tailings
(metric tons)
59,724,000
44,793,000
29,862,000
14,931,000
479,682,000
359,761,500
239,841,000
119,920,500
Tailings in Transport
at Downstream
Extent of Modelc
(metric tons)
0
0
0
0
350,668,000
241,756,000
138,981,000
39,767,000
Downstream Extent of
Wedge
(km)
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
Downstream Extent of
Expected Deposition11
(km)
29
27
24
9
30(+)
30(+)
30(+)
30(+)
Notes:
a Maximum volume of tailings and water expected to be stored, allowing for freeboard in tailings storage facility (TSF). This volume was used to estimate metric tons of stored tailings released in a TSF
dam failure, using an average tailings total density of 1.89 metric tons/m3 and an average tailings dry density of 1.42 metric tons/m3.
b 20% value was used in model; values less than 20% are shown to illustrate how weight of mobilized tailings changes with % mobilized.
c Weight of mobilized tailings that would remain in transport assuming 1 m of deposition in the floodplain inundation area.
d Measured downstream from face of dam.
Bristol Bay Assessment
9-18
January 2014
-------
Chapter 9 Tailings Dam Failure
Downstream of this initial sediment wedge, deposition could occur in the channel and the floodplain as
peak flood discharges decreased with increasing distance downstream of the dam, water velocities
returned to baseflow levels, and the potential for tailings deposition increased. In the Pebble 0.25 failure
scenario, release of 20% of tailings material was sufficient to fill the entire North Fork Koktuli River
valley to within 1 km of the confluence with the South Fork Koktuli River with an average depth of 1 m
of tailings material (Table 9-5). In the Pebble 2.0 failure scenario, over 350 million metric tons of
sediment remained available for transport and subsequent deposition beyond the end of the modeled
reach, indicating that tailings would extend into the mainstem Koktuli River (Table 9-5).
Most of the deposition would be very fine material that would be susceptible to resuspension and
deposition with each subsequent natural flow event. Following the dam failure, the stream channel
would seek equilibrium and could remain unstable over several flow events, potentially creating a
braided system in the post-failure depositional zone. As the new valley fluvial geomorphology developed
over time, newly deposited materials on the floodplain, material at the base of the dam, and material
that remained behind the breached dam of the TSF (if not removed or contained by corrective action)
would serve as concentrated sources of easily transportable, potentially toxic material (Section 9.4).
The two possible failure scenarios presented here are well within the range of reported case histories.
For example, when the parameters for the Pebble 0.25 and Pebble 2.0 dam failures were applied to
runout distance equations from Rico et al. (2008), expected runout distances reached the marine waters
of Bristol Bay. In our analyses, we made a simple assumption that deposition depths averaged 1 m (Box
9-4). We emphasize that our tailings dam failure scenarios reflect a range of possible outcomes, but are
not exhaustive. The depth of tailings deposition on the floodplain could be higher or lower, depending
on the amount of tailings mobilized and the runout distance. Based on historical tailings dam failure
data, potential runout distances can range from hundreds to thousands of kilometers (Box 9-1).
9.3.3 Uncertainties
In this chapter, we have evaluated two potential dam failure scenarios, both caused by overtopping.
Although our sensitivity analyses indicate that the repercussions of failure were relatively insensitive to
the initial cause of the failure (Box 9-4), it is important to note that overtopping represents only one of
several potential failure mechanisms (Section 9.1.1).
Also, a significant amount of uncertainty surrounds potential sediment deposition depths and
downstream distributions. Valley topography, rate of the dam failure, and ultimate make-up of the flood
wave sediment concentration and viscosity can affect outcomes and complicate predictive efforts.
Despite the uncertainty associated with the massive quantities of sediment available and the
complexities of hydraulic forces that would act on this sediment, we present reasonable post-failure
sediment deposition outcomes in the two dam failure scenarios. Other outcomes are possible, but all
share the common reality that massive quantities (i.e., millions of cubic meters) of tailings fines would
be deposited in downstream floodplains and channels (Table 9-5).
Bristol Bay Assessment 919 January 2014
-------
Chapter 9 Tailings Dam Failure
Use of a traditional sediment transport model would likely improve estimates of sediment movement
and deposition, especially as the model is extended further downstream. In addition, tributary streams
would input clean water at each confluence. Because of the site-specific data required to implement a
sediment transport model, we limited our model to the 30 km above the confluence of the South and
North Fork Koktuli Rivers (Box 9-4).
9.4 Scour, Sediment Deposition, and Turbidity
The Pebble 0.25 and Pebble 2.0 tailings dam failures described in the preceding section could have
devastating effects on aquatic life and habitat (Figure 9-1). We identified several processes associated
with a tailings dam failure that would pose risks to aquatic habitat. These processes include exposure to
hydraulic scour and bed mobilization, deposition of tailings fines, and mobilization and suspension of
tailings fines affecting downstream water and habitat quality. Effects of suspended sediments are
discussed in Section 9.5.1, and effects associated with potential toxicity are discussed in Section 9.5.2.
Natural background conditions indicate the sediment levels that support the region's current
productivity of salmonid populations, and two available sources provide data on substrate size
distribution and fine sediment concentrations in the study area. Pebble Limited Partnership (PLP)
(2011) reports concentrations of fine sediments from sieved bulk gravel samples collected at three
known salmon spawning sites in the South and North Fork Koktuli Rivers and Upper Talarik Creek
(sample locations are shown in report by PLP [2011: Figure 4 in Appendix 15.IF]). Average
concentration of fines (less than 0.84 mm) was less than 6% for all streams and dates, except for the
August sample from the uppermost South Fork Koktuli River site (gage SGSK3) (PLP 2011: Figure 27 in
Appendix 15.IF), which had nearly 8% fines. The geometric mean grain size was greater than 15 mm at
all sites for both sampling periods, except the uppermost Upper Talarik Creek site (gage SGUT3), where
the mean grain size for both seasons was between 10 and 15 mm (PLP 2011: Figure 26 in Appendix
15.IF). These data led the authors to conclude that gravel quality was generally high and that, based on
published criteria (Shirazi etal. 1981, Chapman and McLeod 1987, Kondolf 2000), salmonid survival to
emergence would be high (presumably above 80%) at all sites except the uppermost Upper Talarik
Creek site, where criteria predicted survival between 50 and 80% (PLP 2011).
Areal coverage of substrate sizes is also available for 77 wadeable stream sites around the Nushagak and
Kvichak River watersheds, including one site each on the South and North Fork Koktuli Rivers and
Upper Talarik Creek (Table 9-6). Substrate sampling at these study sites followed U.S. Environmental
Protection Agency (USEPA) methodology (Peck et al. 2006), in which five particles are systematically
selected across each of 21 evenly spaced transects (from each wetted margin of the channel and three
locations in between). These data indicate that a mix of substrate sizes occurs in these streambeds, with
cobble and gravel generally abundant (Table 9-6).
Bristol Bay Assessment 920 January 2014
-------
Chapter 9
Tailings Dam Failure
sites in the Nushagak and Kvichak River watersheds. Values represent percentage areal coverage based on 105 systematically selected
particles at each site, following U.S. Environmental Protection Agency methods. All data were collected during June.
River or Stream(s)
South Fork Koktuli River
North Fork Koktuli River
Upper Talarik Creek
77 streams
Date
6/8/2010
6/6/2009
6/13/2011
2008 to 2011
Latitude
59.83047
59.84033
59.91820
-
Longitude
-155.27719
-155.71272
-155.27771
-
% Large
Boulder
(>1000
mm)
-
0(±1)
% Small Boulder
(250-1000
mm)
3
-
2
2 (±4)
% Cobble
(64-250
mm)
3
17
30
13 (±13)
% Coarse
Gravel
(16-64 mm)
55
49
29
40 (±15)
% Fine
Gravel
(2-16 mm)
16
24
13
17 (±12)
% Sand
(0.06-2
mm)
0
10
24
17 (±11)
% Fines
(<0.06
mm)
23
-
2
-
Notes:
Dashes (-) indicate values equal to zero.
Sources: Rinella pers. comm., Peck et al. 2006.
Bristol Bay Assessment
9-21
January 2014
-------
Chapter 9 Tailings Dam Failure
9.4.1 Exposure through Sediment Transport and Deposition
The tailings dam failure scenarios evaluated here would result in intense scour and extensive deposition
in the North Fork Koktuli River valley. Deposition would extend from the tailings dam downstream for
many kilometers. Even with our conservative assumption that 20% of the tailings would be released,
deposition would extend to within 1 km of or beyond the confluence with the South Fork Koktuli River, a
distance of approximately 30 km (Table 9-5). This volume of available fine tailings material could result
in many meters of deposition in a sediment wedge across the entire valley near the TSF dam, with lesser
thicknesses of fines deposited to the confluence with the South Fork Koktuli River or beyond. Erosion
and transport of fines would be expected to continue as the channel adjusted to the vastly increased fine
sediment supply.
To translate these tailings dam failures into effects on aquatic habitat and biota, we assumed that the
calculated velocities during the tailings dam failure flood event (Table 9-4) and the associated transport
and deposition of tailings material and collected debris (Table 9-5) would result in a reworking and
mobilization of the existing North Fork Koktuli River channel bed and banks downstream of the TSF.
Given the volume of material that would be exported from the TSF, we assume that portions of the new
valley floor would be predominantly tailings material, with 70% of the particle mass being finer than 0.1
mm. Following recession of the tailings dam failure flood event, we assume that the bed, margins, and
floodplain would be primarily tailings material, with incorporated coarser dam fill and valley fill
material accounting for less than 20%.
Immediately following either a Pebble 0.25 or Pebble 2.0 tailings dam failure, suitable spawning and
rearing habitat for salmon and other native fishes would be eliminated in the North Fork Koktuli River
downstream of the tailings dam. Tributaries of the North Fork Koktuli River, including portions of the
watershed upstream of the confluence of the North Fork Koktuli River tributary containing the TSF,
could also be adversely affected. Temporary flooding of tributary junctions during the tailings dam
failure event, and subsequent sediment deposition at confluence zones causing local aggradation,
steepening, or shallowing of tributary confluences, could make movement of resident and anadromous
fish between tributaries and the mainstem difficult. Recovery of channel dimensions and substrate size
distributions suitable for salmonid spawning and rearing habitat would be contingent upon rates of fine
sediment export and recruitment of gravels and larger substrates from tributaries or pre-failure valley
fill.
The type, magnitude, and frequency of channel adjustments that would occur in the North Fork Koktuli
River valley following a tailings dam failure would depend on available sediment, channel slope, and
discharge. Post-failure streams flowing across the depositional zone would have tremendous supplies of
fine-grained sediments in the channel bed and banks available for transport. Channels would likely
experience rapid channel incision with frequent bank failure, followed by periods of channel widening
and aggradation interspersed with episodic channel avulsion. Given the volume and depth of deposition,
stream channels would likely remain unstable and continue to contribute sediments to downstream
reaches until equilibrium conditions were met. Recovery of suitable structural habitat in the North Fork
Bristol Bay Assessment 922 January 2014
-------
Chapter 9 Tailings Dam Failure
Koktuli River watershed would likely take decades, given the volume of sediment that could be
delivered in the tailings dam failures considered here. Whether the benefits of removing spilled tailings
fines would outweigh the risks of additional adverse impacts resulting from dredging and removal
operations would depend on the nature and distribution of the tailings spill, the duration of risks, and
existing technologies (e.g., Wenning et al. 2006).
The tailings dam failure scenarios evaluated here would have the potential to fill the North Fork Koktuli
River valley with extensive deposits of tailings fines and, in some cases, still carry a substantial volume
of fine sediments farther downstream. The mass of material remaining in transport at the confluence of
the South and North Fork Koktuli Rivers following a Pebble 2.0 tailings dam failure, and thus available
for deposition in the mainstem Koktuli, Mulchatna, and Nushagak Rivers could exceed 350 million
metric tons (Table 9-5). In addition, some of the remaining stored tailings material could mobilize as
pore water seeped from the exposed slopes immediately following the failure event, creating slides and
smaller flow events. Fine sediment could also be mobilized during any subsequent precipitation or snow
melt runoff events that would direct water across the tailings and down valley through the breach
before it was repaired. The depth and distribution of fines in the mainstem Koktuli, Mulchatna, and
Nushagak Rivers cannot be estimated at this time, but given the volume and grain size of these
sediments, it is reasonable to expect that continued pulses of fine sediments would be transported
through and transiently stored in these mainstem rivers during spring snow melt and fall rain events for
many years (Knighton 1984).
9.4.2 Exposure-Response
9.4.2.1 Fish
The State of Alaska standard for accumulation of fine sediment (0.1 to 4.0 mm) is "no more than 5%
increase by weight above natural conditions (as shown by a grain size accumulation graph) with a
maximum of 30% fines in waters used by fish for spawning" (ADEC 2011). Bryce et al. (2010) found that
even small amounts of fines (exceeding 5% fines or 13% sands and fines) in streambed sediments were
associated with declines in sediment-sensitive aquatic vertebrates, including salmonids. The tailings
dam failures evaluated here would completely scour and transport or bury existing substrates in the
North Fork Koktuli River valley under tailings fines, greatly exceeding all sediment criteria for salmonid
spawning. Continued erosion and transport of fines deposited on bars, floodplains, and terraces would
provide a chronic source of additional fine sediments during precipitation events, providing new inputs
of fines during spawning and egg incubation. Thus, exceedance of fine sediment standards in the entire
North Fork Koktuli River would be a likely outcome for years to decades.
Interstitial spaces used by juvenile salmonids for overwintering and concealment are a critical habitat
resource, particularly in northern ice-bound rivers and streams (Bustard and Narver 1975, Cunjak 1996,
Huusko et al. 2007, Brown et al. 2011). Interstitial habitat initially would be eliminated by the tailings
dam failure, and then subject to continued high levels of embeddedness as new channels eroded into the
new valley fill composed of tailing fines. The new sediment regime in the North Fork Koktuli River and
associated transport and storage of massive quantities of fine sediments would essentially eliminate
Bristol Bay Assessment 923 January 2014
-------
Chapter 9 Tailings Dam Failure
interstitial habitat for years to decades, if not longer. Altered valley morphology and substrate
composition would also very likely lead to changes in groundwater flowpaths and interactions with
surface waters. Infiltration and burial of coarse valley fill by fine sediments could greatly reduce
hydraulic conductivity and result in decreased rates of exchange between surface water and
groundwater (Hancock 2002). Because of these habitat changes, suitable spawning environments and
overwintering habitats for salmonids would be greatly diminished in this watershed, likely leading to
severe declines in salmonid spawning success and juvenile survival (Wood and Armitage 1997).
9.4.2.2 Invertebrates
Aquatic macroinvertebrates are an important food source for Chinook and coho salmon, rainbow trout,
Dolly Varden, Arctic grayling, and other fishes that rear in streams of the mine scenario watersheds
(Nielsen 1992, Scheuerell et al. 2007). Two available data sources describe the existing
macroinvertebrate communities for streams in the study area: PLP (2011: Chapter 15.2) and Bogan et al.
(2012). Both documents describe broadly similar communities that are consistent with those reported
from other regions of Alaska (Oswood 1989). Communities are reasonably diverse, with Bogan et al.
(2012) reporting 137 taxa from 38 families, with 9 to 40 taxa occurring at a given site (Chironomidae
were lumped at the family level). Communities are dominated by Diptera (true flies), primarily
Chironomidae (midges), with lesser numbers of Ephemeroptera (mayflies) and Plecoptera (stoneflies)
and relatively few Trichoptera (caddisflies). Macroinvertebrate densities were characteristically
variable, ranging two orders of magnitude (102 to 11,371 organisms per m2) (Bogan et al. 2012).
In addition to the direct impacts on fish described in Section 9.4.2.1, catastrophic sedimentation
associated with tailings dam failure also would likely affect fish populations through habitat-related
reductions in macroinvertebrate food resources (see Section 9.5 for discussion of toxicity-related
effects). Sedimentation can affect benthic macroinvertebrates through abrasion, burial, and reduction of
living space, oxygen supply, and food availability (Jones et al. 2011). Deleterious effects of sedimentation
have been reviewed thoroughly (Wood and Armitage 1997, Jones et al. 2011). Sedimentation typically
leads to reductions in density and taxonomic diversity (Wagener and LaPerriere 1985, Gulp et al. 1986,
Quinn et al. 1992, Milner and Piorkowski 2004), even at sediment loads substantially lower than those
modeled in the tailings dam failure scenarios (Wood and Armitage 1997, Jones et al. 2011). The
conversion of a stable streambed dominated by gravel and cobble to a highly unstable one composed
entirely of fine sediments, as would occur in the tailings dam failures considered here, would certainly
lead to reductions in the biomass and diversity of macroinvertebrate prey available to fish populations.
9.4.3 Risk Characterization
The complete loss of suitable salmonid habitat in the North Fork Koktuli River in the short term (less
than 10 years), along with the likelihood of very low-quality spawning and rearing habitat in the long
term (decades), would likely result in near-complete loss of North Fork Koktuli River fish populations
downstream of the tailings dam. These impacts would persist for multiple salmon life cycles, so salmon
cohorts that are at sea during the tailings dam failure would eventually return to find degraded
spawning and rearing habitat. The North Fork Koktuli River provides complex, low-gradient, high-
Bristol Bay Assessment 924 January 2014
-------
Chapter 9 Tailings Dam Failure
quality habitats that currently support spawning and rearing populations of sockeye, Chinook, and coho
salmon, and spawning populations of chum salmon (Figures 7-2 through 7-5) (Johnson and Blanche
2012). For example, aerial index surveys in the North Fork Koktuli River documented roughly 3,000
Chinook salmon (surveyed in 2005), 2,100 sockeye salmon (surveyed in 2004), 1,750 coho salmon
(surveyed in 2008), and 1,400 chum salmon (surveyed in 2008) (values inferred from figures in report
by PLP [2011: Chapter 15]). The North Fork Koktuli River also supports rearing Dolly Varden and
rainbow trout (Figures 7-7 and 7-8) (ADF&G 2012). The Koktuli River watershed has been recognized as
an important producer of Chinook salmon for the greater Nushagak River Management Zone (Dye and
Schwanke 2009). Total Chinook salmon run-size estimates for the Nushagak River include estimates of
harvest plus escapement of spawners. Estimates based on a variety of techniques, including sonar,
averaged over 190,000 Chinook salmon from 2002 through 2011 (Buck et al. 2012), making the
Nushagak the largest producer of Chinook salmon for the Bristol Bay region. Of all the Chinook salmon
tallied during annual aerial index counts in the Nushagak River watershed between 1969 and 1985
(years that all reported spawning areas were surveyed), on average 29% (range 21 to 37%) were
counted in the Koktuli River system (Dye and Schwanke 2009) (see Section 7.1.2 for a discussion of the
limitations of abundance estimates based on aerial counts). The Mulchatna River accounts for another
12% (range 9 to 17%) of the Nushagak Chinook salmon count, and the Stuyahok River (which drains to
the Mulchatna downstream of the Koktuli River) represents another 18% (range 10 to 27%). Hence,
Chinook salmon production could be significantly degraded by loss of habitat downstream of the tailings
dam, particularly if effects extended downstream into the Koktuli and Mulchatna Rivers and beyond.
Sockeye are the most abundant salmon returning to the Nushagak River watershed, with annual runs
averaging more than 1.9 million fish between 2001 and 2010 (Jones et al. 2012). Spatially extensive
sockeye salmon spawner data are not available for the Nushagak River watershed, so it is impossible to
estimate what proportion of the population spawns in the Koktuli River system. The Nushagak River
watershed supports two genetically and ecologically distinct groups of sockeye salmon (Dann et al.
2011): those that rear in, and spawn in and near lakes (lake-type, as in Semko 1954), and those that
spawn and rear, at least briefly, in rivers and streams (sea-type and river-type, as in Semko 1954,
collectively called riverine-type here). Sockeye salmon in much of the Mulchatna River system, including
the Koktuli River and adjacent Stuyahok River, are riverine-type, and are more closely related to
riverine-type sockeye salmon of the Kuskokwim River drainage than to Nushagak River watershed lake-
type sockeye salmon (Dann et al. 2011). It is likely that these population groups share a similar life
history pattern. Riverine-type sockeye in Kuskokwim River tributaries preferentially rear in off- and
side-channel habitats within floodplain-prone stream reaches (Ruggerone et al. 2011). From 1995 to
2006, an estimated 528,000 adult sockeye salmon annually migrated to spawning areas in the Nushagak
and Mulchatna River systems upstream of the Wood River system (Jones et al. 2012). Of these,
approximately 70% (an annual average of 363,000) appear to be riverine-type sockeye salmon based on
the proportion of sockeye that escaped to the Nushagak/Mulchatna portion of the basin.
Spawning and rearing riverine-type sockeye salmon habitats occur throughout the South and North
Fork Koktuli Rivers downstream to and beyond the confluence of the Mulchatna and Nushagak Rivers
Bristol Bay Assessment 925 January 2014
-------
Chapter 9 Tailings Dam Failure
(ADF&G 2012). The tailings dam failures considered here would likely affect sockeye salmon production
throughout the Koktuli River system, but the proportion of the total Nushagak River sockeye salmon
production that would potentially be affected is unknown (see Section 7.1 for additional information on
fish abundance).
Populations of resident and anadromous fishes present in North Fork Koktuli River headwaters
upstream of TSF 1 or in tributaries at the time of a tailings dam failure would not immediately suffer
habitat losses, but would suffer indirect effects resulting from alteration of the North Fork Koktuli River
valley. Many species in the region's rivers, including resident non-anadromous species, undergo
extensive seasonal migrations (West et al. 1992). Such movements are important for juveniles moving
from natal areas to overwintering habitats, for smolts emigrating to sea, for adult spawning migrations,
or, in the case of resident species, for migration between spawning, foraging, and overwintering areas.
Sediment deposition at tributary mouths in the North Fork Koktuli River valley could adversely affect
passage of juvenile and adult fish into and out of these tributaries. For several years, access to mainstem
river habitats upon which many tributary fishes depend for portions of their life history could be
blocked or severely degraded.
Successful re-colonization of the North Fork Koktuli River by resident fish would depend on whether
unimpaired tributary habitats or downstream areas could function as suitable refugia and source areas
for re-colonization of the North Fork Koktuli River. Resident fish would require sufficient tributary
habitat to complete their entire life history, as it is likely that downstream habitat would be unusable for
multiple generations. Re-colonization of salmon from tributary refugia or downstream areas would
require suitable passage at tributary junctions and suitable migratory corridors throughout the
mainstem. Aquatic macroinvertebrate food resources would also likely be adversely affected in the main
river channel, limiting rearing potential for insectivorous fishes such as juvenile salmonids. Given
estimates of fine-sediment deposition and the unstable, silt and sand bed channels that would likely
form across the valley floor, as well as metal concentrations in these tailing substrates that could inhibit
migratory behavior (Section 9.5.2.1), successful migratory conditions seem unlikely for at least several
years after a tailings dam failure.
The near-complete loss of North Fork Koktuli River fish populations downstream of the TSF and long-
term transport of fine sediment to downstream locations would have significant adverse effects on the
Koktuli and Nushagak River salmon, Dolly Varden, and rainbow trout populations, affecting downstream
fisheries, including subsistence users (Figure 5-2). Spawning and rearing habitat would be eliminated or
impaired by deposition of transported sediment and/or reductions in the invertebrate prey base. Direct
loss of habitat in the North Fork Koktuli River, and impairments further downstream because of
transport and deposition of sediment, could adversely affect a substantial portion of Chinook salmon
returning to the Nushagak River watershed. Assuming that Alaska Department of Fish and Game
(ADF&G) aerial survey counts reflect the proportional distribution of Chinook salmon in the Nushagak
River watershed, habitat destruction of the North Fork Koktuli River valley, downstream transport of
sediment to the Koktuli River mainstem, and subsequent loss of access to or inhibition of migration into
the South Fork Koktuli River would affect the entire Koktuli River component of the Nushagak Chinook
Bristol Bay Assessment 926 January 2014
-------
Chapter 9 Tailings Dam Failure
run. If the deposited tailings material is of sufficient quantity and toxicity (Section 9.5.2) to have effects
on aquatic life and fish migratory behavior in the lower Koktuli, Mulchatna, and Stuyahok Rivers, much
greater proportions of the Nushagak Chinook populations and other resident and anadromous fish
populations could be affected. Adult salmon returning to these rivers could potentially seek other
tributaries for spawning, but successful recruitment of displaced spawners would require access to and
comparable use of spawning and rearing capacity elsewhere in the Nushagak River watershed.
9.4.4 Uncertainties
It is certain that a tailings dam failure such as those evaluated here would have devastating effects on
aquatic habitat and biota, but the distribution and magnitude of effects is uncertain. Uncertainties
associated with the initial events, including the likelihood of dam failures and sediment transport and
deposition processes are discussed in Sections 9.1.3 and 9.3.3. Uncertainties associated with the timing,
feasibility, and effectiveness of remediation of a tailings spill are discussed in Section 9.6.2. Other
uncertainties related to the time frame for geomorphic recovery, the longitudinal extent and magnitude
of habitat impacts downstream of our modeled 30-km reach of the North Fork Koktuli River, and the fish
populations affected are discussed in this section.
We estimate that recovery of suitable structural habitat in the North Fork Koktuli River and off-channel
areas would likely take years to decades, given the scouring action of the flood wave and the volume of
fine-grain sediment that would potentially be delivered under a tailings dam failure. However, the
period for recovery could be substantially longer. Recovery of suitable gravel substrates and
development of channel morphology suitable for salmon habitat could be delayed even further if the
flood wave were to scour sections of the North Fork Koktuli River valley to bedrock, which would then
be buried under massive deposits of tailings fines. Recruitment of gravels and coarser substrates to the
North Fork Koktuli River valley could be delayed by low supplies and/or low rates of transport from
tributaries or unaffected upstream sources. Recovery may also be delayed if riparian vegetation does
not recover because the tailings are toxic to plants (although this causal pathway is not considered in
this assessment).
The tailings dam failure simulations (Section 9.3) were restricted to approximately 30 km of the North
Fork Koktuli River, from the face of the TSF 1 dam downstream to the confluence of the South and North
Fork Koktuli Rivers. Extension of the simulations beyond this confluence would introduce significant
error and uncertainty associated with the contribution of South Fork Koktuli River flows, and would
require a more sophisticated sediment transport model. As a result, we were unable to quantify
sediment transport and deposition in the mainstem Koktuli, Mulchatna, and Nushagak Rivers. However,
given the high volume of tailings fines that could be transported beyond the confluence of the South and
North Fork Koktuli Rivers (Table 9-5), it is highly likely that impacts on fish habitat estimated for the
North Fork Koktuli River would extend some significant distance down the mainstem Koktuli River.
We estimate that the combined effects of direct habitat losses in the North Fork Koktuli River,
downstream in the mainstem Koktuli River, and beyond, as well as impacts on macroinvertebrate prey
for salmon, could adversely affect 25% or more of Chinook salmon returning to spawn in the Nushagak
Bristol Bay Assessment 927 January 2014
-------
Chapter 9 Tailings Dam Failure
River watershed. If the Koktuli River, Stuyahok, and Mulchatna portion of the Nushagak runs are
impacted via downstream transport of tailings fines, the tailings dam failure may affect nearly 60% of
the Chinook run (on average, 59% of the aerial survey counts were from these three watersheds [range
= 48% to 75%]). Uncertainty around this estimate is associated with the downstream extent of habitat
impacts (described above) and the variable and imprecise estimates of the relative abundance of
Chinook salmon in the Nushagak, Mulchatna, and Koktuli River systems. We based our estimate of
proportions on long-term (1969 to 1970 and 1974 to 1985) aerial counts of Chinook salmon collected
and interpreted by ADF&G (Dye and Schwanke 2009), but aerial counts can substantially underestimate
true abundance (Jones et al. 1998).
Because long-term abundance data are lacking for most other fish species and locations in the mine
scenario watersheds, losses caused by a tailings dam failure are not quantified for other species. Our
analysis focuses on a few endpoint species, and does not incorporate considerations of metacommunity
dynamics, which are poorly understood for the region but may be critical to understanding species
responses to environmental change (Westley et al. 2010). Information documenting known occurrence
of non-endpoint fish species in the region's rivers and major streams is available (Johnson and Blanche
2012, ADF&G 2012), but information on their abundances, productivities, and limiting factors is not
currently available.
9.5 Post-Tailings Spill Water Quality
9.5.1 Suspended Tailings Particles
9.5.1.1 Exposure
During a tailings dam failure, aquatic biota would be exposed to a slurry of suspended tailings moving at
up to 10 m/s (Table 9-4). In the Pebble 2.0 scenario, much of this material would still be flowing 30 km
downstream, at the mouth of the North Fork Koktuli River (the limit of the model) (Table 9-5).
For years after a tailings dam failure, settled tailings would be resuspended and carried downstream. At
first, this process would be frequent if not continuous (except when and where the substrate is frozen),
as channel and floodplain structure is established by erosional processes resuspending the tailings
(Section 9.4.1). Gradually, as the tailings flow downstream, a substrate consisting of gravel and cobble
embedded in tailings fines would become established, and the flow velocities necessary to suspend
sediment would increase until they resembled those of an undisturbed stream.
Studies at other tailings-contaminated sites do not usefully address suspended tailings, as they typically
have been carried out long after the spills occurred, are based on events that differ from the one large
spill that would result from a tailings dam failure, and focus on toxic properties of the tailings
(Section 9.5.2.3).
Bristol Bay Assessment 928 January 2014
-------
Chapter 9 Tailings Dam Failure
9.5.1.2 Exposure-Response
Suspended sediment has a variety of effects on fish that are similar to effects of toxic chemicals. Like
chemical effects, the severity of effects increases with concentration and duration of exposure
(Newcombe and Jensen 1996). At low levels, suspended sediment causes physiological and behavioral
effects; at higher levels it causes death. Salmonids avoid turbid waters when possible, which may result
in loss or underutilization of traditional spawning habitats (Bisson and Bilby 1982, Newcombe and
Jensen 1996). However, salmonids must withstand brief periods of high suspended sediment
concentrations associated with spring floods (Rowe et al. 2003). Empirically derived effective exposures
for lethal and sublethal effects (i.e., reduced abundance or growth or delayed hatching) on juvenile and
adult salmonids may be summarized as follows (derived from Newcombe and Jensen 1996):
• 22,026 mg/L for 1 hour
• 2,981 mg/L for 3 hours
• 1,097 mg/L for 7 hours
• 148 mg/L for 1 to 2 days
• 55 mg/L for 6 days
• 7 mg/L for 2 weeks
• 3 mg/L for 7 weeks to 11 months
However, salmon may adapt to migrate through high levels of suspended sediment. For example, during
mid-May to early August, when adult salmon migrate upstream through the lower Copper River (El
Mejjati et al. 2010), suspended sediment concentrations range from 750 to 1780 mg/L (Brabets 1992).
9.5.1.3 Risk Characterization
During and immediately after a tailings spill, exposure to suspended sediment would be far higher than
any of the effects thresholds listed above. Fish could be literally smothered and buried in the slurry.
Because the standard of 1,000 mg/L of suspended sediment is exceeded by ordinary events such as
erosion of construction sites and tilled fields, erosion of tailings from the re-formation of the channel
and floodplain would likely exceed that standard for days at a time, over a period of years. Fish would be
likely to avoid these streams or experience lethality, reduced growth, or reduced abundance. Avoidance
could also block migrating salmon and other fish from their spawning areas in upstream tributaries
during these periods, although salmon have adapted to migration corridors with high suspended
sediment levels. The potential for tailings to be more aversive or toxic than natural suspended sediment
is unknown. Exposure levels would gradually decline over time as tailings are carried downstream,
channel stability increases, and the floodplain becomes revegetated. Rates of these processes are
unknown, but it is reasonable to assume that decades would be required for suspended sediment loads
in the Koktuli and Mulchatna Rivers to drop to levels that occur with normal high flows in stable
channels of the Bristol Bay watershed.
Bristol Bay Assessment 929 January 2014
-------
Chapter 9 Tailings Dam Failure
9.5.1.4 Uncertainties
There can be little doubt that, during and in the years immediately following a tailings dam failure,
suspended sediment concentrations would be sufficient to reduce fish populations for many kilometers
downstream of a failed tailings dam. A major uncertainty, however, is the number of years required to
reduce suspended sediment concentrations to levels that are not adverse. Another major uncertainty is
the downstream extent of the effects. The data and modeling effort required to determine how far the
initial slurry deposition would extend, how far re-suspended sediments would travel, and how long
erosional processes would continue were not feasible for this assessment.
9.5.2 Tailings Constituents
Although the most dramatic effect of a tailings dam failure would be habitat destruction and
modification due to the flow of tailings slurry downstream, exposures to potentially toxic materials in
the slurry would also occur. The toxic effects of a tailings dam failure can be assessed using the
composition of the tailings and of experimental tailings leachates, as well as experience with tailings
spills at other sites (Box 9-5).
9.5.2.1 Exposure
Aqueous Exposures to Impoundment Waters
During a tailings dam failure, aquatic biota would be exposed to water that had been in contact with
tailings during processing and in the TSF. This water would include pore water associated with the
deposited tailings and water overlying the tailings. If the spill was caused by flow through a fault in the
dam or by a seismically induced tailings dam failure (i.e., a failure under "dry" conditions), undiluted
pore water and supernatant water would be released. If the dam was eroded or overtopped by a
flooding event (i.e., a failure under "wet" conditions), the pore and surface water could be diluted by
fresh water. However, this dilution would be trivial relative to the volume of pore water in the tailings.
A spill would have two phases in our tailings dam failure scenarios; other scenarios could differ in
timing and magnitude. At first, tailings slurry would pour through the breach for approximately 2 hours
based on the specified rates of dam erosion and slurry flow. Pore water would then drain from the
tailings that are not released but are above the elevation of the breach. This latter process would be slow
and could continue until the dam was repaired. If a tailings dam failure occurred after the mine site was
abandoned and no corrective action was taken or was delayed, an equilibrium would be achieved in
which rain, snow, and upstream flows were balanced by outflow of leachate through the breach.
Bristol Bay Assessment 930 January 2014
-------
Chapter 9 Tailings Dam Failure
BOX 9-5. BACKGROUND ON RELEVANT ANALOGOUS TAILINGS SPILL SITES
Past deliberate or accidental spills of metal mine tailings into salmonid streams and rivers have occurred by
mechanisms and mining practices other than those evaluated in this assessment. However, these spills provide
evidence concerning the fate of tailings and the nature of exposures to aquatic biota once the tailings are in
streams and floodplains. In the United States, some of these sites are relatively well studied because their
observed effects have led to classification as Superfund sites. Other tailings spills have caused extensive fish kills
and other significant effects but have not generated useful long-term monitoring data. These brief descriptions
provide background information and support the use of evidence from these cases in analyzing risks from potential
tailings dam failures in the Bristol Bay watershed.
Clark Fork River, Montana. The Clark Fork River Operable Unit of the Milltown Reservoir/Clark Fork River
Superfund Site includes 120 river miles (193 km) extending from the river's headwaters to the Milltown Reservoir
just east of Missoula, Montana. Miningfor gold, silver, copper, lead, and zinc began in the Clark Fork watershed in
the late 1800s. Most of the wastes released were tailings from copper mines in Butte and Anaconda, but aqueous
mine discharges and aerial smelter emissions also contributed wastes. Two sedimentation ponds were constructed
by 1918, with a third constructed by 1959. Mine water treatment was initiated between 1972 and 1975. By the
mid-1970s, waste inputs to the Clark Fork River were largely limited to movement of previously released solids. It
became a Superfund site in 1983. Contaminants of concern were arsenic, cadmium, copper, lead, and zinc, but
copper was the focus of assessment and planning because of its high toxicity.
The primary source of exposure is tailings deposited on the floodplains, resulting in aquatic pollution through
erosion and leaching. Large areas with acidic tailings (both acidic and neutral tailings were deposited) are barren of
plant life due to metal toxicity, which contributes to erosion and leaching. The river was fishless from the late
1800s to the 1950s, but has begun to recover. Trout and other fishes continue to exhibit low growth and
abundance, and intermittent fish kills have followed metal pulses from rainstorms or rapid snow melt. However,
sedimentation was also thought to contribute to effects on fish populations through habitat degradation. Detailed
information can be found in the responsible party's remedial investigation (ARCO 1998) and in U.S. Environmental
Protection Agency (USEPA) documents (USEPA 2012a).
Coeur d'Alene River, Idaho. The Coeur d'Alene River in northern Idaho flows from the Bitterroot Mountains to
Lake Coeur d'Alene. From the late 19th to late 20th century, the upper basin was mined for silver, lead, zinc, and
other metals, and much of the ore was smelted locally. Tailings were dumped into gullies, streams, and the river
until dams and tailings impoundments were built beginning in 1901. Plank tailings dams failed in the 1917 and
1933 floods; direct discharge of tailings did not end until 1968. According to USEPA's remedial investigation,
approximately 56 million metric tons (62 million tons) of tailings were discharged to the Coeur d'Alene River. In
1983, the area of the Bunker Hill smelter was added to the Superfund national priority list, and in 1998 the
contaminated river watershed, Lake Coeur d'Alene, and part of the Spokane River were explicitly included.
Metals concentrations above ambient water quality criteria, lethality in tests of ambient waters, and the absence of
some fish species from reaches with high metal concentrations were all attributed to leachates from tailings and
other mine wastes in floodplains and tributary watersheds. In addition, toxicity of bed sediments, which include
tailings, was found in the Coeur d'Alene and Spokane Rivers and their tributaries. Aquatic effects were attributed
primarily to zinc, but cadmium, lead, and copper also reached toxic levels.
More detailed background information can be found in the Ecological Risk Assessment for the Coeur d'Alene Basin
Remedial Investigation/Feasibility Study (USEPA 2001), other USEPA documents (USEPA 2012b) and the National
Research Council's review of USEPA's assessment and management documents (NRC 2005).
Soda Butte Creek, Montana and Wyoming. The headwaters of Soda Butte Creek drain the New World mining
district in Montana before entering Yellowstone National Park. From 1870 to 1953, porphyry deposits were mined
for gold and copper with some arsenic, lead, silver, and zinc. In June 1950, the earthen tailings dam at the
McLaren mine failed, releasing approximately 41 million m3 of water and an unknown mass of tailings into Soda
Butte Creek (Marcus et al. 2011). In 1969, the creek was rerouted around the tailings pile, which was covered and
seeded. In 1989, a Superfund emergency response re-created and riprapped the creek channel to accommodate a
100-year flood. Despite these actions, metal levels remain high in the creek and floodplain sediments and the biota
are impaired. The lack of any decrease in sediment copper despite floods in 1995, 1996, and 1997 and the lack of
macroinvertebrate recovery following remediation of acid drainage in 1992 indicate that the tailings are persistent
and the primary cause of biological impairments. The primary sources of information on effects of the tailings spill
are academic studies (Nimmo et al. 1998, Marcus et al. 2001).
Bristol Bay Assessment 931 January 2014
-------
Chapter 9 Tailings Dam Failure
Once in the stream, potentially toxic constituents dissolved in the water would not settle out. Because
the potentially toxic constituents are not degradable or volatile, they would eventually flow to Bristol
Bay, although they would be diluted along the way. In the Pebble 2.0 tailings dam failure, the peak flow
of mobilized tailings at the confluence of the North and South Fork Koktuli Rivers is estimated to be
approximately 18,600 m3/s (Table 9-4). The Nushagak River at Ekwok (Figure 2-4) would be the first
downstream gaging station at which most of the tailings would have settled out and dilution could be
estimated. Using the annual average and highest monthly average flows (668 and 1,215 m3/s,
respectively), concentrations of dissolved chemicals in the Nushagak River would be 96 and 94% of
those in the spill. Minimum flow is not considered, because a failure is believed to be less likely during
freezing conditions.
We used the tailings humidity cell test results to estimate the composition of the bulk of the aqueous
phase. However, those values are uncertain, because none of the tests performed by PLP represent the
leaching conditions in a tailings impoundment, material other than bulk tailings would be added to the
TSFs, and no model exists to mathematically simulate the leaching process. However, some mixture of
tailings supernatant, which represents the source water for the impoundment (Table 8-4); humidity cell
leachate, which represents aqueous leaching from tailings under oxidizing conditions (Table 8-5); and a
small amount of local water (Table 8-10) can be used to approximate aqueous phase composition.
During mine operation, tailings impoundment surface waters would consist of water used to transport
the tailings (supernatant) and any other waters stored in the impoundments prior to reuse or treatment
and discharge. Hence, the surface water is expected to resemble the PLP's test supernatant (Table 8-4)
with some dilution by precipitation. However, those results do not include process chemicals (e.g.,
xanthates and cyanide) that may be associated with the supernatant but that are not quantified in this
assessment. Supernatant water would be slightly diluted by rain and snow onto the surface of the
impoundment, but peripheral berms should generally prevent dilution by runoff.
The waters released from a tailings spill during mine operation could consist of surface water, surficial
pore water, and a much larger volume of deep pore water. The surficial tailings pore water would be
generated by leaching tailings in the presence of some oxygen. The composition and concentrations of
constituents in that water may be roughly similar to a mixture of those observed in the supernatant and
humidity cell tests (Tables 8-4 and 8-5). Pore water from deep anoxic tailings would have begun
primarily as supernatant, but may have lower metal content due to chemical precipitation under anoxic
conditions. Leachate flowing from an abandoned and failed impoundment would be more oxidized
because the cover water and much of the pore water would have drained away.
Aqueous Exposures from Deposited Tailings
After a tailings dam failure, aquatic biota would be exposed to potentially toxic tailings that covered the
substrate of streams and rivers. Benthic organisms, or aquatic insects and other invertebrates that
burrow into or crawl upon substrates, would be most exposed. Eggs and larvae (fry) of any salmon,
trout, or char that spawned in the contaminated substrate also would be exposed. In either case, the
bioavailable contaminants would be those that are dissolved in the pore water of the deposited tailings.
Bristol Bay Assessment 932 January 2014
-------
Chapter 9 Tailings Dam Failure
Hence, exposure is determined by the rate of leaching of the tailings and the rate of dilution of the
leachate, which depend on hydrological conditions. Unlike the lakes and estuaries that are the usual
sites of sediment pollution studies, streams have a high level of interaction between substrates and
surface waters. Shallow, turbulent water is typically near oxygen saturation. Bedload sediment bounces
and slides downstream during high flows. At high enough flows sediment is suspended, exposing it to
oxygen. Water also flows longitudinally and laterally through bed and floodplain sediments and
vertically between groundwater and surface water.
Because the biologically active zone is oxidized, the tailings leachate to which biota would be exposed
could resemble leachates from the supernatants and humidity cells (Tables 8-4 and 8-5). Ideally, a
leaching test would be performed that simulated conditions in a streambed, but no such test results are
available. In theory, leachate composition could be estimated using a mechanistic model, but no such
model is available. Dilution of the leachate would be minimal in low-flow areas such as pools and
backwaters and during low-flow periods. Dilution would be greatest in high-flow and turbulent
locations such as riffles, in groundwater upwelling or downwelling areas, and during high-flow periods
such as spring runoff and floods. However, high flows would be expected to increase leaching rates,
resulting in complex dynamics (Nagorski et al. 2003).
Although we assume that spilled tailings would be mixed and would have average metal compositions
(Table 9-7), stream processes would be expected to sort them. In Soda Butte Creek (Box 9-5), copper
concentrations in riffles and glides gradually decreased downstream from the tailings spill site.
However, fine sediments in pools had higher copper concentrations than the high-energy segments, and
some of the highest copper concentrations were found in fine pool sediments more than 10 km
downstream (Nimmo etal. 1998).
After a spill, aquatic biota would also be indirectly exposed to tailings deposited on land, primarily in the
floodplains. Erosion of these floodplain-deposited tailings would result in additional deposition in
streams, potentially replacing tailings lost through streambed erosion (Marcus et al. 2001). In addition,
rain and snowmelt would run across and percolate through tailings deposited on floodplains, leaching
metals and carrying them into the stream. Leachate would also form during lateral groundwater
movement through tailings, particularly where tailings deposited in wetlands. Floodplain-deposited
tailings are leached in the presence of oxygen, with episodes of saturation and drainage (ARCO 1998).
Hence, humidity cell leachates would be more relevant to this exposure route than to others, and
leachate concentrations in Table 8-5 may roughly estimate leachate composition from floodplain-
deposited tailings.
Bristol Bay Assessment 933 January 2014
-------
Chapter 9
Tailings Dam Failure
Table 9-7. Comparison of average metal concentrations of tailings (Appendix H) to threshold effect
concentration and probable effect concentration values for freshwater sediments and sums of the
quotients (£ TU). Values are in mg/kg dry weight.
Tailings Constituents
Ag
As
Ba
Be
Bi
Cd
Co
Cr
Cu
Hg
Mn
Mo
Ni
Pb
Sb
Se
Tl
U
V
Zn
Sum
Average
0.7
25
30
0.3
0.6
0.1
8.1
150
680
0.1
360
52
68
15
1.0
1.8
0.3
0.4
87
87
-
TEC"
9.8
-
-
0.99
-
43
32
0.18
630
23
36
-
120
-
TEC Quotient
2.6
-
-
0.10
-
3.5
21
0.56
0.57
2.9
0.41
-
0.72
32
PEC"
33
-
-
5.0
-
110
150
1.1
1200
49
130
-
460
-
PEC Quotient
0.76
-
-
0.02
-
1.3
4.5
0.09
0.30
1.4
0.12
-
0.19
8.7
Notes:
Dashes (-) indicate that criteria are not available.
"" TECs and PECs are consensus values from (MacDonald et al. 2000), except for Mn which are the TEL and PEL for Hyalella azteca 28-day
tests from (Ingersoll et al. 1996).
TEC = threshold effect concentration; PEC = probable effect concentration; TEL = threshold effect level; PEL = probable effect level.
This leachate could have three fates: it could move upward during dry periods and deposit on the
surface as soluble salts (e.g., hydrated metal sulfates); it could move down into buried soils and deposit
as weak acid-extractable compounds (e.g., metal sulfides); or it could sorb to organic matter or move
laterally to the surface channel as dissolved metal ions (Nimik and Moore 1991, ARCO 1998). Runoff
from tailings-contaminated floodplains of the Clark Fork River had high copper levels (67.8 to 8,380
ug/L) (Nimik and Moore 1991, ARCO 1998). Concentrations from spills in the Bristol Bay watershed
would probably be lower than for the acidic Clark Fork tailings and salt accumulation on the surface
would be less because of greater precipitation, but the same processes would occur. Dilution of leachate
that moves into the stream would be highly location- and condition-specific. Once in a stream, leached
metals are likely to remain dissolved because of the highly dilute water chemistry in the region, but
some precipitation or sorption to clays or organic matter would occur, depending on the conditions that
moved the leachate into the stream.
Bristol Bay Assessment
9-34
January 2014
-------
Chapter 9 Tailings Dam Failure
Remobilization of deposited tailings during high flows could result in acute exposures to suspended
tailings and extend the downstream range of exposure to deposited tailings. In the Coeur d'Alene River,
floods occurring in 1995,1996, and 1997—more than 30 years after the last release of tailings—carried
metal-enriched sediment from both the floodplain and streambed more than 210 km downstream (the
furthest extent of the study) (USGS 2005).
Less dramatic increases in flow would cause bedload transport (movement of sediment without
suspension in the water column), which could release sediment pore water (leachate) into the water
column. First, copper could leach from the tailings and accumulate in sediment pore water during low-
flow periods. Then, when flows increase sufficiently to mobilize the sediment, pore water would mix
with surface water, resulting in exposure of aquatic biota and downstream copper transport. Studies in
the tailings-contaminated Clark Fork River found that copper concentrations in interstitial water were
3 to 36 ug/L in depositional areas and 3 to 22 ug/L in riffles (ARCO 1998). Concentrations would differ
for tailings from the Bristol Bay watershed, but this result demonstrates that deposited tailings can have
significant interstitial water concentrations, even in a hydrologically active stream where leaching has
proceeded for decades. If sediment movement was sufficient to mobilize deep anoxic sediments,
precipitated or complexed metals could be mobilized and, depending on local water chemistry,
dissolved.
Solid Phase Exposure to Deposited Tailings
Although the most bioavailable metals in sediment are those dissolved in pore water, it is useful to
consider the whole sediment as a source of exposure. This approach avoids uncertainties associated
with using leaching tests to represent field processes. It is reasonable to consider the average tailings
composition to represent stream sediment to which biota downstream of a spill would be exposed
(Table 9-7). During and after a tailings spill, there may be some sorting of the tailings by size or density
that would result in locally higher metal compositions, but this variability cannot currently be
quantified. Although the material in the failed dam would dilute the tailings initially, particles in the dam
would be larger than the tailings and would settle out in the first few kilometers downstream. Some soil
would be scoured from the receiving stream, but that would be associated with the first wave of the
slurry. Hence, given the volume spilled, tailings in most of the initial depositional area would be
effectively undiluted. After the spill, the tailings sediment would be diluted by clean sediment from
tributaries, but that process would be slow because the volume of tailings deposited in the watershed
would be so large and the watershed is nearly undisturbed except for potential mine facilities. The
background sediment load (1.4 to 2.5 mg/L total suspended solids, Table 8-10) is miniscule compared to
the meters of tailings that would be deposited (Table 9-5). The washing of tailings from floodplains into
streams and rivers would be more important for many years, so the sediments in streams and rivers
below a tailings spill would resemble average tailings.
Dietary Exposures
As discussed in Section 8.2.2, dietary exposures offish to metals have been an issue of concern at mine
sites. An adjustment factor for rainbow trout to account for a dietary component to aqueous exposures
Bristol Bay Assessment 935 January 2014
-------
Chapter 9 Tailings Dam Failure
(0.95) is presented in that section. It may be applied to cases in which both direct aqueous and dietary
exposures may occur, such as flow into a stream through floodplain tailings or from upwelling through
tailings. Dietary exposures with respect to sediment levels may also be estimated. In such cases, direct
aqueous exposures of fish may be negligible, but invertebrates, particularly metal-tolerant insects such
as chironomids, may accumulate metals, carry them out of the sediment, and then serve as sources of
dietary exposure. This phenomenon has been documented in both the Clark Fork and Coeur d'Alene
River basins (Kemble et al. 1994, Farag et al. 1999).
A review of metal bioaccumulation by freshwater invertebrates (mostly Ephemeroptera and Diptera)
derived models for two relevant feeding guilds:
Collector-Gatherer Copper = 0.294 x
Scraper-Grazer Copper = 1.73 x
where x is sediment concentration and copper is tissue concentration, both in ug/g dry weight
(Goodyear and McNeill 1999). Studies of the Soda Butte Creek tailings spill found that copper
concentrations in mixed invertebrates were slightly lower than sediment concentrations (Marcus et al.
2001). Studies of the Clark Fork River give bioaccumulation factors for copper and river invertebrates
ranging from 0.18 to 1.62, with factors generally rising as sediment concentrations declined (calculated
from Brumbaugh et al. 1994, Ingersoll et al. 1994). Equivalent studies in the Coeur d'Alene River give
very similar factors (0.15 to 1.77) (calculated from Farag et al. 1998). These results support the use of an
average bioaccumulation factor of 1.0 (Goodyear and McNeill 1999). This implies copper concentrations
in invertebrates are equal to those in sediments, which in this case are tailings with an average copper
concentration of 680 mg/kg (Table 9-7).
Another method used to estimate the bioaccumulation and toxicity of divalent metals in sediment is the
acid volatile sulfides (AVS)/simultaneously extracted metals (SEM) approach (Ankley et al. 1996).
However, this method requires measurements of AVS and SEM within the sediment of concern. The
source of copper in the tailings is sulfide ores, so one might assume that there is adequate sulfide for the
copper, but experience with tailings spills refutes that assumption. The availability process of concern is
oxidation of the sulfides, not binding of added copper by sediment sulfides. Studies in the Clark Fork
River found that, contrary to expectations of that model, invertebrates accumulated metals at locations
with AVS greater than SEM (Ingersoll et al. 1994). This discrepancy may be due to spatial variability,
high oxidizing conditions in riffles where most invertebrates are found, and the fact that much of the
metals in these sediments are in a form (metal sulfide particles of the tailings) that is very different from
the lake and estuary sediments for which the model was developed. Hence, for practical and empirical
reasons, the AVS/SEM model is not appropriate to estimate bioaccumulation or toxicity in this system.
Persistence of Exposures
Evidence that tailings persist in streams as sources of metal exposures is provided by prior tailings
releases. A review by Miller (1997) found persistence of high metal content sediment in streams after
10 to 100 years. One well-documented case is provided by a tailings dam failure in Soda Butte Creek,
Bristol Bay Assessment 936 January 2014
-------
Chapter 9 Tailings Dam Failure
Montana, in 1950 (Box 9-5) (Marcus et al. 2001). Sediment was still characterized by high copper
concentrations after 48 years despite two 100-year floods, indicating that some tailings were retained
by streams and maintained high metal levels even after decades of leaching. Similarly, the Coeur d'Alene
River basin was contaminated by direct discharge of tailings to floodplains, tailings dam failures, and
mine drainage, which caused extensive damage to the watershed (Box 9-5) (NRC 2005). Treatment of
the mine drainage improved biotic communities, but they were still impaired, apparently as a result of
metals leaching from deposited tailings that entered the river until 1968 (Holland et al. 1994, NRC
2005). At least as late as 2000, metal (cadmium, lead, and zinc) concentrations were elevated in
caddisflies and were more highly correlated with sediment concentrations than with surface water
concentrations, suggesting that deposited tailings were the primary source of exposure (Maret et al.
2003).
A new study has modeled future decline in sediment metals concentrations for the Clark Fork River
(Box 9-5), assuming an exponential decay in concentrations over time due to loss and dilution (Moore
and Langner 2012). Although there was no significant decline over time (1991 to 2009) in downstream
concentrations (which one would expect as tailings wash downstream), concentrations did decline over
time at three individual sites. Based on regression for each of those sites, Moore and Langner (2012)
estimated that average copper concentrations would decline below the probable effect concentration
(PEC) in less than 85 years. At the most contaminated of the three sites, copper was predicted to reach
the threshold effect concentration (TEC) in 163 years. In the Bristol Bay watershed, dilution with clean
sediment would likely be slowed by denser vegetation and less land disturbance. Lower gradients in the
Bristol Bay receiving streams relative to Montana would also tend to slow recovery, as recovery is
primarily achieved by tailings transport downstream. It also should be noted that these time estimates
are not from the date of a spill, but rather from a date decades later, when channel structure had
stabilized and much of the tailings had been carried downstream.
9.5.2.2 Exposure-Response
Aqueous Chemicals
The toxic effects of exposure to a tailings spill can be estimated from aquatic toxicity data. Ambient
water quality criteria and equivalent benchmarks are used to screen the metals in the two types of
tailings leachates (Tables 8-4 and 8-5). Copper is the dominant contaminant in tailings leachates, and
criteria values based on the biotic ligand model (BLM) (described in Section 8.2.2.1) are used as
benchmarks (Table 9-8). Acutely lethal levels for rainbow trout exposed to the humidity cell leachate
and supernatant are estimated to be 93 and 188 ug/L, respectively, based on the BLM.
Bristol Bay Assessment 937 January 2014
-------
Chapter 9
Tailings Dam Failure
Table 9-8. Results of applying the biotic ligand model to mean water chemistries of tailings
leachates and supernatants to derive effluent-specific copper criteria.
Stream
Acute Criterion
(CMC in
Chronic Criterion
(CCC in
Tailings humidity cell leachates
4.8
3.0
Tailings supernatants
7.2
4.5
Notes:
CMC = criterion maximum concentration; CCC = criterion continuous concentration.
Source: USEPA 2007.
Note that these criteria are calculated for the water chemistry of the tailings supernatant and leachate.
This is clearly appropriate for the acute exposures immediately following a tailings dam failure, when
the slurry volume would greatly exceed natural flows. It would also be appropriate for situations like
sediment pore water, where dilution is minimal. However, these criteria would be too high for situations
in which significant dilution occurs, because uncontaminated water has lower hardness and lower levels
of chemicals that reduce copper binding to biotic ligands. Hence, dilution by a factor of 2 would not quite
reduce toxicity of metal-contaminated water by a factor of 2.
Sediment Chemicals
The toxicity of settled tailings may also be estimated from tailings metal concentrations. Various
approaches have been employed to derive sediment quality guidelines, but the most common are the
threshold effect level (TEL) and the probable effect level (PEL). TELs and PELs have been used in
assessments of sites contaminated by mine wastes (USEPA 2001, USGS 2004, 2007). These levels are
derived from distributions of sediment concentrations that do or do not exhibit apparent toxicity in
laboratory or field studies. MacDonald et al. (2000) performed a meta-analysis of published values,
proposed consensus TECs and PECs, and then tested them using additional sediment studies. One of the
sites in the test data set was the tailings-contaminated Clark Fork River. Out of 347 total sediments from
17 rivers and lakes, that validation study found toxic effects in 17.7% of sediments with copper
concentrations less than the TEC, in 64% of sediments with copper concentrations between the TEC and
PEC, and in 91.8% of sediments with copper concentrations above the PEC (MacDonald et al. 2000). The
consensus TECs and PECs are used to evaluate the potential toxicity of tailings should they become
sediment following a spill, because they have the best scientific support.
Dietary Chemicals
Effects may also be estimated from dietary exposures. If the primary source of exposure is dissolved
copper in the water column (e.g., if significant upstream and floodplain leaching occurs), then the
0.95 adjustment factor (Section 8.2.2.1) is applicable. However, if sediment is the primary source of
exposure, a dietary value is needed for consumption of benthic invertebrates. A dietary chronic value for
rainbow trout derived from multiple studies is 646 ug/g (micrograms of copper per gram of dry diet)
(Borgmann et al. 2005), at which concentration survival and growth are observed to decline in multiple
studies.
Bristol Bay Assessment
9-38
January 2014
-------
Chapter 9 Tailings Dam Failure
Analogous Sites
The effects of exposure to leachate from tailings can also be estimated from effects at analogous sites. In
the Clark Fork River, Coeur d'Alene River, and Soda Butte Creek (Box 9-5), both toxicity and observed
field effects on fish and invertebrates have been associated with deposited tailings. However, the
magnitude and nature of effects are so site-specific that quantitative empirical exposure-response
models from these sites cannot reasonably be applied to the tailings dam failures analyzed here.
Nevertheless, the qualitative relationships are applicable.
9.5.2.3 Risk Characterization
Toxicological risks are usually judged by comparing exposure levels to a criterion or other
ecotoxicological benchmark using a risk quotient (Box 8-3), which equals the exposure level divided by
the ecotoxicological benchmark. If the quotient exceeds 1, the effect implied by the benchmark is
expected to occur, but with some uncertainty (Box 8-3). Quotients much larger than 1 suggest larger
effects than those that define the benchmark, with greater confidence that an adverse effect would
occur. Quotients much smaller than 1 suggest that even small effects are unlikely. The criterion
maximum concentration (CMC or acute criterion) and criterion continuous concentration (CCC or
chronic criterion), and equivalent numbers when criteria are not available, are the primary
ecotoxicological benchmarks used in this assessment for aqueous exposures, because they are relatively
well-accepted as thresholds for significant effects. The CMC estimates a concentration at which 5% of
aquatic species experience some mortality among developed life stages in short-term exposures. The
CCC estimates a concentration at which 5% of aquatic species experience decreased survival, growth, or
reproduction in longer-term exposures. Other benchmarks are used to indicate direct toxicity to
salmonids (Tables 8-13 and 8-14).
Acute Toxic Risks
At sites closest to a failed TSF, acutely toxic effects of a tailings spill would be indistinguishable from the
concurrent effects of being smothered by tailings particles. Aquatic life in the range of the tailings slurry
would be devastated by its physical effects. Dissolved components of the spill would continue to flow to
Bristol Bay, beyond the extent of significant particle deposition. Undiluted leachates of both types would
be expected to exceed the acute national criterion for copper (Tables 8-4 and 8-5), which suggests that
they would kill invertebrates. However, even the minimal dilution in the Nushagak River at Ekwok
would dilute leachate from the maximum spill to the national criterion or below. Even copper in
undiluted tailings leachates (5.3 and 7.8 ug/L for the humidity cell and supernatant, respectively) would
be well below levels required to kill post-larval trout in acute exposures (93 and 188 ug/L for the
humidity cell and supernatant, respectively). Hence, in a tailings dam failure, acute exposure to
dissolved copper immediately downstream of the TSF would be sufficient to kill sensitive invertebrates
but not salmonids, but those effects would be eclipsed by the physical effects. Downstream, where
physical effects would be minimal, toxic effects would be reduced or eliminated by dilution.
Bristol Bay Assessment 939 January 2014
-------
Chapter 9 Tailings Dam Failure
Chronic Toxic Risks for Aqueous Exposure
Potential effects of chemicals leaching from tailings in streambed and riverbed sediments and associated
floodplains are addressed by dividing leachate concentrations by chronic water quality criteria and
standards to derive hazard quotients (exposure concentrations divided by effects concentrations).
These hazard quotients can be interpreted as relative degrees of toxicity of leachate constituents or as
indicators of the degree of dilution required to avoid significant toxic effects. The two estimates of
tailings leachate composition give similar results (Tables 8-4 and 8-5). Undiluted leachate of both types
would be expected to exceed the chronic national criterion, but not the Alaskan standard, for copper. If
combined toxic effects of metals are considered (see the Sum of Metals line in Tables 8-4 and 8-5),
chronic toxicity would be expected with both the hardness-based and BLM-based copper criteria, and
acute lethality would be expected with the BLM-based copper criterion. However, direct aqueous
exposures offish to copper are unlikely to be toxic unless concentrations in the actual field leachates are
much higher than the tailings test leachate concentrations.
The quotients with respect to chronic criteria (CCC) imply that dilution by a factor of 2 to 4 would be
sufficient to render leachate nontoxic. Low dilutions would be expected in the years immediately after a
spill, when flows would pass through large volumes of tailings. After tailings have eroded and a more
normal channel and floodplain are established, low dilution of tailings could occur in sediments during
normal flows and in locations where water contaminated by floodplain tailings feeds a stream. In those
situations, sensitive invertebrates could be reduced or eliminated.
Chronic Toxic Risks from Sediment Chemicals
Sediment quality guidelines provide another line of evidence to assess risks from tailings after a tailings
dam failure. Table 9-7 shows that tailings would be expected to cause severe toxic effects on the
organisms that live in or on them. Notably, copper concentration would be approximately 4.5 times the
PEC; chromium and nickel concentrations would also exceed their PECs. The sum of TEC quotients of 32
implies that tailings would need to be diluted by 32 parts clean sediment to one part tailings before toxic
effects would be unlikely (below the TEC). Because the Bristol Bay watershed is relatively undisturbed,
background levels of total suspended solids are low (Table 8-10) and the time required to achieve that
degree of dilution would be very long.
Chronic Toxic Risks from Dietary Chemicals
The most relevant estimate of fish dietary exposure to tailings is provided by bioaccumulation factors
with respect to sediment. The best estimate bioaccumulation factor of 1 implies copper concentrations
in invertebrates of 683 mg/kg. Dividing this concentration by a consensus dietary chronic value for
rainbow trout of 646 ug/g (micrograms of copper per gram of dry diet) (Borgmann et al. 2005) results
in a quotient of 1.1. This implies that the undiluted tailings would produce toxic prey for fish, but the
result is marginal and certainly within the range of uncertainty.
Bristol Bay Assessment 940 January 2014
-------
Chapter 9 Tailings Dam Failure
Chronic Toxic Risks—Analogous Sites
Some well-documented cases indicate that adverse effects of chronic toxicity on aquatic communities in
general, and on salmonids in particular, can occur in streams and rivers that receive tailings spills. These
cases have shown that effects continue indefinitely, but that the nature and magnitude of those effects
vary among sites. In every case in which the ecological consequences of a major metal mine tailings spill
to a stream or river were studied, extensive and long-lasting toxic effects were observed. However, these
tailings are likely to be more toxic than future tailings due to more efficient metal extraction in modern
ore processing.
The most relevant case appears to be Soda Butte Creek in Montana, where a tailings spill from a
porphyry gold and copper mine occurred in 1950 (Box 9-5). In the Soda Butte Creek case, the copper
content of macroinvertebrates was positively correlated (r2 = 0.80) and their taxa richness was
inversely correlated (r2 = 0.48) with sediment copper (Marcus et al. 2001). Although copper
concentrations generally decreased downstream, sediments and sediment pore waters were toxic to the
amphipod Hyalella azteca for the full 28-km length of the study area (Nimmo et al. 1998).
Macroinvertebrate community effects persisted for at least 40 years after the spill. These effects were
attributed to sediment toxicity (Nimmo et al. 1998), but habitat effects of deposited tailings also may
have contributed. Although they were less well studied, it was clear that trout were also affected: only
two trout were found in the 300-m reach downstream of the spill site in 1993, although prior to mining
Soda Butte Creek was known for "fast fishing and large trout" (Nimmo et al. 1998).
In the Coeur d'Alene River and its tributaries, elevated metals concentrations and effects on both benthic
invertebrates and fish persisted for more than 30 years after tailings releases ended and after treatment
of mine drainage. Some fish species were absent; others were reduced in abundance and experienced
toxic effects from both aqueous and dietary exposures (Farag et al. 1999, Maret and MacCoy 2002, Maret
et al. 2003). Returning Chinook salmon avoided the more contaminated South Fork in favor of the North
Fork (Goldstein et al. 1999), and macroinvertebrate communities and taxa were also impaired (Holland
et al. 1994, Maret et al. 2003).
In the Clark Fork River, a sediment quality triad approach demonstrated that tailings-containing
sediments had high metal levels, were toxic to the amphipod Hyalella azteca, and shifted the
macroinvertebrate community to generally metal-tolerant Oligochaeta (worms) and Chironomidae
(midges) (Canfield et al. 1994). Rainbow and brown trout abundances were low in contaminated
reaches of the Clark Fork, fish kills occurred apparently due to metals washing from floodplain tailings
deposits, and metals in invertebrates were sufficient to cause toxic effects in laboratory tests of trout
(Kemble et al. 1994, Pascoe et al. 1994, ARCO 1998).
9.5.2.4 Uncertainties
All of the lines of evidence concerning risks to aquatic communities from the toxic properties of spilled
tailings have notable uncertainties. In particular, the estimates from test leachates and whole test
tailings underestimate risks because they do not include pyritic tailings.
Bristol Bay Assessment 941 January 2014
-------
Chapter 9 Tailings Dam Failure
Toxic Risks from Aqueous Exposures
The use of leachate and supernatant concentrations to estimate effects of a tailings spill is uncertain
primarily because of issues concerning test relevance to leaching in the field. Leaching of tailings in the
impoundment, streambeds, and floodplains would occur under very different conditions than in
humidity cell tests. In addition, it is possible that tailings could become more acidic over time, as their
acid-neutralizing capacity is consumed or as acid-neutralizing chemicals are dissolved, resulting in
increased metal concentrations. Test leachates are available for bulk tailings but not pyritic tailings.
Finally, the degree of leachate dilution in the field would be highly variable and could be roughly
estimated, at best. These uncertainties concerning exposure are significant in terms of both their
potential size (at least an order of magnitude uncertainty) and in terms of their implications (leachates
from the spilled tailings may be non-toxic or severely toxic given this uncertainty in exposure).
The exposure-response relationships for this line of evidence are also uncertain. As noted in Section
8.2.2, the water quality criteria and standards used in this assessment may not be protective of all
macroinvertebrate taxa that are important prey for fish. However, direct aqueous exposures offish to
copper are unlikely to be toxic unless field concentrations are much higher than test leachate and
supernatant concentrations, so fish toxicity is not an important uncertainty. The uncertainty concerning
exposure-response relationships is smaller than the uncertainty concerning exposure.
Toxic Risks from Sediment
Although the consensus TECs and PECs are the best available effects benchmarks for sediment, their
applicability to tailings in Bristol Bay habitats is uncertain. The studies from which the values are
derived include lakes, reservoirs, and other systems that differ from rivers and streams in the Bristol
Bay watershed. However, the Clark Fork River (a tailings-contaminated salmonid stream) was one of the
confirmation sites for the TECs and PECs, which suggests that they are relevant to this type of situation.
Because the TECs and PECs are geometric means of prior sediment guidelines, the range of guidelines
provides an estimate of uncertainty. Alternate threshold values for copper range from 16 to 70 mg/kg
and probable effect values range from 86 to 390 mg/kg (MacDonald et al. 2000). The average copper
concentration of tailings (680 mg/kg) is well above all of these values, so this uncertainty is immaterial.
Some evidence suggests that these sediment guidelines may not be fully protective. When quotients of
sediment concentrations/TELs (one of the sources of the TECs and a numerically similar value) were
summed to address the combined toxicity of cadmium, copper, lead, and zinc, that value was not a
threshold for effects on stream invertebrates in the Colorado mining belt, and reductions in four
different community metrics occurred below the sum of TEL values (Griffith et al. 2004). This result
suggests that toxicity would be even more severe than the TECs and PECs suggest, but it may be
somewhat confounded by mine drainage.
Bristol Bay Assessment 942 January 2014
-------
Chapter 9 Tailings Dam Failure
Dietary Risks
Effects of dietary exposures would depend on the tailings composition, the copper bioaccumulation
factor for aquatic invertebrates, and the chronic toxic threshold for dietary exposures of rainbow trout.
Tailings composition may differ in practice from the PLP (2011) tests, but that uncertainty is unknown.
Ecological uncertainties are likely to be larger. Bioaccumulation factors for invertebrates and sediment
range from 0.15 to 1.77, even in a single river, which translates to invertebrate body burdens of 102 to
1,210 ug/g. That range encompasses the seven available estimates for the copper toxic dietary threshold
in rainbow trout, which range from 458 to 895 ug/g (Borgmann et al. 2005). This range of
bioaccumulation factors is not surprising given the differences in feeding habits, morphology, and
physiology among invertebrates. Finally, the variation in results among dietary toxicity studies for
copper is large, and the factors controlling dietary toxicity are poorly known.
Analogous Sites
The analogous sites for a potential tailings spill are all salmonid streams or rivers that received large
deposits of tailings from metal mines and that were well studied over an extended period (Box 9-5). A
large source of uncertainty when evaluating effects at those sites is the composition of the tailings. In
general, the Pebble test tailings are less acidic and contain less copper. On that basis, the nature and
magnitude of effects are likely to be less. However, the setting is different in ways that might increase
effects. For example, low hardness and low levels of dissolved materials in the mainstem Koktuli River
receiving waters would make biota of the receiving streams more susceptible to metals than at the
analogous sites. Although these cases are highly uncertain sources of information concerning the
potential toxicity of spilled tailings, they can be used with confidence to identify or confirm important
modes of exposure and the processes leading to exposure. They also demonstrate the persistence of
tailings and the leaching of their metals for multiple decades.
9.5.3 Weighing Lines of Evidence
This risk assessment is based on weighing multiple lines of evidence, and evidence for the various
routes of exposure (summarized in Table 9-9) is complex. For each route, sources of the exposure
estimate and the exposure-response relationship are indicated. All evidence is qualitatively weighed
based on three attributes: its logical implication, its strength, and its quality (Suter and Cormier 2011).
Evidence scored as positive (+) supports the case for adverse effects, whereas evidence scored as
negative (-) weakens the case for adverse effects. A zero (0) score indicates no or ambiguous evidence.
• The logical implication is the same for all lines of evidence: they all suggest that a spill from a
tailings dam failure would have adverse effects.
• The strength of the evidence is based primarily on the magnitudes of the hazard quotients. For
example, if the predicted concentration of copper is twice the median lethal concentration (LCso) for
rainbow trout, that is evidence of acute lethality; if it is 10 times as high, that is stronger evidence. In
Table 9-9, zero signifies a low quotient, + a moderate quotient, and ++ a high quotient.
Bristol Bay Assessment 943 January 2014
-------
Chapter 9
Tailings Dam Failure
• Quality is a complex concept that includes conventional data quality issues, but in this case, the
primary determinant is the relevance of the evidence to the mine scenarios. Because this is a
predictive assessment, none of the evidence is based on observations of an actual spill at the site of
concern. Hence, the evidence is based on the tailings dam failure scenarios, laboratory studies, or
field studies at other sites where tailings have spilled into streams or rivers or where biota were
exposed to other sediments with high copper levels. Separate quality scores are provided for the
exposure estimate and for the exposure-response relationship.
The scores indicated in Table 9-9 are not a substitute for the actual evidence, but rather are intended to
remind the reader what evidence is available and summarize the strength and quality of the different
lines of evidence.
Table 9-9. Summary of evidence concerning risks to fish from the toxic effects of a tailings dam
failure. The risk characterization is based on weighing multiple lines of evidence for different routes of
exposure. All evidence is qualitatively weighed (using one or more +, 0, or - symbols) on three
attributes: logical implication, strength, and quality. Here, all lines of evidence have the same logical
implication, since all suggest a tailings dam failure would have adverse effects. Strength refers to the
overall strength of the line of evidence, and quality refers to the quality of the evidence sources in
terms of data quality and relevance of evidence to the mine scenarios.
Route of Exposure
Sources of Evidence (Exposure/E-R)
Suspended sediment
Assumption/synthesis of laboratory
and field studies
Acute aqueous exposure
Leachate measurements/laboratory-
based criteria
Chronic aqueous exposure
Leachate measurements/laboratory-
based criteria
Chronic sediment exposure
Tailings measurements/sediment
guidelines
Chronic dietary exposure
Tailings measurements and
bioaccumulation factors/mean of
laboratory-based effects levels
All routes in the field
Exposure and effects at analogous
sites
Summary weight of evidence
Logical
Implication
+
+
+
+
+
+
+
Strength
++
0
0
++
0
++
+
Quality
Exposure
0
0
0
+
+
+
0
E-R
+
+
+
0
0
0
0
Results
Adverse effects on fish are certain. Although
the exposure level is unknown, it would
clearly be at effective levels.
Acute lethality to invertebrates close to
tailings storage facilities, but not downstream
and not to fish.
Chronic toxic effects on invertebrates due to
in situ leachate, but effects would end after
some years if diluted by clean sediment.
High likelihood of toxic effects on
invertebrates or fish based on a summary of
field studies.
Marginal dietary copper toxicity to trout would
be eliminated by minimal dilution.
Tailings spilled to streams have persisted and
caused severe effects, but the toxicity of the
tailings is likely to be higher in those cases.
All lines of evidence are consistent with toxic
effects of tailings. Despite the ambiguous
quality and marginal strength of some lines of
evidence, the overall strength is positive.
Notes:
E-R = exposure-response relationship.
Bristol Bay Assessment
9-44
January 2014
-------
Chapter 9 Tailings Dam Failure
9.6 Summary of Risks
9.6.1 Tailings Spill
Following a tailings spill, fish in the receiving stream and the invertebrates on which they depend would
be exposed to deposited tailings, suspended tailings, and tailings leachates. The fine texture of deposited
tailings would make them unsuitable for salmonid spawning and development, and a poor substrate for
the invertebrates that serve as food for developing salmon and resident trout and char. Suspended
tailings would have lethal and sublethal physical effects on fish and invertebrates immediately following
the spill, which are likely to continue with gradually diminishing intensity for years thereafter. The most
toxic constituent of the leachate and tailings would be copper, and exposures would be both direct and
through diet. Copper in leachate and in food is mildly toxic for fish, but copper and other constituents in
the tailings themselves would be moderately toxic to benthic invertebrates and potentially toxic to fish
eggs and larvae spawned in tailings-contaminated streams.
The physical and chemical effects of tailings on fish and invertebrates would be extensive in both space
and time. Elevated levels of suspended tailings would last for years. Deposited tailings and their leachate
would persist at toxic levels for decades. The acute effects of a tailings spill would extend far beyond the
modeled 30-km distance downstream. Based on data from other sites, tailings deposition from a spill
would extend for more than 100 km downstream, resulting in chronic exposures and effects. In addition,
the flow velocity of the receiving rivers, particularly in the spring, would readily transport the fine
tailings particles farther downstream. The mouth of the Koktuli River is 63.6 km from the confluence of
the South and North Fork Koktuli Rivers. From there, the mouth of the Mulchatna River is another 66.5
km, and the mouth of the Nushagak River at Dillingham is another 170.5 km, so contamination of the
entire downstream system would be likely soon after a spill.
We did not explicitly model failures of TSF 2 or TSF 3. The types of risks and effects that would occur
with a dam failure at TSF 2 or TSF 3 would be generally similar to those described for a failure at TSF 1.
The content and toxicity of TSF 2 and TSF 3 are assumed to be similar to that of TSF 1, and the
magnitude and extent of risks would be largely dependent on the volume of material released. One
important distinction, however, should be noted. The South Fork Koktuli River and Upper Talarik Creek
are hydrologically connected via groundwater transfer. In the event of a dam failure at TSF 3, transfer of
contaminated water leached from tailings fines and deposited in the South Fork Koktuli River valley to
the Upper Talarik Creek watershed and Iliamna Lake would be expected.
9.6.2 Remediation of a Tailings Spill
Although streams typically recover from aqueous effluents in less than a decade, the effects of tailings
deposition in streams and floodplains persist for as long as they have been monitored at analogous sites.
For that reason, many tailings-contaminated aquatic habitats in the United States have been or will be
dredged, riprapped, or redirected under the federal Superfund or state cleanup programs. The tailings
dam failure scenarios evaluated here do not consider any mitigating effects of remediation efforts by the
mine operator or other parties. Although such remedial actions have net benefits, they create long-term
Bristol Bay Assessment 945 January 2014
-------
Chapter 9 Tailings Dam Failure
impacts on aquatic habitats. For example, riprapping reduces downstream exposure to tailings and
associated metals by reducing erosion of floodplain tailings, but it also reduces fish habitat complexity
and quality for fish by channelizing the stream or river (Schmetterling et al. 2001).
Remediation under the tailings dam failure scenarios considered here would be particularly difficult and
damaging, because the area of the spill is almost entirely unaffected by other development. One or more
roads would need to be built into this roadless area to bring in equipment and haul out the tailings. At
the upper end of the affected area, the process of removing tailings would do little additional damage,
since the structure of the watershed would have been destroyed. If tailings removal extended to streams
that were not scoured in the initial tailings release, removal would destroy those streams and associated
wetlands. If removal was not undertaken, the substrate of the streams would consist of tailings until
high flows carried them downstream. It may be technically impractical to recover tailings fines that have
been transported past the point of confluence with larger rivers.
In the Pebble 0.25 tailings dam failure scenario considered here, an estimated 45 million metric tons of
tailings solids would be deposited in the North Fork Koktuli River valley (calculated from Table 9-5,
assuming a dry density of 1.42 metric tons/m3). Complete removal of this material would require a
substantial earth-moving effort (e.g., including over 3 million round trips by 20-ton dump trucks).
Recovery and removal would be even more challenging in the Pebble 2.0 tailings dam failure scenario, in
which an estimated 97 million metric tons of tailings solids would be deposited in the North Fork
Koktuli River valley, and an additional 263 million metric tons of tailings solids would be transported
beyond the confluence with the South Fork Koktuli River and into the mainstem Koktuli River (Table 9-
5). Material not deposited on the floodplains would be carried downstream; material deposited in the
floodplains, if not recovered, would be remobilized by future precipitation and would wash
downstream. It is unlikely that tailings in river channels would be recovered, because the fine material
would be rapidly transported by the relatively high flow velocities of the rivers.
Remediation of prior tailings dam failures can serve as case studies. Failures are numerous, but the
degree to which remediation results in restoration of natural resources has not been well documented.
The 1998 failure of the Aznalcollar tailings dam at Los Frailes Mine in Seville, Spain (Box 9-1) has been
described as a case of substantial remediation. However, this kind of successful removal of tailings
would be difficult to replicate in the Bristol Bay watershed. The Aznalcollar area has a drier and warmer
climate, flatter topography, better access from existing roads, and more readily available equipment and
labor compared to the Bristol Bay region. The goal of the Aznalcollar remediation was restoration of
land use, which would not be the primary goal in the Bristol Bay watershed. In addition, potential
releases from TSF 1 would be much larger than the release at Aznalcollar (Box 9-1).
Emergency plans for metal mines in Alaska that have been provided to USEPA do not address
remediation or restoration after a tailings spill has occurred. In fact, no tailings spill has been reported in
Alaska, so it is not clear what remediation or restoration might be required. Given the uncertain toxicity
of the tailings, the difficulty and expense of remediation and restoration, and the damage that would be
done by remediation, it is possible that a spill would be left to be restored by natural processes.
Bristol Bay Assessment 946 January 2014
-------
10.1 Introduction
Because the Bristol Bay watershed is located in one of the last remaining virtually roadless areas in the
United States, development of any mine in the Bristol Bay watershed would require substantial
expansion and improvement of the region's transportation infrastructure. There are few existing
roadways, no improved federal or state highways, and no railroads, pipelines, or other major industrial
transportation infrastructure (Figure 6-6). As described in Section 6.1.3, the mine scenarios evaluated in
this assessment include a 138-km gravel surface, all-weather permanent access road (Figure 6-6)
connecting the mine site to a new deep-water port on Cook Inlet (Ghaffari et al. 2011). This length does
not include road sections within the mine site itself. Approximately 113 km of this corridor would fall
within the Kvichak River watershed.
The transportation corridor area considered in the assessment comprises 32 subwatersheds draining to
Iliamna Lake (Figure 2-7). These subwatersheds, referred to as the corridor subwatersheds, encompass
approximately 2,340 km2 and contain nearly 1,900 km of streams mapped for this analysis (see Chapter
3 for a description of these methods). The seven largest subwatersheds are, from west to east, the
headwaters of Upper Talarik Creek, the headwaters of the Newhalen River, Chekok Creek, Canyon Creek,
Knutson Creek, Pile River, and the Iliamna River. The Newhalen River is the largest river crossed by the
corridor, draining Sixmile Lake and Lake Clark. Sockeye return to spawn in the Newhalen River and
tributaries to Sixmile Lake and Lake Clark. The transportation corridor would cross the Newhalen River
and parallel the north shore of Iliamna Lake (Figure 6-6). It would traverse rolling, glaciated terrain for
approximately 60 road km until reaching steeper hillsides northwest of the village of Pedro Bay and the
shoreline of Knutson Bay. After crossing gentler terrain around the northeast end of Iliamna Lake
(Pedro Bay and Pile Bay), the corridor would cross the Chigmit Mountains (the highest source of runoff
in the Bristol Bay watershed) along the route of the existing Pile Bay Road to tidewater at Williamsport.
From there it would cross Iliamna Bay and follow the coastline to the port site on Iniskin Bay, off Cook
Bristol Bay Assessment
10-1
January 2014
-------
Chapter 10 Transportation Corridor
Inlet. Highly variable terrain and variable subsurface soil conditions, including extensive areas of rock
excavation in steep mountainous terrain, are expected over this proposed route.
Although this route is not necessarily the only option for corridor placement, the assessment of potential
environmental risks would not be expected to change substantially with minor shifts in road alignment.
Along most feasible routes, the proposed transportation corridor would cross many streams (including
unmapped tributaries), rivers, wetlands, and extensive areas with shallow groundwater, all draining to
Iliamna Lake (Figures 10-1 and 10-2).
In this chapter, we consider the risks to fish habitats and populations associated with the transportation
corridor, as illustrated in a conceptual model showing potential linkages among the corridor, associated
sources and stressors, and assessment endpoints (Figure 10-3). We begin with a discussion offish
habitats and populations along the corridor. We then consider potential impacts on these habitats and
populations resulting from its construction and operation. Although the transportation corridor would
include the road and adjacent pipelines (Section 6.1.3), we focus primarily on the road component;
potential pipeline failures are considered in Chapter 11.
Best management practices (BMPs) or mitigation measures would be used along the transportation
corridor to minimize potential risks to salmonids and the ecosystems that support them. Relevant BMPs,
and their likely effectiveness, are discussed in text boxes throughout the chapter.
Bristol Bay Assessment 102 January 2014
-------
Chapter 10
Transportation Corridor
Figure 10-1. Streams, wetlands, ponds, and lakes along the transportation corridor. Streams and rivers are from the National Hydrography
Dataset (USGS 2012); wetlands, lakes, and ponds are from the National Wetlands Inventory (USFWS 2012).
WETLAND INVENTORY
DATA NOT AVAILABLE
\\ ^o
Ilismna^- ^,
Bay
Approximate Location of Pebble Deposit
Transportation Corridor
Transportation Corridor (Outside Assessment Area)
Transportation Corridor Area
Subwatersheds within Area
Rivers and Streams
Freshwater Wetland
Freshwater Pond
Lake
Bristol Bay Assessment
10-3
January 2014
-------
Chapter 10
Transportation Corridor
« Approximate Location of Pebble Deposit
Transportation Corridor
Rivers and Streams
^^^| Freshwater Wetland
Freshwater Pond
0
^m
0
N
A
2
^1=
2
4
1 Kilometers
4
Bristol Bay Assessment
10-4
January 2014
-------
Chapter 10
Transportation Corridor
[ transportation corridor J
stream
crossings
V
culvert blockage
or perching
T- channel erosion
& entrenchment
\/
V
V
4- production and export of
food & other resources
1s competition
& predation
A riparian
vegetation
\/
V
A stream
geomorphology
V
\/
t inhibition of
fish passage
V
V
4- floodplain
connectivity
V
'T' chemical contaminants
(metals, salts, other chemicals]
roadbed cuts, ]
fills & ditches J
V
A downstream water flows
V
LCGCND
additional step in
causal pathway J
additional step in
causal pathway
modifying
factor
Within a shape, '|x indicates an increase in the parameter,
-|. indicates a decrease in a parameter, and A indicates a
change in the parameter.
Arrows leading from one shape to another indicate a
hypothesized cause-effect relationship.
Shapes bracketed under another shape are specific
components of the more general shape under which they
appeal.
4- rearing habitat
(quality orquantity)
4- macroinvertebrate
prey
\/
4, feeding
ability
V
\/
1s physiolog cal
stress
V
4- spawning habitat
(quality orquantity)
Adownstream water
temperatures
A magnitude &
frequency of high flows
^ alteration of channel
morphology & floodplain
connectivity
-f aquatic habitat
fragmentation
4, overwintering habitat
(quality or quantity)
4, i icubalion habitat
(quality or quantity)
V
V
v^fabundance, productivity or diversity)
V
%b other fishes
(abundance, productivity or diversity)
t t
mitigation
Bristol Bay Assessment
10-5
January 2014
-------
This page intentionally left blank
-------
Chapter 10 Transportation Corridor
10.2 Fish Habitats and Populations along the
Transportation Corridor
In Chapter 3, we characterized stream segments in the Nushagak and Kvichak River watersheds by
relative size (mean annual streamflow), channel gradient, and an index of the degree of channel
constraint to describe floodplain potential (proportion of flatland in lowland, where stream segments
with greater than 5% flatland in lowland in each reach's adjacent drainage basin are likely to be
unconstrained and to exhibit floodplain potential). These attributes were selected because they
represent fundamental aspects of the physical and geomorphic stream setting and provide context for
stream and river habitat development and consequent fish habitat suitability (Burnett et al. 2007). Table
10-1 summarizes the proportion of stream channel lengths in the corridor subwatersheds (Scale 5),
classified according to stream size, channel gradient, and floodplain potential. To allow direct visual
comparison of the distribution of stream characteristics in the corridor subwatersheds relative to those
in the entire Nushagak and Kvichak River watersheds (Scale 2), we present cumulative frequency plots
in Figure 10-4. These plots show a frequency curve for each attribute at each geographic scale.
Attributes are grouped into meaningful classes (Chapter 3), denoted by the vertical red classification
bars. For example, the lowest gradient streams are classified as having gradients of less than 1% (Table
10-1), as shown by the vertical classification bar at 1% in Figure 10-4B. Cumulative frequency plots can
be interpreted by evaluating the height at which the frequency curve is intersected by the red vertical
classification bar. In Figure 10-4B, the 1% gradient classification bar intersects the Scale 5 frequency
curve (solid black line) at a cumulative frequency value of approximately 32%. Thus, approximately
32% of the stream kilometers in the corridor subwatersheds (Scale 5) are less than 1% gradient. In
comparison, approximately 64% of the stream kilometers in the Nushagak and Kvichak River
watersheds (Scale 2) are less than 1% gradient.
Streams along the transportation corridor have not been sampled as extensively as streams near the
Pebble deposit. Small to large rivers (2.8 m3/s mean annual streamflow and larger) that would be
crossed by the corridor (Table 10-1) provide spawning and rearing habitat, and are important routes for
adult salmonid migration to upstream spawning areas and juvenile salmonid migration downstream to
Iliamna Lake. Large and small streams with low to moderate gradients (3% or less) provide important
high-quality spawning habitats, primarily for sockeye salmon. These streams also likely provide high-
quality seasonal and some year-round habitats for resident Dolly Varden and rainbow trout. Dolly
Varden are distributed across a much wider range of stream gradients (ADF&G 2012). The majority of
stream length in the corridor subwatersheds consists of small headwater (58%) and medium (31%)
streams, whereas small and large rivers make up 10 and 2% of stream length, respectively (Table 10-1).
A majority (62%) of stream length in the corridor subwatersheds is classified as low to moderate
gradient (32% at less than 1% gradient, and 30% at 1 to 3% gradient) (see Box 3-1 for discussion on
how gradient was calculated). However, the corridor streams are generally steeper and have higher
proportions of stream length without floodplain potential (i.e., less than 5% of flatland in lowland
adjacent to stream) relative to those in the larger Nushagak and Kvichak River watersheds (Table 10-1,
Bristol Bay Assessment 107 January 2014
-------
Chapter 10
Transportation Corridor
Figure 10-4). Streams and rivers with high proportions of length with floodplain potential are more
likely to be unconstrained and to develop complex off-channel habitats that provide a diversity of
channel habitat types and create favorable conditions, particularly for salmonid rearing. However, the
corridor streams are unique within the Nushagak and Kvichak River watersheds in that many of them
are short and originate within the corridor subwatersheds. In addition, they all flow into Iliamna Lake,
which provides high-quality habitat suitable for salmonid rearing and migration among streams.
Table 10-1. Proportion of stream channel length in stream subwatersheds intersected by the
transportation corridor (Scale 5) classified according to stream size (based on mean annual
discharge in m3/s), channel gradient (%), and floodplain potential (based on % flatland in lowland).
Gray shading indicates proportions greater than 5%; bold indicates proportions greater than 10%.
Mean annual discharge
Small headwater and Iliamna
Lake tributary streams3
Medium streams'5
Small rivers0
Large riversd
Gradient
<1%
FP
11%
7%
5%
2%
NFP
3%
2%
2%
0%
>1% and <3%
FP
4%
1%
0%
0%
NFP
12%
10%
3%
0%
>3% and <8%
FP
1%
1%
0%
0%
NFP
17%
7%
0%
0%
>8%
FP
0%
0%
0%
0%
NFP
10%
3%
0%
0%
Notes:
a 0-0.15 m3/s; headwater tributaries of streams crossing the transportation corridor and small streams flowing directly to Iliamna Lake (e.g.,
Eagle Bay and Chekok Creeks).
b 0.15-2.8 m3/s; upper reaches and larger tributaries of streams crossing the transportation corridor, and medium streams flowing directly
into Iliamna Lake (e.g., Chinkelyes and Knutson Creeks).
0 2.8-28 m3/s; middle to lower portions of the Iliamna and Pile Rivers.
d >28 m3/s; the Newhalen River.
FP = high floodplain potential (>5% flatland in lowland); NFP = no or low floodplain potential (<5% flatland in lowland) (see Chapter 3 for
additional explanation).
At the scale of the Nushagak and Kvichak River watersheds, 85% of stream length is classified as less
than 3% gradient (64% at less than 1% gradient and 21% at 1 to 3% gradient), versus 62% in the
corridor subwatersheds. Sixty percent of total stream length in the Nushagak and Kvichak River
watersheds is classified as exhibiting floodplain potential, versus 31% in the corridor subwatersheds
(Figure 10-4). These differences stem in large part from the large portions of the unconfined, low-
gradient lower Nushagak River watershed. Percent of stream length less than 3% gradient is 73 and
91% in the Kvichak and Nushagak River watersheds, respectively; the percent of stream length
classified as floodplain prone is 50% across the Kvichak River watershed and 65% across the Nushagak
River watershed. Thus, stream characteristics in the transportation corridor area are generally more
similar to those in the Kvichak River watershed. Characterization of stream segments for the entire
Nushagak and Kvichak River watersheds, as well as the methods used, are described in Chapter 3.
Bristol Bay Assessment
10-8
January 2014
-------
Chapter 10
Transportation Corridor
Figure 10-4. Cumulative frequency of stream channel length classified by (A) mean annual
streamflow (MAP) (m3/s), (B) channel gradient (%), and (C) floodplain potential (based on % flatland
in lowland) for stream subwatersheds intersected by the transportation corridor (Scale 5) versus the
Nushagak and Kvichak River watersheds (Scale 2). See Section 3.4 for further explanation of MAP,
gradient, and floodplain potential classifications.
100%
§ &
~ c 60%
"^ QJ
fO l
"5 c
E 5 40%
3 OJ
<-» is
ft
_ro
3
TO ^
1 I
3 O)
ro 40%
£
•M
LO
0%
0%
^
r —
^^^^ r r-iln n
MAP Classification
Scale 2
10 20 30
Mean Annual Streamflow (m3/s)
40
50
sno/
^
^ - * "
•Gradient Classification
Scale 2
4%
8% 12%
Channel Gradient
16%
20%
•ScaleS
•Floodplain Potential Classification
Scale 2
-h
-+-
20%
40% 60%
Flatland in Lowland
80%
100%
Bristol Bay Assessment
10-9
January 2014
-------
Chapter 10 Transportation Corridor
These low- to moderate-gradient streams provide important spawning habitat for sockeye. The Kvichak
River watershed includes over 100 separate sockeye salmon spawning locations (Demory et al. 1964,
Morstad 2003), including small tributary streams, rivers, mainland beaches, island beaches, and spring-
fed ponds. The spatial separation and diverse spawning habitat features within the watershed have
influenced genetic divergence among spawning populations of sockeye salmon at multiple spatial scales
(Gomez-Uchida et al. 2011). These distinct populations can occur at very fine spatial scales. For example,
sockeye salmon that use spring-fed ponds and streams approximately 1 km apart exhibit differences in
traits that are consistent with discrete populations, such as spawn timing, spawn site fidelity, and
productivity (Quinn et al. 2012).
Sockeye spawning has been observed at 30 locations along the transportation corridor (Demory et al.
1964). The Alaska Department of Fish and Game (ADF&G) has conducted aerial index counts of sockeye
salmon spawning abundance at these locations in most years since 1955 (Morstad 2003). We recognize
that survey values tend to underestimate true abundance for many reasons. An observer in an aircraft is
not able to count all fish in dense aggregations, and only a fraction of the fish that spawn at a given site
are present at any one time (Bue et al. 1988, Jones et al. 2007). Surveys intended to capture peak
abundance may not always do so. Weather, water clarity, and other factors influencing fish visibility can
also contribute to underestimates. Finally, spawning locations along the corridor occur across a variety
of habitats, including mainland beaches, small ponds, streams, and larger rivers. Aerial survey-based
indices of sockeye salmon spawning abundance vary considerably. Sockeye index counts are highest in
the Iliamna River (averaging over 100,000 spawners), the Newhalen River (averaging over 80,000
spawners), and on beaches in Knutson Bay (averaging over 70,000 spawners) (Table 10-2, Figure 10-5).
In some years, these counts can be very large, as illustrated by the 1960 survey for Knutson Bay that
reported 1 million adults (Demory et al. 1964). Sockeye spawning is associated with upwelling
groundwater areas on beaches along the north and east shores of Knutson Bay, adjacent to the
transportation corridor. In addition, sockeye use of spring-fed ponds has been observed at eight
locations along the corridor. These locations tend to have fewer spawners (approximately 2,700 on
average), but fish using these locations may be adapted to the unique abiotic features of ponds (Quinn et
al. 2012).
Bristol Bay Assessment 1010 January 2014
-------
Chapter 10
Transportation Corridor
Table 10-2. Average number of spawning adult sockeye salmon at locations near the transportation corridor. See Figure 10-5 for the
locations of these areas.
Map
Point
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Area
Upper Talarik
Newhalen River System
Newhalen River System
Newhalen River System
Newhalen River System
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
North East
Iliamna River System
Iliamna River System
Iliamna River System
Iliamna River System
Iliamna River System
Area Name
Upper Talarik Creek
Newhalen River
Little Bear Creek/Ponds
Alexi Creek
Alexi Lakes
Roadhouse Creek
Northwest Eagle Bay Creek
Northeast Eagle Bay Creek/Ponds
Northeast Eagle Bay Creek Ponds
Youngs Creek
Chekok Creek/Ponds
Tomkok Creek
Canyon Creek
Wolf Creek Ponds
Mink Creek
Canyon Springs
Prince Creek Ponds
Knutson Bay
Knutson Creek
Knutson Ponds
Pedro Creek & Ponds
Russian Creek
Lonesome Bay Creek
Pile River
Swamp Creek
Iliamna River
Bear Creek & Ponds
False Creek
Old Williams Creek
Chinkelyes Creek
Type
Stream
River
Ponds
Stream
Lake
Stream
Stream
Stream
Ponds
Stream
Stream
Stream
Stream
Ponds
Stream
Ponds
Ponds
Lake
Stream
Ponds
Ponds
Stream
Stream
River
Stream
River
Ponds
Stream
Stream
Stream
Average Number of Sockeye
Salmon Spawners (1955-2011)
7,021
84,933
527
1,176
7,121
1,052
1,649
3,416
4,766
3,532
1,840
10,882
8,015
4,469
1,144
884
3,797
72,845
1,548
1,200
4,259
2,263
1,026
6,431
1,091
101,306
1,748
1,317
3,726
9,128
Number of Years Spawners
were Counted (Max = 57)
49
34
20
27
33
28
32
38
5
38
32
38
38
26
35
20
34
47
41
39
48
17
6
38
18
53
30
21
27
46
Range
0-70,600
97-730,900
0-1,860
0-13,200
11-38,000
0-4,950
0-17,562
0-18,175
200-11,700
0-26,500
0-8,700
300-56,600
200-48,000
0-28,000
0-6,000
0-5,000
5-34,800
1,000-1,000,000
1-6,600
0-6,350
0-38,150
0-20,000
32-2,675
0-39,200
25-7,700
3,000-399,300
40-10,300
0-13,300
0-38,000
50-44,905
Notes:
Locations are organized from west to east along the corridor.
Sources: Morstad 2003, Morstad pers. comm.
Bristol Bay Assessment
10-11
January 2014
-------
Chapter 10
Transportation Corridor
Figure 10-5. Location of sockeye salmon surveys and number of spawners observed along the transportation corridor. Numbers refer to
map points listed in Table 10-2.
Lake Clark
N
A
10
] Kilometers
10
] Miles
Average Number of Sockeye Spawners
< 1,000
-1,000 to <2.000
-2.000 to < 10.000
>10,000 to <50,000
>50,000
8ay
* Approximate Location of Pebble Deposit
~ Transportation Corridor
" Transportation Corridor (Outside Assessment Area)
| Transportation Corridor Area
Subwatersheds within Area
Cook Inlet
Bristol Bay Assessment
10-12
January 2014
-------
Chapter 10 Transportation Corridor
Less is known about the occurrence or abundance of other salmon species in streams and rivers
crossing or adjacent to the transportation corridor. Chinook, coho, pink, and chum salmon are present in
the Kvichak River watershed, but data for their spatial occurrences are for isolated points in the system
(Johnson and Blanche 2012). In streams intersected by the transportation corridor, sockeye salmon are
in all streams included in the Catalog of Waters Important for Spawning, Rearing, or Migration of
Anadromous Fishes—Southwestern Region (also known as the Anadromous Waters Catalog [AWC])
(Johnson and Blanche 2012) (Figure 10-6). Working from west to east along the corridor, streams with
salmon species in addition to sockeye are as follows: Upper Talarik Creek (Chinook, coho, chum, and
pink salmon), the Newhalen River (Chinook and coho salmon), Youngs Creek (East and West Branches),
Chekok and Tomkok Creeks (coho salmon), Swamp Creek (a tributary to Pile Bay) (Chinook salmon),
and the Iliamna River (Chinook, coho, chum, and pink salmon).
Dolly Varden and rainbow trout distributions have not been surveyed as extensively as salmon
distributions along the transportation corridor (ADF&G 2012). Dolly Varden have been documented in
nearly every sockeye salmon-bearing stream that would be crossed by or adjacent to the corridor, as
well as in locations upstream of reported anadromous salmon use (Figure 10-6). Rainbow trout
presence along the corridor is reported for only a few streams, including Upper Talarik Creek, the
Newhalen River, an unnamed tributary to Eagle Bay, Youngs Creek, Tomkok Creek, and Swamp Creek
(ADF&G 2012). Rainbow trout have also been documented in the Iliamna River (Russell 1977) and
Chinkelyes Creek (Berejikian 1992).
The distributions of both Dolly Varden and rainbow trout along the transportation corridor are likely
much more extensive than reported in the Alaska Freshwater Fish Inventory (AFFI) resident fish
database (ADF&G 2012), which does not account for seasonal movements or low sampling effort.
Sockeye salmon provide an important food subsidy to Dolly Varden and rainbow trout. For example,
Denton et al. (2009) reported Dolly Varden movement into multiple ponds used by spawning sockeye
next to the Pedro Bay village, where they feed on sockeye salmon fry, eggs, and carcass-associated
blowflies. Information on rainbow trout movement between Iliamna Lake and streams intersected by
the corridor is not available, but these movements are likely to occur. Movements between lakes and
tributary streams in response to feeding and spawning opportunities have been documented elsewhere
in Iliamna Lake (Russell 1977), the Alagnak River system (Meka et al. 2003), and in the Wood River lake
system (Ruff et al. 2011).
Bristol Bay Assessment 1013 January 2014
-------
Chapter 10
Transportation Corridor
Figure 10-6. Reported salmon, Dolly Varden, and rainbow trout distributions along the transportation corridor. Salmon presence data are
from the Anadromous Waters Catalog (Johnson and Blanche 2012); Dolly Varden and rainbow trout presence data are from the Alaska
Chinkelyes Creek (Berejikian 1992), although these points are not indicated on this map.
10
Kilometers
5 10
Miles
Note: Sampling intensity is greatly reduced away from the Pebble deposit area
Streams without data points may not have been surveyed: thus, it is unknown
whether or not they contain these species..
Approximate Location of Pebble Deposit
Transportation Corridor (Outside Assessment Area)
Dolly Varden
Rainbow Trout
Transportation Corridor
Transportation Corridor Area
Subwatersheds within Area
Salmon
Bristol Bay Assessment
10-14
January 2014
-------
Chapter 10 Transportation Corridor
10.3 Potential Risks to Fish Habitats and Populations
Only rarely has it been possible to build roads that have no negative effects on streams (Furniss et al.
1991). Roads modify natural drainage networks and accelerate erosion processes, which can lead to
changes in streamflow regimes, sediment transport and storage, channel bank and bed configurations,
substrate composition, and the stability of slopes adjacent to streams. Road construction can increase
the frequency of slope failures by orders of magnitude, depending on variables such as soil type, slope
steepness, bedrock type and structure, and presence of subsurface water. These slope failures can result
in episodic sediment delivery to streams and rivers, potentially for decades after roads are built (Furniss
et al. 1991). All of these potential changes can have important biological consequences for anadromous
and resident fishes by negatively affecting food, shelter, spawning habitat, water quality, and access for
upstream and downstream migration (Appendix G) (Furniss et al. 1991).
In the Bristol Bay region, risks to fish from construction and operation of the transportation corridor
would be complex and potentially significant, largely because of hydrological issues. Field observations
in the mine area (Hamilton 2007, Woody and O'Neal 2010) indicate terrain with abundant near-surface
groundwater and a high incidence of seeps and springs associated with complex glaciolacustrine,
alluvial, and slope till deposits (Appendix G). The abundance of mapped wetlands (Figures 10-1 and 10-
2) further demonstrates the pervasiveness of shallow subsurface flows and high connectivity between
groundwater and surface-water systems in the areas traversed by the transportation corridor
(Appendix G). As noted in Section 3.3, the strong connection between groundwater and surface waters
helps to moderate water temperatures and streamflows, and this moderation can be critical for fish
populations. The construction and operation of the transportation corridor could fundamentally alter
connections between shallow aquifers and surface channels and ponds by intercepting shallow
groundwater flowpaths, leading to impacts on surface water hydrology, water quality, and fish habitat
(Darnell et al. 1976, Stanford and Ward 1993, Forman and Alexander 1998, Hancock 2002).
The lengths of the transportation corridor and their proximities to National Hydrography Dataset (NHD)
streams (USGS 2012) and National Wetlands Inventory (NWI) wetlands, ponds, and small lakes (USFWS
2012) are shown in Tables 10-3 and 10-4, respectively (see Box 10-1 for a description of methods used
to estimate these values). In sum, the length of road within 200 m of NHD streams or NWI aquatic
habitats would be approximately 67 km (Table 10-5). These lengths do not encompass the section of
corridor outside of the Kvichak River watershed (i.e., the watersheds flowing into Cook Inlet). The 200-
m road buffer was derived from an estimate of the road-effect zone for secondary roads (Forman 2000).
The largest impact on sockeye salmon would likely occur where the road would run parallel to the
Iliamna River and Chinkelyes Creek, sites at which many sockeye salmon spawn (Figure 10-2, Inset C).
Other high-impact areas include where the road would run parallel to Knutson Bay, intersecting many
small streams and where groundwater upwelling supports spawning for hundreds of thousands of
salmon (Figure 10-2, Inset B), and where the road crosses wetlands north of Iliamna Lake (Figure 10-2,
Inset A).
Bristol Bay Assessment 1015 January 2014
-------
Chapter 10 Transportation Corridor
In the following sections, we consider potential risks to fish habitats and populations resulting from
construction and operation of the transportation corridor. We focus on risks related to filling and
alteration of wetlands, stream crossings, fine sediments, dust deposition, runoff contaminants, and
invasive species.
BOX 10-1. CALCULATION OF STREAM LENGTHS AND WETLAND AREAS AFFECTED BY
TRANSPORTATION CORRIDOR DEVELOPMENT
We used the National Hydrography Dataset (NHD) (USGS 2012), the National Wetlands Inventory (NWI)
(USFWS 2012), the Alaska Anadromous Waters Catalog (AWC) (Johnson and Blanche 2012), and the Alaska
Freshwater Fish Inventory (AFFI) (ADF&G 2012) to evaluate potential effects of the transportation corridor on
hydrologic features and fish populations.
The length of stream downstream of each crossing was estimated from NHDflowlines. Stream length by
subwatershed, based on 12-digit hydrologic unit codes, was calculated as the total distance from each
crossing to Iliamna Lake. In the multiple instances where stream crossings were tributaries to a single main
channel, the mainstem length was only counted once (Table 10-3). Downstream lengths reported in Table
10-6 include mainstem lengths downstream of tributary crossings. In cases where the corridor crossed
tributaries of a mainstem channel, the mainstem length is included in both crossings.
Mean annual streamflow of NHD streams upstream of the transportation corridor was estimated using
methods described in Box 3-2.
The channel gradient of NHD stream segments intersected by and upstream of the corridor was estimated
using a 30-m National Elevation Dataset digital elevation model (DEM) (Gesch 2007, Gesch etal. 2002,
USGS 2013). A drainage network was developed from a flow analysis using the DEM and slope was
estimated usingthis drainage network. The OEM-based drainage network paralleled the NHD stream
flowlines and therefore, using the toolset in the spatial analyst extension in ArcGIS, slope from the drainage
network was transferred to NHD reach segments. A 12% slope was used to calculate stream length likely to
support fish (Table 10-6). Stream length upstream of the corridor with less than 12% slope was based on
the NHD stream length to the first instance where slope was greater than 12%. The analysis of upstream
fish habitat was extended to include streams in subwatersheds in the Headwaters Newhalen River, Tomkok
Creek, Pile River, and Iliamna River.
For the analysis of road length intersecting and within 100 or 200 m of either a stream or wetland
(Tables 10-3 through 10-5), each stream (NHD) or wetland (NWI) was buffered to a distance of 100 m and
200 m and the length of corridor within these ranges was summed. Similarly, for the area of wetlands,
ponds, and small lakes within 100 m and 200 m of the road corridor, the road corridor was buffered and the
area of wetlands, ponds, and small lakes within that buffered area was summed across the length of road.
For the area of wetlands, ponds, and small lakes directly filled by the road corridor, we assumed a road
width of 9.1 m.
The characterization of both stream length and wetland, pond, and small lake area affected is likely a
conservative estimate. The NHD may not capture all stream courses and may underestimate channel
sinuosity, resulting in underestimates of affected stream length. Additionally, the AWC and the AFFI do not
necessarily characterize all potential fish-bearing streams due to limited samplingalongthe corridor. The
characterization of wetland, pond, and small lake area is limited by the resolution of the available NWI data
product. In this analysis, the transportation corridor often bisects wetland features and the wetland area
falling outside the 200-m boundary was assumed to maintain its functionality. We were also unable to
determine the effect that the transportation corridor may have on wetlands that had no direct surface water
connection, but that may be hydrologically connected via groundwater pathways. Together, these limitations
likely make our calculations an underestimate of the effect that transportation corridor development would
have on hydrologic features in this region. These estimates could be improved with enhanced, higher-
resolution mapping, increased sampling of possible fish-bearing waters, and ground-truthing of surface-
water and groundwater connections.
Bristol Bay Assessment 1016 January 2014
-------
Chapter 10
Transportation Corridor
Table 10-3. Proximity of the transportation corridor to National Hydrography Dataset streams (USGS 2012).
HUC-12 Name or Description
Headwater, Upper Talarik Creek
Upper tributary stream to Upper Talarik Creek
Tributary to Newhalen River portion of corridor
Headwaters, Newhalen River
Outlet, Newhalen River
Roadhouse Creek
Iliamna Lake
Eagle Bay Creek
Youngs Creek Mainstem (Roadhouse Mountain HUC)
Youngs Creek East Branch
Chekok Creek
Canyon Creek
Knutson Creek
Outlet, Pile River
Middle Iliamna River
Chinkelyes Creek
HUC-12 Digit
190302060702
190302060701
190302051404
190302051405
190302051406
190302060907
190302060914
190302060905
190302060903
190302060904
190302060302
190302060902
190302060901
190302060104
190302060205
190302060206
Total length across all HUCs
Percentage across all HUCs
Proximity to Streams
Not nearby
(km)
5.4
4.3
7.8
2.6
4.2
0.8
29.3
3.1
3.0
1.4
1.8
1.1
1.2
2.1
4.5
9.6
82.1
73%
<100 m
(km)
0.8
0.2
1.9
0.4
1.5
1.2
4.3
0.5
0.1
1.0
0.3
0.1
0.3
0.6
1.1
0.8
15.3
14%
100-200 m
(km)
1.2
0.1
1.2
0.4
0.8
1.3
4.1
0.8
0.2
0.6
0.3
0.2
0.4
0.7
0.7
2.1
15.2
13%
Total
(km)
7.4
4.6
10.9
3.4
6.5
3.3
37.7
4.4
3.4
3.0
2.5
1.4
2.0
3.4
6.4
12.5
113
100%
Notes:
HUC = hydrologic unit code.
Bristol Bay Assessment
10-17
January 2014
-------
Chapter 10
Transportation Corridor
Table 10-4. Proximity of the transportation corridor to National Wetlands Inventory wetlands, ponds, and small lakes (USFWS 2012).
HUC-12 Name or Description
Headwater, Upper Talarik Creek
Upper tributary stream to Upper Talarik Creek
Tributary to Newhalen River portion upstream of corridor
Headwaters, Newhalen River
Outlet, Newhalen River
Roadhouse Creek
Iliamna Lake
Eagle Bay Creek
Youngs Creek Mainstem (Roadhouse Mountain HUC)
Youngs Creek East Branch
Chekok Creek
Canyon Creek
Knutson Creek
Outlet, Pile River
Middle Iliamna River
Chinkelyes Creek
HUC-12 Digit
190302060702
190302060701
190302051404
190302051405
190302051406
190302060907
190302060914
190302060905
190302060903
190302060904
190302060302
190302060902
190302060901
190302060104
190302060205
190302060206
Total length across all HUCs
Percentage across all HUCs
Proximity to Wetlands
Not nearby
(km)
0.2
1.7
4.0
2.3
1.1
0.7
28.3
1.3
0.9
0.3
1.8
0.8
1.0
0.3
2.7
7.7
55.0
49%
Intersects
(km)
1.9
0.3
0.4
0.1
2.4
0.3
1.8
0.7
0.2
0.5
0.2
0.0
0.1
1.2
0.6
1.4
12.2
11%
<100 m
(km)
4.0
1.4
3.9
0.4
1.7
1.8
3.9
1.7
1.1
0.8
0.3
0.2
0.6
1.5
1.7
1.9
27.0
24%
100-200 m
(km)
1.2
1.2
2.6
0.5
1.4
0.5
3.7
0.8
1.2
1.5
0.2
0.3
0.3
0.5
1.3
1.5
18.5
16%
Total
(km)
7.4
4.6
10.9
3.4
6.5
3.3
37.7
4.4
3.4
3.0
2.5
1.4
2.0
3.4
6.4
12.5
113
100%
Notes:
HUC = hydrologic unit code.
Bristol Bay Assessment
10-18
January 2014
-------
Chapter 10
Transportation Corridor
Table 10-5. Proximity of the transportation corridor to water, in terms of the length occurring within 200 m of National Hydrography
Dataset streams (USGS 2012) or National Wetlands Inventory wetlands, ponds, and small lakes (USFWS 2012).
HUC-12 Name or Description
Headwater, Upper Talarik Creek
Upper tributary stream to Upper Talarik Creek
Tributary to Newhalen River portion upstream of corridor
Headwaters, Newhalen River
Outlet, Newhalen River
Roadhouse Creek
Iliamna Lake
Eagle Bay Creek
Youngs Creek Mainstem (Roadhouse Mountain HUC)
Youngs Creek East Branch
Chekok Creek
Canyon Creek
Knutson Creek
Outlet, Pile River
Middle Iliamna River
Chinkelyes Creek
HUC-12 Digit
190302060702
190302060701
190302051404
190302051405
190302051406
190302060907
190302060914
190302060905
190302060903
190302060904
190302060302
190302060902
190302060901
190302060104
190302060205
190302060206
Total length across all HUCs
Percentage across all HUCs
Proximity to Streams, Wetlands, Ponds, and Small Lakes
Not nearby
(km)
0.1
1.5
3.8
2.3
1.1
0.0
22.1
0.9
0.7
0.3
1.5
0.8
0.7
0.3
1.9
7.3
45.4
40%
Within 200 m
(km)
7.3
3.1
7.0
1.1
5.4
3.3
15.5
3.5
2.7
2.8
1.0
0.5
1.2
3.1
4.5
5.2
67.3"
60%
Total
(km)
7.4
4.6
10.9
3.4
6.5
3.3
37.7
4.4
3.4
3.0
2.5
1.4
2.0
3.4
6.4
12.5
113
100%
Notes:
HUC = hydrologic unit code.
a Reported length is the sum of the road length within 200 m of a National Hydrography Dataset stream or National Wetlands Inventory wetland reported in Tables 10-3 and 10-4, respectively. In cases
where the same section of road is near both types of water bodies, section is only reported once. Therefore total length is less than sum of lengths in Tables 10-3 and 10-4.
Bristol Bay Assessment
10-19
January 2014
-------
Chapter 10 Transportation Corridor
10.3.1 Filling and Alteration of Wetlands, Ponds, and Small Lakes
10.3.1.1 Exposure
Approximately 10% (12 km) of the transportation corridor would intersect mapped wetlands, ponds,
and small lakes (Table 10-4). An additional 24% (27 km) would be located within 100 m of these
habitats, and another 16% (19 km) would be located within 100 to 200 m (Table 10-4). In total,
approximately 51% (58 km) of the corridor would fill or otherwise alter wetlands, ponds, and small
lakes. These habitats encompass 2.3 km2 (1.6, 0.1, and 0.6 km2 of wetlands, ponds, and small lakes,
respectively), or nearly 11% of the total area within 100 m of the transportation corridor. The area of
NWI-mapped aquatic habitats within 200 m of the corridor would be 4.7 km2 (3.3, 0.2, and 1.2 km2 of
wetlands, ponds, and small lakes, respectively). These areas do not include NWI-mapped aquatic
habitats that would be covered by the mine footprints in the mine scenarios (Chapter 7). The area of
these habitats filled by the roadbed would be 0.11 km2 (i.e., approximately 12 km of road, assuming a
road width of 9 m).
10.3.1.2 Exposure-Response
The distribution of salmonids in wetlands, ponds, and small lakes along the transportation corridor is
not known. However, these aquatic habitat losses can result in the loss of resting habitat for adult
salmonids and of spawning and rearing habitat in ponds and riparian side channels. These habitats can
provide refuge habitats (Brown and Hartman 1988) and important rearing habitats for juvenile
salmonids by providing hydraulically and thermally diverse conditions. In addition, by damming and
diverting surface flow and inhibiting subsurface flow, road construction could block or limit access by
fish to important habitats. Beaver ponds associated with small streams, ponds, and wetlands can be
important winter refugia for coho salmon (Nickelson et al. 1992, Cunjak 1996). Beaver ponds provide
high-quality habitat for salmon rearing, because they provide macrophyte cover, low-flow velocity, and
increased temperatures and trap organic materials and nutrients (Nickelson et al. 1992, Collen and
Gibson 2001, Lang et al. 2006).
These habitats can also provide enhanced foraging opportunities (Sommer et al. 2001). Floodplain
wetlands and ponds can be an important contributor to the abundance and diversity of food (and
foodwebs) upon which salmon depend (Opperman et al. 2010). Within aquatic habitats that are not
blocked and are still accessible, the road bed could alter hydrology and flow paths from these habitats to
the stream network. These alterations could mobilize minerals and stored organic carbon, and expose
soils to new wetting, drying, and leaching regimes, thereby leading to changes in vegetation, nutrient
and salt concentrations, and water quality (Ehrenfeld and Schneider 1991). These changes in wetland
dynamics and structure could affect the availability of these habitats to fish and the contribution of
nutrients, organic material, and a diverse array of macroinvertebrates from headland wetlands to higher
order streams in the watershed (i.e., streams receiving wetland drainage) and downstream waters
(Shaftel et al. 2011, Dekar et al. 2012, King et al. 2012, Walker et al. 2012).
Bristol Bay Assessment 1020 January 2014
-------
Chapter 10 Transportation Corridor
10.3.1.3 Risk Characterization
Filling wetlands would eliminate habitat for salmonids and would indirectly alter wetlands in ways that
could reduce the quality, quantity, and accessibility of habitat for fish. Effects on fish production cannot
be estimated given available data; however, the loss of long riparian side channels to culvert or bridge
crossings that do not span the entire floodplain could be locally significant. These wetlands provide
important spawning and rearing habitats and resting areas for migrating adults. Other wetlands such as
shallow ponds may also provide habitat, but all wetlands serve to moderate variation in flow and
maintain water quality.
10.3.2 Stream Crossings
The transportation corridor would cross approximately 64 streams in the Kvichak River watershed. Of
these streams, 20 are listed as supporting anadromous fish in the AWC (Johnson and Blanche 2012) at
the crossing (Table 10-6). An additional 35 are likely to support salmonids (Table 10-6), and a number
of these are anadromous downstream of the crossing. In total, the transportation corridor would cross
55 streams known or likely to support salmonids.
The physical effects of roads on streams and rivers often propagate long distances from actual stream
crossings, because of the energy associated with moving water (Richardson et al. 1975). Thus, alteration
of hydrology and sediment deposition by road crossings can change channels or shorelines many
kilometers away. The transportation corridor could affect 272 km of stream between its road crossings
and Iliamna Lake (Table 10-7). Fish may also be affected in the approximately 780 km of streams
upstream of the transportation corridor that are likely to support salmonids (based on surveys and
stream gradients less than 12%, Table 10-8). In this assessment, we assume streams with segment
gradients less than 12% both downstream and upstream of the corridor-stream crossing are likely to
support salmonids (i.e., salmon, rainbow trout, or Dolly Varden). The amount of upstream length that
may be salmonid habitat is calculated as stream length to the first reach segment with a gradient greater
than 12%. This criterion is used as an upstream limit for salmonid habitat, as Dolly Varden can be
dispersed across a wide range of channel gradients (Wissmar et al. 2010) and have been observed in
higher-gradient reaches (average 12.9% gradient) throughout the year in southeastern Alaska (Bryant
et al. 2004).
Bristol Bay Assessment 1021 January 2014
-------
Chapter 10
Transportation Corridor
Table 10-6. Road-stream crossings along the transportation corridor, upstream lengths of streams of different sizes likely to support
salmonids (based on stream gradients of less than 12%), and downstream lengths to Iliamna Lake. Bold reach codes are those assumed to
be bridged.
HUC-12 Name or Description
Headwaters Upper Talarik
Creek
Upper Tributary to Upper
Talarik Creekb
Tributary to Newhalen River0
Headwaters Newhalen River
Outlet Newhalen River
Roadhouse Creek
Iliamna Lake-Eagle Bay
Eagle Bay Creek
Youngs Creek Mainstem
(Roadhouse Mountain HUC)
NHD Reach Code at Road-
Stream Crossing
19030206007354
19030206007015
19030206007159
19030206007175
19030205007587
19030205007593
19030205007598
19030205007606
19030205007602
19030205007615
19030205000002
19030205013069
19030205013055
19030205013057
19030205013041
19030206010623
19030206010628
19030206010629
19030206006712
19030206006678
19030206006677
19030206006644
19030206006671
19030206006663
19030206006654
19030206006598
AWC
(*Salmonid
Potential)
Y*
Y*
Y*
N *
N*
N *
N *
Y*
Y*
N*
Y*
N
N*
N *
N *
N*
N *
N*
N
Y*
N
N *
N *
Y*
Y*
Y*
Upstream Fish Habitat Length (km)
Small
Headwater
Streams3
3.5
97.4
1.4
3.7
5.7
3.8
3.6
6.8
2.8
3.1
67.7
0.0
6.2
1.8
3.2
0.7
0.4
0.7
0.0
0.9
0.0
1.6
0.4
11.3
4.0
25.7
Medium
Streams3
0.0
37.6
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
45.2
0.0
2.6
0.0
0.0
0.0
0.0
0.0
0.0
1.5
0.0
0.0
5.5
0.0
0.0
16.3
Small
Rivers3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Large
Rivers3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
13.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total
3.5
134.9
1.4
3.7
5.7
3.8
3.6
6.8
2.8
3.1
126.1
0.0
8.8
1.8
3.2
0.7
0.4
0.7
0.0
2.4
0.0
1.6
5.9
11.3
4.0
42.0
Downstream Length
to Iliamna Lake
(km)
57.6
57.0
55.6
66.0
45.9
41.7
44.5
37.2
34.8
29.4
26.4
1.1
1.3
3.7
3.7
2.4
3.6
2.2
15.7
9.6
10.3
11.1
6.4
6.3
6.4
10.4
Bristol Bay Assessment
10-22
January 2014
-------
Chapter 10
Transportation Corridor
Table 10-6. Road-stream crossings along the transportation corridor, upstream lengths of streams of different sizes likely to support
salmonids (based on stream gradients of less than 12%), and downstream lengths to Iliamna Lake. Bold reach codes are those assumed to
be bridged.
HUC-12 Name or Description
Youngs Creek East Branchd
Chekok Creek
Canyon Creek
Iliamna Lake-Knutson Bay
Knutson Creek
NHD Reach Code at Road-
Stream Crossing
19030206006553
19030206006533
19030206032854
19030206006359
19030206006336
19030206006337
19030206006236
19030206006331
19030206006329
19030206006327
19030206006325
19030206006322
19030206006320
19030206006321
19030206006318
19030206006317
19030206006316
19030206006315
19030206006314
19030206006251
19030206006255
19030206006280
19030206006239
19030206006248
19030206006231
19030206006230
19030206006228
19030206006227
AWC
(*Salmonid
Potential)
Y*
Y*
Y*
Y*
N*
N *
N *
N*
N *
N*
N*
N
N*
N*
N
N
N *
N *
N*
N *
Y*
N*
N
N*
N
N
Y*
N *
Upstream Fish Habitat Length (km)
Small
Headwater
Streams3
32.9
5.8
36.1
0.0
4.4
0.3
1.0
0.6
0.6
0.2
0.8
0.0
0.1
0.5
0.0
0.0
0.5
0.7
0.7
0.6
0.1
0.4
0.0
0.3
0.0
0.0
0.0
0.0
Medium
Streams3
12.4
0.0
42.5
1.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
3.2
0.0
0.0
0.0
0.0
0.0
0.3
0.9
Small
Rivers3
0.0
0.0
7.9
8.6
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.9
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Large
Rivers3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total
45.3
5.8
86.6
9.8
4.4
0.3
1.0
0.6
0.6
0.2
0.8
0.0
0.1
0.5
0.0
0.0
0.5
0.7
0.7
0.6
5.2
0.4
0.0
0.3
0.0
0.0
0.3
0.9
Downstream Length
to Iliamna Lake
(km)
9.0
5.0
8.4
12.1
3.8
3.6
3.4
4.2
3.9
1.9
2.6
0.1
0.7
0.7
0.8
0.9
0.5
0.6
0.7
1.7
4.4
4.4
2.5
4.7
0.6
0.4
1.5
3.0
Bristol Bay Assessment
10-23
January 2014
-------
Chapter 10
Transportation Corridor
Table 10-6. Road-stream crossings along the transportation corridor, upstream lengths of streams of different sizes likely to support
salmonids (based on stream gradients of less than 12%), and downstream lengths to Iliamna Lake. Bold reach codes are those assumed to
be bridged.
HUC-12 Name or Description
Outlet Pile River
Middle Iliamna River
Chinkelyes Creek
NHD Reach Code at Road-
Stream Crossing
19030206006222
19030206000474
19030206010632
324-10-10150- 2343-30066
19030206000032
19030206005773
19030206005761
19030206005759
19030206005754
19030206005737
AWC
(*Salmonid
Potential)
N *
Y*
Y*
Y*
Y*
N *
N *
N*
N *
N*
Upstream Fish Habitat Length (km)
Small
Headwater
Streams3
0.0
34.1
4.2
Medium
Streams3
3.4
24.9
0.0
Small
Rivers3
0.0
50.0
0.0
Large
Rivers3
0.0
0.0
0.0
Total
3.4
109.0
4.2
NO NHD DATA
27.9
0.3
0.5
0.4
0.7
0.0
36.5
0.0
2.7
0.0
1.0
8.5
40.6
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
104.9
0.3
3.2
0.4
1.7
8.5
Downstream Length
to Iliamna Lake
(km)
6.3
5.7
0.9
1.0
10.2
13.4
14.5
18.0
21.6
22.1
Notes:
Values (lengths) are arranged by 12-digit HUCfrom west (top) to east (bottom) along the transportation corridor. Each upstream value is a sum of NHD stream segment lengths in the HUCs between the
crossing and upper extent of salmonid habitat potential based on 12% gradient. Each downstream value is a sum of stream segment lengths in the HUCs between the crossing and Iliamna Lake.
Because the lengths at each crossing represent contiguous lengths, a portion of stream may be included in more than one crossing.
a Small headwater streams = 0-0.15 m3/s; medium streams = 0.15-2.8 m3/s; small rivers = 2.8-28 m3/s; large rivers = >28 m3/s.
b 190302060701.
c 190302051404.
d 190302060904.
e AWC stream code used, because no corresponding NHD stream code (and no upstream habitat data) available.
NHD = National Hydrography Dataset; AWC = Anadromous Waters Catalog; HUC = hydrologic unit code.
Source: AWC data from Johnson and Blanche (2012); NHD data from USGS (2012).
Bristol Bay Assessment
10-24
January 2014
-------
Chapter 10
Transportation Corridor
Table 10-7. Stream lengths downstream of road-stream crossings, classified by stream size. Stream size was based on mean annual
streamflow; downstream length was measured from the road-stream crossing to Iliamna Lake.
HUC-12 Name or Description
Headwaters Upper Talarik Creek
Upper Tributary to Upper Talarik Creekb
Tributary to Newhalen River0
Headwaters Newhalen River
Outlet Newhalen River
Roadhouse Creek
Iliamna Lake-Eagle Bay
Eagle Bay Creek
Youngs Creek Mainstem (Roadhouse Mountain HUC)
Youngs Creek East Branchd
Chekok Creek
Canyon Creek
Iliamna Lake-Knutson Bay
Knutson Creek
Iliamna Lake-Pedro Bay
Iliamna Lake-Pile Bay
Outlet Pile River
Middle Iliamna River
Chinkelyes Creek
Total length across all HUCS
Percentage across all HUCS
Downstream Length (km)
Small Headwater
Streams3
2.1
0.8
4.1
0.9
3.0
11.4
4.4
2.8
0.0
0.8
2.9
4.8
16.0
1.8
6.8
3.5
1.2
0.0
1.3
68.6
25%
Medium Streams3
9.0
8.3
14.5
0.0
1.3
11.4
11.9
8.1
4.2
8.0
0.0
0.0
2.9
0.0
5.5
4.5
0.7
0.7
4.4
95.4
35%
Small Rivers3
36.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.8
6.5
0.0
2.9
0.0
0.0
3.2
10.2
10.7
75.7
28%
Large Rivers3
0.0
0.0
0.0
8.3
23.7
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
32.0
12%
Total
47.6
9.1
18.6
9.2
28.0
22.8
16.3
10.9
4.2
8.7
8.7
11.3
18.9
4.6
12.3
8.0
5.2
10.9
16.4
272
100%
Notes:
Values (lengths) are arranged by 12-digit HUC, from west (top) to east (bottom) along the transportation corridor. Downstream values are the sum of National Hydrography Dataset stream segment
lengths in the HUCs between the crossing and Iliamna Lake.
a Small headwater streams = 0-0.15 m3/s; medium streams = 0.15-2.8 m3/s; small rivers = 2.8-28 m3/s; large rivers = >28 m3/s.
b 190302060701.
0 190302051404.
d 190302060904.
HUC = hydrologic unit code.
Bristol Bay Assessment
10-25
January 2014
-------
Chapter 10
Transportation Corridor
Table 10-8. Lengths of different stream sizes that occur upstream of road-stream crossings and are likely to support salmonids (based on
stream gradients of less than 12%).
HUC-12 Name or Description
Headwaters Upper Talarik Creek
Upper Tributary to Upper Talarik Creekb
Tributary to Newhalen River0
Headwaters Newhalen River
Outlet Newhalen River
Roadhouse Creek
Iliamna Lake-Eagle Bay
Eagle Bay Creek
Youngs Creek Mainstem (Roadhouse Mountain HUC)
Youngs Creek East Branchd
Chekok Creek
Canyon Creek
Iliamna Lake-Knutson Bay
Knutson Creek
Iliamna Lake-Pedro Bay
Iliamna Lake-Pile Bay
Outlet Pile River
Middle Iliamna River
Chinkelyes Creek
Total length across all HUCS
Percentage across all HUCS
Upstream Fish Habitat Length (km)
Small Headwater
Streams3
69.5
36.5
37.7
55.8
11.9
1.7
2.4
15.6
25.7
32.9
41.9
0.0
11.0
0.6
0.3
0.0
38.3
27.9
1.8
411.7
53%
Medium Streams3
17.8
19.7
15.9
29.3
2.6
0.0
1.5
5.5
16.3
12.4
42.5
1.2
0.0
3.2
0.0
1.2
28.3
36.5
12.2
246.2
31%
Small Rivers3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
7.9
8.6
0.0
1.9
0.0
0.0
50.0
40.6
0.1
109.1
14%
Large Rivers3
0.0
0.0
0.0
13.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
13.1
2%
Total
87.4
56.2
53.6
98.2
14.5
1.7
4.0
21.2
42.0
45.3
92.3
9.8
11.0
5.7
0.3
1.2
116.6
104.9
14.1
780.1
100%
Notes:
Values (lengths) are arranged by 12-digit HUC, from west (top) to east (bottom) along the transportation corridor. Each upstream value is a sum of National Hydrography Dataset stream segment lengths
in the HUCs between the crossing and upper extent of salmonid habitat potential based on 12% gradient.
a Small headwater streams = 0-0.15 m3/s; medium streams = 0.15-2.8 m3/s; small rivers = 2.8-28 m3/s; large rivers = >28 m3/s.
b 190302060701.
0 190302051404.
d 190302060904.
HUC = hydrologic unit code.
Bristol Bay Assessment
10-26
January 2014
-------
Chapter 10 Transportation Corridor
10.3.2.1 Exposure
Based on the assumption that crossings over streams with mean annual streamflows greater than 0.15
m3/s would be bridged (Section 6.1.3), the transportation corridor would include 19 bridges, 12 over
known anadromous streams and 7 over streams likely to support salmonids (Table 10-6). Mean annual
streamflow at a crossing in the Eagle Bay Creek hydrologic unit code (HUC)-12 (reach code
19030206006663) was 0.14 m3/s, but we assumed that this crossing would be bridged because the
stream is anadromous and contains 11.3 km of upstream fish habitat. Culverts would be placed at all
other stream crossings. Given that the transportation corridor would cross a total of 55 streams and
rivers known or likely to support migrating or resident salmonids, culverts would be constructed on 36
presumed salmonid streams.
Bridges would generally have fewer impacts on salmon than culverts, but could result in the loss of long
riparian side channels if they did not span the entire floodplain. Approximately 500,000 bridges listed in
the National Bridge Inventory are built over streams, and many of these, especially those on more active
streams, will experience problems with aggradation, degradation, bank erosion, and lateral channel shift
during their useful life (FHWA 2012).
Where flow restrictions such as culverts are placed in stream channels, stream power increases. This
can lead to increased channel scouring and down-cutting, streambank erosion, and undermining of the
road. Salmonids and other riverine fishes actively move into seasonal floodplain wetlands and small
valley floor tributaries to escape the stresses of main-channel flood flows (Copp 1989). Culverts can
reduce flow to these habitats by funneling flow from the entire floodplain through the culvert and into
the main channel. High water velocities in a stream channel may result from storm and snowmelt flows
being forced through a culvert rather than spreading across the floodplain. Higher velocities cause
scouring and down-cutting of the channel downstream of the culvert. This downstream erosion can
result in perched culverts, impairing fish access to upstream reaches. In addition, it can hydrologically
isolate the floodplain from the channel and block fish access to floodplain habitat. Entrenchment of the
channel also prevents fish from reaching slow-water refugia during high-flow events and reduces
nutrient and sediment cycling processes between the stream channel and the floodplain. Lastly, channel
entrenchment may cause a change in the water table and the extent of the hyporheic zone, with
consequences for floodplain water-body connectivity and water temperatures in the floodplain habitat.
Culverts are deemed to have failed if fish passage is blocked (e.g., by debris, ice, beaver activity, or
culvert perching) or if streamflow exceeds culvert capacity and results in overtopping and road
washout. Reported culvert failure frequencies vary in the literature but are generally high. Values of
30% (Price et al. 2010), 53% (Gibson et al. 2005), and 61% (Langill and Zamora 2002) have been
reported, for an average culvert failure estimate of 48% (i.e., culvert surveys indicate that, on average,
48% block or inhibit fish passage at any given time).
When culverts are plugged by debris or overtopped by high flows, road damage, channel realignment,
and severe sedimentation often result (Furniss et al. 1991). Changes in sediment load due to culvert
Bristol Bay Assessment 1027 January 2014
-------
Chapter 10 Transportation Corridor
failures can change stream hydraulics and geomorphic pressures. Generally, habitat value in the stream
is diminished as the channel becomes wider and shallower and silt is deposited in the streambed.
Stream crossing failures that divert streamflow outside of stream channels are particularly damaging
and persistent (Weaver et al. 1987).
Free access to spawning and early rearing habitat in headwater streams is critical for a number of fish
species, and culverts are common migration barriers. Culvert blockages are usually caused by woody
debris and sometimes by woody material used by beavers to block a culvert and create a pond. In
addition, aufeis—an ice feature that forms when water in or adjacent to a stream channel rises above
the level of an existing ice cover and gradually freezes to produce a thickened ice cover (Slaughter
1982)—can completely fill culverts. When this occurs, water will run over the roadway unless flow is
initiated through the culvert (Kane and Wellan 1985). The ice also reduces the cross-sectional area of
flow so that high headwater conditions (and higher velocities than indicated by the culvert design) are
produced during periods of peak flow. In some cases, considerable ice remains after the breakup period,
particularly upstream of the culvert in the channel and floodplain (Kane and Wellan 1985).
Blockages could persist for as long as the intervals between culvert inspections. We assume that the
transportation corridor would receive daily inspection and maintenance during operation of the mine.
The level of surveillance along the corridor can be expected to affect the frequency of culvert failure
detection. Driving inspections would likely identify a single erosional failure of a culvert that damaged
the road or debris blockage sufficient to cause water to pool about the road, and in such cases temporary
repairs would be made to protect the road. However, long-term fixes may not be possible until
conditions are suitable to replace a culvert or bridge crossing. Further, multiple failures such as might
occur during an extreme precipitation event would likely take longer to repair. These fixes may not fully
address fish passage, which may be reduced or blocked for longer periods. Also, some failures that
would reduce or block fish passage (e.g., gradual downstream channel erosion resulting in a perched
culvert) might not be noticed by a driving inspection. Thus, blockage of migration could persist for an
extended period. Extended blockage of migration would be less likely if daily road inspections included
stops to inspect both ends of each culvert.
After mine operations end, traffic would decrease to that which is necessary to maintain any residual
operations on the site, and inspections and maintenance would likely decrease. If the road was adopted
by the state or local government, the frequency of inspections and quality of maintenance would likely
decline to those provided for other roads. Either of these possibilities could result in a proportion of
failed culverts similar to those described in the literature.
10.3.2.2 Exposure-Response
Blockage of a culvert by debris or downstream erosion would inhibit the upstream and downstream
migration of salmon and the movement of other fish among seasonal habitats. The effects of a blockage
would depend on its timing and duration. A blockage would result in the loss of spawning and rearing
habitat if it occurred during adult migration periods and persisted for several days. It could cause the
Bristol Bay Assessment 1028 January 2014
-------
Chapter 10 Transportation Corridor
loss of a year class of salmon from a stream if it occurred during juvenile migration periods and
persisted for several days or more.
Erosional failure of a road resulting from failure of a culvert would create suspended sediment that
would be carried and deposited downstream. Relationships between the concentration and duration of
elevated sediment concentrations and effects on fish and invertebrates are presented in Section 9.4.2.1.
10.3.2.3 Risk Characterization
The mine scenarios specify that culverts would be installed along the transportation corridor with
adequate size for normal flows of the streams crossed, and that the roadway would be monitored daily
to ensure that failures could be rapidly identified and repaired. Even with these assumptions, inhibition
offish passage and reductions in habitat still could occur. Although culverts would be designed to
certain specifications (Box 10-2), they are not always installed correctly or do not stand up to the rigors
of a harsh environment, as indicated by the failure frequencies cited in Section 10.3.2.1. The
transportation corridor would traverse varied terrain and subsurface soil conditions, including
extensive areas of rock excavation in steep, mountainous terrain where storm runoff can rapidly
accumulate and result in intense local runoff conditions (Ghaffari et al. 2011). Although the road design,
including placement and sizing of culverts, would take into account seasonal drainage and spring runoff
requirements, culvert failures would still be expected. For example, heavy rains in late September 2003
washed out sections of the Williamsport-Pile Bay Road (Lake and Peninsula Borough 2009), and
culverts on this road have been washed out on numerous occasions (PLP 2011: Appendix 7.3A).
Culverts are not always built to specifications and the behavioral responses of migrating salmonid life
stages to culvert-induced changes in flow are not always anticipated correctly. Standards for culvert
installation on fish-bearing streams in Alaska mainly consider fish passage (ADF&G and ADOT 2001).
Additional factors unrelated to fish passage, such as the physical structure of the stream or habitat
quality, are addressed on a project-specific basis during preparation of the Alaska Department of
Transportation and Public Facilities environmental document. Culvert capacities are allowed to be less
than channel capacity (ADF&G and ADOT 2001). In most cases culvert width must be greater than 90%
of the ordinary high-water channel width, but where channel slope is less than 1.0% culvert width must
only be greater than 75% of the ordinary high-water channel width. During flood flows, this reduced
channel width results in slower than normal velocities upstream of the culvert and higher water
velocities exiting the culvert. This could result in scoured downstream channel beds, altered channel
dynamics, and disassociated channels and floodplains. These processes would reduce the capacity of
downstream reaches to support salmonids. High flows in and immediately downstream of the culvert, as
well as the structure of the culvert itself, could inhibit fish passage even if movement is not blocked.
Downstream erosion could result in perched culverts that, if they were not inspected and maintained,
would inhibit and ultimately block fish passage. Floodplain habitat and floodplain/channel ecosystem
processes could be disrupted by channel entrenchment resulting from culvert-induced erosion. These
potential reductions in downstream habitat quality and inhibited fish passage could occur in any of the
36 culverted streams that likely support salmonids.
Bristol Bay Assessment 1029 January 2014
-------
Chapter 10 Transportation Corridor
BOX 10-2. CULVERT MITIGATION
Bridge or culvert installation and maintenance activities in fish-bearing water bodies require a fish habitat
permit. Permit application information requirements for culvert installations in fish streams are detailed in a
memorandum of agreement (MOA) between the Alaska Department of Fish and Game (ADF&G) and the Alaska
Department of Transportation and Public Facilities (ADOT) (ADF&G and ADOT 2001). The MOA provides
guidance to project designers and permitting staff to ensure that culverts are designed and installed to provide
efficient fish passage and to ensure statewide consistency in Title 16 permitting of culvert related work. Title
16 is the statute by which the ADF&G performs Fish Habitat and Special Area permitting.
Fish habitat regulations under Title 16 include the Anadromous Fish Act and the Fishway (or Fish Passage) Act.
• The Anadromous Fish Act (AS 16.05.871-.901) requires that an individual or government agency
provide prior notification and obtain permit approval from ADF&G before altering or affecting "the
natural flow or bed" of a specified water body or fish stream. All activities within or across a specified
anadromous water body—including construction; road crossings; gravel removal; mining; water
withdrawals; the use of vehicles or equipment in the waterway; stream realignment or diversion; bank
stabilization; blasting; and the placement, excavation, deposition, or removal of any material—require
approval from ADF&G's Division of Habitat.
• The Fishway (or Fish Passage) Act (AS 16.05.841), requires that an individual or government agency
notify and obtain authorization from the ADF&G's Division of Habitat for activities within or across a
stream used by fish if it is determined that such uses or activities could represent an impediment to
the efficient passage of resident or anadromous fish.
The MOA describes the procedures, criteria and guidelines used for permitting culvert related work in fish-
bearing waters; these criteria augment but do not replace ADOT's standard design criteria presented in the
Alaska Highway Drainage Manual (ADOT 1995). Culverts are designed and permitted using one of the following
design approaches.
• Tier I—Stream Simulation Design (developed by the U.S. Department of Agriculture, Forest Service
[FSSSWG 2008]). The Tier 1 approach most clearly replicates natural stream conditions, and is
applicable in stream gradients less than 6%. Using this design, culverts a re sized larger than culverts
sized hydraulically for floodwater conveyance alone. The culvert width at the ordinary high water
(OHW) stage waterline must be greater than 90% of the OHW width. The culvert grade should
approximate the channel slope, but in no instance should it deviate more than 1% from the natural
grade. In stream channels with slopes less than 1%, culverts may be installed at slopes less than
0.5% with culvert widths greater than 75% of the OHW width.
• Tier II—FISHPASS Program Design. Under this approach, culverts are designed using a combination
of traditional hydraulic engineering methods and the Alaska Interagency Fish Passage Task Force's
1991 "FISHPASS" computer modeling program (Behlkeetal. 1991). The FISHPASS program
evaluates component hydraulic forces in a culvert against a fish's available power and energy
capabilities.
• Tier III—Hydraulic Engineering Design. The Tier III approach is used where site-specific conditions
preclude use of Tier I and Tier II designs. Under this approach, professionally recognized hydraulic
engineering methods are used to ensure appropriate fish passage characteristics in the culvert.
Culverts and other road crossings that do not provide free passage between upstream and downstream
reaches can fragment populations into small demographic isolates vulnerable to extinction (Hilderbrand
and Kershner 2000, Young et al. 2005). In a study of natural long-term isolates of coastal cutthroat trout
and Dolly Varden in southeastern Alaska, Hastings (2005) found that about 5.5 km of perennial
headwater stream habitat, supporting a census population size of greater than 2,000 adults, is required
for a high likelihood of long-term population persistence. Table 10-6 shows that, of the 55 known or
likely salmonid-supporting streams that would be crossed by the transportation corridor, 39 contain
less than 5.5 km of habitat (stream length) upstream of the proposed road crossings. These 39 stream
Bristol Bay Assessment 1030 January 2014
-------
Chapter 10 Transportation Corridor
crossings contain a total of 68 km of upstream habitat and 493 km of downstream habitat. Seven of
these crossings would be bridged, leaving 32 with culverts. Assuming typical maintenance practices
after mine operations, roughly 48% of these streams, or 15 streams, would be entirely or partially
blocked at any one time. As a result, these streams would likely not be able to support long-term
populations of resident species such as rainbow trout or Dolly Varden.
The risk of culvert failures is somewhat uncertain due to the paucity of literature on culvert failures both
in Alaskan taiga and tundra and for modern mining roads crossing salmonid habitat. The most relevant
studies on potential effects of roads, particularly as they relate to salmon, are from forest and rangeland
roads. These roads may differ in important ways from mining roads. Forested streams inevitably carry
more woody debris that could block culverts. However, forested vegetation types represent 68% of the
potential transportation corridor area mapped by Pebble Limited Partnership (PLP) (2011: Chapter 13).
Mine roads carry much heavier loads than logging roads, but would likely be better engineered. For
example, the transportation corridor in this assessment would be designed to support 190-ton haul
truck travel on the road surface (Ghaffari et al. 2011), compared to an average gross legal weight limit of
approximately 44 tons per log truck (Mason et al. 2008). In any case, the culvert failure frequencies cited
in this assessment are from modern roads and not restricted to forest roads, and represent the most
relevant data available.
10.3.3 Chemical Contaminants
In this section we address three sources of potentially toxic chemicals related to the transportation
corridor: traffic residues, road construction, and road treatment and chemical cargos.
During runoff events, traffic residues produce a contaminant mixture of metals (e.g., lead, zinc, copper,
chromium, and cadmium), oil, and grease that can get washed into streams and accumulate in sediments
(Van Hassel et al. 1980) or disperse into groundwater (Van Bohemen and Van de Laak 2003). It is
unclear if the transportation corridor would have sufficient traffic to contaminate runoff with significant
amounts of metals or oil (although stormwater runoff from roads at the mine site itself is more likely to
contain metal concentrations sufficient to affect stream water quality). Therefore, this risk is not
considered further.
Road construction involves the crushing of minerals for the road fill and bed and the exposure of rock
surfaces at road cuts, which leads to leaching of minerals and increased dissolved solids. Fish mortality
in streams, with effects on populations recorded as far as 8 km downstream, has been related to high
concentrations of aluminum, manganese, copper, iron, or zinc from highway construction activities in
geological formations containing pyritic materials (Morgan et al. 1983). Because it is not clear where
materials for the road will come from or their composition, this risk is not considered further.
Two potentially significant contaminants of aquatic habitats may occur along the transportation
corridor: chemicals released during spills from truck accidents and stormwater runoff of salts or other
materials used for winter road treatment. It should also be noted that increased runoff associated with
roads may increase rates and extent of erosion, reduce percolation and aquifer recharge rates, alter
Bristol Bay Assessment 1031 January 2014
-------
Chapter 10 Transportation Corridor
channel morphology, and increase stream discharge rates (Forman and Alexander 1998). These effects
of stormwater runoff are not assessed, however, because they are highly location-specific and not
quantifiable given available data. Increases in sediment associated with stormwater runoff are
addressed in Section 10.3.4.
10.3.3.1 Exposure
Many chemical reagents would be used to process ore (Box 4-5), and these chemicals would be
transported by road to the mine site. Truck accidents along the transportation corridor could spill
reagents into wetlands or streams. To estimate how much reagent and thus how many transport trucks
would be needed for the mine scenarios, we extrapolated from the number of trucks required to
transport reagents at a smaller gold mine (175 trucks per year at Pogo Mine) to the mine scenarios,
based on the relative annual ore production at the two mines. Assuming 20 tons of reagent per truck and
expected annual production rates of 3,000 tons per day at Pogo Mine (USEPA 2003a) and 200,000 tons
per day in the mine scenarios (Ghaffari et al. 2011), we estimate that transport of reagents would
require approximately 11,725 truck trips per year.
The length of the transportation corridor within the Kvichak River watershed would be 113 km. The
probability of truck accidents and releases was reported as 1.9 x 1Q-7 spills per mile of travel for a rural
two-lane road (Harwood and Russell 1990). Based on this rate, the number of spills over the roughly 25-
year life of the Pebble 2.0 scenario would be 3.9—that is, approximately 4 spills from truck accidents
would be expected during mine operations. Over the roughly 78-year life of the Pebble 6.5 scenario, 12
spills would be expected. Only one-way travel is considered, because return trips from the mine would
be with empty trucks or with a load other than process reagents. Because conditions on the mine road
would be different from those for which the statistics were developed (e.g., more difficult driving and
road conditions), this calculation provides an order of magnitude estimate. The reasonableness of this
estimate is suggested by an assessment of the Cowal Gold Project in Australia, which estimated that a
truck wreck would occur every 1 to 2 years, resulting in a spill every 3 to 6 years (NICNAS 2000).
For 14% of its length (15 km), the transportation corridor would be within 100 m of a stream or river
(Table 10-3), and for 24% of its length it would be within 100 m of a mapped wetland (Table 10-4). If
the probability of a chemical spill is independent of location, and if it is assumed that liquid spills within
100 m of a stream could flow to that stream, a spill would have a 14% probability of entering a stream
within the Kvichak River watershed. This would result in roughly 0.5 stream-contaminating spills over
the 25-year life of the Pebble 2.0 scenario or up to 2 stream-contaminating spills over the 78-year life of
the Pebble 6.5 scenario. Similarly, a spill would have a 24% probability of entering a wetland, resulting
in an estimate of 1 wetland-contaminating spill in the Pebble 2.0 scenario or 3 wetland-contaminating
spills in the Pebble 6.5 scenario. A portion of those wetlands would be ponds or backwaters that support
fish. It should be noted that the risk of spills could be somewhat mitigated by using spill-resistant
containers.
Cyanide for gold processing would be transported as a solid. We assume containment equivalent to that
at the Pogo mine (i.e., dry sodium cyanide pellets inside plastic bags inside wooden boxes inside metal
Bristol Bay Assessment 1032 January 2014
-------
Chapter 10 Transportation Corridor
shipping containers). Hence, even in a truck wreck, a cyanide spill is an unquantifiable but low
probability occurrence. A spill on land could be collected, but during periods of rain or snowmelt it
would rapidly dissolve and wash into surface or groundwater. A spill of pellets into a stream or wetland
would rapidly dissolve and dissociate into free ions or, depending on the pH, hydrogen cyanide. Pellets
spilled into a stream would be transported downstream as described for the copper concentrate
(Section 11.3), but, rather than slurry water and solids, the transported material would consist of
dissolving pellets and increasing cyanide or hydrogen cyanide solution.
In addition to process chemicals, the molybdenum concentrate (primarily molybdenum sulfide) would
be transported by truck. The concentrate would be a dewatered fine granular material contained in bags
packed in shipping containers. Thus, as with cyanide, a spill of molybdenum concentrate is an
unquantifiable but low probability occurrence. A spill on land could be collected. A spill into water
would be transported by streamflow as described for the copper concentrate (Section 11.3). Settled
concentrate would oxidize, forming acidic pore water with dissolved molybdenum to which benthic
invertebrates and fish eggs and larvae could be exposed.
Roads are treated with salts and other materials to reduce dust and improve winter traction. In Alaska,
calcium chloride is commonly used for dust control and is mixed with sand for winter application.
During periods of rain and snowmelt, these materials are washed off roads and into streams, rivers, and
wetlands, where fish and their invertebrate prey can be directly exposed. We found no relevant data for
calcium chloride levels in runoff or streams from roads treated in this way.
10.3.3.2 Exposure-Response
A principle processing chemical of concern would be sodium ethyl xanthate (Section 6.4.2.3). A risk
assessment by Environment Australia estimated that a spill of as little as 10% of a 25-metric-ton-
capacity truck carrying sodium ethyl xanthate into a stream would require a "650000:1 dilution before
the potential hazard is considered acceptable" and that the spill could not be mitigated (NICNAS 2000).
Cyanide has acute and chronic U.S. ambient water quality criteria for freshwater of 22 and 5.2 ug free
cyanide per liter. The geometric mean of 30 median lethal concentration (LCso) values from acute tests
of rainbow trout is 55.7 ug/L (USEPA 1985, 2013). In a 2-hour exposure to 10 ug/L cyanide, swimming
speed of coho salmon was reduced (USEPA 1985). Unlike metals, cyanide is not more toxic to
invertebrates than fish. Standard acute endpoints for invertebrates range from 17 to 210,000 ug/L
(USEPA 1985, 2013).
Molybdenum's aquatic toxicity is relatively poorly characterized. The most directly relevant values are
28-day LC50 values for rainbow trout eggs of 730 and 790 ug/L (Birge 1978, Birge et al. 1979). The mean
of two acute lethality tests with rainbow trout is 1,060,000 ug/L (USEPA 2013). Acute and chronic
values for Daphnia are 206,800 and 4,500 ug/L (USEPA 2013). Hence, molybdenum appears to be much
less toxic than copper. However, the small body of test data and lack of information on the influence of
water chemistry on toxicity make judgments about the effects of aqueous molybdenum much more
Bristol Bay Assessment 1033 January 2014
-------
Chapter 10 Transportation Corridor
uncertain than copper or many other metals. Also unlike copper, there are no whole sediment
benchmarks for molybdenum.
Compounds used to control ice and dust (Hoover 1981) have been shown to cause toxic effects when
they run off and enter surface waters. Dissolved calcium, like sodium, has little influence on the toxicity
of dissolved chloride salts (Mount et al. 1997). Based on that study, the toxicity of the calcium chloride
commonly used in Alaska would be expected to be a little greater than the more studied sodium
chloride, based on total chlorine concentrations. Alaska acute and chronic water quality standards for
chloride are 860 and 230 mg/L, respectively (ADEC 2003). However, these values may not provide
adequate protection from calcium salts. In addition, exceedances of the acute criterion could affect many
species, because freshwater biota have a narrow range of acute susceptibilities to chloride (ADEC 2003).
These standards and the associated federal criteria also may not be adequately protective due to the
absence of tests of critical life stages (e.g., egg fertilization).
Rainwater tends to leach out the highly soluble chlorides (Withycombe and Dulla 2006), which can
degrade nearby vegetation, surface water, groundwater, and aquatic species (Environment Canada
2005). Salmonids are sensitive to salinity, particularly at fertilization (Weber-Scannell and Duffy 2007).
According to Bolander and Yamada (1999), application of chloride salts should be avoided within at
least 8 m of water bodies (including shallow groundwater, if significant migration of chloride would
reach the groundwater table), and restricted if low salt-tolerant vegetation occurs within 8 m of the
treated area. Adverse biological effects are likely to be particularly discernible in naturally low-
conductivity waters such as those of the Bristol Bay watershed, but research is needed to substantiate
this (Appendix G).
10.3.3.3 Risk Characterization
Given the liquid form and toxicity of sodium ethyl xanthate (Section 8.2.2.5), it is expected that a spill of
this compound into a stream along the transportation corridor would cause a fish kill. Runoff or
groundwater transport from a more distant spill would cause effects that would depend on the amount
of dilution or degradation occurring before the spilled material entered a stream. Although other
process chemicals would also be used, xanthate is representative of the chemicals estimated to result in
roughly two stream-contaminating spills over the 78-year life of the Pebble 6.5 scenario.
Cyanide pellets spilled by a truck wreck into a stream would be carried by the current but would rapidly
dissolve into a cyanide solution and would ultimately disperse, volatilize, and degrade in Iliamna Lake.
Spills into a wetland would dissolve in place. Spills on land would be collected unless they occurred
during rain or snowmelt, in which case spilled pellets would dissolve and flow to surface or
groundwater. Data needed to derive a cyanide spill scenario and quantify risks are unavailable, but
given the toxicity of cyanide and its rapid action, effects on invertebrates and fish, including death,
would be likely if a substantial spill into a stream or wetland occurred.
Molybdenum concentrate spilled by a truck wreck into a stream would be carried by the current and
deposited in pools and backwaters and ultimately in Iliamna Lake. Compared to copper concentrate,
Bristol Bay Assessment 1034 January 2014
-------
Chapter 10 Transportation Corridor
relatively little is known about molybdenum concentrate. The solubility of the molybdenum in the Aitik
copper concentrate is undefined but appears to be relatively low (Appendix H: Tables H-8 and H-9), and
molybdenum is much less toxic than copper. The frequency of truck passages is also unknown, so the
spill risk is unquantified. Therefore, the ecological risk from a molybdenum spill is unquantifiable but
appears to be low relative to the risk from a copper concentrate spill (Section 11.3).
Risks to salmonids from de-icing salts and dust suppressants could be locally significant, but would
depend on the amount and frequency of application. The transportation corridor would intersect
55 streams and rivers known or likely to support salmonids, and there would be approximately 272 km
of streams between road crossings and Iliamna Lake (Table 10-7). Additionally, approximately 12 km of
roadway would intersect wetlands within and beyond those mapped by NWI. Runoff from these road
segments could have significant effects on fish and the invertebrates that they consume, particularly if
sensitive life stages are present.
10.3.4 Fine Sediment
10.3.4.1 Exposure
During rain and snowmelt, soil eroded from road cuts, borrow areas, road surfaces, shoulders, cut-and-
fill surfaces, and drainage ditches (as well as road dust deposited on vegetation; see Section 10.3.5),
would be washed into streams and other water bodies. Erosion and siltation are likely to be greatest
during road construction. The main variables determining surface erosion are the inherent credibility of
the soil, slope steepness, surface runoff, slope length, and ground cover. Mitigation measures for fine
sediments are discussed in Box 10-3. It is worth noting that improvements have been proposed for the
road between Iliamna and Nondalton, in part to alleviate erosion and sedimentation problems at some
areas along the road (ADOT 2001).
Bristol Bay Assessment 1035 January 2014
-------
Chapter 10 Transportation Corridor
BOX 10-3. STORMWATER RUNOFF AND FINE SEDIMENT MITIGATION
The Alaska Department of Environmental Conservation (ADEC) administers Alaska Pollutant Discharge Elimination
System (APDES) stormwater general permits for construction activities and multi-sector general permits for
industrial operation activities. ADEC also approves stormwater pollution prevention plans (SWPPPs) that include
stormwater best management practices (BMPs).
A permittee covered under the APDES stormwater general permit for construction activities (ADEC 2011a) must
comply with control measures that are determined by site-specific conditions. ADEC developed the Alaska Storm
Water Guide (ADEC 2011b) to assist permittees with selecting, installing, and maintaining control measures that
may be used for projects in Alaska. Erosion and sediment control measures covered under the stormwater
general permit for construction activities (ADEC 2011a) are summarized below.
Erosion Control Measures
• Delineate the site, specifically the location of all areas where land disturbing activities will occur and
areas that will be left undisturbed (e.g., boundaries of sensitive areas or established buffers).
• Minimize the amount of soil exposed during construction activity by preserving areas of native topsoil on
the site where feasible and sequencing or phasing construction activities to minimize the extent and
duration of exposed soils.
• Maintain natural buffer areas.
• Control stormwater discharges and flow rates, via the following mechanisms:
Diversion of stormwater around the site.
Slow down or containment of stormwater that collects and concentrates at the site.
Avoidance of structural control measure placement in active floodplains, to the degree practicable and
achievable.
Placement of velocity dissipation devices (e.g., check dams, sediment traps, or riprap) along conveyance
channels and where discharges from conveyance channels join water courses.
• Protect steep slopes, via the following mechanisms:
Design and construction of cut-and-fill slopes to minimize erosion.
Diversion of concentrated stormwater flows away from and around the disturbed slopes, using
interceptor dikes, swales, grass-lined channels, pipe slope drains, surface drains, and check dams.
Stabilization of exposed slope areas.
Sediment Control Measures
Sediment control measures (e.g., sediment ponds, traps, filters) should be functional before other land-disturbing
activities take place. These measures may include:
• Storm drain inlet protection measures (e.g., filter berms, perimeter controls, temporary diversion dikes),
that minimize the discharge of sediment prior to entry into the inlet for storm drain inlets located on site
or immediately downstream.
• Water body protection measures (e.g., velocity dissipation devices) that minimize the discharge of
sediment prior to its entry into water bodies located on site or immediately downstream.
• Down-slope sediment controls (e.g., silt fences, temporary diversion dikes) for any portion of the down-
and side-slope perimeters where stormwater would be discharged from disturbed areas of the site.
• Establishment and stabilization of construction vehicle access and exit points, limited to one route if
possible.
• Minimization of dust generation through the application of water or other dust suppression techniques
prior to vehicle exit.
• Stabilization or coverage of soil stockpiles, protection with sediment trapping measures, and, where
possible, location away from storm drain inlets, water bodies, and conveyance channels.
• Design of sediment detention basins to capture runoff or conveyed stormwater and reduce water velocity
to allow sediments to settle out before they can enter streams or other water bodies. Storm flows
eventually pass through an outflow structure leaving the sediment (i.e., solids that can settle) in the
basin. There are important design and management considerations for sediment detention basins for
hard rock mining (USEPA 2003b: Appendix H, Section 6.1.6).
Bristol Bay Assessment 1036 January 2014
-------
Chapter 10 Transportation Corridor
Soil Stabilization
All disturbed areas of the site should be stabilized to minimize on-site erosion and sedimentation and the
resulting discharge of pollutants according to the requirements in ADEC (2011a). Existing vegetation should be
preserved wherever possible.
Many of the BMPs for industrial operations associated with metal miningfocus on sediment and erosion control
and are similar to BMPs used in the construction industry (USEPA 2006). Some of these BMPs pertain specifically
to haul and/or access roads (USEPA 2006).
• Construction of haul roads should be supplemented by BMPs that divert runoff from road surfaces,
minimize erosion, and direct flow to appropriate channels for discharge to treatment areas. Examples of
these BMPs include:
Dikes, curbs, and berms for discharge diversions.
Conveyance systems such as channels, gutters, culverts, rolling dips and road sloping, and/or roadway
water deflectors.
Check dams, rock outlet protection, level spreaders, stream alternation, and drop structures for runoff
dispersion.
Gabions, riprap, native rock retaining walls, straw bale barriers, sediment traps/catch basins, and
vegetated buffer strips for sediment control and collection.
Vegetation to stabilize soils.
• Roads should be placed as far as possible from natural drainage areas, lakes, ponds, wetlands, and
floodplains.
• Width and grade of roads should be as small as possible to meet regulatory requirements and designed
to match the area's natural contours.
All stabilization and structural erosion control measures should be inspected frequently and all necessary
maintenance and repairs should be performed.
10.3.4.2 Exposure-Response
Sediment loading from roads can severely affect streams downstream of the roadbed (Furniss et al.
1991). Salmonids are adapted to episodic exposures to suspended sediment, but survival and growth
can be affected as concentrations or durations of exposure increase (Section 9.4.2.1). Increased
deposition of fine sediment decreases the abundance and production offish and benthic invertebrates
(Section 9.4.2.2). Fine sediments have been linked to decreased fry emergence, decreased juvenile
densities, loss of winter carrying capacity, increased predation on fish, and reduced benthic organism
populations and algal production (Newcombe and MacDonald 1991, Gucinski et al. 2001, Angermeier et
al. 2004, Suttle et al. 2004). In low-velocity stream reaches, an excess of fine sediment can completely
cover suitable spawning gravel and render it useless for spawning, and sediment deposited after
spawning may smother eggs and alevins. Excessive stream sediment loading can also result in channel
braiding, increased width-depth ratios, increased incidence and severity of bank erosion, reduced pool
volume and frequency, and increased subsurface flow. These changes can result in reduced quality and
quantity of available spawning habitat (Furniss et al. 1991).
Increased runoff associated with roads may increase rates and extent of erosion, reduce percolation and
aquifer recharge rates, alter channel morphology, and increase stream discharge rates (Forman and
Alexander 1998). During high-discharge events and in high velocity streams, accumulated sediment
tends to be flushed out and re-deposited in larger water bodies (Forman and Alexander 1998). Because
streams crossed by the transportation corridor connect downstream to Iliamna Lake and ponds,
Bristol Bay Assessment 1037 January 2014
-------
Chapter 10 Transportation Corridor
accelerated sedimentation could have an impact on the concentrated sockeye spawning populations in
these habitats. Accelerated sedimentation could also have a localized impact on the clarity and
chemistry of Iliamna Lake, affecting the photic zone (the depth of light penetration sufficient for
photosynthesis) and thereby primary production and zooplankton abundance, which is critical to
juvenile sockeye salmon.
10.3.4.3 Risk Characterization
Suspended and deposited sediment washed from roads, shoulders, ditches, cuts, and fills would likely
diminish habitat quality in the streams below road crossings. The magnitude of effects cannot be
estimated in this assessment. However, published studies of the influence of silt on salmonid streams
(Section 9.4) indicate that even relatively small amounts of additional sediment could have locally
significant effects on reproductive success of salmonids and production of aquatic invertebrates.
Potential mitigation measures for stormwater runoff, erosion, and sedimentation are discussed in Box
10-3.
10.3.5 Dust
Dust results from traffic operating on unpaved roads in dry weather, grinding and breaking down road
materials into fine particles (Reid and Dunne 1984). These fines are either transported aerially in the
dry season or mobilized by water in the wet season. Dust particles may also include trace contaminants,
including de-icing salts, hydrocarbons, and metals. Following initial suspension by vehicle traffic, aerial
transport by wind spreads dust over long distances, so that it can reach surface waters that are
otherwise buffered from sediment delivery via aqueous overland flow (Appendix G). Dust control agents
such as calcium chloride have been shown to reduce the generation of road dust by 50 to 70% (Bader
1997), but these agents may cause toxic effects when they run off and enter surface waters (Section
10.3.3).
10.3.5.1 Exposure
The amount of dust derived from a road surface is a function of many variables, including composition
and moisture state of the surface, amount and type of vehicle traffic, and speed. An Iowa Highway
Research Board project (Hoover et al. 1973) that quantified dust sources and emissions created by
traffic on unpaved roads found that one vehicle, traveling 1 mile of unpaved road once a day every day
for 1 year, would result in the deposition of 1 ton of dust within a 1,000-foot corridor centered on the
road (i.e., traffic would annually deposit 1 ton of dust per mile per vehicle).
To estimate truck traffic required by the mine scenarios, we extrapolated from vehicle use at a smaller
gold mine (Pogo Mine) based on the rate of ore production at Pogo relative to the mine scenarios.
Estimated production rate at Pogo is 3,000 tons per day (USEPA 2003a), versus 200,000 tons per day in
the mine scenarios (Ghaffari et al. 2011). Overall mine-related vehicle use at Pogo averages between 10
and 20 round trips per day (USEPA 2003a). Approximately 175 truck trips per year (0.5 round trip per
day) are required at Pogo to transport reagents, leaving 19.5 round trips per day for other purposes. The
number of truck trips required for transport of reagents is assumed to be roughly proportional to ore
Bristol Bay Assessment 1038 January 2014
-------
Chapter 10 Transportation Corridor
production, resulting in an estimate of 33 round trips per day to transport reagents in the assessment
mine scenarios. The number of daily round trips for purposes other than reagent transport was
estimated at 19.5 round trips per day, for a total daily traffic estimate of 52.5 round trips in the mine
scenarios. This value is likely an underestimate, as it does not account for potential effects of size
differences between Pogo Mine and the mine scenarios or the number of trips for purposes other than
reagent transport.
The length of the transportation corridor within the Kvichak River watershed would be 113 km. Based
on the estimate from Hoover et al. (1973), the average amount of dust (in tons) generated per mile of
road per year along the transportation corridor within the Kvichak River watershed would be
equivalent to the daily average number of vehicles passing along the corridor (one vehicle making a
round-trip constituting two passages). Using this method, the mine scenarios would generate
approximately 105 tons of dust per mile (59 metric tons per km) annually or approximately
6,700 metric tons annually for the entire length of road within the Kvichak River watershed. This value
may be an underestimate because smaller vehicles typically use rural roads in Iowa, or an overestimate
if roads in Iowa are drier or if dust suppression is effective. Regardless, it indicates that dust production
along the transportation corridor could be substantial.
10.3.5.2 Exposure-Response
Walker and Everett (1987) evaluated the effects of road dust generated by traffic on the Dalton Highway
and Prudhoe Bay Spine Road in northern Alaska. Dust deposition altered the albedo of snow cover,
causing earlier (and presumably more rapid) snowmelt up to 100 m from the road margin and increased
depth of thaw in roadside soils. Dust was also associated with loss of lichens, sphagnum, and other
mosses and reduced plant cover (Walker and Everett 1987). Loss of near-roadway vegetation has
important implications for water quality, as that vegetation helps to filter sediment from road runoff.
Thus, dust deposition can contribute to stored sediment that can mobilize in wet weather, and
deposition can reduce the capacity of roadside landscapes to filter that sediment.
In a study of road effects in Arctic tundra at acidic (soil pH less than 5.0) and less acidic (soil pH at least
5.0) sites, Auerbach et al. (1997) found that vegetation effects were more pronounced at the acidic site.
Permafrost thaw was deeper next to than away from the road at both sites, and could affect road
structure detrimentally. Vegetation biomass of most taxa was reduced near the road at both sites.
Species richness in acidic tundra next to the road was less than half the richness at 100 m away from the
road. Sphagnum mosses, dominant in acidic low arctic tussock tundra, were virtually eliminated near
the road. According to PLP (2011: Chapter 5), approximately 72% of the mine area is composed of well-
drained acidic soils (58% strongly acidic); approximately 34% of the transportation corridor is
composed of well-drained acidic soils (3.5% strongly acidic).
10.3.5.3 Risk Characterization
The main impact of dust from the transportation corridor on salmonids likely would be reduced habitat
quality due to a reduction in riparian vegetation and subsequent increase in suspended sediment and
Bristol Bay Assessment 1039 January 2014
-------
Chapter 10 Transportation Corridor
fine bed sediment, especially during road construction. Potential effects of increased sediment loading
are discussed in Section 10.3.4. Loss of riparian vegetation would also occur at the mine site, but there
the main impact of dust would be a direct increase in fine bed sediment due to mine construction and
operation.
10.3.6 Invasive Species
10.3.6.1 Exposure
Construction and operation of the transportation corridor would increase the probability that new
terrestrial and aquatic species would be transported to and could establish themselves in the Bristol Bay
region. Roads can facilitate introductions via contaminated soil or gravel used in road construction and
maintenance, or via contaminated vehicles, equipment, cargo, and people that travel those roads. For
example, road fill appears to be the mode of introduction and spread for invasive sweetclover (Melilotus
alba) in central and southeast Alaska (Wurtz et al. 2010). Elsewhere, road maintenance further spreads
invasive plants along suitable roadside habitat (Christen and Matlack 2009). Vehicles can carry
contaminated equipment and cargo. Over the 2-year construction of a research station in Antarctica, an
estimated 5,000 seeds from 14 different plant families were introduced on almost 15,000 m3 of cargo
(Lee and Chown 2009). Once docked, seeds on cargo could disperse at almost any location along the
transportation corridor. Finally, people unintentionally introduce and spread invasive species in Alaska
and other Arctic environments on their shoes (Bella 2011, Ware et al. 2012).
Once established along or near the transportation corridor, terrestrial species that thrive in riparian and
floodplain areas could spread to salmon-bearing habitat at any of the points where the road crosses a
river, stream, or other aquatic habitat. In a survey of 2,865 km (1,780 miles) of major highways in
interior and south-central Alaska, 64 of 192 sampled bridge crossings (over 30%) were found to have
sweetclover adjacent to them, and sweetclover had spread to downstream floodplains at 17 of these
bridge crossings (Wurtz et al. 2010). This survey likely underestimates the number of floodplain
invasions, because it did not sample numerous stream crossings serviced by culverts or other locations
along streams where fill had been placed.
Aquatic invasive species, including macrophytes, shellfish, and salmonid pathogens and parasites can
also be introduced along the transportation corridor on equipment that has come into contact with
contaminated waters. Most literature emphasizes recreation equipment (Johnson et al. 2001, Arsan and
Bartholomew 2008); little or no information exists about the incidence of aquatic or riparian species
introductions specifically on construction or mining equipment. Transported equipment contaminated
with aquatic invaders could spread those species to salmon-bearing habitat via direct contact with
anadromous waters during stream crossing construction or during mining activity. Aquatic invaders
could also be carried by water to other salmon-bearing habitats downstream of the initial introduction
locations, including into Iliamna Lake and other parts of the Kvichak River watershed.
The likelihood that an aquatic invasive species will establish and spread successfully can depend heavily
on environmental requirements. For instance, Myxobolus cerebralis, a cnidarian parasite that causes
Bristol Bay Assessment 1040 January 2014
-------
Chapter 10 Transportation Corridor
whirling disease, has already been detected in an Anchorage, Alaska, trout hatchery. This parasite has
very specific abiotic and biotic conditions under which it infects salmonids. If the pathogen is introduced
to a new area, susceptible genetic variants of the secondary host (the oligochaete worm Tubifex tubifex)
must be present, seasonal water temperatures must exceed 10°C with approximately 1,500 degree-days,
and susceptible salmonid species and life-stages must co-occur with the secondary host (Arsan and
Bartholomew 2008). In addition to the hatchery location where whirling disease has already been
found, favorable conditions exist for parasite establishment in two tributaries of Cook Inlet near
Anchorage (Arsan and Bartholomew 2008). However, conditions for whirling disease establishment are
not known for the Bristol Bay region.
10.3.6.2 Exposure-Response
Invasive species can drastically alter the composition of riparian and floodplain vegetation adjacent to
salmon habitats. Invasive sweetclover, purple loosestrife (Lythrum salicaria), and giant knotweed
(Polygonum sachalinense)—all current invaders in Alaska—can replace native riparian species (Blossey
et al. 2001, Urgenson et al. 2009, Spellman and Wurtz 2011). In general, it has been difficult to show a
direct effect of riparian vegetation alteration on fish diversity, abundance, or biomass (Smokorowski
and Pratt 2007), but indirect effects on salmon via aquatic foodwebs have been documented (Wipfli and
Baxter 2010). Giant knotweed was shown to release nitrogen-poor litter into a tributary of the salmon-
bearing Skagit River in Washington, which can have cascading, negative effects on fish by altering their
invertebrate food sources (Urgenson et al. 2009). Purple loosestrife was found to decompose four times
faster than native sedge in the Fraser River, making detritus available in fall rather than winter and
spring, when it was usually used by invertebrates that support salmon production (Grout et al. 1997).
Links between aquatic invaders, particularly macrophytes, and fish performance have been made in
lentic, but rarely in lotic, habitats (Smokorowski and Pratt 2007). Effects of invasive macrophytes range
from increased native fish abundance, to no effect, to detrimental effects on fish and their food sources
via exuded toxic compounds, depending on the invasive species and fish species of interest (Schultz and
Dibble 2012). Streambed coverage of several aquatic macrophyte species, both native and introduced
(including the recent Alaska invader Elodea canadensis), reduced the number of Chinook salmon redds
and the percentage of available spawners observed using infested habitat in northern California (Merz
et al. 2008). This is significant in the regulated, low-flow Mokelumne River in California, where
spawning habitat is considered a limiting resource.
Evidence of the effects of other aquatic invaders on salmonids also exists. Didymo (Didymosphenia
germinata) is a colonial diatom capable of covering stream substrates with thick, slippery mats.
Documented effects of didymo on salmonids vary with location and fish species. Effects of didymo on the
invertebrate communities that serve as fish food sources could ultimately affect salmonid growth and
abundance (Whitton et al. 2009). The aquatic invader that causes whirling disease (M. cerebralis) has
had devastating effects on several wild fisheries in the United States intermountain west (Nehring and
Walker 1996). The disease can cause lesions, neurological defects, skeletal deformities, and death. Both
Bristol Bay Assessment 1041 January 2014
-------
Chapter 10 Transportation Corridor
sockeye salmon and rainbow trout fry are highly susceptible to whirling disease, should conditions be
right for infection.
10.3.6.3 Risk Characterization
The spread of aquatic, riparian, and floodplain invasive species along roads and into salmon-bearing
habitats could occur during construction and operation of the proposed transportation corridor,
although mitigation measures can lower the likelihood of invasion (Box 10-4). Invasion of riparian and
floodplain species is occurring in Alaska via the use of contaminated gravel road fill. In the case of
invasive sweetclover, subsequent dispersal to almost 9% of floodplains downstream of bridges along
one major highway was observed (Wurtz et al. 2010). Assuming similar rates of invasion along both the
transportation corridor and bridges and culverts, 9% of the 64 streams and rivers—5 to 6 streams—
crossed by the corridor in the Kvichak River watershed would experience invasion. Given that 55 of the
64 streams crossed by the transportation corridor are known or likely to support salmonids, alteration
of salmon habitats would be expected in approximately 5 streams. However, this is almost assuredly an
underestimate because it is based on rate of invasion for only one species and assumes that the spread
of that species has reached equilibrium.
BOX 10-4. MITIGATION FOR INVASIVE SPECIES
The use of contaminated gravel road fill in Alaska has fostered the invasion of nonnative riparian and
floodplain plant species. In some cases, the species are subsequently dispersed to floodplains downstream
of road-stream crossings. Introduction and invasion of nonnative riparian and floodplain species may also
occur via contaminated cargo, equipment, and boots. The following steps can help to mitigate the
introduction and spread of invasive species.
• Purchase of fill from existing or new gravel pits certified by the Alaska Department of Natural
Resources Division of Agriculture as weed-free (ADNR 2013).
• Proper and thorough inspection and de-contamination of cargo, equipment, and boots, at the port
and at the mine site.
• Use of new equipment, where possible.
• Use of a process for cleaning, draining, and drying equipment previously used at another site
(including personal gear worn by workers) that is advocated by the Alaska Department of Fish and
Game for recreational equipment.
Should sweetclover, purple loosestrife, giant knotweed or other species invade riparian areas and
floodplains adjacent to salmon-bearing streams and wetlands in the Bristol Bay region, they could
change organic matter inputs into those streams and affect salmon food sources (Blossey et al. 2001,
Urgenson et al. 2009, Spellman and Wurtz 2011). The extent to which salmon growth, diversity, or
abundance would be altered would depend on the extent and intensity of infestation. Once initiated,
these invasions would be difficult to reverse.
Invasions by aquatic species seem less likely but cannot be quantified. The most likely vector is believed
to be construction equipment that has been used at stream crossings in a prior project. Such equipment
could carry microbes or propagules in mud that could be transferred when constructing road and
pipeline crossings in the Bristol Bay watershed.
Bristol Bay Assessment 1042 January 2014
-------
Chapter 10 Transportation Corridor
The spread of invasive species is highly stochastic and there are no good, relevant models for risk
estimation. Therefore, it is not as clear a threat as other issues considered in this assessment. However,
the introduction and spread of invasive species has been a major cause of environmental degradation in
the United States, and mitigation measures could reduce the risks (Box 10-4).
10.4 Overall Risk Characterization for the
Transportation Corridor
Risks to salmonids from filling of wetlands, hydrologic modifications, spillage or runoff of contaminants
and fine sediment, dust deposition, and introduction of invasive species are likely to diminish the
production of anadromous and resident salmonids in many of the 55 streams known or likely to support
salmonids that would be crossed by the transportation corridor. Salmonid spawning migrations and
other movements maybe impeded by culverts in 36 streams, 32 of which contain restricted (less than
5.5 km) upstream habitat. Assuming typical maintenance practices after mine operations, approximately
15 of these 32 streams would be entirely or partly blocked at any time. As a result, salmonid passage—
and ultimately production—would be reduced in these streams, and they would likely not be able to
support long-term populations of resident species such as rainbow trout or Dolly Varden.
Approximately 272 km of streams downstream of road crossings also could be affected.
The migratory barriers and degradation of stream habitat discussed herein could also reduce the high
genetic diversity among sockeye populations reported by Gomez-Uchida et al. (2011) and Quinn et al.
(2012). This loss in diversity may decrease the long-term viability of sockeye salmon and would
negatively affect localized watershed food webs.
Truck accidents may spill xanthates, cyanide, or molybdenum concentrate into streams crossed by the
road. Xanthate and cyanide are highly toxic and could kill fish and invertebrates in the receiving streams
and, depending on the size of the spill, portions of Iliamna Lake. Molybdenum concentrate is much less
toxic and unlikely to cause severe effects.
The exact magnitudes of changes in fish productivity, abundance, and diversity cannot be estimated at
this time, but the species, abundances, and distributions that could be affected are summarized below.
• Sockeye salmon spawning has been observed at 30 locations along the transportation corridor.
Highest average abundances are in the Iliamna River (100,000 spawners), the Newhalen River
(80,000 spawners), and Knutson Bay (70,000 spawners), although abundances can be much higher
(e.g., 1 million adults were reported in 1960 survey of Knutson Bay).
• Chinook, coho, pink and chum salmon have been reported at isolated points in the Kvichak River
watershed, and all four species have been observed in Upper Talarik Creek and the Iliamna River.
• Dolly Varden have been reported in nearly every sockeye salmon-bearing stream that would be
crossed by or adjacent to the corridor, as well as in locations upstream of sites with reported
anadromous salmon use.
Bristol Bay Assessment 1043 January 2014
-------
Chapter 10 Transportation Corridor
• Rainbow trout have been reported in Upper Talarik Creek, the Newhalen River, an unnamed
tributary to Eagle Bay, Youngs, Tomkok, and Swamp Creeks, the Iliamna River, and Chinkelyes
Creek.
10.5 Uncertainties
In this chapter we evaluated the risks to salmonid habitats and populations associated with the
transportation corridor (Figure 10-3). A number of uncertainties are inherent in assessing these risks,
which are summarized below (uncertainties related to the effectiveness of mitigation measures are
discussed in Box 10-5).
• Characterization of streams and wetlands affected by the transportation corridor. The NWI, NHD,
AWC, and AFFI were used to evaluate the effects of the transportation corridor on hydrological
features and fish populations (Box 10-1). These datasets include the following limitations.
o Underestimation of the number of stream crossings and degree of channel sinuosity, resulting in
underestimates of affected stream lengths.
o Underestimation of fish-bearing streams due to limited sampling.
o Potential undercharacterization of wetland area due to limited resolution of available NWI data.
o Underestimation of potential impacts on wetlands bisected by the transportation corridor,
because wetland area outside the 200-m boundary was assumed to maintain functionality.
Overall, these uncertainties likely result in a moderate underestimation of risks to fish.
• Estimation of dust production from the transportation corridor. Our dust production estimate is
based on a study that quantified dust sources and emissions created by traffic on unpaved roads.
Extrapolating that study to the transportation corridor does not take into account variables such as
composition and moisture of the road surface, number and width of tires, and speed. In addition,
road dust generation may be reduced by 50 to 70% by the application of dust control agents such as
calcium chloride. Overall, these uncertainties likely have a negligible effect on risks to fish, but a
moderate effect on our dust production calculations.
• Estimation of chemical spill frequency due to truck accidents. Extrapolation of truck accident
probability from a study of rural two-lane roads does not take into account specific, generally more
difficult road and weather conditions prevalent in the area of the Pebble deposit. However, the risk
of spills could be at least partially mitigated by using spill-resistant containers. Overall, these
uncertainties likely result in a moderate underestimation of risk to fish because of effects on spill
frequency calculations. Frequencies of cyanide and molybdenum concentrate spills were not
estimated due to uncertainties in the mining scenarios.
• Estimation of risks to salmonids from spills. A spill of cyanide, xanthate, or molybdenum
concentrate could occur in various ways and at various locations. The sparse literature on the
aquatic chemistry and toxicology of xanthates and molybdenum makes the consequences of these
Bristol Bay Assessment 1044 January 2014
-------
Chapter 10 Transportation Corridor
events particularly uncertain. Given its high toxicity, we are confident that toxic effects would occur
following a xanthate spill into a stream; we are simply uncertain of the magnitude and extent of
effects.
• Estimation of culvert failure frequencies. These frequencies, derived from the literature, assume
that culverts are designed to specifications but are not always installed correctly and/or do not
stand up to the rigors of a harsh environment. This uncertainty likely has a moderate effect on risks
to fish, with unclear direction. Nonetheless, this does not change overall conclusions reached with
respect to reduction of passage and ultimately production of salmonids or the viability of long-term
populations of resident species.
• Risks from invasive species. Roads serve as corridors for the spread of weeds, pathogens, and other
invasive species. However, the list of potential invaders is ill-defined and the rate of their spread
along an industrial road is unknown.
• Climate change effects. The potential impacts of road construction and operation discussed in this
chapter do not take into account potential effects of climate change. Over the timeframe considered
in this assessment (approximately 80 years), the physical environment of the Bristol Bay watershed
is likely to change substantially as a result of increases in temperature and precipitation (Section
3.8). Increases in rain-on-snow events are likely to increase flood frequency. Such changes could
undermine the structure of the transportation corridor and its stream crossings. The variability and
magnitude of streamflows could also enhance other impacts described in this chapter, including
channel entrenchment and the loss of water-body connectivity. Collectively, these impacts would
likely further reduce the diversity offish habitat, causing a loss of population genetic diversity over
time that would reduce the resiliency of salmon stocks to environmental fluctuations related to
climate change. Overall, these climate-related uncertainties result in a moderate underestimation of
risk to fish.
Bristol Bay Assessment 1045 January 2014
-------
Chapter 10
Transportation Corridor
BOX 10-5. LIKELY EFFECTIVENESS OF MITIGATION MEASURES
Environmental characteristics along the transportation corridor would likely render the effectiveness of
standard or even state-of-the-art mitigation measures highly uncertain.
• Subarctic extreme temperatures and frozen soil conditions could complicate planning for
remediation, with uncertain outcomes due to variable conditions and spill material characteristics.
• Subarctic climatic conditions could limit the lushness and rapidity of vegetation growth or re-growth
following ground disturbance, reducing the effectiveness of vegetated areas as sediment and
nutrient filtration buffers.
• Widespread and extensive areas of near-surface groundwater and seasonally or permanently
saturated soils could limit the potential for absorption or trapping of road runoff, and increase
likelihood of its delivery to surface waters.
• The likelihood of ice flows and drives during thaws could make water crossing structures
problematic locations for jams and plugging.
• The region is seismically active (Section 3.6), and even a small increment of ground deformation
could easily disturb engineered structures and alter patterns of surface and subsurface drainage in
ways that render engineered mitigations inoperative or harmful.
• Remote locations that are not frequented by humans mean that mitigation failures and accidents
could go undetected until substantial harm to waters has occurred unless frequent inspections are
conducted.
Although many possible mitigation measures can be identified and listed in a mitigation plan, they cannot all
be ideally applied in every instance. Mitigation measures are often mutually limiting or offsetting when
applied in the field. As a salient example for the transportation corridor, choosing a road location that
minimizes crossings of streams, wetlands, and areas of shallow groundwater in a landscape that is rich in
those hydrologic features could result in a tortuous alignment that is excessively long and curved to
accommodate the upland terrain. This alignment would greatly increase the total ground area disturbed, and
increased road curvature in either horizontal and vertical dimensions may increase risk of traffic accidents
and consequent spills. It would also increase the length and structural complexity of the road-parallel
pipelines (Chapter 11). Thus, avoidance of sensitive habitat features could elevate other environmental
risks.
Bristol Bay Assessment
10-46
January 2014
-------
As described in Section 6.1.3, the mine scenarios include four pipelines along the transportation
corridor—one each for natural gas, diesel, product concentrate, and return water—and various
pipelines on the mine site. Any of these pipelines could fail and release their contents to the
environment. The risks from failure of product concentrate (Figure 11-1), return water (Figure 11-2),
and diesel (Figure 11-3) pipelines are considered particularly high. These failure scenarios are
evaluated in the following sections. Other pipelines are discussed briefly below.
On the mine site, the largest pipelines would carry tailings slurry from the mill to the tailings storage
facilities (TSFs) and reclaimed water from the TSFs to the mill (Table 6-4). Smaller pipelines would
convey water for processing and other uses and wastewater for treatment or storage. Other pipelines
would carry diesel and natural gas from storage tanks to points of use. On-site pipeline spills have
occurred at porphyry copper mines in the United States and some have resulted in significant aquatic
exposures (Earthworks 2012). Such spills are possible at a future mine and could result in uncontrolled
releases within the mine site; however, these spills are more likely to be contained or controlled without
significant environmental effects than pipeline spills along the transportation corridor. In this
assessment, we decided that leakage from on-site pipelines would be captured and controlled by the
mine's drainage system and either treated prior to discharge or pumped to the process water pond or
TSF.
Natural gas is lighter than air, so any release due to a natural gas pipeline failure would rise and
dissipate. If the gas cloud ignited most of the heat would travel upward, but the initial blast and
subsequent radiation heating could affect the road and nearby environment. During dry periods, a
wildfire could result. Such failures were considered to pose relatively low risks to the assessment
endpoints and are not evaluated further in this assessment.
Bristol Bay Assessment
11-1
January 2014
-------
Chapter 11
Pipeline Failures
Figure 11-1. Conceptual model illustrating potential stressors and effects resulting from a
concentrate pipeline failure.
mitigation
(e.g., product recovery)
LEGEND
| concentrate |
I pipeline J
V
concentrate pipeline
breaker leak
V
1^ product
in aquatic habitats
A
product transport
to aquatic habitats
V
concentrate water
in aquatic habitats
A
1 product resuspension
& transport
additional step in
can sal pathway
Within a shape, 'h indicates an increase in the parameter,
-I indicates a decrease in a parameter, and A indicates a
change in the parameter.
Arrows leading from one shape to another indicate a
hypothesized cause-effect relationship.
Shapes bracketed under another shape are specific
components of the more general shape under which they
appear.
salmonid fishes
(abundance, productivity or diversity)
Bristol Bay Assessment
11-2
January 2014
-------
Chapter 11
Pipeline Failures
Figure 11-2. Conceptual model illustrating potential stressors and effects resulting from a return
water pipeline failure.
mitigation
additional step in
can sal path way
proximate
•»•• — "
modifying
factor
Within a shape, 'I indicates an
increase in the parameter, 4- indicates
a decrease in a parameter, and £
indicates a change in the parameter.
Arrows leading from one shape to
another indicate a hypothesized cause-
effect relationship.
Shapes bracketed under another shape
are specific components of the more
general shape under which they
appeal.
[return waten
I pipeline J
V
return water pipeline
break or leak
V
^ return water
in aquatic habitats
V
V
metals
' xanthates
V
V
1s acute toxicity
V
L- invertebrate
abundance
V
•sb salmonid fishes
(abundance, productivity or diversity)
Bristol Bay Assessment
11-3
January 2014
-------
Chapter 11
Pipeline Failures
Figure 11-3. Conceptual model illustrating potential stressors and effects resulting from a diesel
pipeline failure.
diesel
pipeline
V
diesel pipeline
break or leak
V
V
^diesel
in aquatic habitats
mitigation
(e.g., oil recovery)
1s diesel on land
\
/
^ subsurface transport
of oil
T suspended
oil droplets
additional step in
can sal path way
Within a shape, '}• indicates an increase in the parameter, I-
indicates a decrease in a parameter, and A indicates a change in
the parameter.
Arrows leading from one shape to another indicate a hypothesized
cause-effect relationship.
Shapes bracketed under another shape are specific components of
the more general shape under which they appear.
V
•^ salmonid fishes
(abundance, productivity or diversity)
Bristol Bay Assessment
11-4
January 2014
-------
Chapter 11 Pipeline Failures
11.1 Causes and Probabilities of Pipeline Failures
The U.S. transportation system includes more than 4 million km of pipeline, of which more than
3.8 million km are gas transmission or natural gas distribution mains and more than 280,000 km carry
hazardous liquids, primarily petroleum products (PHMSA 2012). The principal causes of failures along
these pipelines are external corrosion and mechanical damage such as impacts by excavating
equipment. Internal corrosion and material breakdown also may cause pipeline failures, but are less
common. The failure rate from impacts, such as can occur during road, pipeline, or bridge maintenance,
tends to be steady over the lifetime of a pipeline, whereas corrosion failures tend to increase with age of
the pipe.
Pipeline failures include both leaks and ruptures. Leaks are small holes and cracks that result in product
loss but do not immediately prevent the functioning of the pipeline. Ruptures are larger holes or breaks
that render the pipeline inoperable. A study of over 2 million km-yr of pipelines in Canada indicated that
leaks account for 87% of failures and ruptures account for 13% (EUB 1998). A rupture could result in
the immediate release of a significant amount of pipeline material. A leak would allow pipeline material
to escape more slowly than a rupture, but a leak could remain undetected for a much longer time,
ultimately releasing quantities comparable to or exceeding a rupture.
The most extensive pipeline failure statistics are derived from oil and gas industry data (Table 11-1).
The industry's record of pipeline failures is directly relevant to the oil and gas pipelines considered in
the pipeline failure scenarios. The failure rate of metal concentrate slurry pipelines is unknown, because
few such pipelines are in operation and no published failure rates are available for those that are in
operation. Oil pipeline failure rates are used as the best available estimate, although it is possible that
the erosive or corrosive nature of the product concentrate slurry would increase pipeline failure rates.
Although the range of published annual failure rates for U.S. oil and gas pipelines spans more than one
order of magnitude (0.000046 to 0.0011 per km-yr) (URS 2000), the range for pipelines most similar to
the assessment pipelines along the transportation corridor is much narrower. For example, the failure
rate is 0.0010 failure/km-yr for pipelines less than 20 cm in diameter (OGP 2010), 0.0015 failure/km-yr
for pipelines in a climate similar to Alaska (Alberta, Canada) (ERCB 2013), and 0.00062 failure/km-yr
for pipelines run by small operators (those operating total pipeline lengths less than 670 km) (URS
2000). The geometric mean of these three values yields a failure probability of 0.0010 failure/km-yr.
This overall estimate of annual failure probability, coupled with the 113-km length of each pipeline as it
runs along the transportation corridor within the Kvichak River watershed, results in an 11%
probability of a failure in each of the four pipelines each year. Thus, the probability of a pipeline failure
occurring over the duration of the Pebble 2.0 scenario (i.e., approximately 25 years) would be 95% for
each pipeline. The expected number of failures in each pipeline would be about 2.2, 2.8, and 8.6 over the
life of the mine in the Pebble 0.25, 2.0, and 6.5 scenarios, respectively. The chance of a large rupture in
each of the three pipelines over the life of the mine would exceed 25%, 30%, and 67% in the Pebble
0.25, 2.0, and 6.5 scenarios, respectively. In each of the three scenarios, there would be a greater than
Bristol Bay Assessment 115 January 2014
-------
Chapter 11
Pipeline Failures
99.9% chance that at least one of the three pipelines carrying liquid would fail during the project
lifetime.
Table 11-1. Studies that examined pipeline failure rates.
Study
OGP 2010
(oil pipelines)
OGP 2010
(gas pipelines)
Caleyo 2007
URS 2000
(56 U.S. oil pipeline
operators)
ERCB 2013
Km-Years
Analyzed
667,000
2,770,000
34,595
28,270
1,268,370
285,000
380,331
395,479
386,930
398,253
406,974
Pipeline or Failure Parameter Assessed
Diameter <20 cm
Diameter 20-36 cm
Wall thickness <5 mm
Wall thickness 5-10 mm
1970-2004
2000-2004
Mexican gas pipelines
Mexican oil pipelines
Highest failure rate
Average failure rate
Minimum failure rate
10 smallest operators (<670 km)
10 largest operators (>6,900 km)
2000, Alberta, Canada
2007, Alberta, Canada
2008, Alberta, Canada
2009, Alberta, Canada
2010, Alberta, Canada
2011, Alberta, Canada
Annual Failure Rate
(per km-year)
0.0010
0.00080
0.00040
0.00017
0.00041
0.00017
0.0030
0.0052
0.0011
0.00028
0.000046
0.00062
0.00020
0.0033
0.0022
0.0021
0.0016
0.0015
0.0015
Although data are insufficient to determine failure probabilities specific to the metal mining industry,
the record suggests that pipeline failures at mines are not uncommon. A review of 14 operating
porphyry copper mines in the United States (including all operating U.S. porphyry copper mines except
two that have been operating for less than 5 years) found that all had experienced pipeline spills or
accidental releases and that pipeline failures have continued into 2012 (Earthworks 2012).
It may be argued that engineering can reduce pipeline failures rates below historical levels, but
improved engineering has little effect on the rate of human errors. Many pipeline failures, such as the
cyanide water spill at the Fort Knox mine (Fairbanks, Alaska) that resulted from a bulldozer ripper blade
hitting the pipeline (ADEC 2012), are due to human errors. Perhaps more important, human error can
negate safety systems. For example, on July 25 and 26, 2010, crude oil spilled into the Kalamazoo River,
Michigan, from a pipeline operated by Enbridge Energy. A series of in-line inspections had showed
multiple corrosion and crack-like anomalies at the river crossing, but no field inspection was performed
(Barrett 2012). When the pipeline failed, more than 3 million L (20,000 barrels) of oil spilled over 2 days
as operators repeatedly overrode the shut-down system and restarted the line (Barrett 2012). The spill
was finally reported by a local gas company employee who happened to witness the leak. The spill may
have been prevented if repairs had been made when defects were detected, and the release could have
been minimized if operators had promptly shut down the line.
Bristol Bay Assessment
11-6
January 2014
-------
Chapter 11 Pipeline Failures
11.2 Potential Receiving Waters
The transportation corridor pipelines evaluated in the assessment would cross approximately
64 streams and rivers in the Kvichak River watershed, 55 of which are believed to support salmonids
and all of which could convey contaminants to Iliamna Lake. This number of crossings is much larger
than the number of hydrologic units presented in Tables 10-3 through 10-5, because hydrologic units
may contain multiple watersheds and each watershed may include crossings of multiple tributaries.
For approximately 14% of their length (15 km), these pipelines would be within 100 m of a stream or
river (Table 10-3), and for 24% of their length (27 km) they would be within 100 m of a mapped
wetland, pond, or small lake (Table 10-4). This proximity would create the potential for spilled slurry to
flow into surface waters either directly or via overland flow. Some of the affected ponds support
salmonids, but the number and distribution of salmonids in the area's wetlands, ponds, and small lakes
are unknown. Approximately 272 km of streams, as well as Iliamna Lake, are downstream of these
pipeline crossings (Table 10-7).
Although exposure pathways for all failure locations are considered, the quantitative analysis addressed
two stream crossings along the assessment's transportation corridor: Chinkelyes Creek and Knutson
Creek. Channel velocities for these creeks were calculated to estimate the time it would take for a spill to
reach Iliamna Lake. Information from the Pebble Limited Partnership (PLP) (2011: Chapter 15.3) was
used to develop channel width and depths. Streamflows were calculated from precipitation models used
to determine mean annual runoff for the assessment's stream culvert analysis (Section 10.3.2). These
mean annual Streamflows applied to the basic channel geometry yielded channel velocities and thus
travel times from each crossing to Iliamna Lake.
From the Chinkelyes Creek crossing, the creek flows 14 km to a confluence with the Iliamna River that
continues for 7.6 km to Iliamna Lake. Lake levels can be seasonally high and create a backwater effect in
the lower 3.5 km of the Iliamna River; however, most of the year the river flows freely for the entire
distance to the lake shore. From the Knutson Creek crossing, the creek flows 2.6 km to Iliamna Lake. As
Knutson Creek approaches the lake, the creek is steeper than the Iliamna River and it flows freely into
the lake year-round. Total travel times to Iliamna Lake are estimated to be 170 minutes and 19 minutes
for a Chinkelyes Creek and a Knutson Creek spill, respectively (Table 11-2). More details concerning
these and other stream crossings are presented in Section 10.3.2.
Bristol Bay Assessment 117 January 2014
-------
Chapter 11
Pipeline Failures
Table 11-2. Parameters for concentrate pipeline spills to Chinkelyes Creek and Knutson Creek.
Parameter
Spill into Chinkelyes Creek
Chinkelyes Creek
Iliamna River
Spill into Knutson
Creek
Knutson Creek
Water Flow
Discharge (m3/s)
Velocity (m/s)
Channel Length (km)
1.8
2.2
14
22
2.0
7.6
3.4
2.2
2.6
Pipeline Drainage and Dilution
Flow rate while draining (m3/s)
Flow rate while pumping (m3/s)
Release time— draining (minutes)
Release time— pumping (minutes)
Volume of slurry spilled (L)
Mass of concentrate solids spilled (metric tons)
Volume of aqueous phase spilled (L)
Maximum fully mixed dissolved copper
concentration (Mg/L)
Quotient3, acute copper criterion
Quotient3, chronic copper criterion
Travel time to confluence (minutes)b
0.11
0.04
9.3
5.0
75,000
66
58,000
37
13
22
110
-
-
-
-
3.3
1.3
2.1
64
0.07
0.04
5.6
5.0
37,000
32
28,000
16
5.9
9.6
19
Pipeline and Slurry Specifications
Length from top of nearest hill to valve (m)
Elevation drop (m)
Viscosity of slurry (cP)
Density of slurry (metric tons/m3)
2100
150
-
810
25
9.5
1.7
Notes:
Dashes (-) indicate that spill is not directly into Iliamna River, which receives flow from Chinkelyes Creek.
a See Box 8-3 for a description of how risk quotients were calculated.
b Confluence with Iliamna River for Chinkelyes Creek; confluence with Iliamna Lake for the Iliamna River and Knutson Creek.
11.3 Concentrate Pipeline Failure Scenarios
11.3.1 Sources
A full pipeline break or a defect of equivalent size in the copper (+gold) concentrate pipeline (Table 6-4)
at the Chinkelyes Creek or Knutson Creek crossing would release slurry into these water bodies. This
kind of failure could result from mechanical failure of the pipe due to ground movement, vehicle impact,
maintenance error, or material failure. Parameters for the concentrate pipeline failure scenarios are
summarized in Table 11-2.
In the concentrate pipeline failure scenarios, a single complete break of the pipeline would occur at the
edge of the stream, just upstream of an isolation valve. These valves would be placed on either side of
major crossings (Ghaffari et al. 2011) and could be remotely activated. Pumping would continue for
5 minutes until the alarm condition was assessed and an operator shut down the pumps. The estimated
total slurry volume draining to the stream would equal the pumped flow rate times 5 minutes, plus the
volume between the break and local high point in the pipeline (i.e., the nearest watershed boundary)
Bristol Bay Assessment
11-8
January 2014
-------
Chapter 11 Pipeline Failures
(Table 11-2). During the entire spill, gravity drainage would govern the flow rate based on calculations
for free-flowing pipes.
The product concentrate would have a density of 3.8 metric tons/m3 and would sink rapidly if released
into a water body at low flows. The slurry water would have a density near 1.0 metric ton/m3 and would
mix readily with surface waters. No analyses of product concentrate or concentrate transport water are
available for the Pebble deposit or any other ore body in the region. To estimate the concentration of
metals and other constituents in the concentrate, we used analyses from the Aitik (Sweden) porphyry
copper mine as described in Appendix H.
The fine particles of product concentrate would, like spilled tailings (Section 9.3), degrade habitat
quality for fish and benthic invertebrates. However, these potential physical effects would be much
lower in magnitude than for a tailings dam failure because of the much lower volume of material, and
would be less important than potential toxic effects. Thus, we focus on toxic effects rather than effects of
sediment deposition on habitat.
11.3.2 Exposure
In these concentrate pipeline failure scenarios, 66 metric tons of product concentrate would be released
into Chinkelyes Creek or 32 metric tons into Knutson Creek. Based on its size and the well-established
relationship between particle size and particle mobilization and transport (commonly represented by
the Hjulstrom diagram), the concentrate would be transported in suspension by streamflows greater
than approximately 20 cm/s and would be transported as bedload between approximately 1 and
20 cm/s. Estimated mean velocities of the streams (2.2 m/s for Chinkelyes Creek and Knutson Creek and
2.0 m/s for the Iliamna River) are consistent with those described for these streams (PLP 2011) and are
well above the transport velocities. Therefore, the fine sand-sized concentrate would be carried
downstream during typical or high flows, even given that the concentrate is denser (3.8 metric tons/m3)
than typical rock (2.8 metric tons/m3 for granite) and would move less readily. Concentrate would be
deposited in any backwaters, pools, or other low-flow locations. If the spill occurred during a period of
high flow, it would be carried downstream immediately, reaching Iliamna Lake within 3 hours (via
Chinkelyes Creek and Iliamna River) or 0.5 hour (via Knutson Creek). Because flood flows are a potential
cause of pipeline failure at stream crossings, this is a reasonable possibility. If the spill occurred during
low flows, concentrate that is not collected would be spread downstream by erosion during subsequent
typical or high-flow periods, eventually entering Iliamna Lake. Concentrate that entered the lake could
mix into sand and gravel beaches used by spawning sockeye salmon. These transport and deposition
processes cannot be quantified with existing data and modeling resources.
The estimated annual failure rate of one per 1,000 km per year (Section 11.1) results in an estimated
failure rate of 0.11 per year for the 113 km of concentrate pipeline within the Kvichak River watershed.
If the probability of a pipeline failure is independent of location, and if it is assumed that spills within
100 m of a stream could flow to that stream, a spill would have a 14% probability of entering a stream
within the Kvichak River watershed. This would result in an estimate of 0.015 stream-contaminating
concentrate spills per year, or 1.2 stream-contaminating concentrate spills over the duration of the
Bristol Bay Assessment 119 January 2014
-------
Chapter 11 Pipeline Failures
Pebble 6.5 scenario (approximately 78 years). In other words, we expect roughly 1 such spill in the
Pebble 6.5 scenario. Similarly, a spill would have a 24% probability of entering a wetland, resulting in an
estimate of 0.026 wetland-contaminating spills per year or 2 wetland-contaminating spills in the
Pebble 6.5 scenario. A portion of those wetlands would be ponds or backwaters that support fish.
Spills from the pipeline failure would contaminate 2.6 km of Knutson Creek or 14 km of Chinkelyes
Creek and 7.6 km of the Iliamna River with product concentrate and leachate (the slurry water that has
leached ions from the product concentrate) before entering Iliamna Lake. The potential extent of
wetland, pond, and small lake contamination cannot be readily estimated.
As with a tailings spill (Chapter 9), lexicologically relevant exposures could occur via multiple routes
following a concentrate pipeline spill. During and immediately following a spill, organisms would be
acutely exposed to leachate and suspended particles. After a spill, product concentrate deposited on a
stream or lake bed would result in chronic aqueous exposures to pore water and acute aqueous
exposures during resuspension events. Unlike the tailings spill, which would inevitably enter a stream
and its floodplain, a slurry spill might directly enter a stream, pond, or wetland; it might flow over land
to a nearby water body; or it might flow across the landscape without reaching water. Terrestrial slurry
deposits are likely to be collected by the operator, so rain and snowmelt are unlikely to leach those
concentrate deposits and significantly contaminate streams. However, spilled leachate from the pipeline
slurry could enter a stream, wetland, pond, or lake by overland or groundwater flow. Contaminated
groundwater could upwell through the gravels and cobbles of streams or deltaic gravels and sands in
Iliamna Lake, and benthic invertebrates and fish eggs and larvae could be exposed to toxic
concentrations if sufficient dilution did not occur.
11.3.2.1 Aqueous Phase Chemical Constituents
The concentrate slurry is estimated to contain 77% water by volume, with dissolved constituents that
include dissolved salts of the product and trace metals, as well as process chemicals. Copper is the
primary ecotoxicological concern, because it is the principal product and is highly toxic to aquatic life.
Analyses of aqueous filtrate from samples taken on 3 different days, from a concentrate pipeline at a
porphyry copper mine with a separation process similar to that considered in the mine scenarios
(Section 6.1.2), reported copper concentrations of 500, 664, and 800 ug/L (Adams pers. comm.). The
mean of these values (655 ug/L) was used as the estimated copper concentration in this assessment.
Due to its relatively high toxicity, sodium ethyl xanthate is the highest risk ore-processing chemical that
could occur in the product concentrate slurry. We were unable to find an estimate of process chemical
concentrations in the concentrate slurry, but xanthate concentration would be 1.5 mg/L if we assume
that it occurs in the concentrate slurry at the same concentration as in tailings slurry (NICNAS 1995).
Unlike the metals, xanthate would degrade, but because its environmental half-life is approximately 260
hours (at pH 7 and 25°C) (NICNAS 2000) it could persist long enough to cause significant exposures
until diluted in Iliamna Lake.
Bristol Bay Assessment 1110 January 2014
-------
Chapter 11 Pipeline Failures
Flows in the potential receiving streams vary considerably. Measurements in streams along the
transportation corridor in 2004 and 2005 yielded a maximum observed flow of 58,000 L/s in the
Iliamna River and a minimum observed flow of 2.8 L/s in an unnamed stream (PLP 2011). Thus, full
mixing of spilled leachate could result in as much as a 33-fold dilution, but in smaller streams dilution
effectively would not occur. Of 12 monitored streams along the transportation corridor, only two had
observed flows in August 2004 (an estimate of summer low flow) that were greater than the estimated
flow of the aqueous phase of the slurry (PLP 2011: Table 7.3-10).
11.3.2.2 Solid Phase Chemical Constituents
If spilled product concentrate entered a stream, wetland, pond, or small lake directly or by overland flow
or erosion, it would flow for some distance, settle, and become substrate for invertebrates and possibly
salmon eggs and fry. In streams, it would be carried downstream by the current and would collect in
pools, behind debris, and in other localized low-flow areas. Some would settle into the cobble substrate
until high flows mobilized the bed. Much of the product concentrate could wash into Iliamna Lake,
where it could contribute to the substrate for spawning sockeye salmon.
Metal concentrations in the solid phase are expected to be similar to those of the Aitik product
concentrate (Table 11-3). Settled concentrate would be leached, resulting in direct aqueous exposure of
benthic invertebrates and fish eggs and larvae that inhabit the substrate to concentrations similar to
leachates from the Aitik product concentrate (Table 11-4). Local accumulation in streams could result in
local exposures to nearly pure concentrate and leachate. However, concentrate in Iliamna Lake would be
distributed and diluted to an extent that could not be estimated. Dietary exposure of fish is not
considered, because invertebrate abundance would be greatly diminished due to sediment toxicity, even
with considerable dilution by clean sediment.
11.3.3 Exposure-Response
Acute water quality criteria (criterion maximum concentrations [CMCs]), chronic criteria (criterion
continuous concentrations [CCCs]), and equivalent benchmark values are used as thresholds for
aqueous toxicity. Consensus sediment quality guidelines are used as thresholds for sediment solids
toxicity. These aqueous and sediment benchmark values are discussed in Section 8.2 and Section 9.5,
respectively. The biotic ligand model (BLM) generates low acute and chronic water quality criteria and
other toxicity values because of the extreme water chemistry of the leachate and receiving waters
(Section 8.2.2). However, the parameters are all within calibration range of the model (except for
alkalinity and dissolved organic carbon, which were set to minimum values because they were absent
from the leachate; this slightly raises criteria values) (HydroQual 2007).
In addition to the product concentrate and its dissolved constituents, the slurry would contain process
chemicals. Sodium ethyl xanthate is sufficiently toxic that it has been used as a pesticide (NICNAS 2000).
Exposure-response information for xanthates is summarized in Section 8.2.2.
Bristol Bay Assessment 1111 January 2014
-------
Chapter 11
Pipeline Failures
Table 11-3. Comparison of mean metal concentrations in product concentrate from the Aitik
(Sweden) porphyry copper mine (Appendix H) to threshold effect concentration and probable effect
concentration values for fresh water. Values are in mg/kg dry weight.
Concentrate
Constituents
Ag
As
Ba
Bi
Cd
Co
Cu
Ga
In
Mn
Mo
Ni
Pb
Sb
Te
Th
Tl
U
V
Zn
Sum of metals
Concentrations
>10
12
59
45
2.4
54
>10,000
0.88
2.4
345
1,100
72
65
43
4.1
1.5
0.2
2.2
23
2,200
-
TEC"
9.8
-
0.99
32
-
630
-
23
36
-
-
-
120
-
TEC Quotient"
1.2
-
2.4
>310
-
0.55
-
3.1
1.8
-
-
-
18
>340
PEC"
33
-
5.0
150
-
1,200
-
49
130
-
-
-
460
-
PEC Quotient"
0.36
-
0.48
>67
-
0.29
-
1.5
0.50
-
-
-
4.8
>75
Notes:
Dashes (-) indicate that values are not available.
a TECs and PECs are consensus values from MacDonald et al. (2000), except for Mn values, which are the TEL and PEL for Hyalella azteca 28-
day tests from Ingersoll et al. (1996).
b See Box 8-3 for a description of how risk quotients were calculated.
TEC = threshold effect concentration; PEC = probable effect concentration; TEL = threshold effect level; PEL = probable effect level.
11.3.4 Risk Characterization
Toxicological risk characterization is performed primarily by calculating risk quotients based on the
ratios of exposure concentrations to aquatic toxicological benchmarks (Box 8-3). However, it also
includes consideration of actual concentrate spills, the potential for remediation, and site-specific
factors.
11.3.4.1 Concentrate Pipeline Failure Scenarios
The concentrate pipeline failure scenarios and resulting spills would release 58,000 L of leachate to
Chinkelyes Creek or 28,000 L to Knutson Creek (Table 11-2). Risks to aquatic biota would result from
direct exposure to the aqueous phase of the slurry, the deposited concentrate, and in situ leachate from
the concentrate.
Bristol Bay Assessment
11-12
January 2014
-------
Chapter 11
Pipeline Failures
Table 11-4. Aquatic toxicological screening of leachates from Aitik (Sweden) product concentrate
(Appendix H) based on acute (criterion maximum concentration) and chronic (criterion continuous
concentration) water quality criteria or equivalent benchmarks, and quotients of concentrations
divided by benchmark values. Values are in ug/L unless otherwise specified.
Analyte
pH (standard units)
Spec, conductivity (|jS/cm)
Alkalinity (mg/L)
S04 (mg/L)
Si02(mg/L)
Ag
Al
As
Ba
Ca
Cl
Cd
Co
Cr
Cu
F
Fe
K
Mg
Mn
Mo
Na
Ni
Pb
Sb
Se
U
Zn
Sum of metals
Concentrations
5.4
260
0
120
59
<1
840
<1
38
27,000
800
3.5
140
<1
8,400
1,600
210
4,000
4,500
640
<2
890
480
11
13
7.3
11
1,300
-
Acute/Chronic Benchmarks
6.5-9
-
0.90V-
750/87
340/150
46,000/8,900
19/11
1.7/0. 22b
89/2.5
500/65b
12/7.9"
0.05/0.03=
-
350/-
-
760/690
32,000/73
-
410/46"
54/2. lb
14,000/1,600
-/5.0
33/15
100/100"
-
Quotients3
-
-
-------
Chapter 11 Pipeline Failures
less chronic, toxicity based on the copper criteria. These criteria are based on toxicity to sensitive
invertebrates, so the food base for salmonids could be severely reduced.
In all three streams, the diluted values are below the BLM-derived acute lethal levels for rainbow trout,
so a fish kill would not be expected (Table 8-13). Therefore, copper is not predicted to cause a kill of
adult salmonids in the receiving streams once mixing has occurred, but localized mortality could occur
in the mixing zone in the absence of avoidance behavior. However, fully diluted concentrations in
Chinkelyes Creek are above the chronic toxicity value for rainbow trout, suggesting that fry would be
affected. Concentrations at mean flow in Knutson Creek are a little below the rainbow trout chronic toxic
level (22 versus 26 ug/L), suggesting that effects on fry could occur at low flows.
Sodium ethyl xanthate, after fully mixing in Chinkelyes and Knutson Creeks, would occur at
approximately 0.1 and 0.07 mg/L. These values are at the low end of observed acutely lethal
concentrations for aquatic biota and below the observed median lethal concentrations for rainbow trout
(Section 8.2.2.5). Hence, the processing chemicals could contribute to acute toxicity in sensitive species.
The occurrence of acute toxicity depends on the exposure duration relative to the concentration. The
5.6- to 9.3-minute exposure duration (Table 11-2) may be sufficient to cause acute injury or lethality to
invertebrates or fish in receiving streams, given the high concentrations of copper (the rate of toxic
response is a function of the concentration) and given that the chronic effects of copper on fish include
lethality to fry. However, it would be more likely to cause acute effects in backwaters and ponds that
retained spilled water, and those areas are important rearing habitat for salmon (Appendix A).
Where the 32 to 66 metric tons of concentrate settled, sediment and benthic invertebrates and fish eggs
and fry would be exposed. The Aitik concentrate exceeds the sediment probable effect concentration
(PEC) for copper by more than a factor of 67 (Table 11-3). Hence, based on experience with other high-
copper sediments, any product concentrate from the Pebble deposit would be certain to cause toxic
effects on benthic organisms, including invertebrates and fish eggs and larvae. Because copper is
aversive to salmonids (Goldstein et al. 1999, Meyer and Adams 2010), the chronic leaching of copper
from deposited product concentrate may prevent returning salmon from using a contaminated stream
or river.
Exposure to pore water in sediments consisting of spilled product concentrate would be chronic. The
screening assessment performed here on Aitik concentrate leachate suggests that spilled concentrate
would cause severe toxic effects (Table 11-4). The 8,400 ug/L of dissolved copper in leachate would be
sufficient to kill benthic or epibenthic invertebrates and fish eggs and fry.
At mine closure, concentrate and return water pipelines would be removed. Therefore, these risks
would be limited to the approximately 78-year maximum operational life of the mine in the Pebble 6.5
scenario.
Bristol Bay Assessment 1114 January 2014
-------
Chapter 11 Pipeline Failures
11.3.4.2 Analogous Mines
No Alaskan mine has a product concentrate pipeline, but the 316-km, 175-mm-diameter product slurry
pipeline for the Bajo de la Alumbrera porphyry copper-gold mine in Argentina provides an analogue for
the pipeline considered here. It was reported that a 6.5-magnitude earthquake on September 17, 2004,
caused a break in the pipeline, releasing an unknown quantity of concentrate that caused the Villa Vil
River to overflow for approximately 2 km (Clap 2004, Mining Watch Canada 2005). The operators
reported that the 2004 spill was controlled in less than 2 hours and water for drinking and irrigation
was not contaminated (Minera Alumbrera 2004). They do not mention an earthquake, do not explain
why control required 2 hours, and attribute the failure to "an existing outer mark on the pipe" (Minera
Alumbrera 2004). Other pipeline failures with concentrate spills were reported in 2006 and 2007, but
not in other years (Minera Alumbrera 2004, 2005, 2006, 2007, 2008, 2009, 2010). They claimed that
those releases were small due to automatic shutoff, that concentrate did not reach water, and that "no
hazard is involved in concentrate handling since it is a harmless product consisting of ground rock"
(Minera Alumbrera 2006). Composition of this ground rock included 28% copper and 32% sulfur
(Minera Alumbrera 2006).
Operators subsequently built collection pits at pumping stations, monitored streams at pipeline
crossings, and brought water into the community of Amanao in part to mitigate effects of "potential
pipeline failure" (Minera Alumbrera 2008, 2010). They stated based on monitoring that pipeline
crossings of streams have no adverse effects on biodiversity, but they do not report monitoring to
address the effects of or recovery from the 2004 spill (Minera Alumbrera 2010). Although the interval
during which Minera Alumbrera has provided sustainability reports is too short to reliably estimate an
annual failure probability, it is notable that, despite International Organization for Standardization
14001 certification of the pipeline, it failed and released concentrate in 3 of 7 years.
More recently (July 25, 2012), a joint broke on the product slurry pipeline for the Antamina copper and
zinc mine in Peru (Briceno and Bajak 2012, Taj and Cespedes 2012). It released 45 metric tons of slurry
over 2 hours, of which 3 metric tons escaped the containment area. Local villagers intervened to stop the
flow of slurry to the nearby Rio Fortenza. A mine spokesperson stated that the river showed no signs of
contamination and the material was only an irritant, although a company document called the
concentrate very toxic (Taj and Cespedes 2012). An Associated Press photo shows workers in white
protective suits apparently cleaning a channel. News reports and Minera Antamina's press releases on
the event emphasized human health effects: 210 people received medical treatment and 45 were
hospitalized, apparently due to inhalation of aerosolized slurry. People reported a strong pesticide odor,
which suggests significant concentrations of a xanthate collector chemical, but no analyses have been
reported. Ecological effects are unknown. Antamina is a modern mine (operation began October 1,
2001) where sustainability is said to be given a higher priority than cost or profitability (Caterpillar
Global Mining 2009). As in the mine scenarios evaluated here, the pipeline is buried except at bridges
and is monitored using a parallel fiber optic system.
Bristol Bay Assessment 1115 January 2014
-------
Chapter 11 Pipeline Failures
Product concentrate spills from pipeline failures have also occurred at the Bingham Canyon mine in
Utah. Between May 31 and June 2, 2003, operators reported to the U.S. Coast Guard's National Response
Center a spill of 70 tons (63.5 metric tons) of product concentrate from a pipeline failure. On October 2,
2009, they reported a pipeline leak that spilled 1,400 gallons (5,300 L) of copper concentrate.
Although the Alumbrera, Antamina, and Bingham Canyon cases do not provide evidence concerning the
ecological effects of a concentrate spill, they do support the plausibility of pipeline failures leading to
concentrate spills. Our estimated pipeline failure rate of one per 1,000 km-year (Section 11.1) implies a
failure rate of 0.32 per year for the 316-km Alumbrera pipeline, which is similar to the 0.43 observed
rate at Alumbrera from 2004 to 2010. These cases indicate that concentrate pipeline failures do occur at
modern copper mines operated by large international mining companies, and that they can result in
spills that are potentially larger than our assumptions indicate.
11.3.4.3 Concentrate Spill Remediation
Remediation of a product concentrate spill would be less problematic than remediation of a tailings spill.
The concentrate is valuable, it would be spilled near a road, and the volume would be much smaller than
a potential tailings spill. Hence, remediation would likely occur relatively quickly if the spill occurred on
land or in a wetland, by excavating or dredging the concentrate and trucking it back to the mine.
However, because concentrate would be carried downstream by high or typical flows in the receiving
streams, substantial recovery of material spilled into a stream is unlikely except possibly during low-
flow periods (less than one-ninth of mean flows). The proportion recovered by dredging would depend
on the circumstances, the rapidity of response, and the balance between the desire to minimize habitat
damage and to reduce potential toxic effects. If the spill was associated with high flows, it is likely that
little of the material would be recovered from a stream even if the entire stream was dredged. Dredging
in Iliamna Lake might be feasible if concentrate was not too dispersed or diluted by other sediment.
11.3.4.4 Weighing and Summarizing the Evidence
Past experience with pipelines in general, and with the Alumbrera, Antamina, and Bingham Canyon
product concentrate pipeline failures in particular, suggests that pipeline failures and product spills
would be likely in the Pebble 6.5 scenario. A concentrate spill into a stream is likely to kill invertebrates
and early fish life stages immediately. If it is not remediated (and remediation of streams may not be
possible), it would certainly cause long-term local loss offish and invertebrates. The settled concentrate
would become sediment, which would be toxic to fish and invertebrates in the receiving streams for
many years. Ultimately, this settled concentrate would reach Iliamna Lake, where it could be toxic to the
eggs and larvae of sockeye salmon until it was sufficiently mixed with or buried by clean sediment. The
length of streams affected in the scenarios would be 14 km of Chinkelyes Creek and 7.6 km of the
Iliamna River for a release to Chinkelyes Creek, or 2.6 km of Knutson Creek for a release there. The area
of the lake that would experience toxic effects cannot be estimated at this time.
The weighing of these lines of evidence is summarized in Table 11-5. For each route of exposure, sources
of the exposure estimate and the exposure-response relationship are indicated. All evidence is
Bristol Bay Assessment 1116 January 2014
-------
Chapter 11
Pipeline Failures
qualitatively weighed based on three attributes: its logical implication, its strength, and its quality
(Suter and Cormier 2011). For logical implication, possible scores are indicated as (+) for results
supportive of adverse effects on the endpoint populations, (-) for results contrary to adverse effects on
assessment endpoints, and (0) for neutral or ambiguous results. In this case, the logical implication is
that the concentrate pipeline failure scenario evaluated here would have adverse effects. The strength of
the evidence is based primarily on the magnitudes of the hazard quotients (exposure concentrations
divided by effects concentrations): a low quotient is indicated as (0), a moderate quotient as (+), and a
high quotient as (++). Quality is a more complex concept. It includes conventional data quality issues,
but in this case the primary determinant is the relevance of the evidence to the mine scenario. Separate
quality scores are provided for the exposure estimate and for the exposure-response relationship. The
scores are intended to remind the reader what evidence is available and show the pattern of strength
and quality of the several lines of evidence and to transparently present our weighing process and
results.
Table 11-5. Summary of evidence concerning risks to fish from a product concentrate spill. The risk
characterization is based on weighing four lines of evidence for different routes of exposure. All
evidence is qualitatively weighed (using one or more +, 0, -symbols) on three attributes: logical
implication, strength, and quality. Here, all lines of evidence have the same logical implication—that
is, all suggest a concentrate spill would have adverse effects. Strength refers to the overall strength of
the line of evidence, and quality refers to the quality of the evidence sources in terms of data quality
and relevance of evidence to the spill scenario.
Route of Exposure
Source of Evidence (Exposure/E-R)
Dissolved copper
Measurements from analogous
mine and dilution
model/Laboratory-based
benchmarks
Concentrate particles
Undiluted concentration from
analogous mine/Field-based
benchmarks
Concentrate leachate
Leachatefrom analogous mine/
Laboratory-based benchmarks
Actual spills
Amount spilled/None
Summary weight of evidence
Logical
Implication
+
+
+
0
+
Strength
++
++
++
0
++
Quality
Exposure
+
0
0
0
0
E-R
++
+
++
0
+
Results
Lethality to invertebrates is certain
and sensitive larval fish may also
be killed.
The concentrate would clearly form
toxic sediment but its distribution
in unclear.
Invertebrates and fish in sediment
would experience toxic effects
unless the concentrate was highly
diluted.
The record indicates that
concentrate spills occur but
exposure and effects have not
been studied.
A spill is likely to occur and toxicity
to aquatic biota is highly likely.
Notes:
E-R = exposure-response relationship.
Overall, available lines of evidence for effects of a concentrate spill are positive (i.e., supportive of the
hypothesis that acute and chronic toxic effects would occur) (Table 11-5). The quality of the exposure-
response information is good, but the quality of the exposure information for the deposited concentrate
and its leachate is uncertain because of the uncertain potential for dispersal in streams. The analogous
Bristol Bay Assessment
11-17
January 2014
-------
Chapter 11 Pipeline Failures
spills provide no information on exposure or effects beyond confirming that concentrate spills do occur.
However, this evidence supplements the more extensive experience with oil pipelines (Section 11.1),
which suggests that a spill is likely.
If the spill could be remediated, some fraction of the concentrate (but none of the leachate) could be
recovered and the extent of chronic (but not acute) toxic effects would be diminished. The proportion of
concentrate recovered would depend on spill location, time of year, diligence of the operator, and the
amount of physical damage due to remediation that is considered acceptable. Concentrate spilled into
streams would be unlikely to be recovered unless streamflows were particularly low. Recovery of the
concentrate would require excavation of streambeds, wetlands, or uplands, depending of the location of
the spill. When determining how thoroughly to excavate and, in particular, how far downstream to
dredge the stream, reduction in toxicity would need to be balanced against habitat destruction.
The effects of a spill on salmonid populations would depend on the receiving waters. Streams along the
transportation corridor that might receive a spill (described in Section 10.1) are quite variable.
Chinkelyes Creek receives an average of more than 9,000 spawning sockeye salmon and flows to the
Iliamna River, which receives an average of more than 100,000 sockeye spawners (Table 10-2). Knutson
Creek receives an average of roughly 1,500 sockeye spawners and flows to Knutson Bay, which receives
an average of 73,000 beach spawning sockeye (Table 10-2). Not all of those salmon spawn below stream
crossings, but copper leaching from concentrate spills could be aversive to salmon and thereby reduce
spawning production along the entire stream lengths. Also, the concentrate deposited in Knutson Bay
would persist and could render a considerable area unsuitable for spawning and rearing for years. In
any case, these values indicate that a non-trivial number of spawners and potential salmon production
would be at risk.
Potential effects on those salmon and other fishes in the receiving waters would include the following.
• Reduced production of salmon fry and parr and all life stages of other salmonids from the loss of
invertebrate prey due to extensive acute lethality during the spill and persistent chronic toxicity in
areas where the concentrate deposited.
• Loss of a year-class of salmon and other salmonids due to direct acute toxicity during and
immediately following a concentrate spill.
• Loss of salmon spawning habitat due to avoidance of copper in areas of deposition and possibly in
the entire stream, if aqueous concentrations from leaching concentrate were sufficiently high.
• Persistent chronic toxicity to salmonid eggs and fry in areas of concentrate deposition, where it is
not aversive to spawning adults.
11.3.5 Uncertainties
Based on multiple lines of evidence, it is certain that a spill from a product concentrate pipeline into a
stream would cause toxic effects. However, there are uncertainties regarding individual pieces of
evidence, which are summarized below.
Bristol Bay Assessment 1118 January 2014
-------
Chapter 11 Pipeline Failures
• The composition of the product concentrate and its leachate are uncertain, because they are based
on a surrogate material and because leaching test conditions are inevitably somewhat artificial.
Copper concentrations in North and South American copper concentrates generally fall in the 200 to
340 mg/kg range, so variance of a factor of 2 is a reasonable estimate for potential variance in
Pebble deposit concentrate from Aitik concentrate. Hence, uncertainty concerning the major source
of toxicity is not large, and therefore it is implausible that the concentrate and its leachate would be
nontoxic to aquatic biota. An informal internet search for copper concentrate compositions suggests
that minor metals differ by an order of magnitude among copper concentrates. Thus, it is possible
that metals other than copper may be significant contributors to toxicity.
• The copper concentration of the aqueous fraction of the slurry is also based on analyses from an
existing mine. However, estimates based on the existing mine are realistic, given that the ore type
and processing are believed to be very similar and that the leachate was formed during actual
operations rather than in a test. Therefore, this uncertainty is estimated to be at least a factor of 2
but no more than 5. Effects on invertebrates are certain, but effects on fish may not occur or may be
more severe than estimated.
• The composition of the aqueous fraction of the slurry is unknown for constituents other than
copper. Although it is certain that copper is by far the most toxic metal in the slurry, the composition
of other constituents is unknown. Sodium ethyl xanthate is highly toxic and might increase the
toxicity of a spill. Combined metal toxicity would make some difference but is unlikely to change the
qualitative conclusions.
• The 5-minute time to shut-off is uncertain, and this estimate appears to be conservative. For
example, Trans Canada's risk assessment for the Keystone XL pipeline assumed that the time to
detection would range from 90 days for a small leak (1.5% of pumping volume) to 9 minutes for a
large leak (50% of pumping volume), and that an additional 2.5 minutes would be required for the
shutdown sequence (DNV Consulting 2006, O'Brien's Response Management 2009). This suggests
that a large spill like the one assessed here would leak for 11.5 minutes based on a state-of-practice
design from an experienced company, which is more than twice our assumed duration.
• The 5-minute time to shut-off depends on successful operation of a remote shutoff system. The
potential for a larger spill if the shutoff failed (e.g., if an earthquake damaged the pipeline and the
shutoff system) or was overridden by the operators is unknown. There are precedents for large
spills but not enough data to quantify the risk.
• The frequency and location of spills are also uncertain. The extensive experience with oil and gas
pipelines provides probabilistic estimates, but these estimates vary considerably among studies.
The more directly relevant experiences with concentrate pipelines at Alumbrera, Antamina, and
Bingham Canyon mines suggest that estimates based on oil and gas pipeline failure rates are
consistent with mining-related pipeline failures.
Bristol Bay Assessment 1119 January 2014
-------
Chapter 11
Pipeline Failures
11.4 Return Water Pipeline Failure Scenarios
A spill from a return water pipeline would result in an acute aqueous exposure (Table 11-6), as
discussed above for a product concentrate spill. The return water is expected to be the same as the
aqueous phase of the concentrate slurry (i.e., it would not be treated at the port), although estimated
flow rates would differ. Hence, copper concentration in the return water is assumed to be the mean of
analyses of the aqueous phase of slurry from a Rio Tinto mine (655 ug/L). Both acute and chronic
criteria would be exceeded, but because of the short spill duration and the absence of a persistent solid
phase, toxic effects would not be expected to be so severe as a product concentrate spill. Effects would
be most likely in low-flow habitats such as backwaters, ponds, and bays. We know of no analogous
return water pipeline failures that might be used to assess this risk; however, experience with pipelines
in general suggests that multiple failures and spills would occur over the life of the mine, and at least one
would be expected to occur at or near a stream (Section 11.1).
Table 11-6. Parameters for return water pipeline spills to Chinkelyes and Knutson Creeks.
Parameter
Spill into Chinkelyes Creek
Chinkelyes Creek
Iliamna River
Spill into Knutson Creek
Knutson Creek
Water Flow
Discharge (m3/s)
Velocity (m/s)
Channel Length (km)
1.8
2.2
14
22
2.0
7.6
3.4
2.2
2.6
Pipeline Drainage and Dilution
Flow rate while draining (m3/s)
Flow rate while pumping (m3/s)
Release time— draining (minutes)
Release time— pumping (minutes)
Volume spilled (L)
Maximum concentration dissolved copper (Mg/L)
Travel time to confluence (minutes)3
0.09
0.03
8.6
5.0
56,000
39
110
-
-
3.5
64
0.06
0.03
5.1
5.0
27,000
17
19
Pipeline and Return Water Specif ications
Length from top of nearest hill to valve (m)
Elevation drop (m)
Viscosity of return water (cP)
Density of return water (metric tons/m3)
2100
150
-
810
25
1
1
Notes:
Dashes (-) indicate that spill is not directly into Iliamna River, which receives flow from Chinkelyes Creek.
8 Confluence with Iliamna River for Chinkelyes Creek; confluence with Iliamna Lake for the Iliamna River and Knutson Creek.
11.5 Diesel Pipeline Failure Scenarios
As with the product concentrate pipeline, effects of a diesel pipeline failure would depend on many
factors, including pipeline design, location of the pipeline failure along the transportation corridor, and
time of year at which the pipeline failure occurred. Parameters for the diesel pipeline failure scenarios
are presented in Table 11-7.
Bristol Bay Assessment
11-20
January 2014
-------
Chapter 11
Pipeline Failures
Table 11-7. Parameters for diesel pipeline spills to Chinkelyes and Knutson Creeks.
Parameter
Spill into Chinkelyes Creek
Chinkelyes Creek
Iliamna River
Spill into Knutson Creek
Knutson Creek
Water Flow
Discharge (m3/s)
Velocity (m/s)
Channel Length (km)
1.8
2.2
14
22
2.0
7.6
3.4
2.2
2.6
Pipeline Drainage and Dilution
Flow rate while draining (m3/s)
Flow rate while pumping (m3/s)
Release time— draining (minutes)
Release time— pumping (minutes)
Volume— total (m3)
Volume % diesel to water in stream at spill
Mass of diesel in stream at input (mg/L)
Maximum concentration dissolved diesel (mg/L)
Distance traveled during release (km)
Travel time to confluence (minutes)3
0.035
0.005
13
5
30
2.2%
17,000
1.9-7.8
1.7
110
-
-
-
-
1,500
1.7-7.2
64
0.023
0.005
7.9
5
12
0.83%
6,500
1.9-7.8
1.1
19
Pipeline and Diesel Specifications
Length from top of nearest hill to valve (m)
Elevation drop (m)
Viscosity of diesel at 15°C (cP)
Density of diesel at 15°C (metric tons/m3)
2100
150
-
810
25
2
0.85
Notes:
Dashes (-) indicate that spill is not directly into Iliamna River, which receives flow from Chinkelyes Creek.
8 Confluence with Iliamna River for Chinkelyes Creek; confluence with Iliamna Lake for the Iliamna River and Knutson Creek.
11.5.1 Sources
11.5.1.1 Pipeline Failure
The volume of material released from a pipeline leak would depend on the type of failure, rate of loss
from the pipe, pumping rate, leak duration, pipe diameter, distance to the nearest shutoff valves, and
time until those valves are closed. For the purposes of this assessment, we evaluate a full break or a
defect of equivalent size in the diesel pipeline that occurs at a stream crossing, thereby releasing fuel
into that aquatic ecosystem. This could occur as a result of mechanical failure of the pipe from ground
movement, vehicle impact, material failure or other cause. Characteristics of the pipeline are described
in Table 6-4. We analyzed spills to two streams that would be crossed by the transportation corridor,
Chinkelyes Creek and Knutson Creek (Section 11.2).
11.5.1.2 Diesel Fuel Composition
In the diesel pipeline failure scenarios, the pipeline would contain fuel from one of the Alaskan refineries
and would have a composition similar to those presented by Geosphere and CH2M Hill (2006). Diesel
fuel is a mixture of many hydrocarbon compounds, and its composition is a function of the petroleum
feedstock source and the refining process. The type and amount of water-soluble hydrocarbons in the
Bristol Bay Assessment
11-21
January 2014
-------
Chapter 11 Pipeline Failures
diesel determine the dissolved aqueous concentration when mixed with water. The most soluble
compounds in diesel are the volatile aromatic hydrocarbons benzene, toluene, ethylbenzene and xylene
(together, BTEX). Most diesel fuels have a low proportion of these soluble compounds and therefore
have low solubilities. The bulk of diesel fuel is made up of heavier hydrocarbons that are essentially
insoluble. A study of the composition of four diesel fuels from Alaskan refineries showed that the fuels
had less than 2% BTEX and resulting diesel solubilities of 1.89 to 7.81 mg/L.
In the analysis of concentrations and solubilities, we incorporate all hydrocarbon compounds in the
diesel samples and calculate the solubility based on Raoult's Law to account for effects of the mixture on
the solubility of individual compounds.
11.5.2 Exposure
11.5.2.1 Background
A failure of the diesel pipeline in these scenarios could occur in the buried or above-ground portions. An
above-ground failure would occur at a bridged stream or river crossing. An underground failure would
result in diesel leaking into the soil and flowing down-gradient (e.g., as in the Trans-Alaska pipeline
failure described in Section 11.5.3.3). If the underground failure occurred below a stream, it would float
upward and into the surface water. An above-ground failure would release diesel directly to a river or
stream, a wetland, or upland soil.
The behavior of diesel fuel in fresh water is less well-studied than the behavior of crude oil or diesel in
marine environments. Diesel fuel has a density of less than 1.0 metric ton/m3 and floats on water. It
typically dissolves or evaporates within a day. In turbulent stream reaches, diesel would form small
droplets suspended in the water column.
The soluble fraction would mix into the streamflow, be transported by advection and dispersion, and
flow with the water. Solubility decreases with temperature, so in colder temperatures a smaller amount
is dissolved in the stream. The soluble fraction is attenuated through dilution (advection and
dispersion), biological activity, photodegradation, and aeration in turbulent streams, but is renewed by
dissolution from the floating oil. The soluble compounds are also susceptible to evaporation from the
floating oil, which typically occurs at a faster rate than dissolution. The soluble fraction compounds have
relatively short residence times in water and sediments (Hayes et al. 1992) and can be reduced to below
detection levels in a few days or weeks, depending on site-specific conditions.
Diesel components that are lighter than water and have low solubility tend to spread on the surface and
form a thin film or sheen less than 0.1 mm thick. As the diesel spreads, it is more susceptible to
destruction by evaporation, dissolution, and photodegradation but is also more likely to contact and
attach to suspended sediments and shorelines. Most of the spilled diesel would flow with the stream
until it reached Iliamna Lake and dissipated. The pour point of diesel (the temperature below which the
oil will not flow) is approximately -7°C (20°F); thus, if the spill occurs during cold weather, the diesel
would be less likely to spread and would instead form globs or strings and become suspended within the
Bristol Bay Assessment 1122 January 2014
-------
Chapter 11 Pipeline Failures
water column. For example, a 1999 cold-weather diesel spill in the Delaware River resulted in more than
90% of the diesel forming globules that were not visible from the surface (Overstreet and Gait 1995).
Oil dispersed in the water column can adhere to fine-grained suspended sediments that settle and
deposit on stream edges and bottoms in low-energy areas. Depending on the source of the diesel, there
may be a significant portion of compounds that are heavier than water and therefore sink, sorb to
sediments, and persist longer than the dissolved fraction. In wetlands or pools and slack water areas of
streams, a large percentage of spilled diesel can be deposited in the sediments.
When spilled on ice, diesel is viscous and forms tar-like accumulations on the surface. Lighter diesel
components can penetrate the ice, become trapped within the ice structure, and be released as the ice
melts. If the spill is trapped below the ice, as is more likely with buried pipelines, it would spread and
stick to the underside of the ice in thin layers. Because cold temperatures reduce the solubility of diesel
components, less would be dissolved in the stream water (NOAA and API 1994). As the ice breaks up
and melts, the diesel would be released from the ice and mix with the stream water.
Because of its low viscosity (except in cold weather), diesel spilled onto the land tends to be rapidly
absorbed by soil so that an above-ground spill on land could soon resemble an underground spill. In this
area, where the groundwater surface tends to be shallow, spilled diesel would flow on top of the
groundwater and a fraction would dissolve in that groundwater. It would then flow down-gradient to
any nearby stream, possibly passing through wetlands on the way. Upon reaching a stream, it could pass
into the channel through the gravels in which salmon, trout, and Dolly Varden spawn. In some locations,
it might flow to Iliamna Lake and pass through a deltaic spawning beach used by sockeye salmon. Diesel-
contaminated soil could episodically contaminate water when the water table rises following rain or
snow melt. The extent to which fish eggs or fry are exposed by this route would depend on the specific
structure of the spill site. Given the abundance of streams, wetlands, and shallow groundwater in the
area crossed by the diesel pipeline, some variant of this exposure route is likely. However, saturated
soils and particularly those that are frozen could result in overland rather than groundwater flow of
diesel fuel.
The primary cause of toxicity to aquatic organisms in oil spills is direct exposure to the dissolved
fraction. Exposure via this route would occur immediately following a direct spill to a stream or wetland
as the oil dissolved, resulting in an acute exposure. Longer exposures to dissolved oil could result from
slow releases of oil from terrestrial spills, flows from oiled wetlands, or the gradual dissolution of oil
sorbed to sediments or plant materials. Oil spills can indirectly expose aquatic organisms to low
dissolved oxygen as microbes decompose the oil.
Which of these transport and exposure processes would occur in a diesel spill depends on the spill
location. The number and nature of water body crossings are the same as for the other pipelines
(Section 11.2).
Bristol Bay Assessment 1123 January 2014
-------
Chapter 11 Pipeline Failures
11.5.2.2 Transport and Fate
In the diesel pipeline failure scenarios, a pipeline failure would result in release of diesel directly into
either Chinkelyes or Knutson Creek at mean streamflows (Table 11-7). The spill at Knutson Creek would
release 12,000 L of diesel into approximately 1.6 million L of stream water, resulting in a 1:130 dilution.
At Chinkelyes Creek, the spill would release approximately 30,000 L of diesel into 1.5 million L of stream
water, resulting in a 1:49 dilution. At a typical diesel density of 850 g/L, this would result in 6,500 and
17,000 mg diesel/L water in Knutson and Chinkelyes Creeks, respectively. Both of these dilutions are
less than the minimum aqueous volume required to get below the saturation of the diesel, if the
dissolved hydrocarbons are well-mixed. This conclusion is based on calculation of the minimum volume
of water required for diluting each component to a concentration below saturation. For benzene, the
minimum volume of water required for dilution below saturation is 169 to 225 L benzene/L diesel; all
other components would require higher dilutions. Thus, it is reasonable to assume that at both spill
locations the diesel would be at saturation (i.e., at concentrations between 1.89 and 7.81 mg/L) in the
receiving waters. Concentrations in the Iliamna River would be lower due to depletion of benzene. The
benzene concentration would fall below its Raoult's saturation limit, resulting in a diesel concentration
of 1.7 to 7.3 mg/1 and a saturation of 92 to 94%.
11.5.3 Exposure-Response
Diesel is considered to be one of the most acutely toxic petroleum products (NOAA 2006), but its
composition is variable. Although a model exists for estimating the acute aquatic toxicity of petroleum
products from their chemical composition (Redman et al. 2012), the composition of diesel that would be
piped to the mine is unknown. For example, the compositions of water-soluble fractions of two brands
of Alaskan diesel fuel were found to be C4-C6 non-aromatic hydrocarbons (0.4-1.2 mg/L), benzene
(0.03-0.2 mg/L), toluene (0.03-0.2 mg/L), and C2 benzenes (0.005-0.1 mg/L) (Guard et al. 1983). Given
this variance in composition, data from laboratory tests and field studies of various whole diesel oils are
used in this section to indicate the range of toxic effects observed in response to different exposures.
11.5.3.1 State Standards
According to Alaska water quality standards (ADEC 2011), total aqueous hydrocarbons in the water
column may not exceed 15 ug/L and total aromatic hydrocarbons in the water column may not exceed
10 ug/L. The standards state (ADEC 2011): "There may be no concentrations of petroleum
hydrocarbons, animal fat, or vegetable oils in shoreline or bottom sediments that cause deleterious
effects to aquatic life. Surface waters and adjoining shorelines must be virtually free from floating oil,
film, sheen, or discoloration."
11.5.3.2 Laboratory Tests
Laboratory tests of the toxicity of petroleum and its derivative fuels to aquatic organisms are performed
with either an oil-water dispersion or a dissolved solution, called the water-soluble fraction. Dispersions
are created by adding oil to water at prescribed ratios and mixing. The vigor and duration of mixing is
variable, ranging from gentle mixing with a stirring rod to extended mixing with a magnetic stirrer. The
Bristol Bay Assessment 1124 January 2014
-------
Chapter 11 Pipeline Failures
resulting dispersion may have an oil layer on the surface as well as suspended oil droplets, although
most tests attempt to avoid suspended material. The oil layer may be left in the test container, but more
often the aqueous material is drawn off for the test. Results may be expressed as mg diesel/L or volume
percent diesel. Water-soluble fractions are created by mixing oil and water to create a nominally
saturated solution. The aqueous solution is drawn off and should be filtered to remove any suspended
oil droplets. It is then diluted in water to create the test media. Results may be expressed as mg
hydrocarbons/L or percent water-soluble fraction. In theory, one could also use toxicity data for each of
the component chemicals in diesel fuel and estimate the combined effect based on individual effects, but
that approach was judged to be impractical given uncertainties about diesel fuel composition in the
scenarios and the paucity of toxicity data.
Potentially relevant results of tests of diesel dispersions and water-soluble fractions are summarized in
Table 11-8. Results range over 4 orders of magnitude, and are highly variable even within an individual
species or test type. This range results from differences in test procedures and diesel fuel compositions.
Tests with biodiesel, synthetic diesel, sub-organismal endpoints, salt water, and dispersants were not
included.
11.5.3.3 Analogous Spills: Diesel in Streams
Diesel spills into streams and wetlands are not uncommon, but their biological effects are seldom
determined and published. Relevant diesel spill case studies are summarized in Table 11-9 and
discussed in the text below. None of these studies were conducted in the Bristol Bay region, so they
provide only a general indication of the nature and duration of effects expected from an instream diesel
spill. We found no publications describing biological effects of diesel spills in relevant wetland habitats.
Multiple diesel spills have been associated with construction of the Trans-Alaska Pipeline, but biological
effects were studied only for a 1972 spill from a broken underground pipeline that released 3,750 L to
Happy Valley Creek (during spring streamflows of 14 m3/s). Biological effects of the spill were studied
downstream in the Sagavanirktok River (Nauman and Kernodle 1975, Alexander and VanCleve 1983).
Invertebrate abundance declined by 89% after the spill (Nauman and Kernodle 1975), and stonefly and
caddisfly nymphs were eliminated from the stream (Alexander and VanCleve 1983). Recovery was not
reported.
A pipeline spill into Camas Creek, Montana, of oil that "most strongly resembled diesel fuel" resulted in
low abundance and low richness of the invertebrate community with few mayfly, stonefly, and caddisfly
taxa (Van Derveer et al. 1995). After remediation that included stream diversion and extensive removal
of contaminated soil below the spill and recovery for approximately 1 year, taxa richness and abundance
at the spill site were 60 to 70% of the upstream reference site, whereas at sites farther downstream
from the remediation activities taxa richness and abundance were less than 15% and 10% of the
reference site levels, respectively.
A tanker truck wreck in Trinity County, California, resulted in the flow of approximately half of a 15,000-
L tank of diesel fuel into Hayfork Creek, a tributary of the Trinity River (Bury 1972). The oil was spilled
Bristol Bay Assessment 1125 January 2014
-------
Chapter 11
Pipeline Failures
on land and reached the stream after 36 hours. An area 1 to 2.5 miles below the spill was surveyed,
because it had been previously studied. Numerous dead organisms were collected, including 4,469
vertebrates (rainbow trout and other fishes, tadpoles, snakes, turtles, and a bird) and uncounted
thousands of macroinvertebrates. Recovery was not monitored.
A 1980 pipeline break released 340 m3 of Number 2 fuel oil to a small tributary of Mine Run Creek,
which ultimately flows to the Rapidan River, Virginia (Bass et al. 1987). The operator reported collecting
240 m3 of oil. Monitoring was initiated 4 months after the spill, so acute effects were not observed.
Standing crop, density, and diversity of macroinvertebrates were reduced in Mine Run Creek
downstream of the tributary, and caddisflies were particularly affected. Effects were still observed at
16 months, when the study ended.
Table 11-8. Toxicity of diesel fuel to freshwater organisms in laboratory tests.
Species
Life Stage3
Test Endpoint
Concentration
Source— Notes
Water-Soluble Fraction
Rain bow trout
Rain bow trout
Dap/i n/a magna
Microcystis
aeruginosa
Selenastrum
capricornutum
Pseudokirchneriella
subcapitata
Free-swimming
embryos
2 months after
yolk resorption
1st instar
Culture
Culture
Cultures
9-day LCso
48-hour LCso
48-hour ECso
4-hour carbon
fixation
4-hour carbon
fixation
96-hour ICso
8 mg/L
2.43 mg/L
6.7%
100%
100%
58.7%
Schein et al. 2009-total dissolved
hydrocarbon concentration
Lockhartet al. 1987— total hydrocarbon
concentration
Giddings et al. 1980— percent water soluble
fraction
Giddings et al. 1980— significant inhibition
as percent water soluble fraction
Giddings et al. 1980— significant inhibition
as percent water soluble fraction
Pereira etal. 2012— inhibition of growth as
percent water soluble fraction
Aqueous Dispersion
Coho salmon
Coho salmon
Pink salmon
Rain bow trout
Rain bow trout
Rain bow trout
Rain bow trout
Fathead minnow
Dap/i n/a magna
Dap/i n/a magna
Dap/i n/a magna
Chironomidae
Selenastrum
capricornutum
Juvenile
Fry
Juvenile
Juvenile
Fry
Swim-up fry
Juvenile
Juvenile
Juvenile
Unspecified
Juvenile
Larvae
Culture
96-hour LCso
96-hour TLm
96-hour LCso
96-hour LCso
14-day LCso
72-hour LCso
96-hour LCso
96-hour LCso
24-hour LCso
96-hour LCso
48-hour LCso
96-hour LCso
72-hour ELso
10,299 mg/L
2,870 mg/L
74 mg/L
3,017 mg/L
44.9 mg/L
133.52 mg/L
31 (6.6-65) mg/L
57 mg/L
1.78 mg/L
20.0 mg/L
36 (2-210) mg/L
346 mg/L
20 (1.8-78) mg/L
Wan et al. 1990-soft water
Hebertand Kussat 1972
Wan et al. 1990-soft water
Wan et al. 1990-soft water
Mos etal. 2008
Khan etal. 2007
API 2003— mean and range of three tests
API 2003
Khan etal. 2007
Das and Konar 1988
API 2003-mean and range of 12 tests
Das and Konar 1988
API 2003— mean and range of seven results
from three endpoints (inhibition of cell
density, biomass, or growth) and three tests
Notes:
a As described by the authors.
LCso =median lethal concentration; ECso = median effective concentration; ICso = median inhibitory concentration; TLm = equivalent to LCso;
ELso = median effective level.
Bristol Bay Assessment
11-26
January 2014
-------
Chapter 11
Pipeline Failures
Table 11-9. Cases of diesel spills into streams. For comparison, the diesel pipeline failure scenarios
evaluated here would release 30 and 8 m3 of diesel into receiving streamf lows of 1.8 and 3.4 m3/s
for spills into Chinkelyes Creek and Knutson Creek, respectively.
Case
Happy Valley Creek, AK
Camas Creek, MT
Hayfork Creek, CA
Mine Run Creek, VA
Reedy River, SC
Cayuga Inlet, NY
Westlea Brook, UK
Hemlock Creek, NY
Diesel Released (m3)
3.7
Unknown
15
240
3,600
26
9.8
0.5
Receiving Streamflow (m3/s)
14
0.42
4.1
1.2
6.4
1.8
1.34
0.76
Observed Effects
Significant declines in the abundance
and species richness of invertebrates
Low invertebrate abundance and
richness
Large kill of vertebrates and
invertebrates
Reduced invertebrate abundance and
diversity
Near-complete fish kill
Fish kill and reduced abundance,
reduced invertebrate abundance and
species composition
Fish kill, invertebrates severely affected
No significant effects on invertebrates
Notes:
'" Mean flow from NHDPIus v2; others as reported by the authors.
In 1996, a pipeline ruptured and released 22,800 barrels (3.6 million L) of diesel into the Reedy River,
South Carolina (Kubach et al. 2011). That spill resulted in a severe fish kill for 37 km downstream to the
confluence with a reservoir. Recovery of the fish community, based on non-metric multidimensional
scaling, occurred after 52 months.
In 1997, a train wreck spilled an estimated 26,500 L of diesel into Cayuga Inlet, a tributary stream of
Cayuga Lake, New York (Lytle and Peckarsky 2001). Despite containment efforts, a kill occurred, which
reduced fish (including rainbow trout) abundance by 92% and invertebrate abundance by 90%.
Invertebrate density recovered within 1 year, but species composition had not recovered after
15 months.
In 2005, 9,800 L of diesel spilled into Westlea Brook in Wiltshire, UK (Smith et al. 2010). Due to its urban
location, response was rapid, and approximately 7,000 L were recovered. However, the spill killed
approximately 2,000 fish and a few frogs and birds. Invertebrate surveys showed that
macroinvertebrates were severely affected and impacts were discernible for 4 km. Recovery occurred
within the 13.5-month sampling period for all but the most affected site.
A tank of home heating oil (described as similar to diesel) leaked 500 L and an unknown amount
entered Hemlock Creek, New York (Coghlan and Lund 2005). Three days after the spill, a survey of
benthic invertebrates below the spill site found no significant reduction in the Hilsenhoff index (Coghlan
and Lund 2005). The authors concluded that their techniques were sufficiently sensitive and no
significant effects resulted from this small spill.
Bristol Bay Assessment
11-27
January 2014
-------
Chapter 11 Pipeline Failures
11.5.3.4 Analogous Spills: Crude Oil in Salmon Spawning Streams
The Exxon Valdez oil spill infiltrated the beaches of tidal Alaskan streams that provide spawning habitat
for pink salmon (Rice et al. 2007). Water draining over the buried oil dissolved hydrocarbons, exposing
salmon eggs and resulting in embryo histopathology and mortality for at least 2 years after the spill. The
type of oil spilled and the circumstances of the spill are different from the diesel pipeline failure
scenarios, but the studies described by Rice et al. (2007) demonstrate that oil buried near spawning
habitats can be a source of potentially toxic exposures for years.
11.5.4 Risk Characterization
Toxicological risk characterization is performed primarily by calculating risk quotients based on the
ratios of exposure levels to aquatic toxicological benchmarks (Box 8-3). However, it also includes
consideration of actual diesel spills, the potential for remediation and recovery, site-specific factors, and
the overall weight of evidence.
To characterize risks from a potential diesel spill, we weighed four lines of evidence based on different
exposure estimates and sources of exposure-response relationships. The first two lines of evidence
relate modeled estimates of dissolved hydrocarbon concentrations to laboratory test results for
dissolved fractions of diesel oil and to state water quality standards. Because the diesel pipeline failure
scenarios are sufficient to saturate the two potential receiving streams, we assume dissolved
concentrations equal the solubilities of the Alaskan diesels (1.9 and 7.8 mg/L). Estimated concentrations
in the Iliamna River are a little lower (1.7 and 7.2 mg/L) due to limited concentrations of soluble
chemicals in diesel. These exposure levels are similar to the two median lethal concentration (LCso)
toxicity values for rainbow trout (2.43 and 8 mg/L) (Table 11-8) and far higher than the state standard
(0.015 mg/L). Based on these estimates of soluble hydrocarbon concentrations, invertebrate kills would
be highly likely and some salmonid mortality would be expected in the diesel pipeline failure scenario at
either location.
The next line of evidence relates exposure (expressed as the amount of oil added to the stream) to
laboratory test results for diesel dispersed in water. This line of evidence is based on the assumption
that diesel added to a flowing stream is equivalent to diesel added to water and stirred. Exposure levels
within the receiving water would be 17,000 mg/L for Chinkelyes Creek, 1,500 mg/L for the Iliamna
River, and 6, 500 mg/L for Knutson Creek (Table 11-7). The laboratory LCso tests for diesel dispersions
are shown in Table 11-8, and strongly suggest that an oil spill would result in acute lethality of fish and
invertebrates, even if turbulent mixing in a stream is not as efficient as stirring. In addition, tests of the
alga Selenastrum capricornutum found that multiple growth and production endpoints were reduced by
50% at 20 mg/L (API 2003), which is also well below the estimated exposure.
The published history of freshwater diesel spills provides the final line of evidence. Diesel spill volumes
at the two locations considered in these diesel pipeline failure scenarios—30 m3 at Chinkelyes Creek
and 12 m3 at Knutson Creek—fall within the range of the cases described in Table 11-9 that caused
effects on stream and river biotic communities. In addition, the sizes of the receiving streams in these
Bristol Bay Assessment 1128 January 2014
-------
Chapter 11 Pipeline Failures
failure scenarios and those in the case studies are similar. If we calculate a crude index of exposure by
dividing the amount of diesel spilled by streamflow, values for the two scenarios (17 and 3.5) fall in the
middle of the range of cases (0.26 to 560).
Only the case of a very small spill (less than 500 L into Hemlock Creek, NY) caused no significant
biological effects. Other diesel spills caused fish and invertebrate kills and reduced invertebrate
abundance and diversity. Invertebrate community effects persisted for several months to more than 3
years. Exposures and effects may be more persistent in Alaska's cold climate, but the only Alaskan study
did not monitor recovery. Based on past diesel spills in streams, the diesel spills evaluated in this
assessment—and any other spill that released more than a trivial amount of diesel to a stream—would
be expected to cause an immediate loss of fish and invertebrates. The community would be likely to
recover within 3 years, but the time to recovery in Bristol Bay streams is uncertain.
11.5.4.1 Weighing and Summarizing the Lines of Evidence
The diesel pipeline failure probability used in this assessment is based on one line of evidence, the
record of actual oil pipelines. However, the predicted effects of a diesel spill are based on four lines of
evidence. All lines of evidence lead to the conclusion that a diesel spill into a stream would result in an
invertebrate and fish kill and reductions in abundance and diversity (Table 11-10). In the diesel pipeline
failure scenarios evaluated here, the lengths of affected stream would be roughly 22 km (Chinkelyes
Creek and the Iliamna River) or 2.6 km (Knutson Creek). Because these distances are short relative to oil
degradation rates, effects would be likely to extend to Iliamna Lake. Effects in the lake are not estimated
here, but are unlikely to extend far beyond the area of input due to dilution. In Knutson Creek, however,
flow to Knutson Bay could result in mortality of congregated spawning salmon, their eggs, and other fish
attracted by salmon eggs as a food source (Appendices A and B). Based on the monitoring of diesel spills
in streams, effects on stream communities would be likely to persist for one to several years. Although
each line of evidence has associated uncertainties and weaknesses (Section 11.5.5), they all support
these general conclusions.
The weighing of these lines of evidence is summarized in Table 11-10, using the same methods
described in Section 11.3.4.4. Overall, available lines of evidence for effects of a diesel spill are
supportive of the hypothesis that acute toxic effects would occur following a diesel pipeline failure
(Table 11-10). The quality of the exposure-response information is good (+) for all routes of exposure
based on reported observations in case studies, because the information is realistic; the quality of
information is considered good (+) for exposure via dissolved and hydrocarbons based on laboratory
acute tests, because the information reflects multiple tests. The quality of the exposure information for
the dissolved and dispersed hydrocarbons is considered ambiguous (0) because of the uncertain
relationship between the laboratory preparations and modeled stream exposures. The quality of the
exposure-response information is considered very good (++), because it is based on the Alaska water
quality standard, an official standard. The analogous spills, as a whole, are considered very strong (++)
evidence that a diesel spill would cause toxic effects in streams.
Bristol Bay Assessment 1129 January 2014
-------
Chapter 11
Pipeline Failures
Table 11-10. Summary of evidence concerning risks to fish from a diesel spill. The risk
characterization is based on weighing four lines of evidence for different routes of exposure. All
evidence is qualitatively weighed (using one or more +, 0, -symbols) on three attributes: logical
implication, strength, and quality. Here, all lines of evidence have the same logical implication—that
is, all suggest a diesel spill would have adverse effects. Strength refers to how strongly the line of
evidence indicates effects, and quality refers to the quality of the evidence sources (i.e., data quality
and relevance to the diesel pipeline failure scenario).
Route of Exposure
Source of Evidence (Exposure/E-R)
Dissolved hydrocarbons
Model/laboratory acute tests
Dissolved hydrocarbons
Model/laboratory-based standard
Dispersed hydrocarbons
Diesel-to-water ratio/laboratory acute
tests
All routes in actual spills
Amount spilled/observed effects
Summary Weight of Evidence
Logical
Implication
+
+
+
+
+
Strength
+
++
++
++
++
Quality
Exposure
0
0
0
+
0
E-R
+
++
+
+
+
Result
Modeled dissolved diesel
concentrations are clearly lethal to
invertebrates and approximately
lethal to trout.
Modeled dissolved diesel
concentrations greatly exceed State
standard.
Diesel oil/water ratios in the spills
and in tests suggest lethality to
invertebrates and trout.
Diesel spills in other streams cause
acute biological effects.
The effects by four lines of evidence
are consistent and the observed
effects are strong. The greatest
uncertainty is the relation of
laboratory to field exposures.
Notes:
E-R = exposure-response relationship.
The specific effects of a diesel spill on salmonid populations would depend on the individual receiving
waters. Streams along the transportation corridor that could receive a spill are described in Section 10.1.
Chinkelyes Creek receives on average roughly 9,000 spawning sockeye salmon and flows to the Iliamna
River, which receives on average more than 100,000 sockeye spawners (Table 10-2). Knutson Creek
receives 1,500 sockeye spawners and flows to Knutson Bay, which receives an average of 73,000 beach
spawning sockeye (Table 10-2). Not all of those salmon spawn below the stream crossing, but these
values indicate that a non-trivial number of spawners and their potential production are at risk. In these
scenarios, a spill would likely disrupt spawning if it occurred during the spawning season and would
potentially kill adults. In other seasons, it would likely kill fry, and would certainly kill invertebrates on
which salmon fry and all stages of other salmonids depend.
11.5.4.2 Duration of Risks
Diesel and natural gas pipelines would be retained after mine closure as long as fuel was needed at the
mine site (e.g., for monitoring, water treatment, and site maintenance). Therefore, the diesel pipeline
risks would continue indefinitely.
11.5.4.3 Remediation
Remediation of freshwater oil spills is discussed in detail in a review by the National Oceanic and
Atmospheric Administration (NOAA) and American Petroleum Institute (API) (1994). For diesel spills in
Bristol Bay Assessment
11-30
January 2014
-------
Chapter 11 Pipeline Failures
small rivers and streams, remediation via booms, skimming, vacuum, berms, and sorbents results in the
least environmental impact. Diesel is difficult to remediate by conventional techniques because its
components seep into soil, dissolve in water, or evaporate relatively quickly, making it is less
containable than typical crude oil. Also, booms, although useful, are imperfect tools for containing
floating oil. Booms were deployed after the diesel spill in Cayuga Inlet (Table 11-9), but within 24 hours
a slick was reported on Cayuga Lake, 16 km downstream (Lytle and Peckarsky 2001). Even when
recovery of diesel fuel was rapid and approximately 70% effective, as in the Westlea Brook spill (Table
11-9), the rapidly dissolved component was sufficient to cause severe acute effects (Smith et al. 2010).
There has been relatively little study on remediation of oil spills in freshwater wetlands. For diesel, the
NOAA and API (1994) review recommends natural recovery, sorbents, flooding, and low-pressure cold-
water flushing as least adverse options. Wetlands also have been remediated by burning, which can
remove floating oil and destroy oiled vegetation that is likely to die from effects of the oil. Burning can
cause severe but localized and short-term air pollution and, if improperly controlled, can result in fires
that spread beyond the oiled area. However, burning does not destroy the dissolved fraction, which
would move to streams or the lake and is primarily responsible for aquatic toxicity.
Cold winter weather complicates remediation of diesel spills (NOAA and API 1994). Spills into water at
temperatures below the oil's pour point can result in the formation of viscous tar-like particles that are
difficult to recover. Ideally, a spill onto ice could congeal on the surface where it might be relatively
easily recovered if action is prompt; however, diesel oil can penetrate ice, and solar absorption by the oil
can result in freeze-thaw cycles that create a complex material. Spills that flow under ice deposit on the
undersurface. Standard procedures for oil remediation do not address those conditions.
11.5.5 Uncertainties
Based on weighing multiple lines of evidence, it is certain that a diesel pipeline spill into a stream would
cause acute toxic effects. However, the following uncertainties apply to individual pieces of evidence.
• The composition of diesel oil is highly variable. As a result, the fate and toxicity of diesel spills are
inherently uncertain unless the specific source is known and analyzed; the source does not change
over time; and any physical, chemical, and biological tests are performed with that specific oil. This
uncertainty cannot be resolved without case-specific studies of a sort that are not normally
performed. This and other uncertainties concerning test results could cause errors of at least one
order of magnitude in the risks estimated from laboratory toxicology.
• Measurement of petroleum hydrocarbons in water is performed using a variety of methods. Because
the results of hydrocarbon analysis are method-specific, significant uncertainty can be introduced
when these results are compared to benchmarks generated using different analytical methods. This
contributes to the overall uncertainty of toxicity test results.
• Invertebrate and fish losses are likely if a diesel spill occurs at a stream, but the magnitude and
nature of these losses would be highly uncertain. Some mortality would occur for some species, but
the species and number of organisms affected cannot be specified. This uncertainty would take a
Bristol Bay Assessment 1131 January 2014
-------
Chapter 11 Pipeline Failures
major case-specific research program to resolve. The inability to exactly define the expected
ecological effects occurs in all risk assessments, but is worse for the diesel spill than for other
contaminants such as copper.
• The ability of the laboratory toxicity tests to predict responses to diesel in the field is highly
uncertain due to the variety of preparation methods, the simplicity of laboratory exposures relative
to the complexity of oil spills in streams, and the lack of field validation studies.
• Variation in sensitivity to diesel among species appears to be high relative to other aquatic
pollutants. Remarkably, even in the same test series, different salmon species range in sensitivity
over two orders of magnitude (Table 11-8).
• Spills into wetlands are likely to have severe and persistent effects due to low rates of flow, but no
relevant studies of diesel spills in freshwater wetlands are available to confirm even that very
general hypothesis.
• The applicability of previous diesel spills considered in Table 11-9 to streams in the Bristol Bay
region is uncertain, given that all of the spills occurred elsewhere. However, the effects observed in
the one Alaskan case are not dissimilar from those in temperate regions. The most likely differences
are slower loss of oil and longer recovery times. Therefore, effects are likely to be more severe in
Alaska than in the temperate cases.
• The principle uncertainty in this analysis is the number and location of spills into aquatic
ecosystems, given the probability of a pipeline failure. We can say with some certainty that a diesel
spill of a non-trivial volume into a stream would have adverse ecological effects. We can also say that
a spill is likely, based on the record of oil pipelines in general and large recent spills from oil
pipelines (e.g., into the Kalamazoo River, as described in Section 11.1). However, we cannot predict
with any certainty where such as spill may occur.
• Although the diesel spill cases suggest that streams are likely to recover within 3 years, time to
recovery is seldom reported. Where it has been reported, it apparently depends on the conditions
and the recovery metric used, and ranges from a year to several years.
Bristol Bay Assessment 1132 January 2014
-------
Large-scale mining, as described in the mine scenarios (Table 6-1), could have both direct and indirect
effects on wildlife and Alaska Native cultures (Figures 12-1 and 12-2). In this chapter, we primarily
consider indirect effects, focusing on how wildlife and Alaska Native cultures may be affected by any
mining-associated changes in salmon resources. Direct effects on these endpoints—defined here as
effects that are independent of impacts on fish populations—could be significant, and would need to be
fully evaluated as part of a comprehensive environmental impact statement for any proposed future
development. However, these direct effects are generally considered outside the scope of the current
assessment (Chapter 2) and are only mentioned briefly here (Box 12-1). Potential cumulative effects
that multiple mines in the region may have on wildlife and Alaska Native cultures are discussed in
Chapter 13.
12.1 Effects on Wildlife
As discussed in Chapters 7 through 11, a large-scale mine and its associated transportation corridor
would likely affect the abundance, productivity, and diversity of Pacific salmon. These changes in salmon
resources could stem from direct habitat losses and downstream flow alterations resulting from the
mine footprint, or from changes in the physical and chemical habitat characteristics resulting from mine
operations and potential accidents or failures. Wildlife species in the Nushagak and Kvichak River
watersheds that depend on salmon could be affected by decreases in salmon abundance. Interactions
between salmon and other fish and wildlife, and the potential for disruption of these interactions, are
complex (Section 5.2.5). In this section, we qualitatively consider how a decrease in salmon abundance
may affect wildlife—that is, salmon-mediated effects on wildlife—via the loss of salmon as a food source
and the loss of marine-derived nutrients (MDN) as a source of productivity.
Bristol Bay Assessment
12-1
January 2014
-------
Chapter 12 Fish-Mediated Effects
BOX 12-1. POTENTIAL DIRECT EFFECTS OF MINING
The salmon-mediated effects considered in this assessment represent only one component of potential
large-scale mining impacts on wildlife and Alaska Natives. Both wildlife and Alaska Natives would likely
experience direct impacts, the magnitude and extent of which could be significant. For example, direct
impacts on wildlife would include loss of terrestrial and aquatic habitat, reduced habitat effectiveness (e.g.,
in otherwise suitable habitats adjacent to mine area), habitat fragmentation, increased stress and
avoidance due to noise pollution, and increased conditioning on human food (Figure 12-1).
Direct effects of large-scale mining on Alaska Native populations could result from multiple stressors,
including noise pollution, air emissions, changes to water supply and quality, an influx of new residents, and
induced development. Mine construction and operation also would have direct economic and social effects,
both positive and negative, on Alaska Native cultures. For example, an influx of new residents in response to
mine development could decrease the local population percentage of Alaska Natives and have a
corresponding effect on local culture. A shift from part-time to full-time wage employment in mining or mine-
associated jobs would provide additional employment opportunities and income, but would affect
subsistence-gathering capabilities by reducing the time available to harvest and process subsistence
resources.
At this time, it is difficult to determine what, if any, effects routine operations at the Pebble deposit would
have on drinking water sources in the Nushagak and Kvichak River watersheds. Private wells are a primary
drinking water source for many residents of the Nushagak and Kvichak River watersheds, and communities
also rely on groundwater for their public water supply. The extent to which surface water influences the
quality or quantity of the groundwater source for these wells is unknown. There are also communities in the
area that rely on surface water sources, which may be more susceptible to mine-related contamination.
Although a thorough evaluation of potential direct effects of large-scale mining on wildlife and Alaska Native
populations is beyond the scope of this assessment, these examples illustrate just a few of the complex
ways in which wildlife and Alaska Natives could be affected by large-scale mine development.
Lower salmon production would likely reduce the abundance and production of wildlife in the mine
area and presumably in the range areas of the affected species, but the magnitude of those effects cannot
be quantified. The Bristol Bay region is home to a complex foodweb that includes salmon and salmon
predators and scavengers (Box 5-3, Figure 12-1). Annual salmon runs provide food for brown bears,
wolves, bald eagles, other land birds, and water birds, and it is likely that these species would be directly
affected by a reduction in salmon abundance. Waterfowl prey on salmon eggs, parr, and smolts and
scavenge on carcasses. Salmon carcasses are an important food source for bald eagles, water birds, other
land birds, other freshwater fish, and other terrestrial mammals. Aquatic invertebrate larvae also
benefit from carcasses and are an important food source for water birds and land birds.
Bristol Bay Assessment 122 January 2014
-------
Chapter 12
Fish-Mediated Effects
Figure 12-1. Conceptual model illustrating potential effects on wildlife resulting from effects on salmon.
. ...
A~ ~~\ /~~ "A / ^ '' N /
1 transportation mine \ ' power generation & ' i otherancillary pOrt j
k corridor J I j transmission facilities ) l facilities . l /
\ / \ / S ^ . ^ f
f
1 1
i i iii
V V V v V V V
1 ^road-related ' ^migration ] ' ^conditioning ' ' ^terrestrial ] ' ^ noise i ' si/ air i ^environmental mpacts
, mortality , barriers i , on human food , , habitat loss i , Slight ' , quality [ on freshwater habitats
---- f ------ -f ---- - - -1 -___ -J ____.._ ^_J__ __ __J_
II 1 III
II 1 III
1_ U 1 _l i J
/ t" direct effects
xx^ on wildlife
LEGEND
/ S, proximate
^ source ) stressor
additional step in / biotic N
can sal path way V^resporvse^X
Within a shape, "\~ indicates an increase in
the parameter, -|- indicates a decrease in a
parameter, and A indicates a change in the
parameter.
Arrows leading from one shape to another
indicate a hypothesized cause-effect
relationship.
Shapes bracketed under another shape are
specific components of the more general
shape under which they appear .
Dashed arrows and outlines indicate topics
outside the scope of assessment.
)
[7
f si/ adults
1 ^^^ — __— --
\/
^
\
V
.^ — "" si/ salmon ~^^v
V^_^ (abundance, productivity or diversity) ^^
i
) (T si/ carcasses ]) C si/ eggs ) (^ si/ juveniles ^)
^c
1
't by wildlife
/ V
si/ terrestrial distribution of si/ aquatic distribution of
marine-derived nutrients marine-derived nutrients
V
[*>
si/ terrestrial
vegetation
j
V
r si/ ca ri bouj)
V
(T si/ brown bears^) v
(si/ W
V
Is— \ j|f \f
si/ riparian
vegetation -i- aquatic food resources
V
r sb moose J)
V \/ V
/• — "N /^r c^\ / — r ~~"~^\
C si/ bald eagles ) « , / si/ other \ f si/ other N
' /"*-» — , , ^-^ >' ( innrl birds J \/ vfre^iwaterfishev \/
/olves ) t si/ water birds J Xi • L- < x ' ' s^^ i -..h-.^ ^^v
^-•^f ^^-j_ ^ A si/ marine fishes >> A -J/ other ^\
V & mammals J ^terrestrial mammals/'
Bristol Bay Assessment
12-3
January 2014
-------
Chapter 12
Fish-Mediated Effects
4/ other fishes
(abundance, productivity or diversity)
A
salmon
(abundance, productivity or diversity)
marine-derived
nutrients
ecosystem
productivity
additional step in
human
response
4- food resources
Within a shape, T indicates an increase in the
parameter, ,|- indicates a decrease in a
parameter, and A indicates a change in the
parameter.
Arrows leading from one shape to another
indicate a hypothesized cause-effect
relationship.
wildlife
(abundance, productivity or diversity)
Shapes bracketed under another shape are
specific components of the more general
shape under which they appear.
xt-commercia
fishing
4-subsistence
equipment
village subsistence
(quality or quantity)
J/ social relations
4/ family activities&
traditional ways of teachi
•I- reciprocal
Food sharing
4, orthodoxy ) (^ animistic
beliefs
ralienation&
purposelessness
sense of place
& identity
4, physical &
mental health
village population
A aska Native cuture
Bristol Bay Assessment
12-4
January 2014
-------
Chapter 12 Fish-Mediated Effects
Salmon predators and scavengers then deposit MDN on the landscape, as either carcasses or excreta.
These nutrients contribute to the plant production that supports caribou, birds, and other terrestrial
wildlife. Caribou are, in turn, prey species for wolves and brown bears. The link between increased
vegetation and MDN distributed by brown bears has been documented (Hilderbrand et al. 1999, Helfield
and Naiman 2006), but additional research is needed to confirm and quantify the links between moose,
caribou, and MDN.
Factors such as the magnitude, seasonally, duration, and location of salmon losses would influence the
specific wildlife species affected and the magnitude of effects. Generally, the loss of salmon as food
resources in any area of the mine scenario watersheds would be expected to create displacement or loss
of wildlife species dependent on those food resources. If the loss were of a sufficient magnitude and
duration, there may be additional indirect effects, such as loss of vegetation from lack of MDN and
consequent loss of food resources for species such as moose and caribou. Should riparian vegetation be
reduced by long-term loss of MDN, there would be decreased food resources for moose, particularly in
the Nushagak and Mulchatna River systems, which have large riparian zones (Brna and Verbrugge
2013).
Seasonally of salmon resources is also important for wildlife species. Brown bears, wolves, bald eagles,
and other species depend on salmon for a large fraction of their summer diet. Mine failures that reduced
or eliminated a salmon substock would be expected to reduce or displace wildlife species that depend
on those particular salmon nutrients during important life-history periods such as breeding, nesting,
and pre-winter feeding.
Alaska Natives have expressed concerns that wildlife may be affected by consuming contaminated fish.
Two potential contaminants of concern are copper and selenium. The primary aquatic contaminant of
concern from a porphyry copper mine is copper, which can cause both acute and chronic toxicity to
salmon and other fishes (Chapter 8). However, copper is relatively weakly accumulated by fish in both
aqueous and dietary exposures and does not bioaccumulate. In fact, in the Clark Fork River, copper
concentrations were lower in fish than in invertebrates, and lower in invertebrates than in periphyton
(ARCO 1998).
Data for copper toxicity to wildlife are not available, because direct toxicity has not been a problem—the
indirect effects of reduced aquatic prey are likely to be greater than direct toxic effects. However, the
dietary maximum for pigs and poultry of 200 mg/kg dry weight can be used as a surrogate benchmark
(Eisler 2000). If we use the highest reported fish bioconcentration factor (290 for fathead minnows,
from an unpublished manuscript cited in USEPA 1985) as a conservative value, we obtain a safe water
concentration of 690 ug/L. This safe level for wildlife is much higher than both the toxic levels for
aquatic biota (Section 8.2.2.1) and the estimated instream concentrations for a wastewater treatment
plant failure (Table 8-20). It is a little higher than the estimated concentration in the concentrate
transport water and return water (655 ug/L, Section 11.3.2), but is not a concern because of dilution and
the short duration of exposures to spills. The copper concentration in the product concentrate leachate
is 12 times that value. However, a product spill would be localized at the mine site or along the
Bristol Bay Assessment 125 January 2014
-------
Chapter 12 Fish-Mediated Effects
transportation corridor. Most important, if copper concentrations were so high that the leachate was not
diluted by a factor of 12, there would not be a sufficient aquatic community remaining after a spill to
attract species such as mink, river otters, or belted kingfishers, which may forage in a particular stream
or river for fishes such as trout, char, or salmon fry. Hence, copper toxicity to fish-eating wildlife is
improbable.
Selenium is a well-characterized avian toxicant that has been a concern for waste rock leachate at other
mines (USEPA 2011). It does biomagnify and the primary route of exposure for fish is diet (Chapman et
al. 2010). However, selenium bioaccumulation depends on biogeochemical conditions that occur in
slowly flowing (lentic) ecosystems such as ponds and wetlands, and is much less prevalent in the
streams that are the likely receptors for effluents with elevated selenium levels. In addition, the
selenium concentrations in wastes identified in the mine scenarios would be relatively low. Mean
selenium (the appropriate measure for a biomagnifying chemical) expected in waste rock leachates is
below the water quality criterion, and leachates from tailings and product concentrate are only 1.5 times
the criterion concentration (Section 8.2.2.2). Hence, minimal dilution would bring these concentrations
down to safe levels. Although fish and birds are sensitive to selenium, mammals are not. The body
burdens of adult salmon are almost entirely due to marine exposures; because salmonid eggs take up
contaminants relatively slowly, body burdens of eggs and larvae would also be expected to reflect
marine sources. Local sources of selenium are therefore not relevant, and bioaccumulation of selenium
would only be a concern for wildlife consuming resident fish or older salmon fry. Fish-eating birds
feeding on resident fish also may forage in more than one stream or other water body, potentially
providing further dilution. For these reasons, aqueous selenium is unlikely to pose a risk to wildlife via
fish consumption.
12.2 Effects on Alaska Natives
As discussed in Chapters 7 through 11, routine development and operation of a large-scale mine, as well
as potential mine accidents or failures, would likely affect salmon resources in the Nushagak and
Kvichak River watersheds. The importance of salmon to Alaska Native cultures is well documented
(Section 5.4, Appendix D). Because these cultures are so intimately related to the local landscape and the
resources it provides, any changes to salmon or other subsistence resources would likely result in
changes to the cultures. The magnitude of these changes could be assumed to be dependent on the
magnitude and duration of both the loss of subsistence resources and the disruption to the landscape
itself. Changes in salmon resources may affect indigenous health, welfare, and cultural stability in
several ways (Appendix D).
• Because the traditional diet is heavily dependent on wild foods, particularly salmon, diets would
move from highly nutritious wild foods to increased reliance on purchased processed foods.
• Social networks are highly dependent on procuring and sharing salmon and wild food resources, so
the current social support system would be degraded.
Bristol Bay Assessment 126 January 2014
-------
Chapter 12 Fish-Mediated Effects
• The transmission of cultural values, language learning, and family cohesion would be affected
because meaningful family-based work takes place in fish camps or similar settings for traditional
ways of life.
• Values and belief systems are represented by interaction with the natural world through salmon
practices, clean water practices, and symbolic rituals. Thus, core beliefs would be challenged by a
loss of salmon resources, potentially resulting in a breakdown of cultural values, mental health
degradation, and behavioral disorders.
• The region exhibits a high degree of cultural uniformity tied to shared traditional and customary
practices, so significant change could provoke increased tension and discord both between villages
and among villagers.
Human health and cultural effects related to potential decreases in salmon resources would depend on
the magnitude of these reductions. A small reduction in salmon quality or quantity may not have
significant impacts on subsistence food resources, human health, or cultural and social organization, but
a significant reduction in salmon quality or quantity would certainly have significant negative impacts
on these salmon-based cultures.
Salmon-mediated effects from potential accidents and failures associated with large-scale mining would
likely have much greater effects on human welfare and Alaska Native cultures than the effects from
routine operations. It should be assumed that any negative impact on salmon quantity or quality
resulting from mine failures or accidents would affect human health and welfare, both from loss of or
change in food resources and from cultural disruption. Because all aspects of Alaska Native cultures in
the Nushagak and Kvichak River watersheds are closely tied to salmon and other fishes, cultural
vulnerability to long-term environmental disruption is very high (Appendix D). A major failure or
accident that resulted in long-term disruption of salmon habitat and ongoing toxicity to salmon or their
food would significantly affect both subsistence resources and cultural identity. Potential causes of
salmon-mediated effects on Alaska Native cultures would differ across the two watersheds. For example,
villages near the transportation corridor could be negatively affected by road and culvert failures
(Chapter 10) or pipeline spills (Chapter 11). Villages downstream of the mine would be more affected by
any water collection, treatment, and discharge failures (Chapter 8), and impacts from these failures
would likely be much greater than impacts from routine operations.
This assessment focuses on potential effects on Alaska Native cultures, but other groups would also be
particularly vulnerable to mining-associated impacts on salmon. Many of the non-Alaska Natives that
reside in the area practice a subsistence way of life and have strong cultural ties to the landscape that go
back generations (Box 12-2). Many seasonal commercial anglers and cannery workers also depend on
these resources and have strong, multi-generational cultural connections to the region.
In this section, we discuss the range of potential salmon-mediated effects on Alaska Native cultures from
large-scale mining in the Nushagak and Kvichak River watersheds. We also reference key impacts that
other mining, oil, and gas development activities in Alaska—including northwest Alaska's Red Dog Mine
Bristol Bay Assessment 127 January 2014
-------
Chapter 12 Fish-Mediated Effects
and oil and gas development on Alaska's North Slope—have had on Alaska Native cultures, especially in
terms of losses of or changes to subsistence resources. Although not directly applicable to large-scale
mining, information about oil and gas extraction activities provides insight into potential effects of large-
scale mining on Alaska Native culture in the Nushagak and Kvichak River watersheds.
BOX 12-2. TESTIMONY ON POTENTIAL EFFECTS OF MINING ON ALASKA NATIVE CULTURES
The U.S. Environmental Protection Agency (USEPA) held a series of public meetings to collect input on the
May 2012 draft of this assessment. Many Alaska Natives, includingtribal Elders and other tribal leaders,
provided testimony on concerns about potential effects of large-scale mining in the Bristol Bay watershed,
as well as the desire for economic development. The following are selected quotes representative of this
testimony. To view the full public meeting transcripts, visitwww.epa.gov/bristolbay.
• "Salmon has been part of our Native spiritual food; and without the food and waters we will die slowly,
we'll be here existing, but our spirit will begone."
• "I urge you to pay especially close attention to the voices of our Elders across Bristol Bay. They have
instilled in them the deepest of our roots, and our God given way of life; our culture that has been slowly
fading away. It is the adaptation to modern civilization that we have embraced so far that is causing our
cultures to become lost."
• "We support the science of this document as it in turn supports what the Elders of this area and their
traditional knowledge have said all along. We are preparing our boats, we are mending our nets, we are
cleaning our smokehouses and we are sharpening our knives. However, by testifying at these meetings
and missing out on one day or maybe several days of preparation for those who attended multiple
meetings here in the region, we hope that this will prevent, with the help of the EPA, missing out on a
lifetime of salmon and missing out on a way of life that we have treasured for thousands of years."
• "Bristol Bay is much different. Everyone who lives here has a deep and strong sense of place. There is a
powerful connection to the lands and waters and resources of Bristol Bay. It is a connection that starts
before birth. It is genetic. It is handed down through the generations and it is also learned from a very
young age. A connection told in stories from parents and Elders and experienced firsthand. Toddlers
accompany parents and grandparents fishing, hunting, berry picking. They participate at home to store
that food and save it. It is part of the family experiencing for anyone who grows up in Bristol Bay, and as
a result, this land its water and its resources become a part of who you are. This is a connection without
a price tag and it cannot be replaced. If it lost, it is lost forever."
• "My family does subsistence. We likeourfish. But still, nobody is going to give my boys jobs. Nobody is
going to pay my bills. I'm not for or against. I want clean water, but we need jobs around here. Who is
going to pay for my bills?"
• "I work for Pebble. I have a big family who loves the outdoors and enjoy their subsistence way of life.
Subsistence is good, but it is not paying for my bills and does not clothe my kids."
12.2.1 Subsistence Use
As discussed in Chapter 5, subsistence foods make up a substantial proportion of the human diet in the
Nushagak and Kvichak River watersheds and likely contribute a disproportionately high amount of
protein and certain nutrients. The percentage of salmon harvest in relation to all subsistence resources
ranges from 29 to 82% in the villages (Appendix D).
The mine scenario footprints would have some effects on subsistence resources. Although no
subsistence salmon fisheries are documented directly in any of the mine scenario footprints, other fish
are harvested in these locations, and the areas are identified as being important for the health and
abundance of subsistence resources (PLP 2011: Chapter 23). Negative impacts on downstream fisheries
Bristol Bay Assessment 128 January 2014
-------
Chapter 12 Fish-Mediated Effects
from headwater disturbance (Section 7.2) could affect subsistence salmon resources beyond the mine
footprints. Those residents using the mine area and immediate areas downstream of the mine pit and
tailings storage facilities (TSFs) for subsistence harvests would be most affected (Figure 5-2). Access to
subsistence resources is also important. A reduction in downstream seasonal water levels caused by
mine-related withdrawals during and after mine operation could pose obstacles for subsistence users
who are dependent on water for transportation to fishing, hunting, or gathering areas.
There could also be effects from the footprint of the transportation corridor. A review of Alaska
Department of Fish and Game data (Appendix D: Table 13) indicates that some residents use the area
along the transportation corridor considered in the assessment for subsistence salmon harvest. Of the
villages in these watersheds, reliance on salmon is highest in Pedro Bay, where salmon provides 82% of
per-capita subsistence harvest. The estimated annual per-capita subsistence harvest for Pedro Bay, the
village closest to the transportation corridor considered in this assessment, was 306 pounds in 2004.
Thus, this village is particularly vulnerable to losses of salmon resources.
The effects of the transportation corridor on subsistence resources would be complex and
unpredictable. Based on the analysis in Chapter 10, we anticipate that routine transportation operations
would have some negative effects on salmon habitat in streams along and downstream from the
transportation corridor. Some subsistence users in these areas could be affected. The corridor also
would increase accessibility of the area, which could increase subsistence use of nearby streams but also
create greater competition for resources.
The initial effect of a mine accident or failure on Alaska Native cultures would be the loss or decrease of
subsistence salmon resources downstream. It is not possible to quantify the magnitude of subsistence
resources that would be lost, nor is it possible to evaluate the geographic extent of disruption to
subsistence resources. However, this assessment provides examples of the potential magnitude of
salmon impacts from failures. One such example is the potential effect of a tailings dam failure on
Chinook salmon in the Nushagak River. As described in Chapter 9, a tailings dam failure at TSF 1 could
significantly affect Koktuli River Chinook runs, which constitute up to 29% of the larger Nushagak River
Chinook runs. Stuyahok River and Mulchatna River Chinook runs, which constitute up to 17 and 10% of
the Nushagak River Chinook runs, respectively, could also be affected. The Alaska Native villages on the
Nushagak River (Koliganek, New Stuyahok, Ekwok, and Dillingham) (Figure 2-4) are culturally and
nutritionally dependent on Chinook salmon. Thus, a tailings dam failure would have negative and
potentially significant effects on the ability of subsistence users to harvest salmon downstream of the
mine area.
It is not possible to predict the magnitude of effects from the loss of salmon as a subsistence food, nor is
it possible to predict what level of subsistence resource loss would be necessary to overcome the
adaptive capacity of these cultures. On a physical level, the loss of salmon as a highly nutritious wild food
and the consequent substitution of purchased foods would have negative effects on individual and
public health (Appendix D). Salmon is especially valued around the world for nutrition and disease
prevention. Dietary transition away from subsistence foods in rural Alaska carries a high risk of
Bristol Bay Assessment 129 January 2014
-------
Chapter 12 Fish-Mediated Effects
increased consumption of processed simple carbohydrates and saturated fats. This has occurred in
urban communities that have low availability and high cost of fresh produce, fruits, and whole grains
(Kuhnlein et al. 2001, Bersamin et al. 2006). Also, alternative food sources may not be economically
viable and are certainly not as healthy. Compounding the shift to a less healthy diet, the physical benefits
of engaging in a subsistence lifestyle would be reduced (Appendix D).
In addition to the salmon-mediated effects of large-scale mining considered here, there could be effects
from the loss of non-salmon subsistence resources, such as land mammals, birds, and other fishes.
Subsistence use of the mine area is high and centers on hunting caribou and moose and trapping small
mammals (Braund and Associates 2011 in PLP 2011). Because no subsistence salmon fisheries are
documented in the mine scenario footprints, direct loss of non-salmon subsistence food resources likely
would represent a greater direct effect than loss of salmon harvest areas in the mine footprints. Tribal
Elders have expressed concerns about ongoing mine exploration activities directly affecting wildlife
resources, especially the caribou herd range (Appendix D).
Experience with existing development in Alaska supports the contention that development of a large-
scale mine operation would directly affect subsistence resources within and around the mine scenario
footprints during routine operations and in perpetuity, from both loss of habitat and disturbance related
to routine operations. For example, the supplemental environmental impact statement for the Red Dog
Mine (USEPA 2009) documented multiple subsistence impacts, including reduced harvest of beluga by
Kivalina harvesters, likely related to port activities. Related to transportation corridors, traffic along the
Delong Mountain Regional Transportation System road was found to cause "limited, localized" effects on
caribou movement and distribution, and nine caribou fatalities occurred because of traffic collisions.
Kivalina harvesters and harvest data also indicated that traffic along the road has likely resulted in
fewer caribou harvested by Kivalina harvesters than would otherwise be the case.
A study of the cumulative environmental effects of oil and gas activities on Alaska's North Slope included
a summary of hearings held with North Slope residents, who are predominantly Alaska Natives.
Community members provided testimony on both positive and negative effects of these activities. North
Slope residents recognize that oil production in the region has brought benefits such as money to spend
on community facilities, schools, modern water and sewer systems, village clinics, child emergency
shelters, and behavioral outpatient and residential programs that provide mental health care and
counseling for substance abuse and domestic violence. However, they also reported that traditional
subsistence hunting areas have been reduced, the behavior and migratory patterns of key subsistence
species have changed, and there is increased incidence of cancer and diabetes and disruption of
traditional social systems.
Residents also reported experiencing significant increases in the time, effort, and funding necessary to
respond politically and administratively to the increased number of projects proposed in their
communities (NRC 2003). The stress of integrating a new way of life with generations of traditional
teachings and the associated impacts of rapid modernization and loss of tradition is known as
acculturative stress. This stress has been linked to a wide variety of health outcomes, ranging from
Bristol Bay Assessment 1210 January 2014
-------
Chapter 12 Fish-Mediated Effects
impaired mental health and social pathology (such as substance abuse, violence, and suicide) to
cardiovascular disease and diabetes. For the Inupiat on the North Slope of Alaska, the greatest defense
against acculturative stress is the continued practice of the bowhead whale hunt, which involves the
entire community (NMFS 2013).
Changes in diet and nutrition are common potential effects of oil and gas exploration and production
activities where populations rely on subsistence resources. These changes can lead to a number of
important public health outcomes. For example, a traditional diet has been shown to be strongly
protective against chronic diseases for indigenous populations. A shift away from subsistence diets is
associated with food insecurity, or the inability to secure sufficient healthy food for a family. Studies of
food insecurity and health have found a variety of detrimental health impacts, including obesity, poor
psychological function among children, poor cardiovascular health outcomes, and lower physical and
mental health ratings (NMFS 2013). The high cost of store-bought food, the costs associated with
harvesting of subsistence resources, and the year-to-year variation in subsistence resource availability
are all implicated in the high food insecurity rates experienced by many northern indigenous
populations.
Alaska Native residents also report subtle changes in species harvested by subsistence hunters,
including changes in color, texture, and taste of the flesh and skin of several subsistence species.
Transportation corridors associated with resource extraction activities can also increase competition for
local subsistence resources. For example, hunting by non-local residents along the Dalton highway has
been reported to have increased after the development (and later public opening) of the road (NRC
2003).
The experiences of subsistence users near Red Dog Mine and Alaska's North Slope indicate that localized
changes in resource movement can affect that resource's availability and predictability to subsistence
users, even when the overall pattern or abundance of the resource may not be affected by development
activities. From a biological standpoint, changes in caribou related to the Red Dog Mine may be viewed
as minimal. However, because residents rely on only a portion of the expansive range of the Western
Arctic caribou herd to harvest caribou, small and localized changes in caribou availability can have large
effects on subsistence uses. Subsistence users have observed changed or diverted migration routes,
reduced harvests of caribou, decreased size of caribou individuals and groups, and increased disease
and infection since mine operations began, and cite both mine-related and other causes (USEPA 2009).
The Exxon Valdez oil spill also resulted in reduced subsistence activities (Palinkas et al. 1993). These
reductions resulted from the closure of many areas to subsistence activities, local concerns over
subsistence food safety, voluntary abstinence from consumption after the spill, and reduced time for
subsistence activities by Alaska Natives who participated in cleanup efforts.
12.2.2 Perception of Food Security
Even a negligible reduction in salmon quantity or quality related to mining could decrease use of salmon
resources, based on the perception of subtle changes in the salmon resource. Interviews with tribal
Bristol Bay Assessment 1211 January 2014
-------
Chapter 12 Fish-Mediated Effects
Elders and culture bearers indicate that perceptions of subtle changes to salmon quality are important
to subsistence users, even if there are no measureable changes in the quality and quantity of salmon
(Appendix D). Aside from actual exposure to environmental contamination, the perception of exposure
to contamination is also linked to known health consequences, including stress and anxiety about the
safety of subsistence foods and avoidance of subsistence food sources (Joyce 2008, CEAA 2010, Loring et
al. 2010), with potential changes in nutrition-related diseases as a result. These health results arise
regardless of whether there is contamination at a level that could induce toxicological effects in
humans—rather, the effects are linked to the perception of contamination (NMFS 2011).
Literature on impacts from oil and gas development on Alaska's North Slope and ongoing operations at
Red Dog Mine demonstrates that even perceived contamination could have a real effect on subsistence
harvesters. In a recent survey, 44% of Inupiat village residents reported concern that fish and wildlife
may be unsafe to eat (Poppel et al. 2007, NMFS 2011). Residents of Kivalina and Noatak, the
communities closest to Red Dog Mine, also have expressed concerns about food safety, potential
contamination of subsistence resources, and corresponding changes in subsistence foraging (USEPA
2009). Kivalina residents are concerned about potential contamination of the Wulik River, which is used
both for subsistence and as the drinking water source for the village. These concerns persist even
though studies by the Alaska Department of Health and Social Services found that heavy metal
concentrations in drinking water were low and did not pose a risk (USEPA 2009).
12.2.3 Economic Impacts
Alaska Natives, as well as other local residents, participate in the salmon-based market economy,
primarily via commercial fishing and tourism. Subsistence harvests also represent a significant
economic value to local residents. A decrease in salmon that affected either of these sectors would be
particularly burdensome to local residents dependent on the commercial fishery for income and the
subsistence fishery for food. The necessity of purchasing expensive foods from outside the region, in
conjunction with more limited opportunities to obtain paid seasonal employment in the region, could be
extremely difficult for families. In many cases, income from commercial and recreational fishing
provides money to purchase equipment for subsistence fishing, so lost or reduced income from
commercial fishing would affect subsistence harvests even if subsistence fishing remains possible. For
those able to benefit economically from mining and induced development, there would be increased
cash resources to purchase equipment and supplies, resulting in more efficient subsistence activities.
However, increased full-time employment could decrease the time available for subsistence activities
and thus the social relationships based on these activities. Some residents have expressed a desire for
jobs and development related to large-scale mining and a market economy, whereas other residents
have expressed concerns that this type of economic shift would be detrimental to their culture
(Box 12-2, Appendix D).
Although large-scale mining would inject some market-based economic benefits for some period of time,
resource extraction experiences in other rural Alaska areas suggest it would likely have only modest
direct employment benefits in the local region (Goldsmith 2007). At the Red Dog Mine, ownership of the
Bristol Bay Assessment 1212 January 2014
-------
Chapter 12 Fish-Mediated Effects
resource empowered the NANA Regional Corporation, Inc. (NANA) to negotiate a development
agreement with strong protections and benefits to Northwest Alaska Natives (Storey and Hamilton
2004). NANA shareholders account for approximately 56% of the mine's 464 full-time employees and
91% of its 78 part-time employees. Although first preference in hiring and most of the training slots go
to shareholders, shareholders disproportionately occupy the mine's lower-skilled positions (Storey and
Hamilton 2004). Additionally, the supplemental environmental impact statement showed that
employment at the Red Dog Mine may have facilitated the relocation of community residents to
Anchorage for lifestyle or economic reasons (Storey and Hamilton 2004, USEPA 2009).
A disproportionately low number of Inupiat people are employed by the oil and gas industry on Alaska's
North Slope, although this may partially result from the large percentage of young people in the
population (NRC 2003). The Alaska Department of Labor reported that, of the 7,432 people who
reported working in the oil and gas sector on the North Slope in 1999 (and worked for companies that
collected and reported residency information), only 64 lived in the state's Northern Region (i.e., the
Nome, North Slope, and Northwest Arctic boroughs). A variety of factors affected both male Inupiat
willingness to work in the oil fields and the desire of companies in Prudhoe Bay to hire them (Kruse et
al. 1983, NRC 2003).
There may be decreased participation in a subsistence way of life for those benefiting from any
employment opportunities. The cash economy and the subsistence economy are intertwined, and
subsistence is a full-time job for those fully engaged in it. However, it is necessary to supplement
subsistence with cash from part-time wage labor or commercial fishing to defray the costs of
subsistence activities (Appendix D). Despite differences in the types of subsistence and traditional
cultural practices between the people of the Nushagak and Kvichak River watersheds and the people of
the North Slope, studies from the North Slope region can provide some insights. A study of Alaska's
North Slope Inupiat people found that there is an inverse relationship between active subsistence
harvesting and wage labor time for the individual worker, but that cash from employment is often used
for subsistence inputs (e.g., gasoline, boats, ammunition) (Kerkvliet and Nebesky 1997).
One of the mitigation measures that can address the impact of full-time employment on subsistence
activities is the implementation of subsistence leave policies. For the development of the Red Dog Mine
supplemental environmental impact statement, interviews were conducted that included questions
asking Noatak and Kivalina residents about their employment history related to the Red Dog Mine and
their employer's subsistence leave policies. Responses were mixed regarding whether or not
interviewees were aware of a subsistence leave policy and whether or not the policy worked. Some of
the companies did not have subsistence leave policies, so workers conducted subsistence activities
during their weeks off or would take personal time. Where the companies did have policies for
subsistence leave, an average of 46% of respondents were unsure whether or not the policy worked
(USEPA 2009).
The creation of mining-related jobs for local residents and attendant increases in the region's cash
economy are often mentioned as potential benefits of large-scale mining development. However,
Bristol Bay Assessment 1213 January 2014
-------
Chapter 12 Fish-Mediated Effects
increases in personal income may not be the best measure of benefits in a subsistence-based culture and
should be considered over the long-term, as oil, gas, or mineral resources are exhausted and future
opportunities—including subsistence resources—are potentially damaged. These types of damages
persist, even when resource extraction ceases (NRC 2003).
12.2.4 Social, Cultural, and Spiritual Impacts
The inability to harvest salmon from portions of these watersheds would result in some degree of
cultural disruption, which goes well beyond a loss of food supply. Boraas and Knott (Appendix D) state,
"The people in this region not only rely on salmon for a large proportion of their highly nutritional food
resources; salmon is also integral to the language, spirituality, and social relationships of the culture."
On a cultural level, a significant loss of salmon would result in negative stress on a culture that is highly
reliant on this resource. Boraas and Knott (Appendix D) discuss and document several of the social
values and activities that are integrated with subsistence, such as sharing and generalized reciprocity,
fish camp, steam baths, gender and age equity, and wealth. Likewise, they document how spirituality
and psychological health of the cultures are integrated with the natural world, especially salmon. Of
particular importance is the sharing and passing along of traditional knowledge to future generations.
This knowledge transfer occurs in several ways but one critical component is fish camp. According to
Boraas and Knott (Appendix D):
Families typically view fish camp as a good time when they can renew bonds of togetherness by
engaging in the physical work of catching and processing salmon. Family members who don't live
in the villages often schedule vacation time to return home to fish camp, not just for the salmon
but for family. The importance of sharing in vigorous, meaningful work cannot be overestimated. It
creates cross-generational bonds between children, their parents, aunts, uncles, and/or
grandparents that today are rare in Western culture because there are so few instances in which
meaningful, multi-generational work occurs.
Some interviewees expressed fear of the future, that a traditional prophecy of "bad times" told by
Elders might be coming true due to economic development resulting in cultural loss characterized
as "anomie," the loss of meaningfulness, sense of belonging, and direction in life. Anomie
increases cultural and individual risk for social ills such as depression and suicide, alcoholism and
drug abuse, domestic violence, and aggressive behaviors. Healing practices can include those
used for trauma and post-traumatic stress disorders, including traditional practices that reconnect
the individual to society and the natural environment through meditative rituals. Culture camps
and other methods of cultural revitalization can be both preventative and healing for children and
adults of indigenous cultures.
Acculturation is a commonly used concept to describe the psychological and cultural impacts of rapid
modernization and loss of tradition. Identity and involvement in cultural activities provide numerous
benefits to Alaska Natives. Participation in subsistence activities and consumption of subsistence foods
include cultural, traditional, and spiritual activities that involve the entire community. One of the
greatest risks to the Alaska Native communities in the Nushagak and Kvichak River watersheds with
respect to acculturation would arise from a major and persistent decline in the subsistence salmon
fishery. For the people on the Nushagak River who consider themselves the "King Salmon" people, any
impact on the Chinook salmon fishery would stress their community and the cultural traditions that
bind them together.
Bristol Bay Assessment 1214 January 2014
-------
Chapter 12 Fish-Mediated Effects
Studies on disruption to Alaska Native cultures from resource extraction industries illustrate the
potential social and cultural impacts of large-scale mining on a key subsistence resource in the
Nushagak and Kvichak River watersheds. Land use by Alaska Natives on the North Slope has been
mostly non-intensive, leaving few traces on the landscape outside the established villages. In contrast,
oil development has altered the landscape in ways that will persist long after resource extraction
activities have ceased. Testimony repeatedly cited "scars on the land" that result from industrial
development, and indicated that these scars have altered both the physical and spiritual elements of the
landscape and thus the very basis of Alaska Native cultures on the North Slope (NRC 2003).
Alterations to the North Slope physical environment have had aesthetic, cultural, and spiritual effects on
human populations (NRC 2003). These alterations have resulted primarily from the construction of
roads, pipelines, buildings, and power lines and from off-road travel. Hunters report that they do not
hunt in the oil fields for aesthetic reasons. North Slope residents have reported that the imposition of a
huge industrial complex on the Arctic landscape was offensive to the people and an affront to the spirit
of the land.
North Slope residents report that there has been a vast increase in the time, effort, and funding
necessary to respond politically and administratively to the ever-multiplying number of projects
proposed in the region (NRC 2003). Local residents must attend industry-related meetings and hearings
and review documents, because they believe that decisions will be made that can significantly affect
both their daily lives and future generations. Additionally, North Slope residents stated that increasing
anxiety about offshore and onshore development is widespread in North Slope communities. Hunters
worry about contamination of the food they consume and know that their health will suffer if they are
unable to eat as their ancestors did. They worry about not being able to provide for their families, or
about the added risk and expense if essential and traditional foods are harder to find. Elders who are no
longer able to provide for themselves worry about the challenges younger hunters face. Families worry
about the safety of hunters who must travel farther and more often if game is not easily accessible (NRC
2003).
According to the National Research Council (2003), increased alcoholism, drug abuse, and child abuse
have resulted from the stresses inherent in integrating traditional and new ways of life. Health effects
also are apparent, as the incidence of diabetes has increased with higher consumption of non-
subsistence foods (NRC 2003). The North Slope Borough bears the costs of these social stresses and
provides services such as counseling, substance abuse treatment, public assistance, crisis lines, shelters,
and other social service programs. It also supports search and rescue services and the police force that
respond to domestic violence and other situations arising when communities are subjected to long-term
and persistent stress. The borough supports biologists, planners, and other specialists who review and
offer recommendations on lease sale, exploration, and development project documents that are
produced each year, and bears the expense of traveling to Fairbanks, Anchorage, Juneau, Seattle, and
Washington, DC, where agencies with permitting authority make decisions that affect their way of life
(NRC 2003).
Bristol Bay Assessment 1215 January 2014
-------
Chapter 12 Fish-Mediated Effects
The goal of a more recent study on the effects of oil and gas development on subsistence harvesters on
the North Slope (Braund and Associates 2009, Braund and Kruse 2009) was to enhance benefits and
mitigate impacts of development. This study reported that, despite raising concerns about oil
development as early as 1975, the Inupiat have, until recently, been successful in maintaining their
subsistence lifestyle. Since 2003, North Slope active harvesters have been experiencing impacts of oil
development at higher rates and report that their wellbeing has declined. This has led to social
problems, including higher rates of drug and alcohol abuse and suicide.
A study that looked at the social, cultural, and psychological impacts of the Exxon Valdez oil spill
determined that the psychosocial impacts of contamination were as significant as the physical impacts
on the environment (Palinkas et al. 1993). Reported issues included declines in traditional social
relations with family members, friends, neighbors, and coworkers; perceived increases in the amount of
and problems associated with drinking, drug abuse, and domestic violence; a decline in perceived health
status; an increase in the number of physician-verified medical conditions; and increased post-spill rates
of generalized anxiety disorder, post-traumatic stress disorder, and depression (Palinkas et al. 1993).
Community-wide pre-occupation with the spill and cleanup affected traditional social relations and
resulted in conflicts and divisiveness, arguments about environmental effects of the spill, issues of fault
and responsibility, issues of whether or not to work on the cleanup, and related monetary and
employment issues (Palinkas et al. 1993). There were pervasive fears and increased fundamental
concerns about cultural survival for many residents in the affected Alaska Native villages.
Palinkas et al. (1993) documented the profound impact that exposure to the oil spill had on social
relations, traditional subsistence activities, the prevalence of psychiatric disorders, community
perceptions of alcohol and drug abuse and domestic violence, and the physical health of Alaskan Native
and non-native residents of the affected communities. Although the specifics of the Exxon Valdez oil spill
may be quite different, a large-scale or long-term failure of mine waste collection, treatment, or
containment systems would produce a similar reduction of subsistence activities, and similar social and
cultural effects could be expected.
12.2.5 Mitigation and Adaptation
It is not likely that any direct or indirect loss of subsistence use areas resulting from the mine footprints
could be avoided. In the mine scenarios, the mine pit, waste rock piles, and TSFs would remain on the
landscape in perpetuity and thus represent permanent habitat loss for salmon and other subsistence
resources. Some measures could be put in place to prevent and respond to accidents and spills. Small
spills and releases that are contained in a timely manner may not affect the salmon subsistence
resource. However, large-scale releases, even with active remediation, would have long-term effects on
the salmon subsistence resource and Alaska Native cultures. Because the Alaska Native cultures in this
area have significant ties to specific land and water resources that have evolved over thousands of years,
it would not be possible to replace the value of lost subsistence use areas elsewhere, or to relocate
residents and their cultures, making compensatory mitigation infeasible (Appendix J).
Bristol Bay Assessment 1216 January 2014
-------
Chapter 12 Fish-Mediated Effects
The ability of Alaska Native cultures to adapt to losses of subsistence use areas or to the larger impacts
of a mine failure or accident is unknown. Several studies have considered adaptation related to
subsistence resources. Holen (2009) studied the adaptations related to the Nondalton subsistence
fishery and identified two major socio-cultural factors that could potentially affect the long-term
resilience of the fishery: children and young adults are not actively participating in subsistence salmon
fishing as they have in the past, and because summer is often when seasonal employment is available,
some residents miss the subsistence fishing season because of work obligations. These factors interrupt
the inter-generational transfer of existing knowledge and wisdom and suggest that permanent cultural
change can result from cultural disruption. On Alaska's North Slope, the issue has not been a question of
whether Alaska Natives adapt to oil and gas development, but rather the consequences of that
adaptation (NRC 2003). There are two potential problems: the loss—sometimes quickly—of traditional
languages, patterns of behavior, economic activities, skills and capital improvements that are no longer
relevant; and the use of human and financial capital and non-renewable resources by the new
development (NRC 2003).
As the cash economy develops and Alaska Natives become involved in discussions about how changes
associated with oil and gas development affect their cultures, they increasingly must use English as their
primary language. They lose fluency in their native language and the traditional ecological knowledge
embedded in that language. Many North Slope residents expressed concern about the loss of their
traditional way of life, while at the same time enjoying the benefits of the cash economy (NRC 2003).
However, over-adaptation can also occur, leaving communities less able to survive in their
environments when extraction activities decline or stop. The significant tax revenues that oil and gas
development have provided North Slope Borough residents are now declining, and the current standard
of living for North Slope residents will be impossible to maintain unless significant external sources of
local revenue are found. If borough revenues decline, residents may face lower standards of living, or be
forced to find other sources of economic activity or migrate to different areas (NRC 2003).
Offshore exploration and development and the announcement of offshore sales have resulted in
perceived risks to Inupiat culture that are widespread and intense. People of the North Slope have a
centuries-old nutritional and cultural relationship with the bowhead whale, and most view offshore
industrial activity as a threat to bowheads and thus their cultural survival.
12.3 Uncertainties
The preceding sections provide a qualitative overview of how wildlife and Alaska Natives may be
affected by mining-associated changes in salmon resources. Because we mention but do not evaluate
direct effects of mining on wildlife and Alaska Natives (Box 12-1), this assessment represents a
conservative estimate of how these endpoints could be affected by mine development and routine
operations. We focused on a limited suite of wildlife species (Section 5.3), but additional species also
could be affected by changes in salmon resources. We also did not consider mining-related changes to all
subsistence species.
Bristol Bay Assessment 1217 January 2014
-------
Chapter 12 Fish-Mediated Effects
In addition to these scope-related limitations, there are several uncertainties inherent in our
consideration offish-mediated effects on wildlife and Alaska Natives.
• The magnitude of salmon-mediated effects on wildlife, subsistence resources, and indigenous
cultures is uncertain and cannot be quantified at this time. Ultimately, the magnitude of overall
impacts will depend on many factors, including the location and temporal scale of effects, cultural
resilience, the degree and consequences of cultural adaptation, and the availability of alternative
subsistence resources.
• Interactions between salmon and other wildlife species are complex and reciprocal, and the
assessment did not comprehensively evaluate all potential linkages between endpoints. Many of
these linkages have not been well-documented or researched (e.g., potential relationships between
MDN, riparian vegetation, and moose and caribou), but may be significant. Therefore, this
assessment likely underestimates salmon-mediated risks to wildlife.
• The magnitude of effects on Alaska Native cultures resulting from any mining-associated changes in
salmon resources is unknown, but other studies related to resource extraction industries (North
Slope, Red Dog Mine) or environmental contamination (Exxon Valdez) in Alaska confirm that there
certainly would be changes in human health and Alaska Native cultures.
• The cumulative effects of mining and climate change represent a significant uncertainty in the
region (Section 3.8, Box 14-2). Residents of the Kvichak River watershed have observed that social
and cultural changes are occurring in an environment where they are also seeing rapid climate
changes (Holen 2009). These changes, which include climate variability and unpredictable weather,
make it difficult to plan for subsistence activities (Appendix D). On Alaska's North Slope, climate
change and oil and gas development together result in greater cumulative effects on the
environment and Inupiat cultural traditions (Braund and Associates 2009). The cumulative effects
of climate change and potential effects on subsistence resources from large-scale mining are
unknown.
Despite these uncertainties, the inability to mitigate or replace subsistence resources or cultural values
lost to effects of large-scale mining is certain because of the significant and long-standing ties that Alaska
Native cultures have to specific land and water resources in these watersheds.
Bristol Bay Assessment 1218 January 2014
-------
Thus far, this assessment has focused on the potential effects of a single mine, described by a range of
mine scenarios. Although the Pebble deposit represents the most imminent and likely site of mine
development in the Nushagak and Kvichak River watersheds, the development of a number of mines of
varying sizes is plausible in this region. Several known mineral deposits with potentially significant
resources are located in the two watersheds (Table 13-1), and active exploration of deposits is occurring
in a number of claim blocks (Figure 13-1). If the infrastructure for one mine is built, it would likely
facilitate the development of additional mines. Thus, the potential exists in these watersheds for the
development of a mining district that could include a number of mines, their associated infrastructure,
and resulting induced development. In this chapter, we briefly consider potential cumulative effects of
the establishment of large-scale mining in the Nushagak and Kvichak River watersheds on Pacific
salmon. In addition to addressing the potential impacts of multiple mines, we briefly consider induced
development and potential increases in the accessibility of the watersheds' currently roadless areas.
13.1 Cumulative and Induced Impacts
13.1.1 Definition
National Environmental Policy Act regulations define cumulative impacts as "the impact on the
environment [that] results from the incremental impact of the action when added to other past, present,
and reasonably foreseeable future actions regardless of what agency (Federal or non-Federal) or person
undertakes such other actions. Cumulative impacts can result from individually minor but collectively
significant actions taking place over a period of time." Assessing the cumulative impacts of multiple
mines requires considering the impacts of their combined footprints, as well as the cumulative risks of
leaks, spills, and other accidents and failures associated with each individual mine.
Bristol Bay Assessment
13-1
January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
Induced effects contribute to the cumulative effects of an action, and are those effects that are "caused
by the action and are later in time or farther removed in distance, but are still reasonably foreseeable"
(43 CFR 1508.8(b)).
Figure 13-2 illustrates how cumulative and induced effects could follow the initiation of large-scale
mining in the Nushagak and Kvichak River watersheds. The original mine—with its associated
transportation corridor, port, power generation facilities, and other infrastructure—likely would initiate
the accumulation of impacts across the watersheds. Mineralized areas in the region (Figure 13-1) are
currently without development infrastructure (e.g., roads, utilities, and airports), which creates an
expensive barrier to development. Thus, it is reasonably foreseeable that infrastructure development for
an initial mine could make mining cost-effective for other, smaller mineral deposits, facilitating further
accumulation of impacts. In addition, the initial and subsequent mines would increase accessibility of
the region, causing other induced development and associated impacts.
As environmental effects on freshwater habitats accumulate, the magnitude of total impact on the
region's fisheries would increase. Increased spatial dispersion of effects, both within and across
watersheds, means that individual effects may go unnoticed but still cumulatively affect the greater
Bristol Bay salmon fishery. Cumulative effects associated with multiple mines would potentially reduce
biodiversity of the overall salmon population (Section 5.2.4) and its resilience to natural and
anthropogenic disturbances, thereby exacerbating total effects on salmon.
13.1.2 Vulnerability of Salmonids to Cumulative Impacts
Throughout the range of Pacific salmon, most ecosystems outside of the Bristol Bay watershed face the
cumulative effects of multiple land and water uses within and across watersheds, resulting in a variety
of stressors that occur in combination. Anadromous and resident fish stocks in these watersheds are
subject to persistent disturbance-induced stresses, the effects of which accumulate through the river
network. For example, sedimentation of spawning beds from accelerated erosion, loss of rearing habitat
from filling of streamside wetlands, and reduced out-migration success from downstream
channelization are separate effects that together have a cumulative impact on fish in a river system. The
effect of each stressor accumulates regardless of whether factors occur at the same time, or even in
temporal proximity. Because Pacific salmon, Dolly Varden, and rainbow trout are migratory, at least
within a given stream system, adverse impacts can even accumulate when fish are absent from a
particular reach. The overall result of these cumulative effects has been the reduction and even
extinction of many salmonid populations.
Historical salmon losses have resulted from the cumulative impacts of many land use activities over a
time span of 150 years (NRC 1996). Salmon depend on adequate supplies of clean, cool water
throughout the freshwater portions of their lifecycles. In addition, well-aerated streambed gravels are
essential for spawning. These instream conditions depend on the overall health of the entire watershed
(NRC 1996). Human development in the watershed can adversely affect these conditions by increasing
sedimentation, raising water temperature, degrading water quality, changing water flow, and reducing
water depths (NRC 1996).
Bristol Bay Assessment ^ „ 2 January 2014
-------
Chapter 13
Cumulative Risks of Multiple Mines
In the Pacific Northwest, the four principle factors responsible for the degradation of salmon stocks have
been referred to colloquially as the "four H's": habitat degradation and loss, hydroelectric dams and
other impoundments, harvest practices, and hatchery propagation (Ruckelshaus etal. 2002). Of these
factors, habitat degradation and loss is the most likely to affect salmon stocks in the Nushagak and
Kvichak River watersheds. In the Pacific Northwest, habitat degradation and loss related to human land
use have obviously been a major factor in salmon declines by reducing population productivity, adult
densities, and early-life-stage production over large geographic areas (Ruckelshaus et al. 2002).
Table 13-1. Mining prospects (in addition to the Pebble deposit) with more than minimal recent
exploration in the Nushagak and Kvichak River watersheds. See Figure 13-1 for prospect locations.
Prospect
AUDN
Big Chunk North
Big Chunk South
Fog Lake
Groundhog
Humble
Iliamna
Kamishak
Kaskanak
Kisa
Northern Bonanza
Shotgun
Sleitat
Pebble South/PEB (38/308 Zones/BOO)
Stuy
Resource
Porphyry copper
Porphyry copper
Porphyry copper
Gold, copper
Porphyry copper
Porphyry copper
Porphyry copper
Porphyry copper
Porphyry copper
Gold
Gold
Gold
Tin, tungsten
Porphyry copper
Porphyry copper
References
Millrock Resources 2011
Millrock Alaska 2012a
U5 Resources 2010
AH EA 2012
Big Chunk Corp. 2012
Liberty Star 2012a
Alix Resources 2008
USGS 2008
Kennecott Exploration Co. 2011
Szumigala etal. 2011
Millrock Resources 2011
Millrock Alaska 2012b
Bristol Exploration Co., Inc. 2011
AERI 2008
Full Metal Minerals 2008, 2012
Golden Lynx 2009
Northern Bonanza Trust 2011
TNR Gold Corp. 2011, 2012
ADNR2012a
Thor Gold Alaska, Inc. 2011
Full Metal Minerals 2008, 2012
PLP 2011
Szumigala etal. 2011
Stuy Mines, LLC2010
Bristol Bay Assessment
13-3
January 2014
-------
Chapter 13
Cumulative Risks of Multiple Mines
Figure 13-1. Claim blocks with more than minimal recent exploration in the Nushagak and Kvichak
River watersheds.
Northern Bonanza
Shotgun
Kisa
KVICHAK ^
Groundhog PortAlsworth
•Nondalton
Iliamna /
Pebble South/PEB •
Kokhanok
Igiugij
Bristol Bay
N
A
25
25
50
] Kilometers
50
] Miles
Fog Lake
Kamishak
Cook Inlet
\
Approximate Pebble Deposit Location
Towns and Villages
Watershed Boundary
Existing Roads
Active Mining Claims
Bristol Bay Assessment
13-4
January 2014
-------
Chapter 13
Cumulative Risks of Multiple Mines
Figure 13-2. Conceptual model illustrating potential cumulative effects of multiple mines.
initial mine
/ \
/ \
/ \
/
Cither ancillary
facilities
transportation j
corridor J
power generation &
transmission facilities
T economic feasibility of
additional mine development
f additional j f additional ancillary 1 f additional "I
I mines J L facilities J I infrastructure J
T accessibility
of region
T effective
mpermeable surface
induced
development
accumulation of environmental impacts
on freshwater habitats
LEGEND
Within a shape. [* indicates an increase in the
parameter, I- indicates a decrease in a parameter,
and A indicates a change in the parameter.
Arrows leading from one shape to another
indicate a hypothesized cause-effect relationship.
Shapes hi acketed under another shape are
specific components of the more general shape
under which they appear.
•f- spatial extent
of impacts
"1" magnitude of
impacts
effects on T- number of
intraspecificfish stocks
V
4-fish population
resiliency
V
V
fish effects
Bristol Bay Assessment
13-5
January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
13.1.3 Nature and Extent of Past, Present, and Future Impacts
In cumulative impact analyses, the U.S. Environmental Protection Agency (USEPA) generally evaluates
any past, present, and reasonably foreseeable future actions that are spatially and temporally linked and
thus can act in combination on the resource (s) of interest. In the Bristol Bay watershed, the contribution
of past or present actions to degradation of salmon habitat and populations is minimal. To date, the
Nushagak and Kvichak River watersheds have experienced minimal cumulative stresses associated with
human activity, and their ecosystems are relatively undisturbed by significant human development
(Section 3.7). Large-scale, human-caused modification of the landscape—a factor contributing to the
extinction risk for many native salmon populations in the Pacific Northwest (Nehlsen et al. 1991)—is
absent from these watersheds, and development consists of only a small number of towns, villages, and
roads. There are no hydroelectric dams and reservoirs and no salmon hatcheries. In fact, harvesting is
the only one of the four H's responsible for devastating salmon stocks of the Pacific Northwest that is
applicable to the Nushagak and Kvichak River watersheds. There have been periods of poor harvesting
practices and overfishing in Bristol Bay in the past. However, when Alaska achieved statehood, the
Alaska Department of Fish and Game (ADF&G) assumed management of the fishery and established the
primary objective of restoring the runs to their former abundance (Box 5-4, Appendix A). The ADF&G
management strategy, based on maximum sustainable yield, is considered a success in maintaining
sustainable salmon harvests (Hilborn et al. 2003, Hilborn 2006). Indeed, since the late 1970s, sockeye
salmon catch, spawning stock, and total returns have been at record levels. Although there has been
some concern that harvest of returning salmon has reduced ecosystem productivity in this region,
Schindler et al. (2005) found that paleoecological analysis of returns to Lake Nerka in the Wood River
system did not suggest decreased salmon production due to commercial fishing.
Reasonably foreseeable future development can be predicted based on project approvals, planning
documents, and data on local trends. The Bristol Bay Area Plan for State Lands (BBAP) (ADNR 2005)
provides information on reasonably foreseeable mining in the Nushagak and Kvichak River watersheds.
Although the Pebble deposit represents the most imminent and likely site of mine development in the
watersheds, the development of several mines of varying sizes is plausible in this region. Several known
mineral deposits with potentially significant resources are located in the two watersheds, and active
exploration of deposits is occurring in a number of claim blocks (Figure 13-1, Table 13-1). Reasonably
foreseeable non-mining development is discussed in Section 13.3.
13.2 Cumulative Impacts from Multiple Mines
Construction of mining and transportation infrastructure at and for the Pebble deposit would
substantially reduce development costs for surrounding prospects and could facilitate creation of a
mining district. Based on planning documents and current patterns of mineral exploration in the
Nushagak and Kvichak River watersheds, it is possible to identify a scenario for potential mine
development in the region over the next 50 to 100 years. Although this scenario is plausible given
available information, it should be kept in mind that it is impossible to predict with certainty what
Bristol Bay Assessment ^ „ g January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
mining activities will occur in the region in the future, the order in which mines will be developed, or the
specific impacts of those mines.
The BBAP assigns land use designations to discrete areas of state-owned or selected lands called
management units. These designations represent the uses and resources for which the Alaska
Department of Natural Resources will manage the units. In the Nushagak and Kvichak River watersheds,
the BBAP assigns the land use designation "Mi" (Minerals) to seven management units (Shotgun, Sleitat,
Kemuk, Fog Lake, and units 06-23, 06-24, and 10-02 in the Pebble deposit area), which total more than
1,300 km2 (ADNR 2005). The Mi designation applies to areas "associated with significant resources,
either measured or inferred, that may experience minerals exploration or development during the
[BBAP's] planning period" of 20 years. The BBAP also allows for mining on units with other designations
(e.g., General Use, Public Recreation and Tourism-Dispersed) that far exceed the area of those
designated Mi, although the BBAP does not describe them as having the same known potential for
mining.
Since 2008, there has been exploration in all of the BBAP Mi-designated management units, as well as in
claim blocks with several other designations. Table 13-1 lists mine prospects experiencing more than
minimal recent exploration activity. These target areas could be future mine sites, if exploration
identifies marketable quantities of metals. Other mineral claim blocks exist in the watersheds, but at the
time of this writing they had experienced limited exploration in recentyears (ADNR 2012b, 2012c,
2012d).
Any potential mine site would presumably include a mine pit and an adjacent waste rock disposal area
(Chapter 6). For analysis purposes, we assume the size of other ore bodies in the area would be more
typical of worldwide porphyry copper deposits than the Pebble deposit (Table 4-2). We used the Pebble
0.25 scenario, which is comparable to a median-size porphyry copper mine, to characterize the
footprints of the major mine components (mine pit, waste rock piles, and tailings storage facilities
[TSFs]) for additional mines. The Pebble 0.25 scenario represents 250 million tons (230 million metric
tons) of ore, resulting in a mine pit and waste rock disposal area of approximately 1.5 and 2.3 km2,
respectively (Table 6-2). Mines affiliated with or close to an existing mine (e.g., a mine at the Pebble
deposit) may be able to use the TSFs, mill, and other infrastructure constructed for that mine. Mines not
affiliated with or more distant from a previously developed mine would require one or more TSFs (an
additional 6.8 km2 in the Pebble 0.25 scenario) (Table 6-2), as well as a mill and other operational
infrastructure as described in Box 6-1. Thus, for potential mines distant from Pebble, we calculated the
footprints of the major mine components both with and without TSFs. Any additional mines would also
require construction of transportation infrastructure, including access roads, pipelines, and possibly
port facilities.
To examine the potential scope of cumulative impacts from large-scale mining, we consider
development of additional mines at six potential sites where there was notable activity and/or
investment in drilling or other exploration in 2011 to 2012: Pebble South/PEB (PLP/NDM claim block),
Big Chunk South, Big Chunk North, Groundhog, AUDN/Iliamna, and Humble prospects (Figure 13-1).
Bristol Bay Assessment ^ „ -, January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
This list does not include four other prospects designated as Mi in the BBAP: Shotgun, Sleitat, Kemuk,
and Fog Lake. Kemuk is an older name for the Humble prospect; the other three prospects are not
porphyry copper deposits. Because exploration of the six selected prospects began approximately 15 to
25 years later than exploration at the Pebble deposit, proposals to develop mines at these sites could be
20 years or more in the future (Millrock Resources 2011, ADNR 2012e, 2012f, and 2012g, Liberty Star
2012a). We describe the waters, fishes, and subsistence resources that could be affected by mines at
these locations (Tables 13-2 through 13-7). The sources of information for these tables are the Alaska
Anadromous Waters Catalog (Johnson and Blanche 2012), the Alaska Freshwater Fish Inventory
(ADF&G 2012), and a series of technical papers from ADF&G on subsistence harvest and use by villages
in the Nushagak and Kvichak River watersheds (Fall et al. 1986, Schichnes and Chythlook 1991, Fall et
al. 2006, Krieg et al. 2009). We also estimate the stream lengths and wetland areas that could be
eliminated by the footprint of the major mine components at each site. Box 13-1 describes the
methodology for estimating these impacts; results across the six potential sites are summarized in Table
13-8. Itis important to note that we did notestimate the size of the groundwater drawdown zones
around dewatered pits at the six additional mines as we did at the Pebble site, so our estimates of
habitat loss are conservative. Inclusion of the drawdown zone in the Pebble 0.25 scenario increases
stream and wetland losses by roughly 50%. Similar increases in habitat loss estimates, subject to
variations in the local geology, would be expected at each of the other mine sites.
13.2.1 Pebble South/PEB
13.2.1.1 Description
The Pebble South/PEB prospect, which is part of the PLP/NDM claim block, includes the 38 and
308 Zone prospects, approximately 15 km south west of the Pebble deposit (Ghaffari etal. 2011). The
BOO prospect is 4km south of 308 Zone on claims held by Full Metal Minerals (USA), Inc. (Ghaffari etal.
2011, Full Metal Minerals 2012). Full Metal entered into an option agreement with PLP/NDM; if
completed, the option will result in at least 60% PLP/NDM interest in the claims (Full Metal Minerals
2012).
Due to its proximity to the Pebble deposit, we assume that any future mines at Pebble South/PEB would
use the TSFs, mill, and other operational infrastructure initially built for mining at the Pebble deposit.
Thus, we anticipate that the primary additional development associated with this prospect would be a
mine pit, waste rock areas, and a transportation corridor to existing operational infrastructure.
Bristol Bay Assessment ^ „ o January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
BOX 13-1. METHODS FOR ESTIMATING IMPACTS OF OTHER MINES
To estimate the extent of aquatic habitat that each mine would eliminate, we overlaid typical footprints of
the major mine components (mine pit, waste rock piles, and tailings storage facilities) onto the stream,
water body, and, where available, wetland densities for each entire claim block. In this way the analysis is
less affected by uncertainty about the precise location of a potential mine within the claim block. Since the
Pebble South/PEB prospects are associated with the PLP/NDM claims, we used the aquatic area densities
for the PLP/NDM claims to assess that potential mine.
We derived the boundaries of the prospects from the Alaska Department of Natural Resources' State Mining
Claims dataset (ADNR 2012h). We then determined stream and water body density using the National
Hydrography Dataset (NHD) for Alaska (USGS 2012). For wetland density, we used the National Wetlands
Inventory (NWI) coverage for the Groundhog, AUDN/lliamna, and Pebble South/PEB (part of the PLP/NDM
claim block) prospects (USFWS 2012). The NWI covered 95, 69, and 58% of these prospects, respectively.
For these three areas, we compared NHD water body density to NWI wetland density to ascertain the
efficacy of using the former as a surrogate for wetland density in areas where NWI coverage was lacking.
This analysis revealed that the NHD water body dataset severely underestimates wetland density in areas of
overlap: NWI wetland density was roughly 10 to 14 times the NHD water body density. Thus, for the three
prospects with no NWI coverage (Big Chunk South, Big Chunk North, and Humble), we calculate a range of
wetland impacts, using NHD water body density as a lower bound and roughly 14 times that density as an
upper bound. We also provide a range of wetland impacts for the three claim blocks with NWI coverage,
because the NHD water body density in the NWI-covered area of all three prospects was lower than for the
claim block as a whole. For those three claim blocks (Pebble South/PEB, Groundhog, and AUDN/lliamna),
the higher estimate applies the wetland-to-water body differential from the area of NWI/NHD overlap to the
full claim block's higher water body density.
For Pebble South/PEB, we used a direct impact area that represents only the typical mine pit and waste rock
disposal area, based on our assumption that any mine at that site would use the mill, tailings storage
facility, and other facilities at an initial mine at the Pebble deposit (Section 13.2.1). We applied a similar
assumption as a lower bound for Big Chunk South, Big Chunk North, and Groundhog: for all three, the upper
bound represents a stand-alone mine, with no shared facilities (Sections 13.2.2 through 13.2.4). We
assume no sharing of mine facilities for AUDN/lliamna or Humble, based on their more remote locations
(Sections 13.2.5 and 13.2.6).
13.2.1.2 Potentially Affected Waters, Fishes, and Subsistence Uses
Table 13-2 summarizes information on the waters, fishes, and subsistence uses potentially affected by a
mine at the Pebble South/PEB prospect The 38 and 308 Zones occur near the south edge of the South
Fork Koktuli River watershed, within the Mulchatna River watershed of the Nushagak River watershed.
The upper reaches of some streams on Sharp Mountain likely are too steep to provide fish habitat and
there have been few fish surveys in this area to date, even in the lower reaches of those streams. Drilling
on the BOO prospect has been just south of the divide between the Nushagak and Kvichak River
watersheds, in the uppermost portion of the Lower Talarik Creek watershed, which flows to Iliamna
Lake. BOO is located approximately 2 km southwest of the fish-bearing stream that drains the south side
of Sharp Mountain. No fish survey data are available for the immediate area of the BOO prospect, and the
streams may be spatially intermittent.
Connecting to infrastructure at an existing mine at the Pebble deposit likely would involve following the
South Fork Koktuli River upstream. This route would presumably involve crossing the river in addition
to a number of tributaries, water bodies, and wetlands.
Bristol Bay Assessment ^ „ p January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
Waters of both the Nushagak and Kvichak River watersheds could be affected at this site, although no
information is available on fish resources in the tributaries of the Kvichak River that would be affected
by mining of the BOO prospect. The tributaries of the Nushagak River that would be affected by mining
of the 38 and 308 Zone prospects are known to contain four Pacific salmon species and Dolly Varden.
People from seven Alaska Native villages hunt for bear, moose, caribou, other mammals, and birds in
these areas, but no subsistence fishing has been reported (Table 13-2). Based on the average stream
density in the area of the prospect, the footprint of the major mine components would eliminate 4.1 km
of streams and between 0.71 and 1.2 km2 of wetlands (Table 13-8).
13.2.2 Big Chunk South
13.2.2.1 Description
The Big Chunk South prospect may be of the same geologic origin as the Pebble deposit. The claim block
abuts the north edge of the PLP/NDM claim block, approximately 20 km north of the potential Pebble
mill site (AREA 2012, Big Chunk Corp. 2012) and approximately 24 km northwest of the village of
Nondalton.
Based on its proximity to the Pebble deposit, a future mine at Big Chunk South may use the TSFs, mill,
and other facilities built for potential mining at the Pebble deposit, under a joint venture or other
agreement with PLP/NDM. In late 2012, Big Chunk partner Liberty Star settled debt and terminated
joint-venture negotiations with NDM (Liberty Star 2012b). For the purposes of this assessment, we
consider Big Chunk South under two scenarios: one in which it shares some facilities built for a mine at
the Pebble deposit and the other in which it operates as a fully separate, stand-alone mine, with no
shared facilities other than the transportation corridor connecting the Pebble deposit site to Cook Inlet
(Table 13-8). In this chapter, we refer to this transportation corridor as the assessment corridor. A mine
at Big Chunk South presumably would connect to the roads and pipelines of the assessment corridor
somewhere near its western terminus, currently estimated to be approximately 14 km south of the Big
Chunk South claim block.
Bristol Bay Assessment ^ „ ^ Q January 2014
-------
Chapter 13
Cumulative Risks of Multiple Mines
Table 13-2. Waters, fishes, and subsistence uses potentially affected by a mine at the Pebble South/PEB prospect.
Location
38/308 Zones and
BOO Prospect
Transportation
Corridor to Pebble
Sited
Downstream of
Mine(s) and
Transportation
Corridor
Affected Waters
Watersheds and Named Subwatersheds
Nushagak River
X
X
Mulchatna River
X
X
X
CD
>
cc
3
^£
0
^
X
X
X
(D
>
cc
3
^£
0
^
Ji
£
£
I
X
X
Kvichak River
X
X
(D
Ji
3
CB
c
E
CO
X
X
Lower Talarik Creek
X
X
Waters3
14 unnamed streams
2 unnamed lakes and
several ponds
Unknown extent of
wetlands
4 unnamed first-order
streams
Numerous ponds
Unknown extent of
wetlands
South Fork Koktuli River
>4 unnamed tributaries
Unknown extent of
wetlands
South Fork Koktuli River
Lower Talarik Creek
Fish Species'1
Unknown Lamprey
Longnose Sucker
Northern Pike
±i
(D
V)
T3
1
Round Whitefish
Unknown Whitefish
Coho Salmon
/\_
Sp, J
Chinook Salmon
J
Sockeye Salmon
J
Chum Salmon
A
Pink Salmon
CD
.C
o
u
B
<
Rainbow Trout
Dolly Varden
A, J
Arctic Grayling
J
+->
o
.Q
3
CO
Ninespine Stickleback
A, J
Threespine Stickleback
Slimy Sculpin
A, J
Unknown
A-
Sp
A-
Sp
A, J
A, J
A-
Sp
A-
Sp
A, J
A, J
X
A-
Sp, J
A-
Sp, J
A-
Sp, J
J
A-
Sp, J
A-
Sp, J
A-
Sp, J
X
X
X
X
A, J
A, J
A-
Sp,
J
A, J
X
A, J
A-
Sp,
J
X
A, J
A, J
A
A-
Sp,
J
A, J
A, J
Subsistence Usesc
Village
Dillingham
Ekwok
Igiugig
Iliamna
Newhalen
Nondalton
Port Alsworth
Dillingham
Ekwok
Igiugig
Iliamna
Newhalen
Nondalton
Port Alsworth
Igiugig
Iliamna
Kokhanok
Newhalen
NewStuyahok
Nondalton
Target Species/Group
Salmon
X
Other Salmonids
X
X
X
X
1
!a
u_
0
1
X
X
(C
(D
ca
|
m
X
X
X
o
o
E
X
X
X
X
X
X
X
Caribou
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Other Mammals
X
X
X
X
X
X
X
X
Waterfowl
X
X
X
X
in
•c
in
0
1
X
X
X
X
Notes:
a "Tributary" indicates that the channel flows into the stream listed above it; "stream" indicates that the receiving water is offsite as identified in the columns to the left. For waters downstream of the mine and transportation corridor, fishes and subsistence uses are noted only where different from the waters' prior listing.
b A = adult; Sp = spawning; J = juvenile; x = unknown life stage. "Unknown" indicates apparent lack of surveys.
c Subsistence uses are not separated by watershed. Uses are noted only for areas of direct impacts (i.e., mine and/or transportation corridor), if they occur there; otherwise, they are noted for areas downstream. "Downstream" applies only to the drainage immediately downstream (e.g., Lower Talarik Creek, but not Iliamna Lake or the
Kvichak River). Data for Dillingham and Ekwok are less detailed and more dated than for other villages, so use patterns may have changed.
d Assumes existing mine infrastructure at the Pebble deposit; hypothetical routing of the new corridor minimizes distance, topographic gradients, and stream crossings and assumes water body crossings would be avoided.
Sources: Siedelman etal. 1973, Russell 1974, Russell 1975, Fall etal. 1986, Schichnesand Chythlook 1991, Fall etal. 2006, Kriegetal. 2009, ADF&G 2012, Johnson and Blanche 2012.
Bristol Bay Assessment
13-11
January 2014
-------
Chapter 13
Cumulative Risks of Multiple Mines
Table 13-3. Waters, fishes, and subsistence uses potentially affected by a mine at the Big Chunk South prospect.
Location
Claim Block
Transportation
Corridor to
Pebble Sited
Downstream
of Mine(s) and
Transportation
Corridor
Affected Waters
Watersheds and Named Subwatersheds
Nushagak River
X
X
Mulchatna River
X
X
0
>
(£
1
O
^
X
X
North Fork Koktuli River
X
X
Kvichak River
X
X
X
X
Iliamna Lake
X
X
X
X
Newhalen River
X
X
Ji
CD
O
(D
Ji
5
X
X
Chulitna River
X
X
Upper Talarik Creek
X
X
Waters3
>66 unnamed streams
Nikabuna Lakes
130 other lakes and
ponds
Extensive wetlands
North Fork Koktuli
River
>7 unnamed
tributaries
Unknown extent of
wetlands
Upper Talarik Creek
X3 unnamed
tributaries
Chulitna River
North Fork Koktuli
River
Upper Talarik Creek
Fish Species'1
Northern Pike
A, J
A, J
Longnose Sucker
J
A
Humpback Whitefish
A
Least Cisco
A, J
A, J
Pygmy Whitefish
J
Round Whitefish
J
Coho Salmon
A
Sp,
J
A-
Sp,
J
Chinook Salmon
J
J
A-Sp
A-
Sp,
J
Sockeye Salmon
A
X
X
A-
Sp,
J
A-
Sp,
J
Chum Salmon
A-Sp
A-
Sp,
J
Pink Salmon
X
CD
6
O
«
<
X
X
Rainbow Trout
J
A
Dolly Varden
A, J
A, J
A, J
Arctic Grayling
A, J
J
A, J
A
Burbot
A, J
Ninespine Stickleback
A, J
A
A, J
Threespine Stickleback
A, J
A
Slimy Sculpin
A, J
A, J
A, J
Subsistence Usesc
Village
Dillingham
Ekwok
Newhalen
Nondalton
Port Alsworth
Dillingham
Ekwok
Iliamna
Newhalen
Nondalton
Port Alsworth
Igiugig
Iliamna
Newhalen
New
Stuyahok
Nondalton
Port Alsworth
Target Species/Group
Salmon
X
Other Salmonids
1
in
Li-
CD
.C
+->
O
X
X
X
CO
1
m
X
X
X
(D
i
X
X
X
X
X
X
X
Caribou
X
X
X
X
X
X
X
X
X
X
X
Other Mammals
X
X
X
X
X
X
Waterfowl
X
X
X
X
(/)
•c
in
(D
.C
+->
O
X
X
X
X
Notes:
a "Tributary" indicates that the channel flows into the stream listed above it; "stream" indicates that the receiving water is off-site, as identified in the columns to the left. For waters downstream of the mine and transportation corridor, fishes and subsistence uses are noted only where different from the waters' prior listing.
b A = adult; Sp = spawning; J = juvenile; x = unknown life stage.
c Subsistence uses are not separated by watershed. Uses are noted only for areas of direct impacts (i.e., mine and/or transportation corridor), if they occur there; otherwise, they are noted for areas downstream. "Downstream" applies only to the drainage immediately downstream (e.g., the Chulitna River, but not Lake Clark or the
Newhalen River). Data for Dillingham and Ekwok are less detailed and more dated than for other villages, so use patterns may have changed.
d Assumes existing mine infrastructure at the Pebble deposit; hypothetical routing minimizes distance, topographic gradients, and stream crossings and assumes water body crossings would be avoided.
Sources: Fall etal. 1986, Schichnesand Chythlook 1991, Stickman etal. 2003, Fall etal. 2006, Woody and Young 2006, Kriegetal. 2009, ADF&G 2012, Johnson and Blanche 2012.
Bristol Bay Assessment
13-12
January 2014
-------
Chapter 13
Cumulative Risks of Multiple Mines
Table 13-4. Waters, fishes, and subsistence uses potentially affected by a mine at the Big Chunk North prospect.
Location
Claim Block
Transportation
Corridor to Big
Chunk South'1
Downstream
of Mine(s) and
Transportation
Corridor
Affected Waters
Watersheds and Named Subwatersheds
Nushagak River
X
X
X
X
Mulchatna River
X
X
X
X
Keefer Creek
X
X
(D
>
(£
1
0
^
X
X
0
>
cc
c
CB
35
X
X
North Fork Swan
River
X
X
Kvichak River
X
X
X
(D
Ji
5
CD
C
E
CO
X
X
X
Newhalen River
X
X
X
^
CB
0
CD
Ji
3
X
X
X
Chulitna River
X
X
X
Waters3
Chulitna River
>JO tributaries
xLOO lakes and ponds
Unknown extent of wetlands
Keefer Creek headwaters
>7 tributaries
X35 lakes and ponds
Unknown extent of wetlands
Unnamed stream
Chulitna River
x4 tributaries
Unknown extent of wetlands
Chulitna River
Keefer Creek
North Fork Swan River
Fish Species'1
Arctic-Alaskan
Brook Lamprey
Northern Pike
Longnose Sucker
Humpback
Whitefish
Least Cisco
Pygmy Whitefish
Round Whitefish
Coho Salmon
Chinook Salmon
Sockeye Salmon
Chum Salmon
Arctic Char
Dolly Varden
Arctic Grayling
Burbot
Ninespine
Stickleback
c
|o.
3
o
(/)
>*
(75
Unknown
Unknown
Unknown
Unknown
A, J
A, J
A, J
A
A, J
J
J
A-
Sp
A-
Sp,
J
A-
Sp
J
X
X
Sp
A-
Sp
X
X
A, J
J
A, J
J
A, J
A, J
A, J
A, J
Subsistence Usesc
Village
Dillingham
Ekwok
Nondalton
Port Alsworth
Dillingham
Ekwok
Nondalton
Port Alsworth
Iliamna
Newhalen
New
Stuyahok
Nondalton
Port Alsworth
Target Species/Group
Salmon
Other Salmonids
IB
.G
en
Li-
CD
1
X
X
CD
&
C
O
to
X
X
Moose
X
X
X
Caribou
X
X
X
X
X
X
X
X
X
Other Mammals
X
X
X
X
X
X
X
Waterfowl
X
X
X
en
•c
in
(D
1
X
X
Notes:
a "Tributary" indicates that the channel flows into the stream listed above it; "stream" indicates that the receiving water is off-site, as identified in the columns to the left. For waters downstream of the mine and transportation corridor, fishes and subsistence uses are noted only where different from the waters' prior listing.
b A = adult; Sp = spawning; J = juvenile; x = unknown life stage. "Unknown" indicates an apparent lack of surveys.
c Subsistence uses not separated by watershed. Uses noted only for areas of direct impacts (i.e., mine and/or transportation corridor), if they occur there; otherwise, noted for areas downstream. "Downstream" applies only to the drainage immediately downstream (e.g., North Fork Swan River, but not the Swan or Koktuli Rivers). Data for
Dillingham and Ekwok are less detailed and more dated than for other villages, so use patterns may have changed (see Section 13.3.6.2).
d Waters, fishes, and subsistence uses limited to those upstream of the Big Chunk South claim block. See Table 13-3 for resources potentially affected by connecting a transportation corridor through and beyond that block. Hypothetical routing minimizes distance, topographic gradients, and stream crossings and assumes water body
crossings would be avoided.
Sources: Fall etal. 1986, Schichnesand Chythlook 1991, Stickman etal. 2003, Fall etal. 2006, Woody and Young 2006, Kriegetal. 2009, ADF&G 2012, Johnson and Blanche 2012.
Bristol Bay Assessment
13-13
January 2014
-------
Chapter 13
Cumulative Risks of Multiple Mines
Table 13-5. Waters, fishes, and subsistence uses potentially affected by a mine at the Groundhog prospect.
Location
Claim Block
Transportation
Corridor to
Pebble Sited
Affected Waters
Watersheds and Named Subwatersheds
Nushagak River
X
Mulchatna River
X
(D
>
(£
1
0
^
X
(D
^ .>
£*
€i
0 0
z ^
X
Kvichak River
X
X
X
X
X
X
X
X
X
(D
^
5
CD
c
E
CO
X
X
X
X
X
X
X
X
X
Upper Talarik
Creek
X
X
Newhalen River
X
X
X
X
X
X
X
X
^
CD
0
(D
^
3
X
X
X
X
X
X
X
X
Chulitna River
X
X
X
X
X
X
X
X
^£
(D
0
o
^
o
o
cc
X
X
Groundhog Creek
X
£
5
uo
c
3
X
Koksetna River
X
X
^£
(D
(D
0
^
O
CD
CD
X
Waters3
Chulitna River
>72 unnamed tributaries
Joe Nort Lake
58 unnamed lakes and ponds
Extensive wetlands
Rock Creek
X39 unnamed tributaries
9 unnamed lakes and ponds
Extensive wetlands
Groundhog Creek
>20 unnamed tributaries
30 unnamed ponds and lakes
Unknown wetland extent
^5 unnamed tributaries
Unknown wetland extent
Unknown wetland extent
Black Creek
>7 unnamed tributaries
one unnamed lake
Unknown wetland extent
Unknown wetland extent
>2Q unnamed tributaries
42 unnamed ponds and lakes
Unknown wetland extent
^13 unnamed tributaries
7 unnamed ponds
Unknown wetland extent
>.! unnamed tributary
Unknown extent of wetlands
Fish Species'1
Northern Pike
Longnose Sucker
Humpback
Whitefish
Least Cisco
Round Whitefish
Coho Salmon
Chinook Salmon
Sockeye Salmon
Arctic Char
Rainbow Trout
Dolly Varden
Arctic Grayling
J
Burbot
A, J
Ninespine
Stickleback
c
|o.
^
u
(/)
>*
(75
A, J
A, J
Unknown
Unknown
Unknown
Unknown
Unknown
A-
Sp,
J
J
A, J
A, J
A, J
Subsistence Usesc
Village
Dillingham
Ekwok
Kokhanok
Newhalen
Nondalton
Port Alsworth
Dillingham
Ekwok
Kokhanok
Newhalen
Nondalton
Port Alsworth
Target Species/Group
Salmon
Other Salmonids
Other Fishes
CD
&
C
O
m
X
X
X
X
o
0
X
X
X
X
Caribou
X
X
X
X
X
X
X
X
X
X
X
X
Other Mammals
X
X
X
X
X
X
Waterfowl
X
X
-------
Chapter 13
Cumulative Risks of Multiple Mines
Table 13-5. Waters, fishes, and subsistence uses potentially affected by a mine at the Groundhog prospect.
Location
Downstream
of Mine(s) and
Transportation
Corridor
Affected Waters
Watersheds and Named Subwatersheds
Nushagak River
X
Mulchatna River
X
(D
>
(£
1
0
^
X
(D
^ .>
£*
€i
0 0
z ^
X
Kvichak River
X
X
X
X
X
(D
Ji
5
CD
C
E
CO
X
X
X
X
X
Upper Talarik
Creek
X
Newhalen River
X
X
X
X
^
CB
0
(D
Ji
3
X
X
X
X
Chulitna River
X
X
X
X
J£
(D
0
o
^
o
o
cc
X
Groundhog Creek
£
5
uo
c
3
X
Koksetna River
X
J£
(D
(D
0
J£
U
CB
CD
Waters3
Chulitna River
Rock Creek
Unnamed tributaries
Koksetna River
Unnamed tributaries
Unnamed tributaries
Fish Species'1
Northern Pike
A, J
Longnose Sucker
A, J
}
\
A, J
Humpback
Whitefish
A
Least Cisco
A, J
Round Whitefish
}
A, J
A, J
Coho Salmon
A-
Sp,
J
Chinook Salmon
A
A, J
J
Sockeye Salmon
X
A-
Sp
Arctic Char
X
Rainbow Trout
A, J
Dolly Varden
Arctic Grayling
A, J
A
A, J
J
Burbot
A, J
J
Ninespine
Stickleback
A, J
c
|o.
^
u
(/)
>*
(75
A, J
A, J
Subsistence Usesc
Village
Igiugig
Iliamna
Newhalen
Nondalton
Port Alsworth
Target Species/Group
Salmon
X
Other Salmonids
X
Other Fishes
X
CD
&
C
O
m
o
0
X
X
Caribou
X
Other Mammals
X
X
Waterfowl
X
X
X
-------
Chapter 13
Cumulative Risks of Multiple Mines
Table 13-6. Waters, fishes, and subsistence uses potentially affected by a mine at the AUDN/lliamna prospect.
Location
^
o
o
ca
E
CD
0
Transportation
Corridor to
Levelockd
on Corridor to
4 unnamed tributaries
Numerous unnamed lakes and
ponds
Unknown extent of wetlands
Yellow Creek
>70 unnamed tributaries
Numerous unnamed lakes and
ponds
Extensive wetlands
Unnamed stream
3 unnamed ponds
Unknown extent of wetlands
>1 unnamed tributary
Unknown extent of wetlands
Kvichak River
Unknown extent of wetlands
Alagnak River
>1 unnamed tributary
Unknown extent of wetlands
Coffee Creek
>4 unnamed tributaries
Unknown extent of wetlands
Kvichak River (>4 crossings)
>8 unnamed tributaries
Unknown extent of wetlands
Ole Creek
Unknown wetland extent
Pecks Creek
Unknown extent of wetlands
>15 unnamed tributaries
Unknown extent of wetlands
Fish Species'1
Arctic-Alaskan Brook
Lamprey
Longnose Sucker
Northern Pike
Alaska Blackfish
Rainbow Smelt
Humpback Whitefish
Round Whitefish
Unknown Whitefish
Unknown
}
A, J
Unknown
Coho Salmon
X
Chinook Salmon
A-Sp
A-Sp
Sockeye Salmon
A-Sp
A-Sp
Chum Salmon
A-Sp
A-Sp
A-Sp
Pink Salmon
Arctic Char
X
X
Rainbow Trout
Dolly Varden
Lake Trout
Arctic Grayling
1
=
m
Ninespine Stickleback
Threespine Stickleback
Unknown Stickleback
Coastrange Sculpin
Slimy Sculpin
Unknown Sculpin
Unknown
J
A, J
A, J
Unknown
Unknown
A
A, J
A
A
X
X
A
A-Sp,
J
X
A-Sp
A
A
A
A-Sp
X
A
A
X
A
A
X
A
A
A
A, J
Unknown
A
A
A
A
A
A
X
X
X
A
X
X
X
A
X
A-Sp
A
A-Sp
X
A-Sp
A
X
A-Sp
A
X
X
X
X
A
A
A
A
Subsistence Usesc
Village
Dillingham
Ekwok
Igiugig
Kokhanok
Koliganek
Levelock
New
Stuyahok
Ekwok
Levelock
New
Stuyahok
Ekwok
Igiugig
Koliganek
Levelock
Ekwok
Igiugig
Iliamna
Kokhanok
Koliganek
Target Species/Group
Salmon
X
X
X
Other Salmonids
X
X
X
X
1
in
Li-
CD
.c
+->
O
X
X
X
JB
CO
CO
uo
3
CD
m
X
X
X
CO
CD
(/)
X
X
CO
1
m
X
(D
i
X
X
X
X
X
X
Caribou
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Other Mammals
X
X
X
X
X
X
Waterfowl
X
X
X
X
X
X
X
(/)
•c
in
(D
.C
+->
O
X
X
X
X
Bristol Bay Assessment
13-16
January 2014
-------
Chapter 13
Cumulative Risks of Multiple Mines
Table 13-6. Waters, fishes, and subsistence uses potentially affected by a mine at the AUDN/lliamna prospect.
Location
Transportation Corridor
to Newhalen6 (Continued)
Downstream of Mine(s) and
Transportation Corridor
Affected Waters
Watersheds and Named Subwatersheds
Kvichak River
X
X
X
X
X
(D
(D
O
(D
(D
1
Alagnak River
Levelock Creek
Jensen Creek
X
(D
(D
0
_0
(D
0
O
(D
O
Pecks Creek
3
CO
c
E
CO
X
X
Lower Talarik Creek
X
Upper Talarik Creek
X
Pete Andrews Creek
X
Newhalen River
X
Waters3
Lower Talarik Creek
>3 unnamed tributaries
Wetlands at least along stream
Upper Talarik Creek
Wetlands at least along streams
Pete Andrews Creek
>1 unnamed tributary
Wetlands at least along stream
Newhalen River
Unknown extent of wetlands
Jensen Creek
Fish Species'1
Arctic-Alaskan Brook
Lamprey
Longnose Sucker
J
Northern Pike
A
A
Alaska Blackfish
Rainbow Smelt
Humpback Whitefish
A
Round Whitefish
A, J
A
A, J
Unknown Whitefish
Coho Salmon
J
A-Sp,
J
X
A
Chinook Salmon
X
A-Sp
A-Sp
Sockeye Salmon
A-Sp,
J
A-Sp,
J
A-Sp
A-Sp,
J
Chum Salmon
A-Sp
A
Pink Salmon
X
Arctic Char
A
X
X
A
Rainbow Trout
A, J
*
A, J
Dolly Varden
A
A
Lake Trout
A
Arctic Grayling
A
A
A, J
CO
Ninespine Stickleback
A, J
Threespine Stickleback
A, J
Unknown Stickleback
A
A, J
Coastrange Sculpin
Slimy Sculpin
A, J
A, J
Unknown Sculpin
X
Subsistence Usesc
Village
Levelook
Newhalen
New
Stuyahok
Nondalton
Port
Alsworth
Levelock
Target Species/Group
Salmon
X
X
X
X
Other Salmonids
X
X
1
\z.
.c
O
X
X
JB
CO
CO
uo
"55
m
X
CO
(D
X
m
|
CD
X
X
(D
i
X
X
Caribou
X
X
X
X
Other Mammals
X
X
Waterfowl
X
X
X
X
in
(D
+->
O
X
X
Notes:
a "Tributary" indicates the channel flows into the stream listed above it; "stream" indicates the receiving water is off-site, as identified in the columns to the left. For waters downstream of the mine and transportation corridor, fishes and subsistence uses are noted only where different from the waters' prior listing.
b A = adult; Sp = spawning; J = juvenile; x = unknown life stage. "Unknown" indicates an apparent lack of surveys.
c Subsistence uses are not separated by watershed. Uses are noted only for areas of direct impacts (i.e., mine and/or transportation corridor), if they occur there; otherwise, they are noted for areas downstream. "Downstream" applies only to the drainage immediately downstream (e.g., Yellow Creek for the mine, but not the Kvichak River). Data for Dillingham
and Ekwokare less detailed and more dated than for other villages, so use patterns may have changed.
d Hypothetical routing minimizes distance, topographic gradients and stream crossings and assumes water body crossings would be avoided.
e Hypothetical routing follows that shown in the Southwest Alaska Transportation Plan (ADOT 2004), but avoids stream and water body crossings, where possible. Waters and fishes for the transportation corridor to Naknek are limited to those in the Kvichak River watershed (i.e., Coffee Creek and north); subsistence uses do not include those by villages outside
the watershed.
Sources: Levelock Village Council 2005, Fall etal. 2006, Kriegetal. 2009, ADF&G 2012, Johnson and Blanche 2012.
Bristol Bay Assessment
13-17
January 2014
-------
Chapter 13
Cumulative Risks of Multiple Mines
Table 13-7. Waters, fishes, and subsistence uses potentially affected by a mine at the Humble prospect.
c
.0
_l
CO
E
o
Transportation Corridor to
Aleknagik11
Affected Waters
Watersheds and Named Subwatersheds
Nushagak River
X
X
X
X
X
X
X
X
X
X
X
Nuyakuk River
X
X
Napotoli Creek
X
Klutuk Creek
X
Kokwok River
X
X
X
X
X
Kenakuchuk Creek
X
Nameless Creek
X
Koggilung Creek
lowithia River
I
'(£.
I
X
Muklung River
X
Kvichak River
Bear Creek
Waters3
Unnamed stream and >6 tributaries
17 unnamed lakes and ponds
Wetlands at least along main stem
Unnamed stream and >2 tributaries
8 unnamed lakes and ponds
Wetlands at least along streams
Unnamed stream and X3 tributaries
2 unnamed ponds
Unknown wetland extent
Napotoli Creek
>43 unnamed tributaries
48 unnamed lakes and ponds
Wetlands at least along main stem and
primary tributary
Klutuk Creek
>17 unnamed tributaries
Numerous unnamed lakes and ponds
Unknown wetland extent
Kenakuchuk Creek
>3 unnamed tributaries
7 unnamed lakes and ponds
Unknown wetland extent
Unnamed stream
Approximately 1.3 km2 (300 ac) of wetlands
Unnamed stream and >17 tributaries
Numerous unnamed lakes and ponds
Kokwok River
Extensive wetlands
Nameless Creek
X3 tributaries
Extensive wetlands
Muklung River
>± tributary
Extensive wetlands
Fish Speciesb
Arctic-Alaskan
Brook Lamprey
J
Longnose Sucker
Northern Pike
Alaska Blackfish
Rain bow Smelt
Round Whitefish
A, J
Unknown Whitefish
Coho Salmon
J
J
J
A-Sp, J
A-Sp, J
Chinook Salmon
J
A-Sp, J
A-Sp
Sockeye Salmon
J
A-Sp
Chum Salmon
A-Sp
A-Sp
Pink Salmon
Arctic Char
X
X
Rainbow Trout
Dolly Varden
A, J
A, J
Arctic Grayling
A, J
1
3
CO
Ninespine
Stickleback
A, J
A, J
Threespine
Stickleback
Coastrange Sculpin
A
A, J
A
Slimy Sculpin
A
A, J
A, J
c
"5
0
to
c
c
c
A, J
A, J
A, J
A, J
Unknown
Unknown
Unknown
J
A-Sp
J
A-Sp
J
A-Sp
A-Sp
A-Sp
A-Sp
A-Sp
A-Sp
X
X
A, J
A, J
A, J
Unknown
A, J
A, J
A, J
Subsistence Uses0
Village
Dillingham
Ekwok
Koliganek
NewStuyahok
Aleknagik
Ekwok
NewStuyahok
Target Species/Group
Salmon
Other Salmonids
X
Other Fishes
X
-------
Chapter 13
Cumulative Risks of Multiple Mines
Table 13-7. Waters, fishes, and subsistence uses potentially affected by a mine at the Humble prospect.
c
.0
_l
Transportation Corridor to Levelock6
Downstream of Mine(s) and
Transportation Corridor
Affected Waters
Watersheds and Named Subwatersheds
Nushagak River
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Nuyakuk River
X
Napotoli Creek
X
Klutuk Creek
X
Kokwok River
X
X
X
X
X
Kenakuchuk Creek
X
Nameless Creek
Koggilung Creek
X
X
lowithia River
X
X
ir
1
Muklung River
Kvichak River
X
X
Bear Creek
X
X
Waters3
Kokwok River
>2 unnamed tributaries
Extensive wetlands
>10 unnamed streams
Unknown wetland extent
Unnamed stream
Unknown wetland extent
Nushagak River (>1 channel)
>2 unnamed tributaries
Unknown wetland extent
Koggilung Creek
X3 unnamed tributaries
Extensive wetlands
Bear Creek
X3 unnamed tributaries
Unknown wetland extent
Unnamed tributary
Napotoli Creek
Klutuk Creek
Kokwok River
Kenakuchuk Creek
Unnamed tributary
Unnamed tributary
Nushagak River
Koggilung Creek
Lowithia River
Bear Creek
Fish Speciesb
Arctic-Alaskan
Brook Lamprey
Longnose Sucker
Northern Pike
Alaska Blackfish
Rain bow Smelt
Round Whitefish
Unknown Whitefish
Unknown
A, J
A
J
J
J
J
A, J
A, J
A, J
J
A, J
A, J
A, J
A, J
A, J
J
X
Coho Salmon
A-Sp
A-Sp, J
J
A-Sp, J
Chinook Salmon
A-Sp
A-Sp
A-Sp, J
Sockeye Salmon
A-Sp
A-Sp
Chum Salmon
A-Sp
A-Sp
Pink Salmon
A-Sp
A-Sp
Unknown
Unknown
A-Sp
A-Sp
A-Sp
J
A-Sp
A-Sp
J
A-Sp
A-Sp
A-Sp
J
A-Sp
Arctic Char
X
X
X
Rainbow Trout
Dolly Varden
Arctic Grayling
1
3
CO
Ninespine
Stickleback
Threespine
Stickleback
Coastrange Sculpin
Slimy Sculpin
c
"5
0
to
c
c
c
Unknown
A, J
J
J
J
A, J
A, J
J
J
A, J
J
A, J
A, J
A, J
A, J
A, J
A, J
J
A
A, J
J
Unknown
A, J
J
A
A, J
A
A, J
A, J
X
J
J
A, J
J
J
J
J
A-Sp
A-Sp
A, J
A, J
A, J
J
A
A, J
J
A, J
A, J
A, J
Subsistence Uses0
Village
Dillingham
Ekwok
Koliganek
Levelock
NewStuyahok
Aleknagik
Dillingham
Koliganek
Newhalen
Portage Creek
Target Species/Group
Salmon
X
X
X
X
X
Other Salmonids
X
X
X
X
X
Other Fishes
X
X
X
X
-------
This page intentionally left blank
-------
Chapter 13
Cumulative Risks of Multiple Mines
Table 13-8. Streams, water bodies, and wetlands potentially eliminated by additional large-scale mines in the Nushagak and Kvichak River
watersheds.
Mine
Pebble South/PEBd
Big Chunk South6
Big Chunk North
Groundhog
AUDN/lliamna
Humble
TOTALS
Claim Block
Size
(km2)
1,380
142
119
317
183
280
2,420
Mine Area3
(km2)
3.87
3.87
10.7
3.87
10.7
3.87
10.7
10.7
10.7
36.9
57.4
Streams
Density
(km/km2)
1.07
1.18
1.45
1.23
1.19
1.07
Length Eliminated11
(km)
4.1
4.6
12.6
5.6
15.5
4.8
13.2
12.7
11.4
43.2
69.5
Water Bodies
Density
(%)
3.14
6.11
4.18
1.24
6.01
0.66
Area Eliminated
(km2)
0.12
0.24
0.65
0.16
0.45
0.05
0.13
0.64
0.07
1.28
2.06
Wetlands
Density0
(%)
18.3
30.5
6.1
83.5
6.1
83.5
4.2
57.2
4.2
57.2
15.8
17.0
15.8
17.0
57.3
75.3
0.7
9.1
Area Eliminated
(km2)
0.71
1.18
0.24
3.23
0.65
8.93
0.16
2.21
0.45
6.11
0.61
0.66
1.69
1.82
6.13
8.05
0.07
0.97
7.9
27.1
Notes:
a Mine area is based on the Pebble 0.25 scenario and includes footprint of major mine components (mine pit, waste rock piles, and tailings storage facility). Where two values are presented for a mine,
the small value represents the footprint assuming the mine uses an existing tailings storage facility at the Pebble deposit, whereas the larger value represents the footprint assuming the mine uses its
own tailings storage facility.
b Length eliminated = footprint of major mine components x stream density.
c For claim blocks with NWI coverage (i.e., Pebble South/PEB, Groundhog, and AUDN/lliamna), minimum density = NWI wetland density and maximum density = (differential between National Wetlands
Inventory (NWI) wetland density and National Hydrography Dataset (NHD) water body density in area of NWI wetland coverage) x NHD water body density for entire claim block. For claim blocks with
no NWI coverage, minimum density = NHD water body density and maximum density = (maximum differential between NWI wetland density and NHD water body density) x NHD water body density.
d Claim block size for entire PLP/NDM block; water body density includes portion of Iliamna Lake.
e Water body density includes portions of Nikabuna Lakes.
Bristol Bay Assessment
13-21
January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
13.2.2.2 Potentially Affected Waters, Fishes, and Subsistence Uses
Table 13-3 summarizes information on the waters, fish, and subsistence uses potentially affected by a
mine at Big Chunk South prospect. The 142-km2 Big Chunk South claim block is entirely within the
drainage of the Chulitna River, which flows into Lake Clark National Park and Preserve and then into the
lake itself, at Turner Bay, 40 km northeast of the block. A segment of the strongly meandering river,
including the Nikabuna Lakes system, runs along the entire 27-km north boundary of the block. Current
National Wetlands Inventory (NWI) mapping does not include the Big Chunk South claim block;
however, based on aerial photography and U.S. Geological Survey (USGS) topographic mapping,
extensive wetlands appear to be associated with the river, extending upstream along several tributaries.
Stream density at Big Chunk South is 1.18 km/km2, and the majority of streams in the block are
headwater tributaries. Water body density—more than 6% of the block—is the highest of all the
potential mine sites, due in part to the Nikabuna Lakes. The relatively flat valley of the Chulitna River
occupies most of the claim block. Elevations range from approximately 90 m along the river to 320 m in
the more rugged south-central part of the block.
To connect to Cook Inlet, a road and pipelines from the Big Chunk South claim block would likely ascend
the valley of an unnamed Chulitna River tributary and then cross the headwaters of the North Fork
Koktuli River to join the end of the assessment corridor, approximately 14 km south of the block.
To date, very few known fish surveys have been conducted in either the Big Chunk South claim block or
the middle or upper Chulitna River (Table 13-3). Fall etal. (2006) describe the Chulitna River valley as
one of the most important subsistence areas for Nondalton. Among the salmonids included as
assessment endpoints, only Dolly Varden are found in the claim block. One village subsistence fishes and
hunts for bear, moose, caribou, other mammals, and waterfowl in the claim block; four other villages
hunt for caribou; and one other village hunts for moose. As an average across the claim block, stream
loss would range from 4.6 to 12.6 km, and wetland area eliminated would range from 0.24 to 8.9 km2,
depending on whether or not a TSF would be constructed on site (Table 13-8).
13.2.3 Big Chunk North
13.2.3.1 Description
The Big Chunk North prospect is approximately 11 km northwest of the Big Chunk South prospect, 34
km north west of the Pebble deposit, approximately 48 km northwest of Nondalton and 9 6 km northeast
of Koliganek. A mine at Big Chunk North would potentially use TSFs, mill, and other facilities at either
Big Chunk South or Pebble. We consider the impacts of both a mine with shared facilities and one
without (Table 13-8). In both cases, we anticipate that a mine at Big Chunk North would connect to the
assessment corridor.
13.2.3.2 Potentially Affected Waters, Fishes, and Subsistence Uses
Table 13-4 summarizes information on the waters, fishes, and subsistence uses potentially affected by a
mine at Big Chunk North. Like the Pebble deposit, the 119-km2 Big Chunk North claim block straddles
Bristol Bay Assessment 1322 January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
the drainage divide between the Nushagak and Kvichak River watersheds. Northwest of Buck Mountain,
in the northwest corner, the block contains ponds and streams that are part of the headwaters of Reefer
Creek, a tributary of the Mulchatna River, as well as the uppermost reaches of the North Fork Swan
River in the Koktuli River watershed. Nearly 90% of the block, though, is a high-density network of
streams, ponds, and wetlands that form the headwaters of the Chulitna River, which rises immediately
north of the block and flows into Lake Clark National Park and Preserve, and ultimately Lake Clark itself.
The system departs the block along its south boundary. National Hydrography Dataset (NHD) stream
density in the Big Chunk North claim block (1.45 km/km2) is higher than in any of the other areas of
potential mines we consider, and more than one-third higher than in the PLP/NDM claim block (USGS
2012). Water body density (>4%) is also fairly high compared to other mine sites, but lower than at Big
Chunk South and AUDN/Iliamna.
A transportation corridor to service a mine in the Big Chunk North claim block would presumably follow
the Chulitna River valley south and eastward to the Big Chunk South block, where it would link up to the
corridor described in Section 13.2.2.1.
To date, no known fish surveys have been conducted in the Big Chunk North claim block or along its
potential transportation corridor, nor any freshwater fish surveys anywhere in Reefer Creek. Table 13-4
summarizes what little information is available on fish presence in the waters potentially affected by a
mine in this claim block. As in the south block, we do not have data on population sizes. At least one of
the stream systems—the North Fork Swan River—has an abundance of beaver dams, indicating it may
provide important overwintering and rearing habitat (Johnson and Blanche 2012). Four villages hunt in
the claim block but no subsistence fishing is reported (Table 13-4). Across the claim block, stream loss
would range from 5.6 to 15.5 km, and wetland loss would range from 0.16 to 6.1 km2, depending on
whether a TSF would be constructed on site (Table 13-8).
13.2.4 Groundhog
13.2.4.1 Description
The 317-km2 Groundhog claim block abuts the northeast corner of the PLP/NDM block, approximately
10 km west of Nondalton and 20 km north-north west of Iliamna. At present there does not appear to be
a relationship between the Groundhog claim holders and PLP/NDM. Nevertheless, given its proximity to
the potential Pebble facility (approximately 6 km), we consider it both as a separate, stand-alone mine
and as one that shares some facilities associated with potential mine development at the Pebble deposit
(Table 13-8), including the assessment corridor. The current route for such a corridor is approximately
4 km south of the claim block and 13 km from the recent target area for exploration drilling at
Groundhog (AREA 2011). A connector would presumably follow one of the Upper Talarik Creek
tributaries down to the corridor.
13.2.4.2 Potentially Affected Waters, Fishes, and Subsistence Uses
Table 13-5 summarizes information on the waters, fishes, and subsistence uses potentially affected by a
mine at Groundhog. Similar to Big Chunk North, nearly 90% of the Groundhog prospect lies in the
Bristol Bay Assessment ^ „ -3 January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
drainage of the Chulitna River, which passes through the narrow, central part of the block and flows into
the Lake Clark National Park and Preserve and ultimately Lake Clark itself. South of Groundhog
Mountain, the claim includes a number of headwater tributaries to Upper Talarik Creek and the North
Fork Koktuli River. A very small portion of the block—less than 1 km2 in the southeast corner—drains to
tributaries of the Newhalen River, which connects Sixmile Lake, below Lake Clark, to Iliamna Lake. A
road from the claim block to the assessment corridor would presumably follow one of the Upper Talarik
Creek tributaries.
Based on NHD mapping, stream density in the Groundhog block (1.23 km/km2) is the second highest of
those we consider; water body density (>1%) is much lower than at all other sites except Humble
(Table 13-8). The NWI maps extensive wetlands along the Chulitna River and Rock Creek, as well as
along lower Groundhog Creek, the drainage in which the most recent exploration of the prospect has
occurred (AREA 2011, USFWS 2012).
Surveys offish use in waters of the Chulitna River drainage have been limited to date (Table 13-5).
There is more information for the headwater tributaries in the Upper Talarik Creek and North Fork
Koktuli River watersheds in the southern part of the claim block, which both support salmonids. The
Upper Talarik Creek tributary system originates in the same series of lakes and ponds as Groundhog
Creek, in the Chulitna River watershed, at an elevation of approximately 460 m. Two Pacific salmon
species and Dolly Varden have been reported in the claim block, and one village hunts there (Table 13-
5). On average across the claim block, stream loss would range from 4.8 to 13.2 km, and wetland loss
would range from 0.61 to 1.8 km2, depending on whether or not a TSF would be constructed on site
(Table 13-8).
13.2.5 AUDN/lliamna
13.2.5.1 Description
The AUDN/lliamna prospect is approximately 3 5 km west of Iliamna Lake and 90 km south west of the
Pebble deposit. It is in the vicinity of the native villages of Levelock, New Stuyahok, Ekwok, and Igiugig
(Figure 13-1). The bulk of the claims associated with this prospect were newly established in 2012 by
Millrock Alaska, the same company that owns the Humble claims. Millrock began exploration of the
prospect in 2012 and describes it as being part of the porphyry copper-gold belt that includes the
Pebble deposit (Millrock Resources 2011). Their AUDN claims surround others staked earlier and still
held by a partnership that includes TNR Gold Corp., the owner of the Shotgun claims. TNR Gold Corp.
calls their project "Iliamna."
In light of the higher development costs associated with this prospect's distance from other potential
mines, we assume that a future mine at this prospect would require a joint venture involving both claim
blocks and would be self-contained (i.e., it would not share any facilities with other mines) (Table 13-8).
The closest potential port site would be on the lower, tidally influenced reach of the Kvichak River, at or
near Levelock. Naknek, approximately 56 km south of the prospect, already operates as a port and could
be an alternative location, should establishing a port on the Kvichak River prove infeasible.
Bristol Bay Assessment ^ „ -4 January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
Development of a mine at AUDN/Iliamna could trigger the involvement of the Alaska Department of
Transportation and Public Facilities (ADOT) in building the Levelock-to-Naknek portion of the Cook
Inlet-to-Bristol Bay (CIBB) and Dillingham/Bristol Bay (DBB) corridors described in the Southwest
Alaska Transportation Plan (SWATP) (ADOT 2004). The SWATP anticipates that such construction
would not occur within its 20-year planning period; however, it noted that "changing circumstances"
such as "discovery of high-value resource that could potentially be accessed economically through
development of [such roads could] trigger consideration of an earlier implementation" schedule. A mine
at AUDN/Iliamna could be such a trigger, and ADOT involvement in road construction could defray mine
development costs. ADOT is currently investigating such a project (the Ambler Mining District Access, in
northwestern Alaska) as part of the state's "Roads to Resources" program (ADOT 2011a). For similar
reasons, Cook Inlet could be another alternative location for a port, using the Levelock-to-Newhalen
portion of the CIBB and DBB corridors; shipping distances to Canada and the lower 48 states appear to
be shorter from Cook Inlet than from Naknek. For the purposes of this assessment, we consider all three
possible transportation corridors.
13.2.5.2 Potentially Affected Waters, Fishes, and Subsistence Uses
Table 13-6 summarizes information on the waters, fish, and subsistence uses potentially affected by a
mine at AUDN/Iliamna. The 183 km2 AUDN/Iliamna claim block occupies a low, relatively flat area of the
Kvichak River watershed, including a portion of the glacial outwash plain at the western edge of the end
moraine that originally formed Iliamna Lake (Detterman and Reed 1973). The Yellow Creek system
drains approximately 90% of the claim block and flows into the Kvichak River approximately 15 km
upstream of Levelock. Jensen Creek drains the southern 11 km2 of the block and enters the Kvichak
River 10 km above Levelock; some maps incorrectly identify it as Yellow Creek (Levelock Village Council
2005). Both stream systems are strongly meandering. The remainder of the claim flows to a third
tributary system that enters the Kvichak River above Yellow Creek. Residents of Levelock use the
AUDN/Iliamna area most for subsistence, but other villages also fish or hunt in the area or along the
potential transportation corridors (Table 13-6).
Stream density at AUDN/Iliamna is higher than at PLP/NDM and Humble, but lower than at most of the
other blocks. Despite its lack of any large lakes (the largest is less than 0.5 km2), water body density at
AUDN/Iliamna (>6% of the block) is nearly twice that of the PLP/NDM claim block (which encompasses
4.9 km2 of Iliamna Lake) and almost the same as at Big Chunk South (which includes 2.7 km2 of the
Nikabuna Lakes). NWI mapping, which covers almost 70% of the block, also shows extensive wetlands
encompassing approximately 57% of the block. Four species of Pacific salmon are cataloged in
numerous streams on the claim block and seven villages hunt there for a variety of mammals and birds
(Table 13-6). Across the claim block, stream loss would average 12.7 km, and wetland loss would range
from 6.1 to 8.1 km2 (Table 13-8).
The short transportation corridor from AUDN/Iliamna to Levelock would potentially involve only one
stream crossing, of an unnamed tributary to Levelock Creek. To connect to Naknek or Cook Inlet (via the
conceptual CIBB route and the assessment corridor), the road would first have to cross the Kvichak
Bristol Bay Assessment 1325 January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
River at or near Levelock. To reach Naknek, it would also have to cross the Alagnak River and Coffee
Creek, as well as several tributaries. Instead of crossing the Alagnak River, the route to the assessment
corridor would follow it briefly, before crossing Ole and Pecks Creeks, as well as crossing the Kvichak
River at least one more time (at or near Igiugig). From there, it would pass the west and north shores of
Iliamna Lake, crossing Lower and Upper Talarik Creeks, at least 16 other streams, and the Newhalen
River, before joining the assessment corridor near Iliamna.
13.2.6 Humble
13.2.6.1 Description
Accounts characterize the Humble prospect as geologically and geochemically similar to the Pebble
deposit (Szumigala etal. 2011, Millrock Alaska 2012b). It is approximately 135 km south west of the
Pebble deposit, 60 km north west of AUDN/Iliamna, and 20 to 30 km west to northwest of the villages of
Koliganek, New Stuyahok, and Ekwok. Wood-Tikchik State Park, the largest state park in the United
States, is approximately 13 km northwest and 29 km west of the claim block.
Due to Rumble's distance from the Pebble deposit and the AUDN/Iliamna prospect, we do not anticipate
sharing of facilities between the mines (Table 13-8). The Dillingham-Aleknagik Road (46 km to the
southwest) is the closest link to existing road infrastructure and port facilities, the latter of which are
another 40 km further at Dillingham. ADOT is currently pursuing federal permits for construction of a
bridge over the Wood River, which presumably would be the southernmost link in a transportation
corridor from Humble to Aleknagik (ADOT 2011b, USCG 2012). Alternatively, development of a mine at
Humble could trigger the involvement of ADOT in building all or portions of the CIBB and DBB corridors,
which could serve to connect Humble to a port on Cook Inlet. A road from the Humble claim block to the
conceptual DBB route presented in the SWATP would be shorter than a mine road connecting to
Aleknagik (ADOT 2004).
13.2.6.2 Potentially Affected Waters, Fishes, and Subsistence Uses
Table 13-7 summarizes information on the waters, fishes, and subsistence uses potentially affected by a
mine at Humble. The 280-km2 claim block's east- and south-flowing streams are entirely within the
Nushagak River watershed (Figure 13-1), and the mainstem channels are all strongly meandering. More
than 40% of the block drains to Napotoli Creek, which enters the Nushagak River approximately 14 km
upstream of Koliganek. The northernmost 36 km2 of the block flow to one unnamed tributary of the
Nushagak River and two unnamed tributaries of the Nuyakuk River, which connects Tikchik Lake, in
Wood-Tikchik State Park, to the Nushagak River.
In the south, nearly one-quarter of the block is in the Klutuk Creek watershed, which flows into the
Nushagak River immediately downstream of Ekwok. The remainder of the block—approximately 20%—
drains to Kenakuchuk Creek and other tributaries of the Kokwok River, which enters the Nushagak
River 8 km downstream of Ekwok. Stream density in the claim block is approximately the same as at the
Pebble deposit (1.07 km/km2), which is lower than at all of the other sites considered here (Table 13-8).
Bristol Bay Assessment 1326 January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
NHD water body density at Humble is less than 25% of that at PLP/NDM and only slightly more than
half of that at Groundhog, the next most similar prospect (USGS 2 012). A long band of ponds occupies
the divide between Klutuk Creek and the Kokwok River tributaries. Current NWI mapping does not
extend to the Humble block. At minimum there appear to be wetlands along most of the larger stream
corridors (e.g., Naptoli, Klutuk, and Kenakuchuk Creeks) and in an approximately 4-km2 area in the
southwest corner of the block, based on aerial photography and USGS topographic mapping. Four Pacific
salmon species and Dolly Varden are present in many streams in the claim block, and four villages hunt
for a variety of mammals (Table 13-7). On average across the claim block, stream loss would be 11.4 km,
and wetland loss would range from 0.07 to 0.97 km2 (Table 13-8).
A potential route for a transportation corridor from the Humble claim block to Aleknagik that minimizes
distance and topographic gradient would cross the Kokwok River, just downstream of Kenakuchuk
Creek, ascend the valley of Nameless Creek (a Kokwok tributary), skirt Wood-Tikchik State Park, and
then follow the Muklung River (a tributary of the Wood River) before turning north of Marsh Mountain
for Aleknagik. The overland route to Cook Inlet would presumably cross the Kokwok River further
downstream from the route to Aleknagik, connecting to the Dillingham/Bristol Bay corridor near the
lowithla River, another Nushagak River tributary. The SWATP's conceptual corridor would cross the
Nushagak River downstream of Ekwok. After crossing Koggilung Creek, the route would pass into the
Kvichak River watershed, reaching the river at or near Levelock. Section 13.2.5.2 describes the route
from Levelock to the assessment corridor.
Despite its relatively low stream density, a large number offish-bearing streams traverse the Humble
claim block (Table 13-7). Information on local population sizes is not available. The Napotoli and Klutuk
Creek systems contain numerous beaver complexes, as well as frequent seeps and springs, that may
provide important overwintering habitat for juvenile salmonids (Johnson and Blanche 2012: nomination
forms 04-158, 04 160, 04-171, 04-890, 06 753, 06-754,11-369-11-372,11-381,11-382, and 11-384-
11-386). Residents of Aleknagik, Dillingham, Ekwok, Koliganek, and New Stuyahok have historically
used the claim block, the potential transportation corridor area west of Levelock, and/or downstream
areas for subsistence fishing, hunting, and/or gathering (Table 13-7). Information for Aleknagik,
Dillingham, and Ekwok is less detailed than for the other villages. The reports for Dillingham and Ekwok
are more than 20 years old, so their subsistence use patterns may have changed, particularly for caribou
hunting, since that species' population and migration routes have shifted (Brna and Verbrugge 2013).
13.2.7 Potential Impacts of Multiple Mines
In the preceding sections, we examined the waters, fishes, and subsistence uses that could be affected by
the footprints of the major mine components at six prospects that could be developed after initial
development of a large mine at the Pebble deposit. For the purposes of this assessment, we consider the
cumulative impacts of these six mines—that is, potential effects on assessment endpoints resulting from
the establishment of these six additional mines and their associated transportation corridors. These
influences would likely accumulate over time and space, potentially having widespread and extensive
effects on the region's populations offish, wildlife, and human residents.
Bristol Bay Assessment 1327 January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
13.2.7.1 Habitat Eliminated
Table 13-8 summarizes direct losses of aquatic habitat to the footprints of the potential six additional
mines. Total stream length eliminated by these footprints would range from approximately 43 to 70 km,
and total water body and wetland area lost would range from approximately 1.3 to 2.1 km2 and 7.9 to
27 km2, respectively. Loss of these areas to mine footprints would result in extensive losses of
floodplain, riparian habitat, and wetland areas. Waters on these claim blocks include the Chulitna River
and Rock, Jensen, Yellow, Napotoli, Klutuk and Kenakuchuk Creeks, as well as over 250 unnamed
tributaries and over 50 unnamed lakes and ponds. Although not all support salmon, many do.
Elimination of substantial habitat across the watersheds would contribute to diminishing the biological
complexity of salmon stocks and the portfolio effect, which would likely increase annual variability in
the size of Bristol Bay salmon runs (Section 5.2.4).
13.2.7.2 Flow Alteration
Water Withdrawal and Retention
Routine operations at additional mines would also likely degrade or destroy downstream habitat due to
water withdrawal and management of precipitation at the mine facilities (Chapters 6 and 7). Mines
require water for mill operation and transport of tailings and concentrate. The required withdrawal and
retention of surface water and groundwater would effectively reduce the size of the watershed
contributing to downstream flow. Mine pit dewatering would further reduce the contributing watershed
by creating a cone of depression (Section 6.2.2). Streams, wetlands, and ponds within this cone of
depression that receive water through groundwater would dry up, discontinuing any contributions to
downstream waters. Groundwater flow down the valley would also be disrupted, potentially affecting
spawning and wintering habitat downstream (Chapter 7).
Increased Effective Impervious Surface
Research in other areas has shown that even low levels of industrial, commercial, and residential land
uses can cause significant degradation and reduce ecological function in downstream water bodies
(Booth and Jackson 1997). Comparisons between stream condition and level of development have
consistently demonstrated a correlation between stream degradation and watershed imperviousness
(Booth etal. 2002). Greater frequency and intensity of floods, erosion of streambeds, displacement of
sediments, poor water quality, increased water temperature, and reductions in channel and habitat
structure all have been associated with increases in impervious surface (Booth and Jackson 1997, Booth
et al. 2002). In humid regions, approximately 10% effective impervious surface area (i.e., impervious
surface area that is directly connected to stream channels) generally causes demonstrable loss of
aquatic ecosystem function (Booth and Jackson 1997).
Additional mines and potential transportation corridors at each of the prospects would convert
additional land to impervious surface, which could increase total effective impervious surface area
above 10% in several subwatersheds in the region (e.g., Groundhog, Jensen, and Napotoli Creeks and the
Nikabuna Lakes). Mine and road operators and village governments may be able to limit or even
Bristol Bay Assessment 1328 January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
eliminate downstream damage from impervious surface. If operators could moderate discharges to
simulate natural flows and prevent higher than natural stormflows, downstream channels could be
maintained in relatively stable condition. Achieving this goal would require sufficient on-site water
retention capacities to allow a slow meting out of storm-generated flows, rather than contemporaneous
discharge of all surface runoff generated by storm events.
Road Crossings
In addition to increasing runoff rates due to their impervious surfaces, the transportation corridors
associated with additional mines and other induced development would increase the likelihood of flow
regime changes, including channel modification, wherever they crossed streams, wetlands, and other
water bodies (Chapter 10, Appendix G). Such alterations frequently affect salmonids and other fishes by
blocking access to habitat and/or physically degrading the habitat itself. Cumulatively, these additional
transportation corridors would add at least five river crossings (South and North Fork Koktuli, Chulitna,
Kokwok, and Nushagak or Muklung) and a minimum of 27 smaller stream crossings (including at least
one of Upper Talarik Creek) (Tables 13-2 through 13-7). Overland access to AUDN/Iliamna from a port
at Naknek or Cook Inlet (rather than near Levelock) would add at least one and possibly as many as four
crossings of the Kvichak River, at least one of another river (Alagnak or Newhalen), and a minimum of
six at other streams. These crossings would substantially increase the potential for hydraulic alterations
resulting in upstream and downstream habitat degradation (Chapter 10).
13.2.7.3 Water Quality Degradation
Chapters 8 through 11 discuss the potential water quality impacts resulting from water treatment and
discharge, tailings dam failure, road construction and operation, and pipeline spills at a single mine at
the Pebble site. Additional mines and transportation corridors would have these same potential impacts.
Routine Operations
Routine operations at six additional mines would result in approximately 37 to 57 km2 of ground
disturbance (Table 13-8) depending on whether or not TSFs would be built at each of the mines. Rivers
and streams in which water quality could be affected include the Chulitna River and Rock, Jensen,
Yellow, Napotoli, Klutuk, and Kenakuchuk Creeks (Tables 13-2 through 13-7). The transportation
corridors for these mines would potentially span the width of the Nushagak and Kvichak River
watersheds from Cook Inlet to Dillingham and potentially affect the Alagnak, Kvichak, Kokwok, and
Nushagak Rivers and Coffee and Bear Creeks. Salmon are reported in streams in the areas of all of these
prospects except Big Chunk North and Big Chunk South.
Accidents and Failures
Chapters 8 through 11 describe the probabilities and consequences of a variety of accidents and failures
in the mine scenarios. Although the probability of such failures at an individual facility at any given time
is low, the cumulative probability of failures increases as the number of facilities increases. For example,
historical data suggest a greater than 99% cumulative probability of failure in one of the four pipelines
Bristol Bay Assessment 1329 January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
over the life of the Pebble 2.0 scenario (Section 11.1). Additional pipelines at additional mines would
increase the overall probability of failure at some location in the Nushagak and Kvichak River
watersheds each year. Similarly, the chances of a road failure with significant consequences for
downstream waters would be substantial with development of a single mine, and would increase as the
length of road in the watersheds increases (Section 10.3.2).
Although failure of any single TSF dam would be a low-probability event, it could be catastrophically
damaging to fisheries in the receiving waters if it were to occur. The presence of multiple large-scale
mines would increase the probability of at least one TSF dam failure occurring in the watersheds over
the mine lifetimes and post-closure periods roughly in proportion to the total number of dams, and thus
increase the chance of long-term adverse downstream effects. A TSF dam failure at one the additional
mines would likely be similar in nature to the Pebble 0.25 dam failure scenario described in Section 9.3,
although the magnitude of adverse impacts would vary with location, TSF size, and the degree of failure.
Salmon-bearing waters into which slurry from TSF dam failures at additional mine sites could flow
would include the North Fork and mainstem Koktuli Rivers; the Kvichak, Nushagak, Kokwok, and
Chulitna Rivers; Upper Talarik, Rock, Napotoli, Klutuk, and Yellow Creeks; Iliamna, Nikabuna, and Long
Lakes; and Lake Clark.
Another potential source of pollutant discharges results from human errors in characterizing the mining
environment (e.g., its geochemistry or hydrology) and/or anticipating long-term needs for pollutant
control (Box 13-2). Human error, as well as mechanical failure, can also result in water bypassing a
treatment system. Similar unintended failures in human judgment could result in unanticipated
discharges of pollutants from mine sites in the Nushagak and Kvichak River watersheds, particularly as
the number of additional sites increases. The cumulative effect of such incidents in the Nushagak and
Kvichak River watersheds would likely be a steady decline in productivity in these systems as the
affected reaches increase in length and number.
As discussed in Section 14.1.2.5, common mode failures—that is, multiple failures with a common
cause—could result from incidents such as earthquakes or severe storms. The potential for common
mode failures would be compounded if a mining district were created, increasing the chance that a
single severe event could result in multiple failures and adversely affect multiple salmon-bearing waters
at one time.
Bristol Bay Assessment ^ „ -Q January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
BOX 13-2. EXAMPLES OF MINE CHARACTERIZATION ERRORS
Errors in mine site characterization or anticipation of long-term needs for pollutant control can contribute to
mine-related pollutant discharges. Examples at existing mines include the following incidents.
• At the Red Dog Mine in northwest Alaska, treatment of waste rock runoff for metals elevated dissolved
solids in runoff to the point that it had to be directed to the tailings storage facility (TSF) rather than
discharged. Compounding this problem, failure to implement planned surface-water diversions early in
mine development resulted in unpredicted rapid filling of the TSF. Unscheduled discharges from the TSF
were necessary to prevent dam overtopping.
• At the Greens Creek Mine in southeast Alaska, plans included reclamation of the dry stack TSF to prevent
acid drainage. However, mine life exceeded the anticipated timeframe, delaying reclamation and
resulting in acid drainage from the tailings. Anew understanding of site geochemistry indicates that
perpetual water treatment will be necessary even after reclamation, a substantial change from the
original design. Moreover, mine operators have discovered that local wetland chemistry resulted in a
treatment system that redissolves metals before discharge, requiring construction of a new water
treatment facility to address this unanticipated source of pollution.
• Human error resulted in an uncontrolled discharge from a TSF at the Nixon Fork Mine, in interior Alaska,
in January 2012 (see Box 8-1 for a description of events). This unpermitted discharge does not appear to
have reached nearby streams at the time of this writing and may have caused no environmental harm.
The passage of time would be another component influencing cumulative impacts of large-scale mining
in these watersheds. Although time could enable the recovery of streams affected by accidents or
failures, it could also increase the likelihood that a particular accident or failure would occur. We assume
that post-closure site management considerations (Section 6.3) would generally apply to each additional
mine, although the specifics would be based on design and operational assumptions of each mine and
thus differ from site to site. Closure at each mine would typically require hundreds to thousands of years
of monitoring, maintenance, and treatment of any water flowing off-site. Given the magnitude of these
timeframes, we would expect multiple and more frequent system failures in future years. In light of the
relatively ephemeral nature of human institutions over these timeframes, we would expect that
monitoring, maintenance, and treatment would eventually cease, leading to increased release of
contaminated waters downstream.
13.3 Cumulative Impacts from Induced Development
Induced development, or development resulting from the introduction of industry, roads, and
infrastructure associated with a specific activity to a region, is an iterative phenomenon. Opportunities
for employment at mines or in mine-related services would contribute to growth in nearby communities
and would increase demand for housing, community infrastructure, and amenities such as recreation in
and around those localities. Independent of growth due to mine-related employment, improved
accessibility due to road and port construction would reduce the cost of shipping fuel and freight to
areas near such infrastructure. Reduced shipping costs would make construction, business operation,
recreation/tourism, and general cost of living more affordable, which would facilitate increased growth.
Induced development following the advent of large-scale mining would bring welcome economic
opportunities to the region. The potential road systems described in Section 13.2, which could extend
Bristol Bay Assessment ^ „ -^ January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
completely across the Nushagak and Kvichak River watersheds (approximately 250 miles), would be a
major driver of induced development. Currently, access to sites in the Nushagak and Kvichak River
watersheds is by air, boat, snow machine, or foot and is typically facilitated and regulated by the tourism
industry. Transportation corridors associated with large-scale mines likely would increase vehicle
access throughout the two watersheds, thereby increasing both lawful and unlawful unmanaged access
to currently remote sites. Access by all-terrain-vehicles and snow machines would be greatly enhanced
by a road system, and areas in the two watersheds that are essentially never visited by humans would
become accessible. This increased access would extend fishing and hunting pressure and make areas
along any transportation corridor more susceptible to trespassing, poaching, and illegal dumping (ADOT
2001).
13.4 Potential Effects on Assessment Endpoints
13.4.1 Fishes
Based on the general locations of likely development and the fish species present in these areas (Tables
13-2 through 13-7), we can estimate some potential impacts on fish, such as the extent of direct habitat
losses to typical footprints of the major mine components (43 to 70 km of streams), the approximate
number of streams crossed by transportation corridors (5 rivers and 27 streams), and the potential for
additional habitat loss or degradation that would result from accidents or failures (Section 13.2.7.3). By
identifying general areas where development is reasonably foreseeable (Tables 13-2 through 13-7), we
can also, to some extent, consider those losses by habitat type (e.g., headwater streams) and fish species.
In addition to the effects associated with the sheer quantity of lost or degraded habitat, the impacts of
large-scale mining could cumulatively threaten biological complexity of the Nushagak-Kvichak salmonid
stock complex (Section 5.2.4). Impacts on genetically distinct populations of salmon across the
watersheds can reduce biological complexity (i.e., the portfolio effect) and lead to salmon population
declines. As described in this chapter, reasonably foreseeable development during an 80-year timeframe
could span the Nushagak and Kvichak River watersheds, encompassing many geographically and
hydrologically distinct waters. For anadromous fish, the potentially affected waters would include at
least 10 different rivers and 20 feeder stream systems. As described by Schindler etal. (2010), each
river stock includes tens to hundreds of locally adapted populations distributed among tributaries and
lakes. Given the extent of stream losses and habitat degradation, it is reasonable to assume that losses of
genetic and life-history diversity would occur with the development of multiple large-scale mines. Even
non-anadromous fish would be subject to genetic diversity losses wherever development blocked
movement to different spawning, rearing, and overwintering habitats within a stream and/or stream-
lake system, thereby isolating portions of the population (Appendix B) (Charles et al. 2000).
Although we can estimate potential habitat losses and anticipate resulting losses of population diversity,
we cannot quantify specific fish losses or their significance in terms of overall populations given the
Bristol Bay Assessment ^ „ 22 January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
current lack of abundance data and uncertainty about the precise locations and magnitude of future
developments.
13.4.2 Wildlife and Alaska Native Culture
As the extent of development in the Nushagak and Kvichak River watersheds increased, so would
development-related effects on wildlife and on Alaska Native culture. The six additional mines would
affect a wide range of wildlife, including both resident and highly migratory species (Tables 13-2
through 13-7). As with fish, data are insufficient to predict wildlife population impacts.
As for Alaska Natives, 13 of the 14 villages in the watersheds would experience some impact on
traditional subsistence use areas from additional development (Figure 13-3). Levelock and Igiugig
would experience the most impacts, in that additional development would have direct and/or indirect
impacts on all categories of their subsistence resources (due primarily to proximity of the potential
AUDN/Iliamna mine and a transportation corridor from that mine to Newhalen). Iliamna, Koliganek,
Newhalen, New Stuyahok, and Nondalton would also have a large number of subsistence resource
categories affected.
Development of a mining district could have broader cultural impacts related to the diminishing role of
subsistence in village life, via any reductions in the areas or fish and wildlife populations available for
subsistence activities (Chapter 12, Appendix D). Employment opportunities and reduced cost of living
resulting from large-scale mining could address some current concerns associated with subsistence,
chiefly the cost of fuel needed to access these resources. At the same time, loss of subsistence areas or
populations would likely be accompanied by increased westernization resulting from increased access,
development, tourism, and prosperity. Both of these factors could erode the current subsistence
cultures, at least to some extent.
13.5 Summary
The industrial complex and transportation corridor associated with potential mine development at the
Pebble deposit would constitute the second largest human population center in the Bristol Bay region
and its longest road system. Figure 13-2 illustrates how cumulative impacts from an initial mine,
multiple subsequent mines, and induced development could result from the introduction of large-scale
mining in the region.
Bristol Bay Assessment ^ „ „„ January 2014
-------
Chapter 13
Cumulative Risks of Multiple Mines
Figure 13-3. Location of claim blocks in relation to subsistence use intensity for salmon, other
fishes, wildlife, and waterfowl in the Nushagak and Kvichak River watersheds. See Box 5-2 for
discussion of subsistence use methodology.
Cook Inlet
Manokotak
Clark's Point
South Naknek
Bristol Bay
N
A
25
25
50
] Kilometers
50
] Miles
« Approximate Pebble Deposit Location
| | Watershed Boundary
Existing Roads
Nonsurveyed Towns and Villages
* Surveyed Towns and Villages
Active Mining Claims
Subsistence Use Intensity
High
Low
Bristol Bay Assessment
13-34
January 2014
-------
Chapter 13 Cumulative Risks of Multiple Mines
Infrastructure (i.e., roads, airports, ports, and utilities) is virtually absent from the mineralized areas of
the Nushagak and Kvichak River watersheds. The transportation corridor, deep-water port, power
generation facilities, and other infrastructure that would be required for mining the Pebble deposit
likely would increase the economic feasibility of developing and operating other, smaller mines. Over
the approximately 25-year life of the Pebble 2.0 scenario mine, and certainly over the approximately 78-
year life of the Pebble 6.5 scenario mine, it is reasonably foreseeable that a number of prospects recently
under active exploration (Pebble South/PEB, Big Chunk South, Big Chunk North, Groundhog,
AUDN/Iliamna, and Humble) could be developed. Each of the mines would need road access to a port
facility and airport. Some of the additional mines could use the assessment corridor, with only relatively
short extensions, but more distant sites such as AUDN/Iliamna and Humble would require development
of extensive additional pipeline, road, or railroad systems. The mines and transportation corridors
described herein are not certain, but the roads are part of state planning documents—and a large-scale
mine could easily be the trigger that starts this pattern of development in motion.
Mines at these sites would cause their own direct impacts, which would accumulate over a much greater
portion of the Nushagak and Kvichak River watersheds and increase the number of distinct salmon
populations affected. These effects could cumulatively threaten the biological complexity of the
Nushagak-Kvichak salmon stocks and the portfolio effect, potentially contributing to salmon population
declines. The genetic and life-history diversity within and among Bristol Bay salmon stocks will likely be
critical for maintaining the resiliency of the population under a future environment characterized by
climate change. Thus, the potential effects of additional mines on salmon genetic diversity could
exacerbate the impacts of climate change on salmon populations in the watershed.
It is reasonably foreseeable that the infrastructure, particularly the transportation corridors, associated
with large-scale mining could induce further development in the region. Existing communities, the
tourism industry, and the recreational housing market could benefit if large-scale mining expanded
through the watersheds. Unmanaged access to currently roadless wilderness areas also could expand.
Improved access would increase hunting and fishing pressure, as well as competition with existing
subsistence users; increase damage from off-road vehicle, boat, and foot traffic in currently inaccessible
areas; facilitate poaching, dumping, trespassing, and other illegal activities; and lead to scattered
development in the watersheds.
Bristol Bay Assessment ^ „ -c January 2014
-------
This chapter summarizes the risk analysis results, organized by assessment endpoint, for a potential
mine at the Pebble deposit. For each endpoint, it integrates the various sources of risk, including those
from routine operations and accidents and failures, different physical and chemical exposures, and
different pathways of exposure and mechanisms of effects. In addition, it combines multiple types of
evidence, including evidence from analysis of the mine scenarios and from knowledge of analogous
mining operations. Limitations and uncertainties in the risk characterization are also summarized.
Finally, these results are extrapolated to the cumulative effects of multiple mines. See Chapters 7
through 13 for the derivation of these conclusions.
14.1 Overall Risk to Salmon and Other Fishes
14.1.1 Routine Operation
During routine operations, mining would be conducted according to modern conventional practices,
including common mitigation measures at the mine site and along the transportation corridor. Toxic
effects would be minimized by collection of nearly all water from the site and treatment of collected
water to meet state standards and national criteria before discharge. However, toxic effects would still
occur, primarily due to the inevitable leakage of leachates. In addition, habitat loss and modification
would occur due to destruction of streams and wetlands and water withdrawals. As a result, local
populations of salmonids would decline in abundance and production. Compensatory mitigation of these
losses in the Bristol Bay watershed would be problematic at best (Appendix J).
14.1.1.1 Mine Footprint
Even in the absence of accidents or failures, the development of a mine at the Pebble deposit would
result in the destruction or modification of streams, wetlands, and ponds. Local habitat loss would be
Bristol Bay Assessment
14-1
January 2014
-------
Chapter 14 Integrated Risk Characterization
significant, because losses of stream habitat leading to losses of local, unique populations would erode
the population diversity key to the stability of the overall Bristol Bay salmon fishery (Schindler et al.
2010).
• In the Pebble 0.25, 2.0, and 6.5 scenarios, 38, 89, and 151 km of streams, respectively, would be
lost to (eliminated, blocked, or dewatered by) each mine footprint (the area covered by the mine pit,
waste rock piles, tailings storage facilities [TSFs], drawdown zone, and plant and ancillary facilities).
This translates to losses of 8, 22, and 36 km of streams known to provide spawning or rearing
habitats for coho salmon, sockeye salmon, Chinook salmon, and Dolly Varden.
• Altered streamflow resulting from retention and discharge of water used in mine operations, ore
processing, transport, and other processes would reduce the amount and quality offish habitat.
Streamflow alterations exceeding 20% would adversely affect habitat in an additional 15, 27, and
53 km of streams in the Pebble 0.25, 2.0, and 6.5 scenarios, respectively, reducing production of
coho salmon, sockeye salmon, Chinook salmon, rainbow trout, and Dolly Varden. Reduced
streamflows would also result in the loss or alteration of an unquantified area of riparian floodplain
wetland habitat due to loss of hydrologic connectivity with streams.
• Off-channel habitats for salmon and other fishes would be reduced due to losses of 4.5,12, and
18 km2 of wetlands and 0.41, 0.93, and 1.8 km2 of ponds and lakes to the Pebble 0.25, 2.0, and 6.5
mine footprints, respectively. These losses would reduce availability of and access to hydraulically
and thermally diverse habitats that provide foraging opportunities and important rearing habitats
for juvenile salmon.
• Indirect effects of stream and wetland losses would include reductions in the quality of
downstream habitat in the three headwater streams draining the mine footprints, affecting the same
species as the direct effects. Modes of action for these effects would include the following.
o A reduction in food resources would result from the loss of organic material and drifting
invertebrates exported from the 38 to 151 km of streams lost to the mine footprints.
o The balance of surface water and groundwater inputs to downstream reaches would change.
Shifting from groundwater to surface-water sources is expected to reduce winter habitat (i.e.,
unfrozen stream reaches) and make streams less suitable for spawning and rearing.
o Water treatment and discharge, resulting in reduced passage through groundwater flowpaths,
are expected to alter summer and winter water temperatures and make streams less suitable for
Pacific salmon, rainbow trout, and Dolly Varden.
These indirect effects on the abundance and production of salmonids cannot be quantified due to lack of
data. However, it is expected that one or more of these mechanisms would diminish fish production
downstream of the mine footprints in each watershed.
Bristol Bay Assessment 14,2 January 2014
-------
Chapter 14 Integrated Risk Characterization
14.1.1.2 Water Collection, Treatment, and Discharge
Water in contact with tailings, waste rock, or the pit walls would leach copper and other metals. Our
assessment evaluates the discharge of treated wastewater and the realistic expectation that leachate
would escape the waste rock pile and TSF water collection systems in the three mine size scenarios.
Routine discharges from the wastewater treatment plant (WWTP) to the South and North Fork Koktuli
Rivers should be non-toxic due to treatment to achieve permit requirements. However, they may be
somewhat toxic due to combined effects of multiple chemicals, poorly known and unregulated
contaminants, and untested species in the receiving waters.
The retention and collection of leachates are inevitably incomplete. In our routine operations scenario,
leakage in the Pebble 2.0 and Pebble 6.5 scenarios would be sufficient to cause toxic levels of copper
and, to a much lesser extent, other metals in the streams draining the mine footprints. The most severe
effects, including death of salmonids, would occur in the South Fork Koktuli River, which would receive
leachate from the acid-generating waste rock. Upper Talarik Creek would experience death of
invertebrates only below the station at which it receives interbasin transfer from the South Fork Koktuli
River. The North Fork Koktuli River would experience death of invertebrates below TSF 1. Death or
inhibited reproduction of aquatic invertebrates, which are food for fish, is estimated to occur in 21, 40 to
62, and 60 to 82 km of streams in the Pebble 0.25, 2.0, and 6.5 scenarios, respectively. Avoidance of
streams by salmonids would occur in 24 and 34 to 57 km of streams in the Pebble 2.0 and Pebble 6.5
scenarios, respectively. Death or reduced reproduction of salmonids would occur in 3.8 and 12 km of
streams in the Pebble 2.0 and Pebble 6.5 scenarios, respectively.
The magnitude and extent of these predicted effects suggest the need for mitigation measures beyond
the conventional practices assumed in the routine operations scenario to reduce the input of leached
copper and other metals. A design based on conventional practices may be sufficient for a typical
porphyry copper mine (i.e., equivalent to the Pebble 0.25 scenario), but not the massive Pebble 2.0 and
6.5 scenarios. Simply improving the efficiency of the capture wells or adding a larger wall or trench is
unlikely to achieve water quality criteria in those scenarios. Additional measures might include lining
the waste rock piles, reconfiguring the piles, or processing the acid-generating waste rock as it is
produced.
In the event of TSF 1 overfilling, supernatant water would be released via a spillway. If the water was
equivalent to the test tailings supernatant, 2.6 km of stream would be avoided by fish and 3.4 to 23 km
would be toxic to invertebrates, independent of other sources.
14.1.1.3 Road Construction and Operation
The assessment's transportation corridor, including a road and four pipelines, would cross
approximately 64 streams and rivers, of which 55 are known or likely to support migrating and resident
salmonids. Nearly 272 km of streams between the road and Iliamna Lake would be affected. Risks to
salmonids from the construction and operation of the transportation corridor are as follows.
• Loss and alteration of habitat through filling of wetlands for the road.
Bristol Bay Assessment 14,3 January 2014
-------
Chapter 14 Integrated Risk Characterization
• Increased suspended and deposited sediment washed from the road, shoulders, ditches, cuts, and
fills.
• Increased stormwater runoff leading to increased suspended sediment, fine-bed sediment, salts,
and, at the mine site, metals.
• Increased dust leading to a direct increase in fine-bed sediment in the mining area, and an indirect
increase along the entire transportation corridor via reduced riparian vegetation.
• Possible introduction of invasive species, particularly plants and fish pathogens.
All of the above sources and stressors would likely lead to degraded or reduced habitat for salmon and
other fish.
14.1.2 Accidents and Failures
Any complex activity such as the mine described in this assessment inevitably experiences accidents and
failures. The number of ways in which failures and accidents can occur—their magnitudes, their
locations, and the circumstances of their occurrence—are effectively infinite. Hence, a complete and
specific assessment of risks from potential accidents and failures is not possible. Rather, a few failure
scenarios are presented, which emphasize the consequences of failures rather than the means by which
they are initiated. These scenarios address potential failures that could occur during mine operations or
after mine closure in perpetuity: failure to treat contaminated water; tailings dam failure; failures of
roads and culverts; wreck of a truck carrying a process chemical; and failure of diesel, product
concentrate, or return water pipelines. Many other potential failures are not analyzed, including failures
of the on-site pipelines, spills of ore-processing chemicals on site, failures of tailings dams on streams
other than the North Fork Koktuli River, wildfires, waste rock slides, or failures at the port.
The probabilities and consequences of the failures analyzed in the assessment are summarized in Table
14-1. The derivation of these estimates is discussed in Box 14-1, and the interpretation of failure
probabilities is discussed in Box 9-3. Probabilities of occurrence were estimated using the best available
information. Some estimates are qualitative, because no applicable data are available. Those that are
quantitative are somewhat uncertain and their interpretation is not straightforward. For example, the
range of annual probabilities of a tailings dam failure is based on design expectations rather than actual
performance data, which are unavailable for recently constructed large earthen dams. The actual
observed frequency of tailings dam failures is near the upper end of that range, which suggests that the
range is reasonable at that bound. However, the lower bound (1 in 250,000 per year) is purely
aspirational, in that it has no empirical basis.
Bristol Bay Assessment 14,4 January 2014
-------
Chapter 14
Integrated Risk Characterization
Table 14-1. Probabilities and consequences of potential failures in the mine scenarios.
Failure Type
Tailings dam
Product concentrate pipeline
Concentrate spill into a stream
Concentrate spill into a wetland
Return water pipeline spill
Diesel pipeline spill
Culvert, operation
Culvert, post-operation
Truck accidents
Water collection and treatment,
operation
Tailings storage facility spillway
release
Water collection and treatment,
managed post-closure
Water collection and treatment,
after site abandonment
Probability3
4 x 1O4 to 4 x 1O6 per dam-year =
recurrence frequency of 2,500 to
250,000yearsb
10'3 per km-year = 95% chance
per pipeline in 25 years
1.5 x 1Q-2 per year = 1 stream-
contaminating spill in 78 years
2.6 x lO-2 per year = 2 wetland-
contaminating spills in 78 years
Same as product concentrate
pipeline
Same as product concentrate
pipeline
Low
3 x 10-1 to ~6 x 10-1 per culvert;
instantaneous = 11 to 22 culverts
1.9 x 10-7 spills per mile of travel =
4 accidents in 25 years and 2
near-stream spills in 78 years
0.93 = proportion of recent U.S.
porphyry copper mines with
reportable water collection and
treatment failures
No data, but spills are known to
occur and are sufficiently frequent
to justify routine spillway
construction
Somewhat higher than operation
Certain, by definition
Consequences
More than 29 km of salmonid stream would be
destroyed or degraded for decades.
Most failures would occur between stream or wetland
crossing and might have little effect on fish.
Fish and invertebrates would experience acute
exposure to toxic water and chronic exposure to toxic
sediment in a stream and potentially extending to
Iliamna Lake.
Invertebrates and potentially fish would experience
acute exposure to toxic water and chronic exposure to
toxic sediment in a pond or other wetland.
Fish and invertebrates would experience acute
exposure to toxic water if return water spilled to a
stream or wetland.
Acute toxicity would reduce the abundance and
diversity of invertebrates and possibly cause a fish kill
if diesel spilled to a stream or wetland.
Frequent inspections and regular maintenance would
result in few impassable culverts, butfor those few,
blockage of migration could persist for a migration
period, particularly for juvenile fish.
In surveys of road culverts, 30 to 61% are impassable
to fish at any one time. This would result in 11 to 22
salmonid streams blocked at any one time. In 10 to 19
of the 32 culverted streams with restricted upstream
habitat, salmon spawning may fail or be reduced and
the streams would likely not be able to support long-
term populations of resident species.
Accidents that spill processing chemicals into a stream
or wetland could cause a fish kill. A spill of
molybdenum concentrate may also be toxic.
Water collection and treatment failures could result in
exceedance of standards potentially including death of
fish and invertebrates. However, these failures would
not necessarily be as severe or extensive as estimated
in the failure scenario, which would result in toxic
effects from copper in more than 60 km of stream
habitat.
Spilled supernatant from the tailings storage facility
could result in toxicity to invertebrates and fish
avoidance for the duration of the event.
Post-closure collection and treatment failures are very
likely to result in release of untreated or incompletely
treated leachates for days to months, but the water
would be less toxic due to elimination of potentially
acid-generating waste rock.
When water is no longer managed, untreated
leachates would flow to the streams. However, the
water may be less toxic.
a Because of differences in derivation, the probabilities are not directly comparable.
b Based on expected state safety requirements. Observed failure rates for earthen dams are higher (about 5 x 10 4 per year or a recurrence
frequency of 2,000 years).
Bristol Bay Assessment
14-5
January 2014
-------
Chapter 14 Integrated Risk Characterization
BOX 14-1. FAILURE PROBABILITIES
Table 14-1 presents probability estimates and consequences of different kinds of failures. Here, we explain the
derivation of these estimates. As much as possible, multiple methods are used within a failure type to determine
how robust the estimates may be. The methods differ among failure types and the results are not strictly
equivalent, but they do convey the likelihood of occurrence. More details can be found in Chapters 8 through 11.
Tailings dam failure. The most straightforward method of estimating the annual probability of failure of a tailings
dam is to use the failure rates of existing dams. Three reviews of earthen dam failures produced an average rate
of 1 failure per 2,000 dam-years (i.e., a recurrence frequency of 2,000 years), or 5 x 1Q-4 per year. The argument
against this approach is that it does not reflect current engineering practice. The State of Alaska's guidelines
suggest that an applicant follow accepted industry design practices such as those provided by U.S. Army Corps of
Engineers and the Federal Energy Regulatory Commission. Both regulatory agencies require a minimum factor of
safety of 1.5 for the loading condition corresponding to steady seepage at the maximum storage facility. An
assessment of the correlation of dam failure probabilities with safety factors against slope instability suggests an
annual probability of failure of 1 in 1,000,000 years for Category I Facilities (those designed, built, and operated
with state-of-the-practice engineering) and 1 in 10,000 years for Category II Facilities (those designed, built, and
operated using standard engineering practice). This corresponds to risks of 10'4to 10'6 per year. The advantage of
this approach is that it addresses current regulatory expectations and engineering practices. The disadvantage is
that we do not know whether standard practice or state-of-the-practice dams designed with safety factors would
perform as expected. Slope instability is only one type of failure; other failure modes, such as overtopping during a
flood, would increase overall failure rates. Slope stability failures account for about one-fourth of tailings dam
failures, so the probability of failure from all causes could be estimated to be 1 in 250,000 (Category I) to 1 in
2,500 (Category II). The mine scenarios include up to three tailings storage facilities (TSFs), two with multiple
dams, so the annual probability of any dam failing would be approximately equal to the annual probability of a
single dam failure times the number of dams.
Pipeline failure. A review of observed pipeline failure rates for oil and gas pipelines yields an average annual
probability of failure per kilometer of pipeline of 10'3or a frequency of 1 failure per 1,000 km per year. This
average risk comes very close to estimating the observed failure rate of the copper concentrate pipeline at the
Minera Alumbrera mine, Argentina. This annual failure probability, over the 113-km length of each pipeline within
the Kvichak River watershed, results in a 0.11 probability of a failure in each of the four pipelines each year, or a
recurrence frequency of 8.5 years. If the probability of a failure is independent of location, and if it is assumed
that spills within 100 m of a stream could flow to that stream, a spill would have a 0.14 probability of entering a
stream within the Kvichak River watershed. This would result in an estimate of 0.015 stream-contaminating spills
per year or 1 stream-contaminating spill over the duration of the Pebble 6.5 scenario (approximately 78 years).
Similarly, a spill would have a 0.24 probability of entering a wetland, resulting in an estimate of 0.026 wetland-
contaminating spills per year or 2 wetland-contaminating spills over the duration of the Pebble 6.5 scenario.
Water collection and treatment failure. During mine operation, collection or treatment of leachate from mine
tailings, pit walls, or waste rock piles would be incomplete and could fail in various ways. In the routine operations
scenario, leachate from the unlined TSFs and waste rock piles would not be fully collected. Equipment and
operation failures and inadequate designs would also result in failures to avoid toxic emissions. Reviews of mine
records found that 93% of operating porphyry copper mines in the United States reported a water collection or
treatment failure (Earthworks 2012). Improved design and practices should result in lower failure rates, but given
this record it is unlikely that failure rates would be lower than 10% over the life of a mine. During operation,
failures should be brief (less than 1 week) unless they involve a faulty system design or parts that are difficult to
replace. After a mine is abandoned (potentially many years after closure), water management would end and the
discharge of untreated water would become inevitable but may not be problematic.
TSF spillway release. Releases of supernatant water from TSFs through spillways are unintended but are not
uncommon (e.g., the release at Nixon Fork Mine described in Box 8-1). However, data on the frequency of such
releases are unavailable. They are apparently sufficiently common that inclusion of a spillway in a tailings dam is a
standard practice. Hence, it is judged likely that a release would occur over the 78-year life of the mine in the
Pebble 6.5 scenario.
Culvert failure. Culvert failure is defined as a condition that blocks fish passage. Empirical data for culvert failures
are not based on rates of failure of culverts but rather on instantaneous frequencies of culverts that were found to
have failed in road surveys. The frequencies in recent surveys range from 0.30 to 0.61 (3 to 6 x 10'1) per culvert.
In the Kvichak River watershed, 35 streams that are believed to support salmonids (salmon, trout, or Dolly
Varden) have culverts, so at anytime 11 to 22 culverted streams would be expected to have blocked fish passage
at the published frequencies. The proportion of failed culverts during mine operation should be much lower.
Bristol Bay Assessment 14,6 January 2014
-------
Chapter 14 Integrated Risk Characterization
It is important to remember that this is an assessment of mine scenarios. It is based on modern
conventional mining practices, especially the plan proposed for the Pebble site by Northern Dynasty
Minerals (Ghaffari et al. 2011). However, like any predictive assessment, it is hypothetical. Although the
major features of the scenarios will undoubtedly be correct (e.g., a pit at the location of the ore body,
waste rock deposited near the pit, the generation of a large volume of tailings), some specifics would
inevitably differ. This would be true of any scenario, including a mining plan submitted for permitting or
even a plan approved by the state. All plans are scenarios, and although each new plan is expected to be
closer to actual operations than the ones before, unforeseen circumstances and events and new
technologies inevitably compel changes in practice.
14.1.2.1 Tailings Dam Failure
Failure of a tailings dam would have a one in 2,500 to one in 250,000 probability of occurrence per year
for each TSF. Probability of a tailings dam failure increases with an increase in the number of dams. The
Pebble 0.25 and Pebble 2.0 scenarios include one TSF, and the Pebble 6.5 scenario includes three. Two of
these TSFs would have multiple dams. However, the probability of a spill from these TSFs would not
increase in proportion to the number of dams for an individual TSF, because failures would not be
independent events. The failure of one dam on a TSF would relieve pressure on others, reducing the
probability of multiple failures; conversely, common mode failures could occur, increasing the
probability of multiple failures. The dam failure analyses in this assessment simulated the release of
20% of the tailings (a conservative estimate) from the failure of a 92-m (Pebble 0.25) and a 209-m
(Pebble 2.0) dam at TSF 1.
Failure of the TSF 1 dam would result in the release of a flood of tailings slurry into the North Fork
Koktuli River, scouring the valley and depositing tailings. The complete loss of suitable salmonid habitat
in the North Fork Koktuli River (29 km of habitat in the Pebble 0.25 scenario and more than 30 km, our
model limit, in the Pebble 2.0 scenario) in the short-term (less than 10 years). The high likelihood of
very low-quality spawning and rearing habitat in the long-term (decades) would result in the nearly
complete loss of mainstem North Fork Koktuli River fish populations below the dam. Even salmon at sea
during the failure would not find suitable spawning habitat on their return to the North Fork Koktuli
River as adults. The river currently supports spawning and rearing populations of sockeye, Chinook, and
coho salmon, spawning populations of chum salmon, and rearing populations of Dolly Varden and
rainbow trout. Suspended mine tailings sediments would continue for an unknown (due to model and
data limitations) distance farther down the Koktuli River, and probably into the Mulchatna and
Nushagak Rivers, causing degraded spawning habitat and reduced food resources. Fish anywhere in the
flowpath below a tailings dam failure would be killed or forced downstream. Fish migrating into
tributaries of affected rivers would be inhibited from migration for some period of time, which our
model could not predict.
Following the slurry flood, deposited tailings would continue to erode from the North Fork Koktuli and
Koktuli River valleys. After many years, a new channel with gravel substrate and a natural floodplain
structure would become established. However, that recovery would come at the expense of the
Bristol Bay Assessment 14,7 January 2014
-------
Chapter 14 Integrated Risk Characterization
downstream Mulchatna and Nushagak Rivers, as much of the spilled tailings initially deposited in the
North Fork Koktuli and Koktuli Rivers would be resuspended by erosion and transported down the
drainage. This process could not be modeled with existing data and resources, but would be inevitable if
a tailings spill occurred.
High concentrations of suspended tailings would occur following a tailings dam failure, but over time
they would decline as erosion progressed. For some years, periods of high streamflow would be
expected to suspend sufficient concentrations of tailings to cause avoidance, reduced growth and
fecundity, and possibly even death offish. Migration to and from any affected tributaries would be
impeded if streamflow from the tributaries was not sufficient to adequately dilute suspended sediment
concentrations, meaning that fish would not reach spawning grounds, winter refugia, or seasonal
feeding habitats.
Deposited tailings would degrade habitat quality for both fish and the invertebrates they eat. Pacific
salmon, Dolly Varden, and rainbow trout spawn in gravels, and their eggs and larvae require sufficient
space within the gravel for water to circulate. Juvenile salmonids require even larger clear spaces for
concealment from predators and for overwintering habitat. Tailings would fill those interstitial spaces.
An increase in fines of more than 5% causes detectable effects on salmonid reproduction. Until
considerable erosion occurred and a gravel-bedded channel was re-established, female salmonids would
be unable to clean the gravel to spawn. Even where gravel was available, high deposition from upstream
erosion of tailings could smother eggs and larvae. Recovery of suitable substrates via mobilization and
transport of tailings fines would take decades, and would affect much of the watershed downstream of
the failed dam.
In addition to degrading fish habitat, deposited tailings would be potentially toxic. Based largely on their
copper content, deposited tailings would be toxic to benthic macroinvertebrates, although existing data
concerning fish toxicity is less clear. Estimated pore water concentrations are below published
thresholds for chronic effects in fish, but directly relevant tests of salmonid early life stages have not
been conducted. The combined effects of copper toxicity and poor habitat quality (particularly low
dissolved oxygen concentrations) caused by fine sediment are unknown. Dietary exposures of salmonids
via invertebrate prey exposed to tailings are estimated to be marginally toxic.
In sum, a TSF 1 dam failure would have severe direct and indirect effects on aquatic resources, and
specifically on salmonids. In the short-term (less than 10 years), certainly the North Fork Koktuli River
below the TSF 1 dam failure location and very likely much of the Koktuli River would not support
salmonids. For a period of decades, those waters would provide very low-quality spawning and rearing
habitat, likely resulting in the nearly complete loss of North Fork Koktuli River fish populations.
Deposition, resuspension, and redeposition of tailings would likely cause serious habitat degradation in
the Koktuli River and downstream into the Mulchatna River. Ultimately, spring floods and stormflows
would carry some portion of the tailings into the Nushagak River. Effects would be qualitatively the
same for both the Pebble 0.25 and Pebble 2.0 dam failures, although effects from the Pebble 2.0 dam
failure would extend farther and last longer.
Bristol Bay Assessment 14,8 January 2014
-------
Chapter 14 Integrated Risk Characterization
The Koktuli River watershed is an important producer of Chinook salmon for the larger Nushagak
Management Zone. The Nushagak River watershed is the largest producer of Chinook salmon in the
Bristol Bay region, with an average annual escapement of nearly 190,000 Chinook salmon from 2002
through 2011 (Buck et al. 2012). Assuming Alaska Department of Fish and Game aerial survey counts
(Dye and Schwanke 2009) reflect the proportional distribution of Chinook salmon within the Nushagak
River watershed, the tailings dam failure would eliminate 29% of that run due to loss of the Koktuli
River salmon population; an additional 10 to 20% could be lost because tailings deposited in the
Mulchatna River would affect its tributaries. Sockeye salmon are the most abundant salmon returning to
the Nushagak River watershed, with annual runs averaging more than 1.9 million fish. However, the
proportion of sockeye and other salmon species that originates in the Koktuli and Mulchatna River
watersheds is unknown. Similarly, populations of rainbow trout and Dolly Varden of unknown size
would be lost for decades.
Remediation of a tailings spill would be difficult and problematic. The affected area is roadless, and the
rivers are too small to float a dredge. If the spill occurred after mine closure, people and equipment to
repair the dam and begin remediation would be absent. Remediation may be slow to start due to the
need to develop a plan, create a facility to receive the recovered tailings, build roads, and bring in
personnel and equipment. Even in the Pebble 0.25 dam failure, complete removal of this material would
require a substantial earth-moving effort, including over 3 million round trips by 20-ton dump trucks.
Dredging tailings from rivers and streams would cause considerable habitat damage.
The dam failures evaluated in the assessment used TSF 1 as a plausible location. Failure of the other
tailings dams at TSF 2 and TSF 3 were not modeled, but would have similar types of effects in the South
Fork Koktuli River and downstream.
14.1.2.2 Wastewater Treatment Plant Failure
In the WWTP failure scenario, untreated wastewater would be discharged. The most severe effects,
including lethality to invertebrates and fish, would occur in the South Fork Koktuli River where
untreated effluent would mix with toxic waste rock leachate. The North Fork Koktuli River, where the
untreated waste would mix with tailings leachate, would experience lethality to invertebrates and,
depending on the season, reduced growth or survival of early fish life stages. In this scenario, Upper
Talarik Creek would receive no wastewater discharge and would experience no additional effects. The
WWTP failure is estimated to result in lethality or reduced reproduction of invertebrates in 78 to 100
km of streams in all three mine sizes. For salmonids, it is estimated to cause avoidance of 74 to 97 km of
streams, sensory inhibition in 70 to 92 km, reduced reproduction in 61 to 84 km, and mortality in 31 km
in the Pebble 6.5 scenario. Direct effects on fish would be less extensive in the Pebble 2.0 scenario, with
avoidance in 64 to 87 km, sensory inhibition in 27 km, reduced reproduction in 11 km, and kills in 3.8
km—and would be limited to avoidance in 27 km of streams in the Pebble 0.25 scenario.
Bristol Bay Assessment 14,9 January 2014
-------
Chapter 14 Integrated Risk Characterization
14.1.2.3 Culvert Failure
The most likely serious failure associated with the potential transportation corridor would be blockage
or failure of culverts. Culverts commonly fail to allow fish passage. They can become blocked by debris
or ice that may not stop water flow but that create a barrier to fish movement. Fish passage also may be
blocked or inhibited by erosion below a culvert that "perches" the culvert and creates a waterfall, by
shallow water caused by a wide culvert and periodic low streamflows, or by excessively high channel
gradients. If blockages occurred during adult salmon immigration or juvenile salmon emigration and
were not cleared for several days, production of a year-class (i.e., fish spawned in the same year) would
be lost from or diminished in the stream above the culvert.
Culverts can also fail to convey water due to landslides or, more commonly, floods that wash out
undersized or improperly installed culverts. In such failures, the stream would be temporarily
impassible to fish until the culvert is repaired or until erosion re-establishes the channel. If the failure
occurs during a critical period in salmon migration, effects would be the same as with a debris blockage
(i.e., a lost or diminished year-class).
Culvert failures also would result in the downstream transport and deposition of sediment. This could
cause returning salmonids to avoid the stream, if they arrived during or immediately following the
failure. More likely, the deposition of fine sediment from the washed-out culvert would smother
salmonid eggs and larvae, if they were present, and would degrade the downstream habitat for
salmonids and the invertebrates that they eat. It would also change stream hydraulics and channel
morphology, generally diminishing habitat value.
Blockage offish passage at road crossings would be infrequent during operation, because our scenarios
assume daily inspection and maintenance. However, after mine operations end, the road may be
maintained less carefully or maintenance may be transferred to a state or a local governmental entity. In
that case, the proportion of culverts that are impassable would be expected to revert to the levels found
in published surveys (30 to 61% inhibit fish passage at any time) (Langill and Zamora 2002, Gibson et al.
2005, Price et al. 2010). Of the 45 culverts that would be required, 36 would be on streams that are
believed to support salmonids. Hence, 11 to 22 streams would be expected to lose passage of salmon or
resident trout or Dolly Varden and some proportion of those would have degraded downstream habitat
resulting from sedimentation caused by road washout.
Of the 36 culverted salmonid streams, 32 contain restricted (less than 5.5 km) upstream habitat.
Assuming typical maintenance practices after mine operations, approximately 10 to 19 of the 32
streams would be entirely or partially blocked at any time. As a result, isolation of resident species such
as rainbow trout or Dolly Varden in such short stream segments would likely result in failure of the
populations, if that isolation was sustained.
It should be noted that high streamflows in and immediately downstream of a culvert and the structure
of the culvert may inhibit fish passage even if movement is not blocked. Culvert-induced erosion could
cause channel entrenchment, disrupting floodplain habitat and floodplain/channel ecosystem processes.
Bristol Bay Assessment 14-10 January 2014
-------
Chapter 14 Integrated Risk Characterization
14.1.2.4 Truck Accidents
Trucks would carry ore-processing chemicals to the mine site and molybdenum product concentrate to
the port. Truck accident records indicate that truck accidents near streams are likely over the long
period of mine operation. These accidents could release sodium ethyl xanthate, cyanide, other process
chemicals, or molybdenum product concentrate to streams or wetlands, resulting in toxic effects on
invertebrates and fish. However, the risk of spills could be mitigated by using impact-resistant
containers.
14.1.2.5 Pipeline Failure
The primary product of the mine would be a copper concentrate that would be pumped as a slurry in a
pipeline to a Cook Inlet shipping facility. Water that carried the sand-like concentrate would be returned
to the mine site in a second pipeline. Based on the record of pipelines in general, and metal concentrate
pipelines in particular, one near-stream failure and two near-wetland failures of each of these pipelines
would be expected to occur over the duration of the Pebble 6.5 scenario (approximately 78 years). In
either case, metal-contaminated water would be released, potentially killing fish and invertebrates in
the affected stream over a relatively brief period. The aqueous phase of the concentrate slurry would be
lethal to sensitive invertebrates and potentially to fish larvae, but a kill of adult fish is not expected. If
the concentrate pipeline spilled into a stream, concentrate would, depending on streamflows, settle and
form bed sediment, be carried downstream and deposited in low-velocity areas, or be carried to Iliamna
Lake and deposited near the shore. Deposited concentrate is predicted to be highly toxic based on its
high copper content and the acidity of its leachate. Unless the receiving stream was dredged, causing
physical damage, this sediment would persist for decades before ultimately being washed into Iliamna
Lake. Potential concentrations in the lake could not be predicted; however, near the pipeline route,
Iliamna Lake contains important beach spawning areas for sockeye salmon that could be exposed to a
spill. Sockeye also spawn in the lower reaches of streams that could be directly contaminated by a spill.
Spills from a diesel pipeline are estimated to have the same probability of occurrence as concentrate
spills. Based on multiple lines of evidence, a spill in the diesel pipeline failure scenario would be
sufficient to kill invertebrates and possibly fish. Remediation is expected to have little success, but
recovery would likely occur within 3 years.
14.1.2.6 Common Mode Failures
Multiple failures could result from a common event, such as an earthquake or a severe storm with heavy
precipitation (particularly heavy rain on snow). Failures resulting from such an event could include
multiple tailings dam failures that spill tailings slurry to streams and rivers, road culvert washouts that
send fine sediment downstream and potentially block fish passage, and product slurry and return water
pipeline failures resulting from culvert washout and scouring of the streambed or a slide of the roadbed.
The effects of these accidents individually would be the same as discussed previously, but the co-
occurrence of these failures would cause cumulative effects on salmonid populations and would make
any mitigative response more difficult.
Bristol Bay Assessment 14-11 January 2014
-------
Chapter 14 Integrated Risk Characterization
Over the perpetual timeframe that the tailings, mine pit, road, and waste rock piles would be in place,
the likelihood of multiple extreme precipitation events, earthquakes, or combinations of these events
becomes much greater. Multiple events further increase the chances of weakening and eventual failure
of facilities that are still in place.
14.2 Overall Loss of Wetlands, Ponds, and Lakes
Wetlands are a dominant feature of the landscape in the Pebble deposit area and are important habitats
for salmon and other fish. Ponds and riparian wetlands provide spawning, rearing, and refuge habitat
for both anadromous and resident fish. Other wetlands moderate streamflows and water quality, and
can influence downstream delivery of dissolved organic matter, particulate organic matter, and aquatic
macroinvertebrates that supply energy sources to fish. In the Pebble 0.25, 2.5, and 6.5 scenarios, 4.5,12,
and 18 km2 of wetlands and 0.41, 0.93, and 1.8 km2 of ponds and lakes, respectively, would be filled or
excavated. In addition, an unquantifiable area of riparian floodplain would be lost or would suffer
substantial changes in hydrologic connectivity with streams, due to reduced flow from the mine
footprint. Another 0.11 km2 of wetlands would be filled in the Kvichak River watershed by the roadbed
of the transportation corridor. By interrupting flow and adding silt and salts, the roadbed would also
influence approximately 4.7 km2 of wetlands, ponds, and lakes occurring within 200 m of the roadbed.
Finally, a diesel or product concentrate spill could damage wetlands and eliminate or degrade their
capacity to support fish.
14.3 Overall Fish-Mediated Risk to Wildlife
Interactions between salmon and wildlife and the potential for disruption of these interactions are
complex. Annual salmon runs provide food for brown bears, bald eagles, other land birds, and wolves. In
addition, wildlife abundance and production are enhanced by the marine-derived nutrients that salmon
carry on their spawning migration. Those nutrients are released into streams when the salmon die,
enhancing the production of other aquatic species that feed wildlife. Salmon predators deposit nutrients
on the landscape, fertilizing the vegetation and increasing the abundance and production of moose,
caribou, and other wildlife.
The effects of reduced Pacific salmon, Dolly Varden, and rainbow trout production on wildlife would be
complex, may not be linearly proportional, and cannot be quantified at this time. Factors such as the
magnitude, seasonally, duration, and location of salmon losses would determine the specific species
affected and the magnitude of effects. However, some degree of reduction in wildlife would be expected
due to the mine footprint and routine operations in each mine size scenario. Because salmon provide a
food source for brown bears, wolves, bald eagles, and other birds, it is likely these species would be
directly affected by a reduction in salmon abundance. Indirect effects on water birds and land birds
through a loss of aquatic invertebrates and on moose and caribou through a loss of marine-derived
nutrients to vegetation are likely, but research is needed to document those linkages.
Bristol Bay Assessment 14-12 January 2014
-------
Chapter 14 Integrated Risk Characterization
Fish-eating wildlife species are also potentially exposed to contaminants bioaccumulated by fish.
However, analyses based on the concentrations of metals in waste rock, tailings, and product
concentrate leachates suggest that toxic effects to wildlife via this route of exposure are unlikely.
14.4 Overall Fish-Mediated Risk to Alaska Native Cultures
Alaska Natives are particularly vulnerable to any changes in the quantity or quality of wild salmon
resources, due to the importance of salmon in terms of both subsistence and cultural identity. Any
change in salmon resources would likely change the diet, social networks, cultural cohesion, and
spiritual well-being of the Alaska Native cultures in the region. These changes could, in turn, result in the
following.
• Effects on human health from loss of a highly nutritious subsistence food and the physical and
mental benefits of a subsistence way of life.
• Degradation of a social support system based on food sharing.
• Decrease in family cohesion and cultural continuity from a loss of family-based subsistence work.
• Mental health degradation from the disruption of spiritual practices and beliefs centered on salmon
and clean water.
Human health and cultural effects related to decreases in salmon resources would vary with the
magnitude of these reductions and cannot be predicted quantitatively. Some fish-mediated effects on
Alaska Native cultures are likely due to the mine footprint or routine operations in any of the mine size
scenarios considered. At minimum, there would be a loss of subsistence use areas and the risk of
decreased use offish because of a perceived change in quality of the fish due to mine operations. Along
the transportation corridor, complex and unpredictable changes to subsistence use would result from
increased access (by both Alaska Natives and others) and possible habitat changes. If significant failures
of water treatment or other infrastructure that greatly affect salmon resources occur during or after
mine operation, large-scale impacts on both subsistence food resources and the cultural, social, and
spiritual cohesion of the local indigenous cultures would occur.
Because the Alaska Native cultures in the Bristol Bay watershed have significant ties to specific land and
water resources that have evolved over thousands of years, it is not possible to replace the value of any
subsistence use areas lost to mine operations elsewhere. As a result, compensatory mitigation,
restoration, or replacement in the case of a failure would be difficult, if not impossible.
It should be noted that, although this assessment focuses on potential effects on Alaska Native cultures,
many of the non-Alaska Natives that reside in the area also practice a subsistence way of life and have
strong long-term cultural ties to the landscape that go back generations. In addition, a large group of
seasonal commercial fishers and cannery workers depend on these resources and have strong, multi-
generational cultural connections to the region. These groups also would be vulnerable to negative
impacts on salmon.
Bristol Bay Assessment 14-13 January 2014
-------
Chapter 14 Integrated Risk Characterization
14.5 Summary of Uncertainties and Limitations in the
Assessment
This assessment makes various reasonable assumptions about the mining, processing, and transporting
of the porphyry copper resources in the Pebble deposit and elsewhere in the Nushagak and Kvichak
River watersheds. If those resources are mined in the future, actual events would not be identical to the
mine scenarios considered here. This is not treated as a source of uncertainty, because it is an inherent
aspect of any predictive assessment. Even an environmental assessment of a mining company's
proposed plan would be an assessment of a scenario that undoubtedly would differ from actual events.
As discussed in the individual chapters, this assessment does have uncertainties and limitations in the
extent to which the potential effects of these scenarios can be estimated. Major uncertainties are
summarized below.
• The estimated annual probability of a tailings dam failure is uncertain and based on design goals
rather than historical experience. Actual failure rates could be higher or lower than the estimated
range of probabilities.
• The proportion of the tailings that would spill in the event of a dam failure could be larger than the
largest value modeled (20%). However, even this conservative assumption results in an initial
outflow beyond the 30-km limit of the model in the Pebble 2.0 dam failure scenario.
• The ultimate fate of spilled and deposited tailings in the event of a dam failure could not be
quantified. From principles of geohydrology and review of analogous cases, we know that slurry
would erode from areas of initial deposition and move downstream over more than a decade.
However, the data needed to model that process and the resources to develop the model are not
currently available.
• It is uncertain whether and how a tailings spill into a remote roadless area would be remediated,
how long it would take to remediate, and to what extent remediation could reduce effects
downstream of the initial slurry runout.
• The effects of mining on fish populations could not be quantified because of the lack of quantitative
information concerning Pacific salmon, Dolly Varden, and rainbow trout populations and their
responses. The occurrence of salmonid species in the region's rivers and major streams is generally
known, but not their abundances, productivities, or limiting factors. Estimating changes in
populations would require population modeling, which requires knowledge of life-stage-specific
survival and production as well as knowledge of limiting factors and processes that were not
available for this case. Further, it requires knowledge of how temperature, habitat structure, prey
availability, density dependence, and sublethal toxicity influence life-stage-specific survival and
production, which is not available. Obtaining that information would require more detailed
monitoring and experimentation. Salmon populations naturally vary in size because of a great many
factors that vary among locations and years, and collecting sufficient data to establish reliable
Bristol Bay Assessment 14-14 January 2014
-------
Chapter 14 Integrated Risk Characterization
salmon population estimates takes many years. Thus, we used estimated effects of mining on habitat
as a reasonable surrogate for estimated effects on fish populations.
• Standard leaching test data are available for test tailings and waste rocks from the Pebble deposit,
but these results are uncertain predictors of the actual leachate composition from a tailings
impoundment, tailings deposited in streams and on their floodplains, and waste rocks piles. Test
conditions are artificial, and the materials tested may not be representative. In particular, the pyritic
tailings were not tested. Additionally, data and resources were insufficient to allow geochemical
modeling of water quality expected in the TSF or downstream of the mine site under varied
chemical and hydrological conditions, or to model expected pit water chemistry at closure.
• The effects of tailings and product concentrate deposited in spawning and rearing habitat are
uncertain. It is clear that they would be harmful to salmonid eggs, alevins, or sheltering fry due to
both physical and toxicological effects, but the concentration in spawning gravels required to reduce
reproductive success of salmonids is unknown.
• The actual response of Alaska Native cultures to any of these scenarios is uncertain. Interviews with
tribal Elders and culture bearers and other evidence suggest that responses would involve loss of
food resources and cultural disruption, but it is not possible to predict specific changes in
demographics, cultural practices, or physical and mental health.
• Although some tailings would eventually reach the estuarine portions of the Nushagak River and
even Bristol Bay, exposures at that distance could not be estimated. Therefore, risks to salmonids
resulting from marine and estuarine contamination could not be addressed.
• The assessment is limited by its focus on the effects of mining on salmonids and consequent indirect
effects of diminished fish resources on wildlife and people. Direct effects of mining on humans,
wildlife, and terrestrial ecosystems, as well as induced development associated with mine-related
activities, are not evaluated in this assessment.
• Some sources, such as air pollution from a power plant, were not addressed because they are less
related to the Clean Water Act or because they were judged to pose less risk to salmonids.
• Climate change will affect both the probability and magnitude of mine-related failures, as well as
change the habitat quality and biology of salmonids. These climate effects are highly uncertain, but
their likely qualitative influences are described in Box 14-2.
Bristol Bay Assessment 14-15 January 2014
-------
Chapter 14 Integrated Risk Characterization
BOX 14-2. CLIMATE CHANGE AND POTENTIAL RISKS OF LARGE-SCALE MINING
Climate change in the Bristol Bay region (Section 3.8) will likely result in changes in snowpack and the timing of
snowmelt, a greater chance for rain-on-snow events, and an increase in flooding. These changes are likely to affect
multiple aspects of any large-scale mining in the area, including mine infrastructure, the transportation corridor,
water treatment and discharge, and post-closure management (Pearce et al. 2011).
Mine infrastructure (e.g., buildings, waste rock piles, tailings storage facilities [TSFs], and water retention facilities)
and the transportation corridor likely would be affected by extreme weather events resulting in increased flooding
(Instanes et al. 2005, Pearce et al. 2011). These components would need to be designed for potential increases in
flood frequency and magnitude and changes in storm patterns, because these changes could weaken structural
integrity, increase embankment instability, and accelerate erosion (Instanes et al. 2005, Pearce et al. 2011).
Water management would be a major challenge at the mine site, and changes resulting from climate change could
exacerbate the challenge. Climate change would contribute to future changes in temperature, precipitation,
evapotranspiration, hydrology, and seasonal flooding and drying patterns. Changes in water availability and
groundwater recharge would affect the amount and timing of water available and the hydrologic gradients of
groundwater in and around the mine site, thereby requiring changes to water management in the mine pit and other
areas of the mine site.
Under future climate conditions, the return intervals of various sized storms could change (e.g., medium-sized
storms could become more frequent or the frequency of current 100-year storms could change). Possible increases
in flood magnitudes would require the need to plan for larger and more frequent flood events at the mine site. This
in turn may affect the likelihood of a tailings dam failure, overtopping of ponds, and/or flooding of water
management facilities. Failure to plan for these conditions could result in unintended environmental releases. For
example, Minto Mine, a copper-gold mine in Canada, was forced to release untreated water into the Yukon River
system in 2008, due to torrential rains and the mine's inability to manage this increased water (Pearce et al. 2011).
Mine infrastructure would need to be designed to account for projected climatic changes, such that its structural
integrity can be maintained in perpetuity, even under potentially more extreme climatic conditions. In addition, any
mine reclamation plan would need to consider changing climate conditions and how those changes will directly
affect fish and wildlife populations. The following list includes infrastructure and operations design, maintenance,
and management that would need to consider climate change.
• Mine footprint. Climate change may affect water availability both within and across seasons (e.g., via changes
in snowmelt patterns; amount, type, and timing of precipitation; frequency of large storms; and groundwater
inputs). Water processing associated with these changes would alter flow and temperature in downstream
water bodies.
• Water treatment and discharge. Climate change might result in greater volumes of water requiring treatment,
changes in the dilution provided by receiving streams, changes in temperature that would affect management of
discharged water, and potential overload due to lack of storage and treatment capacity.
• TSF failure. TSFs may be exposed to greater volumes of water, and the probability of dam failure due to
overtopping may increase with changes in precipitation patterns and/or rapid snowmelt.
• Transportation corridor. Greater flood frequencies and an increase in erosion and sedimentation are likely to
affect streams and wetlands along the transportation corridor.
• Culvert, pipeline, and bridge failures. These failures may be more likely due to changes in precipitation
patterns, rapid snowmelt, larger water volumes, debris issues, and sedimentation.
• Cumulative effects of multiple mines. Climate change issues are likely to affect any mining operation in the
area. As the number of mines increases, the likelihood of having mine and climate change interactions might
increase.
Climate change could obscure or complicate efforts to monitor habitat and fish population responses to mine-related
activities. Survey and monitoring designs would need to take potential climate change effects into account through
strategic measurement of stream and lake temperatures, precipitation, water flow, and fish populations throughout
the Bristol Bay watershed. Monitoring design could be aided by models able to downscale climate effects and
project changes at watershed scales. Population monitoring should take salmon adaptation and metapopulation
dynamics into account, and be cognizant of the many interacting processes influencing populations. Protecting
salmon sustainability in an uncertain future will require adaptability of both management and monitoring strategies
(Schindler etal. 2008).
Bristol Bay Assessment 14-16 January 2014
-------
Chapter 14 Integrated Risk Characterization
14.6 Summary of Uncertainties in Mine Design and Operation
In addition to uncertainties in the assessment, uncertainties are inherent in planning, designing,
constructing, operating, and closing a mine. Such uncertainties are inherent in any complex enterprise,
particularly when it involves an incompletely characterized natural system. However, the large scales
and long durations of any effort to mine the Pebble deposit make these inherent uncertainties more
prominent.
• Mines are complex systems requiring skilled engineering, design, and operation. The uncertainties
facing mining and geotechnical engineers include unknown geological features, uncertain values in
geological properties, limited knowledge of mechanisms and processes at the site, and human error
in design, construction, and operation. Vick (2002) notes that models used to predict the behavior of
an engineered system are "idealizations of the processes they are taken to represent, and it is well
recognized that the necessary simplifications and approximations can introduce error in the model."
Engineers use professional judgment in addressing uncertainty (Vick 2002).
• Accidents are inherently unpredictable. Though systems can be put into place to reduce system
failures, seemingly logical decisions about how to respond to a given situation can have unexpected
consequences resulting from human error—for example, the January 2012 overtopping of the
tailings dam at the Nixon Fork Mine near McGrath, Alaska (Box 8-1). Further, unforeseen events or
events that are more extreme than anticipated can negate the apparent wisdom of prior decisions
(Caldwell and Charlebois 2010).
• The ore deposit would be mined for decades, and wastes would require management for centuries
or even in perpetuity. Engineered mine waste storage systems have only been in existence for about
50 years, so their long-term behavior is not known. The performance of modern technology in the
construction of tailings dams is untested and unknown in the face of centuries of extreme events
such as earthquakes and major storms.
• Human institutions change. Over the long time span of mining and post-mining care, generations of
mine operators must exercise due diligence. Priorities are likely to change in the face of financial
crises, changing markets for metals, new information about the resource, political priorities, or any
number of currently unforeseeable changes in circumstance. The promises of today's mine
developers may not be carried through by future generations of operators whose sole obligation is
to the shareholders of their time (Blight 2010). Similarly, governments that are expected to assume
responsibility when mining companies fail may not appropriately manage mine sites or the funds in
performance bonds.
14.7 Summary of Risks in the Mine Scenarios
Even if the mining and mitigation practices described in the mine scenarios were performed perfectly,
an operation of this size would inevitably destroy or degrade habitat of salmonids. The mine scenario
footprints would eliminate, block, or dewater streams known to support spawning and rearing habitat
Bristol Bay Assessment 14-17 January 2014
-------
Chapter 14
Integrated Risk Characterization
for coho, Chinook, and sockeye salmon and Dolly Varden (Table 14-2). Wetlands would be filled or
excavated in 4.5,12, and 18 km2 of the mine footprints in the Pebble 0.25, 2.0, and 6.5 scenarios,
respectively; an additional 0.41, 0.93, and 1.8 km2 of ponds and lakes would also be lost. Altered
streamflows resulting from water use would significantly degrade additional stream reaches (Table 14-
2) and an unquantifiable area of wetland habitat. Leachates and other wastewater would be collected
and treated to meet standards, but leakage would be sufficient to cause direct toxic effects to fish in up
to 57 km of streams and indirect effects due to loss of invertebrate food species in up to 82 km of
streams (Table 14-2). In addition, the temperature and distribution of effluents could further degrade
habitat. Streams between the transportation corridor and Iliamna Lake would receive silt and deicing
chemicals, which would reduce habitat quality.
Table 14-2. Summary of estimated stream lengths potentially affected in the three mine size
scenarios, assuming routine operations.
Effect
Eliminated, blocked, or dewatered
Eliminated, blocked, or dewatered— anadromous
>20% flow alteration3
Direct toxicity to fisha
Direct toxicity to invertebrates3
Downstream of transportation corridor
Stream Length Affected (km)
Pebble 0.25
38
8
15
0
21
Pebble 2.0
89
22
27
24
40-62
Pebble 6.5
151
36
53
34-57
60-82
272
Notes:
8 Stream reaches with streamflow alterations partially overlap those with toxicity.
This assessment considered failures of a tailings dam; product concentrate, return water, and diesel
pipelines; roads and culverts; and water collection and treatment. Tailings dam failures are improbable
in that they have a low rate of occurrence, but some sort of failure becomes likely in the extremely long-
term. A tailings dam failure could destroy salmonid habitat in more than 30 km of the North Fork
Koktuli River and associated wetlands for years to decades. Product concentrate and diesel pipeline
failures near streams would be expected to occur during the life of a mine. Both would cause acute lethal
effects on invertebrates and fish, and the concentrate could create highly toxic sediment. A truck wreck
near a stream could introduce highly toxic chemicals causing acute lethality to fish and invertebrates.
Culvert failures would be common, unless a more rigorous than usual maintenance program were
maintained, and could block fish passage and degrade downstream habitat. Failures to collect and treat
leachates and other wastewaters could cause releases ranging from short-term and innocuous to long-
term and highly toxic to fish and invertebrates.
14.8 Summary of Cumulative Risks of Multiple Mines
To provide realism and detail, this assessment largely addresses the potential effects of a single mine, at
three different sizes, on the Pebble deposit. However, the development of multiple mines of various sizes
is plausible in the Nushagak and Kvichak River watersheds. Several known mineral deposits with
Bristol Bay Assessment
14-18
January 2014
-------
Chapter 14 Integrated Risk Characterization
potentially significant resources are located in the two watersheds, and active exploration is underway
at a number of claim blocks. The construction of roads, pipelines, and other infrastructure for one mine
would likely facilitate the development of additional mines. Thus, the development of multiple mines
and their associated infrastructure may affect fish populations, wildlife, and Alaska Native villages
distributed across these watersheds.
Outside of the Bristol Bay watershed, most ecosystems that support Pacific salmon have been modified
by the cumulative effects of multiple land and water uses. Anadromous fish are particularly susceptible
to regional-scale effects, because they require suitable habitat in spawning areas, rearing areas, and
along migration corridors. Because Pacific salmon, Dolly Varden, and rainbow trout migrate among
freshwater habitats seasonally or between life stages, loss or degradation of habitat in one location can
diminish the ability of other locations to support these species. As a result of their particular
susceptibility, anadromous salmonid fisheries have declined in most of their range due to the combined
effects of habitat loss and degradation, pollution, and harvesting.
The Nushagak and Kvichak River watersheds are relatively undisturbed, and their ecosystems have not
yet experienced these cumulative stresses associated with human activity. Bristol Bay salmon runs are
resilient because the abundance, diversity, and quality of Bristol Bay habitats result in large and diverse
salmon populations. Fluctuations in habitat availability or quality across the watersheds caused by
natural processes typically result in temporary loss or reduction of a discrete portion of habitat, but
these fluctuations are compensated for by Bristol Bay's diverse salmon populations. In contrast, the
effects of mining may be long-lasting and extensive, eliminating habitat for extended periods and
potentially killing or otherwise eliminating fish populations. Such effects may remove component
populations permanently or for long periods of time, weakening the overall population's ability to resist
and rebound from disturbance.
To examine the potential cumulative risks of multiple mines, we consider development of additional
mines at the Pebble South/PEB, Big Chunk South, Big Chunk North, Groundhog, AUDN/Iliamna, and
Humble prospects. The AUDN/Iliamna and Humble prospects are located approximately 90 and 135 km,
respectively, southwest of the Pebble deposit. All of the other prospects are within 25 km of the Pebble
deposit and may be of the same geological origin. Construction of mining infrastructure at the Pebble
deposit would substantially reduce development costs for surrounding prospects and could facilitate
creation of a mining district that could include these sites.
Impacts from the footprint of major mine components and associated accidents and failures would be
similar to those projected in the mine scenarios. The footprints of the major mine components would
eliminate substantial amounts of stream and wetland habitats, both directly and through dewatering.
Total stream length eliminated by these components would range from 43 to 70 km, and wetland area
lost would range from 7.9 to 27 km2. Further habitat loss and degradation would result from flow
alteration. Each additional mine would increase flow alteration from water removal and retention,
increased impervious surface, and road crossings.
Bristol Bay Assessment 14-19 January 2014
-------
Chapter 14 Integrated Risk Characterization
The consequences of leachate collection or treatment failure would depend on the chemical nature of
the rock or tailings over which it flows. Because porphyry copper deposits tend to straddle the threshold
between acid-generating and non-acid-generating, some of the waste rock and a portion of the tailings at
any of these additional mines would be reasonably likely to be acid-generating. Each additional facility
would increase the likelihood of collection and treatment failures, which would increase the frequency
of discharge of untreated leachate or other wastewater in the Nushagak and Kvichak River watersheds,
with each event resulting in an increment of impact. Longer roads and pipelines associated with
additional mines, coupled with a greater number of stream crossings, would increase the frequency of
events such as culvert failures, pipeline breaks, and truck accidents that would damage aquatic systems,
incrementally decreasing habitat value over an extensive area. In the long-term, cessation of
maintenance and treatment would likely result in the degradation of fisheries in waters downstream of
each mine. Extreme natural events such as earthquakes and floods could cause failures of dams, roads,
pipelines, or WWTPs at multiple mines.
Induced development is that which results from the introduction of industry, roads, and infrastructure.
It is reasonably foreseeable that infrastructure from large-scale mining in the Nushagak and Kvichak
River watersheds, particularly the transportation corridors, would induce further development in the
region. Existing communities, the tourism industry, and the recreational housing market could benefit if
large-scale mining expanded throughout the watersheds. Unmanaged access to currently roadless
wilderness areas also could expand. Improved access would increase hunting and fishing pressure, as
well as competition with existing subsistence users; increase damage from off-road vehicle, boat, and
foot traffic in currently inaccessible areas; facilitate poaching, dumping, trespassing, and other activities;
and lead to scattered development in the watersheds.
Bristol Bay Assessment 14-20 January 2014
-------
15.1 References by Chapter
15.1.1 Executive Summary
ADF&G (Alaska Department of Fish and Game). 2012. Alaska Freshwater Fish Inventory Database.
Available: http://www.adfg.alaska.gov/index.cfm?adfg=ffinventory.main.
Johnson, ]., and P. Blanche. 2012. Catalog of Waters Important for Spawning, Rearing, or Migration of
Anadromous Fishes—Southwestern Region, Effective June 1, 2012. Special Publication No. 12-08.
Anchorage, AK: Alaska Department of Fish and Game.
Ruggerone, G. T., R. M. Peterman, and B. Dorner. 2010. Magnitude and trends in abundance of hatchery
and wild pink salmon, chum salmon, and sockeye salmon in the North Pacific Ocean. Marine and
Coastal Fisheries: Dynamics, Management, and Ecosystem Science 2:306-328.
USFWS (U.S. Fish and Wildlife Service). 2012. National Wetlands Inventory. Available:
http://www.fws.gov/wetlands/Wetlands-Mapper.html.Accessed: December 11, 2012.
USGS (U.S. Geological Survey). 2012. National Hydrography Dataset. High Resolution, Alaska. Available:
ftp://nhdftp.usgs.gov/DataSets/Staged/States/FileGDB/HighResolution. Accessed: October 16,
2012.
Personal Communications
Baker, T. Area Fishery Research Biologist, Bristol Bay Salmon Program. Alaska Department of Fish and
Game, Anchorage, AK. July 8, 2011—email to Rebecca Shaftel.
Bristol Bay Assessment
15-1
January 2014
-------
Chapter 15 References
15.1.2 Chapter 1—Introduction
Chambers, D., R. Moran, and L. Trasky. 2012. Bristol Bay's Wild Salmon Ecosystems and the Pebble Mine:
Key Considerations for a Large-Scale Mine Proposal. January. Wild Salmon Center and Trout
Unlimited.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company), Vancouver, BC.
PLP (Pebble Limited Partnership). 2013. About Pebble. Available:
http://corporate.pebblepartnership.com/about. Accessed: April 1, 2013.
USEPA (U.S. Environmental Protection Agency). 2002a. Clinch and Powell Valley Watershed Ecological
Risk Assessment. EPA/600/R-01/050. Washington, DC: Office of Research and Development,
National Center for Environmental Assessment. Available
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=15219.
USEPA (U.S. Environmental Protection Agency). 2002b. Ecological Risk Assessment for the Middle Snake
River, Idaho. EPA/600/R-01/017. Washington, DC: Office of Research and Development, National
Center for Environmental Assessment. Available:
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29097&partner=ORD-NCEA.
USEPA (U.S. Environmental Protection Agency). 2002c. WaquoitBay Watershed Ecological Risk
Assessment: The Effect of Land-Derived Nitrogen Loads on Estuarine Eutrophication. EPA/600/R-
02/079. Washington, DC: Office of Research and Development, National Center for Environmental
Assessment. Available: http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=15221.
USEPA (U.S. Environmental Protection Agency). 2008. Predicting Future Introductions ofNonindigenous
Species to the Great Lakes (Final Report]. EPA/600/R-08/066F. Washington, DC: Office of Research
and Development, National Center for Environmental Assessment. Available:
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=190305.
USEPA (U.S. Environmental Protection Agency). 2011. The Effects ofMountaintop Mines and Valley Fills
on Aquatic Ecosystems of the Central Appalachian Coalfields. EPA/600/R-09/138F. Washington, DC:
Office of Research and Development, National Center for Environmental Assessment. Available:
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=225743.
15.1.3 Chapter 2—Overview of Assessment
Chambers, D. M., and B. Higman. 2011. Long-Term Risks of Tailings Dam Failure. Bozeman, MT: Center for
Science in Public Participation. Available:
http://www.csp2.org/files/reports/Long%20Term%20Risks%20of%20Tailings%20Dam%20Failu
re%20-%20Chambers%20%26%20Higman%200ctll-2.pdf.
Bristol Bay Assessment 15-2 January 2014
-------
Chapter 15 References
Earthworks. 2012. U.S. Copper Porphyry Mines Report: The Track Record of Water Quality Impacts
Resulting from Pipeline Spills, Tailings Failures and Water Collection and Treatment Failure.
Washington, DC.
Forman, R. T. T., and L. E. Alexander. 1998. Roads and their major ecological effects. Annual Review of
Ecology and Systematics 29:207-231.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company), Vancouver, BC.
PLP (Pebble Limited Partnership). 2011. Environmental Baseline Document 2004 through 2008.
Anchorage, AK. Available: http://www.arlis.org/docs/vol2/Pebble/2004-2008EBDIndex.pdf.
USEPA (U.S. Environmental Protection Agency). 1998. Guidelines for Ecological Risk Assessment.
EPA/630/R-95/002F. April. Washington, DC: Office of Research and Development.
USFWS (U.S. Fish and Wildlife Service). 2012. National Wetlands Inventory. Available:
http://www.fws.gov/wetlands/Wetlands-Mapper.html.Accessed: December 11, 2012.
USGS (U.S. Geological Survey). 2012. National Hydrography Dataset. High Resolution, Alaska. Available:
ftp://nhdftp.usgs.gov/DataSets/Staged/States/FileGDB/HighResolution. Accessed: October 16,
2012.
Woody, C. A., and B. Higman. 2011. Groundwater as Essential Salmon Habitat in Nushagak and Kvichak
River Headwaters: Issues Relative to Mining. Unpublished report. Available:
http://www.fish4thefuture.com/pdfs/Groundwater_and_SalmonFINAL27Augll.pdf.
15.1.4 Chapter 3—Region
ACIA (Arctic Climate Impact Assessment). 2004. Impacts of a Warming Arctic: Arctic Climate Impact
Assessment. Cambridge, UK: Cambridge University Press.
ADF&G (Alaska Department of Fish and Game). 2012. Alaska Freshwater Fish Inventory Database.
Available: http://www.adfg.alaska.gov/index.cfm?adfg=ffinventory.main.
Araki, H., B. Cooper, and M. S. Blouin. 2009. Carry-over effect of captive breeding reduces reproductive
fitness of wild-born descendants in the wild. Biology Letters 5(5):621-624.
Armstrong, J. B., D. E. Schindler, K. L. Omori, C. P. Ruff, and T. P. Quinn. 2010. Thermal heterogeneity
mediates the effects of pulsed subsidies across a landscape. Ecology 91:1445-1454.
Barnett, T. P., J. C. Adam, and D. P. Lettenmaier. 2005. Potential impacts of a warming climate on water
availability in snow-dominated regions. Nature 438:303-309.
Bristol Bay Assessment 15-3 January 2014
-------
Chapter 15 References
Battin, ]., M. W. Wiley, M. H. Ruckelshaus, R. N. Palmer, E. Korb, K. K. Bartz, and H. Imaki. 2007. Projected
impacts of climate change on salmon habitat restoration. Proceedings of the National Academy of
Sciences of the United States of America 104(16):6720-6725.
Beer, W. N., and J. J. Anderson. 2001. Effect of spawning day and temperature on salmon emergence:
interpretations of a growth model for Methow River chinook. Canadian Journal of Fisheries and
Aquatic Sciences 58:943-949.
Berggren, T. J., and M. J. Filardo. 1993. An analysis of variables influencing the migration of juvenile
salmonids in the Columbia River basin. North American Journal of Fisheries Management 13:48-63.
Bjornn, T. C., and D. W. Reiser. 1991. Habitat requirements of salmonids in streams. Pages 83-138 in W.
R. Meehan (ed.), Influences of Forest and Rangeland Management on Salmonid Fishes and their
Habitats. Bethesda, MD: American Fisheries Society.
Bryant, M. 2009. Global climate change and potential effects on Pacific salmonids in freshwater
ecosystems of southeast Alaska. Climatic Change 95:169-193.
Burgner, R. L. 1991. Life history of sockeye salmon (Oncorhynchus nerka}. Pages 1-118 in C. Groot and L.
Margolis, (eds.), Pacific Salmon Life Histories. Seattle, WA: University of Washington Press.
Burnett, K. M., G. H. Reeves, D. J. Miller, S. Clarke, K. Vance-Borland, and K. Christiansen. 2007.
Distribution of salmon-habitat potential relative to landscape characteristics and implications for
conservation. Ecological Applications 17(1):66-80.
Christensen, J. H., B. Hewitson, A. Busuioc, A. Chen, X. Gao, I. Held, R. Jones, R. K. Kolli, W.-T. Kwon, R.
Laprise, V. Magaa Rueda, L. Mearns, C. G. Menndez, J. Risnen, A. Rinke, A. Sarr, and P. Whetton. 2007.
Regional Climate Projections. In Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.
Tignor and H.L. Miller (eds.), Climate Change 2007: The Physical Science Basis. Contribution of
Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change.
Cambridge, UKC and New York, NY: Cambridge University Press.
CRU (Climatic Research Unit). 2012. Data. University of East Anglia, UK. Available:
http://www.cru.uea.ac.uk/data.
Demory, R. L., R. L. Orrell, and D. R. Heinle. 1964. Spawning Ground Catalog of the Kvichak River System,
Bristol Bay, Alaska. Issue 488 of Special Scientific Report—Fisheries. Washington, DC: U.S.
Department of Interior, Bureau of Commercial Fisheries.
Detterman, R. L. 1986. Glaciation of the Alaska Peninsula. Pages 151-170 in Hamilton, T. D., K. M. Reed,
and R. M. Thorson (eds.), Glaciation in Alaska. Alaska Geological Society.
Detterman, R. L., and B. L. Reed. 1973. Surficial Deposits of the Iliamna Quadrangle, Alaska. Geological
Survey Bulletin 1368-A. U.S. Geological Survey. Available:
http://www.dggs.alaska.gOV/webpubs/usgs/b/text/bl368a.pdf.Accessed: December 17, 2012.
Bristol Bay Assessment 15-4 January 2014
-------
Chapter 15 References
Eaton, J. G., and R. M. Scheller. 1996. Effects of climate warming on fish thermal habitat in streams of the
United States. Limnology and Oceanography 41:1109-1115.
Eiler, J. H., B. D. Nelson, and R. F. Bradshaw. 1992. Riverine spawning by sockeye salmon in the Taku
River, Alaska and British Columbia. Transactions of the American Fisheries Society 121(6):701-708.
Feddema, J. J. 2005. A revised Thornthwaite-type global climate classification. Physical Geography
26:442-466.
Gallant, A. L., E. F. Binnian, J. M. Omernik, and M. B. Shasby. 1995. Ecoregions of Alaska. U.S. Geological
Survey Professional Paper 1567. Washington, DC.
Gesch, D. 2007. Chapter 4-The National Elevation Dataset. Pages 99-118 in Maune, D. (ed.), Digital
Elevation Model Technologies and Applications: The DEM Users Manual. 2nd edition. Bethesda, MD:
American Society for Photogrammetry and Remote Sensing.
Gesch, D., M. Oimoen, S. Greenlee, C. Nelson, M. Steuck, and D. Tyler. 2002. The National Elevation
Dataset. Photogrammatic Engineering and Remote Sensing 68(1):5-11.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company), Vancouver, BC.
Haeussler, P. J., and R. W. Saltus. 2004.26 km Offset on the Lake Clark Fault Since Late Eocene Time.
Studies by the U.S. Geological Survey in Alaska, 2004. U.S. Geological Survey Professional Paper
1709-A.
Haeussler, P. J., and C. F. Waythomas. 2011. Review of the Origin of the Braid Scarp Near the Pebble
Prospect, Southwestern Alaska. U.S. Geological Survey Open File Report 2011-1028.
Hamon, W. R. 1961. Estimating potential evapotranspiration./ourna/ of the Hydraulics Division,
Proceedings of the American Society of Civil Engineers 87:107-120.
Hauer, F. R., J. S. Baron, D. H. Campbell, K. D. Fausch, S. W. Hosteller, G. H. Leavesley, P. R. Leavitt, D. M.
McKnight, and J. A. Stanford. 1997. Assessment of climate change and freshwater ecosystems of the
Rocky Mountains, USA and Canada. Hydrological Processes 11:903-924.
Healey, M. 2011. The cumulative impacts of climate change on Fraser River sockeye salmon
(Oncorhynchus nerka] and implications for management. Canadian Journal of Fisheries and Aquatic
Sciences 68:953-953.
Heming, T. A., J. E. Mclnerney, and D. F. Alderdice. 1982. Effect of temperature on initial feeding in
alevins of Chinook salmon (Oncorhynchus tshawytschd). Canadian Journal of Fisheries and Aquatic
Sciences 39:1554-1562.
Bristol Bay Assessment 15-5 January 2014
-------
Chapter 15 References
Hilborn, R., T. Quinn, D. Schindler, and D. Rogers. 2003. Biocomplexity and fisheries sustainability.
Proceedings of the National Academy of Sciences of the United States of America 100:6564-6568.
Hinzman, L. D., N. D. Bettez, W. R. Bolton, F. S. Chapin, M. B. Dyurgerov, C. L. Fastie, B. Griffith, R. D.
Hollister, A. Hope, H. P. Huntington, A. M. Jensen, G. J. Jia, T. Jorgenson, D. L. Kane, D. R. Klein, G.
Kofinas, A. H. Lynch, A. H. Lloyd, A. D. McGuire, F. E. Nelson, W. C. Oechel, T. E. Osterkamp, C. H.
Racine, V. E. Romanovsky, R. S. Stone, D. A. Stow, M. Sturm, C. E. Tweedie, G. L. Vourlitis, M. D.
Walker, D. A. Walker, P. J. Webber, J. M. Welker, K. S. Winker, and K. Yoshikawa. 2005. Evidence and
implications of recent climate change in Northern Alaska and other Arctic regions. Climatic Change
72:251-298.
Hughes, N. F., and L. M. Dill. 1990. Position choice by drift-feeding salmonids: model and test for Arctic
grayling (Thymallus arcticus] in subarctic mountain streams, Interior Alaska. Canadian Journal of
Fisheries and Aquatic Sciences 47:2039-2048.
Irvine, J. R., M. Fukuwaka, T. Kaga, J. H. Park, K. B. Seong, S. Kang, V. Karpenko, N. Klovach, H. Bartlett, and
E. Volk. 2009. Pacific Salmon Status and Abundance Trends. Document 1199, Rev. 1. North Pacific
Anadromous Fish Commission.
Johnson, J., and P. Blanche. 2012. Catalog of Waters Important for Spawning, Rearing, or Migration of
Anadromous Fishes—Southwestern Region, Effective June 1, 2012. Special Publication No. 12-08.
Anchorage, AK: Alaska Department of Fish and Game.
Kisslinger, C. 1976. A review of theories of mechanisms of induced seismicity. Engineering Geology
10:85-98.
Koehler, R. D., and R. D. Reger. 2011. Reconnaissance Evaluation of the Lake Clark Fault, TyonekArea,
Alaska. Preliminary Interpretative Report 2011-1. Alaska Division of Geological & Geophysical
Surveys.
Kovach, R. P., A. J. Gharrett, and D. A. Tallmon. 2012. Genetic change for earlier migration timing in a pink
salmon population. Proceedings of the Royal Society B-Biological Sciences 279:3870-3878.
Kyle R. E., and T. P. Brabets. 2001. Water Temperature of Streams in the Cook Inlet Basin, Alaska, and
Implications of Climate Change. Water Resources Investigations Report 2001-4109. U.S. Geological
Survey.
Lawson, P. W., E. A. Logerwell, N. J. Mantua, R. C. Francis, and V. N. Agostini. 2004. Environmental factors
influencing freshwater survival and smolt production in Pacific Northwest coho salmon
(Oncorhynchus kisutch}. Canadian Journal of Fisheries and Aquatic Sciences 61:360-373.
Lea, P. D., and C. F. Waythomas. 1990. Late-Pleistocene eolian sand sheets in Alaska. Quaternary
Research 34:269-281.
Bristol Bay Assessment 15-6 January 2014
-------
Chapter 15 References
Lea, P.O., S. A. Elias, and S. K. Short. 1991. Stratigraphy and paleoenvironments of Pleistocene nonglacial
deposits in the southern Nushagak lowland, southwestern Alaska, USA. Arctic and Alpine Research
23:375-391.
Lisi, P. ]., D. E. Schindler, K. T. Bentley, and G. R. Pess. 2013. Association between geomorphic attributes
of watersheds, water temperature, and salmon spawn timing in Alaskan streams. Geomorphology
185:78-86.
Lisle, T. 1989. Sediment transport and resulting deposition in spawning gravels, North Coastal
California. Water Resources Research 25:1303-1319.
Liston, G. E., and C. A. Hiemstra. 2011. The changing cryosphere: pan-Arctic snow trends (1979-2009).
Journal of Climate 24(21}:5691-5712.
Luck, M., N. Maumenee, D. Whited, J. Lucotch, S. Chilcote, M. Lorang, D. Goodman, K. McDonald, J. Kimball,
and J. Stanford. 2010. Remote sensing analysis of physical complexity of North Pacific Rim rivers to
assist wild salmon conservation. Earth Surface Processes and Landforms 35:1330-1343.
Mantua, N., I. Tohver, and A. Hamlet. 2010. Climate change impacts on streamflow extremes and
summertime stream temperature and their possible consequences for freshwater salmon habitat in
Washington State. Climatic Change 102:187-223.
Markon, C. J., S. F. Trainor, and F. S. Chapin III (eds.) 2012. The United States National Climate
Assessment—Alaska Technical Regional Report. Circular 1379. Reston, VA: U.S. Geological Survey.
Martins, E. G., S. G. Hinch, S. J. Cooke, and D. A. Patterson. 2012. Climate effects on growth, phenology,
and survival of sockeye salmon (Oncorhynchus nerka}: a synthesis of the current state of knowledge
and future research directions. Reviews in Fish Biology and Fisheries 22:887-914.
Mellina, E., R. D. Moore, S. G. Hinch, J. S. Macdonald, and G. Pearson. 2002. Stream temperature responses
to clearcut logging in British Columbia: the moderating influences of groundwater and headwater
lakes. Canadian Journal of Fisheries and Aquatic Sciences 59:1886-1900.
Mohseni 0., and H. G. Stefan. 1999. Stream temperature/air temperature relationship: a physical
interpretation. Journal of Hydrology 218(3-4):128-141.
Montgomery, D. R., and J. M. Buffington. 1997. Channel-reach morphology in mountain drainage basins.
Bulletin of the Geological Society of America 109(5):596-611.
Montgomery, D. R., E. M. Beamer, G. R. Pess, and T. P. Quinn. 1999. Channel type and salmonid spawning
distribution and abundance. Canadian Journal of Fisheries and Aquatic Sciences 56(3):377-387.
Montgomery, D. R., J. M. Buffington, N. P. Peterson, D. Schuett-Hames, and T. P. Quinn. 1996. Stream-bed
scour, egg burial depths, and the influence of salmonid spawning on bed surface mobility and
embryo survival. Canadian Journal of Fisheries and Aquatic Sciences 53(5):1061-1070.
Bristol Bay Assessment 15-7 January 2014
-------
Chapter 15 References
Mote, P. W., E. Parson, A. F. Hamlet, W. S. Keeton, D. Lettenmaier, N. Mantua, E. L. Miles, D. Peterson, D. L.
Peterson, R. Slaughter, and A. K. Snover. 2003. Preparing for climatic change: the water, salmon, and
forests of the Pacific Northwest. Climatic Change 61:45-88.
Nehlsen, W., J. Williams, and J. Lichatowich. 1991. Pacific salmon at the crossroads—stocks at risk from
California, Oregon, Idaho, and Washington. Fisheries 16:4-21.
Parks, B., and R. J. Madison. 1985. Estimation of selected flow and water-quality characteristics of Alaskan
streams. U.S. Geological Survey.
Peacock, S. 2012. Projected twenty-first century change in temperature, precipitation, and snow cover
over North America in CCSM4. Journal of Climate 25(13):4405-4429.
Pess, G. R., T. P. Quinn, D. E. Schindler, and M. C. Liermann. 2013. Freshwater habitat associations
between pink (Oncorhynchusgorbuscha], chum (0. keta] and Chinook salmon (0. tshawytscha] in a
watershed dominated by sockeye salmon (0. nerka] abundance. Ecology of Freshwater Fish, doi:
10.1111/eff.l2088.
PLP (Pebble Limited Partnership). 2011. Pebble Project Environmental Baseline Document 2004 through
2008. Anchorage, AK. Available: http://www.arlis.org/docs/vol2/Pebble/2004-2008EBDIndex.pdf.
Power, G., R. S. Brown, and J. G. Imhof. 1999. Groundwater and fish—Insights from northern North
America. Hydrological Processes 13:401-422.
PRISM Climate Group. 2012. Oregon State University. Available: www.prism.oregonstate.edu.
Quinn, T. P. 2005. The Behavior and Ecology of Pacific Salmon and Trout. Seattle, WA: University of
Washington Press.
Quinn, T. P., H. B. J. Rich, D. Gosse, and N. Schtickzelle, and J. Grant. 2012. Population dynamics and
asynchrony at fine spatial scales: a case history of sockeye salmon (Oncorhynchus nerka] population
structure in Alaska, USA. Canadian Journal of Fish and Aquatic Science 69:297-306.
Rains, M. C. 2011. Water sources and hydrodynamics of closed-basin depressions, Cook Inlet Region,
Alaska. Wetlands 31:377-387.
Rand, P. S., B. A. Berejikian, A. Bidlack, D. Bottom, J. Gardner, M. Kaeriyama, R. Lincoln, M. Nagata, T. N.
Pearsons, M. Schmidt, W. W. Smoker, L. A. Weitkamp, and L. A. Zhivitovsky. 2012. Ecological
interactions between wild and hatchery salmonids and key recommendations for research and
management actions in selected regions of the North Pacific. Environmental Biology of Fishes
94:343-358.
RAP (Riverscape Analysis Project). 2011. Available: http://rap.ntsg.umt.edu. Accessed: December 2011.
Rich, H. B., Jr., T. P. Quinn, M. D. Scheuerell, and D. E. Schindler. 2009. Climate and intraspecific
competition control the growth and life history of juvenile sockeye salmon (Oncorhynchus nerka] in
Iliamna Lake, Alaska. Canadian Journal of Fisheries and Aquatic Sciences 66(2):238-246.
Bristol Bay Assessment 15-8 January 2014
-------
Chapter 15 References
Rogers, L. A., and D. E. Schindler. 2011. Scale and the detection of climatic influences on the productivity
of salmon populations. Global Change Biology 17:2546-2558.
Ruff, C. P., D. E. Schindler, J. B. Armstrong, K. T. Bentley, G. T. Brooks, G. W. Holtgrieve, M. T. McGlauflin, C.
E. Torgersen, and J. E. Seeb. 2011. Temperature-associated population diversity in salmon confers
benefits to mobile consumers. Ecology 92:2073-2084.
Salathe, E. P. 2006. Influences of a shift in North Pacific storm tracks on western North America
precipitation under global warming. Geophysical Research Letters 33(19):L19820.
Schindler, D. E., R. Hilborn, B. Chasco, C. P. Boatright, T. P. Quinn, L. A. Rogers, and M. S. Webster. 2010.
Population diversity and the portfolio effect in an exploited species. Nature 465:609-612.
Selkregg, L. L. 1974. Southwest Alaska, Alaska Regional Profiles. Anchorage, AK: University of Alaska,
Arctic Environmental Information and Data Center.
Serreze, M. C., J. E. Walsh, F. S. Chapin III, T. Osterkamp, M. Dyurgerov, V. Romanovsky, W. C. Oechel, J.
Morison, T. Zhang, and G. Barry. 2000. Observational evidence of recent change in the northern high-
latitude environment. Climatic Change 46(1-2):159-207.
Shallin Busch, D., M. Sheer, K. Burnett, P. McElhany, and T. Cooney. 2011. Landscape-level model to
predict spawning habitat for lower Columbia River fall Chinook salmon (Oncorhynchus tshawytscha}.
River Research and Applications 29(3):297-312.
SNAP (Scenarios Network for Alaska and Arctic Planning). 2012. Historical and Projected Decadal
Averages of Annual Total and Annual Mean Temperature, Precipitation, and Evapotranspiration.
Available: http://www.snap.uaf.edu/data.php. Accessed: February 7, 2013.
Stafford, J. M., G. Wendler, and J. Curtis. 2000. Temperature and precipitation of Alaska: 50 year trend
analysis. Theoretical and Applied Climatology 67(l-2):33-44.
Stanford, J. A., M. S. Lorang, and F. R. Hauer. 2005. The shifting habitat mosaic of river ecosystems.
International Association of Theoretical and Applied Limnology—Proceedings 29:123-136.
Steen, R. P., and T. P. Quinn. 1999. Egg burial depth by sockeye salmon (Oncorhynchus nerka}:
implications for survival of embryos and natural selection on female body size. Canadian Journal of
Zoology 77(5y.836-841.
Stewart, I. T., D. R. Cayan, and M. D. Dettinger. 2004. Changes in snowmelt runoff timing in western
North America under a 'business as usual' climate change scenario. Climatic Change 62:217-232.
Stilwell, K. B., and D. S. Kaufman. 1996. Late Wisconsin glacial history of the northern Alaska Peninsula.
Arctic and Alpine Research 28:475-487.
Tang, J., M. D. Bryant, and E. L. Brannon. 1987. Effect of temperature extremes on the mortality and
development rates of coho salmon embryos and alevins. Progressive Fish-Culturist 49:167-174.
Bristol Bay Assessment 15-9 January 2014
-------
Chapter 15 References
Taylor, E. B. 1990. Environmental correlates of life-history variation in juvenile Chinook salmon,
Oncorhynchus tshawytscha (Walbaum). Journal of Fish Biology 37:1-17.
USGS (U.S. Geological Survey). 2010. Significant United States Earthquakes, 1568-2009. June. Available:
http://nationalatlas.gov/atlasftp.html?openChapters=chpgeol#chpgeol. File (quksigx020.tar.gz).
Accessed: April 13, 2012.
USGS (U.S. Geological Survey). 2012. National Hydrography Dataset. High Resolution, Alaska. Available:
ftp://nhdftp.usgs.gov/DataSets/Staged/States/FileGDB/HighResolution. Accessed: October 16,
2012.
USGS (U.S. Geological Survey). 2013. National Elevation Dataset (2 arc second), ArcGrid. August 26.
[n59wl55.zip, n59wl56.zip, n59wl57.zip, n59wl58.zip, n60wl54.zip, n60wl55.zip, n60wl56.zip,
n60wl57.zip, n60wl58.zip, n60wl59.zip, n60wl60.zip, n61wl53.zip, n61wl54.zip, n61wl55.zip,
n61wl56.zip, n61wl57.zip, n61wl58.zip, n61wl59.zip, n61wl60.zip] Available:
http: //viewer.nationalmap.gov/viewer/.
Wahrhaftig, C. 1965. Physiographic Divisions of Alaska. Geological Survey Professional Paper 482. U.S.
Geological Survey.
Walsh, J. E., W. L. Chapman, V. Romanovsky, J. H. Christensen, and M. Stendel. 2008. Global climate model
performance over Alaska and Greenland, journal of Climate 21:6156-6174.
Webb B. W., and F. Nobilis. 1997. Long-term perspective on the nature of the air-water temperature
relationship: a case study. HydrologicalProcesses 11(2):137-147.
Wendler, G., L. Chen, and B. Moore. 2012. The first decade of the new century: a cooling trend for most of
Alaska. The Open Atmospheric Science Journal 6.
Winter, T. C. 2001. The concept of hydro\ogic \andscapes. Journal of the American Water Resources
Association 37:335-349.
Wissmar R. C., R. K. Timm, and M. D. Bryant. 2010. Radar-derived digital elevation models and field-
surveyed variables to predict distributions of juvenile coho salmon and Dolly Varden in remote
streams of Alaska. Transactions of the American Fisheries Society 139:288-302.
Zamzow, K. 2011. Investigations of Surface Water Quality in the Nushagak, Kvichak, and Chulitna
Watersheds, Southwest Alaska 2009-2010. Sutton, AK: Center for Science in Public Participation.
Personal Communications
Wiedmer, M. Malma Consulting. September 2011—conversation with Joe Ebersole regarding
distribution of salmon across the Nushagak and Kvichak River watersheds and factors controlling
that distribution.
Bristol Bay Assessment 15-10 January 2014
-------
Chapter 15 References
15.1.5 Chapter 4—Type of Development
Alto, K., S. Broderius, and L. L. Smith, Jr. 1977. Toxicity ofXanthates to Freshwater Fish and Invertebrates.
Prepared for State of Minnesota Environmental Quality Council Regional Copper Nickel Study.
Anderson, E. D., R. G. Eppinger, and K. D. Kelley. 2009. Using regional geochemistry, geology,
aeromagnetics, Landsat, and digital elevation models (DEM) to define favourable areas for
porphyry-style mineralization in southwestern Alaska. Pages 345-348 in D. R. Lentz, K. G. Thorne, K.
L. Beal (eds.), 24th International Applied Geochemistry Symposium, Fredericton, New Brunswick,
Canada. Association of Applied Geochemists.
Blight, G. E. 2010. Geotechnical Engineering for Mine Waste Storage Facilities. Boca Raton, FL: CRC Press.
Borden, R. K. 2003. Environmental geochemistry of the Bingham Canyon porphyry copper deposit, Utah.
Environmental Geology 43(7):752-758.
Brodie, M. ]., L. M. Broughton, and A. M. Robertson. 1991.^4 Conceptual Rock Classification System for
Waste Management and a Laboratory Method for ARD Prediction from Rock Piles. Preprint for the 2nd
International Conference on the Abatement of Acid Drainage, Montreal, Quebec, September 16-18,
1991.
Commonwealth of Australia. 2007. Tailings Management - Leading Practice Sustainable Development
Program for the Mining Industry. Australian Government, Department of Industry, Tourism and
Resources.
Davies, M. P. 2002. Tailings impoundment failures: are geotechnical engineers listening? Geotechnical
News. September: 31-36.
Davies, M. P. 2011. Filtered dry stacked tailings - the fundamentals. Pages 189-197 in D. van Zyl (ed.),
Tailings and Mine Waste 2011: Proceedings of the 15th International Conference on Tailings and Mine
Waste, Vancouver, Canada, November 6-9, 2011.
Fuerstenau, M. C. 2003. Principles of Mineral Processing. Littleton, CO: Society for Mining, Metallurgy &
Exploration, Incorporated.
Fuerstenau, M. C., G. Jameson, and R. H. Yoon (eds.). 2007. Froth Flotation: A Century of Innovation.
Littleton, CO: Society of Mining, Metallurgy, and Exploration.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company), Vancouver, BC.
ICOLD (International Commission on Large Dams). 2001. Tailings Dams, Risk of Dangerous Occurrences,
Lessons Learnt from Practical Experiences. United Nations Environmental Programme, Bulletin 121.
Bristol Bay Assessment 15-11 January 2014
-------
Chapter 15 References
John, D. A., R. A. Ayuso, M. D. Barton, R. J. Blakely, R. J. Bodnar, J. H. Dilles, F. Gray, F. T. Graybeal, J. C.
Mars, D. K. McPhee, R. R. Seal, R. D. Taylor, and P. G. Vikre. 2010. Porphyry Copper Deposit
Mode/.Scientific Investigations Report 2010-5070-B. Reston, VA: U.S. Geological Survey.
Kesler, S. E., S. L. Chryssoulis, and G. Simon. 2002. Gold in porphyry copper deposits: its abundance and
fate. Ore Geology Reviews 21(1-2):103-124.
Khoshdast, H., and A. Sam. 2011. Flotation frothers: review of their classifications, properties and
preparation. The Open Mineral Processing Journal 2011(4):25-44.
Koerner, R. M., Y. G. Hsuan, and G. R. Koerner. 2011. GRI White Paper #6 on Geomembrane Lifetime
Prediction: Unexposed and Exposed Conditions, Updated 2011. Available: http://www.geosynthetic-
institute.org/papers/paper6.pdf. Accessed: April 20, 2012.
Levit, S., and D. M. Chambers. 2012. Comparison of the Pebble Mine with Other Alaska Large Hard Rock
Mines. Sutton, AK: Center for Science in Public Participation.
Lipman, P. W., and D. A. Sawyer. 1985. Mesozoic ash-flow caldera fragments in southeastern Arizona and
their relation to porphyry copper deposits. Geology 13:652-656.
Logsdon, M. J., K. Hagelstein, and T. I. Mudder. 1999. The Management of Cyanide in Gold Extraction.
Ottawa, ON: International Council on Metals and the Environment.
Lusty, P. A. J., and S. D. Hannis. 2009. Commodity Profile: Copper. Internal Report OR/09/041. Keyworth,
Nottingham: British Geological Survey.
Marsden, J. 0., and C. I. House. 2006. The Chemistry of Gold Extraction. 2nd edition. Littleton, CO: Society
of Mining, Metallurgy, and Exploration.
Martin, T. E., M. P. Davies, S. Rice, T. Higgs, and P. C. Lighthall. 2002. Stewardship of Tailings Facilities.
Commissioned by the MMSD project of the International Institute for the Environment and
Development (IIEA) and the World Business Council for Sustainable Development (WBCSD),
London, England.
McLeod, H., and L. Murray. 2003. Tailings Dam Versus a Water Dam, What Is the Design Difference? ICOLD
Symposium on Major Challenges in Tailings Dams, June 15, 2003.
National Inventory of Dams. 2005.2005 United States National Inventory of Dams Database. Available:
http://geo.usace.army.mil/pgis/f?p=397:l:0::NO. Accessed: October 6, 2005.
Novagold. 2012. Donlin Creek Gold Project Technical Report. January 20.
O'Kane, M., and C. Wels. 2003. Mine Waste Cover System Design - Linking Predicted Performance to
Groundwater and Surface Water Impacts. Pages 341-349 in Proceedings of the 6th Annual
International Conference on Acid Rock Drainage, Cairns, Queensland, Australia, July 12-18, 2003.
Bristol Bay Assessment 15-12 January 2014
-------
Chapter 15 References
Porter, K. E., and D. I. Bleiwas. 2003. Physical Aspects of Waste Storage from a Hypothetical Open Pit
Porphyry Copper Operation. Open File Report 03-143, Version 1.0. U.S. Geological Survey.
Price, W. A. 2009. Prediction Manual for Drainage Chemistry from Sulphidic Geologic Materials. 579.
Report prepared by CANMET - Mining and Mineral Sciences Laboratories for the Mine Environment
Neutral Drainage (MEND) Program, Natural Resources Canada.
Price, W. A., and J. C. Errington. 1998. Guidelines for Metal Leaching and Acid Rock Drainage at Mine Sites
in British Columbia. British Columbia Ministry of Energy and Mines.
Rico, M. G., G. Benito, and A. Diez-Herrero. 2008. Floods from tailings dam failures. Journal of Hazardous
Materials 154(l-3):79-87.
Rowe, R. K. 2005. Long-term performance of contaminant barrier systems. Geotechnique 55(9):631-678.
Available: http://www.jcigs.org/8ICG/SL/Geotechnique%20Rankine%20V55%20N9%20631-
678%20Rowe%202005.pdf. Accessed: October 12, 2012.
Scott Wilson Mining. 2005. Guelb Moghrein Copper Gold Project, Environmental and Social Impact
Assessment, Akjoujt, Mauritania.
Singer, D. A., V. I. Berger, and B. C. Moring. 2008. Porphyry Copper Deposits of the World: Database and
Grade and Tonnage Models, 2008. Open File Report 2008-1155. U.S. Geological Survey.
Singer, D. A., W. D. Menzie, D. M. Sutphin, D. L. Mosier, and). D. Bliss. 2001. Mineral Deposit Density- An
Update. U.S. Geological Survey Professional Paper 1640-A. Reston, VA: U.S. Geological Survey.
Slaughter, C. B., G. W. Freethey, and L. E. Spangler. 1995. Hydrology of the North Fork of the Right Fork of
Millar Creek, Carbon Country, Utah, before, during, and after underground coal mining. Water
Resources Investigation Report 95-4025. Denver, CO: U.S. Geological Survey. Available:
http://pubs.usgs.gov/wri/1995/4025/report.pdf.
SME (Society for Mining, Metallurgy, and Exploration). 2011. SME Mining Engineering Handbook. 3rd
edition. Darling, P. (ed.). Denver, CO.
Stange, W. 1999. The process design of gold leaching and carbon-in-pulp circuits. The Journal of the
South African Institute of Mining and Metallurgy January/February (1999):13-25.
State of Idaho. 1992. Mill Tailings Impoundments. Pages 16-23 inBest Management Practices for Mining
in Idaho. Idaho Department of Lands. Available:
http://www.idl.idaho.gov/Bureau/Minerals/bmp_manuall992/pl6-ch4.pdf. Accessed: September
18,2011.
USAGE (U.S. Army Corps of Engineers). 2011. Permit # POA-1992-574-M19, Fish Creek. Issued March 4,
2011.
Bristol Bay Assessment 15-13 January 2014
-------
Chapter 15 References
USDA (U.S. Department of Agriculture). 1995. Anatomy of a Mine from Prospect to Production. General
Technical Report INT-GTR-35 revised. Ogden, UT: Forest Service, Intermountain Research Station.
Available: http://www.fs.fed.us/geology/anatomy_mine.pdf.
USEPA (U.S. Environmental Protection Agency). 1994a. Technical Resource Document: Extraction and
Beneficiation of Ores and Minerals, Volume 4, Copper. EPA 530-R-94-031
USEPA (U.S. Environmental Protection Agency). 1994b. Technical Report: Design and Evaluation of
Tailings Dams. EPA 530-R-94-038. Washington, DC: Office of Solid Waste.
Vick, S. G. 1983. Planning, Design, and Analysis of Tailings Dams. New York, NY: John Wiley & Sons.
Vigneault, B., M. Desforges, and J. Geer. 2009. Mining Reagents and By-Products (e.g. Thiosalts) as
Potential Toxicants in Mine Effluents. Metal Mining Investigation of Cause Workshop, Session 2:
Potential Causes of Effects. Available: https://www.ec.gc.ca/esee-
eem/default.asp?lang=En&n=495426Dl-l&offset=6&toc=show.
Whittaker, B. N., and D. J. Reddish. 1989. Subsidence—Occurrence, Prediction and Control. Amsterdam,
Netherlands: Elsevier.
15.1.6 Chapter 5—Endpoints
ADF&G (Alaska Department of Fish and Game). 2011. Moose Species Profile. Juneau, AK.
ADF&G (Alaska Department of Fish and Game). 2012. Alaska Freshwater Fish Inventory Database.
Available: http://www.adfg.alaska.gov/index.cfm?adfg=ffinventory.main.
Alaska Department of Health and Social Services. 2005. The Prevalence of Diabetes Among Adult
Alaskans, 2002-2004. Epidemiology Bulletins 1-8.
Alaska Federation of Natives. 2010. The Right to Subsist—Federal Protection of Subsistence in Alaska.
Available: http://www.nativefederation.org/wp-content/uploads/2012/10/2010-april8-the-right-
to-subsistence.pdf.
Armstrong, R. H., and J. E. Morrow. 1980. The Dolly Varden charr, Salvelinus malma. Pages 99-140 in E.
K. Balon (ed.), Charrs: Salmonid Fishes of the Genus Salvelinus. Springer.
Augerot,X. 2005.Atlasof Pacific Salmon: The First Map-Based Status Assessment of Salmon in the North
Pacific. Portland, OR.
Baker, T. T., L. F. Fair, R. A. Clark, and J. J. Hasbrouck. 2009. Review of Salmon Escapement Goals in Bristol
Bay, Alaska, 2009. Alaska Department of Fish and Game, Division of Sport Fish, Research and
Technical Services.
Ballew, C., A. Ross, R. S. Wells, V. Hiratsuka, K. J. Hamrick, E. D. Nobmann, and S. Bartell. 2004. Final
Report on the Alaska Traditional Diet Survey. Anchorage, AK.
Bristol Bay Assessment 15-14 January 2014
-------
Chapter 15 References
Bartz, K. K., and R. J. Naiman. 2005. Effects of salmon-borne nutrients on riparian soils and vegetation in
Southwest Alaska. Ecosystems 8:529-545.
Bersamin, A., S. Zidenberg-Cherr, J. Stern, and B. Luick. 2007. Nutrient intakes are associated with
adherence to a traditional diet among Yup'ik Eskimos living in remote Alaska native communities:
the CANHR Study. International Journal of Circumpolar Health 66:62-70.
Blair, G. R., D. E. Rogers, and T. P. Quinn. 1993. Variation in life history characteristics and morphology of
sockeye salmon in the Kvichak River system, Bristol Bay, Alaska. Transactions of the American
Fisheries Society 122:550-559.
Brna, P. J., and L. A. Verbrugge (eds.). 2013. Wildlife Resources of the Nushagak and Kvichak River
Watersheds, Alaska. Final Report. Anchorage, AK: Anchorage Fish and Wildlife Field Office, U.S. Fish
and Wildlife Service.
Burger, C. V., and L. A. Gwartney. 1986.^4 Radio Tagging Study of Naknek Drainage Rainbow Trout.
Anchorage, AK: U.S. National Park Service, Alaska Regional Office.
Butler, L. G. 2004. Unit 9 moose management report. Pages 113-120 in C. Brown, (ed.), Moose
Management Report of Survey and Inventory Activities: 1 July 2001-30 June 2003. Juneau, AK: Alaska
Department of Fish and Game.
Butler, L. G. 2009. Unit 9 and 10 Wolf Management Report. Pages 66-70 in P. Harper (ed.), Wolf
Management Report of Survey and Inventory Activities 1 July 2005-30 June 2008. Juneau, AK: Alaska
Department of Fish and Game.
Cederholm C. J., M. D. Kunze, T. Murota, and A. Sibatani. 1999. Pacfic salmon carcasses: essential
contributors of nutrients and energy for aquatic and terrestrial ecosystems. Fisheries 24:6-15.
Chan, H., K. Fediuk, S. Hamilton, L. Rostas, H. Caughey, G. Egeland, and E. Loring. 2006. Food security in
Nunavut, Canada: barriers and recommendations. International Journal of Circumpolar Health
65:416-431.
Christie, K. S., and T. E. Reimchen. 2008. Presence of salmon increases passerine density on Pacific
Northwest streams. Auk 125:51-59.
Colombi, B. J., and J. F. Brooks. 2012. Keystone Nations: Indigenous Peoples and Salmon Across the North
Pacific. Santa Fe, NM: School for Advanced Research.
Crane, P., M. Lisac, B. Spearman, E. Kretschmer, C. Lewis, S. Miller, and J. Wenburg. 2003. Development
and Application of Microsatellites to Population Structure and Mixed-Stock Analyses of Dolly Varden
from the Togiak River Drainage, Final Report (Study No. 00-011). Anchorage, AK: U.S. Fish and
Wildlife Service, Conservation Genetics Laboratory.
Bristol Bay Assessment 15-15 January 2014
-------
Chapter 15 References
Cunningham, C. ]., R. Hilborn, J. Seeb, and T. Branch. 2012. Reconstruction of Bristol Bay Sockeye Salmon
Returns Using Age and Genetic Composition of Catch. AFS-UW-1202. Seattle, WA: University of
Washington, School of Aquatic and Fishery Sciences.
Darimont, C. T., T. E. Reimchen, and P. C. Paquet. 2003. Foraging behavior by gray wolves on salmon
streams in coastal British Columbia. Canadian Journal of Zoology 81:349-353.
DeCicco, A. L. 1992. Long-distance movements of anadromous Dolly Varden between Alaska and the
USSR.j4rctzc45(2):120-123.
DeCicco, A. L. 1997. Movements of postsmolt anadromous Dolly Varden in northwestern Alaska.
American Fisheries Symposium 19:175-183.
Denton, K. P., H. B. Rich, Jr., and T. P. Quinn. 2009. Diet, movement, and growth of Dolly Varden in
response to sockeye salmon subsidies. Transactions of the American Fisheries Society 138(6) :1207-
1219.
Dewailly, E., C. Blanchet, S. Lemieux, L. Sauve, S. Gingras, P. Ayotte, and B. J. Holub. 2001. n-3 Fatty acids
and cardiovascular disease risk factors among the Inuit of Nunavik. The American Journal of Clinical
Nutrition 74:464-473.
Dewailly, E., C. Blanchet, S. Gingras, S. Lemieux, and B. J. Halub. 2002. Cardiovascular disease risk factors
and n-3 fatty acid status in the adult population of James Bay Cree. The American Journal of Clinical
Nutrition 76:85-92.
Din, J. N., D. E. Newby, and A. D. Flapan. 2004. Omega 3 fatty acids and cardiovascular disease—fishing
for a natural treatment. British Medical Journal 328:30-35.
Dittman, A. H., and T. P. Quinn. 1996. Homing in Pacific salmon: mechanisms and ecological basis. Journal
of Experimental Biology 199:83-91.
Duhaime, G., and N. Bernard (eds.). 2008. Arctic Food Security. Canadian Circumpolar Institute (CCI)
Press, University of Alberta and Centre inter-universitaire d'etudes et de recherche autochtones
(CIERA), Universite Laval.
Dye, J. E., and C. J. Schwanke. 2009. Report to the Alaska Board of Fisheries for the Recreational Fisheries of
Bristol Bay, 2007,2008, and 2009. Alaska Department of Fish and Game, Division of Sport Fish,
Research and Technical Services.
Dye, J. E., C. J. Schwanke, and T. A. Jaecks. 2006. Report to the Alaska Board of Fisheries for the
Recreational Fisheries of Bristol Bay, 2004,2005, and 2006. Special Publication No. 06-29. Alaska
Department of Fish and Game, Divisions of Sport Fish and Commercial Fisheries.
Ebbesson, S. 0. E., and M. T. Tejero. 2007. Fatty acid consumption and metabolic syndrome components:
the GOCADAN study. Journal of the Cardiometabolic Syndrome 2(4):244-249.
Bristol Bay Assessment 15-16 January 2014
-------
Chapter 15 References
Eliason, E., T. Clark, M. Hague, L. Hanson, Z. Gallagher, K. Jeffries, M. Gale, D. Patterson, S. Hinch, and A.
Farrell. 2011. Differences in thermal tolerance among sockeye salmon populations. Science
332:109-112.
Fair, L. F. 2003. Critical elements of Kvichak River sockeye salmon management. Alaska Fishery Research
Bulletin 10:95-103.
Fair, L. F., C. E. Barzil, X. Zyhang, R. A. Clark, and J. W. Erickson. 2012. Review of Salmon Escapement Goals
in Bristol Bay, Alaska, 2012. Fishery Manuscript Series No. 12-0. Anchorage, AK: Alaska Department
of Fish and Game.
Fall, J. A., D. L. Holen, B. Davis, T. Krieg, and D. Roster. 2006. Subsistence Harvests and Uses of Wild
Resources in Iliamna, Newhalen, Nondalton, Pedro Bay, and Port Alsworth, Alaska 2004. Technical
Paper No. 302. Juneau, AK: Alaska Department of Fish and Game, Division of Subsistence.
Fall, J. A., T. M. Krieg, and D. Holen. 2009. Overview of the Subsistence Fishery of the Bristol Bay
Management Area. Special Publication No. EOF 2009-07. Anchorage, AK: Alaska Department of Fish
and Game, Division of Subsistence.
Fancy, S. G. 1983. Movements and activity budgets of caribou near oil drilling sites in the Sagavanirktok
River Floodplain, Alaska. Arctic 36:193-197.
Gende, S. M., and M. F. Willson. 2001. Passerine densities in riparian forests of Southeast Alaska:
potential effects of anadromous spawning salmon. Condor 103:624-629.
Gende, S. M., R. T. Edwards, M. F. Willson, and M. S. Wipfli. 2002. Pacific salmon in aquatic and terrestrial
ecosystems. BioScience 52(10):917-928.
Gerrard J. M., P. Gerrard, W. J. Maher, and D. W. A Whitfield. 1975. Factors influencing nest site selection
of Bald Eagles in northern Saskatchewan and Manitoba. Blue Jay 33:169-176.
Goldsmith, D., A. Hill, T. Hull, M. Markowski, and R. Unswort. 1998. Economic Assessment of Bristol Bay
Area National Wildlife Refuges: Alaska Peninsula, Becherof, Izembek, Togiak. U.S. Department of the
Interior, Fish and Wildlife Service.
Gresh, T., J. Lichatowich, and P. Schoonmaker. 2000. An estimation of historic and current levels of
salmon production in the northwest Pacific ecosystem: evidence of a nutrient deficit in the
freshwater systems of the Pacific Northwest. Fisheries 25(1):15-21.
Gustafson, R. G., R. S. Waples, J. M. Myers, L. A. Weitkamp, G. J. Bryant, 0. W. Johnson, and J. J. Hard. 2007.
Pacific salmon extinctions: quantifying lost and remaining diversity. Conservation Biology 21:1009-
1020.
Gwartney, L. A. 1985. Naknek Drainage Rainbow Trout Study in the Katmai National Park and Preserve.
King Salmon, AK: Alaska Department of Fish and Game, Division of Sport Fish, and U.S. Department
of Interior, National Park Service.
Bristol Bay Assessment 15-17 January 2014
-------
Chapter 15 References
Hansen, A. J. 1987. Regulation of bald eagle reproductive rates in southeast Alaska. Ecology 68:1387-
1392.
Hartman, W. L., and R. L. Burgner. 1972. Limnology and fish ecology of sockeye salmon nursery lakes of
the world. Journal of the Fisheries Research Board of Canada 29:699-715.
Heard, W. R. 1991. Life history of pink salmon (Oncorhynchus gorbuscha). Pages 119-230 in C. Groot and
L. Margolis (eds.), Pacific Salmon Life Histories. Vancouver, BC: UBC Press
Helfield, J. M., and R. J. Naiman. 2006. Keystone interactions: salmon and bear in riparian forests of
Alaska. Ecosystems 9:167-180.
Hilborn, R. 2006. Fisheries success and failure: the case of the Bristol Bay salmon fishery. Bulletin of
Marine Science 78(3):487-498.
Hilborn, R., T. Quinn, D. Schindler, and D. Rogers. 2003. Biocomplexity and fisheries sustainability.
Proceedings of the National Academy of Sciences of the United States of America 100:6564-6568.
Holen, D., and T. Lemons. 2010. Subsistence Harvests and Uses of Wild Resources in Lime Village, Alaska,
2007. Technical Paper No. 355. Anchorage, AK: Alaska Department of Fish and Game, Division of
Subsistence.
Holen, D., T. M. Krieg, and T. Lemons. 2011. Subsistence Harvests and Uses of Wild Resources in King
Salmon, Naknek, and South Naknek, Alaska, 2007. Technical Paper No. 360. Anchorage, AK: Alaska
Department of Fish and Game, Division of Subsistence.
Holen, D., J. Stariwat, T. M. Krieg, and T. Lemons. 2012. Subsistence Harvest and Uses of Wild Resources in
Aleknagik, Clark's Point, and Manokotak, Alaska, 2008. Technical Paper No. 368. Anchorage, AK:
Alaska Department of Fish and Game, Division of Subsistence.
Janetski, D. J., D. T. Chaloner, S. D. Tiegs, and G. A. Lamberti. 2009. Pacific salmon effects on stream
ecosystems: a quantitative synthesis. Oecologia 159(3):583-595.
Johnson, J., and P. Blanche. 2012. Catalog of Waters Important for Spawning, Rearing, or Migration of
Anadromous Fishes—Southwestern Region, Effective June 1, 2012. Special Publication No. 12-08.
Anchorage, AK: Alaska Department of Fish and Game.
Johnson, S. W., J. F. Thedinga, and A. S. Feldhausen. 1994. Juvenile salmonid densities and habitat use in
the main-stem Situk River, Alaska, and potential effects of glacial flooding. Northwest Science
68(4):319-325.
Johnson, J. S., E. Nobmann, E. Asay, and A. Lanier. 2009. Dietary intake of Alaska Native people in two
regions and implications for health: the Alaska Native Dietary and Subsistence Food Assessment
Project International Journal of Circumpolar Health 68:109-122.
Bristol Bay Assessment 15-18 January 2014
-------
Chapter 15 References
King, R. S., C. M. Walker, D. F. Whigham, S. Baird, and J. A. Back. 2012. Catchment topography and
wetland geomorphology drive macroinvertebrate community structure and juvenile salmonid
distributions in southcentral Alaska headwater streams. Freshwater Science 31(2):341-364.
Kline T. C., J. J. Goering, 0. A. Mathisen, P. H. Poe, P. L. Parker, and R. S. Scalan. 1993. Recycling of
elements transported upstream by runs of Pacific salmon. 2. 615N and 613C evidence in the Kvichak
River watershed, Bristol Bay, southwestern Alaska. Canadian Journal of Fisheries and Aquatic
Sciences 50(ll):2350-2365.
Krieg, T., M. Chythlook, P. Coiley-Kenner, D. Holen, K. Kamletz, and H. Nicholson. 2005. Freshwater Fish
Harvest and Use in Communities of the Kvichak watershed, 2003. Technical Data Report No. 297.
Juneau AK: Alaska Department of Fish and Game, Division of Subsistence.
Krieg, T. M., J. A. Fall, M. B. Chythlook, R. LaVine, and D. Koster. 2007. Sharing, Bartering, and Cash Trade
of Subsistence Resources in the Bristol Bay Area, Southwest Alaska. Technical Paper No. 326. Juneau,
AK: Alaska Department of Fish and Game, Division of Subsistence.
Krieg, T. M., D. L. Holen, and D. Koster. 2009. Subsistence Harvests and Uses of Wild Resources in Igiugig,
Kokhanok, Koliganek, Levelock, and NewStuyahok, Alaska 2005. Technical Paper No. 322. Dillingham,
AK: Alaska Department of Fish and Game, Division of Subsistence.
Krueger, C. C., M. J. Lisac, S. J. Miller, and W. H. Spearman. 1999. Genetic Differentiation of Rainbow Trout
fOncorhynchus mykissj in the Togiak National Wildlife Refuge, Alaska. Alaska Fisheries Technical
Report No. 55. Anchorage, AK: U.S. Fish and Wildlife Service, Fish Genetics Laboratory.
Larkin, G. A., and P. A. Slaney. 1997. Implications of trends in marine-derived nutrient influx to South
Coastal British Columbia salmonid production. Fisheries 22:16-24.
Lindley, S. T., C. B. Grimes, M. S. Mohr, W. Peterson. J. Stein, J. T. Anderson, L. W. Botsford, D. L. Bottom, C.
A. Busack, T. K. Collier, J. Ferguson, J. C. Garza, A. M. Grover, D. G. Hankin, R. G. Kope, P. W. Lawson, A.
Low, R. B. MacFarlane, K. Moore, M. Palmer-Zwahlen, F. B. Schwing, J. Smith, C. Tracy, R. Webb, B. K.
Wells, and T. H. Williams. 2009. What Caused the Sacramento River Fall Chinook Stock Collapse?
NOAA-TM-NMFS-SWFSC-447. National Oceanic and Atmospheric Administration, National Marine
Fisheries Service, Southwest Fisheries Science Center.
Lisac, M. J. 2009. Seasonal Distribution and Biological Characteristics of Dolly Varden in the Goodnews
River, Togiak National Wildlife Refuge, Alaska, 2005-2006. Alaska Fisheries Technical Report No. 103.
Dillingham, AK: U.S. Fish and Wildlife Service, Togiak National Wildlife Refuge.
Lisac, M. J., and R. D. Nelle. 2000. Migratory Behavior and Seasonal Distribution of Dolly Varden Salvelinus
malma in the Togiak River Watershed, Togiak National Wildlife Refuge. Dillingham, AK: U.S. Fish and
Wildlife Service.
Bristol Bay Assessment 15-19 January 2014
-------
Chapter 15 References
McGlauflin, M. T., D. E. Schindler, L. W. Seeb, C. T. Smith, C. Habicht, and J. E. Seeb. 2011. Spawning habitat
and geography influence population structure and juvenile migration timing of sockeye salmon in
the Wood River lakes, Alaska. Transactions of the American Fisheries Society 140:763-782.
Meka, J. M., E. E. Knudsen, D. C. Douglas, and R. B. Benter. 2003. Variable migratory patterns of different
adult rainbow trout life history types in a southwest Alaska watershed. Transactions of the American
Fisheries Society 132:717-732.
Miller, S. D., and D. W. McCollum. 1994. Nonresident Hunters: Their Characteristics and Attitudes toward
Wildlife. Anchorage, AK: Alaska Department of Fish and Game.
Miller, S. D., G. C. White, R. A. Sellers, H. V. Reynolds, J. W. Schoen, K. Titus, and V. G. Barnes. 1997. Brown
and black bear density estimation in Alaska using radiotelemetry and replicated mark resight
techniques. Wildlife Monographs 133:1-55.
Milner, A. M., and R. G. Bailey. 1989. Salmonid colonization of new streams in Glacier Bay National Park,
Alaska, USA. Aquaculture and Fisheries Management 20:179-192.
Minard, R. E., M. Alexandersdottir, and S. Sonnichsen. 1992. Estimation of Abundance, Seasonal
Distribution, and Size and Age Composition of Rainbow Trout in the Kvichak River, Alaska, 1986-1991.
Fishery Data Series No. 92-51. Anchorage, AK: Alaska Department of Fish and Game, Division of
Sport Fish.
Moore, J., and D. Schindler. 2004. Nutrient export from freshwater ecosystems by anadromous sockeye
salmon (Oncorhynchus nerka}. Canadian Journal of Fisheries and Aquatic Sciences 61:1582-1589.
Morstad S., and C. E. Brazil. 2012. Kvichak River Sockeye Salmon Stock Status and Action Plan, 2012: A
Report to the Alaska Board of Fisheries. Anchorage, AK: Alaska Department of Fish and Game,
Division of Sport Fish, Research and Technical Services.
Murphy, N. J., C. D. Schraer, E. Boyko, L. Bulkow, B. Doty, and A. Lanier. 1995. Dietary change and obesity
associated with glucose intolerance in Alaska Natives. Journal of the American Dietetic Association
95:676-682.
Nehlsen, W., J. Williams, and J. Lichatowich. 1991. Pacific salmon at the crossroads—stocks at risk from
California, Oregon, Idaho, and Washington. Fisheries 16:4-21.
Nickelson, T. E., J. D. Rodgers, S. L. Johnson, and M. F. Solazzi. 1992. Seasonal changes in habitat use by
juvenile coho salmon (Oncorhynchus kisutch] in Oregon coastal streams. Canadian Journal of
Fisheries and Aquatic Sciences 49:783-789.
Nobmann, E. D., R. Ponce, C. Mattil, R. Devereux, B. Dyke, S. 0. E. Ebbesson, S. Laston, J. MacCluer, D.
Robbins, T. Romenesko, G. Ruotolo, C. R. Wenger, and B. V. Howard. 2005. Dietary intakes vary with
age among Eskimo adults of northwest Alaska in the GOCADAN study, 2000-2003. The Journal of
Nutrition 135:856-862.
Bristol Bay Assessment 15-20 January 2014
-------
Chapter 15 References
NRC (National Research Council). 1996. Upstream: Salmon and Society in the Pacific Northwest.
Washington, DC.
Olson T., and J. Putera. 2007. Refining Techniques to Survey Harvested Brown Bear Populations in Katmai
National Park and Preserve and Lake Clark National Park and Preserve. January.
Ostberg, C. 0., S. D. Pavlov, and L. Mauser. 2009. Evolutionary relationships among sympatric life history
forms of Dolly Varden inhabiting the landlocked Kronotsky Lake, Kamchatka, and a neighboring
anadromous population. Transactions of the American Fisheries Society 138(1):1-14.
Pinsky, M. L., D. B. Springmeyer, M. N. Goslin, and X. Augerot. 2009. Range-wide selection of catchments
for Pacific salmon conservation. Conservation Biology 23:680-691.
PLP (Pebble Limited Partnership). 2011. Pebble Project Environmental Baseline Document 2004 through
2008. Anchorage, AK. Available: http://www.arlis.org/docs/vol2/Pebble/2004-2008EBDIndex.pdf.
Quinn, T. P. 2005. The Behavior and Ecology of Pacific Salmon and Trout. Seattle, WA: University of
Washington Press.
Quinn T. P., H. B. Rich, D. Gosse, and N. Schtickzelle, and J. Grant. 2012. Population dynamics and
asynchrony at fine spatial scales: a case history of sockeye salmon (Oncorhynchus nerka] population
structure in Alaska, USA. Canadian Journal of Fisheries and Aquatic Sciences 69(2):297-306.
Rahr, G., J. Lichatowich, R. Hubley, and S. Whidden. 1998. Sanctuaries for native salmon: a conservation
strategy for the 21st century. Fisheries 23:6-36.
Reynolds, J. B. 2000. Life History Analysis ofTogiak River Char through Otolith Microchemistry. Fairbanks,
AK: Alaska Cooperative Fish and Wildlife Research Unit, University of Alaska Fairbanks.
Ruggerone, G. T., R. M. Peterman, and B. Dorner. 2010. Magnitude and trends in abundance of hatchery
and wild pink salmon, chum salmon, and sockeye salmon in the North Pacific Ocean. Marine and
Coastal Fisheries: Dynamics, Management, and Ecosystem Science 2:306-328.
Russell, R. 1977. Rainbow Trout Life History Studies in Lower Talarik Creek - Kvichak Drainage.
Anchorage, AK: Alaska Department of Fish and Game.
Russell, R. 1980. A Fisheries Inventory of Waters in the Lake Clark National Monument Area. King Salmon,
AK: Alaska Department of Fish and Game, Division of Sport Fish.
Salo, E. 0.1991. Life history of chum salmon (Oncorhynchus ketd). Pages 231-310 in C. Groot and L.
Margolis (eds.), Pacific Salmon Life Histories. Vancouver BC: UBC Press.
Salomone, P., S. Morstad, T. Sands, M. Jones, T. Baker, G. Buck, F. West, and T. Kreig. 2011. Bristol Bay
Area Management Report. Anchorage, AK.
Scanlon, B. 2000. The Ecology of Arctic Char and the Dolly Varden in the BecharofLake Drainage, Alaska.
Fairbanks, AK: University of Alaska.
Bristol Bay Assessment 15-21 January 2014
-------
Chapter 15 References
Scheuerell, M., J. Moore, D. Schindler, and C. Harvey. 2007. Varying effects of anadromous sockeye
salmon on the trophic ecology of two species of resident salmonids in southwest Alaska. Freshwater
Biology 52:1944-1956.
Schindler, D. E., P. R. Leavitt, C. S. Brock, S. P. Johnson, and P. D. Quay. 2005. Marine-derived nutrients,
commercial fisheries, and production of salmon and lake algae in Alaska. Ecology 86:3225-3231.
Schindler, D. E., R. Hilborn, B. Chasco, C. P. Boatright, T. P. Quinn, L. A. Rogers, and M. S. Webster. 2010.
Population diversity and the portfolio effect in an exploited species. Nature 465:609-612.
Sepulveda, A. J., D. S. Rutz, S. S. Ivey, K. J. Dunker, and J. A. Gross. 2013. Introduced northern pike
predation on salmonids in southcentral Alaska. Ecology of Freshwater Fish 22(2):268-279.
Smith, R. J., K. A. Hobson, H. N. Koopman, and D. M. Lavigne. 1996. Distinguishing between populations of
freshwater and saltwater harbor seals (Phoca vitulina] using stable isotope ratios and fatty acid
profiles. Canadian Journal of Fisheries and Aquatic Sciences 53:272-279.
Solazzi, M., T. Nickelson, S. Johnson, and J. Rodgers. 2000. Effects of increasing winter rearing habitat on
abundance of salmonids in two coastal Oregon streams. Canadian Journal of Fisheries and Aquatic
Sciences 57-.906-914.
U.S. Census Bureau. 2010. Alaska QuickFacts from the United States Census Bureau. Available:
http://quickfacts.census.gov/qfd/states/02000.html.
USEPA (U.S. Environmental Protection Agency). 1998. Guidelines for Ecological Risk Assessment.
EPA/630/R-95/002F. April. Washington, DC: Office of Research and Development.
Valkenburg, P., R. A. Sellers, R. C. Squibb, J. D. Woolington, A. R. Aderman, and B. W. Dale. 2003.
Population dynamics of caribou herds in southwestern Alaska. Rangifer Special Issue 14:131-142.
Walsh, P., J. Reynolds, G. Collins, B. Russell, M. Winfree, and J. Denton. 2010. Application of a double-
observer aerial line-transect method to estimate brown bear population density in southwestern
Alaska. Journal of Fish and Wildlife Management 1:47-58.
Wentworth, C. 2007. Subsistence Migratory Bird Harvest Survey: Bristol Bay 2001-2005. Unpublished.
Anchorage, AK: U.S. Fish and Wildlife Service.
Willson, M. F., and K. C. Halupka. 1995. Anadromous fish as keystone species in vertebrate communities.
Conservation Biology 9:489-497.
Willson, M. F., S. M. Gende, and B. H. Marston. 1998. Fishes and the forest: expanding perspectives on
fish-wildlife interactions. Bioscience 48:455-462.
Wipfli, M. S., and C. V. Baxter. 2010. Linking ecosystems, food webs, and fish production: subsidies in
salmonid watersheds. Fisheries 35(8):373-387.
Bristol Bay Assessment 15-22 January 2014
-------
Chapter 15 References
Wolfe, R. ]., and R. J. Walker. 1987. Subsistence economies in Alaska: productivity, geography, and
development impacts. Arctic Anthropology 24(2):56-81.
Wong D., and C. Wentworth. 1999. Subsistence Waterfowl Harvest Survey, Bristol Bay 1995-1998.
Unpublished. Anchorage, AK: U.S. Fish and Wildlife Service.
Woody, C. A., and S. L. O'Neal. 2010. Fish Surveys in Headwater Streams of the Nushagak and Kvichak
River Drainages, Bristol Bay, Alaska, 2008-2010. Anchorage, AK: Fisheries Research and Consulting.
Woolington, J. D. 2004. Unit 17 moose management report. Pages 246-266 in C. Brown (ed.), Moose
Management Report of Survey and Inventory Activities: 1 July 2001-30 June 2003. Juneau, AK: Alaska
Department of Fish and Game.
Woolington, J. D. 2009. Mulchatna caribou management report, Units 9B, 17,18 south, 19A & 19B. Pages
11-31 in P. Harper (ed.), Caribou Management Report Of Survey And Inventory Activities, 1 July 2006-
1 July 2008. Juneau, AK: Alaska Department of Fish and Game.
Personal Communications
Baker, T. Area Fishery Research Biologist, Bristol Bay Salmon Program, Alaska Department of Fish and
Game, Anchorage, AK. July 8, 2011—email to Rebecca Shaftel.
15.1.7 Chapter 6—Mine Scenarios
ADEC (Alaska Department of Environmental Conservation). 2012. Situation Report: Fairbanks Gold
Mining, Inc. Cyanide Water Spill. Anchorage, AK: Alaska Department of Environment and
Conservation, Division of Spill Prevention and Response, Prevention and Emergency Response
Program.
ADF&G (Alaska Department of Fish and Game). 2013. Invasive Species. Overview.
http://www.adfg.alaska.gov/index.cfm?adfg=invasive.main. Accessed: March 4, 2013.
BC (Province of British Columbia, Canada). 2001. Ambient Water Quality Guidelines for Manganese.
Ministry of Environment.
BC (Province of British Columbia, Canada). 2008. Ambient Water Quality Guidelines for Iron. Ministry of
Environment.
Biesinger, K. E., and G. M. Christensen. 1972. Effects of various metals on survival, growth, reproduction,
and metabolism of Daphnia magna. Journal of the Fisheries Board of Canada 39:1691-1700.
Blight, G. E. 2010. Geotechnical Engineering for Mine Waste Storage Facilities. Boca Raton, FL: CRC Press.
Blight, G. E., and A. B. Fourie. 2003. A Review of Catastrophic Flow Failures of Deposit of Mine Waste and
Municipal Refuse. In L. Picarelli (ed.), Proceedings of the International Workshop on Occurrence and
Mechanisms of Flow-like Landslides in Natural Slopes and Earthfills, Sorrento, Italy, May 14-16,2003.
Bristol Bay Assessment 15-23 January 2014
-------
Chapter 15 References
Blowes, D. W., C. J. Ptacek, and J. Jurjovec. 2003. Mill tailings: hydrogeology and geochemistry. Pages 95-
116 in J. L. Jambar, D. W. Blowes, A. I. M. Ritchie (eds.), Environmental Aspects of Mine Wastes:
Mineralogical Association of Canada Short Course Series. Volume 31.
Boutet, C., and C. Chaisemartin. 1973. Specific toxic properties of metallic salts. Comptes Rendus Des
Seances De La Societe De Biologie EtDe Ses Filiales 167(12): 1933-1938.
CCME (Canadian Council of Ministers of the Environment). 1999. Canadian Water Quality Guidelines for
the Protection of Aquatic Life: Molybdenum. Winnipeg, MB.
CCME (Canadian Council of Ministers of the Environment). 2009. Canadian Water Quality Guidelines for
the Protection of Aquatic Life: Boron. Winnipeg, MB.
CEA (Chugach Electric Association). 2011. Facilities. Available: http://www.chugachelectric.com/inside-
chugach/the-company/facilities. Accessed: January 3, 2013.
Chapman, P.M., H. Bailey, and E. Canaria. 2000. Toxicity of total dissolved solids associated with two
mine effluents to chironomid larvae and early life stages of rainbow trout. Environmental Toxicology
and Chemistry 19:210-214.
Eddmaps. 2013. Species Information. Alaska Exotic Plant Information Clearinghouse. Available:
http://www.eddmaps.Org/alaska/species/.Accessed: March 4, 2013.
Environment Canada and Health Canada. 2010. Screening Assessment for the Challenge: Antimony
trioxide (antimony oxide). Ottawa, ON. Available: http://www.ec.gc.ca/ese-
ees/default.asp?lang=En&n=9889ABB5-l.
Environment Canada and Health Canada. 2011. Screening Assessment for the Challenge: Cobalt, Cobalt
Chloride, Sulfuric Acid, Cobalt (2+) Salt (1:1), and Sulfuric Acid, Cobalt Salt. Ottawa, ON. Available:
http://www.ec.gc.ca/ese-ees/default.asp?lang=En&n=8E18277B-l.
Etcheverry, D. 2012. Aquatic Invasive Plants: A New Focus for Alaskan Invasive Species Management.
Presentation at the 13th Annual Invasive Species Conference, University of Alaska Fairbanks.
http://www.uaf.edu/files/ces/cnipm/annualinvasivespeciesconference/13thAnnualMeetingProcee
dings/AISC-Elodea-D.Etcheverry_2012.pdf. Accessed: March 4, 2013.
Gammons, C. H., L. N. Harris, J. M. Castro, P. A. Cott, and B. W. Hanna. 2009. Creating lakes from open pit
mines: processes and considerations, with emphasis on northern environments. Canadian Technical
Report of Fisheries and Aquatic Sciences 2826.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company). Vancouver, BC.
Bristol Bay Assessment 15-24 January 2014
-------
Chapter 15 References
Goldman, C. 1960. Primary productivity and limiting factors in three lakes of the Alaska Peninsula.
Ecological Monographs 30:207-230.
Hidalgo, P., and I. G. R. Gutz. 2001. Determination of low concentrations of the flotation reagent ethyl
xanthate by sampled DC polarography and flow injection with the amperometric detection. Talanta
54:403-409.
HSWCD (Homer Soil and Water Conservation District). 2007. Managing Invasive Plants in Wetlands of the
Kenai Peninsula: Developing a Management Program of Reed Canarygrass Infestations, FY2007
Progress Report Summary. Homer, AK.
Laurence, D. 2011. Establishing a sustainable mining operation: an overview. Journal of Cleaner
Production 19(2-3) :278-284.
Marsik, T. 2009. Basic Study of Renewable Energy Alternatives for Electricity Generation in
Dillingham/Aleknagik Region. Prepared for Nushagak Electric and Telephone Cooperative, Inc., by
Bristol Bay Environmental Science Lab, University of Alaska Fairbanks, Bristol Bay Campus.
Merz, J. E., J. R. Smith, M. L. Workman, J. D. Setka, and B. Mulchaey. 2008. Aquatic macrophyte
encroachment in Chinook salmon spawning beds: lessons learned from gravel enhancement
monitoring in the Lower Mokelumne River, California. North American Journal of Fisheries
Management 28:1568-1577.
ML&P (Municipal Light and Power). 2012. About ML&P. Available:
http://www.mlandp.com/redesign/about_mlp.htm. Accessed: January 3, 2013.
Moore, J., and D. Schindler. 2004. Nutrient export from freshwater ecosystems by anadromous sockeye
salmon (Oncorhynchus nerka}. Canadian Journal of Fisheries and Aquatic Sciences 61:1582-1589.
NDM (Northern Dynasty Minerals). 2006. Pebble Project- Water Right Applications Submitted to the
Alaska Department of Natural Resources. Available:
http://dnr.alaska.gov/mlw/mining/largemine/pebble/water-right-apps/index.cfm.Accessed:
October 12, 2012.
PLP (Pebble Limited Partnership). 2011. Pebble Project Environmental Baseline Document 2004 through
2008. Anchorage, AK. Available: http://www.arlis.org/docs/vol2/Pebble/2004-2008EBDIndex.pdf.
Reeve, A. S., and M. B. Gracz. 2008. Simulating the hydrogeologic setting of peatlands in the Kenai
Peninsula Lowlands, Alaska. Wetlands 28:92-106.
Singer, D. A., V. I. Berger, and B. C. Moring. 2008. Porphyry Copper Deposits of the World: Database and
Grade and Tonnage Models, 2008. Open File Report 2008-1155. U.S. Geological Survey.
Swedish Chemicals Inspectorate. 2008. European Union Risk Assessment Report: Diantimony Trioxide Risk
Assessment. Luxembourg: Office for Official Publications of the European Communities.
USEPA (U.S. Environmental Protection Agency). 1976. Quality Criteria for Water. Washington, DC.
Bristol Bay Assessment 15-25 January 2014
-------
Chapter 15 References
USEPA (U.S. Environmental Protection Agency). 1980. Ambient Water Quality Criteria for Antimony.
EPA/440/5-80-020. Washington, DC: Office of Water.
USEPA (U.S. Environmental Protection Agency). 1997. EPA Can Do More to Minimize HardrockMine
Liabilities. EiDMF6-08-0016-7100223. June 11. Washington, DC: Office of the Inspector General.
USEPA (U.S. Environmental Protection Agency). 1998. Guidelines for Ecological Risk Assessment.
EPA/630/R-95/002F. April. Washington, DC: Office of Research and Development.
USEPA (U.S. Environmental Protection Agency). 2011. A Field-Based Aquatic Life Benchmark for
Conductivity in Central Appalachian Streams (Final Report). EPA/600/R-10/023F. Washington, DC.
USEPA (U.S. Environmental Protection Agency). 2013. ECOTOX User Guide: ECOTOXicology Database
System. Version 4.0. Available: http://www.epa.gov/ecotox. Accessed: October 2013.
USFWS (U.S. Fish and Wildlife Service). 2012. National Wetlands Inventory. Available:
http://www.fws.gov/wetlands/Wetlands-Mapper.html.Accessed: December 11, 2012.
USGS (U.S. Geological Survey). 2012a. National Hydrography Dataset. High Resolution, Alaska. Available:
ftp://nhdftp.usgs.gov/DataSets/Staged/States/FileGDB/HighResolution. Accessed: October 16,
2012.
USGS (U.S. Geological Survey). 2012b. Water Resources of Alaska. USGS Alaska Science Center Water
Resources Office. Last updated: December 12, 2012. Available:
http://alaska.usgs.gov/science/water/index.php. Accessed: April 17, 2013.
Wells, P. S. 2011. Long term in situ behaviour of oil sands fine tailings in Suncor's Pond 1A. Proceedings
from the Tailings and Mine Waste 2011 Conference, Vancouver, British Columbia, November 6-9, 2011.
15.1.8 Chapter 7—Mine Footprint
ADF&G (Alaska Department of Fish and Game). 2012. Alaska Freshwater Fish Inventory Database.
Available: http://www.adfg.alaska.gov/index.cfm?adfg=ffinventory.main.
ADF&G (Alaska Department of Fish and Game). 2Ql3.Anadromous Waters Catalog, Overview. Available:
http://www.adfg.alaska.gov/sf/SARR/AWC/. Accessed: April 9, 2013.
Alexander, R., E. Boyer, R. Smith, G. Schwarz, and R. Moore. 2007. The role of headwater streams in
downstream water quality. Journal of the American Water Resources Association 43:41-59.
Anderson, K. E., A. J. Paul, E. McCauley, L. J. Jackson, J. R. Post, and R. M. Nisbet. 2006. Instream flow
needs in streams and rivers: the importance of understanding ecological dynamics. Frontiers in
Ecology and the Environment 4:309-318.
Armstrong, J. B., and D. E. Schindler. 2013. Going with the flow: spatial distributions of juvenile coho
salmon track an annually shifting mosaic of water temperature. Ecosystems 16(8): 1429-1441.
Bristol Bay Assessment 15-26 January 2014
-------
Chapter 15 References
Armstrong, J. B., D. E. Schindler, K. L. Omori, C. P. Ruff, and T. P. Quinn. 2010. Thermal heterogeneity
mediates the effects of pulsed subsidies across a landscape. Ecology 91:1445-1454.
Arthington, A. H., S. E. Bunn, N. L. Poff, and R. J. Naiman. 2006. The challenge of providing environmental
flow rules to sustain river ecosystems. Ecological Applications 16:1311-1318.
Bartholow, J. M. 2010. Constructing an interdisciplinary flow regime recommendation. Journal of the
American Water Resources Association 46:892-906.
Beechie, T. J., E. Beamer, and L. Wasserman. 1994. Estimating coho salmon rearing habitat and smolt
production losses in a large river basin, and implications for habitat restoration. North American
Journal of Fisheries Management 14:797-811.
Blair, G. R., D. E. Rogers, and T. P. Quinn. 1993. Variation in life history characteristics and morphology of
sockeye salmon in the Kvichak River system, Bristol Bay, Alaska. Transactions of the American
Fisheries Society 122(4):55Q-559.
Bovee, K. D. 1982.^4 Guide to Stream Habitat Analysis Using the Instream Flow Incremental Methodology.
Instream Flow Group Information Paper No. 12, FWS/OBS-2/26. Fort Collins, CO: U.S. Fish and
Wildlife Service.
Bovee, K. D., B. L. Lamb, J. M. Bartholow, C. D. Stalnaker, J. Taylor, and J. Henrikson. 1998. Stream Habitat
Analysis Using the Instream Flow Incremental Methodology. Information and Technical Report
USGS/BRD-1998-0004. U.S. Geological Survey, Biological Resources Division.
Bradford, M. J., G. C. Taylor, J. A. Allan, and P. S. Higgins. 1995. An experimental study of the stranding of
juvenile coho salmon and rainbow trout during rapid flow decreases under winter conditions. North
American Journal of Fisheries Management 15:47-479.
Brannon, E. 1987. Mechanisms stabilizing salmonid fry emergence timing. Canadian Special Publication
of Fisheries and Aquatic Sciences 96:120-124.
Brown, T. G., and G. F. Hartman. 1988. Contribution of seasonally flooded lands and minor tributaries to
the production of coho salmon in Carnation Creek, British Columbia. Transactions of the American
Fisheries Society 117:546-551.
Brown, R. S., W. A. Hubert, and S. F. Daly. 2011. A primer on winter, ice, and fish: what fisheries biologists
should know about winter ice processes and stream-dwelling fish. Fisheries 36:8-26.
Bryant, M. D., N. D. Zymonas, and B. E. Wright. 2004. Salmonids on the fringe: abundance, species
composition, and habitat use of salmonids in high-gradient headwater streams, southeast Alaska.
Transactions of the American Fisheries Society 133:1529-1538.
Bristol Bay Assessment 15-27 January 2014
-------
Chapter 15 References
Bue, B. G., S. M. Fried, S. Sharr, D. G. Sharp, J. A. Wilcock, and H. J. Geiger. 1988. Estimating salmon
escapement using area-under-the-curve, aerial observer efficiency, and stream-life estimates: the
Prince William Sound pink salmon example. North Pacific Anadromous Fish Commission Bulletin
1:240-250.
Callegary, J., C. Kikuchi, J. Koch, M. Lilly, and S. Leake. 2013. Review: groundwater in Alaska (USA).
Hydrogeology Journal 21(l):25-39.
Chadwick, M. A., and A. D. Huryn. 2007. Role of habitat in determining macroinvertebrate production in
an intermittent-stream system. Freshwater Biology 52:240-251.
Collen, P., and R. J. Gibson. 2001. The general ecology of beavers (Castor spp.), as related to their
influence on stream ecosystems and riparian habitats, and the subsequent effects on fish—A review.
Reviews in Fish Biology and Fisheries 10:439-461.
Compton, J. E., M. R. Church, S. T. Larned, and W. E. Hogsett. 2003. Nitrogen export from forested
watersheds in the Oregon Coast Range: the role of N-2-fixing red alder. Ecosystems 6:773-785.
Cunjak, R. A. 1996. Winter habitat of selected stream fishes and potential impacts from land-use activity.
Canadian Journal of Fisheries and Aquatic Sciences 53 (Supplement l):267-282.
Curry, R. A., J. Gehrels, D. L. G. Noakes, and R. Swainson. 1994. Effects of river flow fluctuations on
groundwater discharge through brook trout, Salvelinusfontinalis, spawning and incubation habitats.
Hydrobiologia 277:121-134.
Dekar, M. P., R. S. King, J. A. Back, D. F. Whigham, and C. M. Walker. 2012. Allochthonous inputs from
grass-dominated wetlands support juvenile salmonids in headwater streams: evidence from stable
isotopes of carbon, hydrogen, and nitrogen. Freshwater Science 31:121-132.
Doucett, R. R., G. Power, D. R. Barton, R. J. Drimmie, and R. A. Cunjak. 1996. Stable isotope analysis of
nutrient pathways leading to Atlantic salmon. Canadian Journal of Fisheries and Aquatic Sciences
53:2058-2066.
Dye, J. E., and C. J. Schwanke. 2009. Report to the Alaska Board of Fisheries for the Recreational Fisheries of
Bristol Bay, 2007,2008, and 2009. Alaska Deptartment of Fish and Game, Division of Sport Fish,
Research and Technical Services.
Elliott, G. V., and J. E. Finn. 1983. Fish Utilization of Several Kenai River Tributaries: 1982 Field Report.
Anchorage, AK: U.S. Fish and Wildlife Service, Special Studies.
Estes, C. C. 1998. Annual Summary oflnstream Flow Reservations and Protection in Alaska. Alaska
Department of Fish and Game, Division of Sport Fish, Research and Technical Services.
Ferreira, V., V. Gulis, and M. A. S. Graca. 2006. Whole-stream nitrate addition affects litter decomposition
and associated fungi but not invertebrates. Oecologia 149:718-729.
Bristol Bay Assessment 15-28 January 2014
-------
Chapter 15 References
Freeman, M. C., C. M. Pringle, and C. R. Jackson. 2007. Hydrologic connectivity and the contribution of
stream headwaters to ecological integrity at regional scales. Journal of the American Water Resources
Association 43(1):5-14.
Gesch, D. 2007. Chapter 4-The National Elevation Dataset. Pages 99-118 in Maune, D. (ed.), Digital
Elevation Model Technologies and Applications: The DEM Users Manual. 2nd edition. Bethesda, MD:
American Society for Photogrammetry and Remote Sensing.
Gesch, D., M. Oimoen, S. Greenlee, C. Nelson, M. Steuck, and D. Tyler. 2002. The National Elevation
Dataset. Photogrammatic Engineering and Remote Sensing 68(1):5-11.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company). Vancouver, BC.
Gomez-Uchida, D., J. E. Seeb, M. J. Smith, C. Habicht, T. P. Quinn, and L. W. Seeb. 2011. Single nucleotide
polymorphisms unravel hierarchical divergence and signatures of selection among Alaskan sockeye
salmon (Oncorhynchus nerka] populations. BMC Evolutionary Biology 11(1):48.
Hall, J. V., W. E. Prayer, and B. 0. Wilen. 1994. Status of Alaska Wetlands. U.S. Fish and Wildlife Service,
Alaska Region, Anchorage, AK. Available: http://www.fws.gov/wetlands/Documents/Status-of-
Alaska-Wetlands.pdf.
Hancock, P. J. 2002. Human impacts on the stream-groundwater exchange zone. Environmental
Management 29:763-781.
Hendry, A. P., J. E. Hensleigh, and R. R. Reisenbichler. 1998. Incubation temperature, developmental
biology, and the divergence of sockeye salmon (Oncorhynchus nerka] within Lake Washington.
Canadian Journal of Fisheries and Aquatic Sciences 55:1387-1394.
Henning, J. A., R. E. Gresswell, and I. A. Fleming. 2006. Juvenile salmonid use of freshwater emergent
wetlands in the floodplain and its implications for conservation management. North American
Journal of Fisheries Management 26:367-376.
Hilborn, R., B. G. Bue, and S. Sharr. 1999. Estimating spawning escapements from periodic counts: a
comparison of methods. Canadian Journal ofFisheries and Aquatic Sciences 56(5):888-896.
Hilborn, R., T. Quinn, D. Schindler, and D. Rogers. 2003. Biocomplexity and fisheries sustainability.
Proceedings of the National Academy of Sciences of the United States of America 100:6564-6568.
Hillman, T. W., J. W. Mullan, and J. S. Griffith. 1992. Accuracy of underwater counts of juvenile Chinook
salmon, coho salmon, and steelhead. North American Journal of Fisheries Management 12:598-603.
Hodgson, S., and T. P. Quinn. 2002. The timing of adult sockeye salmon migration into fresh water:
adaptations by populations to prevailing thermal regimes. Canadian Journal of Zoology 80:542-555.
Bristol Bay Assessment 15-29 January 2014
-------
Chapter 15 References
Holt, K. R., and S. P. Cox. 2008. Evaluation of visual survey methods for monitoring Pacific salmon
(Oncorhynchus spp.) escapement in relation to conservation guidelines. Canadian Journal of Fisheries
and Aquatic Sciences 65(2)\212-226.
Huusko, A., L. Greenberg, M. Stickler, T. Linnansaari, M. Nykanen, T. Vehanen, S. Koljonen, P. Louhi, and
K. Alfredsen. 2007. Life in the ice lane: the winter ecology of stream salmonids. River Research and
Applications 23:469-491.
Irons J. G., Ill, S. R. Ray, L. K. Miller, and M. W. Oswood. 1989. Spatial and seasonal patterns of streambed
water temperatures in an Alaskan subarctic stream. Proceedings of the Symposium on Headwaters
Hydrology June:381-390. American Water Resources Association. As cited in Reynolds, J. B. 1997.
Ecology of overwintering fishes in Alaskan freshwaters. Pages 281-302 in A. M. Milner and M. W.
Oswood (eds.), Freshwaters of Alaska: Ecological Syntheses. New York, NY: Springer.
Johnson, J., and P. Blanche. 2012. Catalog of Waters Important for Spawning, Rearing, or Migration of
Anadromous Fishes - Southwestern Region, Effective June 1,2012. Special Publication No. 12-08.
Anchorage, AK: Alaska Department of Fish and Game.
Jones, E. L, III, S. Heinl, and K. Pahlke. 2007. Aerial counts. Pages 399-410 in D. H. Johnson, B. M. Shrier,
J. S. O'Neal, J. A. Knutzen, X. Augerot, T. A. O'Neil, and T. N. Pearsons, (eds.), Salmonid Field Protocols
Handbook: Techniques for Assessing Status and Trends in Salmon and Trout Populations. Bethesda,
MD: American Fisheries Society.
King, R. S., C. M. Walker, D. F. Whigham, S. Baird, and J. A. Back. 2012. Catchment topography and
wetland geomorphology drive macroinvertebrate community structure and juvenile salmonid
distributions in southcentral Alaska headwater streams. Freshwater Science 31(2):341-364.
Lang, D. W., G. H. Reeves, J. D. Hall, and M. S. Wipfli. 2006. The influence of fall-spawning coho salmon
(Oncorhynchus kisutch] on growth and production of juvenile coho salmon rearing in beaver ponds
on the Copper River Delta, Alaska. Canadian Journal of Fisheries and Aquatic Sciences 63:917-930.
Lytle, D. A., and N. L. R. Poff. 2004. Adaptation to natural flow regimes. Trends in Ecology & Evolution
19:94-100.
Matthews, W. J., and E. Marsh-Matthews. 2003. Effects of drought on fish across axes of space, time and
ecological complexity. Freshwater Biology 48:1232-1253.
Meyer, J. L., D. L. Strayer, J. B. Wallace, S. L. Eggert, G. S. Helfman, and N. E. Leonard. 2007. The
contribution of headwater streams to biodiversity in river networks. Journal of the American Water
Resources Association 43(1):86-103.
Morstad, S. 2003. Kvichak River Sockeye Salmon Spawning Ground Surveys, 1955-2002. Regional
Information Number 2A02-32. Anchorage, AK: Alaska Department of Fish and Game.
Bristol Bay Assessment 15-30 January 2014
-------
Chapter 15 References
Nickelson, T. E., J. D. Rodgers, S. L. Johnson, and M. F. Solazzi. 1992. Seasonal changes in habitat use by
juvenile coho salmon (Oncorhynchus kisutch) in Oregon coastal streams. Canadian Journal of
Fisheries and Aquatic Sciences 49:783-789.
Olsen, J. B., S. J. Miller, W. J. Spearman, and J. K. Wenburg. 2003. Patterns of intra-and inter-population
genetic diversity in Alaskan coho salmon: implications for conservation. Conservation Genetics
4(5):557-569.
Parken, C. K., R. E. Bailey, and J. R. Irvine. 2003. Incorporating uncertainty into area-under-the-curve and
peak count salmon escapement estimation. North American journal of Fisheries Management
23(1):78-90.
Pess, G. R., D. R. Montgomery, E. A. Steel, R. E. Bilby, B. E. Feist, and H. M. Greenberg. 2002. Landscape
characteristics, land use, and coho salmon [Oncorhynchus kisutch) abundance, Snohomish River,
Wash., U.S.A. Canadian journal of Fisheries and Aquatic Sciences 59:613-623.
Pess, G. R., S. A. Morley, J. L. Hall, and R. K. Timm. 2005. Monitoring Floodplain Restoration. Pages 127-
166 in P. Roni, (ed.), Monitoring Stream and Watershed Restoration. Bethesda, MD: American
Fisheries Society.
Piccolo, J. J., and M. S. Wipfli. 2002. Does red alder [Alnus rubrd) in upland riparian forests elevate
macroinvertebrate and detritus export from headwater streams to downstream habitats in
southeastern Alaska? Canadian journal of Fisheries and Aquatic Sciences 59:503-513.
PLP (Pebble Limited Partnership). 2011. Pebble Project Environmental Baseline Document 2004 through
2008. Anchorage, AK. Available: http://www.arlis.org/docs/vol2/Pebble/2004-2008EBDIndex.pdf.
Poff, N. L., J. D. Allan, M. B. Bain, J. R. Karr, K. L. Prestegaard, B. D. Richter, R. E. Sparks, and J. C.
Stromberg. 1997. The natural flow regime: a paradigm for river conservation and restoration.
Bioscience 47:769-784.
Poff, N., J. Olden, D. Pepin, and B. Bledsoe. 2006. Placing global stream flow variability in geographic and
geomorphic contexts. Regulated Rivers: Research & Management 22:149-166.
Poff, N., B. Richter, A. Arthington, S. Bunn, R. Naiman, C. Apse, E. Kendy, A. Warner, R. Jacobson, and K.
Rogers. 2009. The ecological limits of hydrologic alteration (ELOHA): a new framework for
developing regional environmental flow standards. Freshwater Biology 55:147-170.
Poff, N. L., and J. K. H. Zimmerman. 2010. Ecological responses to altered flow regimes: a literature
review to inform the science and management of environmental flows. Freshwater Biology 55:194-
205.
Pollock, M. M., M. Heim, and D. Werner. 2003. Hydrologic and geomorphic effects of beaver dams and
their influence on fishes. The Ecology and Management of Wood in World Rivers 37:213-233.
Bristol Bay Assessment 15-31 January 2014
-------
Chapter 15 References
Pollock, M. M., G. R. Pess, and T. J. Beechie. 2004. The importance of beaver ponds to coho salmon
production in the Stillaguamish River basin, Washington, USA. North American Journal of Fisheries
Management 24:749-760.
Postel, S., and B. Richter. 2003. Rivers for Life: Managing Water for People and Nature. Washington, DC:
Island Press.
Power, G., R. S. Brown, and J. G. Imhof. 1999. Groundwater and fish—Insights from northern North
America. Hydrological Processes 13:401-422.
Quinn, T. P. 2005. The Behavior and Ecology of Pacific Salmon and Trout. Seattle, WA: University of
Washington Press.
Quinn, T. P., A. P. Hendry, and G. B. Buck. 2001. Balancing natural and sexual selection in sockeye
salmon: interactions between body size, reproductive opportunity and vulnerability to predation by
bears. Evolutionary Ecology Research 3(8):917-937.
Quinn, T. P., A. P. Hendry, and L. A. Wetzel. 1995. The influence of life history trade-offs and the size of
incubation gravels on egg size variation in sockeye salmon (Oncorhynchus nerka}. Oikos
(1995):425-438.
Quinn, T. P., H. B. Rich Jr, D. Gosse, N. Schtickzelle, and J. Grant. 2012. Population dynamics and
asynchrony at fine spatial scales: a case history of sockeye salmon (Oncorhynchus nerka] population
structure in Alaska, USA. Canadian Journal of Fisheries and Aquatic Sciences 69(2):297-306.
Rains, M. C. 2011. Water sources and hydrodynamics of closed-basin depressions, Cook Inlet Region,
Alaska. Wetlands 31:377-387.
Rand, P. S., C. Kellopn, X. Augerot, M. Goslin, J. R. Irvine, and G. T. Ruggerone. 2007. Comparison of
sockeye salmon (Oncorhynchus nerka] monitoring in the Fraser River Basin, British Columbia,
Canada and Bristol Bay, Alaska, USA. North Pacific Anadromous Fish Commission Bulletin 4:271-284.
Reynolds, J. B. 1997. Ecology of overwintering fishes in Alaskan freshwaters. Pages 281-302 in A. M.
Milner and M. W. Oswood (eds.), Freshwaters of Alaska: Ecological Syntheses. New York, NY:
Springer.
Richardson, J. S., R. J. Naiman, F. J. Swanson, and D. E. Hibbs. 2005. Riparian communities associated with
Pacific Northwest headwater streams: assemblages, processes, and uniqueness. Journal of the
American Water Resources Association 41(4):935-947.
Richter, B. D. 2010. Re-thinking environmental flows: from allocations and reserves to sustainability
boundaries. River Research and Applications 26:1052-1063.
Richter, B., M. Davis, C. Apse, and C. Konrad. 2012. A presumptive standard for environmental flow
protection. River Research and Applications 28:1312-1321.
Bristol Bay Assessment 15-32 January 2014
-------
Chapter 15 References
Rogers, L. A., and D. E. Schindler. 2008. Asynchrony in population dynamics of sockeye salmon in
southwest Alaska. Oikos 117:1578-1586.
Ruff, C. P., D. E. Schindler, J. B. Armstrong, K. T. Bentley, G. T. Brooks, G. W. Holtgrieve, M. T. McGlauflin, C.
E. Torgersen, and J. E. Seeb. 2011. Temperature-associated population diversity in salmon confers
benefits to mobile consumers. Ecology 92:2073-2084.
Schindler, D. E., R. Hilborn, B. Chasco, C. P. Boatright, T. P. Quinn, L. A. Rogers, and M. S. Webster. 2010.
Population diversity and the portfolio effect in an exploited species. Nature 465:609-612.
Shaftel, R., R. King, and J. Back. 2011. Breakdown rates, nutrient concentrations, and macroinvertebrate
colonization of bluejoint grass litter in headwater streams of the Kenai Peninsula, Alaska. Journal of
the North American Benthological Society 30:386-398.
Shaftel, R., R. King, and J. Back. 2012. Alder cover drives nitrogen availability in Kenai lowland
headwater streams, Alaska. Biogeochemistry 107:135-148.
Slaughter, C. W. 1990. Aufeis formation and prevention. Pages 433-458 in W. L. Ryan and R. D. Crissman
(eds.), Cold Regions Hydrology and Hydraulics. New York, NY: American Society of Civil Engineers.
Solazzi, M., T. Nickelson, S. Johnson, and J. Rodgers. 2000. Effects of increasing winter rearing habitat on
abundance of salmonids in two coastal Oregon streams. Canadian Journal of Fisheries and Aquatic
Sciences 57-.906-914.
Sommer, T. R., M. L. Nobriga, W. C. Harrell, W. Batham, and W. J. Kimmerer. 2001. Floodplain rearing of
juvenile chinook salmon: evidence of enhanced growth and survival. Canadian Journal of Fisheries
and Aquatic Sciences 58:325-333.
Szarzi, N. J, S. J. Fleischman, R. A. Clark, and C. M. Kerkvliet. 2007. Stock status and recommended
escapement goal for Anchor River Chinook salmon. Fisheries Manuscript No. 07-05. Alaska
Department of Fish and Game, Divisions of Sport Fish and Commercial Fisheries.
Tank, J. L., E. J. Rosi-Marshall, N. A. Griffiths, S. A. Entrekin, and M. L. Stephen. 2010. A review of
allochthonous organic matter dynamics and metabolism in streams. Journal of the North American
Benthological Society 29:118-146.
USAGE (U.S. Army Corps of Engineers). 2009. Alaska District Regulatory Guidance Letter, RGL ID No. 09-
01.
USAGE (U.S. Army Corps of Engineers). 2011. Memorandum for Regulatory Division, Alaska District, 3
March 2011.
USFWS (U.S. Fish and Wildlife Service). 2012. National Wetlands Inventory. Available:
http://www.fws.gov/wetlands/Wetlands-Mapper.html.Accessed: December 11, 2012.
Bristol Bay Assessment 15-33 January 2014
-------
Chapter 15 References
USGS (U.S. Geological Survey). 2012a. National Hydrography Dataset. High Resolution, Alaska. Available:
ftp://nhdftp.usgs.gov/DataSets/Staged/States/FileGDB/HighResolution. Accessed: October 16,
2012.
USGS (U.S. Geological Survey). 2012b. Water Resources of Alaska. USGS Alaska Science Center Water
Resources Office. Last updated: December 12, 2012. Available:
http://alaska.usgs.gov/science/water/index.php. Accessed: April 17, 2013.
USGS (U.S. Geological Survey). 2013. National Elevation Dataset (2 arc second), ArcGrid. August 26.
[n59wl55.zip, n59wl56.zip, n59wl57.zip, n59wl58.zip, n60wl54.zip, n60wl55.zip, n60wl56.zip,
n60wl57.zip, n60wl58.zip, n60wl59.zip, n60wl60.zip, n61wl53.zip, n61wl54.zip, n61wl55.zip,
n61wl56.zip, n61wl57.zip, n61wl58.zip, n61wl59.zip, n61wl60.zip] Available:
http://viewer.nationalmap.gov/viewer/.
Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Gushing. 1980. The river continuum
concept. Canadian Journal of Fisheries and Aquatic Sciences 37:130-137.
Walker, C. M., R. S. King, D. F. Whigham, and S. J. Baird. 2012. Landscape and wetland influences on
headwater stream chemistry in the Kenai Lowlands, Alaska. Wetlands 32(2):301-310.
Webster, J. R., and E. F. Benfield. 1986. Vascular plant breakdown in fresh-water ecosystems. Annual
Review of Ecology and Systematics 17:567-594.
West, R. L., M. W. Smith, W. E. Barber, J. B. Reynolds, and H. Hop. 1992. Autumn migration and
overwintering of Arctic grayling in coastal streams of the Arctic National Wildlife Refuge, Alaska.
Transactions of the American Fisheries Society 121:709-715.
Wipfli, M. S., and C. V. Baxter. 2010. Linking ecosystems, food webs, and fish production: subsidies in
salmonid watersheds. Fisheries 35(8):373-387.
Wipfli, M. S., and D. P. Gregovich. 2002. Export of invertebrates and detritus from fishless headwater
streams in southeastern Alaska: implications for downstream salmonid production. Freshwater
Biology 47:957-969.
Wipfli, M. S., and J. Musslewhite. 2004. Density of red alder [Alnus ruhrd) in headwaters influences
invertebrate and detritus subsidies to downstream fish habitats in Alaska. Hydrobiologia 520:153-
163.
Wipfli, M. S., J. S. Richardson, and R. J. Naiman. 2007. Ecological linkages between headwaters and
downstream ecosystems: transport of organic matter, invertebrates, and wood down headwater
channels, journal of the American Water Resources Association 43:72-85.
Woody, C. A., and B. Higman. 2011. Groundwater as Essential Salmon Habitat in Nushagak and Kvichak
River Headwaters: Issues Relative to Mining. Unpublished report. Available:
http://www.fish4thefuture.com/pdfs/Groundwater_and_SalmonFINAL27Augll.pdf.
Bristol Bay Assessment ,. co4 January 2014
15-34
-------
Chapter 15 References
Personal Communications
Morstad, S. Fishery Biologist III, ADF&G. September 2011—email of unpublished data to Rebecca
Shaftel.
15.1.9 Chapter 8—Water Collection, Treatment, and Discharge
ADEC (Alaska Department of Environmental Conservation). 2011. Water Quality Standards: Alaska
Department of Environmental Conservation. Report 18-AAC-70.
ADEC (Alaska Department of Environmental Conservation). 2012. Water Quality Standards Amended as
of April 8,2012. Juneau, AK: Alaska Department of Environmental Conservation.
ADNR (Alaska Department of Natural Resources). 2012. Nixon Fork Mine. Available:
http://dnr.alaska.gov/mlw/mining/largemine/nixonfork/. Accessed: April 11, 2012.
Angilletta, M. ]., E. Ashley Steel, K. K. Bartz, J. G. Kingsolver, M. D. Scheuerell, B. R. Beckman, and L. G.
Crozier. 2008. Big dams and salmon evolution: changes in thermal regimes and their potential
evolutionary consequences. Evolutionary Applications 1:286-299.
ANZECC (Australia and New Zealand Environment and Conservation Council). 2000. Australian and New
Zealand guidelines for fresh and marine water quality. Volume 2. Aquatic Ecosystems: Rationale and
Background Information. Canberra, Australia: Australia and New Zealand Environment and
Conservation Council and Agriculture and Resource Management Council of Australia and New
Zealand.
Armstrong, J. B., D. E. Schindler, K. L. Omori, C. P. Ruff, and T. P. Quinn. 2010. Thermal heterogeneity
mediates the effects of pulsed subsidies across a landscape. Ecology 91:1445-1454.
Barry, K. L., J. A. Grout, C. D. Levings, B. H. Nidle, and G. E. Piercey. 2000. Impacts of acid mine drainage
on juvenile salmonids in an estuary near Britannia Beach in Howe Sound, British Columbia.
Canadian Journal of Fisheries and Aquatic Sciences 57:2032-2043.
Beacham, T. D., and C. B. Murray. 1990. Temperature, egg size, and development of embryos and alevins
of five species of Pacific salmon: a comparative analysis. Transactions of the American Fisheries
Society 119:927-945.
Borgmann, U., C. Janssen, R. Blust, K. Brix, R. Dwyer, R. Erickson, L. Hare, S. Luoma, P. Paquin, C. Roberts,
and W.-X. Wang. 2005. Incorporation of dietborne metals exposure into regulatory frameworks.
Pages 153-190 in J. Meyer, W. Adams, K. Brix, S. Luoma, D. Mount, W. Stubblefield, and C. Wood,
editors. Toxicity of Dietborne Metals to Aquatic Organisms. Pensacola, FL: SETAC Press.
Brannon, E. 1987. Mechanisms stabilizing salmonid fry emergence timing. Canadian Special Publication
of Fisheries and Aquatic Sciences 96:120-124.
Brix, K. V., D. K. DeForest, and W. J. Adams. 2011. The sensitivity of aquatic insects to divalent metals: a
comparative analysis of laboratory and field data. Science of the Total Environment 409:4187-4197.
Bristol Bay Assessment 15-35 January 2014
-------
Chapter 15 References
Bryant, M. 2009. Global climate change and potential effects on Pacific salmonids in freshwater
ecosystems of southeast Alaska. Climatic Change 95:169-193.
Buchwaiter, D. B., D. J. Cain, W. H. Clements, and S. N. Luoma. 2008. Using biodynamic models to
reconcile differences between laboratory toxicity tests and field biomonitoring with aquatic insects
(vol. 41, pg 4821, 2007). Environmental Science & Technology 42:3117-3117.
Chapman, G. 1975. Toxicity of copper, cadmium and zinc to Pacific Northwest salmonids. Corvallis, OR:
U.S. Environmental Protection Agency, Western Fish Toxicology Station, National Water Quality
Laboratory.
Chapman, G. A. 1978. Toxicities of cadmium, copper, and zinc to 4 juvenile stages of Chinook salmon and
steelhead. Transactions of the American Fisheries Society 107:841-847.
Chapman, P. M., H. Bailey, and E. Canaria. 2000. Toxicity of total dissolved solids associated with two
mine effluents to chironomid larvae and early life stages of rainbow trout. Environmental Toxicology
and Chemistry 19:210-214.
Clements, W. H., D. M. Carlisle, J. M. Lazorchak, and P. C. Johnson. 2000. Heavy metals structure benthic
communities in Colorado mountain streams. Ecological Applications 10:626-638.
Cohen, B. I. 2010. Eraser River Sockeye Salmon: Past Declines. Future Sustainability? Interim Report.
Vancouver, BC: Commission of Inquiry into the Decline of Sockeye Salmon in the Fraser River.
Cohen, B. I. 2012. The Uncertain Future of Fraser River Sockeye. Volume 2: Causes of the Decline. Final
Report CP32-93/2012E-2. Ottawa, ON: Commission of Inquiry into the Decline of Sockeye Salmon in
the Fraser River.
Courtney, L. A., and W. H. Clements. 2002. Assessing the influence of water and substratum quality on
benthic macroinvertebrate communities in a metal-polluted stream: an experimental approach.
Freshwater Biology 47:1766-1778.
DeForest, D., J. Meyer, R. Genseer, B. Shephard, W. Adams, R. Dwyer, J. Gorsuch, and E. VanGenderen.
201 la. Are ambient water quality criteria for copper protective of olfactory impairment in fish?
Integrated Environmental Assessment and Management 7: 145-146.
DeForest, D., R. Gensemer, E. Van Genderen, and J. Gorsuch. 2011b. Protectiveness of water quality
criteria for copper in western United States waters relative to predicted olfactory responses in
juvenile Pacific salmon. Integrated Environmental Assessment and Management 7:336-347.
DeForest, D. K., and E. J. Van Genderen. 2012. Application of U.S. EPA guidelines in a bioavailability-based
assessment of ambient water quality criteria for zinc in freshwater. Environmental Toxicology and
Chemistry 31:1264-1272.
Bristol Bay Assessment 15-36 January 2014
-------
Chapter 15 References
Earthworks. 2012. U.S. Copper Porphyry Mines Report: The Track Record of Water Quality Impacts
Resulting from Pipeline Spills, Tailings Failures and Water Collection and Treatment Failure.
Washington, DC.
Eisler, R. 2000. Handbook of Chemical Risk Assessment: Health Hazards to Humans, Plants, and Animals.
Volume 1: Metals. Boca Raton, FL: Lewis Publisher.
Environment Canada and Health Canada. 2011. Screening Assessment for the Challenge: Cobalt; Cobalt
Chloride; Sulfiiric Acid, Cobalt (2+) Salt (1:1); and Sulfiiric Acid, Cobalt Salt. Ottawa, ON. Available:
http://www.ec.gc.ca/ese-ees/default.asp?lang=En&n=8E18277B-l.
Erickson, R. ]., D. R. Mount, T. L. Highland, J. R. Hockett, E. N. Leonard, V. R. Mattson, T. D. Dawson, and K.
G. Lott. 2010. Effects of copper, cadmium, lead, and arsenic in a live diet on juvenile fish growth.
Canadian Journal of Fisheries and Aquatic Sciences 67:1816-1826.
European Copper Institute. 2008. Voluntary Risk Assessment of Copper, Copper II Sulphate Pentahydrate,
Copper(I)oxide, Copper(II)oxide, Dicopper Chloride Trihydroxide. Brussels, Belgium: European Copper
Institute for the European Commission.
Farag, A. M., C. J. Boese, D. F. Woodward, and H. L. Bergman. 1994. Physiological changes and tissue
metal accumulation in rainbow trout exposed to foodborne and waterborne metals. Environmental
Toxicology and Chemistry 13:2021-2029.
Farag, A. M., D. E. Woodward, W. Brumbaugh, J. N. Goldstein, E. MacConnell, C. Hogstrand, and F. T.
Barrows. 1999. Dietary effects of metals-contaminated invertebrates from the Coeur d'Alene River,
Idaho, on cutthroat trout. Transactions of the American Fisheries Society 128:578-592.
Fuerstenau, M. C., B. M. Wakawa, R. K. Price, and R. D. Wellik. 1974. Toxicity of selected sulfhydryl
collectors to rainbow trout. Transactions of the Society of Mining Engineers 256:337-341.
Gensemer, R. W., and R. C. Playle. 1999. The bioavailability and toxicity of aluminum in aquatic
environments. Critical Reviews in Environmental Science & Technology 29:315-450.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company). Vancouver, BC.
Griffith, M. B., J. M. Lazorchak, and A. T. Herlihy. 2004. Relationships among exceedences of metals
criteria, the results of ambient bioassays, and community metrics in mining-impacted streams.
Environmental Toxicology and Chemistry 23:1786-1795.
Gustafson, 0. 2012. Active Metal Mines of the Eraser River Basin and Fish - Case Studies. Technical Paper
Series No. 1, White Paper No. 3. Pebble Limited Partnership.
Bristol Bay Assessment 15-37 January 2014
-------
Chapter 15 References
Hansen, J. A., J. Lipton, P. G. Welsh, D. Cacela, and B. MacConnell. 2004. Reduced growth of rainbow trout
(Oncorhynchus mykiss) fed a live invertebrate diet pre-exposed to metal-contaminated sediments.
Environmental Toxicology and Chemistry 23:1902-1911.
Hansen, J. A., J. C. A. Marr, J. Lipton, D. Cacela, and H. L. Bergman. 1999. Differences in neurobehavioral
response of Chinook salmon (Oncorhynchus tshawytscha] and rainbow trout (Oncorhynchus mykiss]
exposed to copper and cobalt: behavioral avoidance. Environmental Toxicology and Chemistry
18:1972-1978.
Hecht, S., D. Baldwin, C. Mebane, T. Hawkes, S. Gross, and N. Scholz. 2007. An Overview of Sensory Effects
on Juvenile Salmonids Exposed to Dissolved Copper: Applying a Benchmark Concentration Approach to
Evaluate Sublethal Neurobehavioral Toxicity. Seattle, WA: National Oceanic and Atmospheric
Administration, Northwest Fisheries Science Center.
Hedtke, S. F. 1984. Structure and function of copper-stressed aquatic microcosms. Aquatic Toxicology
5:227-244.
Hodgson, S., and T. P. Quinn. 2002. The timing of adult sockeye salmon migration into fresh water:
adaptations by populations to prevailing thermal regimes. Canadian Journal of Zoology 80:542-555.
IPCS (International Programme on Chemical Safety). 2004. Concise International Chemical Assessment
Document 63. Manganese and its Compounds: Environmental Aspects. Geneva, Switzerland: World
Health Organization.
ITRC (Interstate Technology & Regulatory Council). 2010. Bingham Canyon Water Treatment Plant,
Kennecott South Zone. Mining Waste Treatment Technology Selection Web. Washington, DC:
Interstate Technology & Regulatory Council, Mining Waste Team. Available:
http://www.itr cweb.org/miningwaste-guidance/cs48_kennecott_south.htm.
Jennings, S., D. Neuman, and P. Blicker. 2008. Acid Mine Drainage and Effects on Fish Health and Ecology:
A Review. Bozeman, Montana: Reclamation Research Group, LLC.
Jensen, A., and B. Johnsen. 1999. The functional relationship between peak spring floods and survival
and growth of juvenile Atlantic salmon (Salmo salar] and brown trout (Salmo trutta}. Functional
Ecology 13:778-785.
Joling, D. 2011. Mine Executive Says Pebble, Fishing Can Co-Exist. Anchorage Daily News (Anchorage,
AK). March 4, 2011.
Kiffney, P. M., and W. H. Clements. 1996. Size-dependent response of macroinvertebrates to metals in
experimental streams. Environmental Toxicology and Chemistry 15:1352-1356.
Le Jeune, A. H., M. Charpin, V. Deluchat, and J.-F. Briand. 2006. Effect of copper sulphate treatment on
natural phytoplanktonic communities. Aquatic Toxicology 80:267-280.
Bristol Bay Assessment 15-38 January 2014
-------
Chapter 15 References
Leland, H. V., and J. L. Carter. 1984. Effects of copper on species composition of periphyton in a Sierra
Nevada, California, stream. Freshwater Biology 14:281-296.
Leland, H. V., and J. L. Carter. 1985. Effects of copper on production of periphyton, nitrogen-fixation and
processing of leaf litter in a Sierra Nevada, California, stream. Freshwater Biology 15:155-173.
Leland, H. V., S. V. Fend, T. L. Dudley, and J. L. Carter. 1989. Effects of copper on species composition of
benthic insects in a Sierra Nevada, California, stream. Freshwater Biology 21:163-179.
Lorz, H., and B. McPherson. 1977. Effects of Copper and Zinc on Smoltification ofCoho Salmon. Corvallis,
OR: U.S. Environmental Protection Agency, Environmental Research Laboratory.
Luoma, S. N., and P. S. Rainbow. 2005. Why is metal bioaccumulation so variable? Biodynamics as a
unifying concept. Environmental Science & Technology 39:1921-1931.
MacDonald, D., J. Sinclair, M. Crawford, H. Principe, and M. Meneghetti. 2011. Potential Effects of
Contaminants on Eraser River Sockeye Salmon. Cohen Commission Technical Report 2. Vancouver,
BC: MacDonald Environmental Sciences, Ltd.
Marchand, E. A. 2002. Minerals and mine drainage. Water Environment Research 74:746-772.
McCullough, D. A., J. M. Bartholow, H. I. Jager, R. L. Beschta, E. F. Cheslak, M. L. Deas, J. L. Ebersole, J. S.
Foott, S. L. Johnson, K. R. Marine, M. G. Mesa, J. H. Petersen, Y. Souchon, K. F. Tiffan, and W. A.
Wurtsbaugh. 2009. Research in thermal biology: burning questions for coldwater stream fishes.
Reviews in Fisheries Science 17:90-115.
Mclntyre, J. K., D. H. Baldwin, J. P. Meador, and N. L. Scholz. 2008a. Chemosensory deprivation in juvenile
coho salmon exposed to dissolved copper under varying water chemistry conditions. Environmental
Science & Technology 42:1352-1358.
Mclntyre, J. K., D. H. Baldwin, J. P. Meador, and N. L. Scholz. 2008b. Chemosensory deprivation in juvenile
coho salmon exposed to dissolved copper under varying water chemistry conditions (42:1352-1358,
2008). Environmental Science & Technology 42:6774-677S (Addition or correction).
Mebane, C. A. 2010. Cadmium Risks to Freshwater Life: Derivation and Validation of Low-Effect Criteria
Values Using Laboratory and Field Studies. Investigations Report 2006-5245, Version 1.2. Reston VA:
U.S. Geological Survey.
Mebane, C. A., and D. L. Arthaud. 2010. Extrapolating growth reductions in fish to changes in population
extinction risks: copper and Chinook salmon. Human and Ecological Risk Assessment 16:1026-1065.
Mebane, C. A., F. S. Dillon, and D. P. Hennessy. 2012. Acute toxicity of cadmium, lead, zinc, and their
mixtures to stream-resident fish and invertebrates. Environmental Toxicology and Chemistry
31:1334-1348.
Bristol Bay Assessment 15-39 January 2014
-------
Chapter 15 References
Mebane, C. A., D. P. Hennessy, and F. S. Dillon. 2008. Developing acute-to-chronic toxicity ratios for lead,
cadmium, and zinc using rainbow trout, a mayfly, and a midge. Water, Air, and Soil Pollution 188(1-
4):41-66. Available: http://dx.doi.org/10.1007/sll270-007-9524-8.
Mellina, E., R. D. Moore, S. G. Hinch, J. S. Macdonald, and G. Pearson. 2002. Stream temperature responses
to clearcut logging in British Columbia: the moderating influences of groundwater and headwater
lakes. Canadian Journal of Fisheries and Aquatic Sciences 59:1886-1900.
Meyer, J. S., and W. J. Adams. 2010. Relationship between biotic ligand model-based water quality
criteria and avoidance and olfactory responses to copper by fish. Environmental Toxicology and
Chemistry 29:2096-2103.
Meyer, J., W. Adams, K. Brix, D. Mount, W. Stubblefield, and C. Wood. 2005. Toxicity ofDietborne Metals to
Aquatic Organisms. Pensacola, FL: SETAC Press.
Mortazavi, S. 2008. Application of Membrane Separation Technology to Mitigation of Mine Effluent and
Acidic Mine Drainage. Mine Environment Neutral Drainage Program (MEND) Report 3.15.1.
Nelitz, M., M. Porterm, K. Wieckowski, D. Marmorek, K. Bryan, A. Hall, and D. Abraham. 2011. Evaluating
the Status of Eraser River Sockeye Salmon and Role Of Freshwater Ecology in Their Decline. Cohen
Commission Technical Report 3. Vancouver, BC: ESSA Technologies, Ltd.
NICNAS (National Industrial Chemicals Notification and Assessment Scheme). 1995. Sodium Ethyl
Xanthate, Priority Existing Chemical No. 5. Full public report. Canberra, Australia: Australian
Government Publishing Service.
NICNAS (National Industrial Chemicals Notification and Assessment Scheme). 2000. Sodium Ethyl
Xanthate, Priority Existing Chemical Secondary Notification Assessment. Report No. 5S. Canberra,
Australia: Australian Government Publishing Service.
O'Neal, S., and C. A. Woody. 2011. Canada's Eraser River: Reasons for Sockeye Salmon Declines with a
Comparison to Bristol Bay. Anchorage, AK: Fisheries Research and Consulting.
Ott, A. G., and P. Weber Scannell. 1994. Fish Monitoring Study, Red Dog Mine in the Wulik River Drainage,
Emphasis on Dolly Varden fSalvelinus malmaj. Summary Report 1990-1993. Anchorage, AK: Alaska
Department of Fish and Game.
PLP (Pebble Limited Partnership). 2011. Pebble Project Environmental Baseline Document 2004 through
2008. Anchorage, AK. Available: http://www.arlis.org/docs/vol2/Pebble/2004-2008EBDIndex.pdf.
Poole, G. C., J. B. Dunham, D. M. Keenan, S. T. Saulter, D. A. McCoullough, C. Mebane, J. C. Lockwood, D. A.
Essig, M. P. Hicks, D. J. Sturdevant, E. J. Materna, S. A. Spalding, J. Risely, and M. Deppman. 2004. The
case for regime-based water quality standards. BioScience 54:155-161.
Power, G., R. S. Brown, and J. G. Imhof. 1999. Groundwater and fish - Insights from northern North
America. HydrologicalProcesses 13:401-422.
Bristol Bay Assessment 15-40 January 2014
-------
Chapter 15 References
Quinn, T. P. 2005. The Behavior and Ecology of Pacific Salmon and Trout. Seattle, WA: University of
Washington Press.
Quinn, T. P., and D. J. Adams. 1996. Environmental changes affecting the migratory timing of American
shad and sockeye salmon. Ecology 77:1151-1162.
Reimer, P. S. 1999. Environmental effects of manganese and proposed freshwater guidelines to protect
aquatic life in British Columbia. M.S. thesis, University of British Columbia, Department of Chemical
and Bio-Resource Engineering.
Roussel, H., L. Ten-Hage, S. Joachim, R. Le Cohu, L. Gauthier, and J.-M. Bonzom. 2007. A long-term copper
exposure on freshwater ecosystem using lotic mesocosms: primary producer community responses.
Aquatic Toxicology 81:168-182.
Rudolph, B. L., I. Andreller, and C. J. Kennedy. 2008. Reproductive success, early life stage development,
and survival of westslope cutthroat trout (Oncorhynchus clarki lewisi) exposed to elevated selenium
in an area of active coal mining. Environmental Science & Technology 42:3109-3114.
Schindler, D. E., R. Hilborn, B. Chasco, C. P. Boatright, T. P. Quinn, L. A. Rogers, and M. S. Webster. 2010.
Population diversity and the portfolio effect in an exploited species. Nature 465:609-612.
Schmidt, T. S., W. H. Clements, K. A. Mitchell, S. E. Church, R. B. Wanty, D. L. Fey, P. L. Verplanck, and C. A.
San Juan. 2010. Development of a new toxic-unit model for the bioassessment of metals in streams.
Environmental Toxicology and Chemistry 29:2432-2442.
Shao, E., J. Wei, A. Yo, and R. Levy. 2009. Application of ultraflltration and reverse osmosis for mine waste
water reuse. Water in Mining Conference, Perth, WA, September 9, 2009.
USEPA (U.S. Environmental Protection Agency). 1985a. Ambient Water Quality Criteria for Copper -1984.
EPA 440/5-84-031. Washington, DC.
USEPA (U.S. Environmental Protection Agency). 1985b. Guidelines for Deriving Numerical National Water
Quality Criteria for the Protection of Aquatic Organisms and Their Uses. Washington, DC: Office of
Research and Development, Environmental Research Laboratories.
USEPA (U.S. Environmental Protection Agency). 1986. Ambient Aquatic Life Water Quality Criteria for
Lead. Washington, DC: Office of Water, Regulations and Standards, Criteria and Standards Division.
USEPA (U.S. Environmental Protection Agency). 1987. Ambient Water Quality Criteria for Zinc - 1987.
EPA-440/6-87-003. Washington, DC: Office of Water, Regulations and Standards, Criteria and
Standards Division.
USEPA (U.S. Environmental Protection Agency). 1998. Alaska's Red Dog Mine Required to Reduce
Pollutant Discharges. Available:
http://yosemite.epa.gov/opa/admpress.nsf/00000000000000000000000000000000/dOd834938
8619c61852570cb0075el66!opendocument.
Bristol Bay Assessment 15-41 January 2014
-------
Chapter 15 References
USEPA (U.S. Environmental Protection Agency). 2004. Draft Aquatic Life Water Quality Criteria for
Selenium - 2004. EPA-822-D-04-001. Washington, DC: Office of Water.
USEPA (U.S. Environmental Protection Agency). 2007. Aquatic Life Ambient Freshwater Quality Criteria -
Copper, 2007 Revision. Washington, DC.
USEPA (U.S. Environmental Protection Agency). 2008. The U.S. Environmental Protection Agency [EPA)
Plans to Re-Issue a Wastewater Discharge Permit to: Tech Cominco Alaska, Inc., Red Dog Mine Near
Kotze, Alaska, and the State of Alaska Proposes to Certify the Permit. Anchorage, AK.
USEPA (U.S. Environmental Protection Agency). 2011. A Field-Based Aquatic Life Benchmark for
Conductivity in Central Appalachian Streams [Final Report). EPA/600/R-10/023F. Washington, DC.
Webb, M., H. Ruber, and G. Leduc. 1976. The toxicity of various mining flotation reagents to rainbow
trout (Salmo gairdneri}. Water Research 10:303-306.
Weber Scannell, P. K. 1991. Influence of Temperature on Freshwater Fishes: A Literature Review with
Emphasis on Species in Alaska. Juneau, AK: Alaska Department of Fish and Game.
Woodward, D. F., W. G. Brumbaugh, A. J. Delonay, E. E. Little, and C. E. Smith. 1994. Effects on rainbow
trout fry of a metals-contaminated diet of benthic invertebrates from the Clark Fork River, Montana.
Transactions of the American Fisheries Society 123:51-62.
Woodward, D. F., A. M. Farag, H. L. Bergman, A. J. Delonay, E. E. Little, C. E. Smith, and F. T. Barrows. 1995.
Metals-contaminated benthic invertebrates in the Clark Fork River, Montana—Effects on age-0
brown trout and rainbow trout. Canadian Journal of Fisheries and Aquatic Sciences 52:1994-2004.
Woody, C. A., and B. Higman. 2011. Groundwater as Essential Salmon Habitat in Nushagak and Kvichak
River Headwaters: Issues Relative to Mining. Unpublished report. Available:
http://www.fish4thefuture.com/pdfs/Groundwater_and_SalmonFINAL27Augll.pdf.
Xu, Y., J. P. Lay, and F. Korte. 1988. Fate and effects of xanthates in laboratory freshwater systems.
Bulletin of Environmental Contamination and Toxicology 41:683-689.
Zamzow, K. 2011. Investigations of Surface Water Quality in the Nushagak, Kvichak, and Chulitna
Watersheds, Southwest Alaska 2009-2010. Sutton, AK: Center for Science in Public Participation.
15.1.10 Chapter 9—Tailings Dam Failure
ADEC (Alaska Department of Environmental Conservation). 2011. Water Quality Standards: Alaska
Department of Environmental Conservation. Report 18-AAC-70.
ADF&G (Alaska Department of Fish and Game). 2012. Alaska Freshwater Fish Inventory Database.
Available: http://www.adfg.alaska.gov/index.cfm?adfg=ffinventory.main.
Bristol Bay Assessment 15-42 January 2014
-------
Chapter 15 References
ADNR (Alaska Department of Natural Resources). 2005. Guidelines for Cooperation with the Alaska Dam
Safety Program. Dam Safety and Construction Unit, Water Resources Section, Division of Mining,
Land, and Waters.
AECOM. 2009. Root Cause Analysis ofTVA Kingston Dredge Pond Failure on December 22, 2008.
Ankley, G. T., D. M. DiToro, D. J. Hansen, and W. J. Berry. 1996. Technical basis and proposal for deriving
sediment quality criteria for metals. Environmental Toxicology and Chemistry 15:2056-2066.
ARCO (Atlantic Richfield Company). 1998. Milltown Reservoir Sediments NPL Site, Clark Fork River
Operable Unit, Final Draft Remedial Investigation Report. Anaconda, MT.
Azam, S., and Q. Li. 2010. Tailings dam failures: a review of the last one hundred years. Geotechnical
News December: 50-53.
Bisson, P., and R. Bilby. 1982. Avoidance of suspended sediment by juvenile coho salmon. North
American Journal of Fisheries Management 2:371-374.
Blight, G. E. 2010. Geotechnical Engineering for Mine Waste Storage Facilities. Boca Raton, FL: CRC Press.
Bogan, D., D. Rinella, and R. Shaftel. 2012. Baseline Macroinvertebrate and Diatom Surveys in Wadeable
Streams of the Kvichak and Nushagak Watersheds, Bristol Bay, Alaska. Anchorage, AK: Alaska Natural
Heritage Program, University of Alaska. Prepared for The Nature Conservancy, Anchorage, AK.
Borgmann, U., C. Janssen, R. Blust, K. Brix, R. Dwyer, R. Erickson, L. Hare, S. Luoma, P. Paquin, C. Roberts,
and W.-X. Wang. 2005. Incorporation of dietborne metals exposure into regulatory frameworks.
Pages 153-190 in J. Meyer, W. Adams, K. Brix, S. Luoma, D. Mount, W. Stubblefield, and C. Wood,
editors. Toxicity of Dietborne Metals to Aquatic Organisms. Pensacola, FL: SETAC Press.
Brabets, T. P. 1992. Hydrologic Data for the Lower Copper River, Alaska, May to September 1991. Open-
File Report 92-89. Anchorage, AK: U.S. Geological Survey.
Breitenbach, A. 2007. Rockfillfor Mines. Infomine E-Book. Available:
http://www.infomine.com/publications/docs/E-Book_Rockfill_for_Mines.pdf.
Brown, R. S., W. A. Hubert, and S. F. Daly. 2011. A primer on winter, ice, and fish: what fisheries biologists
should know about winter ice processes and stream-dwelling fish. Fisheries 36:8-26.
Browne, R. 2011. Tailings Dam Break Model in Nevada - A Practitioner's Perspective. PowerPoint
presentation, August 24, 2011, Colder Associates. Available:
http://www.infomine.com/publications/docs/Browne2011.pdf.
Brumbaugh, W. G., C. G. Ingersoll, N. E. Kemble, T. W. May, and J. L. Zajicek. 1994. Chemical
characterization of sediments and pore-water from the upper Clark Fork River and Milltown
Reservoir, Montana. Environmental Toxicology and Chemistry 13:1971-1983.
Bristol Bay Assessment 15-43 January 2014
-------
Chapter 15 References
Bryce, S. A., G. A. Lomnicky, and P. R. Kaufmann. 2010. Protecting sediment-sensitive aquatic species in
mountain streams through the application of biologically based streambed sediment criteria. Journal
of the North American Benthological Society 29:657-672.
Buck, G. B., C. E. Brazil, F. West, L. F. Fair, X. Zhang, and S. L. Maxwell. 2012. Stock Assessment of Chinook,
Sockeye, and Chum Salmon in the Nushagak River. Fishery Manuscript Series No. 12-05. Anchorage,
AK: Alaska Department of Fish and Game.
Bustard, D. R., and D. W. Narver. 1975. Aspects of the winter ecology of juvenile coho salmon
(Oncorhynchus kisutch] and steelhead trout (Salmo gairdneri}. Journal of the Fisheries Research
Board of Canada 32:667-680.
Canfield, T. ]., N. E. Kemble, W. G. Brumbaugh, F. J. Dwyer, C. G. Ingersoll, and J. F. Fairchild. 1994. Use of
benthic invertebrate community structure and the sediment quality triad to evaluate metal-
contaminated sediment in the upper Clark Fork River, Montana. Environmental Toxicology and
Chemistry 13:1999-2012.
Chambers, D. M., and B. Higman. 2011. Long-Term Risks of Tailings Dam Failure. Bozeman MT: Center for
Science in Public Participation. Available:
http://www.csp2.org/files/reports/Long%20Term%20Risks%20of%20Tailings%20Dam%20Failu
re%20-%20Chambers%20%26%20Higman%200ctll-2.pdf.
Chapman, D. W., and K. P. McLeod. 1987. Development of Criteria for Fine Sediment in Northern Rockies
Ecoregion. Battelle Columbus Laboratories. Prepared for EPA Region 10, EPA 910/9-87-162.
Gulp, J., F. Wrona, and R. Davies. 1986. Response of stream benthos and drift to fine sediment deposition
versus transport. Canadian Journal of Zoology 64:1345-1351.
Cunjak, R. A. 1996. Winter habitat of selected stream fishes and potential impacts from land-use activity.
Canadian Journal of Fisheries and Aquatic Sciences 53 (Supplement l):267-282.
Dalpatram, A. 2011. Estimation of Tailings Dam Break Discharges. USSD Workshop on Dam Break
Analysis Applied to Tailings Dam, Denver, CO, August 24-26, 2011,.
Dann, T. H., C. Habicht, H. A. Hoyt, T. T. Baker, and F. W. West. 2011. Genetic Stock Composition of the
Commercial Harvest of Sockeye Salmon in Bristol Bay, Alaska, 2009. Fishery Data Series 11-21. Alaska
Department of Fish and Game, Division of Sport Fish, Research and Technical Services.
Davies, M. P. 2002. Tailings impoundment failures: are geotechnical engineers listening? Geotechnical
News. September: 31-36.
Davies, M., and T. Martin. 2009. Mining market cycles and tailings dam incidents. Pages 3-14 in Tailings
and Mine Waste '09, Proceedings of the 13th International Conference on Tailings and Mine Waste.
Edmonton, AB: University of Alberta Geotechnical Center.
Bristol Bay Assessment 15-44 January 2014
-------
Chapter 15 References
Davies, M. P., T. E. Martin, and P. Lighthall. 2000. Mine tailings dams: when things go wrong. Pages 261-
273 in Tailings Dams 2000. Las Vegas, NV: Association of State Dam Safety Officials, U.S. Committee
on Large Dams.
Dawson, R. F., and K. A. Morin. 1996. Acid Mine Drainage in Permafrost Regions: Issues, Control Strategies
and Research Requirements. MEND Project 1.61.2. Report CG25047. Prepared for the Department of
Indian and Northern Affairs Canada, Ottawa, ON. Available: http://mend-nedem.org/wp-
content/uploads/2013/01/1.61.2.pdf.
Dye, J. E., and C. J. Schwanke. 2009. Report to the Alaska Board of Fisheries for the Recreational Fisheries of
Bristol Bay, 2007,2008, and 2009. Alaska Deptartment of Fish and Game, Division of Sport Fish,
Research and Technical Services.
El Mejjati, S., J. Bell, J. Botz, A. Faulkner, and S. Maxwell. 2010. Using hydroacoustic methods to
enumerate migrating salmon in the Copper River, Miles Lake sonar project 2007. Fishery Data Series
No. 10-98. Anchorage, AK: Alaska Department of Fish and Game, Division of Commercial Fisheries.
Farag, A. M., D. F. Woodward, J. N. Goldstein, W. Brumbaugh, and J. S. Meyer. 1998. Concentrations of
metals associated with mining waste in sediments, biofilm, benthic macroinvertebrates, and fish
from the Coeur d'Alene River Basin, Idaho. Archives of Environmental Contamination and Toxicology
34:119-127.
Farag, A. M., D. E. Woodward, W. Brumbaugh, J. N. Goldstein, E. MacConnell, C. Hogstrand, and F. T.
Barrows. 1999. Dietary effects of metals-contaminated invertebrates from the Coeur d'Alene River,
Idaho, on cutthroat trout. Transactions of the American Fisheries Society 128:578-592.
FERC (Federal Energy Regulatory Commission). 1991. Embankment Dams. Chapter 4 in Engineering
Guidelines for the Evaluation of Hydropower Projects. Available:
http://www.ferc.gov/industries/hydropower/safety/guidelines/eng-guide/chap4.pdf.
Fitzpatrick, A. 2012. Update: tailings dam collapse at former Gullbridge mine site. The Telegraph (St.
John's, Newfoundland). December 17, 2012.
Gee, D. M. 2008. Comparison of Dam Breach Parameter Estimators. Davis, CA: U.S. Army Corps of
Engineers, Hydrology Engineering Center.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company). Vancouver, BC.
Goldstein, J. N., D. F. Woodward, and A. M. Farag. 1999. Movements of adult Chinook salmon during
spawning migration in a metals-contaminated system, Coeur d'Alene River, Idaho. Transactions of
the American Fisheries Society 128:121-129.
Bristol Bay Assessment 15-45 January 2014
-------
Chapter 15 References
Goodyear, K. L, and S. McNeill. 1999. Bioaccumulation of heavy metals by aquatic macro-invertebrates
of different feeding guilds: a review. Science of the Total Environment 229:1-19.
Griffith, M. B., J. M. Lazorchak, and A. T. Herlihy. 2004. Relationships among exceedences of metals
criteria, the results of ambient bioassays, and community metrics in mining-impacted streams.
Environmental Toxicology and Chemistry 23:1786-1795.
Hancock, P. J. 2002. Human impacts on the stream-groundwater exchange zone. Environmental
Management 29:763-781.
Holland, W. K., F. W. Rabe, and R. C. Biggam. 1994. Recovery of macroinvertebrate communities from
metal pollution in the South Fork and mainstem of the Coeur d'Alene River, Idaho. Water
Environment Research 66:84-88.
Huusko, A., L. Greenberg, M. Stickler, T. Linnansaari, M. Nykanen, T. Vehanen, S. Koljonen, P. Louhi, and
K. Alfredsen. 2007. Life in the ice lane: the winter ecology of stream salmonids. River Research and
Applications 23:469-491.
ICOLD (International Commission on Large Dams). 2001. Tailings Dams, Risk of Dangerous Occurrences,
Lessons Learnt from Practical Experiences. United Nations Environmental Programme, Bulletin 121.
Ingersoll, C. G., W. G. Brumbaugh, F. J. Dwyer, and N. E. Kemble. 1994. Bioaccumulation of metals by
Hyalella azteca exposed to contaminated sediments from the upper Clark Fork River, Montana.
Environmental Toxicology and Chemistry 13:2013-2020.
Ingersoll, C. G., P. S. Haverland, E. L. Brunson, T. J. Canfield, F. J. Dwyer, C. E. Henke, N. E. Kemble, D. R.
Mount, and R. G. Fox. 1996. Calculation and evaluation of sediment effect concentrations for the
amphipod Hyalella azteca and the midge Chironomus riparius. Journal of Great Lakes Research
22:602-623.
Johnson, J., and P. Blanche. 2012. Catalog of Waters Important for Spawning, Rearing, or Migration of
Anadromous Fishes—Southwestern Region, Effective June 1, 2012. Special Publication No. 12-08.
Anchorage, AK: Alaska Department of Fish and Game.
Jones, E. L., T. J. Quinn, and B. W. Van Alen. 1998. Observer accuracy and precision in aerial and foot
survey counts of pink salmon in a southeast Alaska stream. North American Journal of Fisheries
Management 18:832-846.
Jones, J. I., J. F. Murphy, A. L. Collins, D. A. Sear, P. S. Naden, and P. D. Armitage. 2011. The impact of fine
sediment on macro-invertebrates. River Research and Applications 28:1055-1071.
Jones, M., T. Sands, S. Morstad, P. Salomone, G. Buck, F. West, T. Baker, and T. Krieg. 2012.2011 Bristol
Bay Area Annual Management Report. Fishery Management Report No. 12-21. Anchorage, AK: Alaska
Department of Fish and Game.
Bristol Bay Assessment 15-46 January 2014
-------
Chapter 15 References
Kemble, N. E., W. G. Brumbaugh, E. L. Brunson, F. J. Dwyer, C. G. Ingersoll, D. P. Monda, and D. F.
Woodward. 1994. Toxicity of metal-contaminated sediments from the Upper Clark Fork River,
Montana, to aquatic invertebrates and fish in laboratory exposures. Environmental Toxicology and
Chemistry 13:1985-1997.
Knighton, D. 1984. Fluvial Forms and Processes: A New Perspective. London: Edward Arnold.
Kondolf, G. 2000. Assessing salmonid spawning gravel quality. Transactions of the American Fisheries
Society 129:262-281.
MacDonald, D. D., C. G. Ingersoll, and T. A. Berger. 2000. Development and evaluation of consensus-based
sediment quality guidelines for freshwater ecosystems. Archives of Environmental Contamination
and Toxicology 39:20-31.
Marcus, W. A., G. A. Meyer, and D. R. Nimmo. 2001. Geomorphic control of persistent mine impacts in a
Yellowstone Park stream and implications for the recovery of fluvial systems. Geology 29:355-358.
Maret, T. R., and D. E. MacCoy. 2002. Fish assemblages and environmental variables associated with
hard-rock mining in the Coeur d'Alene River Basin, Idaho. Transactions of the American Fisheries
Society 131:865-884.
Maret, T. R., D. J. Cain, D. E. MacCoy, and T. M. Short. 2003. Response of benthic invertebrate assemblages
to metal exposure and bioaccumulation associated with hard-rock mining in northwestern streams,
USA. Journal of the North American Benthological Society 22:598-620.
Miller, J. R. 1997. The role of fluvial geomorphic processes in the dispersal of heavy metals from mine
sites. Journal of Geochemical Exploration 58:101-118.
Milner, A., and R. Piorkowski. 2004. Macroinvertebrate assemblages in streams of interior Alaska
following alluvial gold mining. River Research and Applications 20:719-731.
Moore, J., and H. Langner. 2012. Can a river heal itself? Natural attenuation of metal contamination in
river sediment. Environmental Science & Technology 46:2616-2623.
Morgenstern, N. R. 2011. Improving the safety of mine waste impoundments. Pages 3-10 in Tailings and
Mine Waste '10: Proceedings of the 14th International Conference on Tailings and Mine Waste, Vail, CO.
October 17-20, 2010. Boca Raton FL: CRC Press.
Nagorski, S. A., J. N. Moore, and T. E. McKinnon. 2003. Geochemical response to variable streamflow
conditions in contaminated and uncontaminated streams. Water Resources Research 39:1044-1058.
National Inventory of Dams. 2005.2005 United States National Inventory of Dams database. Available:
http://geo.usace.army.mil/pgis/f?p=397:l:0::NO. Accessed: October 6, 2005.
NDM (Northern Dynasty Minerals). 2006. Pebble Project Tailings Impoundment G - Initial Application
Report. Prepared by Knight Piesold Ltd.
Bristol Bay Assessment 15-47 January 2014
-------
Chapter 15 References
Newcombe, C., and J. Jensen. 1996. Channel suspended sediment and fisheries: a synthesis for
quantitative assessment of risk and impact. North American Journal of Fisheries Management
16:693-727.
Nielsen, J. 1992. Microhabitat-specific foraging behavior, diet, and growth of juvenile coho salmon.
Transactions of the American Fisheries Society 121:617-634.
Nimik, D., and J. Moore. 1991. Prediction of water soluble metal concentrations in fluvially deposited
tailings sediments, Upper Clark Fork Valley, Montana, U.S.A. Applied Geochemistry 6:635-646.
Nimmo, D. R., M. J. Willox, T. D. Lafrancois, P. L. Chapman, S. F. Brinkman, and J. C. Greene. 1998. Effects
of metal mining and milling on boundary waters of Yellowstone National Park, USA. Environmental
Management 22:913-926.
NRC (National Research Council). 2005. Superfund and Mining Megasites. Washington, DC.
Oswood, M. 1989. Community structure of benthic invertebrates in interior Alaskan (USA) streams and
rivers. Hydrobiologia 172:97-110.
Pascoe, G. A., R. J. Blanchet, G. Linder, D. Palawski, W. G. Brumbaugh, T. J. Canfield, N. E. Kemble, C. G.
Ingersoll, A. Farag, and J. A. Dalsoglio. 1994. Characterization of ecological risks at the Milltown
Reservoir Clark Fork River sediments Superfund site, Montana. Environmental Toxicology and
Chemistry 13:2043-2058.
Peck, P. 2007. Avoiding tailings dam failures: good practice in prevention. Workshop on the Safety of
Tailings Management Facilities, November 12,2007. Yerevan, Armenia. Communication on Progress
to the United Nations Convention of the Transboundary Effects of Industrial Accidents.
Peck, D. V., A. T. Herlihy, B. H. Hill, R. M. Hughes, P. R. Kaufmann, D. J. Klemm, J. M. Lazorchak, F. H.
McCormick, S. A. Peterson, P. L. Ringold, T. Magee, and M. R. Cappaert. 2006. Environmental
Monitoring and Assessment Program - Surface Waters Western Pilot Study: Field Operations Manual
for Wadeable Streams. EPA 600/R-06/003. Washington, DC: U.S. Environmental Protection Agency,
Office of Research and Development.
PLP (Pebble Limited Partnership). 2011. Pebble Project Environmental Baseline Document 2004 through
2008. Anchorage, AK. Available: http://www.arlis.org/docs/vol2/Pebble/2004-2008EBDIndex.pdf.
Quinn, J., R. Daviescolley, C. Hickey, M. Vickers, and P. Ryan. 1992. Effects of clay discharges on streams.
2. Benthic invertebrates. Hydrobiologia 248:235-247.
Rico, M. G., G. Benito, and A. Diez-Herrero. 2008. Floods from tailings dam failures. Journal of Hazardous
Materials 154(l-3):79-87.
Rowe, M., D. Essig, and B. Jessup. 2003. Guide to Selection of Sediment Targets for Use in Idaho TMDLs.
Boise, ID: Idaho Department of Environmental Quality.
Bristol Bay Assessment 15-48 January 2014
-------
Chapter 15 References
Ruggerone, G. T., B. A. Agler, S. E. Gilk, D. B. Molyneaux, D. J. Costello, D. E. Orabutt Jr., and D. B. Young.
2011. Habitat and growth of river-type sockeye salmon in the Kuskokwim watershed, Alaska. Pages
49-89 in S. E. Gilk, D. B. Molyneaux, and Z. W. Liller (eds.), Kuskokwim River Sockeye Salmon
Investigations, 2006 and 2007. Fishery Manuscript Series No. 11-04. Anchorage, AK: Alaska
Department of Fish and Game.
Scheuerell, M., J. Moore, D. Schindler, and C. Harvey. 2007. Varying effects of anadromous sockeye
salmon on the trophic ecology of two species of resident salmonids in southwest Alaska. Freshwater
Biology 52:1944-1956.
Schmetterling, D. A., C. G. Clancy, and T. M. Brandt. 2001. Effects of riprap bank reinforcement on stream
salmonids in the western United States. Fisheries 26:6-23.
Semko, R. S. 1954. The stocks of west Kamchatka and their commercial utilization. Izvestiia
Tikhookeanskovo Nauchno-Issledovatelskovo Instituta Rybnovo Khozyaistva i Okeanografii 41:3-109.
Translated in 1960, Fisheries Research Board of Canada Translation Series 288.
Shirazi, M. A., W. K. Seim, and D. H. Lewis. 1981. Characterization of spawning gravel and stream system
evaluation. Proceedings of the Conference on Salmon-Spawning Gravel. Pullman, WA: Washington
Water Resources Research Center.
Silva, F. T., T. W. Lambe, and W. A. Marr. 2008. Probability and risk of slope failure. Journal of
Geotechnical and Geoenvironmental Engineering 134:1691-1699.
Stantec Consulting. 2011. Dam safety review, various former mine sites, former Gullbridge Copper Mine,
NL. Final Inspection Report. Prepared for Newfoundland Department of Natural Resources, File No.
121612646, St. John's, Newfoundland.
Suter, G. W., II, and S. M. Cormier. 2011. Why and how to combine evidence in environmental
assessments: weighing evidence and building cases. Science of the Total Environment 409:1406-
1417.
USAGE (U.S. Army Corps of Engineers). 2003. Slope Stability Engineer Manual. EM 1110-2-1902.
Available: http://140.194.76.129/publications/eng-manuals/EM_1110-2-1902_sec/toc.htm.
USEPA (U.S. Environmental Protection Agency). 2001. Final Coeur d'Alene Basin Ecological Risk
Assessment. Seattle, WA: Region 10.
USEPA (U.S. Environmental Protection Agency). 2007. Aquatic Life Ambient Freshwater Quality Criteria -
Copper, 2007 Revision. Washington, DC.
USEPA (U.S. Environmental Protection Agency). 2012a. Clark Fork River Operable Unit. Available:
http://www2.epa.gov/region8/milltown-reservoir-sediments.Accessed: May 2, 2012.
Bristol Bay Assessment 15-49 January 2014
-------
Chapter 15 References
USEPA (U.S. Environmental Protection Agency). 2012b. Bunker Hill Mining and Metallurgical Superfund
Site (Coeur d'Alene Basin). Available: http://yosemite.epa.gov/rlO/cleanup.nsf/sites/bh. Accessed:
May 2, 2012.
USGS (U.S. Geological Survey). 2004. Integrated Investigations of Environmental Effects of Historical
Mining in the Basin and Boulder Mining Districts, Boulder River Watershed, Jefferson County, Montana.
Reston, VA: U.S. Geological Survey.
USGS (U.S. Geological Survey). 2005. Stream-Sediment Geochemistry in Mining-Impacted Streams:
Sediment Mobilized by Floods in the Coeur d'Alene-Spokane River System, Idaho and Washington.
Washington, DC: U.S. Geological Survey.
USGS (U.S. Geological Survey). 2007. Integrated Investigations of Environmental Effects of Historical
Mining in the Animas River Watershed, San Juan County, Colorado. Reston, VA: U.S. Geological Survey.
Wagener, S. M., and J. D. LaPerriere. 1985. Effects of placer mining on the invertebrate communities of
interior Alaska streams. Freshwater Invertebrate Biology 4:208-214.
Wenning, R. J., M. Sorensen, and V. S. Magar. 2006. Importance of implementation and residual risk
analyses in sediment remediation. Integrated Environmental Assessment and Management 2:59-65.
West, R. L, M. W. Smith, W. E. Barber, J. B. Reynolds, and H. Hop. 1992. Autumn migration and
overwintering of Arctic grayling in coastal streams of the Arctic National Wildlife Refuge, Alaska.
Transactions of the American Fisheries Society 121:709-715.
Westley, P. A. H., D. E. Schindler, T. P. Quinn, G. T. Ruggerone, and R. Hilborn. 2010. Natural habitat
change, commercial fishing, climate, and dispersal interact to restructure an Alaskan fish
metacommunity. Oecologia 163:471-484.
WISE (World Information on Service Energy). 2012. Uranium Project website. Last updated: June 12,
2012. Available: http://www.wise-uranium.org/ctfs.html.
Wood, P., and P. Armitage. 1997. Biological effects of fine sediment in the lotic environment.
Environmental Management 21:203-217.
Personal Communications
Rinella, D. J. University of Alaska, Anchorage. Unpublished data.
15.1.11 Chapter 10—Transportation Corridor
ADEC (Alaska Department of Environmental Conservation). 2003. Alaska Water Quality Criteria Manual
for Toxic and Other Deleterious Organic and Inorganic Substances. May.
ADEC (Alaska Department of Environmental Conservation). 2011a. Alaska Pollutant Discharge
Elimination System General Permit for Discharges from Large and Small Construction Activities.
Permit Number AKR100000. Anchorage, AK.
Bristol Bay Assessment 15-50 January 2014
-------
Chapter 15 References
ADEC (Alaska Department of Environmental Conservation). 20l1.b. Alaska Storm Water Guide.
December. Anchorage, AK: Division of Water.
ADF&G (Alaska Department of Fish and Game). 2012. Alaska Freshwater Fish Inventory Database.
Available: http://www.adfg.alaska.gov/index.cfm?adfg=ffinventory.main.
ADF&G and ADOT (Alaska Department of Fish and Game, Alaska Department of Transportation and
Public Facilities). 2001. Memorandum of Agreement Between Alaska Department of Fish and Game
and Alaska Department of Transportation and Public Facilities for the Design, Permitting, and
Construction of Culverts for Fish Passage, Anchorage, AK.
ADNR (Alaska Department of Natural Resources). 2013. Invasive Plants and Agricultural Pest
Management, Weed Free Gravel. Palmer, AK: Division of Agriculture, Plant Materials Center.
Available: http://dnr.alaska.gov/ag/akpmc/invasives/weed-free-gravel.htm.Accessed: March 22,
2013
ADOT (Alaska Department of Transportation and Public Facilities). 1995. Alaska Highway Drainage
Manual. Revised June 2006. Available:
http://www.dot.state.ak.us/stwddes/desbridge/pop_hwydrnman.shtml.
ADOT (Alaska Department of Transportation and Public Facilities). 2001. Iliamna-Nondalton Road
Improvements. Revised Environmental Assessment and Finding of No Significant Impact. Project No.
STP-0214(3)/51951. Prepared for Federal Highway Administration.
Angermeier, P. L., A. P. Wheeler, and A. E. Rosenberger. 2004. A conceptual framework for assessing
impacts of roads on aquatic biota. Fisheries 29:19-29.
Arsan, E. L., and J. L. Bartholomew. 2008. Potential for dissemination of the nonnative salmonid parasite
Myxobolus cerebralis in Alaska. Journal of Aquatic Animal Health 20:136-149.
Auerbach, N. A., M. D. Walker, and D. A. Walker. 1997. Effects of roadside disturbance on substrate and
vegetation properties in Arctic tundra. Ecological Applications 7:218-235.
Bader, C. D. 1997. Controlling Dust. Erosion Control 4: July/August.
Behlke, C. E., D. L. Kane, R. F. McLean, and M. D. Travis. 1991. Fundamentals of Culvert Design for Passage
of Weak-Swimming Fish. Final Report No. FHWA-AK-RD-90-10. Prepared for Alaska Department of
Transportation and Public Facilities in cooperation with U.S. Department of Transportation, Federal
Highway Administration.
Bella, E. M. 2011. Invasion prediction on Alaska trails: distribution, habitat, and trail use. Invasive Plant
Science and Management 4:296-305.
Berejikian, B. A. 1992. Feeding ecology of rainbow trout with comparisons to Arctic char in Iliamna Lake,
Alaska. M.S. thesis, University of Washington.
Bristol Bay Assessment 15-51 January 2014
-------
Chapter 15 References
Birge, W. J. 1978. Aquatic toxicology of trace elements of coal and fly ash. In J. H. Thorp and J. W. Gibbons
(eds.), Energy and Environmental Stess in Aquatic Ecosystems. U.S. Department of Energy.
Birge, W. J., J. A. Black, A. G. Westerman, and J. E. Hudson. 1979. Aquatic toxicity tests on inorganic
elements occurring in oil shale. In C. Gale (ed.), Oil Shale Symposium: Sampling, Analysis and Quality
Assurance. Cincinnati, OH: U.S. Environmental Protection Agency.
Blossey, B., L. C. Skinner, and J. Taylor. 2001. Impact and management of purple loosestrife (Lythrum
salicarid) in North America. Biodiversity and Conservation 10:1787-1807.
Bolander, P., and A. Yamada. 1999. Dust Palliative Selection and Application Guide. U.S. Department of
Agriculture, U.S. Forest Service. Available:
http://www.dot.state.ak.us/stwddes/research/assets/pdf/dust_sag.pdf.
Brown, T. G., and G. F. Hartman. 1988. Contribution of seasonally flooded lands and minor tributaries to
the production of coho salmon in Carnation Creek, British Columbia. Transactions of the American
Fisheries Society 117:546-551.
Bryant, M. D., N. D. Zymonas, and B. E. Wright. 2004. Salmonids on the fringe: abundance, species
composition, and habitat use of salmonids in high-gradient headwater streams, southeast Alaska.
Transactions of the American Fisheries Society 133:1529-1538.
Bue, B. G., S. M. Fried, S. Sharr, D. G. Sharp, J. A. Wilcock, and H. J. Geiger. 1988. Estimating salmon
escapement using area-under-the-curve, aerial observer efficiency, and stream-life estimates: the
Prince William Sound pink salmon example. North Pacific Anadromous Fish Commission Bulletin
1:240-250.
Burnett, K. M., G. H. Reeves, D. J. Miller, S. Clarke, K. Vance-Borland, and K. Christiansen. 2007.
Distribution of salmon-habitat potential relative to landscape characteristics and implications for
conservation. Ecological Applications 17(1):66-80.
Christen, D. C., and G. R. Matlack. 2009. The habitat and conduit functions of roads in the spread of three
invasive plant species. Biological Invasions 11:453-465.
Collen, P., and R. J. Gibson. 2001. The general ecology of beavers (Castor spp.), as related to their
influence on stream ecosystems and riparian habitats, and the subsequent effects on fish—A review.
Reviews in Fish Biology and Fisheries 10:439-461.
Copp, G. H. 1989. The habitat diversity and fish reproductive function of floodplain ecosystems.
Environmental Biology of Fishes 26:1-27.
Cunjak, R. A. 1996. Winter habitat of selected stream fishes and potential impacts from land-use activity.
Canadian Journal of Fisheries and Aquatic Sciences 53(Supplement l):267-282.
Darnell, R. M., W. E. Pequegnat, F. J. Benson, and R. A. Defenbaugh. 1976. Impacts of Construction
Activities in Wetlands of the United States. Washington, DC: U.S. Environmental Protection Agency.
Bristol Bay Assessment 15-52 January 2014
-------
Chapter 15 References
Dekar, M. P., R. S. King, J. A. Back, D. F. Whigham, and C. M. Walker. 2012. Allochthonous inputs from
grass-dominated wetlands support juvenile salmonids in headwater streams: evidence from stable
isotopes of carbon, hydrogen, and nitrogen. Freshwater Science 31:121-132.
Demory, R. L, R. L. Orrell, and D. R. Heinle. 1964. Spawning Ground Catalog of the Kvichak River System,
Bristol Bay, Alaska. Issue 488 of Special Scientific Report—Fisheries. Washington, DC: U.S.
Department of Interior, Bureau of Commercial Fisheries.
Denton, K. P., H. B. Rich, Jr., and T. P. Quinn. 2009. Diet, movement, and growth of Dolly Varden in
response to sockeye salmon subsidies. Transactions of the American Fisheries Society 138(6) :1207-
1219.
Ehrenfeld, J. G., and J. P. Schneider. 1991. Chamaecyparis thyoides wetlands and suburbanization: effects
on hydrology, water quality and plant community composition. Journal of Applied Ecology 28:467-
490.
Environment Canada. 2005. Profile of Chloride-based Dust Suppressants Used in Canada. Gatineau,
Quebec: Government of Canada, Road Salts Working Group. Available:
http://www.ec.gc.ca/nopp/roadsalt/reports/en/profile.cfm.
ESRI (Environmental Systems Research Institute). 2013. World Imagery. ArcMap 10.0.
FHWA (Federal Highway Administration). 2012. Stream Stability at Highway Structures. 4th edition.
Publication No. FHWA-HIF-12-004 HEC-20. U.S. Department of Transportation.
Forman, R. T. 2000. Estimate of the area affected ecologically by the road system in the United States.
Conservation Biology 14:31-35.
Forman, R. T. T., and L. E. Alexander. 1998. Roads and their major ecological effects. Annual Review of
Ecology and Systematics 29:207-231.
FSSSWG (Forest Service Stream-Simulation Working Group). 2008. Stream Simulation: An Ecological
Approach to Providing Passage for Aquatic Organisms at Road-Stream Crossings. U.S. Department of
Agriculture, U.S. Forest Service.
Furniss, M. J., T. D. Roelofs, and C. S. Yee. 1991. Road construction and maintenance. Pages 29-323 in W.
R. Meehan (ed.), Influences of Forest and Rangeland Management on Salmonid Fishes and Their
Habitats. American Fisheries Society Special Publication 19.
Gesch, D. 2007. Chapter 4-The National Elevation Dataset. Pages 99-118 in Maune, D. (ed.), Digital
Elevation Model Technologies and Applications: The DEM Users Manual. 2nd edition. Bethesda, MD:
American Society for Photogrammetry and Remote Sensing.
Gesch, D., M. Oimoen, S. Greenlee, C. Nelson, M. Steuck, and D. Tyler. 2002. The National Elevation
Dataset. Photogrammatic Engineering and Remote Sensing 68(1):5-11.
Bristol Bay Assessment 15-53 January 2014
-------
Chapter 15 References
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company). Vancouver, BC.
Gibson, R. ]., R. L. Haedrich, and C. M. Wernerheirn. 2005. Loss offish habitat as a consequence of
inappropriately constructed stream crossings. Fisheries 30:10-17.
Gomez-Uchida, D., J. E. Seeb, M. J. Smith, C. Habicht, T. P. Quinn, and L. W. Seeb. 2011. Single nucleotide
polymorphisms unravel hierarchical divergence and signatures of selection among Alaskan sockeye
salmon (Oncorhynchus nerka] populations. BMC Evolutionary Biology 11(1):48.
Grout, J. A., C. D. Levings, and J. S. Richardson. 1997. Decomposition rates of purple loosestrife (Lythrum
salicarid) and Lyngbyei's sedge [Carex lyngbyei) in the Fraser River estuary. Estuaries 20:96-102.
Gucinski, H., M. J. Furniss, R. R. Ziemer, and M. H. Brookes. 2001. Forest Roads: A Synthesis of Scientific
Information. General Technical Report, PNW-509. Portland, OR: U.S. Department of Agriculture.
Hamilton, T. D. 2007. Surficial Geologic Map of the Pebble Project. Report C-l. Anchorage, AK: The Pebble
Limited Partnership.
Hancock, P. J. 2002. Human impacts on the stream-groundwater exchange zone. Environmental
Management 29:763-781.
Harwood, D. W., and E. R. Russell. 1990. Present Practices of Highway Transportation of Hazardous
Materials, August 86 - November 89. Kansas City, MO: Midwest Research Institute; Washington, DC:
Federal Highway Administration.
Hastings, K. 2005. Long-Term Persistence of Isolated Fish Populations in the Alexander Archipelago.
Missoula, MT: University of Montana.
Hilderbrand, R. H., and J. L. Kershner. 2000. Conserving inland cutthroat trout in small streams: how
much stream is enough? North American Journal of Fisheries Management 20:513-520.
Hoover, J. 1981. Emission-Oriented Dust Control and Surface Improvement Processes for Unpaved Roads.
Project 1308. Ames, IA: Iowa State University.
Hoover, J. M., K. L. Bergeson, D. E. Fox, C. K. Denny, and R. L. Handy. 1973. Surface Improvement and Dust
Palliation of Unpaved Secondary Roads and Streets. Final report. Iowa Highway Research Board
Project HR-151.
Johnson, J., and P. Blanche. 2012. Catalog of Waters Important for Spawning, Rearing, or Migration of
Anadromous Fishes—Southwestern Region, Effective June 1, 2012. Special Publication No. 12-08.
Anchorage, AK: Alaska Department of Fish and Game.
Johnson, L. E., A. Ricciardi, and J. T. Carlton. 2001. Overland dispersal of aquatic invasive species: a risk
assessment of transient recreational boating. Ecological Applications 11:1789-1799.
Bristol Bay Assessment 15-54 January 2014
-------
Chapter 15 References
Jones, E. L, III, S. Heinl, and K. Pahlke. 2007. Aerial counts. Pages 399-410 in D. H. Johnson, B. M. Shrier, J.
S. O'Neal, J. A. Knutzen, X. Augerot, T. A. O'Neil, and T. N. Pearsons, (eds.), Salmonid Field Protocols
Handbook: Techniques for Assessing Status and Trends in Salmon and Trout Populations. Bethesda,
MD: American Fisheries Society.
Kane, D. L., and P. M. Wellen. 1985. Hydraulic Evaluation of Fish Passage through Roadway Culverts in
Alaska. Report No. FHWA-AK-RD-85-24 & 85-24A. Water Center/Institute of Northern Engineering,
Fairbanks AK, in cooperation with U.S. Department of Transportation, Federal Highway
Administration.
King, R. S., C. M. Walker, D. F. Whigham, S. Baird, and J. A. Back. 2012. Catchment topography and
wetland geomorphology drive macroinvertebrate community structure and juvenile salmonid
distributions in southcentral Alaska headwater streams. Freshwater Science 31(2):341-364.
Lake and Peninsula Borough. 2009. Multi-Hazard Mitigation Plan. February.
Lang, D. W., G. H. Reeves, J. D. Hall, and M. S. Wipfli. 2006. The influence of fall-spawning coho salmon
(Oncorhynchus kisutch] on growth and production of juvenile coho salmon rearing in beaver ponds
on the Copper River Delta, Alaska. Canadian Journal of Fisheries and Aquatic Sciences 63:917-930.
Langill, D. A., and P. J. Zamora. 2002. An Audit of Small Culvert Installations in Nova Scotia: Habitat Loss
and Habitat Fragmentation. 2422. Dartmouth, Nova Scotia: Canadian Department of Fisheries and
Oceans, Habitat Management Division.
Lee, J. E., and S. L. Chown. 2009. Quantifying the propagule load associated with the construction of an
Antarctic research station. Antarctic Science 21:471-475.
Mason C. L., K. L. Casavant, B. R. Lippke, D. K. Nguyen, and E. Jessup. 2008. The Washington Log Trucking
Industry: Costs and Safety Analysis. University of Washington Rural Technology Initiative and
Washington State University Transportation Research Group. Report to the Washington State
Legislature.
Meka, J. M., E. E. Knudsen, D. C. Douglas, and R. B. Benter. 2003. Variable migratory patterns of different
adult rainbow trout life history types in a southwest Alaska watershed. Transactions of the American
Fisheries Society 132:717-732.
Merz, J. E., J. R. Smith, M. L. Workman, J. D. Setka, and B. Mulchaey. 2008. Aquatic macrophyte
encroachment in Chinook salmon spawning beds: lessons learned from gravel enhancement
monitoring in the Lower Mokelumne River, California. North American Journal of Fisheries
Management 28:1568-1577.
Morgan, E. L., W. F. Porak, and J. A. Arway. 1983. Controlling acidic-toxic metal leachates from southern
Appalachian construction slopes: mitigating stream damage. Transportation Research Record
948:10-16.
Bristol Bay Assessment 15-55 January 2014
-------
Chapter 15 References
Morstad, S. 2003. Kvichak River Sockeye Salmon Spawning Ground Surveys, 1955-2002. Regional
Information Number 2A02-32. Anchorage, AK: Alaska Department of Fish and Game.
Mount, D. R., D. D. Gully, J. R. Hockett, T. D. Garrison, and J. M. Evans. 1997. Statistical models to predict
the toxicity of major ions to Ceriodaphnia dubia, Daphnia magna and Pimephales promelas (fathead
minnow). Environmental Toxicolology and Chemistry 16:2009-2019.
Nehring, R. B., and P. G. Walker. 1996. Whirling disease in the wild: the new reality in the intermountain
west. Fisheries 21:28-30.
Newcombe, C. P., and D. D. MacDonald. 1991. Effects of suspended sediments on aquatic ecosystems.
North American journal of Fisheries Management 11:72-82.
Nickelson, T. E., J. D. Rodgers, S. L. Johnson, and M. F. Solazzi. 1992. Seasonal changes in habitat use by
juvenile coho salmon (Oncorhynchus kisutch) in Oregon coastal streams. Canadian journal of
Fisheries and Aquatic Sciences 49:783-789.
NICNAS (National Industrial Chemicals Notification and Assessment Scheme). 2000. Sodium ethyl
xanthate, priority existing chemical secondary notification assessment. Canberra, Australia: Australian
Government Publishing Service.
Opperman, J. J., R. Luster, B. A. McKenney, M. Roberts, and A. W. Meadows. 2010. Ecologically functional
floodplains: connectivity, flow regime, and scale. Journal of the American Water Resources
Association 46:211-226.
PLP (Pebble Limited Partnership). 2011. Pebble Project Environmental Baseline Document 2004 through
2008. Anchorage, AK. Available: http://www.arlis.org/docs/vol2/Pebble/2004-2008EBDIndex.pdf.
Price, D. M., T. Quinn, and R. J. Barnard. 2010. Fish passage effectiveness of recently constructed road
crossing culverts in the Puget Sound region of Washington State. North American Journal of Fisheries
Management 30:1110-1125.
Quinn, T. P., H. B. J. Rich, D. Gosse, and N. Schtickzelle, and J. Grant. 2012. Population dynamics and
asynchrony at fine spatial scales: a case history of sockeye salmon (Oncorhynchus nerka] population
structure in Alaska, USA. Canadian Journal of Fish and Aquatic Science 69:297-306.
Reid, L. M., and T. Dunne. 1984. Sediment production from forest road surfaces. Water Resources
Research 20:1753-1761.
Richardson, E. V., B. Simons, S. Karaki, M. Mahmood, and M. A. Stevens. 1975. Highways in the River
Environment: Hydraulic and Environmental Design Considerations Training and Design Manual.
Washington, DC: U.S. Department of Transportation.
Ruff, C. P., D. E. Schindler, J. B. Armstrong, K. T. Bentley, G. T. Brooks, G. W. Holtgrieve, M. T. McGlauflin, C.
E. Torgersen, and J. E. Seeb. 2011. Temperature-associated population diversity in salmon confers
benefits to mobile consumers. Ecology 92:2073-2084.
Bristol Bay Assessment 15-56 January 2014
-------
Chapter 15 References
Russell, R. 1977. Rainbow Trout Life History Studies in Lower Talarik Creek - Kvichak Drainage.
Anchorage, AK: Alaska Department of Fish and Game.
Schultz, R., and E. Dibble. 2012. Effects of invasive macrophytes on freshwater fish and
macroinvertebrate communities: the role of invasive plant traits. Hydrobiologia 684:1-14.
Shaftel, R., R. King, and J. Back. 2011. Breakdown rates, nutrient concentrations, and macroinvertebrate
colonization of bluejoint grass litter in headwater streams of the Kenai Peninsula, Alaska. Journal of
the North American Benthological Society 30:386-398.
Slaughter, C. W. 1982. Occurrence and recurrence of aufeis in an upland taiga catchment. Pages 182-188
in Proceedings of the 4th Canadian Permafrost Conference. National Research Council of Canada,
Ottawa, ON.
Smokorowski, K. E., and T. C. Pratt. 2007. Effect of a change in physical structure and cover on fish and
fish habitat in freshwater ecosystems: a review and meta-analysis. Environmental Reviews 15:15-41.
Sommer, T. R., M. L. Nobriga, W. C. Harrell, W. Batham, and W. J. Kimmerer. 2001. Floodplain rearing of
juvenile Chinook salmon: evidence of enhanced growth and survival. Canadian Journal of Fisheries
and Aquatic Sciences 58:325-333.
Spellman, B. T., and T. L. Wurtz. 2011. Invasive sweetclover (Melilotus alba] impacts native seedling
recruitment along floodplains of interior Alaska. Biological Invasions 13:1779-1790.
Stanford, J. A., and J. V. Ward. 1993. An ecosystem perspective of alluvial rivers - connectivity and the
hyporheic corridor. Journal of the North American Benthological Society 12:48-60.
Suttle, K., M. Power, J. Levine, and C. McNeely. 2004. How fine sediment in riverbeds impairs growth and
survival of juvenile salmonids. Ecological Applications 14:969-974.
Urgenson, L. S., S. H. Reichard, and C. B. Halpern. 2009. Community and ecosystem consequences of giant
knotweed (Polygonum sachalinense] invasion into riparian forests of western Washington, USA.
Biological Conservation 142:1536-1541.
USEPA (U.S. Environmental Protection Agency). 1985. Ambient Water Quality Criteria for Cyanide -
1984. Washington, DC.
USEPA (U.S. Environmental Protection Agency). 2003a. Final Environmental Impact Statement, Pogo Gold
Mine Project. Seattle, WA: Region 10.
USEPA (U.S. Environmental Protection Agency). 2003b. EPA and Hardrock Mining: A Source Book for
Industry in the Northwest and Alaska. Appendix H. Erosion and Sedimentation. Seattle, WA: Region
10.
USEPA (U.S. Environmental Protection Agency). 2006. Industrial Stormwater Fact Sheet Series. Sector G:
Metal Mining (Ore Mining and Dressing) Facilities. EPA-833-F-06-022. Office of Water.
Bristol Bay Assessment 15-57 January 2014
-------
Chapter 15 References
USEPA (U.S. Environmental Protection Agency). 2013. ECOTOX User Guide: ECOTOXicology Database
System. Version 4.0. Available: http://www.epa.gov/ecotox. Accessed: October 2013.
USFWS (U.S. Fish and Wildlife Service). 2012. National Wetlands Inventory. Available:
http://www.fws.gov/wetlands/Wetlands-Mapper.html.Accessed: December 11, 2012.
USGS (U.S. Geological Survey). 2012. National Hydrography Dataset. High Resolution, Alaska. Available:
ftp://nhdftp.usgs.gov/DataSets/Staged/States/FileGDB/HighResolution. Accessed: October 16,
2012.
USGS (U.S. Geological Survey). 2013. National Elevation Dataset (2 arc second), ArcGrid. August 26.
[n59wl55.zip, n59wl56.zip, n59wl57.zip, n59wl58.zip, n60wl54.zip, n60wl55.zip, n60wl56.zip,
n60wl57.zip, n60wl58.zip, n60wl59.zip, n60wl60.zip, n61wl53.zip, n61wl54.zip, n61wl55.zip,
n61wl56.zip, n61wl57.zip, n61wl58.zip, n61wl59.zip, n61wl60.zip] Available:
http://viewer.nationalmap.gov/viewer/.
Van Bohemen, H. D., and W. H. J. Van de Laak. 2003. The influence of road infrastructure and traffic on
soil, water, and air quality. Environmental Management 31:50-68.
Van Hassel, J. H., J. J. Hey, and D. L. Garling. 1980. Heavy metals in a stream ecosystem at sites near
highways. Transactions of the American Fisheries Society 109:636-643.
Walker, D. A., and K. R. Everett. 1987. Road dust and its environmental impact on Alaskan taiga and
tundra. Arctic and Alpine Research 19:479-489.
Walker, C. M., R. S. King, D. F. Whigham, and S. J. Baird. 2012. Landscape and wetland influences on
headwater stream chemistry in the Kenai Lowlands, Alaska. Wetlands 32(2):301-310.
Ware, C., D. M. Bergstrom, E. Muller, and I. G. Alsos. 2012. Humans introduce viable seeds to the Arctic on
footwear. Biological Invasions 14:567-577.
Weaver, W.E., O.K. Hagans, and M.A. Madej. 1987. Managing forest roads to control cumulative erosion
and sedimentation effects. Pages 119-124 in Proceedings of the California Watershed Management
Conference, 18-20 November, 1986, West Sacramento, California. Report 11. Berkeley, CA: Wildland
Resources Center, University of California.
Weber-Scannell, P. K., and L. K. Duffy. 2007. Effects of total dissolved solids on aquatic organisms: a
review of literature and recommendation for salmonid species. American Journal of Environmental
Sciences 3:1-6.
Whitton, B. A., N. T. W. Ellwood, and B. Kawecka. 2009. Biology of the freshwater diatom Didymosphenia:
a review. Hydrobiologia 630:1-37.
Wipfli, M. S., and C. V. Baxter. 2010. Linking ecosystems, food webs, and fish production: subsidies in
salmonid watersheds. Fisheries 35(8):373-387.
Bristol Bay Assessment 15-58 January 2014
-------
Chapter 15 References
Wissmar, R. C., R. K. Timm, and M. D. Bryant. 2010. Radar-derived digital elevation models and field-
surveyed variables to predict distribution of juvenile coho salmon and Dolly Varden in remote
streams of Alaska. Transactions of the American Fisheries Society 139:288-302.
Withycombe, E., and R. Dulla. 2006. Alaska Rural Dust Control Alternatives. Report No. SR2006-03-03.
Prepared by Sierra Research, Inc. for Alaska Department of Environmental Conservation.
Woody, C. A., and S. L. O'Neal. 2010. Fish Surveys in Headwater Streams of the Nushagak and Kvichak
River Drainages, Bristol Bay, Alaska, 2008-2010. Anchorage, AK: Fisheries Research and Consulting.
Wurtz, T. L., M. J. Macander, and B. Spellman, T. 2010. Spread of invasive plants from roads to river
systems in Alaska: a network model. Pages 699-708 in J. M. Pye, H. M. Rauscher, Y. Sands, D. C. Lee,
and J. S. Beatty (eds.), Advances in Threat Assessment and Their Application to Forest and Rangeland
Management. PNW-GTR-802. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific
Northwest and Southern Research Stations.
Young, M. K., P. M. Geunther-Gloss, and A. D. Ficke. 2005. Predicting cutthroat trout (Oncorhynchus
clarki) abundance in high-elevation streams: revisiting a model of translocation success. Canadian
Journal of Fisheries and Aquatic Sciences 62:2399-2408.
Personal Communications
Morstad, S. Fishery Biologist III, ADF&G. September 2011—email of unpublished data to Rebecca
Shaftel.
15.1.12 Chapter 11—Pipeline Failures
ADEC (Alaska Department of Environmental Conservation). 2011. Water Quality Standards. 18-AAC-70.
ADEC (Alaska Department of Environmental Conservation). 2012. Situation Report: Fairbanks Gold
Mining, Inc. Cyanide Water Spill. Anchorage, AK: Alaska Department of Environment and
Conservation, Division of Spill Prevention and Response, Prevention and Emergency Response
Program.
Alexander, V., and K. VanCleve. 1983. The Alaska pipeline: a success story. Annual Review of Ecology and
Systematics 14:443-463.
API (American Petroleum Institute). 2003. Robust Summary of Information on Distillate Fuels. Report No.
201-14835B2. Washington, DC: American Petroleum Institute.
Barrett, D. 2012. Notice of Probable Violation and Proposed Civil Penalty: Letter to Mr. Richard Adams,
Vice President, U.S. Operations, Enbridge Energy, Limited Partnership. Kansas City, MO: U.S.
Department of Transportation, Pipeline and Hazardous Materials Safety Administration.
Bristol Bay Assessment 15-59 January 2014
-------
Chapter 15 References
Bass, M. L, J. R. Voshell, and R. W. Young. 1987. Assessment of damage to aquatic communities resulting
from a refined oil product spill. Pages 353-378 in J. H. Vandermeulen and S. E. Hrudey (eds.), Oil in
Freshwater: Chemistry, Biology, Countermeasure Technology. New York: Pergamon Press.
Briceno, F., and F. Bajak. 2012. Townsfolk sickened after Peru toxic spill. BusinessWeek. Available:
http://bigstory.ap.org/article/townsfolks-sickened-after-peru-toxic-spill.
Bury, R. B. 1972. The effects of diesel fuel on a stream fauna. California Fish and Game 58:291-295.
Caleyo, F. 2007. Statistical and tolerance uncertainties in pipeline failure rate estimates. Villahermosa,
Mexico: Noveno Congreso y Expo Internacional de Ductos.
Caterpillar Global Mining. 2009. Minera Antamina: setting the standard, driving the future. Viewpoint:
Perspectives on Modern Mining 2009:12-19.
Clap, L. 2004. Xstrata's "Best Practice" in Argentina. MAC: Mines and Communities. Available:
http://www.minesandcommunities.org/article.php?a=5750&l=l.
Coghlan, S. M., Jr., and J. L. Lund. 2005. Rapid assessment of benthic faunal responses to a small
petroleum spill in a headwaters stream. Journal of Freshwater Ecology 20:777-779.
Das, P. K. M. K., and S. K. Konar. 1988. Acute toxicity of petroleum products, crude oil and oil refinery
effluent on plankton, benthic invertebrates and fish. Environment and Ecology 6:885-891.
DNV Consulting. 2006. Frequency-Volume Study of Keystone Pipeline. Report No. 70015849-2 rev 1.
Prepared for TransCanada Pipelines, Ltd.
Earthworks. 2012. U.S. Copper Porphyry Mines Report: The Track Record of Water Quality Impacts
Resulting from Pipeline Spills, Tailings Failures and Water Collection and Treatment Failure.
Washington, DC.
ERCB (Energy Resources Conservation Board). 2013. ST57-2012: Field Surveillance and Operations
Branch - Field Operations Provincial Summary 2011. Calgary, AB.
EUB (Energy and Utilities Board). 1998. Pipeline Performance in Alberta 1980-1997. Calgary, AB: Alberta
Energy and Utilities Board.
Geosphere and CH2M Hill. 2006. Hydrocarbon Characterization for Use in the Hydrocarbon Risk
Calculator and Example Characterizations of Selected Alaska Fuels, Technical Background Document.
Draft report. Prepared for the Alaska Statement of Cooperation Working Group.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company). Vancouver, BC.
Bristol Bay Assessment 15-60 January 2014
-------
Chapter 15 References
Giddings, ]., B. Parkhurst, C. Gehrs, and R. Millemann. 1980. Toxicity of a coal liquefaction product to
aquatic organisms. Bulletin of Environmental Contamination and Toxicology 25:1-6.
Goldstein, J. N., D. F. Woodward, and A. M. Farag. 1999. Movements of adult Chinook salmon during
spawning migration in a metals-contaminated system, Coeur d'Alene River, Idaho. Transactions of
the American Fisheries Society 128:121-129.
Guard, H. E., N. G. James, and R. B. Laughlin, Jr. 1983. Characterization of Gasolines, Diesel Fuels and Their
Water Soluble Fractions. Oakland, CA: Naval Biosciences Laboratory, Naval Supply Center.
Hayes, M. 0., R. Hoff, J. Michel, D. Scholz, and G. Shigenaka. 1992. Introduction to Coastal Habitats and
Biological Resources for Oil Spill Response. NMRAD-92-4. Seattle, WA: National Oceanic and
Atmospheric Administration, Coastal Monitoring and Bioeffects Assessment Division.
Hebert, G. W., and R. H. Kussat. 1972. A Laboratory Evaluation of the Toxicity of Certain Oils and Chemical
Oil Dispersants to Juvenile Coho Salmon and Staghorn Sculpins. Technical Report 1972-6. Canada
Department of the Environment, Fisheries Service, Pacific Region, Northern Operations Branch.
HydroQual. 2007. Biotic Ligand Model Windows Interface, Version 2.2.3, User's Guide and Reference
Manual. Mahwah, NJ: HydroQual, Inc.
Ingersoll, C. G., P. S. Haverland, E. L. Brunson, T. J. Canfield, F. J. Dwyer, C. E. Henke, N. E. Kemble, D. R.
Mount, and R. G. Fox. 1996. Calculation and evaluation of sediment effect concentrations for the
amphipod Hyalella azteca and the midge Chironomus riparius. Journal of Great Lakes Research
22:602-623.
Khan, N., M. A. Warith, and G. Luk. 2007. A comparison of acute toxicity of biodiesel, biodiesel blends,
and diesel on aquatic organisms. Journal of the Air and Waste Management Association 57:286-296.
Kubach, K. M., M. C. Scott, and J. S. Bulak. 2011. Recovery of a temperate riverine fish assemblage from a
major diesel oil spill. Freshwater Biology 56:503-518.
Lockhart, W. L., R. W. Danell, and D. A. Murray. 1987. Acute toxicity bioassays with petroleum products:
influence of exposure conditions. Pages 335-344 in J. H. Vandermeulen and S. E. Hrudey (eds.), Oil in
Freshwater: Chemistry, Biology, Countermeasure Technology. New York: Pergamon Press.
Lytle, D. A., and B. L. Peckarsky. 2001. Spatial and temporal impacts of a diesel fuel spill on stream
invertebrates. Freshwater Biology 46:693-704.
MacDonald, D. D., C. G. Ingersoll, and T. A. Berger. 2000. Development and evaluation of consensus-based
sediment quality guidelines for freshwater ecosystems. Archives of Environmental Contamination
and Toxicology 39:20-31.
Meyer, J. S., and W. J. Adams. 2010. Relationship between biotic ligand model-based water quality
criteria and avoidance and olfactory responses to copper bv fish. Environmental Toxicology and
Chemistry 29:2096-2130.
Bristol Bay Assessment 15-61 January 2014
-------
Chapter 15 References
Minera Alumbrera. 2004. Minera Alumbrera Sustainability Report for 2004. Catamarca, Argentina:
Minera Alumbrera Ltd. and Xstrata Copper.
Minera Alumbrera. 2005. Minera Alumbrera Sustainability Report 2005. Catamarca, Argentina: Minera
Alumbrera YMAD-UTE and Xstrata Copper. Available:
http://www.alumbrera.com.ar/download/articulos/Sustainability_Rep%20_Def.pdf.
Minera Alumbrera. 2006. Minera Alumbrera Sustainability Report 2006. Catamarca, Argentina: Minera
Alumbrera and Xstrata Copper. Available:
http://www.goldcorp.com/files/docs_sustainability/alumbrera_sustainability_report2006-en.pdf.
Minera Alumbrera. 2007. Minera Alumbrera Sustainability Report 2007. Catamarca, Argentina: Minera
Alumbrera and Xstrata Copper.
Minera Alumbrera. 2008. Minera Alumbrera Sustainability Report 2008. Catamarca, Argentina: Minera
Alumbrera YMAD-UTE and Xstrata Copper. Available:
http://www.alumbrera.com.ar/download/articulos/Al_Sustainab ility_Report_2008.pdf.
Minera Alumbrera. 2009. Minera Alumbrera Sustainability Report 2009. Catamarca, Argentina: Minera
Alumbrera YMAD-UTE and Xstrata Copper.
Minera Alumbrera. 2Q\Q.Minera Alumbrera Sustainability Report 2010. Catamarca, Argentina: Minera
Alumbrera YMDA-UTE, and Xstrata Copper. Available:
http://www.alumbrera.com.ar/download/articulos/sustainability_report_2010.pdf.
Mining Watch Canada. 2005. Wheaton River Profits from Destruction ofBajo La Alumbrera, Argentina.
August 12. Available: http://www.miningwatch.ca/wheaton-river-profits-destruction-bajo-la-
alumbrera-argentina.
Mos, L, G. A. Cooper, K. Serben, M. Cameron, and B. F. Koop. 2008. Effects of diesel on survival, growth,
and gene expression in rainbow trout (Oncorhynchus mykiss] fry. Environmental Science &
Technology 42:2656-2662.
Nauman, J. W., and D. R. Kernodle. 1975. The effect of a fuel oil spill on benthic invertebrates and water
quality on the Alaskan Arctic Slope, Happy Valley Creek near Sagwon, Alaska. Journal of Research of
the U.S. Geological Survey 3:495-500.
NICNAS (National Industrial Chemicals Notification and Assessment Scheme). 1995. Sodium Ethyl
Xanthate, Priority Existing Chemical No. 5. Full Public Report. Canberra, Australia: Australian
Government Publishing Service.
NICNAS (National Industrial Chemicals Notification and Assessment Scheme). 2000. Sodium Ethyl
Xanthate, Priority Existing Chemical Secondary Notification Assessment. Report No. 5S. Canberra,
Australia: Australian Government Publishing Service.
Bristol Bay Assessment 15-62 January 2014
-------
Chapter 15 References
NOAA (National Oceanic and Atmospheric Administration). 2006. Small diesel spills [50-5,000gallons).
Seattle, WA: NOAA Office of Response and Restoration.
NOAA (National Oceanic and Atmospheric Administration) and API (American Petroleum Institute).
1994. Options for Minimizing Environmental Impacts of Freshwater Spill Response. NOAA Hazardous
Materials Response and Assessment Division and API.
O'Brien's Response Management. 2009. TransCanada Emergency Response Plan: Worst Case Discharge
Analysis and Scenarios. Appendix B, Version 21.0.1. O'Brien's Response Management Inc.
OGP (International Association of Oil & Gas Producers). 2010. Risk Assessment Data Directory, Riser &
pipeline release frequencies. Report No. 434-4. March 17.
Overstreet, R., and J. A. Gait. 1995. Physical Processes Affecting the Movement and Spreading of Oils in
Inland Environments. HMRAD 95-7. Seattle, WA: National Oceanic and Atmospheric Administration,
Hazardous Materials Response and Assessment Division.
Pereira. S., V. Araiijo, M Reboucas, F. Vieira, M. de Almeida, F. Chinalia, and I. Nascimento. 2012. Toxicity
of biodiesel, diesel, and biodiesel/diesel blends: comparative sub-lethal effects of water-soluble
fractions to microalgae species. Bulletin of Environmental Contamination and Toxicology 88:234-
238.
PHMSA (Pipeline and Hazardous Materials Safety Administration). 2012. Pipeline Basics. Pipeline Safety
Stakeholder Communications, U.S. Department of Transportation. Available:
http://primis.phmsa.dot.gov/comm/PipelineBasics.htm?nocache=7789. Accessed: February 21,
2012.
PLP (Pebble Limited Partnership). 2011. Pebble Project Environmental Baseline Document 2004 through
2008. Anchorage, AK. Available: http://www.arlis.org/docs/vol2/Pebble/2004-2008EBDIndex.pdf.
Redman, A. D., T. F. Parkerton, J. A. McGrath, and D. M. Di Toro. 2012. PETROTOX: an aquatic toxicity
model for petroleum substances. Environmental Toxicology and Chemistry 31:2498-2506.
Rice, S., J. Short, M. Carls, A. Moles, and R. Spies. 2007. The Exxon Valdez oil spill. Pages 419-520 in R.
Spies, editor. Long-Term Ecological Change in the Northern Gulf of Alaska. Amsterdam, Netherlands:
Elsevier B.V.
Schein, A., J. A. Scott, L. Mos, and P. V. Hodson. 2009. Oil dispersion increases the apparent bioavailability
and toxicity of diesel to rainbow trout (Oncorhynchus mykiss}. Environmental Toxicology and
Chemistry 28:595-602.
Smith, P., D. Snook, A. Muscatt, and A. Smith. 2010. Effects of a diesel spill on freshwater
macroinvertebrates in two urban watercourses, Wiltshire, UK. Water and Environment Journal
24:249-260.
Bristol Bay Assessment 15-63 January 2014
-------
Chapter 15 References
Suter, G. W., II, and S. M. Cormier. 2011. Why and how to combine evidence in environmental
assessments: weighing evidence and building cases. Science of the Total Environment 409:1406-
1417.
Taj, M., and T. Cespedes. 2012. Peru's Antamina mine faces fallout from spill. Reuters.com. Accessed:
August 20, 2012.
URS (URS Corporation). 2000. Final Longhorn Pipeline Environmental Assessment, Volume 4
Responsiveness Summary, Appendix D, Leak Frequencies of Other Pipeline Operations.
Van Derveer, W. D., R. J. Nadeau, and G. L. Case. 1995. A screening-level evaluation of impacts to a
Montana lotic macroinvertebrate community from a fuel oil spill. Pages 391-397 in Proceedings of
the 14th Bien, API et al. Oil Spill [Achieving and Maintaining Preparedness) Internal Conference. Long
Beach, CA. February 27-March 2,1995.
Wan, M. T., R. G. Watts, and D. J. Moul. 1990. Acute toxicity to juvenile Pacific salmonids and rainbow
trout of butoxyethyl esters of 2,4-D, 2,4-DP and their formulated product: Weedone CB and its
carrier. Bulletin of Environmental Contamination and Toxicology 45:604-611.
Personal Communications
Adams, W. General Manager Legacy, Rio Tinto. January 17, 2013—email to Glenn Suter.
15.1.13 Chapter 12—Fish-Mediated Effects
ARCO (Atlantic Richfield Company). 1998. Milltown Reservoir Sediments NPL Site, Clark Fork River
Operable Unit, Final Draft Remedial Investigation Report. Anaconda, MT.
Bersamin, A., B. Luick, E. Ruppert, J. Stern, and S. Zidenberg-Cherr. 2006. Diet quality among Yup'ik
Eskimos living in rural communities is low: the center for Alaska native health research pilot study.
Journal of the American Dietetic Association 106:1055-1063.
Braund, S. R., and Associates. 2009. Impacts of Oil and Gas Development to Barrow, Nuiqsut, Wainwright,
andAtqasuk harvesters. Report prepared for North Slope Borough, Department of Wildlife
Management.
Braund, S. R., and Associates. 2011. Chapter 23: Subsistence Uses and Traditional Knowledge, Bristol
Bay Drainages. In Pebble Limited Partnership. 2011. Pebble Project Environmental Baseline
Document2004 through 2008. Available: http://www.arlis.org/docs/vol2/Pebble/2004-
2008EBDIndex.pdf.
Braund, S. R., and J. Kruse (eds). 2009. Synthesis: Three Decades of Research on Socioeconomic Effects
Related to Offshore Petroleum Development in Coastal Alaska. MMS DCS Study 2009-006.
Brna, P. J., and L. A. Verbrugge (eds.). 2013. Wildlife Resources of the Nushagak and Kvichak River
Watersheds, Alaska. Final report. Anchorage, AK: Anchorage Fish and Wildlife Field Office, U.S. Fish
and Wildlife Service.
Bristol Bay Assessment 15-64 January 2014
-------
Chapter 15 References
CEAA (Canadian Environmental Assessment Agency). 2010. Report of the Federal Review Panel
Established by the Minister of the Environment: Taseko Mines Limited's Prosperity Gold-Copper Mine
Project. Ottawa, ON. Available: http://publications.gc.ca/site/eng/371768/publication.html.
Accessed: May 4, 2011.
Chapman, P. M., W. J. Adams, M. L. Brooks, C. G. Delos, S. N. Luoma, W. A. Maher, H. M. Ohlendorf, T. S.
Presser, and D. P. Shaw. 2010. Ecological Assessment of Selenium in the Aquatic Environment.
Pensacola, FL: Society of Environmental Toxicology and Chemistry (SETAC).
Eisler, R. 2000. Handbook of Chemical Risk Assessment: Health Hazards to Humans, Plants, and Animals.
Volume 1: Metals. Boca Raton, FL: Lewis Publishers.
Goldsmith, S. 2007. The Remote Rural Economy of Alaska. Institute of Social and Economic Research,
University of Alaska Anchorage. April 12. Available:
http://www.iser.uaa.alaska.edu/Publications/u_ak/uak_remoteruraleconomyak.pdf.
Helfield, J. M., and R. J. Naiman. 2006. Keystone interactions: salmon and bear in riparian forests of
Alaska. Ecosystems 9:167-180.
Hilderbrand, G. V., T. A. Hanley, C. T. Robbins, and D. D. Schwartz. 1999. Role of brown bears (Ursus
arctos] in the flow of marine nitrogen into a terrestrial ecosystem. Oecologia 121:546-550.
Holen, D. 2009. The dynamic context of cultural and social sustainability of communities in southwest
Alaska. Journal of Enterprising Communities 3:306-316.
Joyce, A. 2008. Risk and opportunity in British Columbia shellfisheries: the role of limited property
rights in aquaculture development. PhD thesis, University of British Columbia, Vancouver, BC.
Available: https://circle.ubc.ca/handle/2429/7569.
Kerkvliet, J., and W. Nebesky. 1997. Whaling and wages on Alaska's North Slope: a time allocation
approach to natural resource use. Economic Development and Cultural Change 45(3):651-665.
Kruse, J. A., M. Baring-Gould, W. Schneider, J. Gross, G. Knapp, and G. Sherrod. 1983.^4 Description of the
Socioeconomics of the North Slope Borough. Technical Report No. 85. Prepared by the Institute of
Social and Economic Research, University of Alaska for the Minerals Management Service, Alaska
Outer Continental Shelf Region, Anchorage, AK.
Kuhnlein, H., 0. Receveur, and H. Chan. 2001. Traditional food systems research with Canadian
indigenous peoples. International Journal ofCircumpolar Health 60:112-112.
Loring, P. A., L. K. Duffy, and M. S. Murray. 2010. A risk-benefit analysis of wild fish consumption for
various species in Alaska reveals shortcomings in data and monitoring needs. Science of the Total
Environment 408:4532-4541.
Bristol Bay Assessment 15-65 January 2014
-------
Chapter 15 References
NMFS (National Marine Fisheries Service). 2011. Effects of Oil and Gas Activities in the Arctic Ocean, Draft
Environmental Impact Statement. United States Department of Commerce, National Oceanic and
Atmospheric Administration, National Marine Fisheries Service, Office of Protected Resources.
Available: http://www.nmfs.noaa.gov/pr/pdfs/permits/arctic_deis_volume3.pdf.
NMFS (National Marine Fisheries Service). 2013. Effects of Oil and Gas Activities in the Arctic Ocean, Draft
Supplemental Environmental Impact Statement. March. Office of Protected Resources. Available:
http://www.nmfs.noaa.gov/pr/permits/eis/arctic.htm.
NRC (National Research Council). 2003. Cumulative Environmental Effects of Oil and Gas Activities on
Alaska's North Slope. Washington, DC: The National Academies Press.
Palinkas, L. A., M. A. Downs, J. S. Petterson, and J. Russell. 1993. Social, cultural, and psychological
impacts of the Exxon Valdez oil spill. Human Organization 52:1-12.
PLP (Pebble Limited Partnership). 2011. Pebble Project Environmental Baseline Document 2004 through
2008. Anchorage, AK. Available: http://www.arlis.org/docs/vol2/Pebble/2004-2008EBDIndex.pdf.
Poppel, B., J. Kruse, G. Duhaime, and L. Abryutina. 2007. Survey of Living Conditions in the Arctic [SLiCA)
Results. Anchorage, AK: Institute of Social and Economic Research, University of Alaska Anchorage.
Available: http://www.arcticlivingconditions.org. Accessed: May 4, 2011.
Storey, K., and L. C. Hamilton. 2004. Planning for the impacts of megaprojects: two North American
examples. Pages 28-302 in R. 0. Rasmussen and N. E. Koroleva (eds.), Social and Economic Impacts
in the North. Dordrecht, Netherlands: Kluwer Academic Publishers.
USEPA (U.S. Environmental Protection Agency). 1985. Ambient Water Quality Criteria for Copper - 1984.
EPA 440/5-84-031. Washington, DC.
USEPA (U.S. Environmental Protection Agency). 2009. Red Dog Mine Extension Aqqaluk Project, Final
Environmental Impact Statement. October.
USEPA (U.S. Environmental Protection Agency). 2011. A Field-Based Aquatic Life Benchmark for
Conductivity in Central Appalachian Streams [Final Report). EPA/600/R-10/023F. Washington, DC:
Office of Research and Development, National Center for Environmental Assessment.
15.1.14 Chapter 13—Cumulative Risks of Multiple Mines
ADF&G (Alaska Department of Fish and Game). 2012. Alaska Freshwater Fish Inventory Database.
Available: http://www.adfg.alaska.gov/index.cfm?adfg=ffinventory.main.
ADNR (Alaska Department of Natural Resources). 2005. Bristol Bay Area Plan [BBAP] for State Lands.
Resource Assessment and Development Section, Division of Mining, Land, and Water. Available:
http://dnr.alaska.gOV/mlw/planning/areaplans/bristol/.Accessed: December 27, 2012.
Bristol Bay Assessment 15-66 January 2014
-------
Chapter 15 References
ADNR (Alaska Department of Natural Resources). 2012a. Temporary Water Use Permit F2012-275for
TNR Gold Corp (Shotgun Prospect). Available:
http://dnr.alaska.gov/mlw/mapguide/water/twup_start_tok.cfm. Accessed: November 29, 2012.
ADNR (Alaska Department of Natural Resources). 2012b. Case Abstract, ADL 656311, James W. Caswell.
Available:
http://dnr.alaska.gov/projects/las/Case_Abstract.cfm?FileType=ADL&FileNumber=656311.
Accessed: December 10, 2012.
ADNR (Alaska Department of Natural Resources). 2012c. Case Abstracts ADL 701294 et al, Alaska Earth
Resources, Inc., Kody and Koksetna claims. Available:
http://dnr.alaska.gov/projects/las/Case_Abstract.cfm?FileType=ADL&FileNumber=701294.
Accessed: December 10, 2012.
ADNR (Alaska Department of Natural Resources). 2012d. Case Abstracts ADL 644576 et al, TNR Gold
Corp., Winchester claims. Available:
http://dnr.alaska.gov/projects/las/Case_Abstract.cfm?FileType=ADL&FileNumber=644576.
Accessed: December 10, 2012.
ADNR (Alaska Department of Natural Resources). 2012e. Case Abstracts, ADL 552871 etal, Full Metal
Minerals (USA), Inc., South Pebble claims. Available:
http://dnr.alaska.gov/projects/las/Case_Abstract.cfm?FileType=ADL&FileNumber=552871.
Accessed: December 10, 2012.
ADNR (Alaska Department of Natural Resources). 2012f. Case Abstracts, ADL 648509 etal., Alaska Earth
Services, Inc., Groundhog and Nika claims. Available:
http://dnr.alaska.gov/projects/las/Case_Abstract.cfm?FileType=ADL&FileNumber=648509.
Accessed: December 10, 2012.
ADNR (Alaska Department of Natural Resources). 2012g. Case Abstracts ADL 709821 etal., Millrock
Alaska, LLC, AUDN claims. Available:
http://dnr.alaska.gov/projects/las/Case_Abstract.cfm?FileType=ADL&FileNumber=709821.
Accessed: December 10, 2012.
ADNR (Alaska Department of Natural Resources). 2012h. State Mining Claims Dataset, Alaska State Geo-
Spatial Data Clearinghouse. Available: http://www.asgdc.state.ak.us. Accessed: November 19, 2012.
ADOT (Alaska Department of Transportation and Public Facilities). 2001. Revised Environmental
Assessment and Finding of No Significant Impact, Iliamna-Nondalton Road Improvements Project.
Project No. STP-0214(3)/51951. Available: http://www.iliamna-
nondaltonroad.com/documents.shtml. Accessed: December 18, 2012
Bristol Bay Assessment 15-67 January 2014
-------
Chapter 15 References
ADOT (Alaska Department of Transportation and Public Facilities). 2004. Southwest Alaska
Transportation Plan. Anchorage, AK. Available:
http://www.dowlhkm.com/projects/SWAKTP/newjArebsite/documents.htmW.Accessed:
December 19, 2012.
ADOT (Alaska Department of Transportation and Public Facilities). 2011a. Ambler Mining District Access.
Available: http://www.dot.alaska.gov/roadstoresources/projects.shtml.Accessed: December 6,
2012.
ADOT (Alaska Department of Transportation and Public Facilities). 2011b. Aleknagik: Wood River Bridge.
Available: http://www.brooks-alaska.com/aleknagik.Accessed: December 6, 2012.
AERI (Alaska Earth Resources, Inc.). 2008. Affidavit of Annual Labor for Mining [Kamishak claims). 2008-
000296-0. Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20080002960&District=320.
Accessed: November 29, 2012.
AHEA (Alaska Hardrock Exploration Application). 2011. Alaska Hardrock Exploration Application
A113099 by Alaska Earth Sciences, Inc., for Groundhog prospect. January 31.
AHEA (Alaska Hardrock Exploration Application). 2012. Alaska Hardrock Exploration Application
A122793 by Big Chunk Corp., July 11, 2012.
Alix Resources. 2008. News Release. Available:
http://www.mineweb.com/mineweb/content/en/mineweb-fast-news?oid=58213&sn=Detail.
Accessed: January 3, 2013
Big Chunk Corp. 2012. Affidavit of Annual Labor for Mining, Big Chunk Corporation/Liberty Star. Doc. No.
2012-000351-0. Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20120003510&District=320.
Accessed: November 29, 2012
Booth, D. B., and C. R. Jackson. 1997. Urbanization of qquatic systems—Degradation thresholds,
stormwater detention, and the limits of mitigation./ourna/ of the American Water Resources
Association 33(5):1077-1090.
Booth, D., D. Hartley, and R. Jackson. 2002. Forest cover, impervious-surface area, and the mitigation of
stormwater impacts. Journal of the American Water Resources Association 38(3):835-845.
Bristol Exploration Co., Inc. 2011. Affidavit of'Annual Labour for Mining [HSS Claims). Doc. No. 2011-
000463-0. Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20110004630&District=310.
Accessed: November 29, 2012.
Bristol Bay Assessment 15-68 January 2014
-------
Chapter 15 References
Brna, P. ]., and L. A. Verbrugge (eds.). 2013. Wildlife Resources of the Nushagak and Kvichak River
Watersheds, Alaska. Final Report. Anchorage, AK: U.S. Fish and Wildlife Service, Anchorage Fish and
Wildlife Field Office.
Charles, S., R. Bravo De La Parra, J. P. Mallet, H. Persat, and P. Auger. 2000. Annual spawning migration in
modeling brown trout population dynamics inside an arborescent river network. Ecological
Modelling 133:15-31.
Detterman, R. L., and B. L. Reed. 1973. Surflcial Deposits of the Iliamna Quadrangle, Alaska. Geological
Survey Bulletin 1368-A. U.S. Geological Survey. Available:
http://www.dggs.alaska.gOV/webpubs/usgs/b/text/bl368a.pdf.Accessed: December 17, 2012.
Fall, J. A., D. L. Holen, B. Davis, T. Krieg, and D. Roster. 2006. Subsistence Harvests and Uses of Wild
Resources in Iliamna, Newhalen, Nondalton, Pedro Bay, and Port Alsworth, Alaska 2004. Technical
Paper No. 3 02. Juneau, AK: Alaska Department of Fish and Game, Division of Subsistence.
Fall, J. A., J. C. Schichnes, M. Chythlook, and R. J. Walker. 1986. Patterns of Wild Resources Use in
Dillingham: Hunting and Fishing in an Alaskan Regional Center. Technical Paper No. 135. Alaska
Department of Fish and Game, Division of Subsistence.
Full Metal Minerals, Inc. 2008. Affidavit of'Annual Labor for Mining (South Pebble claims). Doc. No. 2008-
000298-0. Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20080002980&District=320.
Accessed: November 29, 2012.
Full Metal Minerals. 2012. Pebble South. Available:
http://www.fullmetalminerals.eom/properties/pebble-south/.Accessed: November 29, 2012.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company). Vancouver, BC.
Golden Lynx, LLC. 2009. Affidavit ofAnnual Labor for Mining (Kisa prospect]. Doc. No. 2009-000419-0.
Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20090004190&District=307.
Accessed: November 29, 2009.
Hilborn, R. 2006. Fisheries success and failure: the case of the Bristol Bay salmon fishery. Bulletin of
Marine Science 78(3):487-498.
Hilborn, R., T. Quinn, D. Schindler, and D. Rogers. 2003. Biocomplexity and fisheries sustainability.
Proceedings of the National Academy of Sciences of the United States of America 100:6564-6568.
Bristol Bay Assessment 15-69 January 2014
-------
Chapter 15 References
Holen, D., J. Stariwat, T. M. Krieg, and T. Lemons. 2012. Subsistence Harvest and Uses of Wild Resources in
Aleknagik, Clark's Point, and Manokotak, Alaska, 2008. Technical Paper No. 368. Anchorage, AK:
Alaska Department of Fish and Game, Division of Subsistence.
Johnson, J., and P. Blanche. 2012. Catalog of Waters Important for Spawning, Rearing, or Migration of
Anadromous Fishes - Southwestern Region, Effective June 1,2012. Special Publication No. 12-08.
Anchorage, AK: Alaska Department of Fish and Game.
Kennecott Exploration Co. 2011. State of Alaska Affidavit of Annual Labor for Mining Groundhog and Nika
Group of484 Claims. Doc. No. 2011-000223-0. Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20110002230&District=320.
Accessed: November 29, 2012.
Krieg, T. M., D. L. Holen, and D. Koster. 2009. Subsistence Harvests and Uses of Wild Resources in Igiugig,
Kokhanok, Koliganek, Levelock, and NewStuyahok, Alaska 2005. Technical Paper No. 322. Dillingham,
AK: Alaska Department of Fish and Game, Division of Subsistence.
Levelock Village Council. 2005. Levelock Watershed Community Planning Project, Final Summary Report.
Available: http://www.commerce.state.ak.us/dca/plans/Levelock-LUP-2005.pdf.Accessed:
December 14, 2012.
Liberty Star. 2012a. Big Chunk Super Project. Available:
http://www.libertystaruranium.com/projects/big-chunk-super-project/. Accessed: November 29,
2012.
Liberty Star. 2012b. Gold Copper Explorer Liberty Star Signs New Agreement with Northern Dynasty
Minerals Ltd, Southwest Alaska. NR 144. Available:
http://www.libertystaruranium.com/2012/ll/14/nr-144-gold-copper-explorer-liberty-star-signs-
new-agreement-with-northern-dynasty-minerals-ltd-southwest-alaska/. Accessed: December 3,
2012.
Millrock Alaska. 2012a. Affidavit ofAnnual Labor for State Mining Claims (AUDN), Millrock Alaska LLC.
Doc. No. 2012-000236-0. Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20120002360&District=310.
Accessed: November 29, 2012.
Millrock Alaska. 2012b. Affidavit of Annual Labor for State Mining Claims [KM), Millrock Alaska LLC. Doc.
No. 2012-000470-0. Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20120004700&District=307.
Accessed: November 29, 2012.
Millrock Resources. 2011. Millrock Provides Exploration Update. Available:
http://www.millrockresources.com/news/millrock_provides_exploration_update. Accessed:
November 29, 2012.
Bristol Bay Assessment 15-70 January 2014
-------
Chapter 15 References
Nehlsen, W., J. Williams, and J. Lichatowich. 1991. Pacific salmon at the crossroads—stocks at risk from
California, Oregon, Idaho, and Washington. Fisheries 16:4-21.
Northern Bonanza Trust. 2011. Affidavit of Annual Labor for Mining. Doc. No. 2011-000248-0. Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20110002480&District=320.
Accessed: November 30, 2012.
NRC (National Research Council). 1996. Upstream: Salmon and Society in the Pacific Northwest.
Washington, DC.
PLP (Pebble Limited Partnership). 2011. Affidavit ofAnnual Labor for Mining. Doc. No. 2011-000252-0.
Available:
http://dnr.alaska.gov/projects/las/Case_Abstract.cfm?FileType=ADL&FileNumber=638795.
Accessed: January 3, 2013.
Portage Creek Village Council and Residents of Portage Creek. 2006. Portage Creek Comprehensive Plan.
Available: http://www.commerce.state.ak.us/dca/plans/PortageCreek-CP-2006.pdf.
Ruckelshaus, M. H., P. Levin, J. B. Johnson, and P. M. Kareiva. 2002. The Pacific salmon wars: what science
brings to the challenge of recovering species. Annual Review of Ecology Systematics 33:655-706.
Russell, R. B. 1974. Rainbow Trout Life History Studies in Lower Talarik Creek-Kvichak Drainage. Annual
Performance Report, 1973-1974. Project F-9-6(Vol. 15), Job No. G-II-E. Juneau, AK: Alaska
Department of Fish and Game.
Russell, R. B. 1975. Rainbow Trout Life History Studies in Lower Talarik Creek-Kvichak drainage. Annual
Performance Report, 1973-1974. Project F-9-7(Vol. 16), Job No. G-II-E. Juneau, AK: Alaska
Department of Fish and Game.
Schichnes, J., and M. Chythlook. 1991. Contemporary Use of Fish and Wildlife in Ekwok, Koliganek, and
New Stuyahok, Alaska. Technical Paper No. 185. Alaska Department of Fish and Game, Division of
Subsistence.
Schindler, D. E., P. R. Leavitt, C. S. Brock, S. P. Johnson, and P. D. Quay. 2005. Marine-derived nutrients,
commercial fisheries, and production of salmon and lake algae in Alaska. Ecology 86:3225-3231.
Schindler, D. E., R. Hilborn, B. Chasco, C. P. Boatright, T. P. Quinn, L. A. Rogers, and M. S. Webster. 2010.
Population diversity and the portfolio effect in an exploited species. Nature 465:609-612.
Siedelman, D. L., P. B. Cunningham, and R. B. Russell. 1973. Life History Studies of Rainbow Trout in the
Kvichak Drainage of Bristol Bay. Annual Performance Report, 1972-1972. Project F-9-5(Vol. 14), Job
No. G-II-E. Juneau, AK: Alaska Department of Fish and Game.
Bristol Bay Assessment 15-71 January 2014
-------
Chapter 15 References
Stickman, K., A. Balluta, M. McBurney, and D. Young. 2003. K'ezghlegh, Nondalton Traditional Ecological
Knowledge of Freshwater Fish. Fisheries Project 01-075. U.S. Fish and Wildlife Service, Fisheries
Information Services. Available:
http://www.kuskokwimcouncil.org/documents/TEK/TEK%20Nondalton%20indians%20on%20fr
eshwater%2Ofish.pdf. Accessed: December 11, 2012.
Stuy Mines, LLC. 2010. Affidavit of'Annual Labor for Mining. Doc. No. 2010-000602-0. Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20100006020&District=320.
Accessed: November 29, 2012.
Szumigala, D. ]., L. A. Harbo, and J. N. Adleman. 2011. Alaska's Mineral Industry 2010. Special Report 65.
Fairbanks, AK: Alaska Department of Natural Resources, Division of Geological and Geophysical
Surveys.
Thor Gold Alaska, Inc. 2011. Affidavit of'Annual Labor for Mining (Sleitat Mountain prospect). Doc. No.
2011-000443-0. Available:
http://dnr.alaska.gov/projects/las/Case_Abstract.cfm?FileType=ADL&FileNumber=650085.
Accessed: November 29, 2012.
TNR Gold Corp. 2011. Affidavit of'Annual Labour for Mining (Shotgun prospect). Doc. No. 2011-000202-0.
Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20110002020&District=403.
Accessed: November 29, 2012.
TNR Gold Corp. 2012. Shotgun, Alaska. Available:
http://www.tnrgoldcorp.com/index.php/projects/shotgun-alaska.Accessed: November 29, 2012.
US Resources, Inc. 2010. Affidavit of Annual Labor for State Mining Claims (Big Chunk claims). Doc. No.
2010-000-591-0. Available:
http://dnr.alaska.gov/ssd/recoff/sag/DocDisplay.cfm?SelectedDoc=20100005910&District=320.
Accessed: January 3, 2013.
USCG (U.S. Coast Guard). 2012. Public Notices of Bridge Projects—Public Notice 17-02-12.
USFWS (U.S. Fish and Wildlife Service). 2012. National Wetlands Inventory. Available:
http://www.fws.gov/wetlands/Wetlands-Mapper.html.Accessed: December 11, 2012.
USGS (U. S. Geological Survey). 2008. Alaska Resource Data File, Fog Lake Prospect. Available:
http://mrdata.usgs.gov/ardf/show.php?labno=IL030. Accessed: January 3, 2013.
USGS (U.S. Geological Survey). 2012. National Hydrography Dataset. High Resolution, Alaska. Available:
ftp://nhdftp.usgs.gov/DataSets/Staged/States/FileGDB/HighResolution. Accessed: October 16,
2012.
Bristol Bay Assessment 15-72 January 2014
-------
Chapter 15 References
Woody, C. A., and D. Young. 2006. Life History and Essential Habitats of Humpback Whitefish in Lake Clark
National Park, Kvichak River Watershed, Alaska. Annual Report FIS 05-403. Available:
http://www.fish4thefuture.com/pdfs/05-403-Lake_Clark_Whitefish_Annual_2005Feb07.pdf.
Accessed: December 11, 2012.
15.1.15 Chapter 14— Integrated Risk Characterization
Blight, G. E. 2010. Geotechnical Engineering for Mine Waste Storage Facilities. Boca Raton, FL: CRC Press.
Buck, G. B., C. E. Brazil, F. West, L. F. Fair, X. Zhang, and S. L. Maxwell. 2012. Stock Assessment of Chinook,
Sockeye, and Chum Salmon in the Nushagak River. Fishery Manuscript Series No. 12-05. Anchorage,
AK: Alaska Department of Fish and Game.
Caldwell, J. A., and L. Charlebois. 2010. Tailings impoundment failures, black swans, incident avoidance
and checklists. Pages 3-39 in Tailings and Mine Waste 2010: Proceedings of the 14th International
Conference on Tailings and Mine Waste, Vail, CO. October 1 7-20, 201 0. Boca Raton, FL: CRC Press.
Dye, J. E., and C. J. Schwanke. 2009. Report to the Alaska Board of Fisheries for the Recreational Fisheries of
Bristol Bay, 2007, 2008, and 2009. Alaska Department of Fish and Game, Division of Sport Fish,
Research and Technical Services.
Earthworks. 2012. U.S. Copper Porphyry Mines Report: The Track Record of Water Quality Impacts
Resulting from Pipeline Spills, Tailings Failures and Water Collection and Treatment Failure.
Washington, DC.
Ghaffari, H., R. S. Morrison, M. A., deRuijeter, A. Zivkovic, T. Hantelmann, D. Ramsey, and S. Cowie. 2011.
Preliminary Assessment of the Pebble Project, Southwest Alaska. February 15. Document
1056140100-REP-R0001-00. Prepared for Northern Dynasty Minerals Ltd., by WARDROP (a Tetra
Tech Company). Vancouver, BC.
Gibson, R. J., R. L. Haedrich, and C. M. Wernerheirn. 2005. Loss offish habitat as a consequence of
inappropriately constructed stream crossings. Fisheries 30:10-17.
Instanes, A., 0. Anisimov, L. Brigham, D. Goering, L. N. Khrustalev, B. Ladanyi, and J. 0. Larsen. 2005.
Infrastructure: buildings, support systems, and industrial facilities. Pages 907-944 in Arctic Climate
Impact Assessment. New York, NY: Cambridge University Press.
Langill, D. A., and P. J. Zamora. 2002. An Audit of Small Culvert Installations in Nova Scotia: Habitat Loss
and Habitat Fragmentation. 2422, Dartmouth, Nova Scotia: Canadian Department of Fisheries and
Oceans, Habitat Management Division.
Pearce, T. D., J. D. Ford, J. Prno, F. Duerden, J. Pittman, M. Beaumier, L. Berrany-Ford, and B. Smit. 2011.
Climate change and mining in Canada. Mitigation and Adaptation Strategies for Global Change
16:347-368.
Bristol Bay Assessment 15-73 January 2014
-------
Chapter 15 References
Price, D. M., T. Quinn, and R. J. Barnard. 2010. Fish passage effectiveness of recently constructed road
crossing culverts in the Puget Sound region of Washington State. North American Journal of Fisheries
Management 30:1110-1125.
Schindler, D. E., X. Augerot, E. Fleishman, N. J. Mantua, B. Riddell, M. Ruckelshaus, J. Seeb, and M.
Webster. 2008. Climate change, ecosystem impacts, and management for Pacific salmon. Fisheries
33:502-506.
Schindler, D. E., R. Hilborn, B. Chasco, C. P. Boatright, T. P. Quinn, L. A. Rogers, and M. S. Webster. 2010.
Population diversity and the portfolio effect in an exploited species. Nature 465:609-612.
Vick, S. G. 2002. Degrees of Belief: Subjective Probability and Engineering Judgement. Reston, VA:
American Society of Civil Engineers.
15.2 GIS Base Map Citations
Alaska Department of Natural Resources. 2006. Alaska Legislatively Designated Areas—State Parks.
Anchorage, AK: Information Resource Management Section. Available:
http://www.asgdc.state.ak.us. Accessed: February 7, 2013.
Alaska Department of Natural Resources. 1998. Populated Alaska Places. Anchorage, AK: Information
Resource Management Section. Available: http://www.asgdc.state.ak.us. Accessed: February 7,
2013.
Alaska Department of Transportation. 2012. Department of Transportation Road System
[DOT_RoadSystem_061212.zip].
Alaska Department of Transportation. 2012. Public Airports in Alaska [Public_Airports_Oct2012.zip].
Environmental Systems Research Institute. 2013. World Base—Light Gray Canvas. ArcMap 10.0.
Environmental Systems Research Institute. 2013. World Shaded Relief. ArcMap 10.0.
Environmental Systems Research Institute. 2007. U.S. States (Generalized). ESRI ArcMap 9.3 USA Census
Data.
Environmental Systems Research Institute. 2007. U.S. and Canada Water Polygons. ESRI ArcMap 10.0
Streetmap.
National Atlas of the United States. 2005. Federal Lands of the United States. Reston, VA. Available:
http://nationalatlas.gov/atlasftp.html. Accessed: February 7, 2013
U.S. Geological Survey. National Hydrography Dataset. High Resolution, Alaska [NHDH_AK_931v210.zip].
Available: ftp://nhdftp.usgs.gov/DataSets/Staged/States/FileGDB/HighResolution.Accessed:
October 16, 2012.
Bristol Bay Assessment 15-74 January 2014
-------
I
- •
United States
Environmental Protection
Agency
Region 10
1200 Sixth Avenue, Suite 900
Seattle, WA 98101
Recycled/Recyclable
Printed with vegetable-based ink on paper that
contains a minimum of 50% post-consumer
fiber and is processed chlorine free.
------- |